User:Dgrtvwnp

Chronic Brucellosis
Classification and external resources

Brucellosis annd Chronic Brucellosis

Introduction.

Brucellosis is a zoonotic illness with a worldwide spread . It occurs in wild and domestic animals. The disease is self sustaining in cattle, sheep, goats, pigs and dogs. Some of the means of perpetuating the disease in these species are: from dam to offspring before or at birth, from dam to offspring via milk, between male to female by sexual contact, by direct physical contact particularly licking, from contaminated environments, and from eating contaminated raw meat particularly placenta and birth products. The most common signs of brucellosis in animals are abortion and lameness. For the former reason B abortus in cattle has been callrd "Contagious Abortion". Brucellosis is caused by the bacterium Brucellae (B). Humans are not a primary host for Brucella and they usually contract the disease zoonotically from domestic animals and their products. It is suggested that about 500,000 humans become infected with brucellosis each year. Brucella is part of the α-2 subgroup of the Proteobacteriaceae and is most closely related to Ochrobactrum and is also related to Agrobacterium and Rhizobactium. Brucellae are facultative intracellular, gram negative coccobacilli that lack capsules, flagellae, and endospores. According to Al-Nasssir (2006) there are 7 known species of Brucella: B melitensis, B abortus, B suis, B canis, B ovis (sheep), B neutomae (rodents), B pinnipediae (seals) and B cataceae (whales, dolphins) Al Dahouk (2003) considers that the Brucella genus divides in six species: B mellitensis, biovars 1-3 (sheep and goats; B abortus, biovars 1-6 (cattle and other Bovidea); B suis biovars 1-6 (biovars 1-3 pigs; biovar 4, reindeer; biovar 5, small rodents; B canis (dog species); B ovis (sheep); B neotomae (desert wood rats); and B maris (seals, dolphins and porpoises). These lists are subject to continuous review as new information becomes available. Recently a new species of Brucella was discovered in the common vole. etc B melitensis, B abortus and B suis are termed smouth strains and B canis and B ovis rough strains. This designation relates to their appearance under the microscope. In smooth strains the bacterial surface layer contains lipopolysaccharide (LPS),Caron (1994) which is not present in the rough strainsHumans are most likely to be affected by B melitensis, B. abortus, B. suis and B. canis. B. melititensis is associated with sheep, goats and camel; B abortus with cattle, camels, Zaki (1948) , yaks and buffalo; B suis with pigs, caribou, Huntley (1963) , hares, Burgisser (1949) , reindeer and desert rats, Thorpe (1963) and B canis with dogs and othe canines, Mantur (2006) . There is little evidence that B ovis can cause illness in humans. For the general public brucellosis is usually contracted from the consumption of contaminated food derived from infected animals such as milk, milk products and meat. Veterinarians and farmers are at high risk of contracting brucellosis by contact with the birth products of infected animals and slaughtermen by contact with animal carcusses, aerisols, and meat products.

In both animals and humans brucellosis may occur in an acute, subacute or chronic forms. In both cases once acquired chronic brucellosis can continue for a lifetime. B melitensis, B abortus, and B suis become chronic in humans but current opinion suggests that B canis seldom does. Scientific evidence indicates that Brucella is an intracellular parisite. In the course of infection Brucella can enter and overcome the defenses of phagocytes and replicates within them. There is also evidence that Brucella can impair the release of cytokines such as INF-γ, Rodriguez-Zapata in 1996 and and TNF-α, Dormand (2002) preventing the action of CD8+. Also it has been shown that Brucella can inpair the function of Natural Killer Cells, Billard (2007) and prevent apoptosis of infected maccrophages, Tolomeo (2003) . These effects favour chronic illness. The infected phagocytes are a recevoir of infection which are protected from the general immune system response and probably inhibit antibiotic effectiveness. Such infected phagocytes can carry the disease to every organ in the body. Once settled within organs Brucella can cause locallised leisons and cysts which may be very small. It is probable that such leisoms can act as a recevoir of infection over long periods abstracted from the influence of the immune system and resistant to antibiotics. In humans B melitensis has the highest prevalence, is the most virulant and is also the most likely to become chronic but milder in B abortus. The onset of brucellosis may be a sudden acute attack or may be slow and insiduous. Acute attacks are usually not difficult to diagnose if the right serology tests are used. At least 2 antibody serology tests are needed measuring both immunoglobulin M (IgM) and immunoglobulins G (IgG). Seroconversion for Ig) usually occurs at about seven days and that for IgG and immunoglobulin A (IgA) slightly later. The antibody response of individual patients may vary considerable with some false negative results. Also the tests can be subject to technical problems giving both false positive and false negative results. The culture test on blood provides a virtulally certain diagnosis but can take up to 6 weeks to complete when a rapid diagnosis is often required. The culture test does not work well for B abortus posibly because with this milder form of the illness patients do not become bacteremic. It can also be used to differentiate between B melitensis, B abortus, and B suis. The culture test can also be used to measure tissue and aspirate samples from infected organs.

Acute attacks of brucellosis usually starts with an influenza like illness with fever, sweating, myalgis, headaches and possibly a high temerature. Brucellosis was once called "undulant fever" because of a perceived fluctuating temperature profile but this does not occur in all cases. Without treatment brucellosis can become established amd become chronic in at least 30% to 50% of cases, Hughs (1897) and Spink . Before antibiotics were introduced the only treatment that worked was repeated injections of vaccine although the usefulness of this treatment has been disputed. Antiboitics were gradually introduced from about 1950 onwards. A range of antibodies that work against gram positive bacteria can be used but doxycyclin, streptomicin, rifampicin, ciprofloxicin and TMP-SMZ have proved the most useful. Combinations of 2 or 3 antibiotics work better than single antibiotics. The most widely used treatments are combinations of doxycyclin plus riframpicin or doxycline plus streptomicin with a duration of at least 8 weeks. For best results treatment should be started as soon as possible after the initial infection. If treatment is delayed and Brucella becomes established with locallised leisons and complications treatment becomes less succesful. At least 10% of patients who are treated with antibiotics relapse within 2 years from the start of the infection. Complications can be severe and affect any organ of the body. The most problematical complications relate to the skeleton, liver, spleen, urinogenital tract, panceas, heart, lungs, and nervous system including the brain. Death can occur in up to 2% of cases and this is usually related to complications affecting the heart and the nervous system. Symptoms of illness vary widely depending on the mode of infection. Chronic brucellosis is usually defined as illness which continues beyond 12 months. The most likely caused of illness becoming chronic are: severe illness with complications, lack of treatment, delayed treatment, and inadequate treatment, Landau (1999) . Patients are ofter described as cured after antbiotic treatment but the extent to which patients make a completely recover is unknown particularlt considering the problems of diagnosing chronic brucellosis as discussed below). Recent research indicates that 70% of patients retain Brucella DNA in their sytems for more than 2 years after treatment despite apparent cure, Pappas (2008) . In up to 50% of cases the onset of brucellosis can be insiduous . This makes diagnosis very difficult and may increase the likelihood of chronic illness.

Chronic brucellosis can be very difficult to diagnose. Problems relate to vague generalised symptoms with nothing specific to brucellosis and false negative serology test results. The symptoms are similar to acute brucellosis but less severe. The main ones are weakness, fatigue, headaches, backache, fever usually without obvious temperature rise and general aches and pains, dispepsia, and depression. Patients often appear to be superficially well with no unusual blood test results. The culture test hardly ever works in chronic brucellosis. Antibody serology tests for IgG and IgA work better than tests for IgM. All tests fail in a significant proportion of cases including imdirect enzyme linked immunosorbent assay (iELISA). There is a widespread opinion that iELISA can diagnose chronic brucellosis better than other tests. This is incorrect. In practice in most cases the results of antibody serology tests in chronic brucellosis are arbitory. Positive results may be considered to prove its presence but negative results do not prove it is absent. It is most likely that as chronic brucellosis progresses, serology antibodies against Brucella disappear from blood probably due to locallisation of infection and therefore lack of interaction with the immune system. The intradermal skin test can give positive results where antibody serology tests fail but is not often used. The skin test only shows that the patient has been in contact with Brucella past or present but not if he is ill. These diagnostic problems have caused great confusion amongst doctors and it is clear that many chronic brucellosis patients go undiagnosed and untreated, Castano (2009) . In such cases the patients history of illness and possible past contact with Brucella may be the most important information. Doctors with no direct experience of brucellosis are likely to get the diagnosis wrong, .

There is no specific antibiotic treatment for chronic brucellosis. Typical treatment regemes for acute brucellosis are used and are mostly unsuccessful. Chronic brucellosis is undoubtedly much more difficult to treat than in the acute illness, Boura (1984) . Prolonged treatment of at least 6 months is probably the best option. The vague symptoms, negative serology tests, and failure of antibody treatments have lead som doctors to question the existance of chronic brucellosis. However the only alternatives offered have been psychiatric/ psychological diadnosis such as neurasthenia, psychoneurosis, , and chronic fatigue syndrome . In somes cases there is a derogatory aspect to these diagnosis which were used in the USA to prevent workers such as laboratory technicians and meat packers getting rightfully deserved compensation. Evans (1961) was the champion of such chronic brucellosis patients. She personally knew the problenms having contracted chronic brucellosis herself and been misdiagnosed several times before getting a positive diagnosis by chance during an unrelated surgical procedure. A small number of other chronic brucellosis patients with negative serology tests have been shown to have brucellosis from tissue samples taken during operations in a similar way to Evans. Quantitative Polymerise Chain Reaction (Q-PCR) is a new technique which has resently shown the presence of Brucella in chronic brucellosis patients up to 30 years after first infection and gives hope of a better diagnosis in the future, Castano et al (2009) .

In this introduction I have briefly discussed many of the issues that affect the understanding and diagnosis of acute and chronic brucellosis. In the text I will discuss these and other issues in more detail. There is a particular emphasis will be given to chronic brucellosis.

History of Brucellosis and Chronic Brucellosis

• c1900 BC. Jacob to Leban, "In all the 20 years I have been with you. You ewes and your she-goats have never miscarried." Genisis 31:38.
• 450 BC. Hippocrates describe intermittent fever in man now considered to be adescription of brucellosis.
• 790 AD. Capasso (1999) and (2002) described leison on the bones from people in Herculaneum killed by the eruption of Vesuvius in 790 AD and which he considered typically of those caused by brucellosis. These occurred in 17.4% of adults. He also found coccoid cells in carbonised cheese consistent with the morpology and size of Brucella.
• 1589. Mascal reported that 50-60% of pregnant cattle in parts of Great Britain abort, Penberthy (1895) .
• 1800-1900. Epidemics of abortions in cattle reported in in Great Britain, France, Germany, and Denmark.
• 1807. The general contageous nature of abortion in cattle was recognised by custom and practice, .
• 1843. Gaylord reported outbreaks of abortions in cattle in New York, Pennsylvania, Delaware, and Virginia in the USA.
• 1861. Marston studies soldiers in the Crimean War sick with typhoid fever. He concluded that some were suffering from a different disease which he called "gastric intermittent fever". He suffered from this illness himself and of his own illness he wrote that the patient is prone to relapses and the disorder is followed by a protracted convalescence with rheumatism of one form or another. This is the first modern description of brucellosis.
• 1867. Chartres of the British army noted that in some cases the onset of illness was sudden and severe with complete prostration while in others its onset was so mild that it was difficult to differentiate between the genuinely sick and the mere pretender but an analysis of the patients history could reveal a history of dispepsia, debility and langour. This is the first reference to chronic brucellosis being related to malingering.
• 1876. Frank demonstrated that infected placenta from aborted cows could cause the disease to healthy pregnant cows.
• 1887. During the study of Mediterranean Fever in Malta, Bruce (1887) observed the presense of an emormous number of micrococci scattered throughout the tissue of the spleen of a solder who had died of the illness which he called micrococcus melitensis (now: Brucella Mellitensis). Within a few months he cultured a small micrococci from similar samles.
• 1897. Hughs provided the first detailed description of acute and chronic brucellosis. He noted that cases of sciatica could continued for more than 18 months and that prolonged cases do not return to their original for 12-24 months and in many cases years.
• 1897. Bang investigated abortion in cattle found a small organism in yellow odourless liquid the layer between the uterus wall and the fetal sac in which there were great numbers of bacterial organisms. After culturing and microscopic examination he considered this to be a small bacillus which he called bacillus abortus (now Brucella abortus) and he concluded that this was the cause of "contageous abortion". The term 2Bang'sDisease" was termed to describe B abortus.
• 1897. Wright and developed a serology agglutination test for brucellosis.
• 1903. Basset-Smith working in Malta for the British Army noted that Brucellosis in humans could continue indefinitely.
• 1904. Zammit as reported by Eyres (1908) showed that goats could be infected with B melitensis showing Brucella in their blood, milk and urine and that goat's milk was the sourve of infection in humans.
• 1905. Zammit observed the chronic nature of B melitensis in goats noting that once a goat was infected with mediterranean fever it never fully recovered and Brucella could be recovered from its organs and milk thereafter. On the basis of his work British service were banned from drinking raw milk in Malta with a consequental dramatic drop in the incidence of the disease.
• 1910. B abortus isolated in the USA by MacNeal .
• 1911. DeBois demonstrated that B melitensis was associated with abortion in goats.
• 1911. It was originally considered that B melitensis was confined to the Mediterranean area but Ferenbaugh (1911) confirmed that the disease was also present in Texas, USA.
• 1914 Traun and Mohler (1913-14) working in the USA were the first to identify Brucella suis as a disease of pigs. Traum recovered B suis from the tissue of aborted piglets.
• 1916. Cooledge fed viable Brucella organisms to human volunteers. The volunteers did not seem to develop the disease but agglutinins were found in their blood.
• 1921. Bevan and Keefer (1924) reported cases of brucellosis abortus in humans.
• 1918. Evans compared micrococcus melitensis and bacillus abortus in terms of morphologically, culturally. biochemically and by agglutination tests that were varients of the same genus. She considered that both organism had siilar pathogenicity in humans and suggested the name bucellosis for the associated illness.
• 1918. Evans found brucella abortus in cows milk. She avocated pasteurisation and was vilified by the milk industry.
• 1920. The relationship between B melitensis and B abortus was confirmed by Meyer and Shaw (1920) and they introduced the name Brucella.
• 1921 Fleishner showed that found the diseases caused by B abortus and B melitensis in monkeys was identical.
• 1924 Evans recorded the first case of a human infected with B abortus from cows milk.
• 1924 Keefer recorded the first case of B suis in humans.
• 1925 Shroeder confirmed that Brucella could be present in milk.
• 1927 Carpenter recovered an organism from patients with undulant fever and injected it into pregnant heifers which then aborted. He thus linked brucellosis abortus in humans and cattle.
• 1928. Huddleson demonstrated that B suis was primarily a pathogen of pigs.
• 1930Morales-Otero (1930) showed that B suis could infect humans.
• 1930Hardy showed on the basis of serology tests that upto 34% of a herd of 611 pigs were carrying sub-clinical (or chronic) infection of B suis.
• 1953. B ovis isolated from rams with epididymitis by Buddle, .
• 1957. B neotomae was isolated from desert wood rats in the Western States of USA by Stoenner, .
• 1966 B canis was identified by Carmichel (1996) for the first time in aborting beagles.
• 1968 The first cases of B canis in humans was described by Morrisset (1969) also reported by The National Communicable Desease Centre of the US.

The opposing argument was put by Evans . Evans was a microbiologist who did pioneering studies on Brucella. She caught brucellosis during laboratory work and had it in a chronic form for most of the rest of her life.

The Genus Brucellae.

Taxonomy.

The classification of Brucella into species and biovars is not fully established. The classical taxonomic designations are: genus, species, and biovar. Strains are further variants of biovars based on lesser mutations. Thus in the case of B abortus, Brucella is the genus, abortus is the species and there are 7 or 8 biovars. Six species of Brucella have been identified by classical biological and biochemical methods, Corbel . These are B abortus, B melitensis, B suis, B ovis, B canis and B neotomae. After the discovery of DNA new genotyping methods were introduced. An early method based on DNA-DNA hybridization is described by Verger (1985) and (1987) . These studies indicated that the genus Brucella is a highly homogenetic with >90% DNA-DNA overlap between species. He proposed that on the basis of this new work Brucella should be treated as a single species and the previously designated Brucella species should be treated as biovars. These new ideas were initially accepted by the International Committee on Systematic Bacteriology; Subcommittee on the taxonomy of Brucella, Corbel (1988) . Thus under this new scheme B abortus was classified as follows: Genus, Brucella; Species, melitensis and Biovar, abortus. These ideas were however rejected by other workers on the basis of classically observed data in relation to host preference, host range, virulence and phenology, ie the observable biological and biochemical charactoristics of the organism. There was also concern the the new classification would lead to confusion in relation to the varying pathogenic risks associated with the different classically determined species of Brucella. For a time both systems were in use but it was decided at a subsequent subcommittee on the taxonomy of Brucella of the International Committee on Systematics of Prokaryotes to revert to the classical system, Osterman (2006) .

The currently established taxonomy of the Brucella genera based on both biotyping and genotyping is:

• Phylum - Proteobacteria.
• Class - Alphaproteobacteria.
• Order- Rhizobiales.
• Family - Brucellaceae.

The class Alphaproteobacteria include families of organisms that are either mammalian or plant pathogens or symbionts. Within this class other organisms affecting mammals are the genera Bartonella, Rickettsia and Ehrlichia which are spread by vector-based transmission. Other genera within the family Brucellaceae are Ochrobactrum and Mycoplana. Unlike other genera of the order Rhizobiales both Brucella and Bartonella can infect mammalian cells, however, whereas Brucella is a facultative intracellular parasite ie capable of living and reproducing either inside or outside cells Bartonella is an obligate intracellular pathogen entirely reliant on intracellular resources for reproduction. According to Paulsen (2002) the Brucella spp genome is 50–100% larger than Bartonella spp. genomes and retains some of the metabolic functions shared by the plant pathogens. Brucella can persist in soil for months under suitable conditions possibly consistent with an ability to metabolically utilize plant-based molecules. Recently it has been found that the newly discovered B microti is able can live in soil, Scholz (2008) like the other two genera in the Brucellaceae family, Ochrobactrum and Mycoplana. The means by which Brucella has the ability to invade mammalian cells may be derived from its biological origin or could have been acquired. Adoption of a host intracellular lifestyle has been shown to include gene loss and possibly horizontal gene transfer, Chain (2005) and Wattam (2009) .

The placement of the genus Brucella in the order Rhizobiales within the class Alphaproteobacteria has been established by Gupta (2005) and Williams (2007) and confirmed on the basis of ribosomal cistron similarities and 16S rRNA sequence comparisons by De Ley (1987) and Yanagi (1993) . Thus Brucella shares a close relation to soil organism such as Ochobactrum spp, and the plant symbionts such as Rizobium spp and Agrobacterium spp. Other possible members of the Brucellaceae family are Pseudochrobactrum reviewed by Kämpfer (2006) and (2007) , Daeguia which was isolated from sludge of a textile dye works as described by Yoon (2008) and Crabtreella a nitrogen-fixing bacterium isolated from rice paddy soil as described by Xie (2006) .

The genus Ochrobactrum is a diverse group of species and is considered the most closely related to Brucella, Velasco (1998) and Lebuhn (2006) . According to current opinion, Leal-Klevezas (2005) and Scholz (2008) Ochrobactrum intermedium is the most closely related to Brucella. In consideration of the observed close relationship between Brucella and Ochrobactrum there is a school of thought that Brucella should be reclassified within the latter genus, Whatmore (2009) .

Michaux (1993) demonstrated that the DNA of B melitensis 16M was divided into 2 chromosomes. To date 15 genomes of Brucella biovars and strains have been sequenced. 14 of these have 2 chromosome pattern. the only exception is B. suis 686 which has one chromosome, Ficht (2010) . There are two possible explanations for Brucella having a second chromosome. These are: division of an original chromosome or capture and modified of a megaplasmid, Jumas-Bilak (1998) . General information derived from related genera which have plasmids with similar gene arrangements favours a plasmid origin. The single chromosome for B. suis 686 has probably been derived from the recombination of the 2 chromasomes.

Biological Phenotyping of Brucella (Biotyping).

Host Preference.

Host preference is considered a most important characteristic of the Brucella species and the main reason that B melitensis, B abortus, etc. are considered to be species rather than biovars. In the host species the disease is self sustaining through the generations by a variety of transitions methods including sexual transmission, vertical transmission, environmental contamination and the ingestion of contaminated food or material. Although Brucella species are host specific the infection can spread to other animal species such as by contact with Brucella shed into the environmental by the host species and by the consumption of contaminated meat from the host species by carnivores, etc. However on the basis of accumulates studies it appears that such infections in these secondary host are self limiting and such disease is not passed on through a series of animal generations. A possible exception is to this is B ovis infection in red deer (Cervus elaphus) contracted form sheep which may be at least partially self sustained.

Although Brucella is generally considered a host specific genus, molecular genotypic analysis indicates that the classical designated Brucella species differ hardly at all from each other whereas the genetic variation in host species are much greater. This seems to mitigate against co-evolution of Brucella species with the host species. Possible explanations are slow evolution of the Brucella species due to limited genetic flexibility or alternatively recent adaptation. Foster (2009) used whole genome comparisons of B abortus, B melitensis, B suis, B canis and B ovis and Ochrobactrum anthropi to determine a molecular clock for evolution Brucella and concluded that the species diverged from their common B ovis ancestor in the past 86,000 to 296,000 years, which precedes the domestication of domestic livestock hosts. This is likely to to simplistic. Interestingly although the date at which humans were first were infected with Brucella cannot be determined, observations made by D'Anastasio (2009) on a 2.4 to 2.8 million year old hominoid Australopithecus africanus from South Africa suggest that lesions on vertebral bodies were consistent with pathological changes due to the initial phases of Brucella infection. The macroscopic, microscopic and radiological appearance of the lytic lesions of the lumbar vertebrae is consistent with brucellosis. This suggesting a longer co-evolution of humans an Brucella possibly related the commencement of meat eating. This might suggest that Brucella infection was already widespread in animals at this time and not confined to sheep.

Virulence of Brucella Specis and Biovars.

The variation of virulence of Brucella species and biovars is an important characteristic of the genus. In general each species has its own host range, Meyer (1964) . The virulence of Brucella can be considered in terms of the ease of transmission to other animals within the same species, the ease of transmission to animals of a different species particularly humans and the severity of the contracted infection in both cases. It is important to note that many cases of Brucella infection in animals and human probable go unrecorded.

• B melitensis biovars 1-3. The host species for B melitensis are domestic sheep and goat. It causes severe infection in these species which can become becomes chronic and long lasting. It transmits easily within the host species and also to a wide range of other species such as cattle, camels and humans. These infections in secondary host are usually severe and often become chronic.
• B abortus biovars 1-7 and 9. The host species for B abortus is domestic cattle It spreads easily between cattle causing severe disease. The main manifestation is abortion. It is easily transmitted to experimental animals such as guinea pigs, domestic animals such as horses, dogs and to humans. The disease in humans is generally less severe than that caused by B melitensis but not in all cases. Chronic disease occurs in cattle and a range of species including humans. Chronic disease in humans is generally less severe than that caused by B melitensis.
• B suis Biovars 1 and 3. The host species for B suis biovars 1 and 3 are the wild boar and the domestic pig. These B suis biovars are similar in virulence to B melitensis and spread easily from pigs to other animals species and to humans. Chronic disease occurs in animals and humans.
• B suis biovar 4. The host species for B suis biovar 4 are reindeer and caribou. It can spread easily to other animals including cattle, deer and carnivores such as bears and wolves. Humans can become infected.
• B suis biovar 2. The host species for B suis biovar 2 are the European hares and the wild boar. It cause severe chronic disease in hares. It can spread from wild boar to domestic pigs. There is insufficient information to decide if this biovar can infect other animals species. No infections in humans have been recorded.
• B neutomae. B neutomae is only known to affects wood rats in the USA. Other species can be infected experimentally in the laboratory but natural infections is not known. Lack of contact may be significant.
• B canis. The host species for B canis is the domestic dog and possibly other Canidae although there little data to support the latter supposition. Disease in dogs can be severe and also becomes chronic. Experimental animals can be infected with B canis and on the basis of limited information it seems to spread easily to other animals such as cats with stray dogs as vectors. There are a limited number of recorded cases of humans infections which have occurred by contact with infected dogs. On the basis of limited information, disease in humans is considered to be mild but early intervention using antibiotic treatment may be a key factor. Some cases of more severe disease have been recorded.
• B ovis. The host species for B ovis is the domestic sheep. The disease in sheep particularly rams is severe and usually chronic. Laboratory animal can be infected but the only other species known to be infected naturally are red deer in New Zealand and white tail deer in the USA. This is probably due to contact with B ovis infected sheep. A great many people have worked with B ovis infected sheep without any obvious signs of infection. A very limited number of infection in humans have been described.
• Newly discovered Brucella species. New species of Brucella which cause infection in marine mammals appears to be widespread and probably causes severe infections. Little is known of the impact of newer discovered species of Brucella on human health although a few cases of infection have been recorded and may be severe.

Tests for Virulence. The main criteria for virulence of Brucella are observations of its affect on animals including humans. Laboratory tests for virulence involve the use of laboratory animals. Guinea pigs have been widely used for this purpose. Typically the guinea pig is inoculated with 1 ml of a suspension of the Brucella isolate under study containing 10⁹ CFU/ml of the organism. The animal is then observed for about 45 days after inoculation and then necropsied. The spleen is examined for the extent of lesions and samples from such lesions bacteriologically cultured. The spleen to body weight is determined in relation to a normal spleen. The diseased spleen is generally enlarged This ratio of diseased to normal spleen size generally increases with the virulence of the species, biovar or strain. A less virulent strain such as B abortus s19 can be used for comparative processes and as a standard.

It is generally noted that when smooth Brucella species such as B melitensis or B abortus are repeatedly re-cultured there can be a change to a rough or mucoidal forms which are less virulent. According to Alton (1988) smooth Brucella species have a tendency to undergo variation during growth. changes in the colonial morphology of a culture are associated with changes in infectivity and antigenicity. Smooth colonies are usually pathogenic, whereas rough variants are less infective and lack the antigenic characteristics.

Microbiological and Biochemical Phenotyping (Biotyping).

Biotyping is the classical methodology which has been used to identify and classify Brucella species and biovars. It has also been used to provided information for the identification of ancestral evolutionary pathways. There are no single definative biotypinng tests that can identify individual Brucella species or biovars. The tests involve optical examination of Brucella isolates and analysis of the impact on them of biochemical and biological reagents. In practice a range of tests are carried out and from an analysis of the combined data it is possible to make a fairly confident determination of the Brucella species and to a lesser extent the biovar. These classical methods which have been used to establish differences between B abortus, B melitensis, B suis, B canis, B ovis and B neotomae and on this basis the latter have been designated as separate species of the Brucella genus. Full details of the tests are described by Alton (1988) and Weyant (1995) .

These methodologies are slow and technically difficult. Biotyping requires handling of live Brucella which potentially makes this work hazardous. Biohazard level 3 containment is necessary. It also requires expert laboratory workers and even then the interpretation may be subjective. The overall limitation is an inability to get beyond a partial resolution between biovars. Unresolved problems include unclassified biovars of B abortus and the problems associated with the inclusion B suis biovar 5 in the same group as the other B suis biovars. Recently a range of new Brucella species have been discovered including the new species associated with marine mammals. It can reasonable said that classification of these new species would be difficult by these classical biotyping methods but fortunately newer genotyping methods are now aiding this process nevertheless traditional biotyping remain a useful resource.

The methods used are:

• The Morphology of Brucella.
• The Antigenic Agglutination Test.
• The susceptibility of Brucella to Phages.
• The Oxidase Test.
• The Urease Test
• The growth of Brucella with or without Carbon Dioxide.
• The growth of Brucella in the presence of Dyes.
• The Production of Sulphur Dioxide by Brucella cultures.
• The affect of Dyes on Brucella Growth
• The Agglutination of Brucella by Antisera A, M and R.
• The oxidation of Aminoacids and Carbohydrate substances by Brucella.
• The analysis of the Fatty Acids of Brucella.
• The susceptibility of Brucella to Antimicrobial Testing.

Identification of Brucella

Morphology.'

Examination of Brucella isolates under the microscope can provide useful information. Brucellae are bacteria. They are gram-negative (i.e. it does not retain crystal violet dye) cocco-bacilli which usually are arranged singularly. They are 0.6 μm to 1.5 μm long and 0.5 μm to 0.7 μm wide. Repeatedly cultured samples may cause changes such as to mucoidal forms. Brucellae are non-spore forming and non-motile. They are resistant to decolourization by weak acids. They require thiamine, nicotinamide and biotin to grow. They are aerobic but some strains require the presence of carbon dioxide. Growth is inhibited by bile acids, tellurite and selenite. Brucellae grow slowly in liquid or on solid media. On solid media colonies can usually become apparent about 3-4 days after culture. They are round and about 1-2 mm in diameter with smooth margins. Viewed from above they are pearly white and convex and viewed obliquely (lit at a 45° degree angle from below) they are a pale honey colour becoming darker over time. Microscopic examination indicates that B abortus, B melitensis, B suis and B neotomae are smooth species. B canis and B ovis are mucoidal and describes as rough. Rough colonies are dry with a granular appearance and are yellow. Mucoidal colonies such as for B canis are transparent greyish in colour and are slimy and/or sticky.

Brucella can be differentiated from other bacteria by a combination of morphology and the agglutination test, Brucella can also be differentiated from other gram-negative bacteria by a range of tests. These include: motility at 20 and 37°C; lactose fermentation; acid production the presence of glucose; heamolysis on blood agar; catalase activity; oxidase activity; urease activity; nitrate reduction and citrate utilisation. General biochemical tests for bacteria can also be carried out including tests with the following reagents: indole production, growth on Simmons citrate, methyl red, Voges-Proskauer, arginine dihydrolase, lysine decarboxylase, ornithine decarboxylase, β-galactosidase, and litmus milk, MacFaddin (1980)

Antigenic Agglutination Test.

This is a test for the formation of an agglutinate in which a suspension of the Brucella from an isolate under test is mixed with ant-serum. Smooth Brucella react with anti-smooth sera and rough Brucella with anti-rough sera. Other gram negative bacteria such as Bordetella bronchiseptica, Campylobacter fetus, Moraxella and Acinetobacter which could be confused with Brucella to not react to this test. This test together with morphological studies can provide a reasonable identification of Brucella. Yersinia enterocolytica 09 can elicit an antibody responce similar to smooth Brucella. but can be can be distinguished on the basis of motility, acid reaction on glucose and their morpholoogy (large rods).

Indentification of Brucella Species.

Phage Typing.

Bacteriophages are viruses that infect bacteria. In the case of Brucella they can be used to diffentiate between species, Meyer (1961) . A number of strains of phage lyse Brucella. Of these only the Tbilisi (Tb) phage and the R/C phage are routinely used. Other Brucella phages are technically difficult to handle. Tb phage lyses B abortus at low concentrations but no other Bucella species. In this case the routine test dilution is the the highest dilution of phage that gives complete lyse (dilution range 10₁ to 10₉). If TB page is used at higher concentrations (x 10⁴) it also lyses B suis and B neotomae. R/C phage lyses B ovis and B canis, Corbel (1984) . The test offers an easy separation of B abortus and B melitensis. It is particularly useful for differentiating B abortus strains which might be construed as B melitensis by other tests. These tests can also adequately distinguish between smooth and rough Brucella species.

The Oxidase Test.

This test used the reagent, NNN¹N¹-tetramethyl-p-phenylene diamine dihydrochloride impregnated into filter paper as described by Steel (1961) . Freshly grown culture is streaked onto the impregnated filter paper. A positive reaction is indicated by the development of a dark purple colour. B ovis and B neotomae give negative results. B abortus, B melitensis, B suis and B canis give positive results. B abortus isolates obtained from Africa give negative results, Verger (1984) .

The Urease Test.

A test has been developed to utilise the urease activity of Brucella. This measures the rate at which urea is converted to ammonia and carbon dioxide. This test can only be used as a general guide because there is some unexpected variation with some Brucella strains. The test can be performed either on agar impregnated with phenol red dye (Christensen's medium) and urea or in liquid media (Bauer's method) also containing these reagents. In both cases a freshly cultured aliquot of the Brucella isolate under test is added and a positive result is indicated by a change of colour from yellow to purple-pink. Some indication of which species of Brucella is being tested can be gauged from the rate at which the colour change takes place. The rate of colour change for B abortus s19 can be utilised as a standard. In general all species of Brucella split urea but the rate at which this occurs varies. B suis, B canis and B neotomae strains split urea rapidly whereas B melitensis strains split urea at a slower intermediate rate although some isolates are exceptional with a more rapid rate. B abortus splits urea slowly and some field strains and B abortus reference strain 544 give negative results. B ovis strains generally gives negative results.

Identification of Brucella Biovars.

Morphology.

Physical study of the morphology of Brucella can be used to differentiate between smooth species: B abortus, B melitensis, B suis and B neotomae and rough species: B ovis and B canis. However it should be noted that smooth forms of Brucella can convert to rough forms by repeated re-culturing.

The Affect of Carbon Dioxide on Growth of Brucella.'

Brucellae are aerobic bacteria. The need for CO₂ to enable growth of Brucella is variable between species and biovars. A original isolate sample of Brucella should be used since re-culturing of an isolate which needs can lead to the development of mutants. Specifically an isolate requiring CO₂ can change to one that does not. Practically, duplicate cultures are performed one in air and the other in air plus CO₂. CO₂ dependence is demonstrated if the isolate grows in the air/ CO₂ atmosphere only. Brucella biovars which require CO₂ to grow are: B abortus biovars 1, 2. 3. 4 and 9 (all variable) and B ovis. Brucella biovars which do not require CO₂ are B melitensis biovars 1, 2 and 3; B abortus biovars 5 and 6; B suis biovars 1, 2, 3, 4 and 5; B neotomae and B canis.

Hydrogen Sulphide Production during Growth of Brucella.

H₂O may be produced during the growth of Brucella. Species and biovars vary in their production of H₂O during growth, Meyer (1961) .

The test is performed as follows: Brucella from the culture under test is inoculated onto a suitable agar slope. Then a dry strip of filter paper impregnated with neutral lead acetate solution is suspended over the gel but not in contact. The atmosphere can be air or air plus CO₂ as necessary. Blackening of the lead acetate paper is indicative of H₂ production. The lead acetate paper is changed every day. Brucella biovars giving negative reaction to this test are: B melitensis biovars 1, 2 and 3; B abortus biovars 5 and 6; B suis biovars 2,3,4 and 5; B ovis and B canis. Brucella biovars giving positive reactions are: B abortus biovars 1, 2, 3, 4 and 9; B suis biovar 1; and B neutomae.

Affect of Dyes on Brucella Growth.

The addition of dyes affects the growth patterns of Brucella on agar. These growth patterns are charactoristic of species and their biovars. The dyes usually used are basic fuchsin and thionine. The requirement for CO₂ is determined first and conditions set accordingly as this can effect the results. If sera is a contituent of the agar this generally increases the rate of growth and for which allowence must be made. The dye is mixed with liquid agar media which is then set in Petri plates. The Brucella culture under test in the form of a suspension (uniform turbidity) is streaked across the agar using a swab. Reference strains of B melitensis, B suis and B abortus are used as appropriate. The plates are then incubated and examined for growth from 3 days onwards.

Brucella growth is positive in the presence of thionin for: B melitensis biovars 1, 2 and 3; B abortus biovars 3, 5, 6 and 9; B suis biovars 1, 2, 3, 4, and 5; B ovis and B canis. Growth is negative for B abortus biovars 1,2 and 4; and B neutomae (concentration dependant). For basic fuschin Brucella growth is postive for: B melitensis biovars 1, 2 and 3; B abortus biovars 1, 3, 4 (some Canadian and British strains are negative) 5, 6, 7, 9; and B suis biovar 3. Growth is negative for: B abortus biovar 2; B suis 1, 2, 4 (most strains) and 5; B neotomae; B ovis (most strains) and B canis (most strains). Safranin O can be further used in relation to B suis. Growth of B suis and B abortus biovar 2 are inhibited by Safranin O but it does not affect B melitensis or the other B abortus biovars. In the case of Brucella live attenuated vaccines, B abortus biovar 1 grows in the presence of thionin blue but B abortus s19 does not. B melitensis biovar 1 grows in the presence of both thionin blue and basic fuchsin but B melitensis Rev 1 does not.

The Agglutination Test using antisera A, M and R.

Useful information concerning the classification of Brucella biovars can be obtained from a comparison of their interaction with antisera. Three antisera are used: smooth antisera A and M and rough antisera R. The antisera is recovered from rabbits which have previously been inoculating with killed Brucella. Antisera A is made using killed B abortus 544, antisera M is made using killed B melitensis 16M, and antisera R using killed B ovis. Antisera A and M interact with smooth forms of Brucella and antisera R with rough forms. The test can be performed by mixing a drop of the diluted antiserum with a drop containing a Brucella suspension. Alternatively, the tube agglutination test procedure can be employed. The presence of agglutination in the test sample is positive identification of Brucella. Results of the tests which determine the dominant antigen as described by Alton (1988) are as follows:

Smooth antisera A reacts positively with B melitensis biovars 1 and 2, B abortus biovars 1, 2, 3, and 6; B suis biovars 1, 2, 3, and 4; and B neutomae and negatively with B melitensis biovar 1; B abortus biovars 4, 5 and 9; B suis bivar 5; B ovis and B canis. Smooth antisera M reacts positively with B melitensis biovars 1 and 3; B abortus biovars 4, 5 and 9; B suis biovars 4 and 5; and negatively with B melitensis biovar 2; B abortus biovars 1, 2, 3, and 6; B suis biovars 1, 2, 3; B neutomae; B ovis; and B canis. Rough antisera R reacts with B ovis and B canis but none of the smooth biovars of B melitensis, B abortus, B suis and B neutomae.

Interaction of Brucella species and biovars with Oxidising Agents.

Meyers (1961) , (1961) , (1961) and (1962) has determined that the different species of Brucella demonstrate different patterns of oxidation rates with different aminoacids and carbohydrate substrates,. The oxidative uptake rates on amino acid (D-alanine, L-alanine, L-asparagine, and L-glutamic acid) and seven carbohydrate substrates (L-arabinose, D-galactose, D-ribose, D-glucose, and meso-erythritolare) can be determined. The method involves determining the oxygen utilisation of a standardised resting cell culture using the Warburg apparatus. The results are calculated as the mls of oxygen uptake per mg nitrogen per hour, QO₂(N). The results can be represented in tables, Alton (1988) .or in graphically, Verger (1977) . In the latter case the levels of utilisation of oxygen were taken as <100, >100<300 and >300. According to Alton (1988) the technical reproducibility of the test is good. The range of results is shown in the Table.

Reagent	B abort.	B melit.	B suis 1	B suis 2	B suis 3	B neut.	B ovis	B canis
	Q O2(N)	Q O2(N)	Q O2(N)	Q O2(N)	Q O2(N)	Q O2(N)	Q O2(N)	Q O2(N)
L-glutanic acid	>100<300	>100<300	>100<300	>100<300	>100<300	>100<300	>300	100<300
L-alinine	>100<300	>100<300	>100	>100	>100	>100	>100	>100
L-aspatagine	>100<300	>100<300	>100	>100	>100	>100<300	>100<300	>100
L-arginine	>100	>100	>100<300	>100<300	>100<300	>100	>100	>100<300
DL-ornithine	>100	<100	>100<300	>100<300	>100<300	>100	>100	>100<300
L-lysine	<100	<00	>100<300	>100<300	>100	>100	>100	>100
L-arabinose	>100<300	<100	>100<300	>100<300	<100	>100<300	>100	>100
D-galactose	>100<300	<100	>100<300	>100<300	<100	>300	<100	<100
D-ribose	>300	<100	>300	>300	>300	>100<300	<100	>300
D-xylose	<100	<100	>100<300	<100	<100	<100	<100	<100
Meso-erythritol	>300	>100<300	>300	>100<300	>300	>300	<100	<100

Analysis of Cellular Fatty Acids by Gas Liquid Chromatography.

Species of Brucella can be differentiated by differences in the composition of cellular fatty acids using gas liquid chromatography. The fatty acids are liberated by saponification of Brucella cells, Weyant (1996) . This is a useful additional method for the classification of Brucella biovars. Dees (1981) <re>[527]Dees SB, Hollis DG, et al. Cellular fatty acids of Brucella canis and Brucella suis. J Clin Microbiol. 1981 Jul;14(1):111-2.</ref> used the technique to distinguish B canis from rough strains of B suis not possible by other procedures. Using this technique Coloe (1984) was able to separate B ovis from B abortus based on specific fatty acid compositions. Ewalt (1990) used gas liquid chromatography to distinguish between field strains of B abortus and the vaccine B abortus s19 by demonstrating the reduced usage of meso erythritol by the latter.

Antimicrobial Susceptibility Testing.

The growth rates of species and biovars of Brucella can be compared on media containing antibiotics such as penicillin. These tests are used to help identify live vaccines B melitensis Rev 1 and B abortus s19. Therefore a typical B abortus biovar 1 isolate grows in the presence of 5 iu per ml of penicillin while s19 does not. B melitensis, biovar 1 does not grow in the presence of streptomycin (2.5 μg per ml) whereas Rev 1 does. B melitensis biovar 1 grows in the presence of penicillin whereas Rev 1 does not.

Molecular Genotyping (Genotyping) of Brucella.

Genotyping of bacteria including Brucella is a rapidly developing field. These techniques can potentially allow a rapid evaluation and understanding of the molecular structure of Brucella and the relationships between species, biovars and strains not possible with traditional biotyping methods. It can also increase understanding of the relationship of Brucella to other bacteria and provide information on the phylogenetic family tree. In these respects biotyping now only offers limited options. Other problems with biotyping include the slowness of the procedure and the significant dangers of handling live Brucella. For genotyping the latter risk is effectively eliminated due to the use of killed Brucella, In general genotyping methods can be carried out rapidly and can generate far more information. Genotyping depends on understanding of the Brucella genome.The recent sequencing of the genome of a number of Brucella species has greatly facilitated geotyping methods. Genotyping is still a new technology and the best techniques are still under development. Some of the techniques so far developed including: DNA (IS711), PCR typing, tandem repeat based typing, multilocus sequencing, and SNP typing. A range of these techniques is discussed below. At this time it is still sensible to use a combination of biotyping and genotyping techneques so that a combined knowledge of the two can be obtained. Some of the benefits of genotyping are as follows:

• It can be used as a rapid test to replace the culture test to demonstrate Brucella infection in animals and humans.
• At a slightly more sophisticated level it can differentiate between species and biovars.
• It is a research tool capable of differentiating between Brucella biovars and strains and tracing their geographical source.
• It can be used to achieve greater understanding of the relationships between different biovars of Brucella and also the relation between Brucella and other bacteria.
• It can probably fulfils the need for a rapid test to be used to identify Brucella as a biological weapon, Christopher (2005) .
• It offers greater knowledge which can be used in relation to the epidemiology of strains of human brucellosis and therefore adds to the understanding of such infections, van Belkum (2003) .
• There is the possibility that these techniques could be used to investigate possible mutations of Brucella in patients with chronic brucellosis, such as change to mucosal forms.

Early genotyping methods such as DNA-DNA hybridisation have shown that Brucella in a genetically homogenetic genus. One of the main faxets of Brucella genotyping methods depend on finding and using loci on the genome which exhibit hetrogenicity, i.e. they are capable of undergoing more rapid mutation. The sequencing of the Brucella genome has allowed the identification of unique loci occur among the different Brucella genomes which are the basis for the development of rapid tests to identify Brucella species, biovars, and strains.

DNA-DNA Hybridisation.

DNA-DNA hybridisation can be used to measures the degree of genetic similarity between different species of bacteria. and determine the extent of their relationship in terms of their arrangement on a phylogenetic tree. This technique can be used to compare DNA from two different Brucella species or compare Brucella to a close relative such as O anthropi. The procedure involves labelling the DNA of one organism and mixing it with unlabeled DNA of the second organism. This mixture is heated at a high enough temperature to disassociate the DNA strands and allow them to reassociate as a hybrid double-stranded DNA.This DNA is bound to a column of hydroxyapatite which is heated in a stepwise process to gradually disassociate the DNA. At each heating stage the eluate is collected for testing.

Simplistically the method is as follows: The reference DNA is double-labelled using DIG- 1 1 -dUTP and biotin- 16-dUTP using nick-translation. DNA-DNA hybridization occurs by mixing the labelled DNA with the subject DNA and heating to 100°C to cause DNA denaturation followed by incubation at 64°C for 16 hrs. The DNA can be separated by thermal elution on a hydroxyapatite column. The DNA eluent is collected on streptavidincoated microtitre plates. The amounts of DNA are detected using anti-DIG antibodies conjugated with alkaline phosphatase and developed with p-nitrophenyl phosphate. The colour development is measured at 405 nm. Methods are described by Verger (1985) and Ziemke (1998) .

As previously indicated, based on DNA-DNA hybridization, Verger (1985) and (1987) showed that the genus Brucella is a highly homogenetic with >90% DNA-DNA overlap suggesting that there is only one Brucella species. By convention it has been considered that species are separate if DNA-DNA overlap is ≤70% . In the case of Brucella Moreno (2002) considered that while DNA-DNA hybridisation suggested a single species a multispecies concept for Brucellla is far more consistent with molecular analyses and taxonomical studies. The DNA-DNA concept of species separation is however under review. Gevers (2005) considered that DNA-DNA hybridization is incapable of reflecting the levels of diversity that are now being uncovered in nature. As discussed below, in the case of Brucella, newer genotyping methods which can resolve Brucella biovars and strains into phylogenetic trees and clusters offers a far more comprehensive understanding of genetic differences. This homogenetic nature of Brucella means that genotyping techniques must be specifically targeted to its properties. Techniques which work for many other more hetrogenetic bacteria do not necessarily work well in this case. For Brucella more specific genotyping methods give better results.

Early work using these genotyping techniques have solved problems which biotyping could not. For example it was not possible using biotyping to determine the status of B canis and B ovis and whether they were true species of Brucella, Meyer (1969) and (1968) . Genotyping has solved this problem. Hoyer (1968) showed by a comparison of DNA-DNA hybridization experiments that polynucleotide sequences of B abortus, B melitensis, B suis and B neotomae showed 100% homology and B ovis 94% homology. In similar work they established homology with B canis', Hoyer (1968) . This work showed that all 6 species were closely related. These results were confirmed by De Lay (1987) who also showed that the Brucella and the plant pathogens Agrobacterium- Rhizobium organisms were from the same phylogeneric origin and derived from the same ancestral route.

Pulsed-field Gel Electrophoresis (PFGE).

For PFGE a low-cleavage-frequency restriction enzymes such as XbaI is used to produce high molecular weight fragments of DNA which are then separated and analysed by electrophoresis. According to Allardet-Servent (1988) the method can distinguish between the species B abortus, B. melitensis, B. suis, B. canis, and B. ovis and partially distinguish between biovars. The use of other enzymes such as Pacl, Michaux (1993) and Pacl and Spel, Michaux-Charachon (1997) enable the species B neomatae to be distinguished. Although these PFGE techniques has been used successfully for typing of other bacteria they are less successful for Brucella because the latter has limited diversity below the species level.

Polymerase Chane Reaction (PCR) Typing.

PCR was developed in the mid 1980’s and is based on a similar principle to that used by cells to replicate DNA. In this case it is used to amplify specified DNA regions, which are usually between 150-3,000 base pairs (bp) in length. In order to amplify the DNA sequence, the following reagents are required: a pair of short priming sequences which are complimentary to the ends of the targeted sequence; a special heat-resistant DNA polymerase, usually Taq polymerase; and a solution of the four DNA bases. These are all mixed together in a test tube which contains a few copies of the targeted DNA sequence. The solution is subjected to repeated cycles of denaturation and amplification to generate sufficient DNA units for analysis. Early attempts to use PCR showed that it could distinguish Brucella from other related genera but could not distinguish between species of Brucella, Baily (1992) , Fekete (1990) and Herman (1992) . This level of identification is adequate for some purposes such as a first test for brucellosis in humans or identifying Brucella contamination of food and other products. PCR has the advantage over classical culture methods that it is more tolerant of contamination and relatively easy to perform. However a method which can distinguish between species are necessary in other work such as Brucella eradication programmes in livestock. A number of different PCR techniques have been explored to identify species. The most widely accepted of these is AMOS-PCR which is based on IS711 an insertion sequence (IS) present in the Brucella genome which is specific to the genus. This method was developed by Bricker (1994) . AMOS-PCR is an acronym related to the species of Brucella which it identifies (B abortus bivars 1, 2 and 4, B melitensis, B ovis an B suis biovar 1). These are the most common species of Brucella encountered in the USA. The assay depends on the polymorphism arising from species-specific localization of the genetic element IS711 on the Brucella chromosome, Ouahrani (1993) and Halling (1993) which is unique to Brucella. The Southern blot method is used to determine the distribution of IS711 in the 6 Brucella species. The EcoRI restriction enzyme which shows that all Brucella species had at least 5 copies of IS711 is used. PCR amplification is carried out in a standard way. A five primer mix is used to amplify the various DNA fractions. The assay uses a common primer attached to the IS711 element and the others primers correspond to species-specific adjacent regions. The identity of the Brucella species is determined by the sizes of the amplified product from primers hybridizing at various distances from the element. This assay gives highly reproducibility for a range of field strains of Brucella and gives negative results for 6 closely related bacteria: Agrobacterium radiobacter, Agrobacterium rhizogenes, Ochrobactrum anthropi, Rhizobium leguminosarum, Rhizobium meliloti, and Rhzodospirillum rubrum. The method was further adapted to include B abortus vaccine strains s19 ad RB51 by Bricker (1995) thus enabling differentiation of field strain of B abortus from the vaccine. This latter is of great importance in relation to determining the source of B abortus outbreaks, Bricker (2002) . The method has been further developed to include B abortus biovars 5,6 and 9 by Ocampo-Sosa (2005) . In general this is a useful technique but it is limited by its failure to detect all species and biovars.

García-Yoldi (2006) developed a newer multiplex PCR assay which is often termed the Bruceladder and which has gained wide acceptance. This is a single tube PCR assay which can differentiate between all 6 classical Brucella species, new marine mammal Brucella species, B abortus vaccines s19 and RB51 and B melitensis vaccine Rev 1. The method is based on a range of unique fragments from the Brucella genomes of different species. The following species- or strain-specific genetic differences were used to design 8 pairs of oligonucleotide:

• A 25-kb DNA deletion leading to the loss of omp31 gene in the reference strains of all B abortus, Halling (2005) and Rajashekara (2004) .
• A 15-kb deletion comprising omp25b and wboAwboB

genes present in the B ovis species, Halling (2005) and Rajashekara (2004).

• A 2.6-kb fragment present in B. suis, but not in B. abortus or B. melitensis, Halling (2005) and Rajashekara (2004).
• A wboA gene disruption by an IS711 element present in the B. abortus vaccine strain RB51, Vemulapalli (1999) .
• A 702-bp deletion in the ery operon in the vaccine strain B. abortus S19, Sangari (1994) .
• A specific mutation in the rpsL gene of the vaccine strain B melitensis Rev1 that differentiates it from the B. melitensis reference strain, Cloeckaert (2002) .
• An IS711 element downstream of the bp26 gene in Brucella spp. isolated from marine mammals, Cloeckaert (2000)
• A 976-bp deletion in chromosome I specific to Brucella canis and a 2.2-kb deletion in chromosome II specific to Brucella neotomae, Rajashekara (2004) .

16S rRNA Gene Sequencing.

16S rRNA is a component of the 30S small subunit of prokaryotic ribosomes. The 16S subunit is entirely composed of ribosomal RNA. The 16SrRNA gene is highly conserved between different species of bacteria. This gene can be used as the basis of a rapid test to distinguish Brucella from othe related bacteria. Gee et al (2004) tested 65 strains of Brucella and compared 10 of these to 17 related bacteria. They concluded that this rRNA technique is reliable and can identify Brucella species with 100% gene similarity and differentiate it from other bacteria including its closest known relatives. The rRNA gene similarity for Ochrobactrum anthropi, Bartonella henselae and Agrobacterium tumefaciens were: 98.8%, 94.9% and 94.2% rRNA respectively. It is considered that a major advantage of this new test procedure is that it can be completed in 1 day compared to about 7 days for traditional biotyping analysis.

For the test 16S rRNA was prepared as follows: A single colony of bacteria was suspended in Tris buffer and heated to 90°C followed by centrifugation . The filtrate which contains the DNA was collected. The 16S rRNA was amplified through repeated cycles using DNA polymerase in the presence of dATP, dCTP, dGTP, and dTTP (nucleotides that are used to synthesize DNA) and eubacterial primers. A primer is a strand of nucleic acid that serves as a starting point for DNA synthesis. The 16S rRNA genes was than sequenced using the primers: F8, R1492 F357, F530, R530, F790, R790, F1068, F1083, F981, and R981, Sacchi (2002) , Moreno (1990) and Eden (1991) . For the non-Brucella bacteria the following additional eubacterial primers were used: R1333, R180, R591, and F1127.

Multilocus Sequencing Analysis. (MLSA).

Whatmore (2007) has described a MLSA method involving the use of 9 distinct genetic fragments. 7 loci are classic housekeeping genes and also a fragment of omp25 encoding a 25 kDa outer membrane protein. The loci are: gap(glyceraldehydes 3-phosphate dehydrogenase), aroA (3-phosphoshikimate 1-carboxyvinyltransferase), glk(glucokinase), dnaK (chaperone protein), gyrB(DNA gyrase B subunit) , trpE(anthranilate synthase), cobQ(cobyric acid synthase), omp25, int-hyp (25 kDa outer-membrane protein) The procedure was used to analysis 160 Brucella strains including B abortus, B melitensis, B ovis, B suis, B canis, B neotomae and marine Brucella samples from seals, porpoises, dolphines. It was demonstrated that this technique could differentiate between the species, but the sequencing of nine fragments is slow and not always practical and therefore the method is not suitable for routine rapid diagnosis procedures. However, use of the phylogenetic framework described here is the basis for the identification of SNPs that can define distinct Brucella species.

Single Nucleotide Polymorphism (SNP) Typing.

SNPs appear to be highly suited to use for studying Brucella and closely related bacteria both for identifying in terms of species and for understanding of their phylogenetic tree. According to Foster (2008) single SNPs are suitable for the identification of Brucella species because they are evolutionarily stable (conserved) and therefore unlikely to mutate into new states. Also they can be readily incorporated into genotyping methods

Other factors which contribute to their usefulness as described by Brumfield (2003) and Morin (2004) . These are as following:

• They are present in large numbers throughout the whole Brucella genome, they are relative stability over evolutionary time.
• They can be easily compare.
• They are included in intergenic regions.

Scott (2007) has describes a multiplex SNPs assay based on the MLSA analysis described by Whatmore (2007) . This technique is based on nine distinct genomic fragments from 160 strains of Brucella representing all the main species. Brucella species specific loci were selected at locations in glk, omp25, and trpE. The SNPs have been carefully selected to define particular species within the Brucella genus. A typical PCR method is used to enable rapid assay. The author considers that this methodology can be used to rapidly and unambiguously identify an isolate as a member of one of the six classical Brucella species or as a member of the recently identified marine mammal group. Budowle (2005) considered that a particular advantage of the use of SNPs for the identification and typing of Brucella compared to alternative methods based on different DNA loci was their unambiguous nature and the multitude of platforms that have already been developed to assay them. A further advantage compared to classical biotyping is the ability of the test to deal with crude extracts making direct use on field samples possible. The method is superior to AMOS-PCR techniques since it can identify a wider range of species. The methodology has been further improved by Gopaul (2008) . In the latter work 7 pairs of short sequence Minor Groove Binding (MGB) probes were designed corresponding to the SNPs. This new assay was able to differentiate all the classical and marine mammals Brucella. Tests on 300 isolates of Brucella showed a good correlation to the results from other classical biotyping and genotypic methods. 3 additional SNPs located within the 16S rRNA gene were added to ensure positive discrimination of Brucella from close phylogenetic relatives. The new method is rapid, simple and unambiguous. It is also considered that it should be possible to extend the number of SNPs to provide a comprehensive assay to encompass new species of Brucella as thay are discovered.

Multiple-locus variable-number tandem-repeat (VNTR) analysis (MLV(VNTR)A).

Analytical techniques based on MLV(VNTR)A can be employed to compare the genetic profiles of microorganisms, such as bacteria. The method is based on the utilisation of loci of the genome which show polymorphism (genetic diversity). In the case of Brucella it can be used to show differences between species and possibly biovars and strains and give information on their relationships. DNA includes repetitive sequenses which are sometimes polymorphic these are called variable tandem repeats (VNTR). Tandem repeats generally mutate at different rates, Vergnaud (2000) . It has been found that in the usefulness of MLV(VNTR)A in the study of other bacteria systems can be limited by rapid evolution and possible homoplasy. However it is considered that VNTR is a viable analytical method for Brucella because its genome is generally conserved i.e. it is homogenetic. The different species of Brucella have variable numbers of VNTRs which can therefore be discriminatory markers. Tandem repeats include both perfect and imperfect elementary units and alleles vary between Brucella strains. Tandem repeats are classified as satellites (megabases of DNA) which can be divided into minisatellites (100s of nucleotides, repeat units >9 bp) and microsatellites (10s of nucleotides, repeat unit >8 bp). Minisatellites are useful for species identification, Vergnaud (2000) . Microsatellites show more diversity and have greater discriminating power and can therefore differentiate biovars and strains. They are therefore useful in resolving questions related to evolutionary scenarios and the establishing phylogeny trees. In general high repeat copies (alleles) are more susceptible to rearrangement. Using MLV(VNTR)A for Brucella it has been found that strains clustered according to geographical origin.

These MLV(VNTR)A are found by study of the genome. The method is essentially the same as that used in forensic science and DNA fingerprinting. The procedure involves selection of suitable tandem repeats based on information derived from the whole genome sequence for Brucella such as from B. suis strain 1330, B. melitensis strain 16 M and B. abortus strain 9–941. A suitable method of identifying the tandem repeats is described by Cloeckaert (1995) . For the purposes of the test, bacteria are cultured, harvested and the cells lysed with proteinase K. The nucleic acids are recovered from the aqueous phase by precipitation using 2 volumes cold ethanol. The DNA can be amplified by PCR using Taq DNA polymerase, a solution of the four DNA bases and suitable primers complimentary to the ends of the targeted RNA sequence. The DNA is unzipping (denatured) at 95°C for 5 minutes. The mixture is then cooled to 60°C for 30 seconds to allow primer bonding to the ends of the DNA (annealing) and 70°C for 1 minute to allow the Taq polymerase to create new copies of each DNA strand. This procedure is repeated 30 times. The DNA is analysed by electrophoresis on agarose gel plates stained with ethidium bromine and detected with UV light.

Two groups of workers have explored this methodology. They have both produced very good results with their similar but different systems there is now a need to work towards a unified system.

The first system utilising MLV(VNTR)A is described by Le Flèche, et al (2006) who used a system based on the work of Denoeud (2004) . This system is based on an evaluation of the whole genomes sequences of B melitensis 16 M, B suis 1330 and B abortus strain 9–941. They utilises a computer program designed by Benson (1999) which can analyse DNA sequences for tandem repeats. By this method 80 possible tandem repeat loci were selected which showed polymorphic in relation to 18 reference biovars of Brucella species and 3 strains of marine mammal Brucella. 15 loci were considered most suitable for Brucella analysis because they gave results which showed clustering of biovars and strain of Brucella similar to those derived from classical biotyping. 8 were user-friendly minisatellite markers (100s of nucleotides, repeat units >9bp) with a good species identification capability and 7 were highly polymorphic microsatellite (10s of nucleotides, repeat unit >8 bp) markers with higher discriminatory power. This system was used to test 257 strains of Brucella and was able to identify 204 genotypes. In general this method gave results which accorded well with information derived from biotyping and other molecular methods, Cloeckaert (2001) and (2003) . Some of the main features identified by this system were:

• In general the method can separation terrestrial strains of Brucella into weakly connected clusters. There are clusters for B suis biovar 1 and B suis biovar 2, B abortus (2 clusters), B melitensis (3 clusters), B ovis. B suis biovar 5 and B neocroti. The clusters for marine Brucella strains are well separated from terrestrial strains.
• The results indicate that there is a definite tendency for isolates of the strains of the same biovar to cluster together according to the geographic source.
• B abortus biovars 1, 2 and 4 form a group. B abortus biovar 3 is in a separate group but 2 strains of B abortus biovar 3 from Africa are separate and closest to 2 strains of B abortus biovar 6 also isolated in Africa. This accords with the results previously found by biotyping which demonstrates differences between B abortus isolates from Africa and elsewhere in relation to growth characteristics and oxidative profiles, Verger (1984) .
• The results indicate that B melitensis occurs in three groups. This fits moderately well with biotyping results. Biovars 2 and 3 are mixed in 2 groups. Biovar 1 forms 2 separate groups. One sample from a dog from Costa Rica which has been classified as B melitensis biovar 2 by biotyping has been shown by genotyping to be distantly related to other B melitensis strain and most closely related to B canis.
• B suis is clustered in 3 groups. Strains of biovars 1, 2 and 4 are in one group and strains of biovar 2 in another group. Biovar 5 appears to be only distantly related to the other biovars. One strain of B suis biovar 3 is different.
• B canis is closely grouped to B suis biovar 4 and loosely connected to B suis biovar 1.This accords with the findings of traditional biotyping as described by Meyer (1990) .

Al Dahouk (2007) used the same MLV(VNTR)A system but including an extra microsatellite to study isolates of B melitensis associated with human infection. These were 28 samples of biovar 1, 55 of biovar 2, 42 of biovar 3 and 3 rough strains. 110 different genotypes were identified. The study showed that the MLVA clustering pattern which were found corresponded to the geographic origin of the strains. In this study B melitensis strains isolated from different patients within the same outbreak or from the same patient before first-line therapy and after a relapse showed identical genotypes. Correlation with biotyping data for the same samples was limited. It is considered that MLV(VNTR)A can significantly contribute to epidemiological trace-back analysis in cases of human brucellosis and therefore could help in studies of surveillance and control.

The second system of MLV(VNTR)A is described by Whatmore (2006) . This describes a VNTR system based on 21 loci. These include the 8 HOOF-Prints (Hypervariable Octameric Oligonucleotide Finger-Prints) described by Bricker (2003) . These HOOF-Print loci had from 7 - 27 alleles, diversity index (DI) at 0.74 to 0.92). The diversity index is described by Simpson (1949) . The other 13 new loci were selected in this work. Of these 13 loci, 7 were short sequence repeats with 5 -8 bp (2 - 15 alleles, DI at 0.71 - 0.91) and 6 were slightly larger repeats with 13 - 40 bp (2 - 6 alleles, DI at 0.31 to 0.54)). The 13 loci are distributed between the two chromosomes of the Brucella genome. This 21 loci system was tested on 121 isolates of Brucella (B abortus, B suis, B melitensis, B canis, B ovis and B neocromati) which had been previously been collected from worldwide sources. From these isolates 121 unique profiles were determined for 117 isolates. Only 2 strains of B neotomae and 2 strains of B melitensis had identical profiles. The main findings of this method were as follows:

• This MLVA techneque is able to differentiate between sample of the same biovar from different geographical areas. It showed that the standard B abortus biovar 3 is different from the strain Tulya from Africa.
• Strains of B melitensis, B abortus, B ovis and B neotomae are in discrete clusters similar to the classification by traditional biotyping.
• B canis is closely related to B suis particularly B suis biovars 1,3 and 4.
• B suis biovar 2 is on a separate branch.
• B suis biovar 5 is only distantly related to the other B suis biovars.
• The 6 slightly longer sequence repeats (13 - 40 bp) taken alone was able to identify all isolates.and arrange them into 17 genotypes correspondibg to traditional biotyping results. It is considered that this could be used as a staight forward system to identify Brucella isolates.
• The test showed that all B abortus biovar 1 strains was a different genotype.
• The techneque was able to separate all the B ovis biovars but lack of varation at 13 of 21 loci indicated that B ovis is highly conserved confirming previous results, Ridler (2005) .
• The procedure demonstrates that isolates of B abortus biovar 1 from the UK and Eire were in a different cluster from the same biovar derived from Portugal. Within the clusters individual isolates could be distinguished even in a restricted geographical area.
• The method demonstrate that an isolate of B melitensis from a patient in the UK was closely related to isolates of B melitensis from Portugal and this was in accordance with his most likely source of infection.
• Isolates of B melitensis in 4 patients in the UK was closely associated with an isolate derived from recovered from livestock in Tansania. The 4 had a history of travel to Eritrea and Somalia.
• Strains clustered according to their species and in accordance with classical biotyping designations.
• In general 6 loci were sufficient to identify species and the full range of loci could identify restricted geographical sources.

In general the MLV(VNTR)A technique is particularly suitable for analysis of Brucella because of it's high homogenicity. For this reason this technique can be used for epidemiological studies including tracing strains in relation to phytology and geographic distribution. According to the authors the technical advantages of MLVA over most existing Brucella typing schemes are overwhelming. These MLV(VNTR)A systems of analysis is able to discriminate between Brucella isolates beyond the ability of classical methods. Both of the two methods described here shows great potential for further development and application to both epidemiological tracing of Brucella transmissions and in determining relationships between isolates worldwide. They offer a great improvement in discriminatory power. Also the method is technically undemanding and the results are reproducible between different laboratories. It seems likely that the method can be adapted to type all isolates. The procedures can also be carried out rapidly and the procedure could be automated. The method of analysis is based on PCR and agarose gel electrophoresis which makes this technique available to many users. An advantage is that crude isolates can be examined. This could possibly include the examination of tissue samples without prior culture. The risks of handling of live cultures is minimised. However at the present time comparison to biotyping procedures remains necessary for research purposes.

The reliability of the test with particular reference to isolate stability, i.e. the rate of mutation, has been assessed by Whatmore (2006) in both in vivo and in vitro experiments. In virtually all cases only minor 1 one-step changes were observed. One isolate underwent a permanent one-step change after 14 passages and 270 days of vitro culture. In general it was concluded that the Brucella isolates were sufficiently stable for the MLVA test. Stability studies carried out in vivo and in vitro showed that VNTR profiles were sufficiently stable such that recovered strains could readily be identified as the input strain.

Full Genome Sequencing of Brucella.

Full Genome Sequencing is a laboratory process that determines the complete DNA sequence of an organism's genome. This entails sequencing all of an organism's chromosomal DNA. This technique has become fundimentally importance in the understnding of the structure and function of such organisms. An important objective of Brucella sequencing is to get more understanding about the difference of virulence between Brucella biovars and strains. The genomes of Brucella that have been sequenced to date are:

• The genome of B suis 1330 was sequenced by Paulsen IT (2002) at the Institute for Genomic Research. The authors considered that phage interaction was a possible cause of divergence from the B melitensis genome. They considered that synteny, i.e. physical co-localization of genetic loci on the same chromosome, demonstrated similarity to the plant symbiont Mesorhizobium. (NCBI genome NC_004311.2 )
• DelVecchio of Integrated Genomics (2001) sequenced the genome of B melitensis strain 16M. This contained 2 circular chromosomes similar to those of other alpha-proteobacteria. It was considered that B melitensis genome showed similarities to the symbiotic Sinorhizobium meliloti. (NCBI genome NC_003317, NC_003318)
• Halling (2005) completed the sequence of B abortus field isolate 9-941. The results for this genome show a striking similarities to those for the genomes of B melitensis and B suis with near identical genetic content and shared genetic elements with each of both B melitensis and B suis. (NCBI genome NC_006932, NC_006933)
• Audic (2009) sequenced the genome of B microti and very surprisingly found that B microti was almost identical to the previously sequenced B suis 1330 with an overall sequence identity of 99.84% in aligned regions.
• Crasta of the Virginia Bioinformatics Institute (2008) sequenced the genome of B abortus vaccine s19 (NCBI genome NC_010742, NC_010740). This study enables a comparison of this attenuated vaccine with more virulent strains B. abortus, 9-941 and 2308 and showed there were 45 different genes for s19. it was considered that 24 of these genes which were different in s19 were associated with loss of virulence. 4 genes of greater than 60 bp had consistent difference in s19 compared to both the virulent strains. These genes in the virulent strains, encode an outer membrane protein and three proteins involved in erythritol uptake or metabolism confirming biotyping results.

According to Foster (2009) other genomes of Brucella which have been sequenced so far are:

• B abortus strain 2308 was sequenced by the Oak Ridge National Laboratory. (NCBI genome NC_007618, NC_007624)
• B canis ATCC 23365 has been sequenced by the Joint Genome Institute (NCBI genome NC_010103, NC_010104)
• B melitensis 63-9 has been sequenced by the U.S. Department of Homeland Security (data - unpublished).
• B melitensis Ether has been sequenced by the U.S. Department of Homeland Security (data - unpublished).
• B ovis ATCC 25840 has been sequenced by the Institute for Genomic Research (NCBI genome - NC_009505, NC_009504)
• Brucella suis 40 has been sequenced by the U.S. Department of Homeland Security (data - unpublished)
• Brucella suis 686 has been sequenced by the U.S. Department of Homeland Security (date - unpublished).
• Brucella suis ATCC 23445 has been sequenced by the Joint Genome Institute & Los Alamos National Lab (NCBI genome - NC_010169, NC_010167)
• Brucella suis Thomsen has been sequenced by the U.S. Department of Homeland Security (data - unpublished).
• Ochrobactrum anthropi ATCC 49188 which may be the closest relation to Brucella biovars, has been sequenced by the Joint Genome Institute (NCBI genome NC_009667, NC_009668).

Foster (2009) describes the use of SNPs based on their occurrence in the whole Brucella genome. This analysis is based on the whole genome of 13 Brucella isolates: 5 strains of B suis, 3 biovars of B melitensis, 3 B abortus biovares, 1 biovar of B canis and 1 biovar of B ovis. The genome of Ochrobactrum anthropi which is a soil bacteria and an opportunist pathogern in human which is considered as the closest relation to Brucella was used as a comparison. The analyses of genomes for SNPs was carried out using MUMmer genome comparison software for pair wise comparison, Kurtz (2004) . This method has identified 20,154 orthologous SNPs shared in all genomes potentially able to distinguish between Brucella species. It was concluded that while methods such as (MLV(VNTR)A can give useful information only whole genome can capture the full significance of genetic variation needed for a full understanding of the phylogeny of Brucella. Also as newer and faster genome sequencing methods become available this method becomes more and more viable as an analytical tool, Hall (2007) . This degree of knowledge is necessary to understand the evolutionary framework of the genus Brucella and to design systems able to differentiate the various biovars and strains and fit them into phylogenetic trees and their evolutionary process. Other systems with reduced sets of markers can only give approximate results.

On the basis of Foster's work a simple phylogenic tree of the Brucella species can be constructed in relation to a Ochrobactrum anthropi root. The Main findings were:

• On the basis of Ochrobactrum anthropi rooting B ovis lineage is basal to the rest of the Brucella lineage.
• B. suis is a highly divergent clade only forming a monophyletic clade when the B. canis genome was included.
• A molecular clock based on this data suggests that most Brucella species diverged from their common B. ovis ancestor between 86,000 to 296,000 years ago, which precedes the domestication of their livestock hosts.
• It is concluded that this methodology can lead to an improved understanding of the ecology, evolutionary history, and host relationships for Brucella.
• It is also concluded that this technique can be used to develop genotyping approaches for rapid detection and diagnostic of Brucella for molecular epidemiological and clinical studies.
• This system is flexible and can incorporate newly sequenced genomes as they become available.

Conclusions.

Classical biotyping has laid the foundation for understanding the taxonomy of Brucella. However there is a limit as to the understanding that these methods can achieve. There are also problems, including the handling of live pathogenic bacteria, slowness of obtaining results and the subjective nature of the analysis. The newer genotyping methods are safer, provide quicker results and the results are objective. They are now providing exciting new information about Brucella including geographical distribution of species and biovars, relations between species, biovars and strains and relations to other closely related genera. These results are new and it can be anticipated that much more information with be forthcoming. Therefore a full appraisal is not yet possible. Perhaps the most important of these new techniques is whole genome sequencing and the subsequent analysis. In recent years new species of Brucella have been discovered and the understanding of them has been greatly facilitated by the use of new genotyping techniques which have enabled rapid identification and classification.

Established Brucella Species.

The phenotype of these species has been established by traditional microbiological and biochemical techniques. However host preference plays a key part in their classification. This classification is the best possible within the scope of the biotyping techniques. Full clarification has not been achieved. In recent years the application of genotyping techniques have give better insight but in most regards have confirmed biotyping classifications. The well established species of Brucella are:

• B melitensis biovars 1, 2 and 3 which are classified as serotypes. The host species are domestic sheep and goats
• B abortus biovars 1, 2, 3, 4, 5, 6, and 9. Biovar 7 is not yet been established. The host species is domestic cattle.
• B suis biovars 1 and 3. The host species are domestic pigs and wild boar.
• B suis biovar 2. The host species are the European hares and wild boar.
• B suis biovar 4. The host species are reindeer and caribou.
• B suis 5. The host species is a Russian rodent.
• B ovis. The host species is the domestic sheep.
• B canis. The host species is the domestic dog and possibly wild canidae.
• B neutomas The only known host is an USA wood rat.

There are no reported biovars of B melitesnsis but there are 3 serotypes. Serotypes are of the same species but are distinguished by a characteristic response to a set of antigens (antisera). There are 8 accepted biovars of B abortus but another 14 have been reported, Meyer (1976) , Harrington (1977) , Ewalt and Farrell (1967) . Biovars are of the same species but differ from each other in biochemical or physiological characters. B abortus seems more labile genetically than the other Brucella species. There are 5 accepted biovars of B suis the most recent has been described by Corbel (1984) . B canis, B neotomae and B ovis occur as single species without biovars or serotypes. The extent to which wild animals are naturally infected with species of Brucella such as B abortus and B melitensis is little understood. Brucellosis is difficult to spot in wild animals and there is often a range overlap with diseased domestic animals.

New Brucella Species.

In recent years new species of Brucella have been discovered and also previously discovered strains with unusual properties are being reassessed as possible new species. Thisnew research has been facilitated by the use of new genotyping methods of analysis. In the past effectively only Brucella species affecting domestic animals have been researched. The reason for this has been the need to alleviate the economic impact on farmers and the distress caused by acute and chronic infection in human. In the light of current knowledge it is likely that there are many more host centred Brucella species to be discovered and there are also possibly undiagnosed related infections in humans. It is already clear that many cases of brucellosis in human associated with classical Brucella species go undiagnosed. However the extent of contact between infected animals and humans is a key variable. The new Brucella species affecting marine animals have recieved most attention but they are not fully classified. Further subdivision into new species is possible.The following is a list of possible new species of Brucella which are as yet not fully authenticated:

• B ceti. Host - whales.
• B pinnnipedialis. Host - seals.
• B microti Host - European voles.
• B inopinata strain B01 which has been found as an infection in one human.
• Brucella strain B02. Host - Australian rodents.
• Brucella strain 83-210. Host - wild native rodent species in Australia.

Prevalence of Brucellosis.

Brucellosis in Developed Countries.

The first important research into brucellosis was carried out by the Mediterranean Fever Commission between 1885 and 1910 to investigate the high incidence of brucellosis melitensis (which he called micrococcus melitensis) amungst British servicemen in Malta. This was a highly effective operation. Their successes included: Bruce (1887) identification and cultured the bacteria which caused the illness. Wright developed of a serology test to test for agglutinates caused by the bacteria which is still used today. Hughs reported a detailed description of the illness in humans including its chronic nature. Zammit as reported by Eyres (1908) and Zammit (1905) showed that goats also had the disease and that it was present in their milk and that humans became infected by drinking such milk. He observed that the disease in goats was long lasting and chronic. By banning the consumption of raw milk by servicemen the achieved a dramatic reduction in the incidence of the illness. The native Maltese who did not observe the new rules continued to be affected by the illness. B melitensis which has been found to affect both sheep and goats has subsequently been found to be present throughout the whole medieranean area. The methodology used to eradicate B abortus in cattle has been appled but eradicating B melitensis has proved very difficult. Among the reasons for this are: different husbandary methods and the greater virulance of B melitensis.

In the USA all 4 types of brucellosis that affect humans are or have been present. B abortus, B melitensis, B suis and B canis. The main problem has been B abortus affecting cattle. B abortus was also the main problem in Northen Europe. This problem was not new, as indicated there was historical evidence for a contageous disease causiong abortions in cattle, goats and sheep going back through the ages. B abortus was identified by Bang (1897) . In the USA it was generally known as contageous abortion because it was known to be transmitted between cattle. Brucellosis in cattle caused significant ecconomic losses for farmers so that there has been sustained efforts to eradicate it from cattle herds. Altogether this has taken about 80 years. Effectively the methofology of eradication had to be learned by trial and error and by developing appropriate technology.

• Early Efforts. In the early part of the twentieth century by individual farmers made attempts to eradicate it from their herds by eliminating infected animals. In practice this was very difficult because the disease cwas insufficiently understood. A major problem was the necessity of introducing new and potentially infected animals into these closed herd. Brucellosis re-introduced in this way could then rapidly spread back through the herd. This was a common problem. In these early days there was also a lack of local and central government support. Slowly governmrnt measures were introduced to help farmers. The aim of such meaasures were to find all infected herds, to prevent restrict existing infection to known recevoirs and to eradicate brucellosis from such recevoirs. States acting individually banned the sale of infected cattle except for slaughter. By 1934 a number os States had set up volunary control and certification programmes to test blood from cattle were started using government laboratories.
• 'Segregation. An important part of the control of brucellosis depends on segregation and slaughter of infected animals. An essential part of this is to provide adequate compensation to animal owener. The greatest risk of spreading the disease is at parturition and this can include apparently healthy cows. THerefore segragation of pregnant cows during birth was essential.
• Tagging. Infected cattle going through markets were a real problem in causing the spread of brucellosis. There was a need to test and trace such animals. There was also a need to trace infected cattle tested at slaughter or in markets back to infeced herds. A comprehensive sysrem of tagging of cattle was essential.
• Serology Test. Serology and milk tests have played a vital role in eradication of brucellosis but it should be said that no test has been developed which works in all cases. In the early days only the Wright's agglutination test (SAT) was the only test available. The main problem of the test was that it was carried out in the laboratory and took 2 days to get a result. A modified version of the test was introduced by Huddleson (1926) which gave similar results to SAT but was quicker and could be done on site if necessary. These tests only measured IgM and missed some infected cattle. The Gompliment Fixation Tesr (CFT) tites was available from 1909 and used in Europe but not th USA. It was re-introduced in the USA in the 1960s. This test finds more infected animals because it measures IgG. The milk ring test which was a rapid test for Brucella infected milk greatly improved the detection of brucellosis in dairy herds. It could be used to test milk on farms both for individual cows and mixed herd milk. In the 1960s 2 new tests were introduced.: the CARD test in the USA and the Rose-Bengal test in Europe. Both tests used an acid buffer and measured IgG. Both could give rapid results under field conditions. These could be used at markets and slaughter houses.
• Vaccination. Many trials have been made to make a vaccine for B abortus in cattle. The main success has been B abortus strain 19 (s19) which is a live attenuated vaccine developed by chance in the USA. This has been widely used. Vaccination can provides protection to brucellosis free cattle but does not cure already infected animals. Whole herd trials with the vaccine caused trouble because pregnant cows aborted. Therefore it has been restricted to calves. Vaccination is probably best used in cases where brucellosis is widespread in a country to reduce the overall level of the diease. Test and slaughter is better in situations where the overall level of nrucelloosis is lower. In more recent work lower dose vaccine has been tried. At lower dose levels the rate of abortion in pregnant cattle is reduced. Such vaccination is continued over multiple generations.
• Elimination. In most North American and European countries B abortus has been eliminated or reduced to very low level by the 1980s. In the USA and Canada there is a localised risk of cattle becoming infected by contact with wild infected bison or elk. There is also a risk associated of B suis from feral swine in the USA, from carabou in Canada and from wild boar in Europe.
• B abortus in Humans. Unlike B melitensis the links between brucellosis in cattle and humans was not found immediately. The link was not nade until Evans (1918) identified the presence of B abortus in milk, and in 1924 recorded the first case of a human infected from cows milk with B abortus. Her discoveries were received with outrage by the dairy industry. She advocated pasteurisation which kills Brucella. This was gradually introduced. Epidemological studies have shown that where B abortus was prevalent in cattle there was a high incidence of people reacting to brucellosis serology or skin tests indicating that they had been in contact with Brucella. The extent to which these people were infected is not known. With or without antibody treatment many infected people go on to develop a chronic form of the illness. The serology tests available then and now are inadequate to diagnose such illness and the symptoms are vague or can mirror other diseases.

Epidemological Studies in Developed Countries.

There is little doubt that before the use of pastuerisation large parts of the population of North America and Europe were exposed to possible infection by B abortus. It is possible to say how many of these developed sub-acute or chronic illness which went undiagnosed. Early studies from the 1930s onwards using intradermal and SAT tests have shown high levels of positive response in humans indicating direct contact with Brucella although they did not necessarily have active disease. There was also a problem of brucellosis in humans due to direct contact with diseased animals. This affected workers on farms and their families, veterinarians, and those involved in the slaughter of diseased animals and preparation of meat. This latter problem was gradually eliminated as diseased animals by test and slaughter in cattle and swine. Currently the brucellosis in humans has been reduced to low levels.

The following are examples of surveys into the level of exposure an disease in humans from1930 onwards:

• Levin (1930) tested 365 individuals using the intradermal skin test using fat free ground up Brucella and found 7% positives and 55.6% of these had clinical symptoms of brucellosis.
• Meyer (1935) found positive intradermal skin tests in 10% of students and 60% of veterinarian.
• Goldstein (1936) found that 10.1% of randomly selected hospital patients gave positive intradermal reactions.
• Gershenfeld (1937) carried out a survey of the incidence of Bang's disease among cattle in the USA. He found that 38.4% of herds tested had positive reactors. In infected herds 14.9% were positive. Of 11,858,859 cattle 8% gave apositive agglutination test.
• Angle (1039) noted that the incidence of positive reactors in cattle in the USA was at 14.9% and considered the possibility that many people were becoming chronically infected from drinking raw milk. He carried out a survey of school children in Kansas City to estimate the possible level of chronic brucellosis illness (ambulatory, subacute and latent). 7,122 children were tested and 642 gave a positive bucellergen skin test of whom 462 were interviewed concerning their state of health including nervous related symptoms, headaches, rheumatism, constipation and fever. The incidence of children with 1 or more symptoms is shown in Table x. The results for specific symptoms is shown in Table y. Nervous symptoms include: depression, irritability and emotional instability. Other symptoms which were mentioned but not corelated include: fatigue, lassitude and anorexia. Symptoms persisted after a fUrther 1 year in 57,5% of cases. A control group of 100 children with negative brucellergen skin tests is used for comparison. The significant lever of symptoms is a surprise but it can be noted that the brucellergen test does not work in all cases. They concluded that chronic brucellosis was a likely cause of the symptoms described and that chronic brucellosis was not uncommon. They also showed that the percentage of positive skin tests increased with the age of the child presumably related to increased exposure to raw milk. Similar results were found by others.

TABLE x

Diagnosis	First Test at 0 Years	Second Test at 1 Year	Control Group.
0 Symptoms	25.1%	26.7%	67%
I Symptom	36.1%	36,1%	20%
2 Symptoms	17.3%	18.2%	10%
3 Symptoms	10.2%	9.4%	3%
4 Symptoms	6.9%	7.2%	0%
% Symptoms	4,3%	4.3%	0%

TABLE y

Symptom	Positive Skin Test	Negative Skin Test
Nervous.	41.3%	26.0%
Headache.	37.3%	15.0%
Rheumatism	34.0%	6.0%
Constipation.	15.3%	2.0%
Fever.	4.8%	1.0%

• Lovell (1940) showed that of 30,892 pregnacies in cattle in England that 9.7% ended abnormally of which 5.4% were abortions and 4,3% were still born. It is assumed that most were due to brucellosis. Stableforth (1951) notes that at the same time 20% of cows had Brucella antibodies in their blood.
• Dustin (1940) considered that chronic brucellosis was prevalent in Southern New England and caused significant disability among the population often preventing physical activity. In a study of 4000 individuals in Rhode Island 10% had clinical and laboratory evidence supporting a doagnosis of chronic brucellosis and 18% had positive reactions to 1 of 3 tests: intradermal skin test, intramuscular test and agglutination test. The intramuscular test involved injecting 0.5cc of the antigen (aqueous solution containing 2,500,000 bacteria. Uninfected persons give only the slightess temperature rise whereas a positive rsult involves a temperature rise of ≥100°F. Concurrent with this research work it has been found that 10% of cattle were positive for B abortus and another 10% were suspicious.
• 1949. Magoffin notes that during a brucelloisis control program in Minesota from 1934-9, 11.4% of cattle gave positive agglutination titres. In 1946 it was 8.2%. It is suggested that the incidence in pigs infected with B suis and B melitensis must have been similar. The incidence in humans was: B abortus at 85.8%; B melitensis at 8.2% ; and B suis at 6.0%.
• Stableforth (1951) indicates that the USA Federal project to eliminate brucellosis started in 1934 when the estimated level of disease in cattle was 10%. Vaccination and test and slaughter were used. Re-infection rate was 2.5%. By 1949 a survey of 5 million animals indicated that the infection rate had dropped to 4% as indicatee by the agglutination test. He notes that the reported incidence of brucellosis in man in 1950 was 3163 but suggests that the real figure was probably 100,000 cases. At this time the the reported level of brucellosis in cattle herds was 16% of herds and individual cows at 4%.
• Damon (1953) reviewed animal brucellosis in rural Indiana. 13,373 farm animals wre found to be reactors (positive agglutination tests) and 28.0% of farms harboured reactors. 4.0% (8.6% with suspects) of swine, 6.9% of cattle, 2.5% 0f horses, and 0.7% of sheep were reactors. 10% of cattle had been vacinated. Infection associated with negative serology tests are not considered
• Saddler (1960) notes that in 1960 in the USA pasteurisation of milk had removed the risk of Brucella infected milk but that raw milk was still consumed in rural areas.
• A snapshot of the incidence of brucellosis abortus in human and cattle is provided by a survey during 1959 in the Isle of Wight, UK is provided by Brodigan (1961) , The Isle of Wight has a mixed rural and urban population of approximately 94,000. Approximately 50% of children age 6 (685) were tested using the skin test. 14.4% were rural children of whom 12.5% were positive. 78.7% were urbabn children of whom 4% were positive. Raw milk was consumed by 37% of urban children and by 67% of rural children, It is considered that the milk drunk by rural children were more likely to be contaminated and also they had more possibility of contact with infected animals. A survey was also carried out of 277 pschiatric patients of whom 18.8% gave positive skin tests. 132 patients were tested using SAT. The results are shown in Table ...... 16 of 20 patients with positive SAT had symptoms of depression. A survey of 1287 blood donors and 108 expectant mothers by the SAT test are also shown in Table .... and gave similar results. Culture tests on 111 pairs of tonsils were all negative. In general, on the basis of skin and serum agglutination tests 17% of the adult poulation gave positive results and but for the proportion of the population that drank raw milk the rate was 70%. The author notes the similar results for a Wiltshire village with a heavily contaminated milk source where the rate of positives was 78%. These are considered subclinical infections without active symptoms. The incidence of positive results for workers with occupational exposed to Brucella are shown in the Table.... Tests on cattle were carried out concurrent with the tests on humans. The overall rate of infection for cattle was 6.3% as tested by MRT and 33 of 520 registered herds were infected. This indicated that brucellosis infection was widespread in cattle at this time.

SAT test	Psychiatric Patients	Blood Doners
<20	86.0%	83%
≥20<40	14.0%	17.0%
≥40<80	7.6%	9.0%
≥80<160	3.0%	5.0%
≥160	0.4%	4.9%

Occupation	Skin Test (+ve)	SAT (+ve)
Vetinarians.	87.5%	87.5%
Public Health Workers	75%	25%
Abattoir Workers	61.5%	30.8%
Sausage Makers	22.7%	18.2%
Bone Meal Workers	80%	40%

• Hendricks (1964) reviewed the incidence of brucellosis in Iowa from 1952 to 1963. During this period he notes a marked decline in the incidence of disease in farmers from 186 cases in 1952 to to 27 cases in 1963. This corresponds to the introduction of the State and Federal sponcered bovine brucellosis eradication programs involving scranning cattle herds in the period 1951-56. During this period the percentage number of cattle testing positive for brucellosis reduced from about 8% in 1951 to about 1.5% in 1963. However during the same period the incidence of brucellosis in abattoir workers increased eratically from about 60 in 1952 to about 120 in 1963. It is suggested that this relates to the high exposure of such workers to pigs infected mainly with B suis and B melitensis. In the same period the incidence brucellosis in other workers (livestock traders and dealers, vetinarians, rural students, rendering plant employees and rural housewives) declined but more slowly than for farmers.
• Morgan (1969) notes that the incidence of brucellosis in humans drops as brucellosis is reduced or eliminated in animals.
• Hendrick (1997) in a survey of milk drinkers in California noted that 3,2% of respondants drank raw milk. 28 US states allow the retail of raw milk. The reasons people drink raw milk are lack of lack of knowledge of the possible dangers and taste preference.
• Medcalf (1986) estimates that in 1934 when a federal-state eradication programme began that across the USA 10% of all cattle were infected by B abortus. The infection rate ranged from 3 to 20%.

B abortus is the only form of the disease to affect domestic animals in he UK. In the UK before 1940 no attempt was made to eradicate brucellosis in cattle in the UK apart from trials. In 1940 s19 was introduced to vaccinate calves and cattle. The trials indicated that test and elimination of reactors is effective but clean herds can easily br reinfected from outside. Lawson (1950) showed that in large scale use of s19 in cattle can significantly reduce the level of abortions to about 2%. However some herds react unsatifactorally to this method. Similar results have been found in other countries.

Brucellosis in Developing Countries.

Although B abortus has been effectively eliminated in cattle in developed countries it remains prevalant in developing countries. Canada, the USA, Northern Europe, Southeast Asia, Australia and New Zealand are free of B melitensis, but it is the most prevulent form of the disease elsewhere. B melitensis appears to have evolved in the mediteranean and middle east area. B melitensis is highly zoonotic and other animals associated with diseased sheep or goats can readily become infected. This includes cows, yaks, camels and buffeloes. This also applies to humans and it is the major cause of brucellosis in humans. Eradication protocols for eradication of B melitensis in sheep and goats are based on those developed for cattle. B melitensis has proved more difficult to control and eradicate in sheep and goats than B abortus in cattle.

Circumstances Favouring the Development of Chronic Brucellosis.

Acute brucellosis is relatively easy to diagnose with antibody serology and Brucella culture tests in combination with the patients hisory of illness and symptoms.

• Jordon (1931) suggests that on the basis of large scale studies in Iowa the ratio of latent or subclinical cases to clinical cases is 8:1.
• Cameron (1934) studied 46 cases of brucellosis of which 10 remained chronically ill for long periods and 6 were confined to bed for upto 2 years.
• Angle (1935) studies 100 brucellosis patients. He divided them into acute and chronic patients. In the latter ceses symptoms persisted for long periods and neurological symptoms were most common.
• Cases of prolonged brucellosis are described by Thames (1935) Marietta (1935) have described cases of chronic brucellosis.
• Kerr (1968) noted that if brucellosis is not adequately treated and the infection is not resolved it may progress to a chronic state due to localization of Brucella in various parts of the body.
• Buchanan (1974) treated 160 patient and diagnosed all within 30 days. He found that patients treated early had fewer complications.
• Sippel (1982) notes that in rural Egypt immediate diagniosis and treatment is seldom possible and patients usually develop long term illness because of inadequate antibiotic treatment and their immunology my include elements of both acute and chronic illness.
• Mousa (1988) observed that 4 patients who received inadequate treatment all relapsed.

Ariza (1992) showed in a study of 75 patients that the extent of focal illness increased with duration of illness before hospitalization/treatment: <7 days, 0%; 8-15%, 19%; 16-60 days, 26%; >60 days, 77%.

• Salata RA (1996) noted that prompt treatment is essential to a good outcome. Spink (1956) in a study of 112 patients found that the complication rate depended on the duration of illness before treatment: up to 30 days; 0%; 31-60 days, 0%; 61-90 days, 25%; >90 days, 80%.
• Colmenero (1996) in a study of 530 patients with brucellosis found that the amount of complication/focal illness depends on the period of illness without treatment: <15 days, 22%; 15-30 days; 32%; 30-90 days, 421%; >90 days, 53% and also linked to therapeutic failures.
• Corbel (1997) wrote that the diagnosis of brucellosis is not difficult if the level of suspicion is high and the presentation is typical. A suitable combined antibiotic regeme (see later) generally leads to full recovery in 85% of cases. If a relapse occurs, a second course of treament is usually sucessfull. In some cases onset is insiduous and then diagnosis is more difficult. The main cause of chronic brucellosis is delayed diagnosis and inappropriate treatment.
• Landau (1999) studied 9 patients with brucellosis mellitensis. They were ill for 102 days before treated. Although treatment was initially relatively successful they all became chronically ill for more than 5 years. It was concluded that delay in antibiotic therapy caused the illness to become chronic.
• Solera (2005) compared 146 patients divided into 2 equal groups treated with doxycline (100 mg po bid) for 30 days or 45 days plus gentamicin (240 grammes daily) for seven days. The patients were free from the complications: neurobrucellosis, spondylitis and endocartitis. In the first 45 days after completion of treatment the relapse rate in the 45 day group was 1.37% compared to 12.3% in the 30 day group. In the next 311 days the relapse rate was similar in both groups at 9.36 in the 30 day group and 11.11 in the 45 day group.The have shown a a high relapse rate with a 30 day treatment duration.
• Rust (2006) comments that chronic Br is most likely to occur in cases where antibiotic care is inadequate.

Transmission of Brucella.

Virulence of Brucella.

Brucella in the Laboratory. Brucella is a dangerous pathogen which requires Biological Safety Level 3 containment when handled in the laboratory. Prior to the introduction of these standards many laboratory workers have become infected. For this reason it has been studied as an agent for germ warfare.

°C, >6 days.===Durability of Brucella in the Environment.=== Brucella can stay active in a contaminated environment for many months depending on conditios. This has been described by Wray (1975) , Nicoletti (1980) and Corbel (2006) and Bercovich (1998) and EC SANCO.C.2/AH/R23/2001. . Brucella can survive in dust, manure, water, manure slurry, aborted fetuses, soil, meat and dairy products for considerable periods. Survival depends on suitable temperature, pH, and humidity. Conditions likely to kill Brucella are heat treatment, disinfection, direct sunlight and dryness. Some idea of conditions under which favour survival in B abortus etc described:

Environment.

• Frozen tissues. <3 years.
• surfaces. 5 hours
• Water. Tap water (-4°C), 114 days; lake water (8°C, pH 6.5), 57 days
• Soil Dried at 18°C, 609 to 72 days; wet <7 days; Humid atmosphere, >2 month. Autumn (90% Humidity)48 to 73 days. wet soil (<10°C), 66 days
• Urine., pH 8,5, 37°C. 16 hours; pH 6,5 68°C, 6 days. urine, 39 days
• Whey. 17 to 24°C' 24 hours; 5°C >6 days.
• Raw milk. 25-37°C 24 hours; 8°C 48 hours; -40°C 2.5 years.
• Manure. Winter, 85-103 days; spring, 120-210 days; summer, 30-180 days; and 50-120 days'
• Manure slurry. Summer, 3 months; Winter, 6 months; tank, 1 to >8 moths (12°C).
• Aborted fetuses. 75 days.
• Wool. 4 months.
• Hay. 2 to 30 days.
• Street dust. 3 - 44 days.
• Wood. In sheds and pens. 4 months.
• Pasture. Sunlight, <5 days' Shade, > 6 days.

Food.

• Milk.. At 0°C, 18 months and at 25 -37°C. 24 hours.
• Cream. At 4°,. 4 - 6 months.
• Icecream. At 0°C, 30 days.
• Butter. At 8°C, 142 days.
• Cheese (In general). 6 - 100 days.
• Cheese (Feta). 4 -16 days.
• Cheese (Pecorino). <90 days.
• Cheese, Roquefort. 20 - 60 days.
• Cheese (Camembert). <21 days.
• Cheese (Cheddar). 6 months.
• Cheese (White). 1 - 3 weeks.
• Whey. At 17 -24°C, pH 4.3 - 5.9, <4 days.
• Whey. At 5°C, pH 5.4 - 5.9, >6 days.

Susceptibility of Animals and Humans.

A numbers of studies have been carried out to determine the threshhold susceptibility of Brucella infection in animals and humans. It is probable that early studies were affected by the diminished activity of Brucella due to too much re-culture of the Brucella which reduced itss virulence.

• Early studies by the Mediterranian Fever Commision in Malta concluded that 5000 viable B melitensis organisms could cause infection in macaque momkeys.
• Early trials in the USA of feeding B abortus to human volunteers in milk by Cooledge (1916) and Morales-Otero were unsucessful. It is suugested that this was partially due to the reduced pathogenicity of B abortus but also to the organisms had been compromised by re-culturing. 2 individuals given B suis bacame ill.
• Nicolle (1923) failed in attempts to inoculate 5 individuals subcutaneously with B abortus.
• Huddelson (1929) noted that in monkeys, B suis caused severe illness whereas B abortus generally caused a milder illness.
• Huddleson (1942) notes that guinea pigs are readily susceptible to Brucella. Disease may be induced by subcutaneous injection of 25 live B abortus or ≤5 B suis bacteria.
• Elberg (1948) considered that the respitatory route of infection was a little less effective than parental injection. Their results with B suis show ID₅₀ for respitory at 36 organisms and ID₅₀ for parenteral injection at 5 organisms. Parenteral ID₅₀ for B abortus and B melitensis were 500 and 50 respectively.
• Berman (1949) found that the degree of exposure determined the incubation period and the rate of abortion with B abortus in cattle.
• Carle (1950) reported than all 17 guinea pigs injected subcutaneously with 19 organisms of B suis developed a agglutination titre of 1:2560 at six weeks.
• According to Renoux (19955) and Elberg (1959) determined the ID₅₀ for B melitensis in goats at 20,000 to 50,000 CFU and 400,000 CFU in sheep.
• Manthei (1959) determined the ID₅₀ (infection in 50% of animal;s) of virulance of B abortus in cattle at 350,000 colony froming units (CFU).
• Ne'eman L (1963) have found that only 11 Brucella are sufficient to cause a brucellosis infection in guinea pigs.
• According to Jubb (1970) referring to the work of Berman (1949) the establishment of brucellosis in cattle depends on the resistance of the animal, its age, reproducive status and degree of exposure.
• Taran (1972) compared the susceptiblity of Brucella inoculated guinea pigs, mice, rats and sheep. Guinea pigs inoculated with 10 -100 cells of B melitensis or B suis developed granulomatous lesions. In mice 10⁴ to 10⁶ organisms and in rats 10⁹ organisms while the sheep used were resistant to the strains of B melitensis and B suis used.
• Garcia-Carrillo (1977) .
• Thorne (1978) that the incubation period in cattle is from 14 to 251 days but the incubation period is shortened by pregnancy.
• Al Dahouk (2003) notes that 10-100 bacteria inhaled as an aerosol can cause infection.

Modes of infection.

General modes of infection in both animals and humans include: ingestion of contaminated food, contact through abraded skin or mucosal membranes with contaminated meat or animal carcasses such as at abattoirs including breathing in aerisols, direct contact with contaminated animals particularly involving abraded skin and mucosal membranes particularly during midwifery (veterinaries and farmers), contact with a Brucella contaminated environment including touchinng conaminated surfaces and breathing dust.

Hardy (1929) showed by experiments with guinea pigs that Brucella could be adsorbed through the skin. Brucella was applied to the skin as follows: shaved guinea pigs with abrasions, 100% infected; guinea pigs shaven without deliberate abrasions, 90%; guinea pigs with cloppied hair, 80%; fed by mouth, 20%. Johnson (1954) noted that the most common mode of infection was by injestion of milk or milk products. Others methods of transmision included by skin contact and inhaling dust such as occurred for stockmen, veterinarians, dairy men, packing-house workers, and laboratory workers. According to Spink (1953) the order of severity of illness caused by Brucella is B melitensis >B suis > B abortud. B melitensis is carried by goats and sheep and in such animals the bacteria is present in milk, urine, feases, and vaginal discharges after abortion. The environment associated with diseased animals may become heavily contaminated with the organism. Other animals can become infected by the injestion of contaminated food, inhaling contaminated dust particles or through skin leasons. Humans exposed to the same environment can also become infected. The disease can also be contracted from urine/feacal contaminated fruit and vegetables, meat, manure and handling animals.

Herr (1994) notes that in sheep and goats transmission is associated with persistent shedding of large amounts of Brucella following abortion or parturition.

Bercovich (1998) suggests that pregnant cows are more likely to become infected than males possibly due to tropism for the uterus, Nicoletti (1980) . The factors affecting resistance to disease are age, sex and cattle breed. Hellmann showed that the rate of infection of cattle with brucellosis in tribal Sudan occurred at a different rate in 2 breeds of cattle.

During a serological Brucella survey in Bahr el Ghazal Province (Southern Sudan), 6.5% of a total of 5982 randomly selected cattle of the Dinka tribe from 303 herds investigated showed positive results. Herds of bigger size were found to be more frequently infected than smaller herds. From 1228 randomly selected cattle of the Fellata tribe (Red Bororo race), 22.5% produced a positive reaction. Here, reactors were detected in 71 from a total of 72 herds tested (98.5%). Among the Dinka cattle, the number of reactors rose steadily in correspondence to the age of the animals. A somewhat more pronounced increase of reactors was observed in animals 4-8 years of age. On the other hand, the number of reactors in the Fellata herds increased rapidly between the age of 2-4 years. Thereafter, this infection rate did not change significantly any more. The results indicated that a significant portion (37.1%) of reactors graded positive (n = 665) expressed only low titers (less than or equal to 1:40) in the SAST. Those titers were even more numerous (49.1%) among the positive reactors belonging to cattle below the age of 4 years.

Infection with Brucella usually results in the induction of both humoral and cell-mediated immune responses, but the magnitude and duration of these responses is affected by various factors including the virulence of the infecting strain, the size of infecting inoculum, pregnancy, sexual and immune status of the host FAO/WHO (1986)

Severity of illness.

Severity in animals.

A number of studies have been carried out to determine the comparative severity of B abortus, B suis and B melitensis in animals. The general conclusion is the B melitensis is the most virulant and B abortus the least. However it is probable that once established the complications associated with B abortus are as severe as the other forms.

• Huddleson (1929) infected monkeys with B abortus, B suis, and B melitensis and considered the the order of severity of infection was B suis> B melitensis> B abortus. The illness caused by B suis was severe and the illness caused by B abortus mild. They orally infected Macaccus rhesus monkeys with B abortus or B suis or B melitensis. They found that B suis and B melitensis were far more viruant than B abortus. The monkeys with B suis and B melitensis had far more severe physical signs of illness. 10 monkeys were exposed to B abortus, 7 to B suis and 4 to B melitensis.

B abortus. 4 monkeys showed no signs of infection. 2 were exposed to infected milk, 1 to an ex-human culture and 1 to fetal material. 1 exposed to infected fetal material became infected. 2 exposed to ex-human culture developed limited infection. 1 became infected after long exposure to infected milk.

• B suis. All 7 monkeys became infected. The sources of infection were: 4, culture ex-hunan; 1, direct contact; 1, hog fetal material; 1, culture from bull testicles.
• B melitensis. All 4 monkeys became infected. Sources of infection were: 2, culture ex-human; 1, culture unknown; and one tissue from cow's udder.
• Braude (1951) carried out experiments to determine the virulance of a range of strains of B abortus, B melitensis and B suis in experimentally infected guinea pigs (18 for each Brucella variant). B melitensis was the most virulent and B abortus the least. 3 of 18 guinea pigs infected with B melitensis died and 1 each for B abortus and Brucella suis. The general physical condition of the guinea pigs with B melitensis was worst (roughening of fur, hair loss, and listlessness) and they gained the least weight (average wight gain: B abortus, 189g; Brucella suis, 126g; B melitensis, 119g). Apart from the one death, the guinea pigs with B abortus were apparently outwardly unaffected by the infection. In the case of Brucella suis apart from the one death the guinea pigs appeared in fairly good health. Abscesses were most common in Brucella suis and macroscopic leisons were observed in 16 of 18 guinea pigs. In the 18 guinea pigs infected with B melitensis macroscopic abscesses were observed in 5 animals and microscopic leisions in 9. In the 18 guinea pigs infected with B abortus macroscopic abscesses occurred in 1 but there was wide spread involvement of tissue with microscopic leisions.

Mackaness (1967) studied delayed hypersensitivity in mice which are less susceptible to brucellosis than humans. From the start of infection they maximum infection and delayed hypersensitivity occured on the 4th day and hypoactivity of of macrophages at 8 days with a fall in the number of of bacteria.

Hardy (1930) studied over 300 cases of B abortus and B suis in Iowa USA and considered that concurrent disease in animals. He noted that in domestic animals the desease could be acute, subacute or chronic and could be extensive. His research was mainly directed at cattle and hogs. The effects in preganancy on the uterus were: chorinitis, placentitis, and endometris and the typical signs were: abortions, permature births, retained placentae, sterility or complete absence of all signs. The udder was subject to chronic inflamatory foci or microscopic leisions with Brucella present in the milk. Generally there wre no obvious leisions in other organs but urine and fecal material could contain Brucella. The risk to humans related to contact with vaginal discharges, fetuses, placentae, urine, feces, carcuses (blood and tissue). Farmers may also be infected through the skin during milking by direct contact or from handling raw contaminated milk via the skin.

Severity in Humans.

Studies have been carried out in relation to the desease in humans. Just as in animals the severity of illness was B melitensis> B suis > B abortus. B abortus seems to give rise to a greater proportion of mild subclinical illnesses but can sometimes be severe.

• Hardy (1930) compared patients with proven B suis and B abortus and found B suis caused more severe illnesses. His results for hunan patients are shown below. The results for B suis (2) represent patients who had been infected by contact with pigs and therefore most likely to be B suis and B abortus (2) were patients infected from cattle and therefore likely to be infected with B abortus.

Severity of Illness.	B suis.	B abortus .	B suis (2).	B abortus (2).
Fatal.	3	0	2	1
Severe.	23	5	19	33
Mild.	4	3	5	36
Ambulatory.	3	5	8	25

• Johns (1933) studied 100 patients in a mental hospital who had potentially drank raw milk from a herd of cattle infected with B abortus. They were reviewed on the basis of clinical investigation and agglutination, CFT and skin tests for brucellosis. On this basis 40% were considered free from disease, 23% as actively infected and 38% to have been infected previously. In all cases clinical symptoms were mild.
• Jordon (1943) compared 2 outbreaks of brucellosis in towns in Iowa due to the consumption of contaminated raw milk. One caused by B abortus and the other by B suis and in each case one dairy was involved. For B abortus it was found that 9 of 13 cows suppying the dairy were reactors (positive agglutination titres). For the B suis outbreak 4 cattle at the dairy were found to be reactors for B suis. In the B abortus outbreak 62 people who drank the infected milk were tested. 71.4% had negative serology agglutination titres. Of 18 with positive titres (29.6%): 17 had tires of 1:5 to 1: 20 and 1 a titre of 1:40 to 1:80 (83 people tested at other dairies at 20.2% positive). Of 81 people tested by brucellergen skin tests 14.1% were positive (99 tests for other dairies at 12.1%). For the B suis outbreak 45 people were tested by the agglutination test and 63.7% were negative. 12 were positive with tires of 1:80 to 1:1256 (the tires of 171 people tested for other dairies had 0% positives). 77 cases of brucellosis developed during the few months after the start of the outbreak. 27 people who had positive titres of 1:160 to 1:1256 were re-tested approximately 1 year late. 7 were negative and 16 had titres of 1:10 to 1:40. 101 people were tested using the brucellergen skin test of whom 73.7% were positive (Of 158 people who drank raw milk from other dairies 10.8% were positive). These results seem to demonstrate the greater susceptibility of people to B suis over B abortus.
• Magoffin (1949) noted that in a survey of human brucellosis cases in Minnesota 1945-8 that the incidence was as follows: Farmers, 38.04%; abattoir and stockyard workers and butchers, 21.17%; general workers (male), 12.94; farm women, 8.62%; general workers (female), 7.85%; farm children, 6.28%; urban children, 3.58%; vetinarians, 0.78%; laboratory workers, 0.78%. Patients with negative serology tests are not included. The likely causes of infection were contact with infected animals and ingestion of raw milk. Fitch (1070) and (1039) had previously demonstrated that raw market milk was contaminated with B abortus. Magoffin notes that this indicates that a signoificant proprtion of the population of Minnesota was repeatedly exposed to viable Brucella in raw milk. Some showed no clinical or laboratoiry evidence of infection; some developed developed serum agglutination titres and positive skin tests and some developed clinical illness and bacteremia.
• Spink (1953) notes that in a direct comparison of B abortus and B melitensis the relapse rates are respectively 20% and 50%. He notes that prompt antibiotic treatment is essential to prevent a chronic and debilitating illness due to B melitensis.
• Corbel (2006) notes that susceptibility to brucellosis in humane depends on a number of effects: the response of the immune system, route of infection, size of inoculum and the species of Brucella. Although 'B melitensis and B suis generally cause more severe illness but the illness in B abortus and B canis can sometimes also be severe. Possible modes of direct infection were via cuts in skin, the conjuncival sac, inhalation of aerosols, injestion of contaminated food (milk etc), blood transfusion, tissue transplant, sexual transmission. The latter 3 are rare.

Intra-Species Transmission of Brucella.

In general. The transnission of Brucella infections between animals of the same species is either oral or sexual or occasionally through the skin. Corbel (2006) notes that in mainly a subacute or chronic disease in animals usually without an apparent initial acute phase. The disease locallises in the placenta and udder of female and the testes and epididymis in males. Typical signs of brucellosis infection are: abortion, premature birth, retained placenta. Cows, etc usually abort only once. Arthritis due to B melitensis occurs in sheep and goats. Bursae and local abscesses may be the only symptoms in horses. Bacteraemia is usually intermittent.

• Contaminated environment. Corbel (2006) Bringing herds or flocks of cattle and sheep together accentuates spread of the disease due to contamination of the environment. In this way brucellosis can be spread to othe species. These include: B abortus in cattle or buffeloes to humans or dogs or horses; B melitensis from sheep or goats to humans or cattle or dogs or horses (rare); B suis from pigs to humans or dogs(rare) or cattle (rare) or horses (rare). Possible modes of contamination are: contaminated dust, dried dung, and contaminated surfaces. Possible modes of infection are: dust or aerisol inhalation, skin and mucosal membranes including conjunctivae. He notes that the incubation period may be long with associated negative serology. Hardy (1930) notes that urine and fecal material may contain Brucella.
• Direct Contact. Laing (1956) suggests that the main route of B abortus infection in cattle is usually oral but it can sometimes occur through the skin. The most important mode of infection is oral contact with contaminated abortion and birth products. Cattle are known to lick such birth products and also vaginal discharges from infected aborted and infected cows, Payne (1959) , Plommet (1973) and (1972) . Poulding (1939) studied 12 brucellosis free goats which were penned with infected female goats which had just aborted. 8 goats became infected with B melitensis of which 6 aborted.
• Sexual Contact. Burrows (1968) considered that the products of abortion play a significant part in oral transmission. Jubb (1970) notes that once a bull is infected, possibly with orchitis, then sexual transmission either naturally or by artificial insemination can occur and thereafter infected cows may become infertile. Natural sexual intercourse in catt,le is less likely to cause Brucella infection than artificial insemination, . Acutely infected bulls shed B abortus in their semen ut becomes intermittent in the chronic phase, McCaughey (1973) . Rankin (1965) notes that although the semen of bull cattle may be infected with B abortus this is not an important means of transmission of brucellosis in cattle. However Schroeder (1922) considered the possibility that semen leaking from infected bulls could contaminate food and The Nat Res Council (1977) considered the posibility that cows could contract brucellosis by licking infected semen leaking from other cows. Oriel (1974) and Roberts (1971) consider that transmission of B suis in pigs sexual has a major role. Infected boars develop epididymo-orchitis and produce Brucella infected semen and can infect many sows. Oriel (1974) considers that transmission of B canis in dogs is mainly sexual.
• Dam to Offspring Transmission. The offspring of an infected dam may be infected either in utero or by Brucella infected colostrum or milk, Catlin (1986) , Lapraik (1975) and Stringfellow (1983) and (1988) . Oriel (1974) observed that sows infected with B suis develop granulomatous leisions and mucosal cysts of the endometrium from which 'Brucella can be cultured. This is a chronic condition lasting for many years. This leads to infertility and fetal death.
• Brucella Contaminated milk. A likely cause of Brucella infection in offspring is the drinking of infected collostrum or milk from the dam. Schroeder (1913) was able to isolate B abortus from the milk of 83.5% of serologically positive cows. Hardy (1930) notes that Brucella is present in cows milk but in limited concentration compared to goats milk. He considered that since contaminateds milk was usually bulked with unaffected milk the dilution factor was sufficient to signifiantly reduce the risk to consumers.
• Insect and Other Parasites. There is no significant evidence that brucellosis can be spread by insects and other parasites although there is evidence that such parasites can carry brucellosis, Ozsan (1982) and Gudoshnik (1958) . Britov (1979) showed that round worms (Trichinella spiralis) could transmit 'Brucella to guinea pigs. Aminals with brucellosis are frequently no bacteremic.

Bang (1897) considered that bull to cow was a primary means of transmission of B abortus through sexual contact but later work indicated that it was not a significant route to infection, Hadley (1916) , Thomse4n (1943) , Rankin (1965) and Mccaughley (1973) . However Metcalf (1986) that an infected bull in a brucellosis clean herd could occasionally transmit brucellosis to cows or heifers and therefore re-infect the herd. Thomsen (1937) noted that the average incubation period for such transmission was 225 days. Cows are also subject to artificial insemination and in this case infection is far more likely apparently because the semen is deposited in the uterus, The uterus is far more susceptible to infected semen, Manthei (1950) and Lamvert (1963) . Therefore it is essential that bulls used for artificial insemination are free from brucellosis, .

Inter-Species Transmission.

Cattle are a natural host for B abortus, sheep and goats for B melitensis and pigs for B suis. These 3 types for brucella can spread amungst othe domestic animal species. Different animal species are often kept in close proximity such as on farms or at markets. This offers the possibility of the disease spreading from one specis to another. Modes of transmission include: direct contact, exposure to a contaminated environment including pasture and eating contaminated food such as dogs eating meat. However in general it is considered that such secondary infections are self limiting, Nicoletti (1980) . This means that such animals do not become a self sustaining host for the disease. The individual animals may however become chronically infected and carriers able to spread the disease.

B abortus.

• Fitch (1930) demonstrated that Brucella can be transfered to horses and (1947) showed that 2.9% of farm horses in Minnesota were reactors to Brucella agglutination titres.
• Huddleson (1929) demonstrated that B abortus could spread to chickens. Anczykowski (1973) concluded that Brucella## did not persist in domestic fowl and it could not be found in droppings of of serologically positive hens on infected farms. WHO (2001) observes that Brucella has been isolates in a few cases of domestic fowl naturally infected. Possible symptoms are: weight loss, reduction in egg production, and diarrhea. Brucella has been isolated fron some wild birds such as raven (Corvus corvix) and crows (Tripanscorax fragilecus).
• There has been close associations between dogs and infected cattle a few cases of B abortus in dogs has been recorded. It is likely that many other cases have gone unreported. It is possible that dogs can be a vector for the disease. Van der Hoeden (1933) studied 23 dogs infected orrally, conjunctivally or intracutaneously with B abortus. The dogs showed no signs of illness although all had possitive SAT results and 1 dog had Brucella infected urine. B a bortus was recovered at autopsy. Morse (1951) and (1953) fed aborted fetuses and placenta to 15 dogs which were later examined post-mortem. B abortus was recovered from the lymph nodes of 9 dogs and i by guinea pig innoculation. Sat response was low or negative. 2 bitches aborted but otherwise there were There were only rare clinical signs illness. Other cases of B abortus in dog has been reported by Philipon (1969) , Schwarz (1954) and Taylor (1975) . Dogs with physical signs of Brucella infection have been describes. Clegg (1988) records a dog with polyarthritis. McErlean (1966) described undulant fever and arthritis in a dog diagnosed as brucellosis by serology tests. Hall (1974) describes a dog with fever, muecle stiffness and arthritis for which Brucella was isolated from blood and urine. Prior (1976) isolated B abortus from the spleen of 2 symptomless farm dogs associated with an infected herd and considered they were possible carriers. Bicknell (1979) reported a case of B abortus in a bitch. Likely cause of the illness was association with diseased cattle and in particular access to Brucella infected aborted fetuses. Post mortem examination allowed isolation of B abortus biotype 2 from the brain, uterus, vagina, and a range of lymph nodes. B abortus was also isolated from urine at 2.45 x 10² organisms per ml, a possible mode of transmissioin to animals or humans. Serology test by SAT, RBPT, CFT and Antiglobulin were all positiver. No signs of brucellosis were observed. Other studies have also shown that outbreaks of brucellosis in cattle can be caused by infected dogs. Forbes (1990) studied dogs on B abortus infected cattle herds. All 14 dogs involved became infected. 1 dogs was shown to shed Brucella. Bruceella infecteed dogs are considered a danger to there owners and in particular children due to close physical contact.
• Luchsinger (1967) reported the spread of B abortus to sheep. It is generally considered that sheep are resistant to B abortus, Allsup (1974) and but nevertheless they can become infected. Once infected they can become long term carriers and can shed B abortus, Luchsinger (1979) and Okoh (1980) , At the same time they may have negative serology results to Brucella.
• Meador (1989) reports B abortus infection in goats. Mathur (1967) notes that there is a low pervalence of brucellosis in goats exposed to B abortus.
• Okoh (1979) describes B abortus infection in cammels. Al-Khalaf (1989) in a study of brucellosis in Kurait found the 14.8% were positive by CFT and RBPT serology tests and B abortus biovar 1 was isolted.

B melitensis.

• Simpson (1940) B melitensis showed that B melitensis could spread from goats to cows and be found in their milk was infected. Zowghi (1985) .describes B melitensis infections of cattle in Iran.
• B melitensis easily spreads between goats and sheep and both are hosts.
• Pinigin (1970) describes cases of B melitensis in dogs.

B suis.

• Simpson (1940) found that B suis from pigs could be transfered to cows and be found in their milk. Jordon (1947) reports 3 cases were B suis spread from hogs to cattle sharing the same lot. The milk from the cattle was infected. Milk infected with B suis was more infectious to humans than B abortus. B suis infection in cattele is also described by Cook (1984) and Norton (1979) .
• Barr (1986) describes B suis infections in dogs.
• Cook (1988) describes a B suis infection in a horse..

Human immunity.

Otero (1929) innoculated human volunteers with B abortus and found that this gave rise more easily to infection than by similar oral tests. Huddleson (1942) reviewed acquired immunity for brucellosis in humans and animals. They noted that some people exposed to the infection develop clinical or subclinical illness and then recover. Some people had no apparent illness but laboratory tests (SAT, skin test and osponophagocytic test) showed that they had been subject to infection. They considered this an acquired immunity. Young animals seem to have an increased immunity. Immunity to one Brucella species extends to the others. Huddleson (1940) studied 349 students at a college during a brucellosis outbreak. The students risk of infection was variable. 41 became clinically ill and 49 showed evidence of infection by SAT, skin tests and opsonophagocytic tests but had no symptoms. They concluded that the latter had developed immunity. As previously noted in cattle Spink (1953) observed that while children under the age of 12 are often exposed to Brucella they are less likely to show overt signs of illness. He also studied family groups affected by brucellosis and noted that not all family members exposed to Brucella became ill although they had positive SAT of skin tests.

Brucella ovis seems to be almost exclusive to sheep and evidence of human infection is limited, Meyer (1982)

Animal to Human Brucella Transmission.

• According to Jordon (1931) the level of positive SAT results indicative of active or latent brucellosis (mainly B suis was 18.2% at a time when McNutt (1936) had determined the incidence of B suis in hogs at 3%. Although many surveys of the illness suggest that either men or women are more affected he notes that in those under nine or over 70 not directly involved with animals the incidence illness was the same in both sexes. Both B abortus and B suis in farming communities and there is a peek in cases during farrowing/ calving season.
• Jordon (1947) surveyed the incidence of human brucellosis infection in Iowa. The infection is most common in those in contact with infected animals: meat processing workers, vetinarians and farm workers. Excluding meat processors, the incidence (1939-43) was: farms, 0.014%; small towns, 0.009%; large towns, 0.005% (mostly passeurised milk). For meat processing workers the incidence was 0.25%.

• Milk and Milk Products.' For members of the general public the most likely route to infection is the ingestion of Brucella contaminated food. The most likely sources of infection are: milk or milk products such as soft cheese and yoghart and raw meat (unusual). This problem has been virtually ellininaled in most developed countries through. a combination of pasteurisation of milk and the elimiation of infected animals, but is still a problem in many countries. An example is provided by Williams (1971) ( who notes the complacency at that time relating to the consumption of raw milk in the UK and the lack of knowledge concerning the significant risk of brucellosis. Brucella is killed by heating including and sterilisation, pasteurisation and cooking.
• Brucella Contaminated Environments.. The environment associated with Brucella infected animals such as pasture, farm and stock buildings and markets can become and remain contaminated with Brucella' for considerable periods of time.. This is a risk to other animals and humans. THe disease can be transfer to humans or other animals though direct physical contact, dust, and aerosols. This risk can affect farm workers and the general public.
• Brucella Infected Animals. Farmers, farm workers, their familues, vetinarians who are in direct contact with Brucella infected animals are at a high risk of contracting brucellosis. The most likely route to infection is through skin abrations and/or physical transfer to muccosal membranes such as mouth, nose or eyes. Thomsen (1931) found that 94% of all new veterinarians working with brucellosis infected cattle herds developed positive SAT tites indicating that they had been infected with Brucella and either had developed a subacute infection or a form of immunity.
• Animal midwifery. Farmers, veterinaries and others involved in midwifery of infected animals are at high risk. Fetuses, placenta msaterial and viginal secretions can be highly contaminated with Brucella. The same risk may apply even if the dam and offspring are apparently healthy. Likely modes of infection are handling calves or aborted fetuses from infected dams, gynaeological examinatio, obstrectric manipulation and rectal examination.
• Meat Processing and Meat Products. Meat contaminated with Brucella has been recognised a significant risk factor to those handling it. Those most likely to be affected are abattoir (packing house) workers. Others at much lesser risk are butchers, restaurant workers and those handling meat in the home. Specific risk factors to facilitate infection are infection via skin cuts and abrasions (Cuts are common in this industry) and breathing in aerisols. Saddler (1960) concluded that for the USA in 1955 a substantial part of retail meat was potentially contaminated with Brucella. The level of infection was at least 1.25% in cattle and 1.15-3.5% in pigs. He notes that normal cuts of meat contain lymph nodes know to be highly contaminated by Brucella and this is particularly the cases with mincd products such as burgers. McNutt (1935) estimated that infected lymph nodes contain approximately 300 organism per milligram of tissue but that Brucella is not distributed evenly throughout the lymph node. Macroscopic leisons are rare in bruvcellosis so there is no obvious way of detecting Brucella infection. Hutchings (1951) demonstrated that B melitensis could survive in hams cured in brime at 32-36°F for 21 days but not if the meat was smoked. Huddleson (1933) found that the viability of Brucella in infected lymph nodes was not affected by being held at -10°F for ≥10 days. Heathman ((1934) studied 1096 abbatoir workers for evidence of Brucella infection. 6.9% of workers had positive SAT titre and 55.6% had positive intradermal test results. Hendricks (1962) reviewed an outbreak of brucellosis in a Iowa packing house where pigs were slaughtered. The greatest concentrated of cases occurred on the killing floor where contact with fresh warm tissue including during the dressing of carcasses and the processing of the offal was intimate. It was noted that aerosols were present on the killing floor and that an airborne mode of infection was a likely mode of infection increasing cases in this area. Harris (1962) was able to demonstrate that B suis was present in small droplet in the air on the killing floor. This is supported by the work of Rosebury (1947) who has demonstrated that B suis is relatively stable in experimentally created aerisols.
• Deseased Animal Carcasses. Knackers and renderers and hauliers who handles Brucella infected animal carcasses and meat parts unfit for human consumption are also at considerable risk of infection.
• Hunters who dress the carcasses of wild animals infected with Brucella can also be at risk athough the incidence is apparently low. Such cases are described by Robson (1993) .
• 'Laboratories Workers. Laboratory workers handling Brucella contaminated materials such as tissue samples and cultures can be at high risk of contracting brucellosis. Special precautions are essential.
• Live attenuated Brucella vaccine. Live virulant Brucella vaccines are used to control brucellosis in domestic animals. B abortus strain 19 has been widely used in brucellosis eradication programs in cattle. According to Tizardd (1990) strain 19 is an immunogenic organism that can induce lifelong immunity and prevent abortion in cattle but has significant virulance and can cause systemic reactions in vaccinated animals including local swelling, high fever, anorexia, listlessness and decresed milk yield. It can cause abortion in pregnant coows and orchitis in bulls. In humans it can induce both acute and chronic desease. Vetinerains who administer the vaccine are particularly at risk often through accidental injection. B abortus Strain 19 may be a risk factor for humans in raw milk and could affect meat.
• Brucella canis. Brucella canis affects mainly dogs but can be transmitted to human because of the close association of dogs with humans. In dogs the disease is both acute and chronic. People particularly children handling pups are at particular risk. The initial illness in humans is often mild and it is suggested it often goes undetected. The usual serology test for Brucella antibodies do not work. B canis is endemic in many parts of the world. Dogs can also catch and transmit B abortus, B suis and B melitensis.

Human to Human Transmission.

There are a number of ways that human to human transmission of brucellosis can occur: blood transfusion, bone marrow transfusion, transplacental/prenatal exposure, sexual contact, from mother to neonatal via beast milk and environmental contamination. Human to human transmission is considerd rare, but is also often difficult to prove. In families where there are multiple cases of brucellosis they are usually exposed to the same risk factors so person to person transmission is difficult to prove. Possible risks to family or brucellosis patients are from blood, fecal matter and body fluids such including urine, saliva and semen. A few cases of human to human transmission have been proven. In general the normal hygene practices used by humans seem likely to guard against this infection.

• Mohler (1913) was able to culture B abortus from tonsils. Vanne (1925) recovereed B melitesis from cattle sputum.
• Amoss (1929) recovered Brucella from the fecal material of 2 patients, one after 16 months of illness.

Prenatal or Placental Transmission. In domestic animals birth problems such as abortion, premature birth, still births and sickly offspring are typical sign of brucellosis. This is also probable true in wild animals but is not often observed. These problems also affect women but the extentto which this occurs is disputed. Treatment of the infected mother with antibiotics can reduce the effects and help pregnacy success rates. A number of cases of infants being congenitally infected with Brucella have been recorded:

• Schreyer (1980) describes the case of a second-trimester pregnancy woman with B melitensis septicemia resulting in intrauterine fetal death and Gram-negative septic shock with diffuse intravascular coagulation.
• Sharif (1990) describes the results of a screening programme in Saudi Arabia. 18 of 513 pregnant women had positive brucellosis titre as well as 24 who had symptoms suggestive of brucellosis. Thirty of the 42 positive cases had titres exceeding 1:160. The incidence of abortion among pregnant women with Brucella titres less than 1:160 was 7.7% contrasting with 17.6% among those with titres above 1:160.
• Chheda (1997) describes a case of congenital brucellosis in a premature infant.
• Malone (1997) reports a case of maternal brucellosis resulting in preterm labor, chorioamnionitis, placental abruption, and delivery of a live-born infant at 25 weeks'. Maternal blood and amniotic fluid cultures were positive for B abortus. Aggressive antibiotic and tocolytic therapy was used to pretect the pregnancy. It was concluded that in practice and contrary to the conventional wisdom that B abortus does not cause poor obstetric outcomes since human pregnancy is resistant to such infection it can lead to significant perinatal morbidity.
• Makhseed (1998) studied 346 pregnant women using ELISA and microagglutination and Brucella culture tests. This included 227 women with preterm delivery, 51 with intra-uterine fetal death, 29 with spontaneous abortion, and 39 with term delivery of a normal baby. Ingestion of raw milk was positive in 18% of the control cases, 28% of the preterm delivery cases, 30% of intrauterine fetal death cases, and 21% of spontaneous abortion cases. Acute or chronic Brucella infection was found in 8% of preterm, 10% of intra-uterine fetal death, and 7% of abortion cases. None of the control cases had evidence of Brucella infection. None of the four groups showed positive culture of the placenta for Brucella. Brucella serology titres were significantly higher in the preterm, abortion, and intrauterine fetal death groups than in the control group.
• Giannacopoulos (2002) also described a case of congenital brucellosis.
• Koklu (2006) describes a cases of a premature infant with transplacentally acquired congenital brucellosis with pulmonary involvement.
• Ozbay (2006) describes the case of a woman with active brucellosis infection with a twin pregnancy at 10 weeks who commpleted a successful pregnancy after treatment with rifampicin (900 mg/day) indicating that early and adequate treatment can prevented the early detrimental consequences of the disease.
• Mesner (2007) describes an interesting case which mirrors widespread accounts of infection of vetinarians during animal midwifery. The infection was passed from a pregnant woman to a preterm infant and thence to 3 doctors. The mother presented with slightly elevates liver enzymes but did not have hepititis A or B. Multiple chances to test for brucellosis were missed. She developed vaginal bleeding and a cesarean delivery was performed involving serious complications. At this stage cultures and serum antibodies for Brucella were positive and the patient received antibiotics and recovered. Earlier blood samples were retested and were also found to be brucellosis positive by SAT titre. The infant was born prematurely at 24 weeks and was subject to intensive care. Blood tests by culture indicated Brucella and the patient was given antibiotics but died at 20 weeks. The doctors who became infected were a obstetrician who took part in the emergency Cesarean delivery, a pediatritian who was involved in resusitation of the infant and a neonatologist. All 3 had symptoms of brucellosis, positive Brucella cultures and elevated Brucella antibody titres. They were treated with antibiotics and had at least partial recovery. None of the 3 was involved in any acident or contamination with blood or body fluids and all wore protective gloves but the pediatitian and the neonatlogist did not ware face masks or googles. 56 other medical staff who had contact with the mother and child were apparently unaffected.
• Kurdoglu (2010) review their experience of risk factors with brucellosis in pregnancy with acute, subacute, chronic, or relapsing brucellosis. The main risk factor leading to illness was consumption of unpasteurized dairy products in 92.3% of the cases. Spontaneous abortion, intra-uterine fetal death, and preterm delivery rates were 24.14, 3.45, and 6.9%, respectively of which only spontaneous abortion rate substantially exceeded that among the general population of pregnant women. It was concluded that brucellosis in pregnancy is associated with increased incidence of spontaneous abortion without an association with the magnitude of serum agglutination titer, the clinical type of brucellosis and hospitalization.

Breast Milk transmission. Cows infected with brucellosis abortus secret B abortus in there milk. This apples to cows with new or latent infections. The same applies to B melitensis in goats and sheep. In these animals brucella frequently becomes locallised in the mammory glands and the related lymph nodes. The milk is an important way of transmission to humans. There is also considerable evidence that women infected with brucellosis also secret Brucella in their milk: Some typical reports are discuued here. In the cases discussed where infants are infected from breast milk there is also the possibility that in some cases the infection could have been prenatal.

• Vierucci (1963) Lubani (1988) discusses a case where a neonatal infant develops brucellosis. It is suggested that the disease is transmitted from the mothers milk since the baby was fed breast milk only when the mother became acutely ill with fever, arthralgia and malaise. Both cases were proved to be B melitensis by blood cultures. The brucella agglutination titre in the breast milk was 1:2560.
• al-Eissa (1990) describes the case of a woman who developed brucellosis 4 weeks after the birth of a child. The child subsequently developed brucellosis melitensis proved by culture test and it was considered this was probably from ingestion of beast milk. The mother's beast milk Brucella serology titre was 1:640.
• Varon (1990) Barroso Espadero (1998) reported 2 cases of brucellosis in infants ascribed to transmission from mothers milk.
• Gasser (1991) and (1893) describe 2 cases of mammary absesses associated with chronic brucellosis melitensis. B melitensis was isolated from a suspected breast tumour of a woman who also showed signs of uveitis. Three weeks after surgical drainage and despite antimicrobial therapy a new abscess developed on the other breast from which a culture also yielded B melitensids. This is a cases of localization of brucellosis without previous penetrating injury may be the only manifestation of chronic infection.
• al-Mafada (1993) describes the isolation of B melitensis from human milk.
• Casalinuovo (1995) describes the isolation of B melitensis biovar 2 from human milk.
• Cokça (1999) describes the case of a pregnant woman with bilateral mammary abscesses due to Brucella melitensis.
• Palanduz (2000) describes the case of a mother and child with brucellosis in which it was consudered that disease was transmitted to the 3-month-old baby via breast milk.
• Tsironi (2003) describes a human mammary absess caused by B melitensis.
• Celebi (2007) describes the case of a family in which one member suffered acute brucellosis melitensis. A woman in the household who was asymptomatic women was screened and B melitensis biovar 3 was isolated from her blood and beast milk.
• Tikare (2008) present a case of meningitis due to B melitensis biotype 1 in a 11-month-old infant considered to be transmitted though breast milk. B. melitensis was isolated from the mother's breast milk.
• Jensenius (2008) describes a cases of mastitis in a woman caused by Brucella.

As previously noted in cattle Spink (1953) observed that while children under the age of 12 are often exposed to Brucella they are less likely to show overt signs of illness. He also studied family groups affected by brucellosis and noted that not all family members exposed to Brucella became ill although they had positive SAT of skin tests.

Sexual transmission. In animals sexual transmission is an important vector for the spread of brucellosis. This is particularly the case for B canis and B ovis. In other animals such as cattle with B abortus this is less inportant. In animals Brucella can be cultured from both seminal fluid and vaginal secretions. In humans Brucella has been cutlured from seminal fluid. However the extent to which bruucellosis can be spread sexually in humans is unknown but it is generally considered a rare mode of transmission. A number of cases have been recorded:

• Ruben (1991) records the case of a laboratory worker who became infected with B melitensis biotype 3 during his work as a microbiologist. His wife's blood also subsequently became infected. The isolate obtained from husband and wife were indistinguishable.In consideration of the risk factors it was comcluded that sexual intercourse was a possible means of transmission.
• Wyatt (1996) considered that sex transmission may occur in cases of brucellosis melitensis.
• Mantur (1996) suggests there is evidence for B melitensis could be spred sexulally..
• Kato (2007) describes a case in which a Japanese man became infected with brucellosis melitensis during his travels. Subsequent to his return his wife also contracted the illness. It is suggested that the husband transmitted the disease to his wife sexually,
• Mantur (2006) carried out a study of 517 brucellosis patients in India. 11 patients suffered from epididymo-orchitis. Of these patients with epididymo-orchitis, Brucella specific antibodies were demonstrated in the testicular fluid of 8 patients and in semen of 5 patients. B melitensis was cultured from the testicul fluid of 7 of 9 patients but semen was found to be sterile in 8 patients. He found that sexual transmission was implacated as a possible means of infection in 6 cases.
• Sözen (2007) notes that epididymoorchitis is the most frequent genitourinary complication of brucellosis and is often unilateral. They describe a male patient diagnosed with epididymoorchitis for whom B melitensis was isolated from blood, bone marrow and ejaculate cultures.

Blood Transfusion. In human brucellosis bacteremia occurs during acute attacks, relapse and probably intermittently during chronic illness. Therefore there is a possibility that blood taken from such a person could cause an infection in the recipient. Introducing Brucella directly into the blood stream is probably the most effective means of infection. Brucelllosis is a disease in which full cure is always in doubt therefore it is better that anyone who has had brucellosis should not donate blood. This is the rule in some countries. In countires where brucellosis is widespread in domestic animals and particularly if raw milk is consumed there is a particular risk.

• Brito-Mena (1948) indicates that brucellosis can be spread accidentally through blood transfusion.
• Magoffin (1949) notes that in an epidemiologic survey of 1627 healthy blood doners at the Minnesota University Hospital he found that 1 of 4 sera were positive at 1:40 for agglutinates.
• Spink (1956) carried out a survey of donated blood in minnesota and found that 20% of donated blood contained Brucella agglutinates.

Medical workers. The risk to medical staff caring for brucellosis patients appears relatively small. Bodily fluids and fecal mmatterial may be risk factors. Risks associated with brucellosis infected mothers may be higher. The following are some interesting example:

• Singer (1991) describes a case of possible transmission during delivery.
• Poulou (2005) describes the case of an obstetrician who was infected during the delivery of an infant suffering from congenital brucellosis. The obstetrician was treated with doxycycline and rifampin and fully recovered. The cause is considered to be transmitted through infectious secretions at the delivery of a transplacentally infected newborn possibly due ingestion of secretions during clearance of the newborn's respiratory tract from saliva and amniotic fluid.
• Mesner (2007) describes the case of a baby born to a mother later found to be ill with brucellosis. The baby was delivered by casarean section. Three doctors became infected: a obstetrician who took part in the emergency Cesarean delivery, a pediatritian who was involved in resusitation of the infant and a neonatologist. All 3 had symptoms of brucellosis, positive Brucella cultures and elevated Brucella antibody titres. None of the 3 was involved in any accident or contamination with blood or body fluids and all wore protective gloves but the pediatitian and the neonatlogist did not ware face masks or googles. 56 other medical staff who had contact with the mother and child were apparently unaffected. See above (Prenatal Transmission).

Testing For Brucellosis in Animals.

Brucellosis is a zoonotic disease which spreads from animals to both other animal species including humans. The main ways that it spreads is by direct or indirect contact with infected animals and by the consumption of contaminated food. In humans most cases of brucellosis are related to such contact with domestic animals such as: cattle, goats, sheep, pigs and dogs. Also, when brucellosis is present in domestic animals it causes ecconomic hardship to farmers due to loss of animals and reduced ecconomic output. For these reasons the greatest efforts has been made into reducing and eliminating the disease in domestic animals thereby reducing the incidence of illness in humans. The greatest effort has been expended in diagnosing and elimination of the disease in cattle. An essential element of this strategy has been to attempt to develop effective and usable tests for brucellosis. The test developed in this way have also been appled to testing for brucellosis in other animals such as small ruminants and humans.on this work.

Testing for Brucellosis in Cattle.

THe most important problem faced in the elimination of brucellosis from cattle is the inadequate performance of tests for the disease. Both intradermal skin tests and serum and milk antibody tests are used. A wide range of serum antibody tests have been developed but none have proved to work in all cases. The response of individual animals differ including that related to the stage of the illness therefore a combinations of tests. Tests can also be carried out on uterine discharge, vaginal mucous, milk or semen. It is highly likely that in some animals antibodies disappear over a period of time from serum rendering all tests ineffective. See more detailed information about tests under 'Serology Antibody Tests'.

Studies of the relationships between B abortus and the bovine antibody response have been made by Allan (1876) Corbell (1972) , Levieux (1974) and Beh (1973) and (1974) . For B abortus infection in cattle IgM antibodies appear within 5 to 15 days in the majority of cases but the amplitude of response depends on many factors. IgG1 appears soon after follwed by IgG2 and IgA. Therefore IgM is used to assess the early stages of infection however IgG1 is the most useful to test herds of cattle, Lamb (1979) and Butler (1986) .

The sensitivity of a test is its ability to correctly identify a diseased animal and is defined as the proportion of animals with the disease which test positive. Specificity is the ability of the test to correctly identify non diseased animals and is defined as is defined as the proportion of animals without the disease with negative test results. A cut off has usually to applied which takes account of both selectivity and specificity. Inevitable a compromise is required. This means the results are subject to interpretation and both false positive and negative results can occur.

The main test used were:

• The Milk Ring Test (MRT).
• The Serum Agglutination Test (SAT).
• The Standeard Plate Test (PAT).
• The Complement Fixation Test (CFT).
• The 2-Mercaptoethanol Test (2-ME).
• The Buffered Antigen Test (BPAT).
• The Rose Bengal Plate Test (RBPT).
• The Card Test (CARD).
• THe Rivanol Test (Riv).
• The Indirect Enzyme Linked Immunosorbtion Assay (iELSIA).
• The Competitive Enzyme Linked Immunosorbtion Assay (cELSIA).
• Delayed type hypersensitivity Skin Test. (DTH).
• Culture test.
• Fluorescence polarization immunoassay (FPA).

The Milk Ring Test (MRT).

MRT was first develop by Fleischhauer (1938) and then further developed by Bruhn (1944) , (1948) and Christiansen (1948) . It is essentially a rapid agglutination test carried out on whole milk or cream. It is considered that agglutinins present in milk are adsorbed onto fat globules. A stained Brucella antigen is added and the mixture allowed to stand for 1 hour at 37°C if a fat/agglutinins/stained antigen forms a ring in the fat layer this indicates brucellosis. Dilutions can be made to determine the extent of infection. This is a quick and easy test suitable for widespread use but it is only suitable for lactating cows. Kerkhofs (1990) the immunoglobulins detected by MRT are IgM and IgA. The main problem is lack of specificity. False positives are associated with colostrum, end of lactation milk, and mastitis and hormone disorders, Bercovich (1979) and Kerkhofs (1990) . False negatives may also occur. Either because the immunoglobulins present fail to combine with the milk fat or because they not present in sufficient concentration. This can occur in pooled mlk of in individual cows because the level of immunoglobulins in milk rapidly declines after abortion or parturition. In such cases MRT may fail to work. Increasing the volume of milk in the test helps 8mls (MRT-8) rather than 1ml (MRT-1), Bercovich (1978) . The test gives rise to none specific reactions, Bercovich (1978) and .

The Serum Agglutination Test (SAT).

THe SAT titre used in cattle is essentially that developed for use in humans disease by Wright (1898) . The test was first used in Denmark for the diagnosis of cattle by Grinstead (1909) . In early work this was the main test for brucellosis in cattle and it is still widely used. When first used there were problems with the test due to lack of set standards, Huddleson (1920) . In the USA standard procedures were determined by Malcolm (1931) which are still used. A positive titre of 1:100 is considered indicative of a reactor cow and titres from 1:50 to 1:100 as suspicious. Hajdu (1974) has found that the test measures IgM, IgG1 and IgG2 and IgA but principally IgM. Chappel (1976) found that SAT was 10 times more sensitive to IgM than IgG2 or IgG2. Its is known that IgM appears first in Brucella infection and then usually falls to low or negative levels within 1 year. It is best used for acute infections, Beh (1973) and (1974) and Levieux (1974) . The test has a nunber of defects including false positive and false negative as discussed elsewhere. For this reason the test is only suitable for herd testing, Alton (1975) rather than individual animals. The main reason for false positive results is the presence of agglutinins to unrelated bacterial antigens from other Gram negative organisms, Hess (1953) , (1953) , Rose (157) Corbel (1984) . If calves are vaccinated before 8 months THe other most likely cause of false positives is the presence of IgM related to previous use of s19 vaccine. This effected is limited in calves are vaccinated before 8 months but significant when older cattle are vaccinated.

The Standard Plate Test (SPAT).

One of the problems with the SAT titre is that it takes at least 48 hours to complete. There was a need for a more rapid test particularly in fields testing. Such a rapid test was developed by Huddleson (1926) using a concentrated antigen. The test was standardised to give simmilar results to the SAT titre. Later a dye stained antigen used for ease of reading. It can be used in feild conditions, Donham (1934) . This test does not use a series of dilutions like the SAT titre but is standardised to give a result equivalent to a SAT titre of 1:100 (positive). The test has been widely used. Stemshorn (1985) considered that could give positive results when SAT was negative by virtue of its use of high saline (8%) and higher serum concentrations. It was also resistant to the prozone effect.

The Acid Plate Antigen Test.

This test works on the principle that the pH of the serum is altered to the isoelectric point of IgM to prevent its agglutination, Rose (1964) and . The test was developed by Rose (1957) . The test has been used widely because it eliminated some non-specific reactions but has the disadvantage that fresh antigen is needed daily. It was considered that it could replace SPT and be used for market screening. The test has been improved by use of a stable buffered antigen. It is the basis of a number of important screening tests which are widely used. These tests are: The CARD Test, the Rose-Bengal test, and the Buffered Acid Brucella Antigen test.

The Buffered Antigen Test (BPAT).

Angus (1984) developed a Brucella plate antigen test buffered to pH 3.63 of testing sera of cattle and pigs. He condsidered that the main advantage was the reduction of nonspecific test reactions. Other information suggests that it tests for IgG1. A 11% suspension of B abortus s1118-3 stained with crystal violet and brilliant gree buffered to 3.63. The antigen is mixed with test sera on a glass plate and incubated at 37°C for 8 minutes.The presence of agglutinates is indication of a positive result.

The Rose Bengal Plate Test (RBPT).

RBPT is also a acid buffered test. It is a development of the acid plate test developed by Rose (1957) by the US Department of Agriculture (1965) and tested by Nicoletti (1967) . At neutral pHs this test can measure the presence of IgM, IgG1 and IgG2 however IgM appears the more active. According to Allan (1975) IgM is 5 times as active as IgG1 and IgG2. At the buffered pH of 3.65 RBPT apparently meassures only IgG1. Morgan (1974) found that the Rose Bengal Plate Test (RBPT) gave similar results to the CF test. However Nicoletti (1967) considered that while the test gave few false negative tests it gave many false positives possibly due in significant part to reaction with IgM in cattle previous s19 vaccination. False positives can be tolerated in a screening test provided false negatives are low therefore this test has been adopted in the UK and Australia as a screening test for bovine brucellosis. Automatic procedures have been developed, Gower (1974) . Allan (1975) and Miller (1873) notes that RBPT is less sensitive than CFT and misses some Brucella positive cattle which CFT finds.

The Card Test (CARD).

In the USA the CARD test is a standard test which has been used in field studies and market testing. It is n=based on the Acid Plate Antigen Test. The antigen is described by Metcalfe (1966) and the card (3 x 5 in.) is suppled by Hynson, Westcott and Dunning of Baltimore. The test has been adopted for use

CARD is a version of the RBPT test developed for use in Canada, Stemshorn (1985) and used in the USA. It works similarly to BPAT and RBPT.

The Rivanol Test (Riv).

Rivinal is the commercial name for the dye, 2-ethoxy-6,9-diaminoacridine lactate. This test is an adaption of the SPT titre. Acrydine dyes such as rivanol can be used to precipitate glycoproteins from serum solutions, Nicoletti (1969) and Huber (1986) . In this case rivanol precipitaes mainly IgM and leaves mostly IgG in the serum. The precipitate is removed by centrifugation. The supernatent is then tested with an antigen by the normal SPT procedure. This is a fairly labour intensive test but is considered to have good specifity for brucellosis. It is used in a confirmatory role.

The Complement Fixation Test (CFT).

The general principle of the CFT involves deactivation of the inherent complement in a sera sample. Appropriate dilutions of the serum are then prepared and carefully measured quantities of the antigen, guinea pig complement and sheep red blood cells added. If antibodies to Brucella are present they bind to the antigen and the complement is fixed, thus preventing lysis of the red blood cells.

The CF test as applied to brucellosis in cattle was fisrt developed by Holth (1909) in Denmark. Trials were carried out in the USA by Larson (1912) and Surface (1912) to look for evidence of Brucella infection in cattle. It was applied in the testing of cattle by Michel (1931) . The test was subsequently widely used in Europe and was considered one of the most effective tests for brucellosis, Jones (1963) . In the USA it was considered that the CF test was complicated and that SAT gave equally reliable results, Firch (1926) . However the test was re-introduced from 1960 onwards because of its superior performance compared to other testsusefulness in diagnosing reactor cattle in problem herds of cattle, Nicoletti (1986) and and Jones (1963) . In general CF test has probably been the most widely used confirmatory test for brucellosis. It can be automated and is usually used as a confirmatory test. According to Levieux (1974) and Sutherland (1984) and (1985) CFT measures IgM and IgG1 which fix complement. Morgan (1969) stated that the detects both IgM and IgG but the latter more effectively. Chappel (1977) and MacMillan (1990) considers that although IgM is more effective in the CF test than IgG1 in practice the test measures more IgG1 probably because IgM is significantly destroyed by the test procedure. Curtain (1971) and Cho (1972) showed that that the complement fixation ability of bovine IgG related entirely to IgG1. IgG2 and IgA do not fix comlement. McNaught (1877) and Plackett (1975) have shown that the test may be adversly affected by IgG2 interference (prozone effect) and by anti-complement activity. CFT finds more positive reactors than most other tests, however Bercovich (1998) notes that results for cattle with chronic brucellosis may be negative or inconclusive. Cordes (1979) and Sutherland (1982) found that CFT gives negative test results for som infected cattle.

The methodology of the test has been standardised by Hill (1963) and MacKinnon (1963) is as follows: The tests for the interaction of complement with utilizes sheep red blood cells (SRBC) along with specific antigen. Serum is heated to destroy its own complement and then incubated with B abortus whole cell antigen and a quantified amount of guinea pig complement and allowed to stand. If Brucella antibodies (IgG1) are present in the serum then these complex with the antigen and the complement is activated (used up). Then a quantified amount of sheep erythrocytes coated with rabbit antibody is added. If Brucella antibodies were not present then the unused complement lyses the sheep erythrocytes to release haemoglobin. If Brucella antibodies were present and the complement was used up then the sheep erythrocytes remain unlysed. Therefore the amount of haemoglobulin release is directly proportional to the concentration of Brucella antibodies. The serum is tested at a range of dilutions and the result is taken as the highest dilution with 50% lysed blood.

The CF test can be performed in a number of ways: warm, cold, he indirect haemolysis test and haemolysis in gel, Plackett (1976) and (1980) , Chappel (1982) and (1982) and Sutherland (1982) indirect haemolysis.

Nicoletti (1978) and (1978) notes that although CFT is relatively specific compared to other tests it gives some false positive values after adult vaccination with B abortus s19.

Although the CF test has been widely used it is subject to serious technical problems:

• The test requires great technical proficiency. It involves the time consuming and difficult preparation of reagents. This also makes it expensive.
• Labile reagents. The use of labile reagents may lead to erratic results particularly if the tests are carried out on a small scale. Lucero (1990) noted that a severe disadvantage of CFT was the use of the use of highly labile reagents (such as complement.
• CFT is affected by anti-compliment activity. If this is present the test cannot be read at lower dilutions. Nielsen (1996) working with cattle considered that sensitivity and specifity varied considerablely due to anti-complement activity. Lucero (1990) notes that CFT mostly identifies IgG but has severe disadvantages such as the the occurrence of anti-complement activity and general technical problems.
• THe prozone effect: Al Dahouk (2003) notes that CFT is affected by the prozone effect and anti-complement activity. This gives rise to false negative results.
• Insensitivity. CFT fixes IgG and moderate amounts of IgM but not IgA. Heremans (1963) showed that CFT measured IgG and limited IgM but no IgA.Dahouk (2003) considered that CFT should be used only as a confirmatory test and she noted that sensitivity and specificity could vary widely.
• Haemonised sample cannot be tested.

====The Anti-globulin Test (AGT). The anti-globulin test can be used with a screening test such as SAT is a possible confirmatory test for brucellosis. It measures IgG2 and therefore offrs an alternative to CFT where the results of the latter is equivical. The Test method is long and labor intensive. It is discussed elsewhere. Studies of its use in bovine brucellosis are described by Kiss (1970) , Beh (1973) and Anczykowski (1986) .

The 2-Mercaptoethanol Test (2-ME).

This an adaption of the SAT titre. There are 2 forms of this test which use either 2-mercaptoethanol, or dithiothreitol, Klein (1981) . Dithiothreitol is preferable because of the toxicity of 2-mercaptoethanol. The effect of the reagent is to reduce disulphide bonds. It measures mainly IgG1 because IgM is reduced and therefore unable to agglutinate. However IgG can also be reduced giving false negative results. but is less sensitive than CFT. It has been tried as a possible means of differentiationg between infected and vaccinated cattle. Studies have been carried out in cattle by Rossi (1969) and Stemshorn (1985) . However Nicoletti (1969) considered that antibodies from s19 vaccinated cattle could interfer. This test has not been widely used.

The Indirect Enzyme Linked Immunosorbtion Assay (ELSIA).

iELISA is the most modern test developed in an attempt to improve on the results of existing tests. It can be possibly be used for screening or as a confirmatory test either using milk or blood. Bercovich (1990) demonstrated that it is a satifactory substitue for the MRT in milk But he found it slightly less sensitive than MRT-8. Kerkhofs (1990) tested tank milk samples from 105 farms using MRT and ElISA IgM, IgG1 and IgG2. ELISA and MRT detected 69 and 40 Brucella positive herds respectively. ELISA IgG1 was the most useful test. ELISA detected brucellosis before MRT up to 6 months. If collostrum was present in milk then false positives could occur.Cargil (1985) developed an ELISA test and considered that it found more reactors in infected herds. His results showed that for vaccinated herds it identified 1.79% of cattle as reactors compared to 1.44% with CFT. The results for non-vaccinated herds were: ELISA, 4.2%; CFT, 3.59%. The ElISA test was less specific than CFT but more sensitive. It has the advantage of being relatively quick to perform. It can be used to measure IgA, IgM and IgG singly or together. Stemshorn (1980) tested 430 sera from infected herds and found good agreement to CFT. However there is a question as to whether it is better than CFT. Van Aert (1984) compared ELISA with CFT and SAT. Of 11 cattle which showed as positive to culture tests post mortem: ElISA, SAT, CFT and RBT foumd 8, 2, 3 and 3 rspectively. However Bercovich 1998 notes that some workers consider that ELISA is less sensitive than CFT. The ELISA test is costly if used on a small scale.

The use of the ELISA test was first reported by Carlsson (1976) using Brucellas LPS.

The following is a simplified version of the test procedure: The antigen used in the standard iELISA is either sLPS or OPS derived from B abortus S1119-3 or S99. A buffered solution of this is coated into all the wells of a standard 96 well microplate which is then incubated at 4°C for 18 hours. The plates may then be stored at -20°C and will be usable for upto 1 year. For use the microplate is washed with phosphate buffer saline (PBST) to remove unbound antigen and then sera samples in PBST (with EDTA/EGTA) at a range of dilutions are added to individual wells. Standard sera samples are also used to quantitatively define the range of the test at high and low values. The plate is incubated at 20°C for 30 minutes and then washed with PBST to remove unboung sera. Then a conjugate of a solution of monoclonal antibody (MAb) (specific to a heavy chain epitope of bovine IgG1) and horseradish peroxidase (HRPO) in PBST is added to all wells and incubated at room temperature for 30 minutes. Finnally unbound conjugate is washwd away and hydrogen peroxide (substrate) and 2,2'-azino-bis-(3-ethylbenzthiazoline-6-sulphuric acid (ABTSI) (chromogen) in citrate buffer (pH 4.5) is added to all wells and the plate shaken for ten minutes to develop the colour. The optical density is read using a spectrophotometer at 405 nm and the results determined against the standards. A typical example of the full test regeme is provided by OIE Terrestial Manual (2009) .

Lindberg (182) .

The Competitive Enzyme Linked Immunosorbtion Assay (cELSIA).

This test is similar to the iELISA but in this case, prior to the test proper the plate with attached antigen is treated with a specific monoclonal antibody (MAb) against the common epitope (C/Y) which is the immunodominant epitope in O-LPS side chain od both B abortus and B melitensis. This technique has been developed for the diagnosis of cattle, sheep and goat. A problem with the iELISA is that it gives false postitive results in relation to cross reation with antibodies to other other bacteria and s19 vaccinated cattle or Rev 1 vaccinated sheep and goats. The use of the MAb decreases the sensitivity of the test but increases its selectivity. It can be used to disciminate between cattle naturally infected with B abortus and those vaccinated with s19 vaccine and sheep and goats naturally infected with B melitensis and those vaccinated with Rev 1. This has been demonstrated by MacMillan (1990) Muñoz (2005) Nielsen (1995) Stack (1999) Weynants (1997) .

Marin (1999) showed that cELISA was much more specific than iELISA in sheep and could be used in the differentiation of naturally infeced and Rev 1 vaccinated sheep.

The following is a simplified version of the test procedure: The antigen used is the same as in the standard iELISA is either sLPS or OPS derived from B abortus S1119-3 or S99. A buffered solution of this is coated into all the wells of a standard 96 well microplate which is then incubated at 4°C for 18 hours. The plates may then be stored at -20°C and will be usable for upto 1 year. For use the microplate is washed with phosphate buffer saline (PBST) to remove unbound antigen and then a solution of the competing MAb in PBST/ETDA was added to each well followed immediately by sera samples in PBST (with EDTA/EGTA) at a range of dilutions. Some wells were treated with standard sera samples are also used to quantitatively define the range of the test at high and low values. The plate is incubated at 20°C with shaking for 30 minutes and then washed with PBST to remove unboung MBa and sera. Then a solution of commercial goat anti-mouse IgG (H and L chain) HRPO comjugatre was added to the wells and the plates incubateed at room temperature for 30 minutes. Finnally unbound conjugate is washed away and hydrogen peroxide (substrate) and ABTSI (chromogen) was added to all wells and the plate shaken for ten minutes to develop the colour. The optical density is read using a spectrophotometer at 405 nm and the results determined against the standards. A typical example of the full test regeme is provided by OIE Terrestial Manual (2009) .

Nielsen (1996) describes the use of a monoclonal antibody (M84) with specificity for an epitope of the O-polysaccharide comsidered to help in the differentiation of s19 vaccine and B abortus infecteed cattle. Much better differentiation was achieved with cELISA than iELISA or BPAT.

cELISA in cattle. Nielsen (1994) and

Fluorescence polarization immunoassay (FPA).

The principles of FPA were discovered by Perrin in 1926 and a test for antibodies to B abortus in cattle developed by Nielsen (1996) AND (1998) . FPA depends on rate of rotation of molecules in solution. The rate of rotation of a molecule is inversely proportional to its size so large molecules rotate more slowly. For the purpose of the test a fluorescent dye such as fluorescein isothiocyanate (FITC) is conjugated to a small Brucella antigen (such as O-polysaccharide (OPS) (av.22kDa)) prepared from B. abortus S1119.3) that is excited by plane-polarized light at the appropriate wavelength. If the antigen binds to an antibody then the speed of rotation changes ant this can be measured using fluorescence polarization. A fluorophore is chosen which has fluorescence lifetime (the time between absorbing a photon and emitting one) on the same time scale as the rate of the molecule’s rotation. The test can be carried out rapidly in real time. It is a homogeneous assay which does not require removal of unreacted reagents and can be performed very quickly using portable equipment either in the laboratory and in the field. A suitable commercial test assay is supplied by Diachemix Whitefish Bay, WT, USA. Typically FPA was conducted in 96-well flat-bottom black polystyrene microtitre plate. Initially, Samples of sera and diluted sera are mixed with buffer and a background reading made by fluorescence polarization. Then antigen conjugated with FITC was added to each well, the solution mixed and then a secod reading taken. The background reading is subtracted from the test reading and the result recorded in millipolarization units (mP). The results are interpreted as follows: <20 mP, negative; ≥10<20, suspect; ≥20, positive. As well as sera FPA can also be used to test whole blood and milk from individual animals or bulk tank samples, Nielsen (2001) . THe test is also cost effective. Dajer (1999) tested 560 bovine sera from Mexico comparing FPA with other tests and found that FPA gave results similar to CFT and better than RBPT or Riv. They considered it a suitable replacement for CFT. Gall (2000) compared BPAT, CFT, CELISA, IELISA and FPA serology tests for the detection of antibodies to B abortus in bison (Bison bison). They concluded that in terms of sensitivity and specificity the FPA test gave the best results. The performance of the CELISA, IELISA was almost as good and better than BPAT and CFT. Data suggests that FPA could differentiat between s19 vaccinated and infected bison. The ease of use, low cost and rapid results made the FPA an attractive test procedure.

According to Corbel (2000) the sensitivity (99%) and specificity (100%) of FPA for bovine brucellosis is similar to that of cELISA.

Nielsen (2000)

Nielsen K. (2003): Fluorescence polarization assay for diagnosis of human brucellosis. Journal of Medical Microbiology; 52: 883–7.

Nielsen (2004) demonstrated that FPA can be used for the diagnosis of B melitensis in sheep and goats.

Delayed type hypersensitivity Skin Test (DTH).

DHT can detect cattle which have had direct contact with Brucella and is highly specific, . Use of a suitable antigen , brucellins, means the cattle do not become sensitised to the test in repeat use. A positive result dos not prove active infection. DTH does not find all cases of brucellosis in cattle. There can be discrepancies between its results and those of serological and bacteriological tests, Nicholetti (1983) and some seropositive and culturally positive cattle give negative results with DTH but it can detect brucellosis where serology tests fail. It can detect brucellosis in heifers born to infected dams, Bercovich (1990) and in cattle with negative or anbiguous serology test results, Bercovich (1989) , (1990) and (1996) . The DTH test is normally used in conjunction with serology tests and is particularly useful to define outbreaks in areas which are generaly free of brucellosis or in which it is present at low levels.

The antigen used in the test is Brucellin which is cytoplasmic protein which is a cold saline protein extract derived from the rough B melitensis 115 strain which is free of O-LPS, Jones (1973) .

THe DTH skin test is considered more sensitive than serology tests in sheep, Ebadi (1983) (1984) and Trap (1982) .

A problem with the test which only applies to sheep is that the test can also give positive results in rams infected with B ovis.

In Brucella free animals the test is 100% negative, Alton (1990) .

Problems with the DTH test are: the necessity of of catching and testing animals on two occasions; the slowness of the test at 3 days; the relatively poor response to the test which is about 25% of that of the similar test for tuberculosis; and possible sensitisation of the animal to othe tests for brucellosis.

γ-INF Test.

This is a test for Cell Mediated Immunity. The test is carried out in vitro so there is no resultant impact on the animal such as with the skin test. The test was originally developed fot bovine tuberculosis, Rothel (1990) and Wood (1990) and (1991) . The usefulness of the γ-INF test in cattle with B abortus has been evaluated by Weynants and by Garrido (1992) and Duran-Ferrer (2004) . In Weynant's work took samples of heperinised whole blood and cultured them with with Brucellin (cytoplasmic protein free of S-LPS) fo 18 to 24 hours at 37°C. γ-INF in the supernatant was then tested using a propriety sandwich ELISA kit. A base line at for stimutation index was established by testing brucellosis free animals. Values above this level were considered positive. He carried out comparative tests on 10 cows experimentally infected with B abortus using RBPT, CFT, EDTA-SAW (an agglutination test), IELISA, DTH skin test and the bacterial culture test. The γ-INF showed the best performance. RBPT and EDTA-SAW identified uo to 9 animals but this was a transitory and dropped off rapidly over time to 1 despite positive culture tests. CFT identified only positives and also dropped off. iELISA identified 10 infected cattle but development of positive results was slow. DTH identified 5 and γ-INF 10 infected animals respectively in both cases results were sustained over time. The γ-INF test was also the first to show positive results of any test. The most interesting aspect of the test is that it can be used to differentiate between naturally innfected animals and those infected with Y enterocolitica 09, Weynants considered 40 cows of which 23 of which gave FPSR in various tests including: EDTA-SAw, RBPT, CFT and iELISA. The γ-INF test gave almost exclusively negative results. In this regard the test is being used in field trials in France and Belgium. In the work of Duran-Ferrer with experimental infections of B melitensis in ewes the test was carried out with both S-LPS or cytosolic protein and results were superior to RBPT, FCT and iELISA.

The Anamnestic Test.

If brucellosis is suspected but serology tests are equivical or negative, an anamnestic vaccination may be considered.B abortus s19 or 45/20 vaccines have been used, Corner (1983) , Reid (1972) . A reaction is then determined by a serology test such as CFT. A positive reaction indicates existing disease.

The Culture Test.

The culture test may be performed by culturing body tissue or secretions such as: blood , milk, virginal discharge. Growth of the bacteria is unequivical proof of infection. The disadvantage of the test is that it is difficult to perform, slow and does not work in all cases. The best results are usually obtained after abortion ot parturition. The lochia is usually cultured but results are affected by contamination. The culture test on blood will only work if the animal os bacteremic which is not usually the case. Milk is often found to contain Brucella by this test. Post mortem tests of lymph nodes, liver, spleen, udder and other organs can show positive results associated with negative serology tests. In this respect the culture test has been widely used in research. The culture test may be used to differentiate between B abortus, B melitensis and B suis such as by the methodology of Meyer (1962) , (1958) , (1959) , (1961) , (1961) (1961) , (1961) and (1962) .

Comparison of the tests.

Stemshorn (1984) and (1985) reviewed the use of a range of tests: BPAT, SAT, SPAT, 2MET, RBPT, CARD, and CFT. None of the test individually or collectively finds all cases of brucellosis in an infected herd. CFT was considered the most effective finding 79% of culturally proven cases of brucellosis. The 3 acidified antigen tests (CARD, BTAT and RBPT) produced similar finding at about ~ 74.5% of which BPAT was considred the best. SAT was considered the least useful finding only 68.9% of cases. The normal procedure was to use an acified buffer test in combination with the CF test. notes that a screening test should be rapid , ecconomical but not highly specific whereas a confirnatory test must be sensitive and specific.

There are few signs of brucellosis in cattle apart from abortion and occasional lameness, F . These signs are not unique to the disease and cannot be used to eliminate brucellosis in cattle therefore laboratory type tests are essential in its control. Serology and milk tests (SAT, MRT, CFT, RBPT, and ELISA or combinations of theese are subject to failure. Therefore some cattle infected with brucellosis give false negative test results and are a source of re-contamination of a herd. Like serology test DTH does not find all cases of brucellosis but it does find cases missed by serology tests.

Sutherland (1985) studied vaccinated and unvaccinated cattle and found that 9 of 27 infected cattle which had previously been vaccinated with s19 or 45/20 vaccines and then challenged withB abortus s544 and which wre later found to be infected had low, delayed, transient or negative test results for CFT. Interestingly these CFT negative cattle produced more live calf births than infected cattle with CFT positive. The ELISA test showed earlier and more peristant positive results on the same cattle. However ELISA gave was positive much longer on vaccinated cattle than CFT and gave rise to more false positive results. However in unvaccinated cattle ELISA can be used either as a screening test or as a confirmatory test to identify susspect cattle missed by CFT.

Elberg (1973) considered that cattle with B abortus were more likely to have IgG type immunoglobulins probably relating to the duration of disease, IgM could predominate early in the disease or if s19 vaccine had been used.

C

According to studies in mice and Winter (1990) both antibody and cell mediation immunity are involved in pretection of mice against Brucella. Antibodies response to antigen in the outer membraneof Brucella, . Further studies by Cloeckaert (1992), Limet (1989) showed that anti-S-LPS were protective while anti-OMP and anti-R-LPS MABs were less potent.

False Positive Serology Reactions in Cattle. The main serological response of cattle to infection with smouth Brucella is to the smooth lipopolysaccharide (S-LPS) in the surface of the bacteria and specifically to its O-PS chain which is the most exposed part, Bowden (1995) Cloeckaert (1990) . All the main serology tests depend on detecting the antibdies developed agianst these antigens. One of the main problems with these tests is that they cannot differentiate between cattle infected with Brucella and cattle (particularly adult cattle) treated with s19 vaccine and also cattle infected with other gram-negative bacteria which have similar O-PS surface structures. This is known as cross reaction. Both these problems give rise to false positive serological reactions (FPSR).

FPSR due to Cross Reactions. In cattle the most important cross reaction is with Yesinia (Y) enterocolitica 0.9. Greater undestanding of the structure of Brucella antigens and the development of newer tests such as ELSIA using MAbs have offered the possibility of more antigen specific tests in an attempt to address the FPST problem, Douglas (1988) . According to Weynants (1997) 4 epitones of O-LPS have been describes, M, B, A, C and C/Y but overlap between epitones occurs. Different Brucella species or strains may express epitopes in different amounts. For example B abortus W99 is a A-dominant strain whereas B melitensis Rev 1 is a M-dominant strain. C is designated as Common and and shares epitopes with both A and M strains. Weynants (1997) proposes the sub-division of C and C/Y as follows: C(M>A), C(M=A), C/Y(M>A), C/Y(M=A), and C(A>M). The M epitope has an association of α-(1-3) and α-(1-3) linkages not present in Y enterocolitica. The similarity in the O-LPS chain of Brucella and Y enterocolitica relates to A and C/Y epitopes. This is a homopolymer of N-formyl-perosamine joined by α-(1-2) and α-(1-3) linkages. Thus B abortus biovar 1 has exclusively in α-(1-2) linkages whereas B melitensis Biovar 1 has α-(1-2) plus α-(1-3) in a 1:4 ratio. The O-LPS are charactorised by overlapping epitopes (antigenic determinants): C (common to all types of Brucella O-chains), M [present in O-chains with α-(1-3) linkages], and A [present in O-chains with no α-(1-3) linkages or with a proportion of α-(1-2) to α-(1-3) linkages higher than 4:1], Cloeckaert (2001) Douglas (1988) and Weynants (1996) . According to the O-chain of Y. enterocolitica O:9is a homopolymer of N-formyl-perosamine in α-(1-2) linkages that is indistinguishable from the O-chain of B. abortus biovar 1 and monoclonal antibodies (MAb) of O-chain react equally with S brucellae and Y. enterocolitica O:9 (C/Y epitopes). However other MAb recognize epitopes of Nrucella not present in Y. enterocolitica O:9 (C epitopes) indicating structural differences, Cloeckaert (1998) Douglas (1988) and Weynants (1997) strongly suggesting subtle structural differences.

The normal form of brucellosis in cattle is B abortus but according Verger (1985) and (1989) cattle in association with sheep and goats readily become infected with B melitensis.

Some serology tests have been devised to detect antibodies to Brucella proteins, M Diaz (1989) and .

All the earlier tests used killed Brucella or LPS derived antegen, however in recent years the study of Brucella has revealed more knowledge of its molecular structure and tests have been based of more sellective structures such as: rough LPS (Core Lipid A epitopes), native hapten polysaccaride, rough LPS group 3 membrane protein complexes, recombinated BP26 protein and cytosolic proteins. These have been evolved mainly in relation to the problem of FPSR. Simpler tests such as RBT, CFT and iELISA S-LPS or native hapten or S-LPS derived polysaccraides give high sensitivity but lower specificity. These tests give rise to FPSR. Tests which use specific antigens such as immunoprecipitation tests with native hapten polysaccraride, countercurrent-immunoelectrophoresis with cytosolic protein give excellent specificity but poorer sensitivity.

Muñoz (2005) concludes that Y enterocolitica 0.9 is the main cause of FPSR in cattle, Gerbier (1997) giving rise to high titres in blood and milk, . The prob;em particularly affects countries where B abortus has effectively been eliminated but where surveillance continues, . Studies in the EEC have indicated that FPSR due to Y enterocolitica 0.9 can be as high as 15% in such cases, Gerbier (1997) Pouillot (1997) and (1998) , Saegerman (1999) and Weynants (1995) .

Steak (2000) developed a cELISA based on a chaotropic agent.

Serology Tests Based on Brucella Proteins.

According to the work of Dubray (1976) the Brucella cell enclosure is a 3 layered structure comprising: an inner cytoplasmic membrane, a periplasmic space and an outer membrane. The outer layer contains: lipopolysaccaride (LPS), ptotein and phospholipids. Cloeckaert (1990) describes the preparation of 7 outer memrane protein (OMP) by sodium dodecyl sulphate extraction, separation by polyacrylamide gel electrophoresis and identification by Western immunoblotting using MAbs. The OMP recognised according to mass were: the porins (25-7 and 36-38 kDa) and the minor proteins (10, 16.5, 19, 31-34, 89 kDa). Further studies using electron microscopy and ELSIA with anti-OMP MAbs indicated that all anti-OMP MAbs attached to these proteins at the surface of rough Brucella but that binding was was much better on rough strains than smooth due to steric hinderance caused by long O-PS side chains in smooth Brucellosis.

Giambartolomei (1996) used ELISA to compared cattle vaccinated with s19 or infected with Y enterocolitica 0.9. The IgG and IgM response against LPS, LPS free cytoplasmic proteins (CYT) and 18 KDa CYT were tested. High titres of anti-LPS IgM and IgG were recorded similar to 'Brucella infected cattle. THe IgG and IgM response to LPS-free CYT and the 18-kDa protein was negative or low in vaccinated or Yersinia-infected cattle but higher in infected cattle. This suggests that cytoplasmic proteins of B. abortus could be useful for the differential diagnosis of bovine brucellosis.

It should also be noted that protein tests include delayed-type hypersensitivity (DTH) based on Brucellergene, a Brucella protein extract, which can be used to test for B melitensis in sheep, Blasco (1994)

Chukwu (1985) and (1987) used lymphocyte transformation (an in vitro whole-blood lymphocyte stimulation procedure) to compare cell-mediated immune response in cattle infected with B abortus s544 and Y enterocolitica 09. A soluble polypeptide derived from B abortus s19 containihg some LPs was used as antigen and Concanavalin A as mitogen. Cattle infected with B abortus developed a high lymphocyte transformation responses while the results were low for Y enterocolitica 09 infected and non infected were low. This was concluded to be a possible test to differentiate between B abortus and Y enterocolitica 09 infections.

The OMP 25-7 kDa (Omp28) antigen has been shown to be an immunodominant antigen in infected cattle, sheep, , goats, and humans, Lindler (1996) . The protein has been redesignated as BP26 based on work on the encoding gene by Rossetti (1998) . showed that this protein antigen has a different antibody responses in infected sheep to those of B. melitensis Rev. 1 vaccinated sheep.

Cloeckaert (1996) . Cloeckaert (2001) and (1996) used iELISA to test for the OMP at 25-7 kDa (Omp28) antigen to compare antibody response for sheep naturally infected with B melitensis and vaccinateed with Rev 1. The response to BP26 was delayed and much weaker than that to O-PS nevertheless it was considered that the BP26 protein could be used as a diagnostic antigen in confirmatory tests to differentiate between infected and B. melitensis Rev.1 vaccinated sheep. The results accorded well with bacteriologically and DTH resulta: bacteriologically and serologically +ve sheep, 100% agreement; bacteriologically -ve but serologically and DTH +ve sheep, 88% agreement; bacteriologically and serologically -ve but DTH +ve sheep, 84%.

Seco-Mediavilla (2003 found that a recombinant fusion protein including the reactive region of BP26 was potentially more specific than the entire BP26 since it avoided some false positive results.

Allergic skin reactions are also a form of protein antigen test. Godfroid (2002) studied heifers infected with bovine brucellosis due to B abortus biovar 1 and B suis biovar 2 and infection due to Y enterocolitica O:9. He concluded that traditional tests such as SAT, RB, CFT, IELISA and CELISA were insufficiently specific for brucellosis where Y enterocolitica O:9 was present. A Yersinia outer membrane proteins (YOPs)-ELISA was able to detect a Yersinia group infection but this did not determine if brucellosis was present, as in dual infection. SAT, RB, CFT, and cELISA were ineffective in persistantly infected animals. iELISA was more effective. The brucellergen skin test was the only proven specific test.

Godfroid (2002) concluded that B. suis biovar 2 from infected wild boars is a risk factor in cattle doth infection and transmission. He considered the infection to be self limiting.

Diaz ,

Iannelli (1976) describes a method of testing for antibodies to Brucella using radial immunodiffusion based on the method of Vaermann (1968) . Bovine serum was place in a well of a plate of agarose impregnated with anitegens from B aborrtus and B melitensis and allowed to diffuse. Precipitation circles indicated the presence of antibodies. It is suggested that the method could be used instead og CFT or to differentiate between the antibodies of infected animals and those of vaccinated animals.

Diaz (1979) and Jones (1980) showed that the antigen, poly B derived from rough biovar B melitensis B115 incorporated in RID plates could distinguish between could be used in RID to distinguish between the sera of infected and s19 vaccinated cattle. Precipitation lines occuring in infected cattle were absent in vaccinated cattle. The RID test is simple rapid could possibly replace the technically demanding and sometimes unreliable CFT test. The RID test takes about 3 hours to perform. S-LPS and Poly B are components of the surface layer of smouth Brucella but only Poly B is present in the rough biovar possibly from cytoplasma. S-LPS is an endotoxin, , and is the major anigen responcible for antibody response in brucellosis, Diaz (1968) and (1972) . Anitibodies to S-LPs occur in early infection and antibodies to Poly B. Therefore antibodies to Poly B indicat continued infection.

Daz (1976) used counter-immunoelectrophoresis (CIEP) to study human patients with clinical brucellosis and other patients with yersinoisis 09. The protein anigens was derived from B melitensis and Yersinia 09. Buffered agarose slides were used. Protein antigens were considered free of LPS. LPS antigens cannot be used because they travel in the same direction as the antibodies. Sera samples were introduced in wells at the anode and the protein antigens at the cathode end. The tests took about 2 hours. Concurrent RB tests for brucellosis showed positive titres for both yersiniosis and brucellosis samples. Precipitation lines for yersiniosis were absent from the brucellosis samples and vice versa. Antibodies to Brucella protein antigens take longer to develop than IGM and IgG. 1-7 precipitation lines were observed. This is a test which could beuseful in animals and humans if cross reactionis considered possible.

Hunter (1972) compared the SAT, RB and CF tests on sera and milk samples and concluded that that non of these tests found all cases of brucellosis.

Blasco (2010)

Eradication of Brucellosis in Domestic Animals.

Eradication of B abortus in Cattle.

The main purpose of the attempts to elimating brucellosis from animals has been to eliminate the ecconomic loss associated with the disease. These ecconomoc losses relate to loss off offsprings, loss of milk and subsequent animal infertility. However the indirect improvement in human health is also a very important factor. In most cases humans catch brucellosis by direct contact with infected diseased animals or from their products, such as milk, milk products and meat. In humans the disease has both health and ecconomic consequences. They are often suffer from a long term illness and incapacity which frequently prevents them from working.

An effective control of animal brucellosis requires the following elements:

• Surveillance to identify infected animal herds.
• Prevention of transmission to non-infected animal herds.
• Eradication of the reservoir to eliminate the sources of infection in order to protect vulnerable animals or herds coupled with measures to prevent reintroduction of the disease.

Three methods of eradication have been used:

• Testing animals and removing reactors.
• Vacination.
• A combination of both test/slaughter and vaccination.

xzc

Bovine brucellosis causes abortion, the birth of weak calves. infertility and reduced milk production. Abortion usually occurs in at 5 to 7 moths of gestation. This often associated with retained placenta and infection of the uterus. Other signs are Swellinng of the testicles in buls and lameness assocated with bursitis.

Socialand ecconomic factore affected farmers with infected cattle include: loss of calves, infertility of animals, decreased milk production , weak calvees, joint infections aqnd higher losses due to culling of unproductive animals.

B abortus in horses is most likely to affect tendons, muscles, and joints with bursa, fistulous withers and pollevil. Horses can spread B abortus back to cattle, Elliot (1977) . Infected horses should be destroyed for the protection of humans. B Abortus is present in Africa, the Middle East, Central and South America and many developingt countries. The introduction of B abortus into a clean herd of cattle causes a surge in abortions of 30 -80%. Calves do not show signs of the illness but may remain infected. This is less evident in dry hot conditions. iELISA has the advantage over CFT that it is more tolerant of the quality of blood samples. Times for serology tests are approximately: RBPT, 10 minutes; CFT, 4 hours; iELISA, 2 hours; MRT, 1 hour. 'B abortus is not the only cause of abortion in cattle. Other causes are: Trichomonas foetus, Neospora caninum, Campylobacter fortus, Listeria monocytogenes, Sarcosporidia, Leptospira, fungi and Rift Valley fever. Abortion can also be caused by poisoning.

Limited trials have been carried out to assess the usefulness of therapeutic agents to treat B abortus in cattle with limited success. Antibiotic tratments have been tried by Radwan (1995) similar to those used for humans. He used intramuscular injections of oxytetracycline ans steeptomycin with some sucess. Such treatments are of limited use because of cost, the possibility of the development of antibody resistance and practical considerations.

Dose rate for calves at 3 to 6 months at 5-8 x 10¹⁰ viable organisms. s19 offers good immunity against a moderate challenge by a more virulant strain of B abortus such as natural strains.

In consideration of human infection it should be noted that RB51 does not produce antibodies detecable by the usual serology tests. Infected cows shed large amounts B abortus for about 10 days after birth or abortion with high risk from the resulting contaminated environment. The incubation period in cattle is variable depending on dose size, sexual marurity, state of pregnancy, and possible previous exposure. For humans it is 5 to 30 days.

Brucella is rapidly destroyed by heat, dryness and direct sunlight. Cows usually abort once and become chronic carriers which shed B abortus intermitently in mi9lk and reproductive secretions for many years. THe amount of B abortus spread at subsequent births is reduced. Infected cows secret B abortus intermittently throughout lactation. Infected bulls may excrete large numbers of B abortus in semen however infection of cows by natural transfer is considered limited. In countries where B abortus;; is endemic at high levels vaccination is necessary to a level where test and slaughter is feasible. Factors which might favour the use of vaccination are: types of herds affected including ranch cattle, a large number of herds affecxted, high and unreallistic costs of culling dependind on socioecconomic conditions, general public health considerations and the diffuiculty of differentiating infectedd and vaccinated cattle. Attempts to use vaccination to entirely eliminate brucellosis have failed. There are 2 ways of elimination brucellosi: culling of whole herds were brucellosis is present or test and slaughter. Culling is the most effective and is preferred since it allows the clensing disinfection of infected premises, a period of quarantine for infected pastures etc followed by restocking with clean cattle. At least 30 days must elapse after decontaminationof premises before restocking can begin. This strategy is best if a small number of cattle herds is involved and allows a rapid turn round.

For test and slaughter all cattle are regularly tested (30 to 60 days) using a range of agreed test. Animals which react positively and their calves are destroyed or sent for immediate slaughter. Cattle shoud be tagged. Segragation of herds according to age is desirable. The herd is considered clear when consective tests are negative but testing of dams and calves is continued plus rotine surveillance. Tracew back of infected animals which have been moved is essential. Other local herds must be surveyed. Test and slaughter fails in 5% of herd. Test and slaughter is preferred if the outbreak has spread more widely from the indentifed infected herd or if pedigree herds are infected. Tests of choice are RBPT, CFT and iELISA and MRT for milk. MRT is cheap and easy to carry out and effective in identifying infected herds. It does not identify infected heifers, recently aborted cows or herds where there are only a few reactor cows. The inradermal skin test (Brucellin) is useful for epidemilogical surveys and for post eradication surveillance where incidence is low and is relatively cheap.

Cleaning agents which are effective in destr0ying Brucella are 0.03% formalin, 1% phenol, 0.01% β-propiolactone, sodium hypochlorite, sodium hydroxide, iodines, quaternary ammonium compounds, ether and chloroform.

All personnel handling infected carcasses should take special precautions including the use of: long protective gloves, protective eye shields, effective sealed breathing masks, inpervious aprons and boots. All these must be washws and disinfected between use. Particular care is required in the handling of udders, uteri, bladders, tongue and bursa (if present). Cuts and skin abrasions occurring during carcass handling should be treated immediately and medical help sought including antibiotic treatment. Anyone with an open wound should be excluded from such work.

There are usually no signs of infection in young sheep and goats and they only show a transient serology response. Older animals are apparently more susceptible particularly with pregnancy.

Most breeds of goat are susceptible to B melitensis but the susceptiblity of different breeds of sheep varies. Milking breeds of sheep are more susceptible, Corbel (1984) . The susceptibility of milking flocks of sheep may in part relate to husnabry conditions including being held in large flocks. As in cattle the disease is spread by intermingling of infected and uninfected anumals and herds.

Cattle in association with B melitensis infected sheep or goats are likely to become infected but it is not known if the disease is self sustaining. B melitensis in cattle appears to be more severe than B abortus. Abortions occur at a similar level but B melitensis may be excreted in milk for years and cause epidemics of illness in humansworking with the cattle or drinking the milk. A similar situation occurs in camels where the milk is similarly infected.

Microscopic examination of smears from vagina swabs, placenta and aborted foetuses stained with the Stamp modification of the Ziehl-Neelsen method can provide a presumptive diagnosis of N melitensis however it should be noted that some other bacteria such as B ovis, Chlamydia psittaci and Coxiella burnetti can lead to mistakes.

THe best opportunity of isolating B melitensis from sheep and goats is from vaginal swabs and milk and is most likely to be found in the spleen and lymph nodes (iliac, mammary and prefemoral), Marin (1996) . Details of methodology for culture of B melitensis is reviewed by EC . Success of the culture methodology depends on the viable number of organisms withinthe sampleand the number of specimens tested from the same animal, Hornitsky (1986) . The culture test may not work in cases of samples with low numbers of viable Brucella.

Antigen for serology tests. The antigen most widely used in the RBPT and CFT tests us a suspention of whole cells derived from B abortus biovar 1. This is an A inmmubodoninant strain, Alton (1988) . Other strains such as B meltensis biovar 1, B abortus biovars 4,5 and 9 and B suis biovar 5 areare m dominant strains, however it is concluded that it is adequate to diagnose B melitensis in sheep, Blanco (1994) . THe outer membrane of smooth Brucella is composed ofphospholipids, proteins and lipopolysaccharide (Smouth), s-LPS. S-LPS is the immunodominant antigen. Antibodies reacting to S-LPS including RBPT, CFT and iELISA. s-LPS is composed of an inner glycolipid moiety (core oligosaccharide and lipid A) and an outer polysaccharide chain (O-LPS). O-LPS is composed of a perosamin homopolymer comprising α-1,2 and α-1,3 linkagesm, Cherwonogrodzky (1990) . The O-LPS is the most significant immunodominant area. O-LPS of A dominant B abortus biovar 1 consists mainly of α-1-2 links with only 2% α-1-3 links while M dominant B melitensis biovar 1 has one α-1,3 link for four α-1,2 link. A type epitopes contain none or few α-1,3 links compared to M type epitopes which also contain α-1,2 links in di-, tri- and tetrasaccharide sstuctures, Brundle (1989) , Mieikle (1989) , Cherwonogrodzky (1990) . Epitopes which contain both A and M immunodominant regions are called common (C) as detected using monoclonal antibodies, Douglas (1988) . These may explain why the B abortus biovar 1 antigen works for B melitensis.

Serology tests in relation to outer membrane and inner cytoplasmic proteins have also been examined The eElectrophoresis patterns for different Brucella species are similar, Diaz (1967) and (1967) . THese tests can be used to distinguish between Brucella and other bacteria with similar LPS which causes false positive reaction with the usual serology tests. There is however cross reaction with Ochrobactrum anthropi which is a known pathogen in humans, Velasco (1997) . Brucella cytoplasmicantigens have been used sucessfully in the allegic diagnosis of brucellosis in sheepm and goats, Festerbank (1982) and (1985) and Ebadi (1983) Loquerie (1984) and Blasco (1994) . According to the work of Trap (1982) , Muhammed (1980) and Diaz-Aparicio (1994) cytoplasmic antigens can be used to diagnose brucellosis in sheep anf goats in precipitation tests, however the use of these antigens in iELISA is insufficiently sensitive to differentiate between animals with and without brucellosis infection, Salih-Alj (1996)

Rose Bengal test in sheep and goats. THr RBPT is considered a standard test for B melitensis in sheep and goats, Joint FAO/Who (1986) but the test as used for B abortus in cattle is less sensitive for B melitensis in sheep and goats, Falade (1978) and many give negative values with positive CT test results but can be improved by adjusting the conditions of the test, Blasco (1994) .

CFT. This test is reasonably effective for the diagnosis of B melitensis in sheep and goats but but as in the case of cattle has the drawbacks of complexity, variability of reagents, prozones and anti-complementary activity of sera, difficulty to perform with hemolysed sera and subjectivity of interpretation at low titres, EC . The test has low specificity when Rev 1 vaccine is used but results are better if the vaccine is applied conjunctivally rather than subcutaneously, Festerbank (1982) .

iELISA. The S-PLS antigens adhere strongly to polystyrene and are therefore the most widely used form of the test. The O-LPS side chain has the most influence in the test. A relatively crude antigen extract can be used without significantly affecting the sensitivity of the test in fact there are some problems with pure antigens due to more false positives and micelle formation causing irregular patterns of absorbtion. In sheep and goats iELISa using an extract of S-LPS from B melitensis and polyclonal IgG H + L (Heavy * light) has beeen found suitable to diagnose sheep and goats, Jimenez de Bagues (1992) , Diaz-Aparicio (1994) and Blasco (1994) .

Tests based on anti-globulins based on sera fractions (anti-γ,and antiIgG (heavy and light). Anti-γ favours the adsorption of IgM which has thge disadvantage of non-specificityas like other tests of IgM. Anti IgG (H * L) detects mainly IgG1 and is more specific. Lamb (1979) and Nielsen (1983) (1984) (1988) . Where tests are specific for a particular immunoglobulin the performance of the test may be affected by steric hinderance preventing adhesion on on adjacent sites and therefore the effectiveness of the test which may fail as a result.

For small ruminants an iELISA test best on S-LPS from B melitensis as antigen and polyclonal conjugate (ant- IgG (H + L) has been used sucessfully by Jimenez de Bagues , Diaz-Aparicio (1994) , Blasco (1994) . As with cattle iELISA is less effective with in low incidence situations due lack of specificity, Jiménez de Bagüés (1992) but this problem may be reduced by the use of protein G as conjugate, Diaz-Aparicio (1994) and Ficapal (1995) or monoclonal anti-ruminant IgG conjugate. in both cases the sensitivity of the test is reduced. the iELSIA lacks specificity where REV 1 vaccine is used.

In the case of Rev 1 vaccinated sheep and goats it is necessary to differentiate between vaccinated and naturally infected animals. RBPT, CFT and iELISA have low specificity in this respect. Tests which can possibly be used are Gel Diffusion or Radial immunodiffusion tests using Native Hapten (NH). These react more strongly to antibodies from infected sheep than to those of Rev 1 vaccinated sheep but this has not been proved in field trials.

cELISA. There is a form of cELISA called Particle Concentration Immunoassay (PCFIA) is a form of cELISA, Nicoletti (1993) . in this case the sub-microne antigen coated polystyrene particles are suspended in solution. This is incubated with serum and an anti-Brucella conjugate which compete for bindind. THe polystryrene beads are separated and analysed. A commerciasl form of this test is available which is reported to test 550 samples per hour. The specificity of this test appears to be lower than the CF test, Greenlee (1994) .

Native hapten. When Brucella is physically degraded such as with hot water at 120°C, S-LPS, native hapten (NH), nucleic acids and denatued proteins are released. THe S-LPS can be used in place of killed Brucella in tests such as iELISA. NH dose not normally adhere to polystyrene unless modifications are used such as acylation when it gives similar results to S-LPS, Alonso-Urmeneta (1988) . Radial immunodiffusuion test can be used to differentiate beween naturally infected cattle vaccinated with s19, . NH is particularly usefull in this respect. This is also true of goats where RID can be used with a M-type NH derived from B melitensis to differentiate between naturally infected animals and those vaccinated with Rev 1, Diaz-Aparicio (1994) . In general the fall off of antibody titres is faster for vaccinated animals than in those naturally infected. This is particularly true if the conjubctival vaccine royute is used.

In general tests used for cattle such as RBPT, CFT have used for small ruminants but they are not entirely satisfactory due to the absence of the M epitope in B abortus. Modifoied test procedures which work better have been suggested by because they are based on B abortus antigens which are have been modified by Diaz-Aparicio (1994) . The iELISA using S-LPS and polyclonal antibodies give good sensitivity but poor specificity but if protein G or anti-IgG1 monoclonal antibody conjugate is used then the speficity is improve.

An ELISA test for goat brucellosis is described by Diaz-Aparicio (1994) .

A possible reason that Elisa using cytosolic proteins does not work very well is may be poor adhesion to polystyrene.

Marin (1999) compared RBPT, CFT, iELISA, cElISA and RID with NH for the identification of natural brucellosis melitensis in sheep and also the differentiation of naturally infected from Rev 1 vaccinated sheep and concluded that the best choice was a combination of RBPT and RID-NH. Although a combination of iELISA and cELISA could be used RBPT and RID-NH was simpler and easier to use.

Cross reaction. B ovis is a rough variant and B melitensis is a smooth variant which co-exist in sheep. It might be expected that test dependant on S-LPS would give negative titres with regard to the antibodies associated with rough variants. However, according to EC SANCO.C.2/AH/R23/2001 (2001) initial studies indicate significant cross reactions when the iELISA test with S-LPS is used leading to difficulties in interpretation. Also allegy and RID tests ma which depend on protein antigens which occur in both smooth and rough variants. Goats do not suffer from B ovis and are therefore not affected.

According to EC SANCO.C.2/AH/R23/2001 (2001) In sheep and goats, the major clinical sign is abortion but other signs may be observed, such as orchitis, epididymitis, hygroma, arthritis, metritis, subclinical mastitis. Animals may develop self-limiting infections or become latent carriers with potential excretion of the bacteria. This stage is frequently associated with a persistent infection of the mammary gland and supramammary and genital lymph nodes with shedding of the organisms in milk and genital secretions. Abortion generally does not occur if the female becomes infected towards the end of pregnancy.

Problem Cattle Herds.

Vaccination of herds of cattle can reduce the incidence of brucellosisto relatively low levels but cannot eliminate it. Vaccination of adult cattles with s19 confused diagnostic tests and was generally discontinued. Vaccination of calves which loose their antiboody response to Brucella at about 6 months as a control methodd has continued. The only successful method of eliminating brucellosis from herds of cattle has been test and slaughter. However this a long and laboriuos process with many set backs. The main problem in this regeme is the inadequacy of serology tests. The term 'problem herds' refers to herds of cattle which have been subject to test and slaughter but in which brucellosis conntinues to recur over a long period. In these herds brucellosis can recur for a number of reasons but the most important is cattle termed 'latent carriers' which carry the disease but cannot be detected by the usual serology tests. Latent carriers can be aready present in a herd, or introduced from another herd and often considered safe, .

Many of the problems associated with eradication of brucellosis from cattle are described by Christie (1969) . He noted that in Northern Ireland prior to an eradication programme which started in 1963 that up to 25% of cattle herds were affected by brucellosis. At this time a typical eradication programme involves tests on bulk milk either at creameries or of individual herds using the Milk Ring Test. Individual infected herds are then dealt with separately. The problems were:

• The difficulties of persuding farmers of the necessity of imposition of strick hygene standards on infected farms.
• The ease of spread of brucellosis within herds and from herd to herd. This can be by direct including milk or indirect contact. The most danderous source of infection occurs at abortion or patition. Isolation during calving and afterwards for 14 days is recommended. He gives the example of infected animals using rented land on brucellosis free area causing a new brucellosis outbreaks.
• With difficult herds it is better to slaughter the whole group because even with sustained effort efforts at eradication success is unlikely.
• Separation of infected animals and their milk is extremely difficult in practice.
• A proportion of pregnant infected cows do not give positive brucellosis tests until after partition so prolonged testing periods are necessary.
• In UK vaccination of adult cattle was banned in 1963.
• Outbreaks of brucellosis in certified free herds in 1989 was at 0.7% and mostly caused by an outside source. Spread to other herds is the norm.

Luchsinger (1963) studied B abortus in herds of cattle in Minnesota during an eradication programme using the culture test. The materials examined were: milk, tissue samples which included: supramammary, retropharyngeal, isschiatic and illiac lymph nodes and sections of spleen, uterus and the 4 udder quarters. The primary means of spread of brucellosis between herds were: introduction of exposed cattle, introduction of chronic undetected carriers and contact transmission by various means from adjacent or neighbouring infected cattle. The normal eradication proceedure included testing with the Milk Ring Test and standard serology tests and the elimination of reactor cattle. The culture test was useful because it identified infected carrier cattle which had negative standard serology tests. Some such infections were found to be due to B abortus strain 19 used to vaccinate calves. The use of vaccines is questioned. It is noted that the culture test is not infallable and misses many cases so it can only be used as a guide.

Nicoletti (1966) studied problem cattle herds in New York and Vermont. The agglutination test had been used to eliminate reactor cattle in an eradication programme but had left problem herds. They isolated B abortus mainly from the milk of cows for which the agglutination test only identified 61%. Other tests were tried but no test found all the culturally positive cattle. the CF test was probably the most effective but was considered inpractical. More positive culture tests were from vaccinated cattle (≤ 20 months) at 82% than unvaccinated. These vaccinated cattle were recevoirs of infection not identified by the agglutination test. This questions the usefulness of this vaccine. Some positive culture tests (4 of 135) were caused by B abortus strain 19.

Morgan (1969) points out that non-pregnant cattle and cows in the early stages of pregnany have longer incubation periods and may be negative for brucellosis tests. Once infected a cow typically only aborts once and subsequent calvings are apparently normal. At this stage the cows have a chronic infection with no clinical signs of illness but with Brucella harboured internally such as in the supramammary lymph nodes and udder with some excretion in milk. As chronic illness pregresses titres from Brucella serology tests fall away to negative or low levels. He notes that brucellosis can spread from a diseased herd to an clean herd on adjacent land even where the fences are good and prevent any direct contact. One possible explanation is the activity of scavengers such as foxes, It is also possivle that blood sucking parisites such as ticks could be involved, Morgan (1964) . Morgan (1969) considered that for heavily infected herds it had been found that it was more successful to slaughter the whole herd rather than attempt a program of test and slaughter which may fail.

Robertson (1871) describes a self contained dairy herd of cattle in which there was a low grade persistent brucellosis infection. From time to time reactors appeared. They describe a typical example of a 4.5 year old cow which had tested negatively on 3 occasions for brucellosis by the RBT, SAT and CFT tests before and after a normal calfing. However after calving a culture test on colostrum and vaginal mucus proved positive on two occasions revealing B abortus Type 1. At 14 days after calving the SAT remained negative, the RBT became weakly positive, and CFT stromgly positive. The cow was slaughtered and tissue from the uterus, vaginal passage and retropharyngeal, prescapular, iliac and supramammary lymph nodes were cultured without finding Brucella. This demonstrates that a synptomless carrier with a low grade infection can exist undetected in a cattle herd. Robertson notes that this is a well document case but is not a unique.

Morgan (1972) and (1974) reviewed the eradication porogramme in the UK. The vaccine B abortus Strain 19 was introduced in the UK in 1942 and used in both calves and mature cattle mainly to reduce the incidence of abortion but also to reduce the incidence of brucellosis. From 1964 a survey of cattle based on culturatl/biological examination of milk and placentae indicated that 25 -30% of cattle herds contained infected animals and that 2% of cattle individually were infected. It should be noted that such tests could only show a proportion of infected cattle. The situation was confused by the presense of vaccinated cattle which could also give positive results, therefore vaccination was restricted to calves from 4-8 months old. In 1967 this was further restricted to female calves between 91 to 100 days old. Abortion is considered the only sign of the disease but that even if the culture tests on the placenta and fetus were positive concurrent blood tests could be negative. Vaccination was replaced in the 1960-70s by a test ans slaughter regeme. Herds were monitored using either the MRT, replaced in 1970 by the RBP test, and the SA test. Cattle that showed as positive by this regeme, 'reactors', were removed from herds and slaughtered with compensation to farmers so that gradually the herds became free of brucellosis and were given accredited status. A serious problem were symptomless carriers with negative serology. This problem was partially solved by the use of the RBP and CF tests which were able to identify more but not all such carriers. Restock was only allowed from brucellosis free herds. This was a gradual process taking 20 years with many setbacks. At the end of this time cattle in the UK were recognised as brucellosis free.

Eradication of B melitensis in Sheep and Goats.

The strategy for eliminating B melitensis in small ruminants is generally based on procedures developed for B abortus in cattle. B abortus s19 vaccine if ineffective against B melitensis in sheep and goat therefore B. melitensis Rev.1 vaccine is used. Young animals are vaccinated and it is considered that this can offer lifelong immunity. THerefore it is reasoned that over a 5-7 yesr period all animals will have been vaccinated and brucellosis suppressed to low levels. Blasco (1997) notes that full dose Rev1 causes abortion in pregnant animals but a low dose strategy which overcomes the problem gives ineffective protection.

Blasco (2010) notes coditions necessary for eradication of B melitensis in sheep and goats:

• A decision making organisation is required.
• A comprehensive vetinerary service is the most important factor with the resources to vaccinate all animals in a short period of time.
• A comprehensive data base of all flocks.
• Adequate financial resources to compensate falmers for slaughter and other costs and therefore secure their co-operation.
• Education, organisation and involvement of shepards.
• A national comprehensive epidemological survey to determine the types and extent of infection in animal species. Including the assessment of different areas. Based on this survey the type of program should be determined:
  • Low prevalence (< 1% of flocks) - test and slaughter with no vaccination.
  • Moderate prevalence (>1<10% of flocks) - Test and slaughter for adult animals and vaccinatio of young (3 - 4 months old) replaclement animals.
  • High prevalence (≥ 10% of flocks) Mass whole flock vaccination of all infected .animal species.

Animals are only likely to infect other animals when they are excrete the bacteria. B melitensis can be transmitted comgenitally in sheep and goats or via collostrum and milk, Grilla (1997) .

Whole Flock Vaccination. Whole flock vaccination is the only feasible method of controlling B melitensis in underdeveloped countires where verinary services are rudimentary and there limited ecconomic resources. According to Blasco (2010) a standard dose of Rev 1 at 1-2 x 10⁹ CPU applied conjunctivally is the best option. However if applied to pregnant femals a this causes an unacceptible loss due to abortion and Rev 1 excreted in milk and therefore a danger to humans. The best times to overcome the problem are: during the pre-breeding period, the late lamb season and lactation. In general whole flock vaccination should be applied to the whole country. Following up mass vaccination must be maintained over a number of years:

• Vaccination of only young animals to reduce adverse factors associated with Rev 1 continuing for upto 10 years with tagging of animals. However there is some doubt as to the overall success of this stategy (see below).
• An alternative strategy is to carry out mass vaccination every 2 years over a sufficient period of years to reduce overall brucellosis prevalence to low levels.

Young Replacement Vaccination. Vaccination of young animals only is an alternative to whole flock vaccination. All young animals new to the flock are vaccinated at 3-4 months old. This is about 15-25% of the flock. This process must be repeated anually though sufficient geberation so that all the flock has been vaccinated. This usually takes from 4 to 8 years depending on the animal species. According to Blasco (1997) there is some question as to the effectiveness of this strategy even in developed countries mainly because of the need for frequent vetinary visita and the difficulty of tracing all the young animals. He also considers that this stratagy will not work in high prevalence ares and this is particularly true of nomadic flocks.

Eradication. Once the prevalence of brucellosis in flocks has been reduced by vaccination to low levels the next stage is to attempt eradication. This requires substancial resources as discussed above. Assuming mass vaccination has been used the the main problem is the viability of serology tests. OLder animals which have been vaccinated can have persistant high antibody titres. Also vaccinated animals in contact with infected animal can exhibit a false anamnestic responce. If such tests are used to identify infected animals for slaughter the number of animals culled may be unacceptible. Blasco (2010) suggests 2 possible stategies:

• For the first 2 years after cessation of mass vaccination, compulsory vaccination of all young animal and the implimentation of animal movemernt controls. Followed after 2 years by a test and slaughter programme based on positive results from CFT and RBPT tests. Such tests have to be repeated as necessary until all animals in flocks are negative.
• This procedure involves re-testing animals 6 to 12 weeks after completion of mass vaccination with CFT and/or RBPT to identify seropositive animals and then further testing such using a gel precipitation test using Native Hapten (NT) antigen, Diaz (1979) and (1999) which can deferrentiate infected animal from others with positive serology antibody tests to an acceptible level. Animals positive to this test are slaughter. The procedure should be repeated until NH tests readings are repeatedly negative.

Surveillance. After eradication of brucellosis to satisfactorily low levels it is essential that a comprehensive surveillance system is put into operation based on regular serology screening using RBPT or iELISA om a representitive sample of the animal population. Testing should take place at least once a year.

Vaccination of Animals.

Huddleson (1942) reviewed the early history of the use of killed and live Brucella vaccines. He concluded that killed Brucella vaccines were of little use. Live Brucella vaccines could reduce the level of abortion but could slso infect of the animals. The main aim of early workers was to reduce the rate of abortion of calves which was of economic importance, rather than cure the disease. Huddleson is critical of these early studies which are generally poorly conducted and did not distinguish between vaccine treatment of infected and non-infected animals. He concluded that in general they did more harm than good.

It has been shown by experimentation that virulant 'Brucella can be grown and maintained in tissue monocyte cultures and Stinebring (1959) showed that virulant B abortus could be transfered fron tissue sample to tissue sample of monocytes of guinea pigs, but in rats resistant to the Brucella this did not occur. The amount of invasion and rate of multification of Brucella seemed to relate to the host's susceptibility and resistance. Braun (1958)) and Pomales-Lebron (1957) showed that in guinea pigs previously inoculated with live Brucella intercellular multiplication was markedly reduced but Braun (1962) showed that if killed Brucella was used multiplication inside monocytes was only slightly reduced. This depression of growth of Brucella in monocytes is part of the basis of successful vaccination.

General experience shows that vaccination can be used to reduce the incidence of brucellosis but not eradicate it. The successful vaccines are of the live attenuateed type. There are safety issues associated with these vaccines because it has been shown that vaccinated anumals can develop chronic infections and the live vaccine can be secreted in milk. Some related incidental of vaccine induced infections of humans have been recorded. Accidental injections mainly involving veterinarians using of the B abortus s19 vaccine have shown that it can cause acute and chronic illness in humans. The vaccines have limitations. McDiarmid (1960) and Manthei (1959) reports that if a vaccinated animal is subject to a high dose of a more virulant strain of Brucella then it may become infected, The extent of residual infection in cattle due to vaccines cannot be easily determined because of the limitations of the serology and bacteriological tests.

The main vaccines used and tried in animals are:

• Killed Brucella vaccines and related products.
• B abortus strain 19 (s19).
• B mellitensis strain Rev 1.
• B abortus rough strain 45/20.
• B abortus rough strain RB51.

Self Immunisation.

It has been suggested by early workers, Bang (1897) and Connaway (1929) that in large herds of cattle with persistant infection the disease becomes less virulent over time and that a herd immunity can develop. Cows only abort onceor twice and then appear well. A modern interpetation of this is that some cattle have more resistance to B abortus and do not develop brucellosis. Others develop a chronic illness (latent carriers) without any obvious signs of illness but remain capable of infecting other cattle and humans.

Live Virulent Brucella Vaccines.

Bang (1906) , McFadyean (1914) carried out trials with live B abortus vaccine either by intravenously or subcutaneousin injections in heifers, goats and sheep prior to breeding. Althogh abortion was prevented in some animals many of the animals developed permanent udder infections and their milk was infected. Trails of live vaccine in calves were carried out by Buck (1930) and Cotton (1934) this reduced abortio levels but the animals became perminantly infected. These type of vaccines were banned, Fitch (1933)

Killed Brucella Vaccines.

A great many attempts have been to test and use vaccines based on killed Brucella. Vaccines based on antigenic fractions including LPS and outer membrane proteins. In gerneral this work has been unsuccessful.

B abortus s19

Huddleson (1922) and Giltner (1929) developed a live culture of B abortus of low virilence and considered that it could be injected into cattle without harmful effects and reduce the level of abortion. However it proved to be unstable.

s19 is a stable attenuated strain of B abortus which was isolated by chance mutation after a virulant strain derived from the milk of a Jersey cow was was left exposed to the light at room temperature in a laboratory for over a year, Buck (1930) . B abortus strain 19 (s19) was developed by Cotton and Buck (1933) and (1933) as a vaccine to reduce the level of abortion in cattle and then subsequently to reduce the incidence of brucellosis. It was considered successful because it suppressed abortion and other Brucella shedding mechanisms. s19 has become the vaccine of choice because of its stable charactor, its low virulence, and its satisfactory level of immunity against B abortus infection. s19 is a smooth strain of B abortus. Smooth refers to the appearance of the bacteria under the microscope due to the presence of O-polysaccharide (O-PS) in its surface layer. This O-PS stimulates a humoral immunoglobulin response. This causes problems because the serological response to s19 is very similar to that from active disease. Therefore the use of s19 is incompatible with the normal test and slaughter regeme which is used to eliminate brucellosis from herds of cattle. For this reason it is not usually used in adult cows, however it can be used in calves because it has been observed that in most cases of s19 vaccinated calves humoral antibodies disappear by their first calving. Restriction of the use of s19 to calves is the normal practice. s19 has been widely used for vaccination of cattle in many countries including the USA. It can be effective in reducing the level B abortus infection but on its own it cannot eliminate the disease. Trials of s19 against B melitensis in sheep and goats have shown it to be ineffective.

In some cattle can become persistently infected with s19 and shed it in their milk. They can also develop signs of brucellosis illness including joint lameness. Bracewell (1980) has suggested that rather than this being causeddirectly by active infection by s19 this could be due to Brucella antigen-containing immune complexes present in the affected joint.

Manthei (1968) notes that s19 causes a transient infection in most cattle irrespective of age. s19 has occasionally been isolated from the milk of vaccinated cattle but he found no evidence that s19 could spread from vaccinated to other cattle. Adult cattle which are vaccinated retain positive serology titres but calves under 8 months do not. He considers that in general vaccinated cattle retain immunity during their normal life span. Approximately 75% of vaccinated cattle have complete protection, 15% have partial protection and 10% have no protection. Re-vaccination does not significantly improve immunity. Manthei (1959) also studied the extent to which exposure to a more viulant strain of B abortus (2308) can overcome vaccinated immunity from s19. His results are shown in Table xx.

Viable cell dose	Vaccinated	Vaccinated	Unvaccinate	Unvaccinated
	Infection	Abortion	Infection	Abortion
	%	%	%	%
350,000	0	0	78	56
700,000	26	20	87	74
15,000,000	61	47	100	100
75,000,000	61	45	100	97
Natural	20	20	56	22

Mohler (1941) vaccinated 70 calves with s19 and compared them with 73 unvaccinated controls. 8 vaccinated calves and 57 unvaccinated calves bacame infected of the latter 55 aborted. Thomsen (1939) evaluated 6 herds of cattle in Denmark of which 66.3% were vaccinated with s19. Of 266 vaccinated animals 13 aborted whereas of 135 controls 34 aborted. Mohler notes that at first B abortus s19 was used for adult cattle as well as calves but later it was restricted to calves. The main reason for this was that in older cattle the resultant persistent antibody response made it impossible to differentiate between vaccinated and diseased cattle. It is now restricted to calves under 8 months in which case the serology response usually disappears after about 6 months. Stableforth (1951) considered that calves vaccinated with s19 between 4 and 8 months gradually loose the immunity to natural infection by the 3rd or 4th calving.

Manthei (1964 notes that both calves vaccinated with s19 and naturally infected produce both IgM and IgG, then called 16-19S and 7S repectively based on separation by ultracentrigation. IgM is produced at 5-7 days after vaccination maximising at 13 to 21 days. IgG is detected at 14 to 21 days and maximises at 38-42 days. IgM disappears more quickly than IgG. IgM levels were higher and persisted longer in infected than vaccinated cows.

B melitensis Rev 1.

Rev 1 is an antenuated B melitensis strain which was introduced by Elberg (1953) , (1955) and (1957) as a vaccine to treat B melitensis in sheep and goats for which s19 is not very effective. It can also be used to combat B abortus in cattle and possibly gives better results than s19, Horwell (1971) . It has also been tried against B ovis in sheep particularly rams, Lantier (1985) . It was developed from a stretomycin depentant mutant of a virulent B melitensis isolate. Rev 1 is a smooth variant and therefore produces a positive serology response with the usual problems of distinguishing this response from that of active disease. For this reason its use is usually restricted to kids and lambs in the same ways that s19 is restricted to calves. If administrated during pregnancy large doses can induce abortion, Needman (1968) and Alton (1967) . A serious problen with Rev 1 is that it is resistant to streptomicin, Moriyón (2004) which is an important antibody in the treatment of human brucellosis. Another problem is that it can revert to a more virulent form. See below. Nevertheless Rev 1 remains the vaccine od choice for sheep and goats afgainst B melitensis.

B abortus rough strain 45/20

The main problem with s19 is that it is a smooth variant with an associated serological antibody response with the associated difficulty of interpreting test results for infected animals. Therefore it is logical that an attenuated rough strain of B abortus would overcome the problem. A rough strain is so called because of its appearance under a microscope due to the absence of O-PS in the surface layer of the bacterium. 45/20 is a live attenuated B abortus vaccine developed by McEwen (1938) and (1940) , and Edwards (1945) . It was made from a virulant strain by 20 passages through guinea pigs and this caused mutation into a rough strain. Unfortunately it has been found that this live attenuated strain can revert to a more virulent smooth form in use. It has also been found that the rough 45/20 colonies retain some degree of O-PS side chain charactoristics, Schurig (1984) and (1991). In general the performance of this vaccine is similarly or slightly inferior to s19 but its main problem was that it lacked complete stability. For this reason its use has been discontinued.

B abortus strain BR51

Although there were significant problems with vaccine 45/20 the related trials did show the benefits associated with a rough form Brucella vaccine and possible use in adult cattle. BR51 is a newer live attenuated rough strain of vaccine developed by Schurig (1991). RB51 is a riframpicin-resistant mutant derives by repeated in vitro passages of virulent strain 2308 of B abortus through a media containing riframpicin and penicillin. It has effectively no O-PS side chain and apparently this mutaion is genetic and partly due to a wboA transposon mutation. Because of the absence of O-PS there is no decernible serology antibody response even if used repeatedly. However Olsen (1996) indicates that there can be an anamnestic response. Nowithstanding this, it provides immune protection while avoiding a serological response and can therefore be used in adult cattle. RB51 has been shown to be a stable mutation by repeated in vivo passges through animals and also by vitro studies, Schurig (1991) and Colby (1997) . It is probable that the main action of RB521 is to induce a protective CMI response including enhancement of cytoxic T cells, Jiménez de Bagüés (1994) . Studies have shown that RB51 is of low persistence and clears quickly in experiments in animal subjects including; mice, guinea pigs, goates and cattle. The way in which RB51 has mutated is not clearly understood but may involve disruption of genes associated with O-PS side chain assembly and synthesis, Vemulapalli (1996) and (2000) . Early trials indicate that it has low virulance to humans in typical accidental inoculations. Cheville (1993) and (1996) considered that it provided similar protetive to s19 in cattle. BR51 has been approved for use in the USA and other countries in preference to s19. It can be used in calves (older than 4 months), pregnant cows (preferally with a reduced dose of 10⁹ organisms). Lord (1998) has shown that RB51 can be effective in pigs against B suis and other work indicates that it also works for B melitensis in goats. However these results have been disputed. In a review of the usefulness of RB51 Moriyón (2004) noted that controlled experiments in calves have demonstrated that reduced doses of RB51 is ineffective, full doses only partially effective and less effective than s19 against severe challenges. He also concluded that it was also ineffective in sheep and possibly goats. Moriyón (2004) also considers that a seriuos defect of RB51 is its resistance to riframpicin which is an antibody which is important in the treatment of human brucellosis. There is the possibility that these riframpicin resistant genes could be of the release to the environment. This particularly applies to attempts to vaccinate wild animals.

The development and study of RB51 in relation to modern knowledge of genetics has lead to greater understanding of the mechanisms which underly the mutations in rough attenuated strains of Brucella. On the basis of this knowledge newer vaccines are being developed however none of these has yet reached full fruition. There is however a good possibility that better rough attenuated Brucella vaccines can be produced in the future.

Persistence of Live attenuated Brucella Vaccines.

One of the most important considerations in the use of live attenuated vaccines is whether they can cause persistant infection in animals and infect humans. Such problems have been encountered in all the vaccines considered here. The question arises as to the level of infection in humans by such vaccines. Any illness might be anticipated to be mild and could be beyond the diagnostic skills of general medical practicioners but further could develop into long term chronic illness which would be virtually undiagnostable.

Persistence of B abortus strain 19.

The following are examples persisence problems associated with s19 vaccine and also safety issues in relation to humans:

• There are numerous examples of veterinarians being infected with s19 mainly due to accidental injections.
• Examples of persistance of s19 in vaccinated cattle has been provided by Nicoletti (1966) , (1978) and (1981).
• Meyer (1967) , (1969) and (1985) examined tissue and fluid samples from cattle and humans and found that Brucella abortus strain 19 persisted in vaccinated cattle and showed that this could be recovered of by conventional bacteriologic methods.
• Ewalt (1979) examined tissue samples from 104 cows in a herd and isolated B abortus strain 19 from 23 of them.
• Thomas (1981) recovered 34 cultures from cattle in the UK identified unequivocally as s19 by metabolism and guinea pig virulence tests. 11 samples were recovered from abortion material, 10 samples were from calves which died from a hypersensitivity reaction within 24 hours of S19 vaccination and the remainder were from milk or internal organs. It is suggested that s19 results in persistent systemic infection in a small proportionof vaccinated calves.
• Roerink (1989) indicates that calves vaccinated with B abortus strain 19 under the age of 8 months return to negative serology tests within 6 months but a high percentage of older animals have long lasting titres which are often life long and suggestive of a chronic infection.
• Corbel (1989) describes 30 cows with symptoms of lameness due to chronic granulomous arttopathy and persistent serology for Brucella ascribed to s19. Intra-articular injection of rabbits and a calf with immune complexes isolated from the synovial fluid of a field case produced clinical and histological signs of arthropathy accompanied by the development of rheumatoid factor.

Persistence of B melitensis Rev 1.

The following are examples associated with persistence of Rev 1 in animals and safety issues for humans:

• Alton (1988) reviewed data for Rev 1 accummulated over 10 years and concluded that it was infectious to humans.
• Pieterson (1988) and Pefanis (1988) and (1988) described B melitensis Rev. 1 mutant (FSA) as a mutant strain of Rev 1 which has been recognised n South Africa. FSA resembles Rev 1 in its reactions to penicillin and streptomycin but reacts closer to a field strain of B. melitensis as regards dye (thionine and basic fuchsin) sensitivity and colony size. A flock of sheep infected with this mutant were examined. 21 of 62 sheep were bacteriologically positive. The organism could be best isolated from the udder, supramammary lymphnodes and uterus. Vertical transmission to 24 lambs was not shown.
• Hunter (1989) examined the mutant strain from Rev 1(FSA). They demonstrated increased virulance showen by a slower clearance rate in guinea-pigsand mice and sheep with a Rev. I mutant strain (FSA) and they showed horizontal transmission in 2 sheep.
• Banai (1990) demonstrated the occurance of a new strain of Rev 1 vaccine in Isreal. The strain was isolated from two brothers in a family that reared sheep and from the milk of one of their sheep. THey concluded that the prevalence of such strains in two distinct geographical zones in Israel provide epidemiological support for the notion that a new variant has been identified.
• Lucero (2006) describe a new strain of B melitensis in Argentina similar to B melitensis Rev 1 vaccine although the vaccine has never been authorised there. 9 such isolates were obtained from humans. The variant has not been identified in animals.

THe problems associated with the virulance of Rev 1 has been oartially solved by standardising on Elberg 101 sRev 1.The best solution is a new better vaccine.

Persistence of RB51.

Te following examples relate to problems of persistence of RB51 and its safety in humans:

• On the basis of vaccination trials with RB51 using heifer calves Olsen (1999) concluded that RB51 could persist in some cattle into adulthood but the incidence and significance of this persistence remains unknown.
• Palmer (1996) and found that RB51 has tropism for the bison placenta, and can cause placentitis, which induces abortion in pregnant bison. The vaccine dose used was similar to that being tested in cattle, but it did not clear as quickly as in cattle. RB51 was also recovered from fetal lymph nodes.
• Palmer (1996) carried out vaccination trials with s19 and RB51 and concluded that RB51 was less abortifacient than s19 but it can infect the bovine placenta, mammary gland, and fetus, can induce placentitis, and, in some cases, can lead to preterm expulsion of the fetus.
• The first case of infection of a human with RB51 has been described by Villarroel1 (2000) The cause of clinical brucellosis in a Chilean verterinarian was shown to be RB51.

Vaccination of Wild Animals against Brucellosis.

Brucellosis is a disease of both domestic and wild animals. Brucellosis can be eliminated in domestic animals by eradication programmes involving test and slaughter, and vaccination but the disease can be re-introduced to domestic animals by contact with infected wild animals. Wild animals may be naturally infected with Brucella or they may become infected by contact with infected domestic animals. Some species of wild animals infected from domestic animals become self-sustaining hosts for the disease such as B abortus in American bison. Therefore in some developed countries the need to control or limit bruucellosis in wild animals is now a priority. For example in North America, elk and bison are infected with B abortus from cattle, and carabou/reindeer are naturally infected with B suis and ferile swine are infecrted with B suis derived from domestic pigs. B canis is also present in wild canidae. Eliminating brucellosis in wildlife presents a formidable problem. Test and slaughter or direct vaccination are unviable and impractical. One idea is to administer vaccinate orally by lacing food pellets with vaccine and distribute these in feeding grounds. Some success related to this idea has been described by Stevens (1996) in mice and Elzer (1998) in cattle. These experiments suggested that a considerable of immunity could be devivered in this way. Trials in the use of vaccines in wild animals has been reviewed by Davis (2002) and Godfroid (2002) . In general attempts to vaccinate wild animals have not been sucessful. Davis (2002) notes that s19 is not safe and does not protect American bison. Similarly it has been reported by Olsen (1997) that RB51 causes leisions similar to s19 in bison and Palmer (1996) found that it also caused placentitis and abortion in bison although these findings are at variance with studies by Davis (2002) and Olsen (1999) . Overall the effectiveness of RB51 in bison appears to be very limited. Similarily disappointing results in studies of vaccination of elk have been described by Cook (2002) and Kreeger 2002) . Overall thes results are discouraging and suggest this approach may not work.

Davis (1990) considered that in general there is no example of eradication of brucellosis in domestic livestock without asccompanying elimation of associated infected wildlife and that there will remasin a residual reservoir in wildlife. The only hopee for control and eradication in wlidlife is vaccination.

Vaccines of Humans against Brucellosis.

At the present time the control of brucellosis in humans depends on elimination of the disease in domestic animals, strict safeguards and hygene in handling of infected animals and their produces and pasteurisation or sterilisation of milk and milk products. These measures have limited the exposure of humans to Brucella infection in developed countries and therefore the rate of disease but are not effective un underdeveloped countries. It has been postulated that vaccines might be the best option to control brucellos in humans in such underdeveloped countries, Schurig (2002) . It has also been suggested that a human vaccine against brucellosis is required to counter the use of Brucella as a biological warfare agent, Kaufmenn (1997) . B suis was among the earliest agents investigated and developed as a bioterrorism weapon in the United States offensive bioterrorism program in the 1950s based on their contagiousb and ease of dispersal and their impact on human and animal health. Rotz (2002) considered that B. abortus, B. melitensis, and B. suis could be used as bioterrorism agents.A vaccine could also be of use to protect traveller.

Killed Brucella and Brucella antigens as vaccines. Until the introducion of antibiotics the use of intradermal vaccines was probably the most widespread treatment for brucellosis and particularly chronic brucellosis. The treatments were based on the use of killed Brucella or extacts therefrom such as the Brucella protein extract, brucellergen. There is reaonable evidence that repeated treatments of this type could relieve symptoms and possibly even effect complete cure. One problem was that such treatments were never standardised. Another problem was that repeat use of these treatments could cause severe local and sytematic reactions. These procedures are discussed under treatments. A reasonable idea us that these treatments stimulated a TH1 immune response including revitalising cytokine activity including INF-γ and TNF-α but they probably did not induce long term immunity.

Some limited work is continuing. Jacues (1991) prepared a Brucella conjugate vaccine by covalently coupling the O-polysaccharide (OPS) obtained from Brucella melitensis 16M, to bovine serum albumin (BSA). This conjugate was non-toxic for mice and considerable protection was confered against challenge with the virulent strain B. melitensis H38. Oliveira (1996) prepared a recombinant Brucella abortus L7/L12 ribosomal protein fused to maltose binding protein and used it to vaccinate mice. They considered it provided a significant degree of protection based om T-cell activation. Rosinha (2002) attempted to identify antigens of Brucella which stimulate a protective T-cell-mediated immune response and proposed glyceraldehyde-3-phosphate dehydrogenase. There is the possibility that such vaccines could also be used as a form of treatment such as in chronic brucellosis where antibiotic treatment is inadequate. In general although this type of work is interesting there is no likelihood of a human vaccine of this type in the near future.

Live Attentuated Brucella Vaccines.

Vershilova (1961) described trails of the attenuated vaccine s19-BA (A biovar of s19 developed by these workers) in USSR. B melitensis associated with the husbandry and slaugher of sheep and goats is the main form of the disease in the USSR (~90%). The trial were carried out between 1852-58. 3 million persons were involved and a reduction of 60% in the incidence of brucellosis is claimed. However despite claiming theat the vaccine was innocuous it is noted that a considerable proportion had ongoing symptoms of brucellosis after vaccination. It is difficult to imagine that these trials in a totalitarian state were in anyway ethical.

Spink (1962) carried out trials of 2 vaccines on volanteers at Minnisota State Prison. 16 were vaccinated with s19-BA and 16 with Rev.1. 2 of the inmates who received s19-BA and 11 of those who received Rev.1 developed accure brucellosisSAt titres developed in all cases and were higher with Rev.1. All apparently recovered after receiving antibiotics.

Some trials of live attenuated vaccine have taken place in European an Asian communist countries. Kolar (1989) describes the use of strain 19-BA (a derivative of s19) as an interdermal vaccine in the USSR. It has been described as providing a limited period of moderate protection. Re-immunisation was necessary with possible severe hypersensitive reactions. Lu (1989) describes the use of 2 attenuated Brucella vaccines: 84-C and 104-M used in intradermally or aerosols forms in China. These trails are described as providing effective protection but provoking severe reactions in some individuals. In either case it is not known if the indiviuals in these trails suffered from long term ongoing symptoms caused by these treatments. Considering that brucellosis infections often become chronic and can resist antibody treatment and can last a lifetime these trial would appear to be reckless and unethical. These vaccination procedures are apparently not now in use.

Schurig (2002) has discussed the possible use of live attenuated Brucella vaccines to protect humans. He notes the low pathogenicity of RB51 in humans but rules it out because of riframpicin resistance. In fact a case of RB51 infection in a veterinarian has since been described. He also suggests that a vaccine based on a rfbK mutant (manB_coremutant) of B melitensis could be considered as a starting point for a human vaccine. He notes the difficulty of testing such vaccines and suggests the use of animal testing including primates. The criteria for a suitably attenuated live vaccine is that it clears easily from the host however we must consider the possibility that there can be some ongoing low grade infection caused by these vaccines. Despite having negative serology tests for brucellosis human suffersa of chronic brucellosis describe ongoing symptoms such as tireness, fategabiliy, and muscle and bone pain. In some such cases their illness has been proved after a long period of illness. It is very likely that animals with long term brucellosis suffer the same type of symptoms but this cannot be proved even in primates. There is also the possibility that vaccinated animals suffer in the same way. Antibiotics do not work in all cases of brucellosis so humans injected with a live Brucella vaccine could face lifelong infections. There are also unexpected factors for example RB51 which is considered relatively safe in cattle is surprisingly much more virulant in American bison. It is difficult not to conclude that any trials of a live attenuated Brucella vaccine in humans would probably be unethical.

M1

Nicoletti (1980) notes that calves may acquire infection in utero or by ingestion of contaminated milk. A significant number of heifer calves infected in early life are serologically negative to tests for Brucella but abort or have an infected calf at first pregnancy. Latent carriers such as these are difficult to diagnose and make eliminatio of brucellosis from herds very difficult.

Nicoletti (1980) notes that the incubation period of brucellosis in cattle, that is the period between exposure to infection and the development of evidence of active infection varies widely. THe fectors involved are gestation, exposure, age, vaccination, and other host-related resistive influences.

Nicoletti (1980) notes that s19 rarely causes permanent infection in vaccinated cattle. It is suggested that there is no pathogenicity for humans except when injected. Vaccination is of no value in cattle which are already incubating brucellosis

Davis (1990) considered that in general there is no example of eradication of brucellosis in domestic livestock without asccompanying elimation of associated infected wildlife and that there will remasin a residual reservoir in wildlife. The only hopee for control and eradication in wlidlife is vaccination.

Nelson (1977) notes that in cattle there is no correlation between desease resistance and serological respomse. The resistance acquired from vaccination is relative immunity only without complete protection.

Morgan (1969) In the UK brucellosis in animals has involved only B abortus. in cattle B abortus has a long incubation period from 6 weeks to six to eight months or more.

Whole herd vaccination. Whole herd vaccinatio of cattle was first tried with s19 vaccine in the 1930s onwards and showed considerable benefits, Haring (1943) . The main problen with whole herd vaccination is that it causes abortion in cows and gives persistent SAT titres similar to infected animals. Another problem was that s19 caused udder infections in cows which were considered a possible danger to human health. Therefore vaccination of calves where these problems were more limited was considered a better option. Use of mass vaccination was reconsidered in the 1970s to deal with large problem herds. If brucellosis is endemic in a country or even a large herd it can reduce the level of infection to low levels. Cattle which are already infected with brucellosis are not cured or affected. Campbell (1976) carried out trials with s19 vaccine in cattle and used newly available serology test to attempt to differentiate between infected and vaccinated animal. monator herds. Nicoletti (1976) and (1978) and Crawford (1978) showed that whole herd vaccination could be carried out with a lower dose of vaccine with reduced adverse effects. The use of mss vaccinations of cattle in undeveloped counrties with limited vetinerian support with reduced vaccine doses could be considered an option to reduce disease levels.

Pathogenesis of Brucella in Humans and Animal.

According to Enright (1990) the manifestations of brucellosis are very similar in domestic animals irrestective of the species of Brucella. In general brucellosis is a chronic infectious disease which which usually persists for the animals life time.

Enright (1990) indicates that experimental work with differing amounts and strains of B abortus have shown that a significant proportion of cattle can innately resist the infection.

He notes that much work with cattle has centred on the disease in relation to genital brucella infections but less to the general nature of the disease in theacute and subacute phases .

The stages of infection by Brucella have been studied in animals particularly cattle and reviewed by Enright (1990) . The stages are:

• Mucosal Infection. The initial infection is an interaction with a mucosal surface. This includes: conjunctival, oral, pharyngeal, intestinal. respitatory, vaginal and preputal mucosa. The bacteria must penetrate the epitheial lining of the mucous menbrane. Invasion of the epithelium of the conjunctiva or lachrymal ducts causes an acute inflammatory reaction in the submucosal tissue and Brucella can be recovered from the area for up to 8 days. This is associated with the presence of monocytes, granulocytes, neutroplils and eosinophils.
• Invasion of local lymph nodes. Brucella which has escaped the submuccosal defences are distributed to local lymph nodes probably carried by granulocytes, monocytes and macrophages. In studies by Payne (1959) and others it has been shown that after inoculation of the conjunctiva with B abortus the parotid and mandibular lymph node became infected within 2 to 8 days. However after oral mucosa challenge infection of the suprapharyngeal lymph node which was more distant from the site of challenge took up to 4 weeks. Infected lymph nodes were often swolllen due to hyperplasia and infiltration of inflammatory cells. Infection was associated with granulotous inflammatory foci and extramedullary hematopoiesis foci.
• Bacteremia and secondary location. Carmichael (1970) considers that if the immune response in the mucosal tissue and related draining lymph nodes fails to erradicate the Brucella then there is the likelihood the it will escape into the blood stream presumably protected intracellularly inside macrophages and neutrophiles from humoral immune defenses, Bacteremia may be prolonged. Payne (1959) has shown that in cattle the spleen is colonised in about 15 days and distal lymph nodes at about 22-29 days. During bacteremia a wide range of tissue can be coloniesd. In cattle the main sites are: lymphoid tissue, mammary gland, and reproductive tract.Inferction can localises most frequently in the bones, joints, eye, testes, epididymis and occasionally the brain.

B abortus in Cattle.

Cattle are a natural host for B abortus but they can become infected with B suis (Huddleston (1929) ) and B melitensis. Just like humans cattle can have brucellosis in acute, subacute and chronic forms. An epidemic of abortions in a cattle herd is usually is indicative of brucellosis and it is therefore often referred to as "contagious abortion". This can often result from the introduction of one infected cow. The stages of infection in cattle are similar to those in humans:

• Infection at a primary site
• Parasitation of phagecytes
• blood infection
• Spread to other organs: spleen, liver, bone, marrow, lymph nodes, kidneys and uterus.

During pregnancy B abortus becomes concentrated in the foetal tissue. It was originally considered by Smith (1961) , (1962) , Pearce (1962) , Williams (1962) that erythritol plays a key role in this localisation of B abortus in foetal tissue of pregnant cows. Similar results have been found by Williams (1963) Keppie (1965) for B melitensis and B suis infections in pregnant sheep. However Berman (1977) noted that some strains of B abortus are not stimulated by erythritol but still cauuse genital infections and abortions. Also Bosseray (1980) has found that rodents which are without detectable levels of erythritol which were infected with B abortus also suffered placental localization and abortions. More recent work by Anderson (1986) suggests that B abortus proliferates readily within the rough endoplasmic reciculum of the chorionic trophoblasts of pregnant goats. It has also been shown by Detilleux (1988) that B abortus grows within the rough endoplasmic reticulum of chick embryos. Payne (1959) , Smith (1919) , Pennell (1937) and Moreno (1981) showed that the likely cause of fetal stress and abortion caused by Brucellais the associated Brucella LPS endoxin which may interfer with progesterone producion. They showed that LPS could cause abortions in cattle, guinea pigs and mice. The aborted calf and placenta are usually highly contaminated with Brucella and are highly infectious to other animals and man. The milk of such cows is also infected.

Thorne (1982) suggests that more serology tests have been developed for brucellosis in animals that for any other disease. This is because of the importance of animals but mainly a reflection of the limitations of the tests. No test is ideal. The main limitation of all tests is the occurance of serologically non-reactive animals (latent carriers). Cheville (1998) suggests that because Brucella is only present in small numbers in lymphoid tissue this is an inactive state that fails to stimulate sufficient antibodies to produce detectable serology response.

Brucellosis will spread rapidly through a herd previously free of the disease. Cunningham (1977) considered that exposure of pregnant cows nearly always lead to abortion. Other workers. Stableforth (1959) , Huddleson (1943) , Plenderleith (1970) and have reported rates of abotion in pregnant cows as: 40-50% up to 100%; 70%; 73%; 85% respectively.

Jubb (1970) notes that bursitis. hygroma and secondary infections of hygromata are fairly common in cattle suffering from brucellosis.

Richardson (1959) and notes that intracellular brucellae are a common finding in brucellosis. In pregnant infected cow uterine and fetal epithelial cells completely fill with Brucella is chaacristic of the disease. During these experiments cells from tissue samples from slaughtered cattle were grown in vitro and exposed tham to B abortus and the growth measured by colony counting. The 'B abortus invaded and multiplied rapidly within the cells. Multiplication of B abortus was curtailed by cell rupture but this lead to infection of other adjacent cells. The tissue types tested were uterine mucosa, testes, spleen, bone marrow, and lung of adult cattle and skin and kidney from fetuses.

Payne (1959) in a study of cattle showed that B abortus first infects the lymphoid tissue and then spreads to the uterus. It can be isolated from the spleen soon after infection but considered this not to be a preferred location. Doyle (1935) noted that in naturally infected cattle the spleen was infected in 10% of cases.

A

Nicoletti (1980) notes that early optimism in the USA that bovine brucellosis could be eliminated have been tempered by the complexity of host parasite relationships and farming practices. Elimination remains a long teerm objective.

According to Nicoletti (1980) the most likely source of B abortus in cattle is from discharges from infected heifers and cows mostly by ingestion, but other possible methods of transmission are inhalition, conjunctival exposure, direct skin contact, intramammary inoculation, reproduction tract, and congenital. The oral route of infection is usual for both male and female cattle often at an early age, Rankin (1965) . Phillippon (1970) showed that B abortus was present in the vaginal discharge of infected non-pregnant heifers. After parturation or abortion, vaginal excretion of Brucella is generally continuous for about 15 days then diminishes and becomes intermittent. Natural infection of ctattle with other Brucella species is unusual. Meyer (1966) found that cattle infected with B melitensis can become carriers and excrete the organism in milk.

Nicoletti (1980) notes that there are considerable differences in the virulence between Brucella species and also biotypes. This can to some extent be demonstrated by in vitro studies but there is no totally satisfactory method of measuring virulence. Once infection has occurred there is the likelihood of Brucella becoming localised in the reproductive organs, udder, and supramammary lymph nodes in cows and in the testicles and related sex glands in bulls. During pregnancy the uterus is infected with rapid multiplication of Brucella during the second and third trimester possibly leading to abortion. The cow usually becomes non infectious until the next pregnancy when there is a similar cycle but the degree of genital tract infection is usually lower Manthei (1950) and 7 0f 8 such cows had B abortus in their colostrum. Brinley-Morgan (1960) demonstated that the excreation of B abortus in the milk of infected cows was intermittent and that the concentration of bacteria varied greatly.

Just like female cattle bulls can be infected from an early age, Rankin (1965) . In bulls B aborts can be isolated from the testes, epididymides, seminal vesticles and ampullae of the ductus and this can affect semen quality but there may be no clinical signs of infection, Lambert (1963). The disease is then likely to become chronic and the excretion of B abortus intermittent, Manthei (1950) . Such bulls are not usually infertile. Although they found no evidence of infection of cows by natural insemination they were able to infect 8 or 12 Brucella-free cows by artificial insemination using semen from an infected bull. Nicoletti (1980) nots thast in general natural or artificial infection usually persist indefinitely although possibly up to 15 percent may recover spontaneously.

Kellar (1976) reviewed the causes of brucellosis in Ontario and that the main cause of outbreaks or sustained infection in herds was:

• Proximity to other infected herds. This includes physical contact at fences and pasture conntact and possibly drinking contaminated water, Dopel (1968) .
• Introduction of cattle from othe herds.
• It was more difficlt to eradicate brucellosis from large herds than smaller ones and more difficult to control closed than open herds.
• Vaccnated herds had a lower incidence of infection but vaccination did not protect individual animals.

Christie (1969) noted that the trend towards larger herds results in greater likelihood of infection and greater prevalence of disease and Nicoletti (1975) found that this equated with greater dificulties in eliminating the disease. Nicoletti (1980) concludes that a a hed size of 250 cows the disease cannot be eliminated by test and slaughter with a prohibitetive level of reactors.

Bercovich (1998) notes that there are no pathognomic signs of infection in cattle however exanination of strained smears of chorionic epithelium for bacteria or examination of vaginal secretions may give some information.

B melitensis in Sheep and Goats.

B melitensis was the first form of the disease to be discovered and charactorised. This is well described by Alton (1990) The natural hosts for B melitensis are goats and sheep. There are 3 biovars of B melitensis. Biovar 1 occurs mainly in goats and biovars 2 and 3 in sheep. Some breeds of sheep are resistant. B melitensis seems to have developed in the mediterranian area through to eastern asia but has spread world wide through human activity however some countries are now clear because of eradication programmes. The course of the disease in sheep and goats is similar to that of B abortus in cattle. In sheep and goats B melitensis can enter the body via the mouth, nose, eyes or skin particularly through is abrasions. Factors affecting likelihood of infection include virulant of the organism, the number of bacteria and the animals health. As with cattle there is a tendency for localisation in fetal tissue and the udder. Abortion is an indicator of disease. Sheep usually abort only once but goats can abort more than once. Lambs or kids born to infected dams appear to recover after 2 months with negative serology but it is likely that at least some of them remain infected. Fetal material from abortions and virginal discharges are highly contaminated and a significant cause of transmission. In goats such discharges can continue for many months and recommence at later pregnancies. The effect in sheep is less pronounced. In general infection of milk in goats persists for two or more lactations but for only one lactationin sheep. Repeat abortions can occur in successive pregnancies . B melitensis may persist in a chronic form in goats for years. While Striedter (1938) showed that in sheep the disease appears to be more self limiting. Andreani (1967) found that male sheep and goats can become infected by physical contact with infected females even without sexual contact. The most common manifestation of B melitensis in male sheep and goats is orchitis which is more common in goats. Other reported manifestations are; hygromas and joint lesions, Renoux (1957) . There is no evidence that male sheep or goats spread the disease by sexual contact. Serology tests are often negative at any stage of the disease. It is considered by Kolar (1984) and Farina (1985) For this reason such tests can only be used to test herds not individuals.

Mollela (1963) notes that placental changes due to B melitensis in experimentally infected sheep are more necrotizing and less exudative than those caused by B abortus.

A main cause of the spread of the disease is due to the practice of bringing together large numbers of animals in close proximity such as transhumance.

Skimi (1981) , Renoux (1955) , Renoux (1956) and Biggi (1956) noted that aglutination titres in sheep and goats are often low or negative even in the presence of confirmed infection. This is particularly treue for non-pregnant and animals in endemic areas.

Renoux (1955) admistered virulant B melitensis via the conjunctiva to Swedish sheep and goats and found that a 50% infective dose was 1 x 10⁴ in goats and 4 x 10³ in sheep.

The strategy for eliminating B melitensis in small ruminants In small ruminants the initial step in brucellosis control is to vaccinate young animals (kept as replacements) with the B. melitensis Rev.1 vaccine. This approach is based on the hypothesis that the Rev.1 vaccine offers life-long immunity and that after implementing the vaccination program for 5-7 years, which is the productive life-span of sheep and goats, the whole population will be vaccinated and fully protected against brucellosis. This method is also recommended to minimize post vaccinal diagnostic problems and to prevent abortion (Blasco, 1997). Blasco, J. M. (1997): A review of the use Brucella meletinsis of Rev.1 vaccine in adult sheep and goats. Prev Vet Med. 31:275-81.

Horrocks (1905) found that B melitensis survived in dried soil at 18°C for 69 to 72 days but only about 7 days. Poliakov (1965) found that B melitensis could survive on wooden walls and on the floors of animal pens fot upto 4 months in cold conditions. Brucella is kill in hours by sunlight.

According to Alton (1990) the serological response using agglutation an CFT tests is variable in sheep and goats. Either a response may be seen after 2 - 4 weeks or it may be totally absent. Further peaks of serelogy titres may occur in pregnant sheep or goats at parturition and to a lesser extent due to invasion of the udder. Localisation in the lymph nodes may occur without serological response.

The tests used in sheep and goats for B metilensis are those developed for cattle.

Rev 1 offer considerable long lasting protection agasinst natural field strains of brucellosis. Vaccination can be used to suppress infection pressure and therefore reduce the disease in humans.

B melitensis is widely spread around the world. THe actual incidence of human brucellosis is unknown but in endemic areas may be as high >200 per 100,000 people. in endemic areas, Lopez-Merino (1909) .

In human brucellosis serological tests play a major role in the diagnosis when the agent cannot be detected by blood culture. Yet, the interpretation of these tests is often difficult, particularly in patients with chronic brucellosis, in re-infections and relapses.

Culture test. Brucellae are an aerobic although some strains require CO² to facilitate growth. THe best pH for growth is 6.8 (range 6.6 - 7.4) and the best temperature is 36-38°C (range 20-40°C). Biotin, thiamin and nicotinamide for growth. Colonies appear a honey colour in daylight and white when observed from above. Brucellae and particularly B melitensis has a tendency during culture and particularly sub-culture comvert to rough and intermediate mucoid forms. Rough strains are less virulent. More details on culture of B melitensis are provided by EC SANCO.C.2/AH/R23/2001 .

As in cattle the main route of infection in sheep and goats is through the mucous membranes of the oropharynx and upper respiratory system and conjunctiva. The mucous membranes of both the mail and female may be a secondary route. The bacteria migrates to the nearest lymph node. Whether brucelloosis becomes extablished depends on the ruminant species, infected, age, immune status of the host, pregnancy status, and the virulence and number of the invading Brucella. If the bacteria becomes established it then diseminates to the organs of the body with bacteremia which can last more than 2 months. B melitensis is likely to become locallised and signs of infection are abortion (main), orchitis, epididymitis, hygroma, arthritis, metritis, and subclinical mastitis. Followed by persistant infection of mammary glands and supramammary and genital lymph nodes with constant or intermittent shedding of the organisms in the milk and genital secretions, Fensterbank (1977) . In pregnant animals bacteremia leads to infection of the uterus. Generally animals abort once but but further infection of the uterus occurs at subsequent pregnancies with further shedding of Brucella. Females in an infected herd are less likely to abort. The udder is a main area for colonisation and Brucella is secreted in milk during more than one lactation cycle and is associated with reduced milk yields without signs of mastitis. In general secretion of B melitensis in milk is for a shorter period. In females the main sign of infection is abortion, the birth of weak offspring,and retained placenta. In males Brucella localises in the testes causing orthitis, epididymitis leading to infertility. Although semen is infected evidence suggests that transmission by mating is low. Lameness occasionally occurs in both sexes.These develop into chronic conditions. Macroscopic or microscopic leisions may be present in lymph nodes, particularly those of the head and reproductive organs, and infected organs. The disease is less severe in sheep than goats. There is sufficient Brucell in milk to facilitate transmission to offspring and humans.

In sheep and goats an serology response occurs at about 2 to 4 weeks but is variable and may be absent. IgM is produced first folloowed by IgG. In chronic infection IgG persists longer but both fall to low levels. In pregnant animals with uterus infections a persistant high serology titre can occur usually after abortion or parturition. The serological response to udder infection is usually limited or not present.In immature animla the serological response is transient or missing.

In sheep and goats the a long term persistent infection is usual. Non pregnant animals exposed to low doses of Brucella may develop a self limiting infection or they may become latent carriers.

B melitensis is enemic in the Mediteranian area, Central Asia as far as Mongolia and North Africa and Arabia.It is also present in South and Central America including: Mexico, Argentina and Peru. It also occurs in India and Africa.

As in cattle the main mode of transmission in sheep and goats is via foetal fluids and vaginal secretions from infected females associated with pregnancy. THe secretions and related risk in goats may contiue for months but is usually shorter in sheep.

Transmission from dam to offspring can occur via itra-utural infection but it is considered that transmission is mainly via infected collostrum and milk. It is possible that a self cur mechanism occurs in many infected lambs, Grillo (1997) , however latent carrier without syns of infection and detectable immune response are also a feature which pose a significant challenge during eradication. In goats infection can be of short duration or persist for years with re-occurance of B melitensis in milk during a series of lactations. In general sheep are more resistant.

Persistence in the environment may be considerable depending on favourable conditions; high humidity, pH>4, low temperature, and absence of direct sunlight. Milk and milk products are a vector for transmission to humans. Pasturisation kills Brucella but is not always used for some products such as soft cheeses. Some idea of the durability of Brucella in milk and milk products has been demonstrated by studies carried out by Davies (1973) , Nicoletti (1989) and Plommet (1988) . Some of the survival data is as follows: milk at 0°C, 18 months; cream at 4°C, 4 to 6 weeks; icecream at 0°C, 30 days; butter at 8°C, 142 days; cheese in general' 6 to 100 days; Feta cheese (B melitensis) 4 to 16 days; Pecorina (B melitensis). <90 days, Roquefort, 20 to 60 days; Camem bert (B abortus), <21 days; Cheddar (B abortus), 6 months; wheyat 5°C' >6 days. In general there is no safe time limit for cheese made from unpasruerised brucella infected milk.

v

B suis in Pigs.

B suis is the main cause of brucellosis in pigs and was first described by Traum (1914) and identified as a separate species by Huddleson (1929) . In nature B suis is a smooth type brucella. The disease in pigs has been reviewed by Alton (1990) . Pigs can become infected with biovars 1, 2, and 3 of B suis. Biovars 1 and 3 are highly pathogenic to humans and cause servere illness. Biovar 2 only occurs in Europe and apparently has low pathogenicity in humans. Desease in domestic pigs is caused by B suis biovars 1 and 3 and is widely distributed in the world but the prevalence is generally low however there are particular problems in South America and South East Asia. THe desease can spread to wild and ferile pigs. It can be spread amung pigs by general contact and ingestion of contaminated birth products and contact with contaminated vaginal discharges or semin. Piglets can be infected by their mothers either pre-natally or from mother's milk. Infection of sexual organs occurs in both sexes and abortion is the most common sign of the disease. Boars may be apparently symptomless but have Brucella present in its semen in which case the seminal vesicles, the prostrate, and the testes are affected sometimes with abscess formation and sclerosis. The main indicator that a herd is infected the sows aborting or producing still borne piglets which may occur at any stage of gestation. Infertility in both bores and sows is also common and bores may exhibit pain in the testicles and be reluctant to mate. Boars pass on the infection through infected semen. Infection in this way is more common than in cattle possibly in part because pigs deposite semin directly in the uterus. During the infection the lymph nodes, liver, spleen, bones, joints, tendons and tendons sheaths may be affected. Lesions include granulomata with hyperplasia, minute abcesses, occasional large abscesses sometimes with caseation. Other possible symptoms are: lameness, swollen joints, bursa, spinal problems possibly including paralysis and absecess formation. B suis in pigs can occur in acute or chronic forms. In the latter case it can be apparently symptomless, have negative Brucella antibody serology but positive culture tests. B suis can be transmitted from pigs to cattle and to sheep and goats and to dogs probably mainly via raw meat. The best method of dealing with B suis is total eradication of herds.

According to Jubb (1981) B suis tends to produce focal granulomatous reactions with significant coagulative necrosis. Such lesions are widespead but most frequently located in bones and joints. Enright (1990) considers that the disease is transmitted either most commonly by coitus or by ingestion. The semen of boars may be infected for extended periods and B suis may be shed fron the uterus of sows for up to 2.5 years. Following infection the disease spreds to the local lymph node. In the case of sexual transmission this is the inguinal and iliac lymph nodes and for ingestion the lymph nodes of the neck and head. Bacteremia may persist for years and give rise to secondary infected sites. According to Jubb (1985) these include: the male and female reproductive organs, the sketeton including joints, mammary glands, lymph nodes, spleen, kidney, bladder and brain. Typical lesions are granulomatous with sheets of histocytes and epithelioid cells with a central area of caseous or coagulative necrosis. The nectotic area is heavily infiltrated with neutrophils.

Bishop (1994) (1994) notes that in B suis in pigs after multiplication at the site of entry and bacteremia, the organism can localise in the skeletal system, joints, mammary glands, lymph nodes, spleen, liver, kidneys, urinary system and reproductive system.

The main cause of B suis in humans is infection during slaughter and meat processing. It can also be transmitted to cattle and thence to humans through milk. Such cases have been described by Beattie (1934) and Borts (1943) .

Ferris (1995) studied 221 pigs in 39 infected herds using bacteriology and serology antibody tests including: CFT, BPAT, Card test, fluorescence immunoassay and SAT. B suis was isolated from 46 animals of which 6 were negative for all serology tests, It was concluded that test and removal programes based on serology tests were unviable in pigs. THe CF test which is widely used for brucellosis screening in other animals (cattle, etc) is affected by interaction of pig and guinea pig complement is not effectiver in pigs. According to Rogers (1989) the CF test only identified 49% of culturally positive pigs.

B suis also occurs naturally in wild species includind: carabo / reindeer, hares and murine species.

Routine serology tests are unreliable in porcine brucellosis but can identify infected herds. Infected herds can also be identified using the DTH skin test.

Sows seem to recover at about 6 months but many remain chronically infected.

Wild boar (Sus scrofa) with B suis biovar 2 can transmit the disease to domestic pigs where contact is possible. Transmission to domestic animals is also possible from hares (Lepus capensis) infected with biovar 2, Godfroid (2002) . The disease occurs across mainland Europe. A charactoristic of B suis biovar 2 if the formation of miliary leasions and purulent nodules in any tissue. B suis biovar 4 affects both wild and domesticated reindeer and carabou (Rangifer tarandus) and is present throughout the artic area including Canada, Alaska and Siberia. According to OIE Terrestrial Manual (2009) these animals are readily susceptible to B suis biovar 4 and signs of the disease in carabou are: fever, physical depression, abortion, retained placentas, metritis, blood stained vaginal discharges, mastitis, bursitis and orchitis. The disease is readily transmited to humans by direct contact or via milk and raw or semi-cooked meat including bone marrow a local delicacy. The study of B suis biovar 4 and the development od appropriate serology test is in the early stages. There are probably more biovars of B suis to be discovered. One such is disease in small rodents reported in the former USSR Russia and called biovar 5 and thisa is possibly similar to the Brucella which affects marine animals. Others possible B suis variants have been reported in Australia and Kenya. All these new variants seem to have differing charactoristics.

In carabou lameness may be accompanied by buritis, tenosynovitis and/or hygromas.

In B suis infections in pigs the culture test is more reliable than antibody serology tests, . For 8 of the 46 culture-positive pigs, results of all 6 tests were negative. None of the available serology tests can identify B suis inferctions in individual pigs. In particular piglets ≤ 3 month can be infected but have only a limited serological response. There is also cross contamination with Y enterocolitica leading to false positive results. THe CF test which is widely used for other species is ineffective because of interaction between pig and guinea pig complement, Rogers (1989) who showed that CFT found only 49% of culure positive pigs.

B suis biovar 2 can spread from wild boar to domestic pig and then it acts similarly to biovars 1 and 3 spreading easily amongst animals.

Currently there are no serology antibody tests that can reliably identify B suis in pigs. This is particularly true of infected piglets of ≤3 months in which serological response is very limited.

Tests like iELISA, cELISA and FPA are only just being tried in porcine brucellosis. Tests such as BBAT and RBPT are possible screening tests.

THe DHT skin test is an alternative to serology tests. but is not reliable for dignosis of individual animals. Thest results for serology antibody tests and skin tests do not exactly coincide. So far the antigen used for the skin test is Brucellin Fraction F which does contain some S-LPS however sensitisation of animals is not known. The tests are best used together since this test is useful indifferentiating between response to B suis and Y enterocolitica.

B suis biovar 4 i s spread among animals by contact with aborted fetuses and birth products but other possible methods of spread are unknown.

In humans B suis biovars 1 and 3 cause a serious, debilitating and sometimes chronic illness affecting a wide range of organs. B suis biovar 4 is also a serious problem in humans. Cases of brucellosis related to B suis biovar 2 aa reported are rare.

According to CFSPH (2009) THe extent of infection in wild and feral swine in South Carolina ranges from 14% to 44%. At one location the infection rate was 28% in 1976, 18% in 1992, and 44% in 1999.

When B suis infects a new herd of pigs the rate of abortions and stillbirths and other signs of the disease such as lameness are high but once it has become epidemic these signs reduce and the manifestations are a slightly reduced farrowing rate, irregular estrus, and some degree of infertility.

The difficulty of diagnosing brucellosis can be exacerbated by other diseases which can cause abortions, orchitis, arthritis posterior paralysis and lameness including Aujeszky's disease, leptospirosis, erysipelas, salmonellosis, streptococcidiosis, swine fever and porcine parvovirus. Positive serology due to cross reaction with Y enterocolitica can cause problems.

Although eradication of infected herds is the preferred option in the cases of high value stock may be test and slaughter.

Brucells canis (B canis).

Introduction

Relatively speaking there has been little research into B canis and it is not well understood. Dogs and wild canidae (although there is little evidence) are considered to be the naturally hosts of B canis and it can spread zoonotically to humans and also to other animals. It is not known if any other animals species are also natural hosts for B canis. There is some evidence that B canis is present in other animals including: domestic cats, bobcats, coyotes and racoons but the research is very limited. B canis was first discovered and isolated fron beagle dogs by Carmichel (1968) and Moore (1967) and (1967) and . Spink (1968) studied an epidemic of B canis at a beagle kennels in the USA and noted that 89% of 142 adult dogs became infected. It was at first thought that B canis infections was limited to beagles it is now known that all breads of dog are susceptible. There have been relatively few epidemiological studies of B canis in dogs and these have been hampered by inadequate serology tests however there is every reason to consider that B canis in dogs is present worldwide (see later). It is known to be present: throughout North, Central and South America, including the USA, Argentina, Chile, Brazil, Peru, Mexico and Canada; Europe, including Germany, France, Britain, Czechosovakia, Poland, Italy and Spain; Africa including Tunisia, Nigeria, South Africa and Madagascar; and South East Asia including Turkey, India, Korea, Japan and Malaysia, Philippines, Taiwan, China but has not been found in Australia and New Zealand to date. Various surveys indicate that infection rates among stray dogs in various countries is about 6 to 10%. B canis is a rough variant of Brucella. This means that it contains rough lipopolysaccharide (R-LPS) in the outer cell wall rather than S-LPS as in smooth variants such as B abortus, B suis and B melitensis. Therefore the established serology tests such as SAT and RBPT developed to detect antibodies to S-LPS from smooth variants do not work for B canis infections since the antibodies are not present. Tests to detect antibodies to R-LPs have been developed using antigens derived from B canis, B ovis (another rough form of Brucella). In general little is known as to the extent of B canis infections in dogs and humans. Studies of other domestic animals infected with B abortus, B susis and B melitensis indicate that epidemics occur when large numbers of animals into close association. This also seems to be the case of B canis in dogs where epidemics of the disease occurs in breeding kennels when up to 75% of puppies can be lost. In such cases owners suffer severe financial loss. Stray dogs in urban areas can also show a significant level of infection and are a possible vector for transmission of the infection of to domestic pet dogs. The level of infection in pet dogs is usually lower than in that in strays. Dogs become infected during mating, by contact with infected birth products including vaginal discharges and from environmental contamination including the urine of male dogs. Symptoms of infection in dogs are variable and infected dogs may show no signs of illness at all. Signs of disease are discussed below. Some more typical indicators of infection are abortion and pup death and a swollen scrotum in males. Untreated dogs are likely to become chronically infected over a long period and this is related to localisation of B canis in lymph nodes and various organs of the body. In male dogs the prostate and epididymides is often affected. Chronically infected dogs can be asymptomatic with low or negative serology results and negative bacteriological culture results (blood) but nevertheless remain a risk to other dogs and humans. Treatment of dogs with antibiotics is only partially successful so that euthinasia is a better option with regard to human safety. The greatest risks to humans seems to be related to: handling infected tissue and culture samples in the laboratory; close association with infected breeding bitches and their pups (alive or aborted) and birth products (including vaginal secretions; contact with infected bitches at estrus; and contamination with the urine of infected male dogs. As with other forms of brucellosis the environment associated with an infected dog may be contaminated and is therefore a risk factor. In general pet dogs live in close association with humans therefore if a dog is infected with B canis there is significant risk of zoonotic transfer to the humans of the household. The extent to which humans are or can be infected by B canis is not known. Tests for B canis in humans are seldom carried out. Some authorities suggest that humans are relatively resistant to infection but this is not known for sure. Limited survey for B canis in general human populations using antibody serology tests have indicate a range of positive results of about 0.5 to 2% may be typical. There are only a limited number (about 50) of recorded cases of acute B canis infection in humans but it is likely that many go unrecorded. It is also likely that as with other forms of brucellosis many infections start insidiously without obvious signs of illness. Most recorded cases indicate a relatively mild infection but in others the disease causes serious illness. No cases of chronic brucellosis have been recorded in humans but this indicates little since tests for B canis are not done.

Signs and Symptoms of B canis in Dogs.

General signs of B canis infection in dogs can be fatigue, decreased appetite, weigh loss, lack of interest in breeding, reluctance to exercise, premature ageing, poor coat quality and unusual behaviour. However in many or most dogs these signs are absent or easily overlooked. . Fever is not generally present and it has been suggested that this is because this rough variant of Brucella produces relatively low levels of endotoxins, Spink (1968) . Although Jones (1976) has shown that heat killed whole B canis and B ovis have similar levels of lethality in mice to that from killed B abortus. This was also true of R-LPS derived from B ovis. Lameness, stiffness and back pain are also possible signs of infection and may indicate diskospondylitis, Kerwin (1992) . Lymphadenitis is common in infected dogs. The lymph nodes most likely to be swollen are the retropharyngeal which is associated with an oral route of infection and the iliac nodes which are associated with infection related to sexual contact. The mode of response is lymoid hyperplasia rather than inflamitary. An enlarged spleen and/or hepatomegaly may be present in some cases. More serious effects include: diskospondylitis; uveitis, Saegusa (1977) ; endophthalmitis, polygranulomatous dermatitis; endocarditis, Ying (1999) and meningoencephalitis with associated lesions, Purvis (1981) . Deaths are rare. It is very important to note that the disease is often insidious and a significant proportion of dogs which have been proved to be infected by the bacteriological culture tests show no signs or symptoms of the disease. The most obvious effects of the disease relates to the genito-urinary tract and lymphoid tissue. Male dogs suffer from epididymitis (increased size and hardening), prostratitis, testicular aphrophy, and sterility. Early in the disease the scrotum may be enlarged and painful and this are likely to be associated with constant licking resulting in dermatitis (possibly due to non-hemolytic staphylococci). The quality and quantity of ejaculate may be affect. Sperm may be abnormal or damaged or even absent, Carmichael (1990) . In females the main sign is abortion and still births, infertility and vaginal discharges which usually last up to 6 weeks. Abortion is likely to occur late in pregnancy at the 45-55 days range of a typical 63 day gestation and this is often followed by a bloody and/or gray-green vaginal discharge, Carmichael (1968) . 75% of infected bitches abort. Early embryo death can also occur with fetal resorption at about 10 to 20 days of gestation without any obvious signs. This may wrongly be ascribed to conceptual failure or infertility. Pups which are born at full term may be weak a die within the first month or can appear normal then develop the disease later, Nicoletti (1989) . B canis infections are not apparent in immature dogs or non-pregnant bitches. Clinical manifestations vary from dog to dog. Outbreaks may be readily identified in kennels such as breeding farms and hunting packs because of a sudden spate of aborting bitches.

Brennan (2008) describes a typical outbreak of B canis infection in a Saskatchewan kennel, Canada. 20 of 33 dogs showed positive results to the indirect fluorescence test (IFA) of which 13 (8 females and 5 males) were euthanized and studies further. Bacteriological culture tests were carried out on blood and tissue samples. Positive results for females were: blood, 4/8; lymph node, 4/8; spleen 6/8; tonsils (1 tested), 1/1; mammary gland, 1/7; ovary, 2/8; uterus, 5/8. Positive results for male dogs were: blood, 0/5; lymph node, 2/5; spleen, 0/5; epididymide, 1/5; prostate, 4/4. Tissue samples of 5 female and 5 male dogs were also subject to histological examination. Lymphatic tissues (Peyer’s patches, tonsil, and sub-lumbar lymph node) of both sexes revealed a consistent lymphocytic to lymphohistiocytic follicular hyperplasia of varying severity. For females there were also other lesions included lymphohistiocytic endometritis, occurring in the aborting female (1/3); lymphohistiocytic and neutrophilic endophthalmitis, which included a mild lymphoplasmacytic retinitis (1/5); and meningitis (4/5). Meningeal lesions were very mild and multifocal, consisting of rare perivascular clusters of a few lymphocytes and histiocytes. For male dogs lesions included chronic lymphohistiocytic orchitis, with testicular fibrosis and atrophy (1/5); lymphohistiocytic interstitial epididymitis (3/5), with evidence of intratubular neutrophils and spermatophagic macrophages (1/5); lymphohistiocytic funiculitis (spermatic cord) in 3/4 dogs; and lymphohistiocytic interstitial prostatitis in 4/4 dogs.

Chronic B canis in Dogs.

General information indicates that B canis is both an acute and a chronic disease in dogs even if antibiotic treatment is used. This is related to localisation of B canis in the lymph nodes, spleen, bone marrow, reproductive organs and other organs such as the eye, brain and liver. Chronic disease of the reproductive organs of the male is common. The disease progresses through swelling of the scrotum to hardening of the epididymis and degeneration, fibrosis and atrophy of the testes, Carmichel (1970) . This is associated with a decline in the amount of viable sperm resulting in infertility. Jubb (1985) has observed that localisation in the lumbar vertebra is common and sperm may be morphologically abnormal and of reduced viability. According to Carmichael (1990) dogs with chronic infection may have the following range of symptoms: sub-acute or chronic endomtritis, granulomatous prostatitis, testicular atrophy and fibrosis, hyaline thickening of the basement membrane of glomeruli with minimum cellular infiltration or proliferation, or meningitis diskospondylitis, anterior uveitis and ocular pathology including granulomatous iridocyclitis, exudative retinitis and cone edema. Females may develop reticular cell nodules in the endometrium. Usually infected bitches only abort once and subsequent litters are born alive and apparently healthy.

Wanke (2004) notes that in chronically infected dogs serology titres are often negative but at the same time bacteriological culture tests can proof the disease is present. Some dogs seem to clear brucellosis due to B canis within 1 year but others can become chronically infected and remain bacteremic with shedding of bacteria for at least 5 years. Wooley (1978) describe the isolation of B canis from a dog with negative serology results.

B canis infection can also become chronic in other animals. Egwu (1979) artificially infected rabbits with natural B canis s666 and an L-phase (penicillin induced) form from the same strain and in both cases the desases became chronic.

Other Infections with Overlapping Symptoms.

Other bacteria and viruses which can give rise to similar signs and symptoms are: beta-hemolytic streptococci, Escherichia coli, Mycoplasma, Ureaplasma, Streptomyces, Salmonella, Campylobacter, Canine Herpesvirus, Neospora Caninum and Toxoplasma gondii, Purswell (2004) .

Dogs can also be infected by B abortus, B suids and B melitensis with similar symptoms. These are usually caused by contact with infected stock such as cattle, sheep, goats and pigs or by eating infected raw meat (particularly birth products. Such dogs are likely to show similar symptoms to those with B canis.

Recovery of Dogs from B canis Infection.

It is suggested by Serikawa (1978) and Meyer (1969) that dogs recovery from B canis occur at 1 to 5 years as indicated by the bacteremia stopping and results from antibody serology test dropping to low levels. However this may not be true recovery but only indicate the start of a chronic phase of the disease. The bacteriological culture test (blood) does not find all cases of B canis infection and the natural decline in antibody levels at the end of bacteremia may have limited significance. . Also it is known that in some dogs positive blood cultures can occur in association with negative antibody serology results. Wooley (1978) describe such a case of the isolation of B canis from a dog with negative serology results. The main reason for this seems to be localisation of B canis within tissue.

Transmission of B canis in Dogs.

Likely ways of transmission of B canis from dog to dog from contact with foetal material associated with abortions and stillbirths, through sexual and by general contact with an infected animal. According to Carmichael (1990) dogs can transmit B canis through natural mating, oronasal contact, ingestion of contaminated tissue or fluid. Carmichael (1988) and Serikawa (1979) considered that infection could occur at any mucous membrane including the mouth, nose, eyes and genitalia and they considered that 10⁴ to 10⁶ organisms are sufficient to initiate infection. According to Hollet (2006) infection is likely to start the mucous membranes of the orinasal, the genitalia or the conjuctiva and travel to a local lymph node and start bacteremia within 30 days. Infection is likely to occur in the prostate, testicles and epididymides of male dogs and for females in the foetus, gravid uterus and placenta. B canis carried in the blood and lymph can seed in other organs such as: inter-vertebral disks, kidney and anterior uvea of the eye. Infection through skin abrasions is also possible. High levels of transmission is most likely where dogs are gathered together such as kennels. Stray dogs in urban areas can show high levels of infection and may spread the disease to pets. General surveys indicate that infection in pet dogs is lower but is worse in economically deprived areas. Brown (1978) estimates that in 1976 there were 6.8 million dogs in the USA of which 20% were strays.

Likely modes of transmission from dog to dog are:

• Contact with birth products. Birth products associated with infected bitches can be highly infected with B canis. These include: infected foetuses, foetal fluid, and vaginal discharges which may continue for at least 4 to 6 weeks. These materials can contain up to 10¹⁰ organisms per ml, Carmichael (1990) .
• Lactating bitches.The milk of infected bitches can be highly infected. Puppies may become infected in this way, Spink (1970) , although intra-uterus infection is more likely. Leaking of milk is a possible cause of transmission to other dogs.
• Oestrus Secretions. Vaginal secretions of a bitch during oestrus can be highly infected. This a possible route of infection to both male and female dogs.
• Semen. In males B canis may present in semen in high concentration particularly in the first few months of infection and then intermittently for years thereafter. Concentrations of up to 10⁶ organisms per ml in semen are typical. This happens because B canis can be locallised in the prostrate and epididymis. Transmission may occurs during mating to females and by semen spillage to other dogs.
• Urine. B canis is usually present in the urine of both male and female infected dogs. The concentration in male dog urine is usually much higher at 10⁶ organisms per ml. Urine usually becomes infected at about 4-8 weeks after infection. This is because of the presence of semen and prostate fluids in the urine. Transmission to other dogs is likely through environmental contamination. Serikawa (1979) . It has been shown but not proved that infected urine can be a mean of transmission between dogs kept together for up to 6 months, Carmichael (1988) .
• Oronasal and Ocular Secretions. B canis can be present in low concentrations in saliva, nasal and ocular secretions. In general this is not consided that a significant means of transmission, Weber (1982) .
• Faeces. Contact with infected faeces is a possible means of transmission to other dogs. Egwu (1979) infected rabbits with B canis and showed that it was present in the faecal matter for at least 3 weeks indicating a possible source of further infection.
• Contaminated Environment. The environment associated with an infected dogs may be contaminated with B canis, . Information relating to other forms of Brucella (see above) demonstrates that they can remain viable in a contaminated environment for long periods depending on temperature and other conditions. Cool moist conditions favour survival. Faecal matter, urine and other bodily secretions are likely contaminates.
• Blood transfusions. Dog to dog transfusions may spread B canis.
• Veterinary Procedures. The use of non sterilised instruments during veterinary procedures such as vaginoscopy, artificial insemination, and contaminated syringes are a possible way of spreading B canis.

Dogs infected with B abortus, B suis and B melitensis can also transmit these forms of Brucella to other animals and humans. In a trial von Kiok (1978) infected 9 bitches with B abortus and penned them with 11 Brucella free pregnant heifers. 6 heifers developed brucellosis which was identified by bacteriological culturally and the heifers aborted. The heifers subsequently aborted. In consideration of brucellosis in dogs caused by B abortus, B suis, B melitensis and B canis Nicoletti (1980) concluded only B canis caused a naturally perpetuating disease in dogs whereas disease caused by the others was self limiting.

Control and Elimination of B canis in Dogs.

B canis can affect dogs in kennels, stray dogs and pet dogs. If a dog is found to be infected with B canis the best course of action is euthanasia because of the risks to human and other animals.. Of course pet dog owners are reluctant to do this. Kennels are usually commercial ventures and therefore there is a need to eliminate the disease and resume normal operations as quickly as possible. Wanke (2004) notes that kennel owners are unwilling to accept that asymptomatic dogs are infected with B canis but he considered that it is unethical to breed from such dogs. It is a difficult and costly job to eliminate B canis from a kennel. Control and elimination procedures are similar to those developed for other animals such as cattle and pigs, Greene (1984) and Pickerill (1970) . The cost of eliminating B canis from a breeding kennel is high and includes the loss of revenue from dog sales (75% of affected bitches may abort.), the loss of breeding stock, and the veterinary costs associated with supervising the elimination and surveillance during the follow up. Regimes to eliminated B canis from kennels involve segregation of all dogs and elimination of infected animals based on repeated screening with serology and culture tests. It is also essential to take measures to prevent B canis entering the kennel such as screening of new dogs.

The following procedures for eliminating B canis from a kennel have been proposed by Carmichael (1990) and Hollet (2006) :

Prevention.

• No dog should be introduced into a kennel which might have been exposed to B canis or which showed possible symptoms.
• Dogs introduced from outside should be quarantined for at least 12 weeks. Such dogs should have negative serology titres and be tested twice monthly and show negative results taken one month apart.
• All dogs in the kennel should be tested at least yearly and at any time that there are possible signs of B canis infection such as reproductive problems including abortion.
• All male and female dogs should be tested before mating.
• All abortions, etc, should be considered questionable until proved otherwise.
• Dogs which prove positive to ME-TAT or ME-RSAT should be quarantined and subject to B canis culture tests.
• If infection is found elimination procedures should be commenced.

Eradication.

• If B canis is found in a kennel all dogs must be tested.
• Tests should include ME-RSAT and TAT or AGID-CPAg. Dogs giving suspicious results in these test should be subject to three consecutive culture tests.
• Bitches should be tested in the weeks before oestrus when serology testing works better.
• Dogs known to be infected should not be allowed to breed.
• Dogs testing positive should be remove and euthanized. Alternatively to speed up eradication all dogs considered suspicious by serology tests should be euthanized.
• The quarantining measures should include a complete physical barrier to prevent any form of contact between dogs.
• Infected bitches which have aborted once may subsequently produce live litters but are likely to pass B canis to the pups. Such puppies should be euthanized. A less satisfactory alternative is to keep them alive and monitor them by the culture test.
• Although some veterinarians advocate the infected dogs be spayed or neutered, such dogs are not necessarily free from B canis and can have positive serology titres for years.
• The kennel can be considered clear of disease when tests in 3 consecutive months are negative for all dogs.
• If B canis is present in a kennel humans should take hygienic precautions particularly relating to abortions and mating including gloves and appropriate protective clothing.
• Contaminated areas should be cleaned with disinfectants such as quaternary ammonium salts, sodium hypochloride, bleach, iodoform, ethanol or formaldehyde.

Wanke (2004) suggests an alternative procedure based on the use of antibiotics. This is a regime that might be more appealing to animal lovers but it is accepted that no antibiotic treatment is completely effective. Also chronically infected male and female dogs may be serologically negative and resist antibiotic treatment. Puppies born to infected bitches may harbour the disease so that it can re-appear after puberty to start off a new outbreak. This is also probably a more expensive option for owners. The regime is initiated if possible signs of B canis infection such as abortion or epididymitis are present:

• Quarantine the kennel until B canis has been eliminated.
• Test all dogs by serology and blood culture tests and determine the source of the outbreak if possible.
• Remove infected dogs rather than separate them physically. All such positive dogs should be neutered and treated with antibiotics.
• Treat all remaining negative testing dogs for 1 month with a combined antibiotic therapy such as tetracycline and streptomycin (Note: The use of streptomycin in dogs should be discouraged because of its importance in the treatment of tuberculosis in humans.).
• Test the dogs at monthly intervals to identify new positives which can occur for up to 5 months. The kennel is considered clear when there are 3 consecutive months free from positives.
• The kennels should be thoroughly clean using an effective cleaning agent such as quaternary ammonium salts.
• Continue to test every 3 months for a year.
• Establish a plan to prevent re-infection.

Wanke (2006) reports a trial using enrofloxacin in an infected kennel. He considered that the use of this antibiotiv therapy may permit maintenance of fertility, prevent abortions, and stop transmission of the disease to pups, even with persistent seropositivity in some breeding dogs. The ethicality of this approach must be questioned.

Treatment of Pet Dogs with B canis.

There currently is no reliable treatment for B canis in dogs. Long term antibiotic therapy is used but with variable success rates and in many cases it fails completely. In general B canis is very difficult to eliminate in individual dogs irrespective of the pharmaceutical agents used. Pet owners may demand treatment but this is not in their best interests it is more sensible to euthanize an infected dog. bAn infected dog is a health risk to the owner and others. Dog owners may be unwilling to accept that a dog is infected with B canis and should be euthanized if the dog is asymptomatic and may insist on treatment.. Treatments based on single antibiotic are ineffective, Greene (2006) and while combinations such as tetracyclines and gentomycin can work better, Jennings (1974) failures and relapses are common. Wanke (2004) describes possible antibiotic treatments for dogs. Antibiotic treatment is probably most successful if used early in infection but less so once B canis has become established and localised, Moore (1970) . Treatment with antibiotics causes serology antibody titres to fall but this should not be taken as indicative of successful treatment since some such dogs can relapse in which case serology titres may rise again. Dogs with chronic brucellosis may have negative serology result irrespective of other considerations. A combination of multiple antibiotics including doxycycline or minocycline or long acting oxytetracycline and streptomycin or gentomycin is a typical treatment regime, however it is generally considered that streptomycin should not be used for dogs because of its importance in human treatment. In most cases the antibiotics will reduce the bacteria load in the blood stream but will not fully destroy all the bacteria in the animal. Some vetinarians suggest the the dog should be spayed or neutered, Dillon (1981) to control sexual transmission of the disease but this should not be considered a cure and the dog can still spread the disease by other means to animals and humans. For pet owners the cost of antibiotic treatment and longer term monitoring are considerable and may be unsuccessful so that euthanasia is by far the best option. There is no vaccine available to immunise dogs against B canis infections in dogs.

Diagnosis of B canis in Dogs.

Diagnostic methods for B canis depend on physical examination, bacteroliogical culture and serology tests.

• General Tests and Physical Examination. Physical examination can sometimes offer useful information as to the possible presence of B canis infection in a dog. (Also see also "Signs and Symptoms of B canis in Dogs".)
• The Bacteriological Culture Test. The culture test is the best method for proving a dog has B canis and a positive result gives sure proof of Brucella infection. Further type-testing is necessary to differentiate between Brucella prove that B canis is the cause. The culture test on blood gives positive results for about 50% of infected dogs in the early stages of disease however bacteremia is only intermittent later in the course of infection so the test becomes less reliable. The culture test is usually carried out on blood but in dogs it is also gives useful results for semen, urine, vagina secretions and tissue samples. For example, some dogs have a positive culture tests for urine while at the same time the culture test on serum is negative.It is important to note that a negative culture test result cannot be considered proof of the absence of the disease.
• Polymerase chain reaction (PCR). PCR tests for DNA released by Brucella. Specific sequences of DMA are targeted which are not present in other bacteria. The use of these new tests are in the trial stage and there is not yet a standard test procedure. There is the possibility of these tests can replace the bacteriological culture both to detect Brucella and to identify the type, in this case B canis. They could also possibly be used instead of serology antibody tests.
• Serology Tests. Serology antibody tests are usually used for screening dogs for B canis infection. Tests such as SAT, BPAT, RBPT, developed for B abortus, B melitensis and B suis infections in animals and humans do not work for dogs with B canis. This is because these serology tests rely on for antibodies against smouth lipopolysaccharides, S-LPS which is not present in B canis. Tests specific to the rough lipopolysaccharides R-LPS which is present in B canis have been developed. Antibodies against B canis can be present from 2 weeks post infection, Weber (1977) . The main serology tests which are in use are: the rapid slide agglutination test (RSAT) (with or without 2-ME), the tube agglutination test (TAT) (with or without 2-ME), the agar gel immunodiffusion (AGID) and iELISA. The success of antibody serology tests depends on the stage of the disease. They work best in the early stages but may fail completely in the chronic phase. Dogs can also sometimes become infected with B abortus or B suis or B melitensis. This is either by association with infected stock or eating contaminated meat. Therefore it may be useful to test a dog with appropriate S-LPS tests such as SAT, BPAT or RBPT.

The tests describe here depend on a rough LPS antigen usually strain (M-) variant of B canis.

General Tests.

The signs and symptoms of B canis in dogs is discussed above. The most obvious signs are abortion in females and swollen testes in males and swollen lymph nodes in both sexes.

• General blood tests are of little use for diagnosing B canis in dogs. Akhtardanesh (2011) tested 95 dogs (pets at 62 and kennels dogs at 35) for B canis of which showed 15.8% had positive serology antibody tests to B canis. General blood tests (haematological and biochemical) showed no significant difference between seropositive and seronegative dogs.
• Examination of semen can be useful. In the first 3 months after infection abnormal sperm morphology and inflammatory cells can be present. In chronically infected dogs the sperm and immature sperm may be greatly reduced in number and be present with head to head agglutination.
• X-ray examination can be useful in some cases to such as inflammation occurring at spinal disc spaces as well as joint and bone abnormalities related to osteomyelitis.
• Ophthalmologic examination may show evidence of uveitis with associated lesions.
• Ultrasound examination may indicate genitalia problems and testes atrophy in male dogs.

The Bacteriological Culture Test.

The culture test can be used to conclusively prove that a dog is infected with B canis. Tests can be carried out on tissue samples and blood or other bodily secretions. The success rate of the blood test is reasonably good in dogs. Dogs usually become bacteremic at 14 to 21 days from the start of infection and may remain so for many months or may be intermittently bacteremic for long periods. Early in the infection about 50% of dogs give positive culture tests but the test is less effective in the longer term. The number of organisms circulating in the blood is usually small so multiple samples may be required to get a positive result. Antibiotic treatment greatly reduces the chances of a positive culture test result. The culture test if positive proves infection but a negative test does not prove the dog is B canis free. Despite a dog ceasing to be bacteremic B canis is likely to persist and become localised in internal organs including: the spleen, bone marrow, prostate gland, and epididymides from which it may be intermittently released for years. According to Carmichael (1990) B canis can be readily recovered from tissue samples including: lymph nodes, spleen. liver, and male reproductive organs. In abacteremic and chronically infected dogs the most likely sites for recovery are: the prostrate and epididymides in males and spleen, lymph nodes in both sexes. B canis can be recovered from still born and aborted foetuses (particularly from the animal's stomach contents), from lymph tissue, urine (more often positive in males than females), vaginal secretions which can contain up to 10¹⁰ organisms per ml and from the milk of lactating bitches, Brown (1975) and Pollock (1979) . It can also be recovered from tissue samples from joints (such as associated with discospondylitis) and eye lesions. Urine can be positive in conjunction with negative blood samples. About 50% of dogs infected with B canis initially become bacteremic and current knowledge indicates that this can last for at least 5 years, Carmichael (1984) . Spink (1968) . noted that B canis could be recovered from lymph tissue of apparently well dogs. The time from infection to bacteremia is usually about 2 - 4 weeks. In the ongoing phase of the illness B canis becomes localised such as in the prostrate and epididymides with continuous or spasmodic release lasting for up to years. Semen and urine can be similarly affected.

Suitable methods for isolating B canis are described by Carmichael (1996) . From a bacteriological view B canis is considered most like B suis but it has the unusual feature of being naturally muccoidal (M+) and in culture forms ropey growths which are sticky and difficult to handle. Type testing methods are described by Corbel (1984) . It is an aerobic and grows slowly in enriched media such as (Albimi) at 35-37°C but is inhibited by CO₂. At least 9 days should be allowed. The colonies are mucoid and ropy in media suspension. Cultures are very sticky and difficult to handle particularly below pH 6.8 but are less mucoid at pH 7.4. Growth is slow like other Brucella species. B canis is always recovered in the rough form.

Polymerase Chain Reaction (PCR).

PCR is a new test as applied to the detection of B canis. It can be considered a direct test like the bacteriological culture test because it tests for DNA derived directly from B canis. At this time the use of the procedure is still experimental,there are no standardised methods and the results can be variable in sensitivity. Specificity depends on choosing DNA sequences only present in Brucella. In theory PCR could replace both serology antibody tests and bacteriological culture tests. PCR could be a better option than serology antibody tests because a positive result is a definite proof of infection whereas antibody tests have known problems. General information suggests that the test works better in tissue samples than in blood and better in whole blood than on serum. Some examples of the methodology currently being assessed are given:

• Possible methods for the diagnosis of B canis by PCR are described by Keid (2007) and .

• Keid (2007) describes the use of a pair of primers directed to 16S-23S rDNA interspacer (ITS) was designed directed to Brucella genetic sequences in order to develop a PCR putatively capable of amplifying DNA from any Brucella species. Tests with dogs indicated that this could detect DNA from B canis in low concentrations.

• Imaoka (2007) describe a combinatorial polymerase chain reaction (PCR) procedure to identify four major species of the genus Brucella simultaneously involving four pairs of primers targeting the genes encoding a cell surface protein (BCSP31) and outer membrane proteins (omp2b, omp2a and omp31). able to identify B canis.

Some of the findings of studies using PCR to detect B canis in dogs are as follows:

• Keid (2007) compared the usefulness of PCR assay in semen and blood from 52 naturally infected male dogs with serology antibody tests and bacteriological culture. The positive results were: RSTAT, 30.7%; ME-RSAT, 9.6%; bacteriological culture (blood), 26.9%; bacteriological culture (semen), 21.1%; PCR (blood), 34.6%; PCR (semen), 34.6%. The results indicate that both PCR (blood) and PCR (semen) could be useful confirmatory tests.

• Keid (2007) tested the use of PCR on vaginal swabs from 144 bitches and compared the results or blood using PCR, bacteriological culture and serology antibody tests. The positive results were: RSAT, 45.8%; ME-RSAT, 15.9%; bacteriological culture (blood), 34.0; %; bacteriological culture (vaginal swab), 4.1%; PCR (blood), 37.5%; PCR (vaginal swab), 36.2%, total PCR (blood and vaginal swab), 50.7%. The results indicated the usefulness of PCR analysis in B canis infection.

• Keid (2010) compared the value of PCR (whole blood) and PCR (serum) for dogs with confirmed or suspected B canis infection and concluded that tests of serum using PCR are of little value. It detected less positives than bacteriological culture (whole blood), PCR (whole blood), RSAT and ME-RSAT.

• Aras (2010) evaluated and compared PCR to bacteriological culture on 48 stray dogs (18 males and 30 females) with possible B canis infection in Konia, Turkey. Tests were carried out on inguinal lymph node samples. Both tests showed the same number of positives at 8.3% indicating the PCR could be used as a substitute for bacteriological culture.

• Corrente (2010) describes the use of PCR to test an Italian male dog for B canis. The dog suffered from stiffness, decreased physical activity, lethargy, reduced pelvis limb reflexes and faecal incontinence manifest as epididymitis, infertility and diskospondylitis. The dog had a long history of chronic protatitis. Bacteriological culture tests and real time PCR assay using 4 sets of primers targeting the genes BCSP31, omp2 and omp31 to identify Brucella spp and a combinatorial Brucella-typing PCR which yielded an amplification pattern specific for B canis. RSAT (B canis M- antigen) and CFT serology tests were also carried out. The culture tests on blood, urine and on a inter-vertebral disk fragment were all negative. PCR on the same samples was positive for urine and disk fragment but negative for whole blood, however was positive for the buffy coat (a layer between the plasma and red blood cells of centrifuged blood) possibly suggesting a intra-macrophage localization for the Brucella. RSAT detected B canis antibodies and CFT was weakly positive at 1:8. The work demonstrates that PCR can be a powerful weapon in the study of B canis infection but the negative result for blood indicates that like other tests the PCR is not infallible.

Serology Antibody Tests.

In General. Serology tests are an essential tool in the diagnosis of B canis in dogs but currently (2011) there are no properly standardised tests and there are problems associated with all the tests. However according to Hollet (2006) there are no standard tests for B canis and problems occur with all the tests currently in use and he notes that in practice titres from serology fluctuate even with persistent bacteremia and are not a guide to the stage of the illness. Carmichael (1990) notes that in dogs with B canis antibody serology tests are unreliable for the first 4 weeks after infection even if the dog is bacteremic but are thereafter remain present while the dog is bacteremic. A dog can remain bacteremic for up to 5 years but when this stops serology antibody titres decline and become negative or equivocal irrespective of the test used. Antibodies may continue to be present in chronically infected dogs for up to 16 months but then usually decline to negativity. This has been demonstrated for abacteremic dogs with localised infections of the prostrate or epididymides. Dogs treated with antibiotics may retain low agglutinate titres and it should be noted that antibiotic treatment in dogs is considered ineffective. A negative antibody titre against B canis appears to have no bearing on recovery from the disease and can often be coexistent with bacteremia for long periods. Antibody treatment may reduce or suppress antibody titres.

Choice of Antigen. The choice of antigen has had an important role in the development of serology antibody tests for B canis infections. As already indicated serology antibody tests developed for smooth type Brucella do not work for B canis. Tests for B canis have been developed with antigens derived from rough forms of Brucella. Antigens isolated from the rough forms: B ovis, B canis and B abortus 45/20 vaccine can all be used. Antigens are made from B canis (Rm 6/66) or (M-) were developed by Carmichael (1968) because natural B canis cultures are difficult to handle due to their viscous nature. A sonical extracted antigen was first used but gave rise to many false positive results. In general antigens derived from B ovis and B canis (RM 6/66) give high levels of false positives (62%) while results with B canis (M-) are more precise, Wanke (2004) . Better results were obtained by Myers (1972) who extracted a hot saline R-LPS antigen from B ovis and Diaz (1973) by a similar method recovered a R-LPS antigen from B abortus s45-20 (a rough strain vaccine). Carmichael (1990) has shown that R-LPS can be used as the antigen to detect antibodies to B canis. In all cases the antigen reacted only with sera from animals infected or vaccinated with rough strain Brucella. Problems have been found with the B ovis antigen which gives significant levels of false positive results. Such false positive results are mostly due to antibodies developed against other gram negative bacteria which share antigens with B ovis. Such gram negative bacteria may include: Streptococci, Staphilococci, Bordella bronchiseptica, Pseudomonas, Pasteurella multocida, and Moraxella spp, Zoha (1982) , and Carmichael 1987) . Better results are obtained with a B canis antigen.

The Tube Agglutination Test (TAT). This test probably measures mainly IgM. The TAT test was developed by Carmichael (1968) . The test uses a R-LPS antigen but is otherwise similar to the SAT test for smooth Brucella used in cattle and humans. Positive results can be obtained from 2 to 4 weeks after infection, Carmichael (1968) . The test is described by Alton (1975) and Carmichael (1968) . Briefly, a range of dilutions of sera (typically 1:50 - 1:6400) in phosphate-buffered saline solution at pH 7.2 are prepared. The preferred antigen is a suspension of saline heat-killed B canis. This antigen is added to the sera and the mixture incubated 24 to 48 hrs at 50°C. Positive results are assessed against a standard and the highest dilution showing a 3+ agglutination is recorded as the result. The differentiation between positive and negative results is subjective. The reading taken as positive is either 1:50. 1:100 or 1:200. Fredrickson (1974) , Rhoades (1980) and Carmichael (1990) consider titres of ≥1:50 ≤ 1:100 are suspicious, titres of ≥1:100 are positive and a titre ≥1:200 is evidence of active infection. Some consider that there is a direct correlation between a titre of ≥1:200to and a positive culture test result. However it is probable that screening dogs at tires of ≥1:200 misses significant numbers of infected dogs. Tires are often negative or low results can occur in the early stages infection so it is essential that such titres are repeated at say 2 week intervals. This test and/or ME-TAT is often used as a confirmatory test for RSAT or ME-RSAT titres.

Damp (1973) described an alternative method to TAT using a micro-plate method which she considered gave similar results. Kimura (2008) also describes a micro-plate agglutination test and considered it more sensitive and simpler to perform than TAT with a potential for large scale use.

There are a number of problems associated with the use of TAT to diagnose B canis in dogs:

• Dog blood has been found haemolyses readily in which means the test cannot be used.
• It cannot detect low titres in chronically infected dogs despite positive blood cultures.
• It has been shown to be ineffective with some bacteremic dogs which only develop a low agglutination response.
• It is affected by the prozone effect (See SAT).
• The procedure is slow requiring >48 hrs to get a result.
• The test gives false positives due to cross reactions. Antibodies to other bacteria with similar LPS can be detected.

ME-TAT. ME-TAT is a variant of the TAT test in which the serum is first treated with 2-meracaptoethanol. This tests were introduced to overcome problems with that TAT test and to reduce false positive results (>50%). This procedure is similar to the ME-SAT test that has been used cattle and humans (See: Testing of Brucellosis in Humans.). The basis of those tests was that 2-meracaptoethanol breaks down -sulphur-sulphur- bonds in antibodies. This primarily affects IgM which is therefore not tested and in its absence the test then measures mainly IgG instead. In the case of humans some researchers placed considerable importance on this test. They suggested that it could differentiate between active and inactive brucellosis in both cattle and humans and this idea has been transcribed to the tests for B canis in dogs. In reality the impact of 2-meracaptoethanol is to damage not just IgM but also to a lesser extent IgA and IgG which reduces the tests sensitivity. Although it is suggested that there is a 80% correlation to results of the culture test the designation of active and inactive brucellosis may be useful but is not empirical. In the case of humans infected with other Brucella better results are obtained from more sensitive tests for IgG such as AHG and CFT. The cross reactions with antibodies to other bacteria are associated with IgM rather than IgG so the main virtue of ME-TAT (and ME-RSAT) is a reduction in false positive results, Moore (1969) . This test remains positive for up to 12 weeks after bacteremia has ended.

The rapid slide agglutination test (RSAT). This is the main test used in the diagnosis of B canis in dogs. It is based on the Plate Agglutination Test used for B abortus in cattle and it is quick and easy to use so that it can be used by veterinarians without recourse to laboratory facilities. It works best in the early stages of infection. The main problem is false positive results. The original method was developed by George (1978) using 'a B ovis antigen. According to Flores-Castro (1977) this test is reputed to give few false negative results, but significant levels of false positives at 50-60% which is due to the presence of antibodies against other gram negative bacteria which share antigens with B ovis.

The methodology for RSAT is described by Carmichael (1987) and Lucero (2000) . It is a rapid screening test based on the Plate Agglutination Test used for B abortus in cattle. The antigen used in the original test was derived from B ovis but better results with less false positives is given with an antigen derived from B canis, Mateu-de-Antonio (1994) . The antigen used in this test is a hot saline heat killed reduced mucoidal (M-) strain of B canis strained with Rose Bengal. A 10 μl aliquots of serum at a range of appropriate dilutions are dropped on to a 25 x 75 mm glass slide and 10 μl aliquots of antigen are added and they are mixed together for 1-2 minutes and then a reading (extent of agglutination) is taken at 10x magnification. Control samples of serum of known titre are used as standards.

According to Badakhsh (1982) RSAT has the following advantages of the test over TAT:

• The test can be used even if the dog's blood haemolyses.
• The test can be readily used by veterinarians.
• It is good for showing negative results in non-infected dogs apparently with few false negative results. (Note: this does not apply to chronically infected dogs.)
• RSAT is a good technique for identifying the early B canis in the early stage s of infectiuon.
• A problem with RSAT is that it gives rise to false positive results. Although the test measures mainly R-LPS there are shared epitopes which lead to some overlap with LPS of other bacteria.

Commercial test kits are available such as Manufactured by Synbiotics Corp. Kansas City, Kansas City, U.S.A but this uses the B ovis antigen. This test gives high levels of false positives even if 2-meracaptoethanol is added. Much better results have been obtained with B canis (M-). Carmichael (1987) and Damp (1973) .

ME-RSAT. ME-RSAT is a variant of RSAT. This test is also widely used to diagnose B canis often as a confirmatory test for RSAT because it shows less false positive results which mainly relate to IgM. A problem is that it mainly measures IgG and these antibodies may not be present for up to the firdt 12 weeks of infection therefore it is usually used as a complimentary test. As in the case of ME-TAT 2-meracaptoethanol is added to the serum sample prior to adding the antigen as describes by Badakhsh (1982) . In this case test 0.1 ml (1 drop) of serum was combined with 0.1 ml of 0.2 M 2-meracaptoethanol then rapidly mixed (≤ 10 seconds). 1 drop of this mixture plus 1 drop of the antigen are combined on a glass or plastic slide and the extent of agglutination immediately determined at 10 x magnification. Badakhsh considered that if a dog which reacts positively to RSAT but negatively to ME-RSAT or ME-TAT then this was indicative of a false positive result. He based this on a good co-relation between the culture test results and 2ME-RSAT results. The problems of limited sensitivity of the ME-SAT titre demonstrated for humans and probably apply to the ME-RSAT titre for dogs as well. Nevertheless this is a useful test in dogs. A possible better combination of tests is RSAT for the early screening test (IgM) and iELISA to test for IgG later on. From general considerations of reliability and ease of use ME-RSAT is preferred to 2-ME-TAT. ME-RSAT usually remains positive for 12 weeks after bacteremia ceased but in some cases has been known to continue for up to 30 months.

Seronegative cases where bacteriological culture has proved the presence of B canis as a localised infection in the prostrate and epididymides have been described by Moore (1969) and Flores Castro (1977)

Agar Gel Immunodiffusion (AGID).

This test method was first applied to the diagnosis of B ovis in sheep by Myers (1970) and then subsequently applied to the diagnosis of B canis in dogs, Myers (1974) . Myers found that it gave similar results to thee previously developed TAT titre. The test method is as follows. Simplistically, an agar gel plate is prepared and holes are punched in it in a predetermined pattern, usually a central hole with six holes arranged around it. A suspension of the B canis antigen is placed in the central hole and samples of sera and standards in the other holes. Both antigen and antibodies are allowed to diffuse outwards and where they met a precipitation line is produced if there are antibodies to B canis in the serum. One problem with the test is that the precipitation lines can be difficult to interpret, however according to Wanke (2004) the precipitation lines due to B canis can be distinguished from those of other bacteria by careful analysis including the incorporation of standards.

The antigen used in the test is significant. Three have been used: a sonically disrupted antigen, a hot saline extracted antigen (cell wall antigen, cwa) and an antigen derived from cytoplasmic protein (cpa). In all cases the antigen is derived from a rough variant of Brucella either from B ovis or B canis or B abortus 20/45 rough vaccine. The choice of antigen is important. The antigen derived by sonically disrupted cells which was used in early work caused false positives due to an overlap with antibodies to antigens present in other Brucella and antibodies to other gram negative bacteria with the same antigens.

• Agar Gel Immunodiffusion using Cell Wall Antigen (AGID-cwa). Better results are obtained if sonically disrupted cell antigen is replaced by an antigen extracted by a hot saline process with less false positive results. However, there is still overlap with other rough Brucella but it does eliminate infection due to smooth Brucella such as B abortus and B melitensis which dogs can acquire due to contact with infected animals or meat. AGID-cwa is ineffective for the first 12 weeks after the start of infection however it remains effective longer (≥12 weeks).

• Agar Gel Immunodiffusion using Cytoplasmic Protein Antigen (AGID-cpa). In this test the antigen is cytoplasmic protein (cpa). This soluble antigen from the interior of the cell is made by gentle agitation of B canis in acidified buffer or by sonic disintegration and subsequent separation by high speed centrifugal separation. LPS is removed by solvent extraction with methylene chloride although residues can remain. Cytoplasmic protein is common to both rough and smooth Brucella as shown by the analytical method SDS-Page, Baldi (1997) . Therefore the test cannot differentiate different types of Brucella. The test method is the same as that used for AGID-cwa. This test is considered to be a useful confirmatory test, Zoha (1982) . AGID-cpa gives positive results where other tests are negative or equivocal. Carmichael (1984) noted that positive results from the test could last to ≥36 weeks after bacteremia has ceased therefore the test was particularly useful in chronic infections. Healthy dogs do not give reactions to the test unlike AGID-cwa. Tests based on cytoplasmic protein are highly specific for Brucella species and therefore false positive reactions due to cross reactions with other bacteria do not occur.

The Compliment Fixation Test (CFT or CF Test). The CF Test which has been widely used for other types of brucellosis has been tried for B canis infections, Alton (1975) and Weber (1977) . However it is not widely used because of the significant anti-complement effects in canine sera.

Indirect Enzyme Linked Immunosorbion Assay (iELISA). These tests are not widely used and are not commercially available. The test can be used to identify sera antibodies to B canis is similar to the procedure used for B abortus in cattle except that the in this case the test must be set up to identify antibodies to R-LPS rather than S-LPS. A suitable method is described by Lucero (2002) . In this case the antigen used is derived from (M-) variant of B canis. This is extracted from B canis in hot saline by the procedure of Meyer. The conjugate is made up of horseradish peroxidase and protein A/G. and the colour is developed using H₂O₂/2,2^'-azino-bis (3-ethylbenzthiazoline-6-sulphonic acid). The colour is measured at OD₄₁₄ with a negative/positive cut off at 0.281 OD. The tests are calibrated with sera of known OD values. This test is for IgG and IgA. It is suggested that this test is best used in conjunction with the RSAT test. The iELISA test can also be used to identify antibodies to cytoplasmic protein or a specific extracted cytoplasmic protein (18 kDa), Baldi (1994) (1997) and . As in the case of AGID-cpa there is also no cross reaction for iELISA-cpa with other bacteria. Wanke (2002) compared the effectiveness of iELISA-cwa and iELISA-cpa with ME-RSAT and concluded that both iELISA-cwa and iELISA-cpa are highly specific and sensitive for the diagnosis of canine brucellosis and can detect the infection by B. canis shortly after the exposure to the pathogen. In general both tests gave better results than ME-RSAT. The choice of R-LPS antigen seems to be important. Cell wall antigens derived from wild strain B canis RM 6/66 gave high levels of false positives whereas results with B canis (M-) were more specific but less sensitive, Mateu-de-Antonio (1993) .

Indirect Immunofluorescence. Indirect immunofluorescence uses a primary and a secondary antibody. The primary antibody attaches to the antigen and the secondary which is conjoined to a fluorophore attaches to the primary antibody which is then read using a microscope. According to Weber (1976) this test is more sensitive and specific than the than agglutination tests such as RSAT. However Wanke (2004) considered the test gives rise to a high rate of false positives. Carmichael (2006) considered that this test has lower sensitivity than other agglutination and iELISA tests.

Rose Bengal Plate Test (RBPT). If it is considered that a dog is has a brucellosis infection then as well as testing for B canis, the possibility should also be considered that the dog could be infected with a smooth form Brucella. Dogs can become infected with B abortus, B suis or B melitensis by association with infected animals or their contaminated environment or by consumption of contaminated meat. The RBP test which was developed to detect smooth type Brucella using a S-LPS antigen or a similar test can be used. Just as with other tests there is the possibility of false negative results from these tests.

Other Tests.

Because the existing B canis antibody tests are inadequate new tests are still being developed. These include: PCR, Antigen coated beads and tests using outer membrane protein analysis (OMPA).

• OMPA has been developed by Brower (2007) and Okwumabua (1995) . Watarai (2007) describe a method in which crude antigens were extracted from B. canis using hot saline and coated on to latex beads. They considered that the method was more specific than TAT using whole bacterial cell antigens but more work is required. The test is easy to use and can be done in 15 minutes.

• Kim (2007) describes an immunochromatographic assay for B canis. This is simple to use and according to the authors gives results comparable to ME-RSAT. Serum and whole blood can be used. A strip of nitrocellulose is used for the immunodiffusion with a control line used as a standard and a test line containing B canis extract to react with B canis antibodies. Serum or whole blood is added through an opening which contains Protein A conjugated to colloidal gold which interacts with the antibodies. The general principle is similar to iELISA. If B canis antibodies are present then they interaction with the antigen at the test line which becomes visibly due to the gold. The interpretation is subjective: +++ is a strong positive, ++ is a moderate positive, + is weakly positive and no line is negative. Some doubt as to the reliability of the test is cast by the work of Behzadi (2011) which showed the test worked in some dogs but not others. It seems possible that only IgM is tested.

Comparision of Serology Test.

Mateu-de-Antonio (1994) compared a range of serology for the detection of B canis in dogs. They found that RSAT (with B canis (M-) antigen), 2-ME-RSAT (B ovis antigen), 2-ME-TAT (cut off ≥1:50). and AGID-cpa gave similar results at about 6.5% of those sera tested but that RSAT (B ovis) and AGID (Hot saline extracted antigen) (HSES) gave much higher results at 20.2 and 14.5% respectively. He considered that this was in agreement with the work of Carmicheal (1990) who had previously noted that RSAT (B ovis) gives >50% false positive results. Mateu-de-Antonio concludes that the higher results for AGID also represented false positives and this was mainly because IgM is measured in the test. In general serum samples for which precipitation lines took the longest time to develop (48-72 hours) were most likely to be negative by 2-Me-TAT and 2-ME-RSAT. It is noted that IgM antibodies are larger in size than IgG and diffuse more slowly. Also if the AGID (HSEA) tests were repeated this time mixing the suspected false positive serum with 2-ME prior to the test which is known to destroy IgM then the results of the tests became negative. Although ME-TAT at ≥1:50 gave similar results to other tests, ME-TAT at ≥1:200 gave much lower results at 2.6% which probably indicated ME-TAT with this cut-off (1:200) missed many positive results. AGID-cpa continued to detect antibodies long after dogs ceased to be bacteremic. This cytoplamic protein is present in all Brucella variants including the smooth strains B abortus, B suis and B melitensis therefore the AGID-cpa test cannot be used to differentiate between them.

Keid (2009) compared the performance of RSAT, ME-RSAT, AGID were compared for the diagnosis of 167 naturally infected dogs as identified by bacteriological culture and PCR on blood, vaginal secretions and semen. The dogs were divided as: infected, non-infected and suspected. The diagnostic sensitivity of RSAT, ME-RSAT and AGID was respectively 70.58%, 31.76%, and 52.94% and the diagnostic specificity of RSAT, ME-RSAT and AGID was respectively 83.34%, 100%, and 100%. In suspect B canis infection positive results were: RSAT, 15%; ME-RSAT, 0% and AGID, 5%. There were considered to be a significant level of false negative results.

Epidemological Surveys of B Canis in Dogs.

Epidemiological surveys of B canis infections in dogs are limited however those that have been done show that there is a world wide distribution of the disease in dogs. A range of different serology antibody tests are used in these surveys which makes comparison difficult. Some tests are also subject to inaccuracies due to false positive and negative results. In general the levels of disease found is a cause for concern particularly in relation to human health. It can be construed from the results that B canis is more prevalent in both pets and strays dogs in economically deprived areas.

North America.

Most epidemiological surveys have been conducted in North America. The prevalence of positive serology tests among strays ranges from 6% to 37% and among pet dogs from 0% to 20%. Most of the higher results are probably affected by false positive results due to cross reaction which can approximately double the results. However false negative results can also occur.

• Flores-Castro (1976) tested 500 dogs in Mexico City using TAT (≥1: 100) and found that 28.0% of the dogs were had positive results. Bacteriological culture (blood) tests identified that 1.6% of dogs were definitely infected.
• Brown (1976) screened 100 stray dogs and 100 pet dogs from Atlanta, Georgia USA for B canis using RSAT and TAT (cut off at ≥1:100). He found positive results among stray and pet dogs at 9% and 1% respectively.
• Lovejoy (1976) tested 235 mature stray dogs and 67 pet dogs in Memphis USA without reference to consideration of their health, and found that 9.4% of strays and 0% of pet dogs were positive by serology tests.
• Flores-Castro (1977) tested a sample of 59 stray dogs in Mexico City. According to RSAT 37.3% of the dogs had positive titres. Some of these results were considered to be due to false positives caused by cross reaction with other bacterial infections. The 22 with positive RSAT titres were tested using TAT and ME-TAT. The positive results for both according to cut-offs were: 10 of 22 at ≥1:100; 5 of 22 at ≥1:200; ≥ 4 of 22 at ≥1:500. Assuming that RSAT found all the possible positive results the overall positive rate based on TAT and ME-TAT were: 16.9% at 1:100 and 8.5% at 1:200%. The 5 dogs (4 male and 1 female) with the highest TAT results also had positive blood and tissue culture tests. The tissue with positive cultures were: spleen, retropharyngeal lymph node, inguinal lymph node, mesenteric lymph node, epididymides and prostrate gland. 2 other male dogs had positive bacteriological culture tests (epididymides) although the corresponding serology evidence was dubious.
• Higgins (1979) showed that the rate of positive serology test results in pet dogs in Quebec Province by the RSAT and 2ME-TAT (≥ 1:50) were 20.2% and 4.3% respectively.
• Boebel (1979) surveyed 2,572 pound dogs in Illinois and Wisconsin for B canis and found 6.7% were positive to RSAT, 1.5% to TAT and 0.2% had positive blood cultures.
• Bosu (1980) tested 2,000 pet and kennel dogs in Ontario Canada for evidence of B canis. 5% were positive by RSAT. Of these 31% were considered to be suspicious by TAT (1:≥50 ≤100) and 1% positive by TAT at 1≥200 respectively. 6% of the RSAT positive samples were also positive by AGID.
• Brown (1982) tested 200 stray dogs in North Georgia USA and 94 stray dogs in Portof Spain, Trinidad for B canis using the RSAT and TAT procedures. Sero-positive results for North Georgia and Port of Spain positives for RSAT were 10% and 39.4% respectively. The seropositive results for TAT in North Georgia according to titre cut off were: ≥ 6.0%,at ≥1:100, 5.0% at ≥ 1:125, 2.0% at ≥ 1:250; and 2.5% at ≥ 1:500 and for Port of Spain were: 5.3% at ≥ 1:100, 5.3% at ≥1:125%; 4.3% at ≥ 1:250; and 3.2% at ≥ 1:500. It was concluded that the numbers of reactor were 6.0% in North Georgia and 5.3% in Port of Spain and that the high values for SAT in Port of Spain were caused by cross reactions related to the poorer state of health of those dogs.

South America.

• Myers (1980) tested for the presence of B canis among stray dogs in Moreno, Argentina. The tests used were: counter immunoelectrophoresis , ME-TAT and AGID. 131 stray dogs were tested of which 30% were ser-positive and 5 of these gave positive bacteriological culture (blood) results. It was concluded that there was a associated risk to humans.
• Boeri (2008) tested 219 (184 females and 35 males) pet dogs from lower social class areas of Buenos Aires. All were negative for smooth Brucella by the BPAT titre. B canis antibodies were discovered in 16 (9 females and 7 males) dogs (7.3%). 175 dogs were tested by blood culture and B canis was recovered in 3 cases (2 females and 1 male).
• Lopez (2009)

tested 224 (220 females and 4 males) pet dogs in Buenos Aires to determine the extent of infection of by Brucellosis. 33 (14.7%) and 24 (10.7%) tested positive for B canis by RSAT and iELISA respectively. Of the 33 dogs which tested positive to RSAT and 17 of these were tested by bacteriological culture (blood) of which 2 gave positive isolates. The 224 dogs were also tested for smooth Brucella using BPAT which showed 1 seropositive and 4 possible positive results but it was considered that these latter might be due to weak cross reactions with B canis antibodies. Stray dogs were considered a likely vector for the disease.

Europe.

Few surveys have been carried out in Europe other than at infected kennels.

• Mateu-de-Antonio (1994) noted that 6.5% of stray dogs in Catalonia Spain were sero-positive for B canis as determined by 2-ME-RSAT (M- B canis antigen). These results were confirmed by ME-RSAT (B ovis), 2-ME-TAT. and AGID-CPAg. In some areas the positive level was 9.2% in strays and 1.9% in domestic pets.
• Ebani (2003) tested 2328 dogs in Pisa, Italy by AGID and ME-RSAT using B ovis antigen. The samples were also tested for B abortus and B melitensis using RBPT. AGID showed 1.07% positive results but of these only 16% (overall 0.17%) were positive by ME-RSAT.
• Oncel (2005) investigated the prevalence of B canis in Istanbul and Izmir provences in Turkey. Of 362 dogs tested by TAT, Me-TAT and iELISA all using a B canis RM 6/66 antigen positive results were obtained for 12.7%, 7.73% and 7.4% respectively.
• Corrente (2010) notes that to date only isolated cases of B canis in dogs has been observed in Europe. They describe the case of a 6-year old male mongrel dog with chronic prostatitis, discospondylitis and locomotor problems. Clinical signs were stiffness, a decreased level of physical activity, lethargy, depressed pelvic limb reflexes and faecal incontinence and a history of chronic prostatitis. There was also lymph node enlargement and testicular atrophy. Bacteriological culture tests on blood, urine and disk tissue were all negative, whereas they were positive using real-time PCR. RSAT and CFT (rough antigen) serology antibody tests. DNA was detected in the buffy coat rather than whole blood. This is apparently a chronic case. The source of infection was not determined.

Asia.

It appears that B canis is widespread throughout Asia.

• Kikuchi (1979) tested 1549 stray dogs for B canis in Miyang Prefecture, Japan. 173 of 1549 dogs (11.2%) were sero-positive to B. canis. B. canis was isolated from 55 from 148 of these positive dogs which represented 3.6% for the total of 1549 dogs. Isolation rates were higher in females most frequently from the uterus in females and from the spleen of both sexes.
• Diker (1987) using ME-TAT reported seroprevalence rates of 6.3% in 222 serum samples obtained from military, stray, and household dogs.
• Shang (1989) tested 12,948 dogs for B canis using samples taken throughout China. The RSAT and R-RBPT titres were used. The positive rates were from 0.3 - 42.7% in different provinces and were generally higher in the south. The rate was higher in older dogs (≥ 2 years) and females (2 to 1). B canis was separated from the spleen, lymph nodes and liver of 644 dogs. It was noted that the higher the serology titre the more likely that a bacteriological culture test would be positive.
• Kimura (2008) tested 485 dogs in a animal shelter in Kanagawa Prefecture, Japan and showed 2.5% with positive (5.9% among strays and 0.9% among pets) by B canis serology using the MPAT test method.
• Mosallanejad (2009) examined 102 blood samples from pet dogs in Ahvaz, Iran. using a commercial Rapid B canis AB Test Kit as manufectured by Anigen, Animal Genetics Inc. Korea (Cat No: RB21-03) and found 4.9% positive for B canis (2 of 43 males and 3 of 59 females. 3 cases (21.4%) had clinical symptoms: 2 of scrotom dermatitis, and 1 of infertility with bacteriological culture positive semen. The other 2 positives had no clinical signs.
• Behzadi (2011) tested 62 pedigree dogs (Doberman, German Shepherd, Rottweiler, Boxer, Bulldog, Terrier, etc.) and 51 local breed mongrels in Fars Province, Iran using a commercial B canis test kit (Cat No: RB21- 03; M/S Anigen, Animal Genetics, Inc., Korea). 19.4% of pedigree dogs and 0% of mongrel dogs were positive. The reason for the failure in mongrels is not given but may relate to the test kit and the stage of the disease.
• Akhtardanesh (2011) tested 95 dogs (62 pets and 35 kennels dogs) for B canis and smooth Brucella (B abortus using IFA (MegaScreen FLUOBRUCELLA, MegaCor, Horbrank, Austria) and B melitensis) using SAT and RBPT. 15.8% were sero-positive for B canis and 29.5% (SAT) and 12.6% (RBPT) were seropositive for smooth Brucella. 3 bitches had had abortions and 3 males dogs were infertile. In general the dogs sero-positive to smooth Brucella were asymptomatic and only 4 of these were farm (sheep) dogs.

Australasia.

• On the basis of Brucella cultures tested at the WHO Brucellosis Centre, CSL, Melbourne, from 1968-1976, Ekers (1978) concluded that B canis had not been found in Australia. Some have claimed that this means that Australia is free of B canis but this is highly unlikely. Not only are their indigenous dingoes (Canis lupus dingo) (probably introduced by man about 50,000 years ago) but also domestic dogs and red foxes (Vulpes vulpes) have been introduced over the last 200 years from other countries without any consideration of possible disease. A more comprehensive surveys among dingoes, foxes and domestic dogs would be required to prove that B canis was absent from Australia. Patten (1987) tested 225 dogs in Papua New Guinea. for agglutinates to B canis and found 1 positive result. One possible case of B canis in a dog has been found in New Zealand.

Africa.

• Adesiyun (1986) tested 249 dogs in Zaria, Nigeria which had been brought to an animal clinic for B abortus and B canis. 34.9% of exotic dog breeds and 28.1% of local dogs were seropositive by RSAT for B canis. The prevalence amongst dogs younger than 1 year and dogs older than 2 years was 33.3% and 22% respectively. The rate in males and females was 29.6% and 26.7% respectively.

B canis in Aninals other than Dogs.

The extent to which animals other than dogs are susceptible to infection by B canis or whether animals other than dogs could be primary host able to sustain the disease is little understood. Very little work has been done. As with other surveys in dogs and humans the results are affected by variability and viability of test methods and related factors such as cross reactions with other bacteria. Some of the information which is available is as follows:

• Wild Animals.' A number of surveys have been carried out to determine the extent of B canis infection in the wild animals of the USA with variable results. Hoff (1974) found serologic evidence that B. canis was present in coyotes, bobcats (Lynx rufus), raccoons (Procyon lotor) and opossums (Didelphis virginiana) from South Texas, USA. Randhawa (1977) found B canis in 269 wild animals (14 species coyotes, raccoons, opossums, badgers, jackrabbits, and feral hogs) in southern Texas, USA. The CARD test, RSAT and ME-TAT (1:50) were used. Rates were form 6% to 9%. Bischof (2005) tested 63 wild raccoons (Procyon lotor) in Nebraska, USA for B canis using a commercial RSAT serology test (Synbiotics). All results were negative.

• Canidae. It is widely reported that B canis is a disease that affects all canidae and although this is a reasonable supposition there is little evidence There are studies of coyotes (Canis latrans) but virtually nothing for other canidae such as wolves, foxes, dingoes and jackals. *Hoq (1978) found that 6.1% of coyotes gave positive test results for B canis by the RSAT test. Holzman (1992) examined 17 dead coyotes in Georgia USA but did not find antibodies to B canis. Gese (1997) tested 110 coyotes (Canis latrans) from Yellowstone National Park, Wyoming (USA). No coyotes had serologic evidence of exposure to brucellosis, either B abortus or B canis. He notes however that coyotes and wolves usually do not develop clinical signs from brucellosis and that the likelihood of reproductive failure is minimal for both canidaes. Bischof (2005) tested 67 free ranging coyotes in Nebraska, USA for B canis using a commercial RSAT serology test (Synbiotics). All results were negative.

• Primates. Experiments have shown that primates are readily susceptible to B canis infection. Moore (1968) was able to infected a female monkey (Macaca mulatta) with B. canis given orally. Percy (1972) experimentally infected 4 monkeys (Macaca arctoides) with B canis (2 intravenously and 2 orally/conjunctival inoculation with 10¹⁰ organisms. General blood tests were unremarkable. Positive blood cultures were obtained from 2 monkeys at 2 weeks and all monkeys at 3 to 4 weeks. At 5 to 10 weeks positive blood culture was intermittent indicating the strong possibility of chronic infection. Antibodies tested by RSAT started at 2 weeks and rose to 1:1280 at 5 weeks. After euthanasia histopahology indicated the following sites of infection: uterus, acute endometritis with copious grey gelatinous material; liver, a few scattered aggregations of hyperplastic Kupffer cells with focal granulomatous hepatitis; spleen, red pulp with focal granulomatous splenitis; heart, sparse discrete foci of interstitial myocarditis. Bacteriological culture tests were carried out on the liver, spleen, lymph nodes, kidney, uterus, and testes, but surprisingly only one sample (uterus) was positive.

• Farm Animals. It is reasonable to suppose that farm animals could be susceptible to B canis infection through contact with inffected dogs or possibly wild animals. Pickerill (1970) reported that non-pregnant sheep, pigs, and cattle were highly resistant to B canis infection by the oral/conjunctival route. Nicoletti (1982) used a CARD test to examine hospitalised and slaughtered horses for the presence of B canis. Positive results were found for 38% of 73 hospitalised horses and 100% of 141 slaughtered horses which also had 27.4% positive results by TAT and ME-TAT. The age of the horses seemed to be a significant factor with more infection in older animals.

• Cats. There seems to be reasonable evidence that cats can carry B canis but whether they can act as a primary host is not known. Pickerill (1970) found that 3 of 14 cats exposed experimentally to 'B canis became infected as indicated by bacteriological haemoculture but they failed to show antibodies as measured by TAT. Randhawa (1977) recorded rates of B canis at 7.1%, 8.9% and 6.7% by the CARD test, the RSAT, 2-ME-RSAT respectively in Texas. Randhawa also found domestic cats from rescue shelters to be serologically positive at a rate of 8.9% by the 2-ME-RSAT test. Larsson (1984) tested 134 cats ( 63 males, and 71 females) of which 69 were adult and 45 kittens in Sao Paulo Brazil using a TAT titre. 3% were positive at a titre of 1:100 (2 males and 2 females). They failed to identify B canis by positive bacteriological haemoculture. Some consideration is given to eliminating possible cross reaction with Pasteurella multocida.

• Rodents. Carmichael (1968) demonstrated that mice, rabbits and guinea pigs can be experimentally infected with B canis. Carmichael (1990) noted that guinea pigs, mice rats were all susceptible to experimental infection with B canis and that rabbits are susceptible and can develop orchitis, and peritoneal abscesses.

• Parasites. Brucella canis has been isolated from ticks (rhipicephalus sanguinis) associated with an infected dog, Peres (1981) .

Transmission of B canis to Humans.

There is no evidence the humans are a natural host for B canis but they can be infected zoonotically by contact with infected dogs and possibly other animals. Most of the research work and known outbreaks relate to the domestic dog. Some other animals including the domestic cats can seropositive to B canis but no cases of zoonotic transfer to humans have been recorded. Very little is known about the extent of B canis in humans. The first cases of B canis in humans was described by Morrisset (1969) and Swenson (1972) subsequent to the discovery of the disease in dogs. There have only been a few epidemiological surveys to test for B canis in humans and these mostly show seropositivity at about 0 - 2% however some higher levels have been found. There are only a limited number of recorded cases of acute infections in humans (41 at last count, 2011) and some experts consider that human are relatively resistant to B canis infection but this is not proven. It seems possible that a high bacterial dose is necessary to trigger an acute attack of B canis in humans such as might occur from handling highly contaminated tissue samples from aborted foetuses or laboratory cultures. Most likely infection in humans occurs via mucous membranes including the oral cavity and conjuctiva. It is likely that many cases may occur unnoticed not least because there is no routine tests for B canis infection in humans. Also, as with other forms of brucellosis, it is likely that the onset of disease may be insidious. For example for B abortus infections in humans the rate of insidious infection is thought to be about 50%. There is also the reasonable possibility that undiagnosed B canis infection in humans can proceed unnoticed to a chronic state. In a number of the cases which have been documented indicate that repalses are common even if antibiotics are use typical of the chronic course known in other forms of brucellosis in humans. Early workers thought that B canis infection in humans was less severe than other forms of brucellosis but this was probably due to the fact that such cases were related to known outbreaks in dogs and were rapidly diagnosed and treated. Current knowledge known cases of acute brucellosis (B canis) in humans indicate that it has the same characteristics and symptoms as brucellosis caused by other types of Brucella and can be severe even life threatening. According to the work of, Swenson (1972) , Shin (1999) , Godoy (1979) and Munford (1975) the symptoms of B canis in humans are similar to those caused by other types of Brucella and may include: malaise, sore throats and pharyngitis, headaches, chills, joint pain, nausea, weight loss, lymphadenopathy and weakness. Tosi (1982) notes that symptoms can range from asymptomatic to typical symptoms associated with other forms of Brucella including: fever, chills, myalgia, weakness, malaise, anorexia, pharynitis, headaches, splenomegaly, lymphadenopathy, night sweats, weight loss, and depression. Complications include ocular lesions Opperman (1969) and endocarditis, Ying (1999) . Problems of diagnosis include lack of testing and vagueness and the variety of symptoms. The difficulty of diagnosis may mask the true infection rate. Possible means of transmission are: laboratory accidents, Carmichael (1996) and direct contact with infected animals particularly relating to midwifery, Carmichael (1980) . The risk factors to dog handlers include: tissue and secretions associated with births, still births and abortions, vaginal and oestral discharges, and urine particularly in male dogs. There is a possible risk of transmission from contaminated surfaces and via aerosols. Those handling infected puppies particularly children may be at risk.

The Table shows a comparison of symptoms of B canis and those of other Brucella (B abortus, B suis, and B melitensis) derived from the work of Rumley (1986) . This shows that in general B canis infection in human is similar to the others forms. The symptoms are possibly less severe but the amount of data involved is too limited to be sure. Rumley considers that B canis infections may be more chronic or indolent in nature.

Symptom	B canis	Other Brucella
	%	%
Fever	76	70-95
Headaches	46	44-51
Weakness	31	42-68
Chills	38	65-95
Malaise	23	65-95
Myalgias	38	62-64
Sweats	16	61-93
Lymphadenopathy	38	10-20
Splenomegaly	<15	<10
Epididymitis	<15	<10
Abscesses	<15	<10

Testing for B canis Infection in Humans. The same methods and procedures are use for humans as for dogs. In dogs the bacteriological culture test is very useful because dogs can stay bacteremic for prolonged periods but in humans the test is not reliable because the numbers of B canis in human blood is low and it is only erratically present. The problems associated with serology antibodies tests used in relation to B canis infection in dogs are also used for humans have the same problems including: haemolysed serum and cross reactions with other bacteria and negative test results in proven cases.

Some Recorded Cases of B canis in Humans:

• Moore (1967) describes a case of a patient with B canis for whom 7 of 10 blood cultures were positive and she also had a TAT titre of 1:250. The family pet dog was presumably the source of the infection. It also had a positive blood culture for B canis and a a TAT titre of 1:500.
• The 1968 Annual Brucellosis Survey for the US. recorded 2 cases of B canis infection in laboratory workers. Both had positive antibody serology titres and B canis was isolated from the blood in one case. The latter had painful enlarged posterior cervical lymph nodes, low-grade fever and headaches. Both apparently responded to treatment with tetracycline.
• Morrisset (1969) and (1970) described 2 cases of B canis infection in humans who worked in a laboratory preparing B canis antigen. The first patient, a 53 year old woman, had mid symptoms including fatigue with a TAT tire of 1:320. The second, a 33 year old man, had severe fatigue, headaches, chills, nausea, an enlarged tender spleen and enlarged cervical, epitrocheal and maxillary lymph nodes. He had a TAT tire of 1:320 and B canis was cultured from his blood.
• Swenson (1972) describes the long and complex illness of a 23 year old women which eventually proved to be brucellosis caused by B canis. She had first been ill at the age of 4 (1951) and again at 8 (1955) with an acute attacks of illness diagnosed as rheumatic fever. In February 1966 she was admitted to hospital for the delivery of her first child and evidence of congestive heart failure was noted. In September 1967 she suffered a miscarriage and mild congestive heart failure was again noted although she was asymptomatic and he heart condition was being treated with digitalis. In July 1968 she had a normal full term pregnancy and evidence of congestive heart failure was minimal. In July 1969 she was admitted to hospital after being ill for 3 weeks with nausea, vomiting, fever (temperature at 39.4°C) and increasing breathlessness. She had an enlarged heart with a diastolic murmur but no lymphadenopathy or hepatosplenomegaly. She became afebrile after penicillin treatment and was better with only mild breathlessness. However in Feb 1970 after the end of the penicillin treatment she became ill again with fever, chills, a sore throat with a moderately inflamed pharynx, and an increasing frontal headaches. Her temperature spiked to 40°C despite receiving penicillin. On this occasion a bacteriological blood culture yielded a gram negative organism which subsequently proved to be B canis. She was treated with antibiotics including streptomycin for 5 weeks and became afebrile. Serology tests (TAT) subsequently proved positive (1:200). The patient remained well for 1 year after the end of treatment. In retrospect it seems possible that this whole illness was caused by B canis infection. There is insufficient information to show if the diagnosis of rheumatic fever was right but brucellosis is a possible viable alternative diagnosis. The other aspects of the illness can also be ascribed to brucellosis. It is known to cause miscarriages in humans and animals and Brucella is known to cause similar heart problems (endocarditis) to those described. The failure of penicillin to work with brucellosis is well documented. The author considers the a possible source of her illness was the family dog which was shown to be positive serology antibody tests for B canis.
• J Fredrickson (1974) notes that cases brucellosis caused by B canis can occur in laboratory workers handling B canis samples and carrying out bacteriological cultures tests, and in kennel workers and pet dog owners.
• Munford (1975) described 2 patients suffering from fever, chills and weight loss and other symptoms found to be caused by B canis. Serology and culture tests indicated that the family dog which was also infected was the most likely sources of their illness. Epidemiological studies using serology and culture tests indicated that 30% of local dogs were positive for B canis.
• Blankenship (1975) describes a 48 year old man who was shown to have B canis by bacteriological culture. The man had a history of cardiovascular disease and mild congestive heart failure but whether this could be ascribed to B canis which can give rise to these type of symptoms is not known. In December 1973 he presented feeling generally ill and had chills, fever (temperature 37.9 - 38.3°C), anorexia, and nausea and a pain at the back of his right foot. Over 3 weeks he lost 15 pounds weight. He had very limited signs of illness: no lymphadenopathy and only occasional hepatosphenomegaly. After 3 weeks he felt better and became afebrile without treatment however bacteriological culture on blood at this time showed a gram negative coccobacillis subsequently identified as B canis. The patient remained reasonably well until March 1974 except for brief relapses with fever and sweats and regained lost weight. However a bacteriological culture test was again positive in April 1974. He was then treated with tetracycline . A subsequent serology test for B canis (RSAT) was positive at 1:500. Further information is not given. It is suggested that the possible cause of the illness was a 2 year old mongrel dog which had produced two puppies 2 months before his illness started. One puppy was still born and the other died soon after birth. The bitch has a negative bacteriological culture test but a positive RSAT serology test at 1:200.
• Rifkin (1978) describes a 55-year-old woman who developed fever, abdominal pain (right upper quadrant), malaise, weakness, and anorexia eight weeks after her dog delivered stillborn pups. Her temperature was 39.4°C but she did not consider that she suffered from fever or chills. She had bilateral swollen inguinal lymph nodes. B canis infection was confirmed by a bacteriological culture test but serology tests (agglutinins) were negative throughout illness and convalescence, including during treatment with tetracycline and streptomycin. Interestingly she had a history of cholelithiasis which might indicate the that this was not an acute case but rather a relapse of a chronic case.
• Godoy (1979) describes the case of a laboratory worker infected with B canis.
• Ramacciotti (1980) describes the case of a veterinarian accidentally contaminated by pus from a bitch which had aborted. The vet developed charactoristic symptoms of brucellosis. He was treated with tetracycline but subsequently relapsed but recovered with further treatment. B canis was confirmed as the cause by bacteriological haemoculture.
• Polt (1982) describes 4 cases of human brucellosis caused by B canis .
  • Patient 1 was a 66 year old man who was hospitalised with night sweats, headache, chills, fever (temperature up to 41°), a distended tender and painful abdominal, jaundice, and an enlarged liver. Liver biopsy indicated granulomatous hepatitis. Liver enzyme assays were abnormal. X-ray examination indicated bilateral lower lobe pneumonia. B canis. He was treated with antibotics but relapsed after 5 months with similar symptoms. He was given further treatment (tetracycline/streptomycin) with a reduction in symptoms.
  • Patient 2 was a 28 year old woman who had previously suffered from headaches for 6 years. She was admitted to hospital with nausea, vomiting, anorexia, headaches, (lasting 2 to 3 days), 6 weeks duration of night temperature rises (up to 38.8°C). Physical examination found nothing unusual. B canis infection was found using a micro-TAT serology antibody test. She was treated by tetracycline but relapsed after 7 months later with bloody diarrhoea but apparently improved on treatment (tetracycline/streptomycin).
  • Patient 3 was a 31 year old man admitted to hospital with a sour throat, myalgias, lethargy, headache, cough and conjunctival burning of 3 weeks duration, night sweats (several months), Physical examination showed nothing unusual. B canis infection was diagnosed by micro-TAT. The patient partially improved and was discharged but the treatment with tetracycline (2 x 21 days treatment) was ineffective and he continued to have symptoms including low grade fever and lethargy. He was further treated with tetracycline/streptomycin but no further information is recorded. His dog was thought to be infected with B canis on the basis of swollen testicles and the author considered this dog as the possible cause of the infection.
  • Patient 4 was 41 year old women hospitalised with fever, a tender right upper quadrant with progressively increasing pain (cholecystitis), low grade fever, nausea, anorexia, a temperature up to 40°C spiking in the afternoon/evening. A liver scan suggested hepatocellular disease. B canis infection was determined by micro-TAT. The patient was treated with ampicillin and remained well for 18 months. The patient had 2 dachshunds which both had positive serology titres for B canis.
• Tosi (1982) describes the case of a 17 month old boy with B canis. The boy presented with a limp affecting the right leg (knee) which was swollen with a limited range of movement. The symptoms fluctuated from day to day. X-ray of the knee showed nothing but there was a suggestion of septic arthritis. A bacteriological culture test produced small gram negative rods later identified as B canis. Standard antibody serology tests for smooth Brucella were negative but a TAT titre for B canis was positive at 1:200. The child was treated with moxalactam for 20 days and the knee returned to normal. No further information on the subsequent progress of the child is given. It is noted that the family had a pet dog (a bitch) which had previously whelped unsuccessfully and this dog was considered a possible source of the infection however both bacteriological culture test on blood and vagina secretions were negative and serology antibody tests for B canis from this dog were negative.
• Schoenemann (1986) describes the case of a 30-year-old woman with B canis infection characterized by granulomatous hepatitis and splenomegaly. The source was an infected dog. She was treated with cotrimoxazole and streptomycin but relapsed twice. After further treatment she became symptom free for 2 years.
• Rumley (1986) describes a 46 year old male patient ill for 4 months with fever, chills, sweating, weight loss, malaise and backache. Hepatosplenomegaly, cardiac murmours, a swollen right lymph node (supraclavicular area) and joint abnormalities were absent. Prior to the start of the illness he had been in contact with a bitch which had aborted. General blood and urine tests showed no specific results to point to B canis. Bacteriological culture (blood) indicated B canis. ME-TAT was positive at 1:200. The symptoms were resolved by treatment with tetracycline and streptomycin.
• Shin (1999) described cases of B canis in humans caused by contact with abortion products and oestrus secretions. The disease was generally considered mild.
• Ying (1999) describes a man with B canis possibly acquired in Kawait which shows that B canis can cause severe complications in humans. A 49 year old man had symptoms of weight loss, heart murmur and fever (temperature at 38°C). Tests indicated fibrinous vegetation on the aortic valve and aortic insufficiency. The patient was treated by aortic valve replacement and with antibiotics and recovered. In other forms of brucellosis heart failure is the most serious complication causing 80% of deaths. Deaths occur in ≤2% of all cases.
• Wallach (2004) describes a case of a laboratory worker who contracted B canis strain (M-) due to exposure during his work in a laboratory. B canis strain (M-) is considered a low virulant strain in dogs but caused acute brucellosis in this patient including: recurrent fever, headaches, arthralgia, weakness and constipation. The patient was given antibiotics and became asymptomatic.
• Nomura (2010) describes the cases of 2 male workers in a pet shop in Nagoya, Japan. Patient 1 was a 71 year old man who developed fever and fatigue and a temperature up to 37.8°C. Physical examination was unremarkable. Patient 2 had similar symptoms except that he had a moderate liver dysfunction. B canis was identified by blood culture and combinatorial PCR in both cases. TAT titres were 1:1280 and 1:320 respectfully. Treatment used in one patient was doxycycline plus streptomycin and in the other it was doxycycline plus rifampin. Overall both were ill for about 7 weeks. The symptoms were considered mild in both cases and both apparently recovered. The likely cause was a bitch which aborted at the shop. 6 other dogs in the shop were positive by blood test and serology tests for B canis.

Epidemiological Studies of B Canis in Humans.

There are very few epidemiological surveys of B canis infection in humans and little is known as to how many people are or have been infected worldwide. The accuracy of the results of such surveys depends on the serology tests used. The results (particularly in early work) may be inflated by false positive results due to cross reactions with other bacteria (see above) but there may also be false negative results because serology tests do not find all cases particularly in longer term infections. Different tests give different results. Also cut off points for tests significantly affect the results. Fore example, TAT at dilution 1:50 finds far more positives than at 1:200 yet neither is more correct than the other. Therefore these epidemiological results are only a guide.

• Lewis (1973) found 0.4% of US military recruits gave positive titres for B canisas tested using TAT (≥ 1:100 titre).
• Hoff (1974) and found that 0.59% of Florida, USA residents had positive titres for B canis.
• Monroe (1975) carried out an extensive study of the presence of B canis antibody in humans based in Oklahoma, USA. Serology test were carried out on a range of groups considered to vary according to their exposure to dogs. The test used was a micro-TAT type method. The percentage positives in this survey are high compared to other surveys possibly because a low threshold at ≥1:12 titre used as a positive cut off. 153 new born babies had a positive rate of 5.7% (maternal antibodies) and 59 of their mothers (38.6%) were also positive (59 of 153 were tested). In 2026 members of the general public (These were: hospital patients, hospital employees, and blood donors) considered to have a range of every day exposure to dogs the positive rate was 67.8%. The positive results were: 32.2% at ≤1:16; ≥ 67.8% at 1:12; ≥ 41.3% at 1:24; ≥13.7% at 1:48; ≥ 2.9% at 1:96; and ≥ 0.6% at 1:192 at. More women showed positive results at 72.4% compared to 62.1% in men. This was ascribed to greater contact with pet dogs. Older people had less positives results than younger people. Among women the positive results were according to age ranges (years) were: 81.3% at 3-14; 81.1% at 15-24; 70.8% at 25-34; 69.7% at 35-44; 60.0% at 45-54; 57.9% at 55-64; and 42.5% at >65. According to Monroe a study by Harrison (1928) in relation to B abortus infections in humans has shown that age does not affect antibody response to Brucella. A possible explanation is that older people have had previous exposure but their antibody response has declined over time. In Table 73 veterinerians who were considered to be a high risk group are compared to the general public (men). There are higher titre in veterinarians and it is noticeable that there are more positives at higher titres. The highest levels at lower titres were present in patients with fever of an unknown origin. These positive results were: 19.6% at ≤1:6; ≥ 80.5% at 1:12; 56.7% at ≥1:24; 27.4% at ≥1:48; 11.5% at ≥1:96; 8.0% at ≥1:192; 5.3% at ≥1:384 at 5.3.
• Flores-Castro (1976) tested 500 dogs and 203 humans in Mexico City using TAT (≥1:100) and found that 28.0% of dogs and 13.3% of humans had positive titres.
• Varela-Diaz (1979) studied 1,065 persons in Corrientes Province, Argentina for B canis using gel diffusion test with a B ovis antigen and found 21 positive results (2%). In a separate survey in 887 rural school children in Neuquén Province they found 2 reactors 0.22%).
• Carmichael (1980) found 6 B canis positive sera out of 1915 humans (0.3%) tested in Germany.
• Poll (1982) used a micro-TAT test using a safranin-dyed B. canis organisms as antigen (cut-off ≥1:160) to test for B canis in 1,147 people and found positive results in 18 (1.57%).
• Boeri (2008) found that 16 of 219 dogs in Buenos Aires had antibodies to B canis. Only 3 of the owners of dogs that tested positive consented to a serological diagnosis and two of them were also had positive titres. This suggests a considerable health hazard.
• Lopez (2009) tested the owners of 5 dogs with either RSAT serology tests or on the basis of possible clinical signs of B canis. 2 of the owners tested positive by RSAT and iELISA suggesting a high incidence of transmission.

Antibody Titre.	Veterinarians.	General Public (Men).
Titre.	%	%
≤ 1:6	27.4	43.1
≥ 1:12	72.6	56.9
≥ 1:24	61.6	29.5
≥ 1:48	31.5	7.0
≥ 1:96	19.2	1.9
≥ 1:192	11.0	1.10
≥ 1:384	8.3	0.0

Brucella ovis (B ovis).

B ovis was first reported in New Zealand by McFarlane (1952) and Buddle (1958) . It was also identified in Australia by Simmons (1953) . There has been relatively little research into this disease and few epidemiological studies but it is probable that it occurs worldwide in sheep flocks. According to the OIE Manual (2009) and other sources the disease has been reported in South America (Brazil, Chile, Peru, Argentina and Uruguay), North American (USA, Canada, Mexico) and Europe (Germany, France, Spain, Italy, Czechoslovakia, USSR, Romania, Hungary, Croatia.), Australasia (Australia and New Zealand) and Africa (South Africa), but probably occurs in most sheep-raising countries. It appears that some breeds of sheep are more resistant than others. In countries where B ovis is endemic in sheep up to 60% of rams can be infected in individual flocks and up to 75% of all flocks, Meyer (1982) and Sancho (1985) . B ovis in sheep does not cause the level of stock losses and the significant economic losses associated with other forms of brucellosis such as B abortus in cattle or B canis in dogs. In particular loss of lambs due to abortion and stillbirths is more limited. Also B ovis is generally considered not to cause infection in humans which means that there is not need to eradicate it to safeguard human health. For these reasons countries with endemically infected flocks and individual sheep farmers have been reluctant to instigate expensive eradication programmes. Reductions in lamb production which might be caused by ewes being serviced by individual sexually unproductive ram may not be important if a flock is served by a higher ratio of rams to ewes. In such circumstances the main economic losses may relate to the replacement of infected rams. Despite these considerations eradication programmes for B ovis in sheep are generally desirable.

B ovis is a chronic disease in sheep and studies show that it is mainly a sexually transmitted disease. Chronic infections are known to continue indefinitely particularly in rams. The disease causes epididymitis of rams and infertility and genital and reproductive organ infections in ewes. Other internal organs may be affected in both sexes. Infected ewes can abort lambs or the lambs can be stillborn or show postnatal weakness. Flocks can become heavily infected. There are a number of methods of positively identifying the presence of B ovis infection including scrotum examination in rams, the bacteriological culture test on semen and serology tests but none of them work in all cases. Physical scrotal examination of rams is useful to initially identify an infected flock, but is unreliable in individual rams. A positive result from the bacteriological culture test on semen is an almost certain diagnosis of infection but the procedure only works in some rams. In other rams B ovis may be absent from semen or only present intermittently. The bacteriological culture tests on blood is not reliable due to the tendency of the bacteria to be localised within the genitalia or internal organs. Serology antibody tests are best way of identifying B ovis infection in flocks and therefore these are used in eradication programmes. They find the greatest number of infected sheep, however they are not totally reliable in individual animals which may give negative results while being proven as infected. Apart from sheep, only are red deer (Cervus elaphus) and white-tailed deer (Odocoileus virginianus) are known to be naturally susceptible but the disease. The disease is probably spread to deer by contact with domestic sheep. Laboratory animals such as rabbits, and guinea pigs can be infected by application of B ovis to their mucous membranes or by inoculation. There is only very limited evidence that humans can become infected by contact with infected sheep.

B ovis is a rough species or biovar of the genus Brucella. Buddle (1956) showed that B ovis does not possess smooth antigens of B melitensis and B abortus and Renoux (1955) showed that B ovis has a similar antigenic characteristics as rough B melitensis.

Signs and Symptoms of B ovis in Sheep.

According to the OIE Manual (2009) B ovis can be a clinical or sub-clinical disease in sheep. The main signs of B ovis infection in sheep are epididymitis and impaired fertility in rams and for ewes infrequent abortion and the birth of stillborn or weak lambs possibly with a vaginal discharge. It should be emphasised that not all infected rams have epididymitis. Buddle (1956) observed that B ovis causes impaired fertility in rams, abortion in ewes and neo-natal mortality in lambs. The principal clinical manifestations are lesions of the epididymes, and testes in the ram and placentitis in the ewe, with abortion or stillborn or post-natal death of lambs. It is possible that B ovis infection may go unnoticed in a flock of sheep if these signs of infection are limited or absent. During the first stage of infection with B ovis a sheep may have a elevated temperature and reduced appetite but this is unlikely to be noticed by the owner. In other ways behaviour appears normal.

Localisation of B ovis in the epididymis and testes of rams gives rise to a problem of low semen quality which can be observed by examination of the semen. Ewes served by such rams often fail to conceive and it can be noted that such ewes often conceive when they com back into season usually about 17 days later. This leads to a protracted lambing season with negative economic consequences. For infected ewes after the birth of a lamb an infected vaginal discharge may be present. This is usually infected by B ovis which can be confirmed by bacteriological culture test.

B ovis infection in Rams.

Epididymitis is the main sign of B ovis infection in rams. This can be identified by scrotal examination. Epididymitis can be unilateral or bilateral. The most notable form of epididymitis is enlargement of the tail of the epididymis at the lower end of the testes. However this is not conclusive proof of B ovis infection since not all infected rams show epididymitis and epididymitis can have other causes including the effects of other bacteria infections. Although epididymitis is a sign of B ovis infection in rams not all rams develop it. Blasco (1990) notes that in a survey of 267 seropositive rams only 125 (46.8%) showed any palpable testicular alteration and on the basis of histological examination he concluded that a significant proportion of B ovis infected rams do not have scrotal lesions. If B ovis infection is suspected further information can be gathered by the examination of semen. The viability of sperm decreases in infected rams and effects include decreased mobility and concentration together with morphological changes. Poor semen quality may occur in rams with B ovis infection irrespective of palpable epididymitis being present. These changes have been characterised by a number of workers. Gunn (1942) showed that the sperm of rams affected by B ovis had disrupted morphology The main fault was separation of the head and tail. McGowan (1960) studied 29 rams and showed that 15 had semen with no sperm mobility, 7 had low mobility and 7 intermediate mobility. Cameron (1976) noted that on average the output of sperm from infected rams was lower than from non-infected rams. The result of epididymitis and testicular lesions in rams is reduced lambing success although this can be masked if multiply rams are used. Laurence (1958) collected evidence showing that generally rams infected with B ovis had decreased fertility. This relates not just to a single mating but also a further reduction in the viability of semen in the course of the normal repeated mating behaviour by rams.

The pathogenesis of B ovis infection of the testes and epididymis of rams has been described by Blasco (1990) . The initial localization of B ovis in the epididymis and testes involves perivascular edema and infiltration of thr peritubular by lymphocytes, monocytes and neutrophils. This is followed by papillary hyperplasia and local hydropic degeneration with formation of intraepithelial cysts. The destruction of epithelium leads spermatazoa leakage and formation of spermatic granulomas which can block the epididymis with subsequent testicular degeneration and fibrosis.

'B ovis Infection in Ewes.

B ovis infection in ewes has not been studied to the same extent as that in rams. Abortion and stillbirths or the birth of weak lambs are the main signs associated with B ovis infection in ewes. There may also be failure to conceive and embryonic resorption. Rates of abortion of 1 to 2 percentage have been recorded in flocks but for experimental infections rates of up to 8% have seen. However according to Bulgin (1990) B ovis infection in sheep flocks causes a reduction in the lambing percentage up to 30% in newly exposed flocks, and 15% to 20% in flocks in which it is endemic. These problems are most likely to occur at the first pregnancy of the ewe and are infrequent in later pregnancies, Hartley (1955) and Clapp (1962) . Although B ovis is known to be a chronic disease in rams because of obvious epididymitis and scrotal atrophy, there are usually no obvious signs of persistent infection in ewes. Buddle (1955) considered that active infection was less persistent in ewes than rams. He studied 26 ewes which had B ovis infected placenta at their first post-infection pregnancy and found that only 3 had diseased placenta at the subsequent pregnancy. Snowden (1958) studied 91 seropositive ewes in flock of ewes in Australia and found that most became seronegative over 3 years. We should treat these ideas with some caution since we know that chronic brucellosis can occur with negative serology titre in many species of animal such as B abortus infection in cattle. Also absence of B ovis from a specific organ such as the placenta does not mean that it is not localised elsewhere. Further, in most other animals species abortion occurs mainly at the first post infection pregnancy and less so at later pregnancies but this cannot be taken as proof of the absence of sub-acute or chronic disease.

Muhammed (1975) inoculated B ovis onto the vagina of 20 ewes in early pregnancy. The ewes were killed at intervals and examined by bacteriological culture and CFT. B ovis was recovered from the mammary and inguinal lymph nodes after 29 days but not the uterus, foetus or foetal membranes. It was recovered from the blood of 9 ewes intermittently from 22 to 91 days but B ovis was not recovered from blood, or any other tissue examined, after 98 days of infection. Grillo (1999) and Marco (1994) showed that B ovis could be recovered the uterus of seropositive ewes. It could also be present in the mammary gland, spleen, and the iliac, cranial, scapular, prefemoral and mammary lymph nodes.

The pathogenesis of B ovis infection in ewes is not well understood. Ewes are most likely to abort if exposed to B ovis in early or mid pregnancy rather than before or in late pregnancy, Ris (1970) and Thoen (1986) . According to Collier (1964) and Meyer (1982) B ovis becomes localized in the placenta and reaches the fetus through chorion vessels. Placentitis causes fetal malnutrition and therefore weak lambs.

Transmission of B ovis in Sheep.

In General.

General information indicates that B ovis infection in sheep is largely a sexually transmitted disease mostly from ram to ewe but also from ram to ram and from ewe to ram. Lambs can be infected intra-uterus or from mothers milk. Environmental contamination as a route to infection seems unimportant. In early work because epididymitis in rams is the obvious manifestation of the infection most work concentrated on the role of rams in spreading the disease however it is now known that ewes can also play an important part. Infection seems to occur readily at mucous membranes including the mouth, conjunctiva, prepuce and vagina such as by contact with infected semen or vaginal secretions. According to the OIE Manual (2009) modes of infection among sheep are: sexual transmission: ram to ewe via semen; ram to ram via semen; ewe to ram via vaginal secretions; and ewe to lamb via milk. They consider that the role of ewes in transmitting B ovis is significant.

The likelihood of rams being infected increases with their age and is probably related to the duration of sexual activity, Walker (1986) , Murray (1969) and Bulgin (1983)

On the balance of observations some breeds are more susceptible than others. Probably Spanish and Merino type breeds are more resistant than other European breeds, Marin (1986) , Ficapal (1998) and De Wet (1984) < De Wet JA and Erasmus JA. Epididymitis of rams in the central and southern districts of the Orange Free State. J S Afric Vet Assoc. 1984. 55. 173-</ref>. According to Blasco (1990) the probability of infection depends on the route of infection, the infecting dose and the breed and age of the sheep. Most outbreaks occur immediately after the breeding season.

Experimental Transmission of B ovis in Sheep.

A number of trial experiments of transmission have been carried out in an attempt to determine the modes of transmission of B ovis in sheep. In early work most emphasis was given to the disease in rams. Buddle (1953) and Hartley (1954) have shown that rams can be infected experimentally via intravenous, subcutaneous and intra-testicular routes and also via the mucous membranes of the mouth, conjunctiva and genitalia. Blasco (1990) notes that possible experimental routes of infection are: orally, intravenously, intratesticularly, conjunctivally, intrapreputially, subcutaneously, through skin lesions, intrarectally and intranasally. For artificial infection the best results are obtained conjunctivally and/or intrapreputially, Afzal (1986) . A typical dose size to initiate 100% infection rate is 5 x 10⁸ to 10⁹ CFU, Fensterbank (1982) and Blasco (1987) .

Biberstein 1964 carried out work on the pathogenesis of B ovis in rams. The rams were inoculated by the conjuncival route and then killed at intervals over 255 days. After 10 days B ovis was recovered from lymph nodes and after 30 days from semen and the reproductive organs with associated histological lesions. B ovis was also found to be intermittently present in the liver, spleen, lung and kidneys. This demonstrates that the course of B ovis infection in sheep is similar to that of other forms of brucellosis such as B abortus in cattle. It can be concluded that B ovis infection starts at a mucous membrane migrates to a local lymph node followed by a break out to give bacteremia and thereafter transport around the body and localisation in the bacteria in bodily organs particularly the reproductive tract.

Fensterbank (1982) and Blasco (1987) have shown that in experimentally infected rams: B ovis can be found in the liver, spleen, kidneys, testes, epididymis, vesicular glands, bulbourethral glands, and ampullae, iliac, prescapular, precrural, submaxillary, parotid and retropharyngeal lymph nodes. Although the genitalia are the main target for infection in some animals the bacteria was not present in these organs but was present in the spleen and iliac lymph nodes.

Ram to Ewe Transmission.

B ovis appears to be a largely sexually transmitted disease and therefore ram to ewe transmission occurs via infected semen and this is probably the most important vector for the spread and maintenance of the infection in flocks of sheep. Buddle (1955) and Hartley (1955) noted that active infection could developed in ewes following mating with naturally infected rams. However he considered that ewes seemed more resistant to active infection than rams. This also seems to apply in pregnancy, Ris (1970) and Collier (1964)

'Ewe to Ewe Transmission.

There is little evidence that can ewes transmit B ovis to each other. Buddle (1955) pastured pregnant ewes with ewes artificially infected with B ovis. He found no evidence of transmission to the non-infected ewes.

Ewe to Ram Transmission.

Transmission of B ovis from infected ewes to uninfected rams is considered to be an important vector for the spread of B ovis in sheep. This can happen in three ways:

• Firstly, infected ewes can infect uninfected rams during mating. Hartley (1955) showed that rams mating with ewes known to be infected with B ovis could become infected. This has also been observed by Ryan (1964) who reports the case of a ram which became infected from contact with a ewe which had failed to conceive when mated to an infected ram four months previously.
• Secondly, it is also likely that rams can become infected intra-nasally because rams indulge in sniffing and nosing of the external genitalia and perineum of B ovis carrier ewes, Plant (1986) .
• Thirdly, a ram free from B ovis can become infected by contact with a ewe which have recently mated with a B ovis infected ram. It is considered that this is caused by infected semen in the ewe's vagina. This has been demonstrated by the work of Van Rensburg (1958) .

Ram to Ram Transmission.

Ram to ram contact seems to be a significant means of spread of the disease particularly in the mating season. This is not apparently derived from a shared environment but rather from direct contact. Thus Buddle (1956) notes that natural transmission of infection from ram to ram has been shown experimentally but that it also occurs readily during the mating season, and also when rams are run together in isolation from ewes. A number of ideas have been put forward to explain this:

• It has been suggested that this might be related to trial mating behaviour in which rams mount each other with possible contact of infected semen with the anal region, Jebson (1954) .
• Another possibility is the infection occurs because rams may sniff and lick infected semen from another ram, Keogh (1958) . According to Clapp (1962) non-infected rams can become infected if they are housed in pens previously occupied by B ovis infected rams where a likely vector is spilled semen.
• More recently, a possible explanation has been suggested by the OIE Manual (2009) . It is suggested that the transmission of B ovis between rams is related to their mating rituals. It is normal for rams to compete for dominance of a flock of ewes by head to head trials of strength and it has been observed that rams which have lost such contests lick the prepuce of the dominant ram as an act of submission and therefore can be in contact with infected semen.

Vertical Transmission.

The reproductive organs including the uterus and placenta of a ewe can become infected with B ovis and also infected ewes can secret B ovis in their milk. Therefore lambs can become infected in the uterus or from infected milk.

According to Thoen (1986) the inflammatory responses of B ovis in infected foetuses is similar to those caused by B abortus in bovine foetuses. Hartley (1961) investigated foetal death in ewes infected with B ovis during pregnancy. He concluded that during the bacteremic phase B ovis localises in the uterus. The main effects are: invasion of the chorionic epithelium with subsequent necrosis, infection of the foetal cotyledons and extensive placental lesions occur which can cause foetal death and abortion usually in the last trimester or the birth of weak lambs, Lawrence (1961) . Hughes (1972) considered that for B ovis infection in sheep the death of at term weak lambs may be more common than abortion. Grillo (1999) artificially infected 40 ewes during pregnancy by the conjunctival route. 3 lambs were still born and 11 died after birth. 46 survived and appeared healthy. He considered that most ewes excreted 'B ovis at lambing and during lactation. Subsequent pregnancy seem to occur normally but the ewes can excreted B ovis in their milk during the second lactation. There is some controversy as to the extent to which lambs can be infected by their mother. According to Meyer (1982) lambs with infected ewe mothers seldom develop active infection even if they drinking infected milk. However there is evidence that some lambs are infected. The presence of B ovis has been demonstrated in a stillborne ram lamb by Hughes (1972) . He showed that B ovis could be cultured from the testes of still born ram lambs born to infected ewes. Rickard- Bell (1963) demonstrated that ram lambs born to ewes that had been mated to B ovis infected rams had a 5% incidence infection 14 months later.

Skin Abrasions.

It is considered that physical abrasion of the skin is a possible means of transmission of B ovis, Blasco (1990) . Clapp (1962) considered that wounds caused by shearing could be a possible means of spreading the infection.

Environmental Transmission.

Environmental contamination is an important means of spread of B abortus, B melitensis, B suis and B canis infection. However, this does not seems to be the case for B ovis. Trials have been carried out involving both rams and ewes in which they were deliberately put in contact with a contaminated environment and these trials have shown that both rams and ewes are very resistant to this possible route to infection:

• Buddle (1955) grazed B ovis free ewes on pasture previously grazed by infected ewes during the latter's lambing but none of the non-infected ewes became infected.
• Hartley (1955) grazed rams in a paddock previously occupied for lambing by infected ewes in which infected residues were considered to be present but none of the rams became infected.
• Keogh (1958) carried out a trial in which infected rams and uninfected rams were alternated on the same pasture every other week for one year but none of the uninfected rams became infected.

Diagnosis of B ovis.

Physical Examination of Rams.

Scrotum Examination. Epididymitis in rams is an obvious possible sign of B ovis infection in sheep. Physical examination of ram's testes by a trained veterinarian can reveal the presence of epididymitis but the procedure is only a guide to possible B ovis infection, Edgar (1959) . The veterinarian carrying out scrotal examination looks for changes in the shape, size and consistency of the epididymis, Gouletsou (2010) . Epididymitis in rams caused by B ovis can be unilateral, bilateral or intermittent. Only about 35% of rams in an infected flock develop physically discernible evidence of epididymitis, Hughs (1984) . Even in a B ovis infected flock up to 50% of cases of epididymitis may be from other causes. Also rams without epididymitis can give positive bacteriological culture tests for B ovis in semen and even more rams have positive serology tests. In some rams palpable epididymitis only occurs for a short period even though later histological examination indicates the presence of B ovis lesions in the testes and the epididymides, Blasco (1990) . In general there is not a good correlation between epididymitis and laboratory test results for B ovis. According to Hughes (1968) the presence of epididymitis in the rams of a flock of sheep can be used as a possible guide to the presence of B ovis but is not useful in individual animals,.

Medical conditions other than B ovis infection can cause epididymitis in rams and this has been described by various researchers. These include: Bulgin (1983) , Burgess (1981) , De Long (1979) , Ekdahl (1968) and Livingstone (1964) . The microorganisms which can cause epididymitis in rams include: include Actinobacillus seminis, A. actinomycetemcomitans, Histophilus ovis, Haemophilus spp., Corynebacterium pseudotuberculosis ovis, B. melitensis and Chlamydophila abortus. Epididymitis can also be caused by sterile, trauma-induced spermatic granulomas.

Examination of Semen. In possible cases of B ovis infection semen quality can be used as a guide to the fertility of rams or lack of it and can show possible signs of infection. Some infected rams retain a significant level of fertility. Methods of determining ram semen quality have been described by Tsakmakidis (2010) . The presence of B ovis in semen has been described by Webb (1980) . They noted that semen samples from infected rams could show the presence of coccobacilli. They artificially infected 10 rams by preputial inoculation and examined semen samples at weekly intervals. All rams developed lesions at 8 weeks after inoculation. Inflammatory cells in semen were present at 8 weeks in all rams and subsequently 86% were positive for B ovis. 52% of semen smear samples showed likely B ovis coccobacilli after 4 weeks and B ovis was bacteriologically cultured from 97% of samples after 5 weeks. However there is difficultly in distinguishing between B ovis and B melitensis and/or Clamydia spp which could also be present. However immunofluorescence can be used to identify B ovis, van Drimmelen (1963) and Ajai CO (1980) . Inflammatory cell possibly present in B ovis infected cells include, neutrophils, monocytes and desquamated epithelial cells.

The method of microscopic examination of semen or smears is based on the Stamp's modification of the Ziehl-Neelsen method and can be useful for a presumptive diagnosis. Brucella species are not truly acid-fast, but they are resistant to decolourization by weak acids, and stain red against a blue background. Brucella are coccobacilli or short rods which can be present singularly, in pairs or in small groups. This test is limited because other microoganisms such as Chlamydophila abortus and Coxiella burnetii give similar results to Brucella in this test. Also it cannot distinguish between B ovis and B melitensis.

The Bacteriological Culture Test.

Bacteriological culture tests can be used to isolate and identify B ovis from semen rams and from the milk and vaginal discharges of ewes and also from tissue samples from rams, ewes and lambs. In rams shedding of B ovisin urine can also occur. For B ovis, the bacteriological culture tests on blood are not usually successful and therefore are of limited usefulness. This is probably because B ovis becomes localised within organs. In practice most test are carried out mainly on ram's semen and sometimes on ewe's vaginal discharge. A positive result with this test is a virtually certain diagnosis of B ovis but does not find all infected rams. Also tests on individual rams at different times can give different results either positive or negative.

• Buddle (1956) notes that the organism may be recovered readily from the semen of naturally and experimentally infected rams even in the absence of palpable lesions of the genitalia and from the epididymes, testes and regional lymph nodes and from the placentas and mammary secretion of infected ewes as well as from the tissues of aborted lambs.

• According to OIE (2009) the preferred tissues to collect in rams at necropsy are the epididymis, seminal vesicles, ampullae and inguinal lymph nodes and for ewes the uterus, iliac and supra-mammary lymph nodes. In aborted or stillborn lambs, the best sites to culture are the stomach (abomasal) contents and lung.

• Worthington (1985) studied rams naturally infected with B ovis over 14 months. Most rams remained chronically infected and excreted the organism in their semen throughout the investigation. Bacteriological culture tests was positive in 87.9% of semen samples but the CF test was negative in 11% of the infected rams. At necropsy B ovis was isolated from the epididymides and related sexual glands. In one case it was also isolated from lung, spleen, kidney and iliac lymph nodes.

Methodology. A number of authorities have recorded methods for the bacteriological culture test. A typical method for the bacteriologic culture from ram semen has been described by Myers (1973) . It is necessary that semen is collected under as aseptic conditions as possible including disinfection of the animals genitals. In this method, the samples were inoculated onto Brucella Albimi agar plates enriched with 10% rabbit serum . They were incubated at 37°C in an atmosphere of 10 - 20% carbon dioxide. The plates were examined daily for 2 weeks. General methodology for the collection, examination of samples and methodology for bacteriological culture and typing of B ovis are described in the OIE Manual (2009) .

Serology Tests for B ovis.

The main tests used in B ovis serology testing are the complement fixation test (CFT), the gel diffusion test (AGID) and the indirect enzyme linked immunosorbent assay (iELISA). Other tests are used occasionally but not routinely. Combinations of tests such as iELISA plus AGID usually provide more accurate results than single tests.

Antigens for Use in Serology Tests.

B abortus, B melitensis and B suis are the most studied of the Brucellae spp. These are termed smooth bacteria because of their appearance under the microscope. This is related to the smooth lipopolysaccharide (S-LPS) in their surface layer. B ovis is termed a rough species and has different 'rough' lipopolysaccharide (R-LPS) in its surface layer. B melitensis although mainly smooth can switch to rough form in some circumstances but current information suggests that B ovis is permanently in a rough form. Therefore tests developed from smooth Brucellae such as antigens derived for B abortus etc containing S-LPS do not work when used to test for B ovis. However serology tests do work with antigens which have been derived from B ovis itself or other rough Brucella such as B canis and B abortus 45/20 vaccine. Methods of preparing antigens are given in detail by Blasco (1990) . Various antigens have prepared and tried as follows:

• Sonically Disrupted B ovis Antigen (SD antigen). This is a process which substantially degrades the B ovis cells to give a mix of antigens. These include surface antegens such as R-LPS and OMP and also antigens from inside the bacterium cell. Preparation involves sonic disruption of acetone killed Brucella for 2 hours followed by dialysis against water for 2 days at 4°C. After centrifugation, the supernatant which contains the antigen is concentrated for use. Cross reactions with smooth Brucellae are possible. Diaz (1967) compared sonically disrupted antigens derived from B ovis, smooth B melitensis and rough B melitensis B115 by immunoelectrophoresis against homologous antibodies in rabbit serum and found that 9 to 11 antigens were present. There was extensive overlap of antigens in all cases.

• Organic Solvent Extracted B ovis Antigen (PCP). This antigen is extracted using a mixed solvent of petroleum ether/chloroform/phenol. The technique is based on similar work used to extract LPS from with enterobacteria, Galanos (1969) . This method yields R-LPs free from protein or nucleic acids. This antigen is widely used but has the problem that it is highly hydrophobic which limits it use in some tests. The general methods is as follows: B ovis cells are suspended in a petroleum ether/chloroform/phenol mixture (Volumes of 80:5:2), homogenize and centrifuged. The process is repeated (3 x in all) and the supernatant is collected in each case and pooled. The pet ether/chloroform is evaporated and the antigen precipitated from the residual phenol solution by careful addition of water and then purified.

• Hot Saline Extracted B ovis Antigen (HS). The antigen prepared by this method is water soluble but is less pure than the solvent extracted antigen and includes outer membrane protein (OMP) and other components. There is an overlap between these latter antigens and some occurring in smooth Brucellae, Gamazo (1989) and Santos (1984) . This antigen is widely used in serology tests but the results must be interpreted with care if there is the possibility that other Brucellae such B melitensis are also present. The method of preparation is as follows: A suspension of B ovis cells in saline solution are autoclaved for 30 minutes, cooled, centrifuged and then purified. The supernatant is collected and purified by dialysis (or similar) with water. Methodology for the preparation of this antigen is described by the OIE Manual (2009) .

• Cold Saline Protein Extract (CS antigen). This antigen is suitable for the allergic skin test and is essentially free from R-LPS. B ovis cells are suspended and stirred in saline solution at 4°C for 72 hours. The protein antigen is recovered in the supernatant after centrifugation and is then purified by dialysis (or similar) against water.

Seroconversion.

If rams are artificially infected with B ovis seroconversion occurs at about 2 -5 weeks and B ovis is detected in semen at about 4 - 9 weeks, Laws (1972) and Webb (1980) . Afzal experimentally infected 6 rams with B ovis. Antibodies were detected after 11 days using iELISA. The titre of IgM and total antibodies rose up to 33 and 41 days respectively then levelled off.

The Compliment Fixation Test (CFT or CF test).

The CF test has been widely used as a serology test for the identification of brucellosis in animals. It has been adapted for use in the identification of B ovis infection in sheep by Clapp (1955) . This method had problems related to anti-complement activity and sensitivity. Improved methods were developed by Biberstein (1958) and Clapp (1961) . Clapp used whole cell suspensions of B ovis but there were problems with this antigen including lack of sensitivity and anti-complement activity. Biberstein (1958) produced an antigen by sonic disintegration of B ovis and with this antigen he reported a good correlation between CFT and the culture test, however in this case there was cross reaction with other Brucella species, Diaz (1968) and Leong (1968) . Since sheep can be infected with both B ovis and B melitensis this is a particular problem. This cross reaction is caused by the presence of antigens from inside the cell rather than the R-LPS from the cell wall. Myers (1972) developed a method of extracting an antigen from B ovis by hot saline extraction which was mainly R-LPS. This HS antigen is widely used in the CF and other tests and has been shown to detect B ovis but not the smooth forms of Brucella including B melitensis. This antigen can also be used to detect antibodies to other rough form Brucella such as B canis and B abortus vaccine 45/20. There is also less anti-complement activity with this antigern. Similar antigens which can be used for the detection of B ovis can also be prepared from B canis and 45/20 vaccine.

This is the general procedure of the CF test: All test reagents are carefully standardised prior to use so that they can be used at standard concentration.. These are: red sheep blood cells (SRBC) sensitised with rabbit anti-SRBC, the HS antigen and guinea-pig complement. Barbital (veronal) buffered saline (VBS) is the chosen buffer/diluent. The natural complement in sera samples is deactivated by heating it at 60°C for 30 minutes. A range of sera dilutions are made in VBS buffer to which is added the HS antigen and guinea-pig complement. Known sera are also tested at the same time as standards. All these solutions are then incubated at 37°C for 30 minutes (Hot incubation) or overnight at 4°C (Cold incubation). Then the sensitised SRBC suspension is added followed by further incubation for 30 minutes at 37°C for 30 minutes. The samples are centrifuged to remove unlysed cells. The positive result is represented by the highest dilution with 50% haemolysis. Fuller description of the method is provided by the OIE Manual (2009) and LQWG (2007) .

Problems associated the CF test as applied to detection of antibodies to B ovis in sera as follows:

• Anticompement activity occurs and is particularly related to contamination and haemolysis of blood samples, Worthington (1982) . According to Clapp (1961) this problem is associated with the presence of whole bacteria and cellular debris contaminating the antigen.

• For some rams, the CF test gives low fluctuating titres. An example is supplied by Webb (1980) . 10 rams were artificially infected with B ovis by preputial inoculation so that all were found to be infected by about 5-8 weeks on the basis of the bacteriological culture tests on semen. Tests using CFT at 9 weeks after infection indicated: 6 rams with positive results, 3 rams with suggestive results and 1 ram with unreliable titres whereas their results by the bacteriological culture on semen was positive.

• False negative results have been describes as occurring in chronic infection in some rams, Worthington (1982) .

• Variations in sensitivity occur depending on the method used including: the effects of different antigens, the variation between hot or cold methods and different methods and times of incubation. HS extracted antigen is the most generally used antigen rather than the SD or the PCP antigens. Cold incubation (4°C / 16 hours) is considered more sensitive than hot incubation (37°C / 40 minutes), Burgess (1982) .

• The CF test is affected by the prozone effect which gives rise to false negative results (See also B abortus).

• There can be some overlap of antibodies between B ovis and B melitensis Rev 1 vaccine. This makes interpretation of the CF test difficult if the vaccine is being used.

The Gel Diffusion Test (AGID).

In general CFT is the most widely used test and AGID is often used as a confirmatory test. It is also used as a confirmatory test for iELISA. AGID measures mainly IgG. The technique of interacting antibodies with antigens within a gel was originally explored by Ouchterlony (1949) . Matthews (1967) demonstrated a technique using gel diffusion to detect antibodies to B ovis as precipitated lines. Myers (1970) developed a gel diffusion method for diagnosing B ovis in rams and considered that it gave results similar to CFT (92.3% agreement). He considered that because of the technical difficulties associated with the CFT, AGID was which is simple to use is more suitable for field work. The antigen required for the test was derived from B ovis. The HS antigen is usually used, Myers (1970) . The PCP antigen is too insoluble for this test.

The following are the main points of the test procedure. A standard 75 x 50 micro-glass slide is coated with 1.25% agar dissolved in buffered saline containing 1:10.000 merthiolate as a preservative to a depth of 2.5 mm. In a typical procedure 3 mm diameter holes are punched in the gel at 3 mm apart with a central hole for the antigen surrounded by 6 holes in an hexagonal pattern. Test sera and positive control sera standards are inserted individually into the 6 holes. The humidity must be controlled throughout the test procedure to prevent the gels drying out. The HS antigen is usually used and should be standardised against known positive sera samples, Blasco (1990) . Incorporating 5% salt in the gel aids the formation of precipitation lines, Jones (1975) . Typically about 0.1 ml of serum and antigen are used. After introduction of the reagents the gels are held for 24 to 48 hours at room temperature. Positive results are indicated by the formation of precipitation lines between the antigen and the sera holes. Faint precipitation line corresponding to minor antigen present in HS antigen may also be present. These also occur for B melitensis and should be taken as indicating of a positive result. The precipitation lines are analysed against a dark background using a hand lens. Serum corresponding to high CFT results produced precipitation bands within 24 hours but serum corresponding to low CFT values can take up to 72 hours. Incubation time do not affect the results. See also the OIE Manual (2009) .

The Indirect Enzyme Linked Immunosorbent Assay (iELISA).

iELISA is a relatively modern test and is used as an alternative to CFT for the surveillance of B ovis in flocks of sheep. It is considered more sensitive but less specific than CFT but is not affected by anti-complement activity. The general view is that it gives slightly better results than CFT or AGID. iELISA can find positive results where CFT fails Its used for the determination of antibodies in sera to B ovis has been described by a number of workers. Ris (1984) , Spencer (1984) , Walker (1985) Walker (1985) and Cho (1987) .

Combinations of tests can be useful. Both iELISA and AGID can be used as secondary results to clarify false positive and negative CFT results. CFT is often used because there are existing laboratory facilities but iELISA and AGID are more simple to use. Hilbink (1993) notes that the CF test can show what are considered false positive results in accredited flocks. These are mostly occur as low titres. iELISA or AGID can be used to help clarify these results. West (1993) showed that iELISA or AGID can demonstrate proven positive results where CFT is negative. He describe 4 rams diagnosed as negative for B ovis by CFT but positive by iELISA and by AGID on semen. On necropsy, 1 ram had gross evidence of epididymitis, the other three had histological lesions of epididymitis and all 4 had a seminal vesiculitis.

As with other tests that choice of antigen is important in iELISA . The SD, PCP and HS antigens have been tried. Afzal (1984) used PCP antigen in an iELISA to test rams in relation to ram epididymitis caused by B. ovis infection. He concluded that iELISA-PCP gave better results than other serology tests and less false positives than iELISA-SD. Blasco (1990) considered that sensitivity and specificity depend on the antigen used. In a comparison of SD, PCP and HS antigens he considered that the HS antigen was the most sensitive and useful however there is cross reaction with positive results for B melitensis infected sheep. iELISA-PCP antigen also gives good results and there is less overlap with B melitensis. The cross reactions associated with the HS antigen in relation to B ovis and B melitensis antibodies in sheep has been noted by Gamazo (1989) and Riezu-Boj (1986) .

The methodology for the use of iELISA for the detection of B ovis in sera is described by the OIE Manual (2009) . The principle of the iELISA test is that B ovis antigen is coated onto the wells of a polystyrene micro-titration plate. Serum is added and if B ovis antibodies are present they attach to the antigen. The presences of the bound antibodies is detected using a conjugated which attaches both to the antibodies and a dye. The intensity of colour development indicates the concentration of antibodies present. A simplified typical method is as follows. The reagents used is the HS B ovis antigen in carbonate buffer at pH 9.6 The sera samples and positive and negative sera standards are suitably diluted in PBST buffer at pH 7.2 The conjugate is either horseradish peroxidase plus anti-ovine IgG (H + L) or horseradish peroxidase plus recombinant protein G, in both cases in ABST buffer. The chromogen is ABTS (22'-azino-di-3-ethylbenzthiazoline) and the colour is developed with hydrogen peroxide. HS antigen is coated into the wells of a polystyrene 96 well flat-bottoms ELISA microtitration plates and then incubation for 2 hours at 37°C. The plate is then washed free of unbound reagent with PBST. Test sera and positive and negative control sera are then added to different wells and the plate incubated for 1 hour at 37°C then washed clear of free reagent using PBST. The conjugate is the added and the plates further incubated for 1 hour at 37°C then washed free of reagent. Finally the chromogen, ABTS, is added and the plates incubated for 1 hour at room temperature. The colour is developed using hydrogen peroxide and read at 414 nm. The difference between positive and negative results for iELISA depends on the use of a cut off point which is based on predetermined standard.

The Immunofluoresence Test.

Cox (1977) described a immunofluoresence test which he tried on affected and non-infected rams. He considered that it gave results similar to CFT and better than AGID. Ajai (1980) used immunofluoresence to identify B ovis in culture smears from ewes and concluded that it could be used as an early diagnosis in sheep.

The Counterimmunoelectrophoresis Test (CIEP).

Myers (1972) examined a immunoelectrophoresis method using saline-extracted antigens of B. ovis and B. melitensis and demonstrated that they showed distinct mobility patterns. Myers (1979) also used CIEP to test sera from naturally and artificially infected rams and considered it gave better results than AGID and CFT using the same antigen. CIEP is quicker than AGID and does not have the technical difficulties of CFT. This test has not been widely used and requires a higher level of laboratory competence than AGID but has a higher potential.

Allergenic Skin Test.

The presence of a present or past infection by B ovis in sheep can be detected by a delayed type hypersensitivity (DTH) reaction using a skin test, Jones (1975) . A soluble protein antigen should be used because LPS antigen elicits a humeral antibody response, Jones (1973) . The protein antigen also produces a response in sheep which are infected with B melitensis and sheep vaccinated with B melitensis Rev 1, Fenserbank (1985) . According to Jones (1975) for sheep infected with B ovis there is no correlation between allergic and serological or bacteriological culture test results.

Polymerase Chain Reaction (PCR).

At the present time PCR is being adapted of use to test for brucellosis in animals and humans and is producing interesting and useful results. PCR is a direct method and can be considered a possible replacement method for the bacteriological culture test and serology tests such as CFT, AGID and iELISA. It can also be used to distinguish between B ovis and B melitensis. Possible methodology has been described by Bricker (1994) , Manterola (2003) , Hinić (2008) , Bounaadja (2009) and Xavier (2010) .

Comparison of the Serology Tests.

There have been many comparisons and evaluations of the usefulness of the serology test used to detect B ovis. In general iELISA is considered more sensitive and more specific than CFT. CFT is widely used because there are existing laboratory facilities. None of these tests are reliable in individual animals. Combinations of tests find more positive results than any single test. The CF test is the most complicated to use and most prone to technical error. The following are comparative studies of the tests:

• Myers (1973) compared the results from the clinical examination of rams, the bacteriological cultural tests, the CF test, and the AGID test and concluded that AGID was probable the best test with 89.7% correlation with CFT. In general he was considered that:
• The bacteriological culture test can find cases not demonstrated by clinical examination.
• The CF test gives positive results for B ovis in rams not found by the bacteriological culture test.
• AGID give similar results to the CF test with possibly more viable positive results. Research indicate that AGID and CFT measure IgG, Myers (1972) .

• Jones (1975) and Myers (1970) and (1973) concluded that CFT-HS and AGID give similar results.

• Rahaley (1983) compared iELISA to CFT for detecting B ovis antibodies in the sera of 220 rams and considered that the iELISA test was more sensitive. It found 11 positives from known infected or vaccinated rams which were negative by CFT. iELISA titres were higher than those of CFT.

• Spencer (1984) compared iELISA to CFT for 16,527 sera samples and concluded that iELISA was more sensitive and more specific to B ovis antibodies. iELISA is able to detect both IgG1 and IgG2 but CFT only measures IgG1.

• Ris (1984) compared iELISA to a cold CFT and concluded that iELISA distinguished better between positive and negative samples and was easier to use.

• Worthington (1984) compared CFT, iELISA and AGID for the diagnosis of B ovis in 109 rams which were shedding B ovis in their semen. The sensitivity as positive results were: CFT, 96.3%; iELISA, 97.2%; AGID. 91.7%. In 141 rams from non-infected flocks the specificities as negative results were: CFT, 99.3%; iELISA, 98.67%; AGID, 100%. He used a trial on 285 rams from infected flocks of which 38% of the rams shed B ovis in their semen to determine predictive values for the three tests. Predictive values of positive tests were: CFT, 75.5%; iELISA, 66.7%; AGID, 72.5%. Predicative values of negative tests were: CFT, 97.1%; iELISA, 97.6%; and AGID 93.8%.

• Lee (1985) tested serum from B ovis infected flocks and non infected flocks. iELISA correlated with results by CFT and bacteriological culture test on semen but iELISA showed more positives than CFT. CFT failed to detect some rams excreting B ovis in semen.

• Worthington (1985) studied 42 rams for 14 months. The rams were chronically infected with B ovis and 87.9% shed B ovis in their semen throughout the study. 11% of 686 sera samples were negative by CFT but 76% and 77% of these were positive by AGID and iELISA respectively. The results for all three tests were erratic in 5 rams which did not shed B ovis in their semen. The five rams were from heavily infected flocks.

• Cho (1987) compared the iELISA to the CF test. In a study of 79 rams from which B ovis had been isolated: iELISA showed 75 positives and 4 suspects while CFT showed 64 positives, 2 suspects and 2 negative results. In another study of 176 rams considered to be uninffected by B ovis: iELISA found 175 negatives and 1 suspect whereas CFT showed 167 negative, 7 suspect and had 3 affected by anti-complement activity. iELISA was considered more specific, more sensitive and technically more easy. They note that CFT is only able to detect IgG1 and not IgG2. The latter does not fix guinea pig complement.

• Marin (1989) compared different antigens for the diagnosis of B ovis by CFT, AGID and iELISA. The PCP and CS antigens gave poorer diagnostic results than those obtained using the HS antigen in all the tests. Sensitivities were: iELISA, 97.6%; AGID, 96.4%; and CFT, 92.7%. AGID detected 2 positive rams not detected by iELISA. All tests were 100% specific for rams known to be free of B ovis.

• According to Blasco (1990) the main advantage of AGID compared to CFT is simplicity of use meaning that a non-specialist laboratory can do the test. It is also relatively easy to interpretation the precipitation lines. The main disadvantages of AGID is the need for standardisation and that it is only indicative test and cannot be quantified.

• According to Blasco (1990) based on the work of Marin (1989) in a comparison of AGID, CFT and iELISA the best results are from the iELISA test however some positive results are obtained by AGID which are not demonstrated by iELISA therefore a combination of AGID and iELISA gives the best sensitivity. A combination of CFT and iELISA is no better than iELISA alone. This is shown in TABLE .

Test	Sensitivity	Specificity
	%	%
AGID	96.4	100
CFT	92.7	100
iELISA	97.6	100
AGID + iELISA	100	100
CFT + iELISA	97.6	100

• West (1991) carried out trials on a flock of 64 rams. 18 had lesions of epididymitis with positive CFT results. 20 other rams had positive CFT of which 14 had positive bacteriological culture tests in their semen. 2 had negative CFT but positive iELISA and positive bacteriology culture on semen.

• The OIE Manual (2009) notes that AGID and iELISA have similar sensitivities but iELISA can be more sensitive than CFT. It is concluded that the best results are obtained with a combination of AGID and iELISA in terms of sensitivity. AGID represented the simplest most practical test for B ovis. At the time of writing because there are no international standards for iELISA and AGID, CFT is still used as the test procedure for international trade.

Epidemiological Studies of B ovis in Sheep.

• Australia. Studies by Gunn (1942) , Miller (1954) , Osborne (1955) of commercial flocks in Australia based on the incidence of epididymitis in rams has indicated that B ovis affects about 6% of rams. It should be noted that this can only be taken as an estimate. The incidence of B ovis as measured by serology tests is usually significantly higher because many rams infected with B ovis do not develop epididymitis. Conversely epididymitis may have other causes than B ovis.

• New Zealand. A survey by veterinarians described by Liberona (1981) indicated that that 12.3% of rams in commercial flocks in New Zealand were infected. Worthington (1988) using serology test data indicated 14% of rams in commercial flocks in New Zealand were seropositive to B ovis. According to Reichel (1996) this had been reduced by eradication programmes to 5.6% of rams.

• Brazil. Surveys by a number of authors have indicated that the prevalence of B ovis in rams in Brazil is between 5% and 35%, Alves (2010) . The tests which were used in these surveys included AGID and PCR.

• Canada. In Alberta the estimated level of B ovis infection is 8.6%, Niilo (1986) based on the results of CF and the bacteriological culture tests. 23 of 124 flocks were tested. 67% of rams in infected flocks were positive to CFT. A small survey indicated that B ovis was also present in Ontario, Buckrell (1985) but it is apparently absent or only present in low levels (≤1%) in Quebec, Arsenault (2004) .

Control and Eradication of B ovis from Sheep.

The same methods are used to control and eliminate B ovis from flocks of sheep as have been developed for other forms of the disease such as B abortus in cattle, B melitensis in goats and sheep and B canis in dogs. Custom and practice indicates that B ovis is much easier to eliminate than B abortus or B melitensis. This probably relates to its lower virulence. The methods that have been tried are:

• Vaccination. Vaccination using B melitensis Rev 1 can reduce the incidence of B ovis infection but not eliminate it. It is best used to reduce the incidence of B ovis infection in flocks of sheep where test and slaughter is considered unlikely to work such as if re-infection is probable. It can also be used to reduce the lever of infection prior to test and slaughter.
• Test and slaughter. In a test and slaughter program animals (in this case mainly rams) are tested by physical examination and serology tests and positive animal are slaughtered. This is continued at regular intervals until the flock is free of infection. Test and slaughter should only be considered if complete eradication is possible without external re-infection.
• Antibiotic treatment. Trials in the use of antibiotics to treat B ovis in sheep has been shown to be partially successful but the risks the development of resistance with negative consequences for human health are of greater importance. Generally antibiotics are too expensive to be used in this way for sheep but their use can possibly be justified in the case of valuable animals such as pedigree rams..

Test and Slaughter Based on Elimination of Rams with Epididymitis.

Originally B ovis in sheep was considered to a form of contagious epididymitis and attempts were made to eliminate the disease by removing rams with this disorder however this has proved to be ineffective because it is now known that not all infected rams show epididymitis. The failure of such trials has been reported by Clapp (1955) and Bagley (1985) . Bagley noted that on the basis of a 5 year study that culling all rams with palpable lesions in a flock progressively reduced the incidence of epididymitis it did not eliminate it. According to Hughes (1968) only about 35% of rams with B ovis develop detectable lesions. It is considered that attempts to control B ovis in a flock by continually removing rams with signs of epididymitis is expensive and is not recommended. Most test and slaughter eradication programmes have concentrated on eliminating infected rams however according to the OIE Manual (2009) , the role of ewes in transmitting B ovis is significant and it is essential that eradication programmes include ewes.

Test and Slaughter using Serology Tests.

Test and slaughter has been found to be the best method of eliminating other forms of brucellosis such as: B abortus in cattle, B melitensis in sheep and goats, B suis in pigs and B canis in dogs. In the case of B ovis these programmes usually concentration removing diseased rams although it is also known that ewes can spread the infection and should be considered to ensure complete elimination. The control of B ovis infection is particularly important in pedigree flocks particularly those which sell rams. Usually an eradication programme involves using serology test at regular intervals and culling infected rams. This is repeated until all test results remain negative for a controlled period. Alternatively in cases where there is a high level of established disease, chronic infection is present and the rams are not pedigree stock culling of all the rams should be considered as the most economic choice, West (2009) . In some cases problems can arise from the presence of infected animals which are negative according to physical examination and serology tests.

Examples of Test and Slaughter programmes.

• Ryan (1964) describes an apparently successful programme of eradication of B ovis from sheep flocks of the Bass Strait Islands of Tasmania. The programme was initiated after high levels of B ovis infection of sheep was discovered in the Flinders and King Islands. Initial evaluation was by scrotal examination followed by serology testing of all rams irrespective of epididymitis using CFT. This showed that 18% of ram flocks were infected. In some flocks with high infection levels all rams were culled and in others a test and slaughter programme was initiated. It was found that this worked best outside the mating season. CFT testing was carried out at monthly intervals and reactors were culled 2 to 3 weeks later. Stock holders were compensated for lost rams. The process of test and slaughter was repeated 4 times at which stage no further infected rams were detected. Further testing 2 to 3 years later revealed a limited number of positive results and a few rams were slaughtered. At this point it was considered that eradication was complete. During the programme restriction were placed on the movement rams aged 6 months or over. Measures were taken to ensure that all new rams shipped to the Islands were free of B ovis.

• West (1978) , (1979) and Bruere (1987) have shown that a combination of scrotal examination and CFT can used to successfully eradicate B ovis from sheep.

• Walker (1985) used an iELISA-HS test together with scrotal palpation to detect, control and eliminate B ovis in an infected flock. All reactor rams were removed. Subsequently rams were free of B ovis and none of the mature breeding rams developed lesions related to epididymitis.

• According to a study described by Blasco (1990) from his own work Blasco (1986) eradication of B ovis from a flock sheep can be achieved using a combination of scrotal palpitation and AGID-HS testing every 6 months in about 2 years. He considered that without associated evidence it is reasonable to assume that if iELISA-HS or iELISA plus AGID-HS were used in this procedure this would be more effective since iELISA is more effective than AGID in identifying infected animals and a combination of iELSIA and AGID finds most cases of B ovis infection.

• Reichel (1994) describes the procedures used to eliminate B ovis from sheep in the Falkland Islands. A combination of serology tests and slaughter were used and the process took from 1977 to 1993. All rams even those which may be intermittently infected must be culled.

• Robles (1998) noted that a single cull based on the results of clinical examination and serological test results was unable to decrease the prevalence of B. ovis in an infected flock of sheep flock. The main problem was false negative results. Better results are obtained if a combination of serology tests and bacteriological culture tests on semen are used, West (1993) .

Vaccination against B ovis in sheep.

In general elimination of B ovis by test and slaughter is preferred to vaccination. However vaccination may be the best option in areas of high incidence of B ovis infection, Blasco (1990) . This can also be used as a preliminary stage of eradication which can reduce B ovis infection to manageable levels prior to use of test and slaughter programmes. Vaccination can also be justified as a cheaper solution on economic grounds in poorer countries.

Killed B ovis Vaccines (Bacterins).

Vaccines based on killed B ovis have proved of limited value offering only a low level of protection, Haughey (1969) and Buddle (1963) .

B abortus s19 Vaccine plus B ovis Bacterins (killed B ovis).

B abortus s19 is a live attenuated vaccine used to vaccinate cattle against B abortus. Trails of the use of B abortus s19 plus B ovis bacterins to vaccinate sheep against B ovis have been carried out by Buddle (1954) , (1957) and (1958) and he considered it gave effective protection to rams against infection in rams. However the method has significant problems:

• There is a general view that this procedure has a low or moderate efficacy.
• The procedure is associated with the development of B ovis antibodies which interfered in subsequent serology tests, Claxton (1968) .
• Skin lesions probable associated with killed B ovis and chronic granulomatous lesions associated with intra-peritoneal vaccination can occur, Bailey (1988) .
• It has been noted that some rams excrete B abortus s19 in their semen, Buddle (1962) <Buddle MB. Production of immunity in rams against brucella ovis infection. NZ Vet J. 1962. 10. 111-15.</ref> and (1963) .
• Some rams which had been subject to the vaccine develop lameness with lesions in tissue from which B abortus s19 was cultured, Kater (1963) and West (1978) .
• B abortus s19 vaccine can be a serious risk to veterinarians or others using it usually by accidental inocculation since it can cause acute and chronic infection in humans.

B melitensis Rev 1 Vaccine.

Rev 1 is the best available vaccine against B ovis in sheep. It is a live attenuated vaccine developed to provide protection against B melitensis infection in sheep and goats (See under B melitensis). It is currently the most effective vaccine against B ovis infection in sheep, however its use is restricted in B melitensis free countries. The vaccine is administrated subcutaneously or conjunctively at a dose rate of 10⁹ CFU in rams at 3 to 5 months old. and produces 75% or more immunity. Trials indicate that immunity provided by Rev 1 is superior to that provided by B abortus s19, Van Heerden (1962) . Rev 1 works well in immature rams but its effectiveness in mature rams has been questioned, Gradwell (1975) . Rev 1 reduces the incidence of B ovis in a flock of sheep but cannot eliminate it. Flocks of sheep can be infected by both B ovis and B melitensis at the same time. In this case Rev 1 can be used as a vaccine against both infections. Rev 1 may cause disease in some rams.

Fensterbank (1982) vaccinated 15 rams with a dose of 8.5 x 10⁸ CFU of B melitensis Rev1. 15 rams which were unvaccinated were used as controls. All the rams were challenged 8 months later with 5 x !0⁸ CFU of B ovis. All were slaughtered 3 months later. 3 of 15 vaccinated rams had organs or lymph nodes infected with B ovis. while all controls were heavily infected. The trial suggests that Rev1 can be used to control B ovis.

Blasco (1987) tested the use of B melitensis Rev 1 for the vaccination of rams against B ovis. 43 rams aged 3 months were vaccinated with a dose of between 5 x 10⁸ and 10⁹ CFU. There were also 6 unvaccinated controls. At 21 months all rams were challenged with a dose of 5 x 10⁸ CFU of B ovis. At slaughter 8 weeks later: 1 vaccinated ram had developed epididymitis and 4 of the unvaccinated rams had developed testicular alterations. At slaughter: B ovis was recovered from 26.6% of vaccinated rams and 100% of controls. The lesions in testes and vesicular glands were more severe in control than vaccinated rams. It was concluded that Rev 1 vaccine gave some measure of control of B ovis in ram lambs.

Although some reports indicate that Rev 1 is avirulent in rams, such as a report by Erasmus (1985) . They repeatedly inoculated rams with Rev 1 but considered that this caused no obvious effects. However in another report Rev 1 was isolated from a vaccinated ram which developed epididymitis, Lantier (1985) . This reflects similar results for the use of Rev 1 in other animals (See above B melitensis).

Problems associated with the use of Rev 1 vaccine against B ovis are as follows:

• Flocks of sheep can be infected concurrently with both B melitensis and B ovis. If Rev 1 is used against B ovis then this leads to positive serology tests results for B melitensis. The problem may be partially addressed by use of conjunctival vaccination in young animals for which in case most of the serology titres return to negative values after about 5 months, Blasco (1990) .
• Rev 1 also produces low CFT titres in relation to serology tests for B ovis, Gradwell (1975) and Erasmus (1985) and therefore interferes with serology testing for B ovis.
• Rev 1 can cause serious infections in humans by accidental inoculation.
• Rev 1 is banned from use in countries where B melitensis has been eliminated.
• Rev 1 can sometimes cause disease in sheep. See above.

Test and Slaughter Eradication Procedures.

Brucellosis caused by smooth Brucella such as B abortus infection in cattle, B melitensis in sheep and goats and B suis in pigs cause serious economic losses to farmers and is a serious danger to human beings. This has led to sustained efforts to eliminate these diseases from domestic stock particularly in developed countries. In the case of B ovis infection in sheep convincing arguments for comprehensive eradication programmes are more difficult to make. The economic losses are less and there is apparently little danger to human health. Therefore there is resistance from farmers to the significant cost and sustained effort involved, Middleberg (1973) . The main economic losses associated with B ovis infection in sheep are a limited loss of viable lambs produced and a probable greater loss related to the replacement of infected rams. This process test and slaughter is time consuming but possible benefits include: a shortened the lambing period, increase of twin rates, and a reduction of unproductive ewes and the need for fewer rams. The main costs are: the cost of culling and replacing rams, the cost of veterinarian services and laboratory cost from serology testing. The whole process can take ≥3 years and there is also a need for subsequent annual check ups.

There are no uniformly accepted eradication procedures. Here are some of the main points:

• Most eradication programmes specifically target the culling of infected rams although some authorities consider ewes are also important.
• Although scrotum examination for epididymitis is a useful indicator of the presence of B ovis in a flock it cannot be used as the basis of an eradication programme.
• A test and slaughter eradication programme depends on the use an one or more effective serology tests. The tests which can be used in eradication programs are: CFT, AGID, and iELISA. Combinations of tests are able to find more positive infections. It is considered that as combination of iELISA and AGID is probably the most useful.
• It is considered that eradication is best carried out outside the breeding season, Ridler (2011) . The best time is probable 4 to 8 weeks after the end of the breeding season. Most infections are noticed after the end of the breeding season.
• Serology tests should be repeated regularly and infected or suspect rams culled each time until all the remaining continue to give negative results. Ridler (2011) suggests that tests should be repeated every 4 to 8 weeks.
• Rams should be tagged for ease of identification.
• All rams should be included in the eradication programme including juveniles, Bulgin (1990) .
• Environmental contamination is of limited importance and infected and non-infected animals can be kept on the same property as long as there is no direct contact.
• Breeding rams should be kept separate from rams in other flocks and rams and ewes kept separate except at breeding times. It is useful if rams are sub-divided and mated to separate groups of ewes to reduce cross exposure.
• When an eradication programme is successful continued success will depend on preventing re-introduction of the disease. Only stock known to be clear of B ovis should be added. For a ram introduced from a non-accredited flocks a single negative serology test is not reliable.
• If possible replacement disease free stock should come from B ovis free accredited stock farms. Accredited schemes for B ovis free flocks of sheep have been introduced in some countries including New Zealand and some USA states.
• Annual testing on B ovis free flocks to look for possible new carriers is essential.

Clean rams should not be mixed with rams from other flocks.

• In a B ovis free flock a ratio of 1.5 to 2 rams per 100 ewes is sufficient whereas 3 to 4 rams per 100 ewes is necessary in infected flocks. This is a possible economic benefit.
• Ineffective eradication programmes lead to a high culling rate of rams which is not cost effective.

Treatment of B ovis using Antibiotics.

Only limited trials of the use of antibiotics to treat B ovis infection in sheep have been carried out. General studies indicate that antibiotic treatments are not particularly successful with only about 50% success rates and often no improvement in fertility. Best results are obtained if treatment is carried soon after the start of infection. The costs involved in antibiotic treatment and the monitoring its use is considered unrealistic for the average sheep with B ovis. From an ethical point of view it is probably wrong to use antibiotics in this way consideration of the possibility of antibiotic resistance developing and the associated risks to human health. In particular streptomycin should not be considered for this type of use. There is insufficient information to determine if antibiotics produces long term improvement. Possibly antibiotic treatment can be considered for animals of greater value such as pedigree rams. Infected rams which have been treated with antibiotics remain sero-positive for B ovis which is a complication factor in test and slaughter regimes.

• Kuppuswamy (1954) carried out trial which demonstrated that a combination of tetracycline and streptomycin could stop the shedding of B ovis in semen a proportion of infected rams during short periods of treatment and follow up He treated 6 rams with chlortetracycline and streptomycin for 21 day and noted that all six stopped excreting B ovis for 113 days (the test period).

• Glauffret (1974) treated 27 B ovis infected rams with tetracycline and streptomycin. They became serologically negative and were free from epididymitis for 8 months.

• Marin (1989) infected 24 rams with B ovis. 12 rams were treated with oxytetracycline and streptomycin and 11 rams became negative by the bacteriological culture test. However in the case of 12 rams treated with oxytetracycline alone only 4 became negative. Treatment failed to resolve epididymitis in 2 rams and antibiotic treatment failed to cure lesions in the epididymis and ampullae.

• Dargatz (1990) carried out antibiotic trials on rams shedding B ovis in semen but without palpable abnormalities of the epididymides. Treatment was with oxytetracycline and streptomycin. Approximately 80% of rams stopped secreting B ovis during a 19 week period.

Chronic B ovis Infection in Sheep.

It is widely accepted that B ovis infection in rams is a chronic disease that last indefinitely. This is based on study of the obvious condition, epididymitis and associated secretion of B ovis in semen. West (1993) considered that brucellosis in sheep, caused by B ovis is primarily a chronic infectious disease of rams with epididymitis as its most characteristic lesion. The chronic disease in rams is characterised by testicular atrophy and enlargement of the tail of the epididymis. Testes which may appear normal show granulomas and calcification on dissection, Badiola (1982) . Buddle (1955) noted that a high percentage of infected rams were capable of secreting infected semen for periods of more than 4 years.

Worthington (1985) studied 42 rams naturally infected with B ovis over a 14 month period. Bacteriological culture tests on semen was positive in 87.9% of samples of semen. CFT was negative in 11% of infected rams but 76% and 77% of these CFT negative sera were positive by AGID and iELISA tests respectively. Rams with negative or fluctuation CFT values had a history of chronic infection. 5 rams with negative bacteriological culture test result on semen had erratic serology test results.

Chronic B ovis in ewes is less well understood. There is a widespread view that ewes recover from the infection on the basis that they lack of observable signs and symptoms of the disease. However some ewes which show infection during pregnancy secrete B ovis in their milk again at a subsequent pregnancy. This is typical of brucellosis caused by other Brucella in other animal species such as B abortus in cattle (See B abortus above). Cattle which show no signs of brucellosis and which have negative serology test results can be shown to carry B abortus at necropsy by positive bacteriological culture test results on internal organs. These cattle are termed latent carriers and can infect their calves and other cattle. This may also be true of ewes. Also it has been found that B ovis localised in the spleen and lymph nodes in sheep. Commonsense would suggest that some ewes can be latent carriers although their capacity to spread the infection is unknown and may not be important commercially.

Susceptibility of Other Animals Species to B ovis.

Natural Infection in Red Deer (Cervus elaphus).

Ridler (2001) has observed that red deer in New Zealand are the only species other than sheep known to have become infected naturally with B ovis (Also see below: White Tailed Deer). Red deer can be infected experimentally by the intravenous, conjunctival, nasal and rectal routes. Infected stags shed B ovis in semen and can have poor quality semen but do not seem to transmit the organism long-term. In pregnant hinds abortion and postnatal deaths seem to be rare or absent. In documented cases of B ovis in red deer the probable source of infectionis seems to be by contact with infected rams or other infected stags. B ovis infection in red deer has similar manifestations to that in sheep but the prevalence is probably low. Trials have indicated that B ovis can be transmitted directly from stag to stag but not from stags to rams, West (1999) and from ram to stag, Ridler (2000) while they are grazing together. Ridler (2000) showed that 5 of 6 stags which were grazed in the same paddock as infected rams became seropositive by the CF test and in 4 of these stags B ovis' was cultured from semen. In general it is likely that direct contact is necessary to cause infection, Hartley (1955) , Keogh (1958) and Ridler (2000) . Ridler (2001) and Snowden (1958) note that stags can become infected with B ovis from the vaginas of infected hinds in a similar way to which it can be transmitted from ewes to rams. According to Ridler in stags the infection localises in the epididymes, seminal vesicles and ampullae with characteristic lesions. In chronic infections the lesions tend to be mild. about 10 to 20% of infected stags, have gross enlargement of the tail of the epididymis and small abscess-like lesions may be present. B ovis can be cultured from the epididymes, seminal vesicles and ampullae of infected stags. In general B ovis infection in deer mimic the disease in sheep however most stags stop shedding B canis in semen within a year of becoming infected possibly indicating the elimination of infection. Data indicates that the overall incidence of B ovis in deer is probably low.

B ovis Infection in White Tailed Deer (Odocoileus virginianus).

White tailed deer are native to the Americas. From limited information B ovis infection in white tailed deer appears to be similar to that in red deer. Barron (1985) artificially infected 8 white tailed deer by the conjunctiva route. The animals were killed at intervals. All became infected but in younger bucks the infection appeared to be transitory. In mature bucks B ovis became established in 1 case continuing to 429 days (until necropsy). Four deer had gross lesions and histopathologic changes of the epididymides. It was shown that transmission was possible between deer. A mature non-infected buck confined for 7 months with an infected buck became infection and developed epididymal lesions.

A survey of wild-white tailed deer by Martinez (1999) suggests that B ovis is not present in the wild animals. He tested sera from 350 white-tailed deer in 1994 from north eastern Mexico and found no evidence of B ovis infection.

Infection in Wild Sheep or Mouflon (Ovis musimon)

There is no evidence that mouflon are infected with B ovis but very little work has been done. Limited trials have been made by Cerri (2002) to infect 8 mouflon the intra-conjunctival inoculation route but with limited success. These mouflon are a form of wild sheep from mountainous regions of Italy. No obvious signs of infection were noted and no lesions were detected by histological examinations from a range of internal organs at necropsy carried out 8 months post infection. However seroconversion did occur after about 2 weeks post infection and transient bacteremia wes also noted. It is concluded that B ovis probably did not infect the mouflon.

Induced B ovis Infection in Laboratory Animals.

According to Blasco (1990) small animals which have been shown to be susceptible to B ovis artificial infection are: rabbits, rats, gerbils, hamsters, mice and guinea pigs. Doses from 10⁴ CFU are sufficient to initiate infection.

Buddle (1956) tested the susceptibility of guinea-pigs and rabbits to B ovis. Guinea-pigs were inoculated with 5 x 10⁹ CFU. The animals showed few macroscopic lesions but in most cases B ovis was recovered from their spleens after 6 weeks. Serum antibody titres to a rough antigen were low at 0 to 1:10. Similarly rabbits were inoculated with 5 x 10⁹ CFU of B ovis. In this case antibody titres against a B ovis antigen were higher at 1:80 to 1:640. Macroscopic lesions were only present in one case but B ovis was recovered in heavy concentration from the epididymes and spleens of all the rabbits. Histological examination indicated changes in the liver, the spleen, kidneys and testes. Guinea-pigs were inoculated intramuscularly with doses ranging from 15 to 15 x 10⁹ CFU of B ovis and killed after 6 weeks. Low antibodies tires against B ovis at 1:10 were present when a doses ≥ 15 x 10² CFU of B ovis was used. B ovis was not recovered from the spleens of guinea-pigs at doses <15 x 10² CFU. In similar experiments in rabbits infection of the spleen was found when doses of 15 x 10³ CFU of B ovis inoculant was used but not below this dose. Antibody titres against a B ovis antigen were 1:10 to 1:20. B ovis was less pathogenic in guinea-pigs and rabbits than B suis at similar doses. Cuba-Capara (1973) also carried out work with gerbils which developed lesions in the epididymus similar to those observed in rams.

B ovis Infection Domestic Goats (Capra aegagrus hircus).

It has been show that goats can be artificially infected with B ovis. García-Carbrillo (1977) tested 15 goats to assess their susceptibility to B ovis. They were inoculated by the conjunctival route with 10⁹ CFU of B ovis. They were killed as follows: 3 at 78 days, 5 at 108 days and 5 at 138 days. B. ovis was recovered from semen of a male goat at 33 and 61 days after inoculation. All 5 goats killed at 78 days had B ovis in their internal organs. 1 goat showed epidydimitis. The goats became seropositive by CFT and AGID after 80 days. 15 rams similarly treated showed more severe symptoms with 7 of 15 with epididymitis. Burgess (1985) attempted to infect 3 lots of 2 goats with semen infected with B ovis. 2 were inoculated in the testes, 2 nasally and 2 in the preputial epithelium. All 6 goats showed positive antibody response by iELISA and 5 by CFT. 1 inoculated in the testes and 1 goat treated nasally subsequently excreted B ovis in their semen.

There are no known confirmed cases of natural B ovis infection in goats however, Vermeulen (1988) describes the cases of possible B ovis infection in Angora he-goat.

Induced B ovis Infection in Domestic Cattle (Bos taurus).

Only limited trials have been carried out to test the susceptibility of cattle to B ovis infection. Buddle (1953) inoculated 2 pregnant cows with B ovis. One was slaughtered prior to calving and one calved normally. There was no evidence of infection in the calves or foetal membranes but B ovis was recovered from the milk of the surviving cow.

Susceptibility of humans to B ovis Infection.

Many people are routinely exposed to B ovis infected sheep and there is virtually no evidence that this can result in acute infections. Evidence from animals indicates that a wide range of animals can become acutely infected if the dose is sufficiently high and this is also probably true of humans. The are some possible examples of infection through exposure as described by Meyer (1982) but B ovis has not been isolated from man. There has been very little work in this area so there is a possibility the humans can and have been subject to sub-acute and chronic infections. There is probably a need for epidemiological surveys on people in contact with sheep past and present to look for the well known symptoms of chronic brucellosis such as tiredness, fatigue, headaches, myalgia and arthralgia. PCR tests for B ovis DNA could also be used.

Brucellosis in Wild Animals.

The species which are affected by Brucella can be divided into special host which can act as a long term recevoir for the disease and secondary hosts which contract the disease by contact with a primary host and for which the disease is probably self limiting, Smith (1990) . In general brucellosis is a chronic illness which lasts for the lifetime of the animal. It is obvious that brucellosis must have been present in wild animals before the process of domestication by humans but it is now virtually impossible to distinguish between wild animals which are long term primery hosts and wild animals which have become long term hosts by contact with domestic animals. An example of a species appaently infect by contact with domestic animals is the American bison which seems to have been infected by imported European cattle brought by settlers from 1880 onwards, Meyer (1995) . The disease is now self sustaining in bison such as has been demonstrated for herds in the Yellow Stone National Park in Wyomin USA and the Wood National Park in Canada. A possible example of a long term infection in wildlife is suggested by Waghela (1986) . He studied brucellosis in wildebeest and African buffalo using serology antibody tests (SAT, CFT, RB). 18% of wildebeest and 30% of buffalo were positive in all age groups and both sexes. He noted that hygromata is a sign in the disease. There is no evidence that Brucella has any self-sustained existance outside an animal host, Corbel (1989) . In order for the disease to become self-sustaining in an animal species it requires a reliable mechanism for passing the disease among animals and/or from one generation to the next. Some of the most important mechanisms are: contact with aborted fetuses, milk and vaginal secretions such as for B abortus in bovids and goats; sexual contact such as for B canis in canids and B ovis in sheep and B suis in pigs; general contact in social animals such as licking and contact with fecal material and vertical transmission from dam to ofspring. Thorne (2001) notes that pathogenisuis of both wild and domestic animals is similar regardless of the species of Brucella. Infections are usually chronic and persist for the lifetime of the host. Thorne (1982) notes that in wild animals with Brucella infection detectable signs are not known and may not occur. Fetal loss may occur but because of habits of wild animals it is not likely to be observed. General possible signs are: artritis, osteomyelitis, synovitis, and bursitis.

The following are the known self sustaining primary hosts for Brucella:

• B abortus biovars 1 to 7 and 9 in bovids such as cattle. African and Asian buffalo are also possible primary hosts.
• B melitensis in caprids such goats.
• B melitensis in domestic sheep.
• B suis biavars 1 and 3 in suidea such as pigs.
• B suis biovar 4 in reindeer and carabou.
• B suis biovar 2 in european hares.
• B canis in canids such as dogs.
• B neotomae in wood rats.
• B ovis in sheep.
• B pinnipediae in seals which is not yet fully classified.
• B cetaceae in whales and dolphins which is not yet fully classified.

According to Thorne (2000) Brucella has been discovered in 91 species of wild animals: marsupialla, 1; lagomorpha, 5; rodentia, 22; cetacea, 5; carnovora, 21; pinnipedia, 3; tubulidentata, 1; perissodactyla, 1; and artiodactyl, 32. Only a few species have been studied in any detail.

Tests for serology antibodies to Brucella used in wildlife are based on those used in domestic animals. Just as in domestic animals and humans, a positive antibody titre does not necessarily indicate active infection. Also just as in almost all Brucella infections in other animal species; human, domestic and experimental animals low or negative antibody levels may develop sometimes concurrent with positive Brucella cultures. In general the tests can be used to identify desease in groups of animals but not individuals.

B abortus in The North Amercan Elk: Elk are are not a natural host for Brucella. It was first observed in elk in Wyoming in the 1930s, Murie (1951) . B abortus infection in elk has become established in the Greater Yellowstone area but not elsewhere in the USA, Lee (1937) and Thorne (1978) . This is because they are continually exposed to reinfection by contact with infected cattle and bison mainly at winter feeding grounds.

Thorne (1978) studied 24 elk cows and 3 elk bulls artificially infected with B abortus all but 2 became infected as demonstrated by serological response and culture tests (blood or at necropsy. The exceptional 3 cows had relatively low serotitres. 18 Bruccella free cows and 6 Bruccella free bulls were intermixed with the infected elk and all but 1 cow became infected. 7 cows aborted. 2 cows developed latent infections. they were serologicallyand hemaculturally negative until 23 months when abortion occurred. Similarly a bull calf became seronegative after 8 weeks but was shown to be infected at 15 months by recovery of B abortus at necropsy indicating latent infection. The route of infection in elk was apparently simmilar in elk to that in cattle by licking or ingestion of infected material. This was confirmed at necropsy by recovery of B abortus from mandibular, suprapharyyngeal, or altantal lymph nodes related to a oral or conjumctival route of entry, . Observations indicate that elk cows normally eat the placenta and attempt to eat aborted fetuses. Other cows were observed licking atn aborted fetus. Venereal transmission semmed unimportant as indicated by breeding trials with infected bulls. Abortion or premature birth or birth of non viable calves occurred in 48% of infected cows and 11% lost a calf at their second pregnancy and 20% of these a third pregnancy proven by culture testing. The likely rates were higher.

Thorne (1978) notes that buritis or hygroma and secondary infection of the joint and synovitis is fairly common in Brucella infected elk (~20%). Such swellings are the second most common sign of Brucella infection after abortion and non-viable calves. These conditions were most prevulent in the chronic phase of the illness. Such swellings could be large or small. Brucella could be recovered from fluid asparate in some cases. Bursa could sponntaneously diminish. In one case Brucella was recovered from the asparate of a new bursa but not from the same but diminished bursa 2 years later. In some cases there is lameness sometimes severe and in more than 1 leg. Abortion has been noted at feeding grounds but is not likely to be observed in the wild because the placenta is usually eaten immediately by the dam and the aborted fetus is eaten either by the dam or scavengers. Thorne (1978) found that 48% of elk cows artificially infected with virulant strain B abortus s2308 aborted. He also found synovitis due to B abortus in 20% of elk on necropsy. According to Thorne (1982) hygromas have moist red-brown internal linings and sometimes contain opaque straw coloured fluid and fibrinous debris. The surface included composites of fibrin and underlying tissue was edemarous and infitrated by neutrophiles, plasma cells, macrophages and sometimes giant cells. A nimber of animals developed chronic brucellosis which continued until they were killed. The longest recorded was 56 months.

The serological response in elk is similar to that in cattle. Morton (1981) studied the serological response of elk with 4 tests including SAT anf CFT. SAT titres appeared at about 3 weeks but declined rapidly at about 6 months. CFT identified that most culturally positive elk at necropsy. Thorne (1982) notes that according to Morton (PhD thesis 1978) elk artificially infected with Brucella produce nonspecific IgM first followed by IgG1. Just as in cattle there may be animals with no serological antibody response with positive bacteriological culture tests. Thorne (1978) noted elk calves born to infected dams could become latent carriers. Also Morton (1981) found some calves which were culturally positive for B abortus at necrpsy but were seronegative suggesting they were latent carriers. Thorne notes that whereas in cattle the udder is a preferred site for Brucella localization, Lambert (1960) this is not the case in elk. In culturally positive cow elk brucella was only recovered from the udder in 22% of cases. Also few samples of Brucella contaminated milk were recovered. It was concluded that calves were infected in utero rather than by nursing.

B abortus in North American Bison. Just like in elk, bison have become infected with brucellosis by contact with infected cattle but in the case of bison the disease is self-sustaining. The first cases were recognised in 1917, Katz (1941) The significant effort in studing both bison and elk in the Greater Yellowstone area relates to the danger they pose in re-infecting cattle from which the desease has been cleared. At the present time bison which enter cattle pasture are shot with significant negative impact on their numbers. Attempts have been made in bothe elk and bison to control brucellosis by vaccination.

Probable signs of brucellosis in American bison are; abortion and retained placenta. Williams (1997) . Other signs in free ranging bison are enlarged testicles and orchitis, Tunnicliff (1935) and Corner (1958) , artritis and hygoma, Tessaro (1987) .

Meyer (1995) considered that transmission of Brucella can occur from feces of calves nursing infected dams and this has been demonstrated by the recovery B suis biovar 4 from the feces of artificially infected calves, Bevins (1996) .

Davis (1991) studies on bison confirmed that B abortus is a sustainable infection in bison and can be transmitted to cattle, Rhyan (2001) .

Attempts to vaccinate bison using B abortus s19 and RBN! have shown problems. Both are associated with the develpoment of granulotous lymphadentis and are not cleared as rapidly as in cattle, . In bison B abortus s19 caused abortions, chronic infection and persistent antibody titres, Davis (1991) . RB51 has also been shown to cause abortions and placentitis, Palmer (1996) .

B suis in Caribou or Reindeer. Caribou and reindeer are a migratory wild and semi-domesticated type of deer present in the artic and sub-artic region of North America and Ei=uroasia respectively. Caribou are the host species for B suis biovar 4. B suis biovar 4 can be psssed from caribou to domesuc animals and carnifors. Neiland (1975) noted that antibodies to B suis biovar 4 could be found in the blood of various domestic and wild carnivors feeding on carabou including: led dogs. wolves, red foxes and grizzly bears. Forbes (1993) penned 16 reindeer infected with B suis biovar 4 with 8 cattle (6 males and 2 females). 2 reindeer aborted. B suis biovar 4 was isolated from 2 female and 2 male cattle. The females showed positive test serolgy tests (agglutination tests, CFT ans ELISA) but the tests were incosistent in the males. This indicates thatb carabou are a brucellosis disease risk to cattle.

Neiland (1968) in a study of free ranging carabou in Alaska noted that signs of the disease were: orchitis-epididymitis with enlarged testicles, bursitis-synovitis with large hygromas known as carabou knee band swollen hocks and metritis-abortion syndromes with possible retained placenta and bleeding.

A number of authors have noted that native hunting peopple in tha artic are vunerable to B suis biovar 4 infection, Huntley (1963) , Neiland (1970) and (1975) .

Miller (1980) showed that a range of rodents could be readily infected with B suis biovar 4. These included: guinea pig (Cavia porcellus), Scandinavian lemming (Lemmus lemmus), brown lemming (L. sibiricus), northern red-backed vole (Clethrionomys rutilis), varying lemmings (Dicrostonyx stevensoni and D. rubricatus), yellow-cheeked vole (Microtus xanthognathus), flying squirrel (Glaucomys sabrinus) and ground squirrel (Citellus parryii). In lemmings the response was dramatic and most animals eventually died. In other rodents pathogenic response was not marked.

Neiland (1970) reported that sled dogs eating B suis biovar 4 infected raw meat can become infected. Examination of an infected bitch indicated positive serology for brucellosds. post mortemm examination indicated B sius colonies in the mandibular, mesenteric, popliteal and mediastinal lymph nodes and spleen. B suis biovar 4 also affects woves. Neiland (1975) also demonstrated that 30 to 90% grizzly bears (Ursus arctos horribilis) associated with diseased carabou had positive serology antibody tests against Brucella (presumably B suis but wolves (Canis lupus) and red foxes (Vulpes fulva) are less commonly infected.

B abortus in Canids. Davis (1979) studied coyotes infected with B abortus biovar 1 in Texas. Wild coyotes were capture and their tissue and blood esxamined. 18% of 52 had positive serology tests. B abortus was cultured from 7 of 43 coyotes. 3 culturally positive coyotes were seronegative by 4 tests including SAT and CFT. This is not the same situation as with seronegativity due to B canis which is a rough strain without LPS. Congenital transmission was found. Th most likely means of infection is scavaging for infected bovine tissue. In further experiments Davis (1988) showed that coyotes artificially infected with B abortus could pass the desease to cattle. In trials infected coyotes were penned with brucellosis free parturient heifers. Some heifers became seropositive and subsequently aborted.

B abortus has also been reported in foxes in Argentina, Szyfres (1966) and (1967) and in the UK McCaughley (1968).

B melitensis in Ibex and Chamois. There are few cases of B melitensis described in wild animals. B melitensis has been reported to occur in ibex and chamois in Europe. Ferroglio (1998) describe the first case of B melitensis in Alpine ibex. A 7 year old buck showed swollen testicles and thich-walled capral joints. Brucella was cultured from both. Garin-Bastuji (1990) describes a B melitesis biovar 3 systematic infection in a chamois in the French Alps. Clinical signs were orchiepididymitis, polyarthritis, blindness and neurological signs. Brucalla was isolated from the eye, testes. kidneys, lung and joints.

B suis in Wild Boar and Ferile Pigs. Wild boar throughout Europe Munoz (2010) , Leuenberger (2007) , Montagnaro (2010) are infected with B suis biovar 2. This the same biovar that affects Eyropean hares. The is a risk that infecred wild boar can spread the disease to domesytic pigs. Cvetnić (2009) in a study in Croatia isolated B suis biovar 2 from wild boar and outdoor reared domestiv pigs. He considered that the domestic pigs had been infected by contact with the infected wild boar. Godfroid (2002) notes that in wild boar infected animals may be bacteriologically positive in the absense of gross lesions.

Feral pigs infected with B suis occur in a number of countries including;, the US and Queensland, Australia. In the USA feral swine are presnt in 38 states and Brucella infected swine have been identified in many including; Florida, , Texas and Hawaii . In feral swine the B suis biovars are those in domestic pigs and B suis biovar 1 has been identified in South Eastern USA. CDC (2009) describe 3 cases of brucellosis associated with the hunting of ferile pigs. In 2 cases the incubation period was 4-6 months. report 2 cases of brucellosis in hunters who killed and dressed ferile pigs. Ronson (1993) reviewed cases of brucellosis for 2 years in Queensland, Australia. 11 cases were due to B suis associated with the hunting and slaughter of feral swine. Irwin (2009) identified 4 cases of brucellosis suis in NW Australia and considered that the infections were related to hunting of feral pigs.

Brucella in Marine Mammals. Brucellosis was first recognised as a disease of marine mammals after the discovery of Brucella in a captaive dolphin by Ewalt (1994) . It was noted that this Brucella had different charactoristics to existing biovars. Since then Brucella has been been found in a wide range of marine mammals, Godfroid (2002) and Foster (1996) . The marine animals shown to be infected include: seals by Forbes LB (2000) , porpoises by Neimanis (2008) , dolphines by Ewalt (1994 Miller (1999) and minke whales by Clavareau (1998) . Van Bressem (2001) studies 58 mixed whales and dolphins from Peru and others stranded on the Spannish coast and found significant levels of Brucella antibodies in both groups. Brucella infection has now been found throughout the worlds oceans.

The mechanism of sustained infection in sea-mammals in not knon but may be similar to that in land animals. such as; venereal transmission, envirnomental contamination, vertical transmission, and ingestion of contaminated food. Interestingly Garner (1997) has found lungworms from Pacific harbour seals infected with Brucella and Salem (1997) has shown that Nile catfish can be infected with B melitensis biovar 3. Antibodiers were detected 7 days after infection and the organism was recovered from visceral organs.

Ewalt (1994) isolated Brucella from the aborted fetus of bottlenosed dolphin which suggests that abortion may occur in this species in a similar way to land animals. Godfroid (2002) notes that in most cases Brucella has been isolated from subcutaneous lesions of marine animals. More recently Dagleish (2008) has isolated Brucella ceti from the testis and epididymis of an adult sexually mature free-living harbour porpoise. Examples of chronic brucellosis is described by Gonzalez (2002) . 3 young stripped dolphins live-stranded in Scotland were shown to be suffering with chronic non-suppurative meningoencephalitis. Brucella was isolated from leisions in the brain. Miller (1999) describes natural abortions in 2 bottlenosed dolphins due to Brucella infection. It is suggested that infected placenta and vaginal/uterine discharges could transmit disease to other ctaceans.

There is little known as to whether these new Brucella variants can be transmitted zoonoticcally to to humans. This is a desease unlooked for until recently. The meat of both seals and whales is butchered and eaten. Brew (1999) reports a case of a laboratory worker who probably became infected with a marine mannal Brucella. In a trial by Bingham (2008) with a human isolate of a marine type Brucella 02/611 used to artificially infect piglets re-isolated it from 3 piglets after 1 month but considered that it did not replicate easily in pigs.

Recently, the species B. ceti, B. pinnipedialis, and B. microti have been described Foster et al., (2007)

Some evidence that marine type brucellosis can be transfered zoonoticly to humans has been provided by Sohn (2003) who described 2 cases of neurobrucellosis in Peruvian men. 1 had negative serology titres. McDonald (2006) reported the isolation and characterization of a strain of Brucella from a New Zealand patient which appears most closely related to strains previously identified from marine mammals. The isolate was identified to be closely related to a Brucella sp. in bottlenose dolphin (Tursiops truncatus) and common seals (Phoca vitulina).

S

There is no evidence that B neotomae which has only been found in desert rats in Utah USA, has ever infected humans, FAO/WHO (1986) .

Meneses (2010) describes the case of a 84 year old woman who having undergone chemotherapy developed due to a bacteremic brucellosis infection. It is considered that the infection had been latent since childhood and had been trigger by a weakened immune state.

Ewalt (1979) examined tissue from 104 cows in a herd for Brucella. B abortus strain 19 was isolated from 22 cows, B abortus biovar 1 was isolated from 9 cows, and both strains were isolated from 2 cows.

Attempts to vaccinate bison using B abortus s19 and RBN! have shown problems. Both are associated with the develpoment of granulotous lymphadentis and are not cleared as rapidly as in cattle, . In bison B abortus s19 caused abortions, chronic infection and persistent antibody titres, Davis (1991) . RB51 has also been shown to cause abortions and placentitis, Palmer (1996) .

Brucellosis in Experimental Animals.

The use of experimental animal, usualy guinea pigs and mice, has played an important role in the understanding of brucellosis. There is reasonable corelation between the finding in these animals with humans and domestic animals. Guinea pigs are readily susceptible to B abortus, B suis, metiltensis, B canis and B neotomae but not B ovis. Enright (1990) notes that guinea pigs are considered very susceptible to Brucella and can be infected; subcutaneously, conjunctivally, intraperitoneal, intranasally, intravenously, vaginally, orally, or by cutaneous scarification.

Huddleson (1929) infected monkeys with B abortus, B suis, and B melitensis and found that the severity of illness was worst for B melitensis and B suis. All infected monkeys were killed after about 1 to 4 months and subject to posttmortem examination. The types of leisions found were as follows: B abortus:

• Lung. Dark grey lesion, small foci;
• Liver. Tubercle-type lesions with supurative centre.
• Spleen. Hyperplasia.

B Suis:

• Lung. Nolules (2-3 cm) consisting of non necrotic reticulo-endothelial cells with yellow / grey content, focal lesions (1-2 cms),microscopic foci, abscesses (2-3 cm).
• Lymph nodes. Microscopic foci, abscesses.
• Liver.. Necrotic and non-necrotic foci, microscopic foci.
• Spleen. Foci (1-2 cms), microscopic foci.
• Kidney. Abscesse (cluster of supurative lesions) 6 mm.

B melitensis.

• Lung. A large abscess (15 mm), greyish nodule (2 mm), necrotic (2-4 mm) and non-necrotic microscopic foci, necrotic neumonia (numerous microscopic foci of reticulo-endothethelial / leucocytes with necrosis and suppuration).
• Liver. Discoloued (greyish mottled) surface raised foci, microscopic foci Some necrotic).
• Spleen. Small and microscopic foci.
• Lymph nodes. Diffuse hyperplasia of teticulo-endotheliel cells.
• Kidney. Small grey and microscopic foci.

Garcia-Carrillo (1990) notes that in experimental animals that as well as inflammatory reactions and degenerative lesions being found in internal organs, Brucella can also be recovered from apparently normal tissue. Granulomatous lesions originally described in guinea pigs by Fabyan (1912) have also been described in humans, mice, rats, rabbits, dogs and monkeys by Watanabe (1960).

Granuloma formation. The intercellular habit of Brucella was first noted by Fabyon (1912) and their multiplication in renal and liver cells in guinea pigs. Such histological changes have been noted by other workers in a variety of animals. In guinea pigs, Jaffe (1922) Cotton (1922) , Meyer (1922) , Smith (1926) , and Braude (1951) . In mice, Feldman (1935 . In monkeys, by Huddleson (1929) , in dogs by Margolis (1945) , and rats and rabbits, by Nyka (1948) . Barrett (1953) found nodular granulomata in the spleen of guinea pigs experimentally injected with B abortus and hepatic granulomata due to B melitensis in cattle. Granulomata have been found in the liver and spleen of humans by Loffler (1930) and Albertini (1937) . These results were confirmed by the work of Spink (1949) reported the distinctive leisions in biopsy sample. It can therefore be assumed that highly distinctive granulamata occur in the liver of most brucellosis patients.

Braude (1951) infected 3 lots of 18 male guinea pigs with a range of strains of B abortus, B melitensis and B suis and observed the effect. He then killed them after 130 days and examined their organs microscopically for abscess and granulous leisions. The total number of guinea pigs with leisions was 63 out of 72 distributed as follows: liver, 31; testicle and epididymis, 17; kidneys, 11; bones, 2; lungs, 2. They were divided into macroscopic abscesses, microiscopic abscesses and granulous leisions. The typical appearance of an abscess was a globule of cheesy material enclosed in a tough thin white casing. In the liver or spleen these were loosely embedded in the surface layer but in the testicle they were firmly bound within the fibrous tunic and rarely as large masses in the pelvis or neck. The abscesses were composed of 3 layers: a core of homogeneous cellular debris mainly granulocytes; an intermediate zone of granulomatous tissue made up of epithelioid and granulocytes cells; and an outer layer of fibrous tissue containing collagen which formed a circular skin around the entire abscess. The proportions of the necrotic core and the intermediate zone were variable. The structure of macroscopic and microscopic abscesses were similar. In both guinea pigs and mice, microscopic abscesses were effectively absent in B abortus and the the granulomatc leisions that did occue did not develop into abscesses and fibrous capsules. Granullomas and absceses in the liver occupied the any portion of the lobules with possible periportal hepatitis due to infiltration of epithelioid cells, granulytes, leucocytes, lymphocytes and plasma cells which could resemble cirrhosis. Abscesses and granulous leisions were also present in the testicles and epididymis. In the kidneys the leisions were different. They wre extended irregularly and difusely through interstitial portion of the cortices. Necrosis, suppuration or fibrosis did not occur in these cases. Visual leisions of the bone are unusual however microscopic examination of bone marrow demonstrated that granulomas wre often present.

Hillaert (1950) and Ueda (1969) found that male guinea pigs experimentally infected with B abortus by subcutaneous, intraperitoneal, intravenous and oral routes develop high concentrations of Brucella in their semen. However experiments to sexually infect female guinea pigs were unsuccessful.

Renoux (1950) and Bosseray (1974) demonstrated vertical transmission and persistence of Brucella melitensis from infected guinea pig dams to their young. Bosseray studied the congenital or neonatal transmission of brucellosis in 13 pregnant female guinea pigs challanged with B abortus strain 544. 5 aborted, and 8 had normal parturition. In one case 2 offstring were stillborn. All the surviving young were tested at 2 months and were found to be infected with Brucella. Also Bosseray (1982) has studied mother to young transmission in a mouse model. Pregnant mice were infected with 'B abortus and the infected outcome of the offspring tested by Brucella culture method post mortem. Approximately 60% were shown to be infected at birth and remained so at 30 days. Infection was lodged main mainly in the liver but was also present in the spleen and lung. Feeding infected milk from infected dams caused a limited increase in infection but it should be noted that infection due to infected milk has been recorded in kid by Renoux (1955) and in calves by Carpenter (1926) Vertical transmission of 'Brucella has also been recorded in goats by Alton (1955) , in mink by Prichard (1971) and in cows See: Chronic Brucellosis in Animals.

Bosseray (1982)

Chronic Brucellosis in Animals

Evans noted that studies have shown that brucellosis becomes chronic in ever species of domestic animal that has been studied including: goats, cattle, swine, and horses. By common consent of veterinarians and farmers such animals never really recover and are thereafter of relatively little value.

Sheep and Goats: In 1922 Bassett-Smith commenting on the work of Zammit noted that a goat once infected with brucellosis never recovered. Although the goat could appear healthy its milk after kidding always contained brucella and brucella was always found in its organs. One goat could infect hundreds of people. Polding found that latent localised infection persists in goats for several years following acute attack. Garrido-Abellan notes that although some goats recover from exposure to brucella melitensis without the disease becoming established, other progress first to bactereania and then to establishment of brucella in the retino-edotholial system includine the udder and spleen and limph gland so that a chronic state of illness becomes established. Signs of illness may be orchitis, epididymitis, hygromia, arthritis and mastitis. More likely the illness is self limiting or the goats become asymptomatic latent carriers and potential excretors. Over time the infection particularly persist in the mammory, supermammory and genital lymph nodes so that brucella is secreted into the milk. Inspection of the organs shows grandulomous inflammatory leisons. The disease follows a similar course to that with brucella abortius in cattle. The antibody immune response is similar to that in humans IgM is usually noted in the first 14 days followed by IgG which then predominates. During the focal/chronic stage of the illness antibody levels drop to negative or low levels mostly IgG.

Taran (1972) compared the susceptiblity of Brucella inoculated guinea pigs, mice, rats and sheep. Guinea pigs inoculated with 10 -100 cells of B melitensis or B suis developed granulomatous lesions. In mice 10⁴ to 10⁶ organisms and in rats 10⁹ organisms while the sheep used were resistant to the strains of B melitensis and B suis used.

Tittarelli (2005) studied the persistence of infection in 46 ewes experimentally infected with B melitensis biovar 3 and monitored through three subsequent reproductive cycles lasting 151 weeks. The infection of the ewes and the recurrent presences of Brucella in milk and vaginal discharges, lasted throughout the duration of the trial. B melitensis was recovered from the tissus of one ewe killed at the end of the trial. B melitensis was recovered from vaginal swabs and milk following parturition in the third reproductive cycle from an ewe that had aborted in the first cycle but was not pregnant in the second cycle. The fluctuating persistence of B melitensis in milk throughout the entire trial indicating the continuous occupational risk and human health risk from consumers of fresh milk and milk products. All the infected lambs had negative antibody tests within 5 months indicating that such tests were of no practicle value. The infected carrier lambs represented a mechanism for persisance of the disease in a flock.

Grilló (1997) studied 41 pregnant sheep with positive serological or allergic tests to B melitensis. The ewes were kept isolated for lambing (62 lambs) and thereafter during lactation. After weaning the ewes were slaughtered. 14 ewes excreted B melitensis during lactation and 17 were found to be infected postmortem. The lambs were reared in a clean pen. B melitensis was not isolated from 7 lambs which died after birth or from 8 seronegative lambs which were slaughtered between two and seven months after weaning. However, one permanently seropositive lamb born to a culture-negative dam was found to be infected postmortum five months after weaning. The remaining 46 lambs were reared until adulthood and slaughtered at intervals for bacteriological study. Four ewe lambs (two born to culture-negative dams) were found to be infected postmortem, but were negative in immunological tests for B melitensis.

Leal-Klevezas (2000) observed that detectable amounts of antibodies are not recorded in the first 12–16 days after artificial inoculation of goats with Brucella abortus. On the other hand, Titteral (2007) noted that when the disease became chronic, the antibody titre could fall to undetectable level which is especially the case in intracellular organisms like Brucella spp. Latent infection without seroconversion complicates this problem particularly in pre-pubertal animals.

Cattle: In early work it was constued on the basis of agglutination tests that cattle could be immune to brucellosis or recover from it. For example: Huddleson (1931) monatored an B abortus infected dairy herd over 4 to 8 years. Of 227 cows, 41 remained immune to disease as indicated by the agglutination test. Their results were as follows: 189 animals with SAT at ≥1:100, 16 reverted to ≤1:25; 58 animals with SAT at 1:100, 18 reverted to ≤1:25; 26 did not develop titre above ≤1:25. Hadley (1931) found that with artificially infected cattle SAT became negative within 2 years. They could not diistinguish between infected animals with negative serum agglutination tires and infection free animals. Other work suggested that agglutination results were unreliable. Evans (1954) notes that cows with no apparent illness can secret Brucella in there milk. Evans also quotes the work of Wight who found during brucellosis eradication programs 33 to 50% of cattle infected with brucella never fully recover. Simpson (1941) notes that many cattle with B abortus excrete the organism in large numbers in the milk and vaginal secretions but without apparently manifesting symptoms of illness (abortion, sterility, mastisits or reduced milk yields). Harris (1943) noted that in cattle a negative serum agglutination titre can occur at the same time as positive Brucella culture. While Reddin (1965) wrote that cattle with low titres agglutinins which show epidermiologic or clinical evidence of infection secrete Brucella in milk and Brucella can also be recovered from tissue. Infection of the udder lymph nodes is associated with infection of the milk. Mann (1973) notes that that cows with negative Brucella serology can abort yet subsequently produce health calves while producing infected milk. In such cattle post-mortem examination could most likely reveal Brucella infection. Stableforth (1959) notes that cows with B abortus which have aborted on the first gestation can successfully subsequently produce full term calves. In the case of cows which only abort once Brucella can be found subsequently in the lymph nodes of the uteris, the udder, and the digestive system and other internal organs, spleen, liver. etc. Such cows which often become serologically negative by a range of Brucella antibody tests will continue to be infected for their lifetime and are often referred to as latent carrier cows. This has been demonstrated by Christie (1968) , Sjollema (1967) , Christie (1969) , Nicoletti (1966) . Morgan (1974) , Robertson (1974) Luchsinger DW .

• Nelson (1966) studied 167 cows from 97 problem herds for B abortus using culture tests. Problem herds are those where brucellosis eradication precedures fail because of the presence of carrier cows which cannot be identified by standard Brucella serology tests. Such cows are similar to humans with chronic brucellosis. The cattle were slaughtered and a range of organs removed and tested for Brucella. The organs tested were: the iliac, lumber, ischiatic, retropharyngeal and supramanmary lymph nodes; spleen; tissue from the 4 udder quarters; and uterus. In some cases positive cultures were only recovered from one site. The recovery rate from different organs was: iliac lymph node, 38%; lumbar lymph node, 25%; ischiatic lymph node, 24%; retropharyngeal lymph node, 35%; suppramammary lymph node, 80%; spleen, 37%; udder, 59%; uterus, 42%; milk, 45%.

• Saddler (1960) reviewed the distribution of Brucella in meat. He noted that on the basis of work by Doyle (1935) , McCullough (1951) and Washko (1952) found that Brucella could be distributed throughout the body of infected cattle. This includeed: liver, prescapular lymph node, popliteal lymph node, iliac lymph node, prefemoral lymph node, deep inguinal lymph node, thigh muscles, vertebral marrow, pharyngeal lymph node, submaxillary lymph node, atlantal lymph node, parotid lymph node, mammary gland, supramammory lymph node, mediastinal lymph node, bronchial lymph node, spleen, hepatic lymph node, tonsils, splenic lymph node, illio cecel lymph node, mesenteric lymph node, sternal marrow, knee fluid, bile. These cows could have negative or low serology antibody titres to Brucella. Similar results were found for the distribution of Brucella in infected pigs.

Carpenter (1926) showed that B abortus could be recovered from the milk of cattle without serum agglutinins.

Stableforth (1951) notes that there is no way to determine if cattle, sheep, goats and pigs with chronic brucellosis suffer the typical symptoms of lassitude, chills, headache and other ailments such as aflict humans.

ter Huurne (1993) notes the significance of latent Brucella infection in calves. Persistent infection of B abortus can occur in utero or after ingesting infected milk. The usual serology tests are negative in these cases usually until near the first calving or abortion, These calves are a source of reinfection in herds.

Bercovich (1998) notes that traditional serology tests are ineffective in the detection of B abortus in apparently healthy single animals with latent infection. The results are inconclusive at best. Such animals jeopardize the Brucella-free status of herds of cattle. He suggests the use of the skin delayed-type hypersensitivity test.

Dolan (1980)

Lapraik (1982) describes the case of an apparently healthy heifer calf from an infected herd was moved to a fresh herd free from brucellosis for 9 years when B abortus was isolated from its milk.

Lapriak (1975) describe a calf born to an dam infected with B abortus biotype 1 which was serologically negative to RB, SAT, and CFT serology tests from 4 to 6 months but yielded B abortus biotype 1 post mortem after 6 months.

Hornitzky (1986) studied 72 cattle at slaughter. B abortus was isolated from 49. They noted that B abortus could be isolated from cattle with negative serology tests by RB and CF tests. ELISA gave the best results.

In general transmission of Brucella by ticks, fleas has not been proven, Meyer (1977) .

Booseray (1987) in a study brucellosis in pregnant mice did not consider that there was a tropism for Brucella in the placenta but rather that concluded that it is a privileged site of Brucella carried there in blood and the prolonged pregnancy in animals such as cattle allowed high growth.

Fitch (1941) found that in cattle for calves born to infected dams the serological response declines to undetectable levels at about 6 weeks.

Morgan (1969) notes that the serological response in cattle to B abortus and B abortus s19 vaccine is similar. IgM is first produced followed by IgG in calfhood vaccination IgG ususally disappears at about 6 months but but IgM may persist much longer. n chronic carrier cattle serology titres fall to low or negative values and the only IgG is detected.

One of the most important caused of transmission of brucellosis in cattle are vaginal discharges of infected cows. This can occur some after infection, Philippon (1970) but the main period of heavy contamination is during a period of weeks following abortion or full term parturition and continues with diminution forr up to 3 months. Thereafter the cows become carriers and continue to secret B abortus intermittently for many years, Herr (1980) , Manthei (1950) and Philippon (1970) .

Calves: In early work it was concluded that calves had innate immunity to B abortus. Huddleson (1942) reviewed brucellosis in calves and noted that the accumulated evidence showed that calves up to a certain age have a high resistence to Brucella.. He fed calves Brucella contaminated milk and they either produced no agglutinating antibodies in their blood or if they did these disappeared within a few months. When these calves were later slaughtered no evidence of Brucella could be found. (Tese results suggest a posible problem with the fed). Carpenter (1924) fed B abortus infected milk to calves and found that the tissue of such calves was infected but after 5 weeks this disappeared. However Birch (1925) showed that such calves did not retain an immunity. In an experiment 9 heifers which had previously been infected as calves were exposed to infection of these 8 became infected. Huddleson (1942) noted that a small percentage of calves become infected with Brucella and remain so thereafter. Other studies by Fitch (1941) and McEwen (1950) generally accepted the idea of innate immunity. This early work is confused. More recent work has shown that such calves have an apparent symptomless chronic infection. The impact of cows with latent brucellosis in their calves has been reviewed by Sunderland (1990) . Calves born to infected cows initially have high antibody serology titres which fall to negativity after 4 to 6 weeks. More recent work by Morgan (1971) , and Robertson (1971) indicated that such calves remain apparently well with negative serology until the presence of infection may be detected during pregnancy and abortion. It has also been shown by Hignett (1964) , Plommet (1971) and Lapraik (1975) that female calves infected in utera can become latent carriers. Plommet studied 22 heifers born to artificially infected cows. He separated them at birth without colostrum and isolated them. Upon masturation they were artificially inseminated and after birth or abortion they and their calves (including aboreted) were killed and examined post-mortum. B abortus was isolated from 4 of the 22 heifers. All calves were serologically negative and remained so until 8 weeks prior to parturation.

It is generally concluded, Cunningham (1977) , Jubb (1970) and Atableforth (1959) , that calves may be infected by an Brucella infected dams; in utero, by asparation of amniotic fluid during parturition or after birth by ingestion of contaminated colostrum or milk. Such calves develop a serotitre against Brucella (partly immunoglobulins derived from colostrum) which is generally rapidly lost. Such calves seem to have resistance to brucellosis. However in some animals the disease is only dormant,

In a review of B abortus infection in cattle Nicoletti (1980) concluded that although it had previously been thought that calves are quite resistant to brucellosis and infections aquired in utero or by ingestion of milk were always temporary, more recent information indicates that a small but significant group of heifer calves with negative serology tests remain infected and abort or show evidence of infection during the first pregnancy. This is termed latent infection and such animals can spread the disease in later life often to other herds, . Wilesmith (1978) concluded that 2.5% of such calves beca\me latent. Nicoletti (1980) concludes that latent infections are more common than previously suggested and difficult to diagnose and present a serious impediment to the elimination of brucellosis in herds.

Hadley (1931) subjected 33 cattle to artificial infection with Brucella of these 13 remained uninfectered although they developed antibodies for a short time. Beach (1935) kept these 13 cattle free from further exposure for 5 years after which they were again subject to artifical infection. Only 1 became infected. Of 22 similar uninfected contols 11 became infected.

Pigs: Hardy (1930) , Feldman (1934) , and McNutt (1934) noted that brucella may exist in the tissue of apparently health swine. Carpenter examined tha blood of mares over a 2 year period and found that that despite lack of clinical signs some of them were infected with brucellosis. Feldman (1933) working with apparently health swine with positive agglutination tests found that Brucella could be isoloated fron internal organs even if leisons were not present. They noted spodylitis in some swine and cultured Brucella. Evans (1954) points out that this parralels lumber pain in humans without objective signs of illness is common in humans.

B suis occur in acute and chronic forms in pigs with lesions in inernal organs. Pigs commonly have negative serology tests and positive culture tests (eg. Manthel (1948) .

Saddler (1960) has corelated results for the distribution of Brucella organisms in the meat of infected pigs based on the work of McNutt (1935) , Huddleson (1933) , Hutchings (1951) , Borts (1946) , and Johnson (1931) and (1933) . Brucella was found to be present in the following lymph nodes: lumber, inguinal, right perfemoral, iliace, prescapular, submaxillary, cervical, suprapharyngeal, supramammamary, gastric, broncial, heptic and mesentric. It was also found in loin muscle, liver, spleen, and kidney. In the study of Borts (1946) B melitensis was found in the tissue of 20 pigs, of which 8 were serologically positive and 12 negative. In a few cases Brucella lesions were found in the spleen and lumber verebra.

Guinea pigs. Guerra (1973) noted that after intravenous injection of Brucella in guinea pigs it concentrates in the spleen and livcer and chronic infection occurs despite the reticuloendotheiial protective response.

Mice. Cheers (1984) has shown thatmice develop chronic infection when challenged with virulent and attenuated vaccine strains of Brucella. It has been shown by Birmingham (1981) and Ho (1982) that Strain 19 caused a chronic infection in CBA/H mice but not in BALB/c mice.

Brucellosis in Children.

Spink (1953) observed that while children under the age of 12 are often exposed to Brucella they are less likely to show overt signs of illness.

Gotuzzo (1987) in a study of 39 families exposed to B melitensis found an infection rate of 50.9%. The likely source of infection was contaminated food. He noted that children under 10 years were less susceptible with an infection rate of 26% compareed with adults at ~60%. Moderate and severe illness was more common in womwn (41.8%) than men (13.5%). Possible explanation s of diffences in attack rate are suggested. In children the lower attack rate may be due the more effective clearance of Brucella by the oropharyngean lymphoid tissue and relatively lower gastric ph. In older parients the higher attack rate may be due to a decline in the effectiveness of the immune system.

Dooley (1931) and (1932) studied 263 boys who had injested Brucella infected raw milk from a known infected herd. These boys showed no manifestations of ilnness although 41% had SAT tires of 1:40 to 1:1200. 15 boys who had drank the milk for 1 to 10 years showed no agglutination titre and all were symptomless. Some other boys retained an agglutination titre of ≥1:320 but remained symptomless. On the basis of his own work and that of others he concluded that young children are less susceptible to Brucella infection than adults. He compared this to similar results on calves reported by White (1926) who noted that irrespective of whether the cow is infected or not calves are born with negative agglutination titre. They have a positive titre from sucking colostrum/milk in the first 24 hours but are again negative by 6 months. but infection can be established after sexual maturity.

Chronic Brucellosis in Humans.

Brucellosis is then established as a chronic illness by infected monocytes, macrocytes and possibly neutrophils transfering bruceela to the reticuloendothelial system : liver, spleen and lymph glands and kidneys and also other organs including bone marrow, the brain and then the development of complications and foci of infections. Geyik (2002) indicates that chronic Br is usually caused by persistent deep foci of infection such as suppurative leisons in the bone, liver, spleen and kidneys. According to Alton (1994) in humans leisons produced by Br consist of minute granulomas composed of epithelioid cells, polymorphonuclear leukocytes, lymphocytes and some giant cells. In Br mellitensis these granulomas are particularly small although the associated toxemia is great. This means their physical detection is extremely difficult.

Meyer (1943)

Sundberg (1947) reviewed 9 patients with active brucellosis (both acute and chronic) in terms of granulomas type leisions. In this series of 9 patients with B abortus: 9 had positive SAT tires, 8 had positive blood cultures. 2 of 8 had positive bone marrow cultures, 4 of 9 tested had granulomas leisions in bone marrow, all 5 of 5 tested by biopsy had granulomas in the liver. the bone marrow of the 4 patients was studied in detail. The granulomas leisions were variable: diffuse or small and nodular and some with diffuse boundaries. The cell type comprising the leisions were: epithelioid, gaint cells (Langhan, epithelioid and rarely foreign body), lymphocytes/monocytes, plasma cells/eosinophils, disintergrated cells and neutrophils (central and peripheral), peripheral zone comprising mainly eosinophils and plasma cells, and fibrinoid material. Leisions varied in size from a large nodule of 0.5 mm diameter to minute leisions representing oonly the agglomeration of 3 to 4 epithelioid cells. Nodules consisted mainly of epithelioid and giant cells. Cell destruction could be considered as partial necrosis and accumulations of neutrophils was abscess like. Caseous necrosis similar to tuberculosis was not found. The smaller leisions were composed mainly of densely packed epithelioid cells peripherally surrounded by lymphocytes and eosinophils and were similar to tuberculiod nodules.

Bone Marrow Eyre (1908) was the first to observe abnormalities lesions associated with B melitensis in human bone marrow. These lesions included red cells, giant cells, mononuclear and lymphoid cells. Smith (1912) and Fabyan (1912) described guinea pigs infected with B abortus showing inflammatory nodular foci in various organs. They noted destructive and proliferation leisions in bone marrow. Smith (1925) studied the effect of B abortus, B melitensis and B suis on guinea pigs. He noted that leisions occuted in the thymus, lymph nodes, spleen and liver The leisions were either minute and greyish or larger with yellosh cheesy foci (For B abortus and B melitensis from 1 to 3 mm and for B suis upto 5 mm). Loffler (1930) noted the presence of fibrinoid material in granulomas leisions. He noted that inflamminatory cellular infultrates were thickest in the peripheral zone. Wohlwill (1932) described the presence of lesions in: spleen, lymph node, liver and bone morrow and Wegener (1935) reported epithelioid nodules in human bone marrow with some necrosis.

Chronic brucellosis caused by localised lesions.

We have seen that chronic brucellosis is an accepted diagnosis in animals even without positive serology tests. Such cases have been proved by post mortem test where brucella lesions have been found in internal organs. Generally cutting humans apart to prove they are ill is not an acceptible option however there are some cases where brucellosis leisions have been found in the internal organs of humans by chance during unrelated surgery. Thorne noted that in chronic brucellosis abscess formation can occur at any time and allow positive cutlure verification. Williams (1982) notes that serology tests are sometimes negative with positive culture tests. Examples of this are presented by Martin (1982) and others.

• Wainwright (1931) isolated Brucella from an ovarian cyst after 5 years illness.
• Amoss (1931) isolated Brucella from the fallopian tube, the Mesoappendix, the appandix and an ovarium cyst after 7 years of illness. In another case of 3 years Brucella was cultivated from hemorrhagic ovarian cysts of the ovaries and from duodenal drainage and the gall bladder. In both cases SAT was negative throughout and no abnormalities were found by X-ray or physical examination.
• Evans (1937) notes the case of a patient who was assessed in 3 hospitals and a leading teaching hospital during a year of illness none of them diagnosed the illness. During a separate operation Brucella was cultured from a focus of infection.
• Spink (1946) reported a patient ill with Br suis for 19 years after which Brucella was isolated from a heptic abscess.
• Weed (1952) reported a patient ill for 13 years prior to isolation of Brucella from a inguinal lymph node.
• Osmundson (1959) reported a case where Brucella was isolated from a spleen after 14 years of illness.
• Shirger (1959) recovered Brucella from an infected speen after 17 years of illness.
• Schmidt (1959) isolated Brucella from the spleen of a patient 10 years after the initial illness.
• Carpenter (1979) described a man of 25 who was ill for 13 years with Br after contact with infected cattle who eventually was diagnosed by culture from a groin abscess. During his illness he was tested 10 times by culture and serology without success. Finally he tested positively by an indirect fluorescent antibody test.
• Alice Evans (1947) contracted brucellosis during laboratory work she was ill with undiagnosed chronic brucellosis for 17 years. She was first ill with a subacute illness for 9 months and was given a diagnosis of neuroasthenia, than came an acute phase of illness diagnosed as brucellosis melitensis by culture from a blood sample. Then she had 5 years of poor health often seriously incapacitated. During tis period she was examined in 4 hospitals and again given a diagnosis of neuroasthenia. Finally she had an operation for another illness during which brucella leisions were found. She points out that these type of misunderstandings that tries tha patient almost beyond endurance.
• Mantur (2006) refers to a patient with epididymo-orchitis with negative serology fot brucellosis. However agglutinates were demonstrated in by RB and SAT in testicular fluid. This was confirmed by culture test.
• Raptis (2007) .
• Payne (1974) describes a man who was ill for 6 years with chronic brucellosis abortus who was tested 7 times with SAT, AHG, and CFT with negative results. Eventually he was diagnosed by a posative culture test.
• Poole (1975) describes a veterinary surgeon ill with mailase without diagnosis or treatment for 3 years. Afeter three years he was trested serologically. His test results were: SAT, <20; CFT, <20; and AHG , 20 but he had a positive culture for B abortus.
• Yorgancigil (2003) describe at patient with recurrent knee effusion for 20 years. After this time he was tested for brucellosis by culture and serology tests, SAT and Coombs, which were negative. Diagnosis for Brucella melitensis was acheived by culture of knee fluid.
• Ariza (2006) report a man ill with brucellosis and apparently cured and living with no apparent risk factors for brucellosis suddenly taken ill after 40 years withhigh fever, chills, hypertension, peripheral hypoperfussion, scleral icterus and painful heparic enlargement, and unknown source os septicemia. Brucella test were: RB, -ve; SAT 1:40; culture, -ve but seroconversion was later noted and brucellosis confirmed.CT indicated a Brucella granuloma which was resectioned. The patient received 5 months of treatment and then returned to health for 4 years.
• Ariza (2006) a woman was diagnosed and apparently cured and remained well for 20 years. During this period she had no risk factors for brucellosis. Then she suffered a relapse with high fever, chills, sweating, arthralgia and weakness, hepatomegaly, splenomegaly and right sided pleural effusion.Tests for brucellosi were negative. After 4 weeks, brucellosis was confirmed by SAT and AHG tests. CT revealed a reactivated Brucella liver granuloma. She recovered unwith antibiotics but became ill again 7 years later after a surgical hepatic resection and antibiotic treatment.
• McDermitt (1994) describes a 39 year old man with fever, inguinal adenopthy and a leision on his knee diagnosed with brucellosis by SAT/AHG tests. The knee leision remained and seven years later a purulent discharging sinus developed . SAT and AHG were again positive for B abortus and culture of resectioned materal grew B abortus.

Al Ali (2008) has noted that in the subacute stage of brucellosis with low antibody titres correspond to inhibited levels of IL-2, IL-12, IFN-γ and TNF-α.

Busila (1975) Reviewed that history of 53 cases of chronic brucellosis over a period of 9 years. The patients demonatrated typical symptoms localised manifestatations including: cardiovasular, hepitic, nervous, osteoarthricular, orchitic, ocular, and cutaneous.

Goldstein (1934) reports 2 cases of brucellosis which were proved by agglutination titres of 1:640, a positive Brucella antigen skin test and a positive Brucella culture from feces but not blood. In one case Brucella was cultures from tonsillar exudate.

Carpenter (1930) and Giordano (1930) reported cases of brucellosis with positive Brucella tests but agglutinsation tests below 1:30 of whom 5 had negative agglutination tires. Goldstein (1934) notes that many such cases have been described by european workers.

Hardy (1935) consider that Brucella can establish a focus of infection in the body and stimulate antibody production apart from manifestations of the disease. The patient is not ill but but an exacerbation can occur at any time.

Baker (1929) describd a case of brucellosis with intermittent hydrarthrosis with positive cutlures for blood and synovial fluid and urine in the presence of negative SAT.

Amoss (1931) describes a case of brucellosis with intermittent hydrarthosis of seven monthsa duration having positive culture of blood and joint fluid but persistently negative blood agglutination titres. He also notes another case of a patient with debiliating symptoms but without agglutination titre win which Brucella was isolated from drainage of bile and gallbladderand a ovarian cyst. 7 years earlier Brucella had beren isolated from blood and an agglutination titre had been present.

Interaction of Brucella with the immune System of Animals and Humans.

In this section we will look at the scientific evidence which supports the existance of chronic brucellosis. This mainly involves the interaction of Brucella with the immune system. More detailed reviews has been undertaken by Baldwin (2006) and Splitter et al (2003) .

Anergy in Chronic Brucellosis

In 1902 Bassett-Smith A noted that phagocyte power of leucocytes in chronic brucellosis was feebly active and deficient. Harris refers to a case of chronic brucellosis of 15 years of illness treated successfully using a brucella abortus vaccine. He noted that the opsonocytophagic test before treatment showed absense of brucella resisance but increase after treatment indicating suppression of the immune response in chronic brucellosis. Thorne noted that chronic brucellosis patients had suppression of the cellular inmmune system. He found that they had lower T-lymphocyte levels and did not respond to a test for delayed hypersensitivity with 2,4-Dinitrochlorobenzene (DNCB). Once their T-lymphocytes were restored they became sensitive to DNCB. He considered that showed anergy. Cheers (1979) a reason for the chonicity in brucellosis was the insensitivity of brucella within "incompetant" macrophages. Boura (1983) noted that because brucella is located primarily in the monocyte and macrophage where they can replocate protected from antibodies and antibiotics. This situation plays a key role in chronic brucellosis. He measured phagocytosis and found this lower in chronic brucellosis patients than controls. Since phagocytosis is an important means of eliminating bacterial parisites this is important in establishing chronicity of brucellosis.

The response of the adaptive immune system to Brucella infection.

It seems that Brucella has the ability to interfer and suppress the function of the immune system using a range of survival strategies. In recent years attempts have been made to explain this anergy. Control of brucellosis in humans depends on cell mediated immunity and the (Th1) immune response involving IL-2, IL-12, IFN-γ and TNF-α. According to Splitter et al (2003) the optimum immune response to Brucella infection of machrophages is as follows:

• Brucella is internalised by machrophages.
• IFN-γ is produced by CD4+, CD8+, and γδ T cells activated by Brucella function in macrophages to hamper Br survival.
• Cytotoxicity of CD8+ and γδ T cells destroy the infected macrophages.
• Th1 type isotopes such as IgG2a and IgG3 opsonise the pathogen to facilitate phagocytosis.

However this process is defective in chronic brucellosis. The function of TNF-α, IL-12 and INF-γ which are essential to the elimination of this parasite are apparantly impaired.

Typical mode of Brucella infection.

Brucella is a facultative intracellular parasite. It has no persistent environmental reservoir for infection so it must survive intracellularly. It infects monocytes, macrocytes, and nonprofessional phagecytes such as epithelial cells and fibroblasts. Ziang (1993) has shown in a murine study that after initial phagocytosis of machrophages the number of internalised Brucella abortus bacteria rapidly fell for about the first 12 hours and 90-99% are killed due to the oxidative burst but therafter Brucella numbers steadily increase. Smooth strains are more resisant to killing than rough variants. Similar results were found with nonprofessional phagecytes . The same results were found in humans by Caron (1994) for B suis and Rittig (2001) for B melitensis.

According to Baldwin (2006) the stages of infection are as follows:

• Phagocytosis of Brucella
• Survival inside the phagocyte against reactive oxygen intermediates/nitric oxide
• Activation of genes in responce to acidic phagocytes environment
• Prevention of phagolysosomal fusion by remodelling of the compartment (brucellosome)
• Replication and therefore establishment of reservoir of infection from extracellular response.

Brucella impairs the Destruction of Infected Phagocytes.

Impairment of TH1 Immune Response and INF-γ function.

INF-γ is a necessary procurser to facilitate the Th1 immune response with the secretion of IL-12, TNF-α, CD4+ T cells, more INF-γ and cytotoxic CD8+ T cells to destroy infected macrophages. Ziang (1993) has shown in a murine study that INF-γ can suppress the growth of Brucellain infected macrophages if added early in the infection but once infection is established it has no effect. Rodriguez-Zapata in 1996 studied acute brucellosis patients before and after treatment and found that in untreated acute patients levls of IL-2 were normal but levels of INF-γ were reduced. After successful treatment levels of INF-γ returned to normal. He also noted that there was a defect in the normal T cell responses. It was concluded that acute brucellosis infection rendered INF-γ production defective. Giambartolomei (2002) exposed peripheral blood mononuclear cells (PBMC) from chronic brucellosis patients to Brucella cytoplasmic protein (CP), Unlike controls they were unresponsive and produced no IFN-γ or IL-2 which were required to facilitates the Th1 pathway. Similar results were found by Zachariouadaki (2003) . Kimikli (2005) investigated the proliferatin response of T cell subsets to phytohemagglutin (PHA) in acute and chronic brucellosis and found that the stimulation of CD+ T cells was suppressed in chronic brucellosis. CD4+ T cells are an important producer of IFN-γ. Dormand J (2002) considers that the outcome of Br infection, recovery or chronic illness, dependes on a balance of factors. It seems that in chronic brucellosis there is insufficient IFN-γ to effect recovery but sufficient to maintain chronicity against death of the host.

Skendros et al (2006) have examined factors related to IL-2 and the Th1 immune response and showed that in ex vivo experiments on humans that CD+/CD25+ T cells are higher in acute brucellosis than controls and lower in chronic patients and concluded that diminished response of CD4+ T-lymphocytes expressing IL-2Rαis associated with chronic brucellosis. They also found that the diminished percentage of CD4+/CD25+ T cells in chronic brucellosis is not associated with inadequate CD80/CD28 co-stimulation. They co-cultured PBMC with PHA and E coli LPS and found a significant reduction of CD4+/CD25+ T cells in chronic brucellosis patients compared to acute brucellosis. Co-culture of PBMC plus Heat killed brucell abortus (HKBA) compared to E coli LPS- culture showed a considerable reduction of CD4+/CD25+ T cells in both cases. This indicats that brucella can down modulate immune response producing T cell anergy and chronic illness.

Impairment of TNF-α fuction.

Dormand (2002) has shown conclusively that human macrophages infected with B. suis produce no TNF-α although they do produce other cytokines such as IL-1, IL-6 and IL-8. However cells treated with killed B suis do produce TNF-α and Brucella supernatants suppresses production of TNF-α in E coli infection. Jubier-Maurin has shown that that TNF-α secretion is blocked by the Brucella wall protein Omp23b. The effect of TNF-α deficiency has been studied in mice. Mice are capable of recovery from Brucella infection in about 3 weeks. Murphy (2001) has shown that following infection in mice INF-γ is rapidly produced together with IL-1, IL-6 and TNF-α. Jubier-Maurin has shown that the Omp23b protein effect which suppresses TNF-α production in humans does not occur in mice. Studies by Jiang , Murphy , Zhan , Zhan have indicated that TNF-α is essential to initiate full anti-Brucella activity in machrocytes.

Murphy has shown that both INF-γ and TNF-α are necessary to clear the infection in mice. Jiang has shown that the absence of TNF-α inhibites INF-γ activity. Zhan and Cheers have shown that mice deficient in the TNF-α gene have a greatly exacerbated initial illness. They have also shown that in brucella infected mice TNF-α positively controls the early production of IL-12 and TNF-α essential to develop the Th1 type macrophage response generating INF-γ, CH4+ and Ch8+ cytotoxic leading to clearing of the parisite. He considered this a significant factor in chronic Br.

Apoptosis resistant monocytes and lymphocytes in chronic brucellosis.

A macrocyte can induce its own apoptosis to prevent intramacrocyte bacterial replication. Roup (1986) obverved that brucellosis infections inhibit spontaneously apoptosis in human monocytes when cultured in vitro. Gross (2000) has shown that prevention of apoptosis is mediated by soluble factors released by infected macrocytes. This is a statergy which favours the development of chronic illness. Tolomeo (2003) has shown that the monocytes and lymphocytes in untreated acute and chronic brucellosis patients are resistant to spontaneous or CH11 induced cell death. During the treatment of acute patients this persisted for 10-20 days but in chronic patients it was more than 45 days. He notes that the Fas.Fasligand mediated extinsic pathways is the main mechanism by which cytotoxic cell (activated CD8+) kill infected cells and this may be due to the overexpression of the A1 Gene. He suggests that chronic brucellosis may be induced by a combination of resistance to apoptosis in moncytes and low grade cytotoxic lymphocyte activation. He also suggests that this is a potential way of monatoring chronic brucellosis patients.

Gross (2000) have shown by various techniques that Brucella infection inhibites the spontaneous apoptosis that occurs in human monocytes in the absence activation signal and also makes macrophagic cells resistent to apoptosis induced by the immune system. They demonstrated that intracellular survival and proliferation of Brucella was corelated to monocyte survival. Both infected and noninfected monocytes were protected fronm apoptosis indicating the impact of soluble factors. They showed that this was not due to TNF-α and that Brucella LPS had only a moderate effect. They noted that IL-1, IL-6 and IL-12 are produced by B suis infected monocytes and could therefore be involved. They also showed some evidence of a possible involvement of the A1 mRNA gene. As well as this they pointed out that Brucella infection inhibits macrophage apoptosis induced by IFN-γ, Fas ligand, or possibly Th2 cytokins. They concluded that inhibition of monocyte / macrophage apoptosis by Brucella can occur at primary infection and after in immunological induced apoptosis processes. They concluded that Brucella like Mycobacterium are intracellular parasites for which monocytes are a preferred host. The bacteria can survive phagocytosis and replicate within the cells and these sites of infection are then protecting by apoptosis.

Dendritic Cells.

The main function if naive dendritic cells (DC) is to identify antigen material and then transfer it to its own surface using MHC to stimulate naive T cells. Contact with the antigen activates DC from naive to mature status able upregulate CD80, CD86 and CD40 and enhance T and B cell activation. DCs have an essential role in promoting a Th1 immune response by release of IL-12 and stimulating the production of IFN-γ. Billard has shown that DCs are very susceptible to invasion by Brucellaand therefore may act as a reservoir of infection to aid development of Brucella within the host. Billard (2007) also noted that Brucella prevents infected DCs from maturing and impaired their ability to present antigen to T cells or to secret IL-12 required for the Th1 immune response. They have shown that this inhibitionary effect involves the Omp25 from Brucella prevented the secretion of TNF-α by DCs. It is shown that by manipulating tha system either adding TNF-α or anti-TNF-α or using Brucella deficive in Omp25 that DC maturation and antigen presentation can take place. This may be an essential pathway for inhibiting the human immune respose and favouring chronicity. This work was done with smooth phenotypes of Brucella. Billard (2007) also looked at the interaction of human DCs with rough strains of Brucella. Smooth Brucella phenotypes express the O-side chain of lipopolysacchraride (LPS) while rough variants do not and rough stains do not usually produce chronic illness. She found that despite the negative effect of Omp25, DCs infected with rough stains induced maturation of DCs, production of IL-12 and stimulation of naive CD4+ T lymphocytes thus favouring clearance of the bacteria. This could indicate a further significant roll for LPS in chronic brucellosis. Further work is required.

B cell and Antibodies.

In the early stages of brucellosis infection B cell /antibodies intervention may eliminate the desease. The main antibodies which are produced are are IgM and IgG3 . Antibodies are elicitated by the O-polysaccharide side chain Brucella LPS. However once intracellular infection is achieved then the role of antiboies is limited. Antibody opsonisation is generally thought to favour the host because of enhanced phagocytic uptake and associated killing but probably not in this case. It is considered that the normal mode of entry of smooth type Brucella into macrophages is via lipid rafts . It is considered that opsomised antibodies (mainly IgG3) interacting with the macrophage FcR facilitates enhanced entry of Brucella . Although this tends to kill the bacteria some remain viable. Baldwin notes that in the classical Type 1 activated macrophage model Toll Like Receptors (TCR) of the macrophage interact with the microbe to induce production of IL-12 an inflamatory agent which stimulation the production of CD4+ and CD8+ T cells and therefore IFN-γ. She considers that the interaction of Brucella with TLR is a poor stimulator of pro-inflammatory cytokines but she considers that this is vital for host survival. She notes that IFN-γ deficient mice die from brucellosis. She also suggests that IgG3 may play a crucial role in favouring chronicity in Brucella infection. The IgG3/microbe complex interacts with the macrophage FcγRl to give Type II activated macrophages leading to the production of IL-10 which induces T cells to secrete IL-4 therfore inhibiting production of IFN-γ. This is an anti-inflamitory process which favours the survival of the bacteria and she postulates that this may be profoundly important. She notes that IFN-γ has an essential role in Brucella infection which affects the survival of the host and its tendency to become chronic. It should be noted howver that Rafiei (2006) found no corelation between duration of disease and IL-10 or IL-12 production in humans.

Spera (2006) in a murine study identified a gene prpA belonging to the proline-racemase family which elicits a B lymphocyte polyclonal activation and that stimilation of splenocytes with PrpA resulted in IL-10 secretion which was not observed with a B abortus-prpA mutant. The latter had a reduced capacity to induce chronic infection. They postulate that this effect could also happen in humans and that in the acute phase of the illness the transient secretion of IL-10 which is anti-inflamiatory would inhibit the inflamitary immune response favouring chronicity.

The response of the innate immune system to brucellosis infection.

Baldwin (2006) suggests that if the production of IFN-γ by the adaptive immune response is compromised by the action of Brucella then the innate immune system is recruited in an attempt to control the infection.

The complement system.

The complement system is a cascade of plasma proteins which interact with bound antibodies and or bacterial surfaces to opsonize kill them. The interaction of complement and brucella is via brucella surface LPS. Complement interactions with phagocytes is via complement receptors. There is no evidence for a specific role for complement in Brucell infection.

Neutrophils

Neutrophils are short lived immune cells which constitute up to 70% of white blood cells. Neutrophils fuction is phagocytosis and to combat brucella they must be activated by opsonisation. Neutrophils are a primary line of defense against Brucella invasion. According to Moreno (c2002) neutrophils are able to ingest and digest Brucella faster and more effectively than other cells. Up to 80% of Brucella is killed in 5 hours without apparent replication. However the survival of Brucella in neutrophils has been observed and therefore neutrophils may tranport Brucella to lymphoid tissue thus aiding spread of the infection in the body. Cheers (1996) observed that TNF/IFN-γ deficient mice challenged by Brucella produced a 10-fold increase in neutrophils and NK cells. In support of a role for neutrophils Dormand notes that Br suis infections are associated with increased levels of IL-8 and Refik (2004) reports that IL-8 is elevated in acute and sub-acute Br patients. A main function of IL-8 is the induction of chemotaxis in neurophils.

Natural killer cells

NK cells are primative T cells which have a significant role in combating acute infections particularly in the early stages. Dornand (2004) has noted that in the early stages only of a B suis infection NK cells can inhibite the development of B suis in infected macrophages. This was established in experiments in which B suis infected human macrophages were co-cultured with human NK cells. They found the inhibiting effect required direct between infected macrophages and NK cells and although both TNF-α and INF-γ are produced during this interaction their effect is limited. The main mode of inhibition was from induction of NK cell cytotoxicity against infected macrophages. However in a slowly developing infection such as brucellosis the role of NK cells seems limited. NK cells are weak producers of IFN-γ. Green notes that during the early stage of infection the role of TNF-α is to activate NK cells to produce IFN-γ with the potential to activate the macrophage Th1 pathway. Since Dormand (2002) has shown that TNF-α secretion is severely inhibited in human brucella infection this means that the cytotoxic activity of NK cells is repressed favouring the progression of the illness. Inhibition of NK cell functions has been confirmed by Salmeron , Splitter (2003) and Rodriquez-Zapata . Dormand (2002) has shown that NK cells can be activated to kill infected macrophages but suggests that the reduced levels of TNF-α impair the killing activity favouring establishment of brucellosis and chronic illness. Zaitserva (1996) suggests that in chronic brucellosis the effectiveness of NK cells may be inpaired by low levels of IL-12.

γδ T cells

γδ T cells are part of the early immune response to infection and are considered a bridge between the innate and adaptive immune response. They exist in primates but not mice. γδ T cells make up 1-5% of circulating T cells of which Vγ9Vδ2 T cells are predominant. They respond very quickly to phosphorylated nonpeptide antigens released by intracellular pathogens. This has been shown for B suis infected macrophages by Ottones et al. and they have also shown that Vγ9Vδ2 T cells thus activated impair intracellular growth of B suis. Bertotto (1993) has shown that the numbers of Vγ9Vδ2 T cells dramatically increase in the peripheral blood of acute brucellosis.

Oliaro (2005) have shown that interactions of Vγ9Vδ2 T cells with B suis involves a contact dependant mechanism and by release of soluble components which kill B suis by a potent bactericidal effect. During contact dependent interaction Vγ9Vδ2 T cells express perforin, granzynes and granulysin . leading to the release of lysic granule which lyse infected cells and cell apoptosis. This study seems to suggest that while there is lysis of infected cells this does not impact on B suis viability. They also studied FasFasL interaction. mediated apoptosis of infected macrophages had no effect on B suis viability. FasL is expressed on the surface of activated Vγ9Vδ2 T cells. They concluded that FasFasL interaction induce apoptosis of infected macrophages had no effect on B suis viability. It is also of note that B suis released from lysed cells can reinfect other monocytes. Vγ9Vδ2 T cells activated by nonpeptideic anigens from infected macrophages release IFN-γ and TNF-α</ref> and they . They showed that IFN-γ has some significant effect in killing Brucella but TNF-α had none. Finally they determined that an undetermined soluble factor produced by Vγ9Vδ2 T cells was the main antimicrobial agent. Dudal (2006) have shown that this soluble factor is LL-37. This is the mature form of cathelicidin express by T cells. They show that it has a bactericidal activity against B suis. Dortmand has postulates that this is part of a balance of factors for and against Brucella elimination which leads to chronic illness.

Conclusions.

There are some doctors who claim there is no such thing as chronic brucellosis. In this section we have reviewed the compelling scientific research which attempts to explain the mechanisms by which brucellosis becomes a chronic illness. We have seen that brucella invades and is able to survive in machrophages as an intracellular parisite but that the cell mediated immune response by which macrophages are activaled to produce IL-12, CD4+ T cells, CD8+ T cells and INF-γ and therefore eliminate the ilness is compromised. It has also ben shown that Brucella can prevent apoptosis of infected mocrocytes favouring chronic illness. Evidence has been compiled which shows that both INF-γ and TNF-α are essential to recovery from brucellosis but that their availablity and function is impaired in human brucella infection. Lack of TNF-α also impairs the function of NK cells which favours establishment of the illness. γδ T cells apparently continue to function and aid survival of the host. It is considered that there is a balance of factors which prevents death of the host but leads to chonicity. In animals chronic illness may be proved by recovery of Brucella from the animals organs. Results from PCR tests show that brucellosis patients who have apparently recovered from the illness still secret Brucella DNA which is indicative of a residual chronic infection.

Testing for Brucellosis in Humans.

Apart from the positive culture test there is no completely reliable and satisfactory diagnostic procedure for brucellosis. Salient points are: less than 10% of chronic brucellosis patients have an abrupt febrile onset as in acute illness, the diagnosis must inevitably rely on the unreliable laboratory test.

General blood tests.

ESR and CRP are widely used to as inflamitary markers but on the basis of accumulated data they seem to be of little use in diagnosing brucellosis. Studies of the value of ESR as an aid to the diagnosis of brucellosis have been carries out by Agnew (1949) , Madkour (1989) , Solera (1995) , Kologlu (2006) and Demiroglu (2006) Most of the patients these studies were acutely ill and typically showed ESR values slightly elevated (>20 mm/hr) in about 40-80% of cases. In the absense of serious complication the results are lower. Colmenero (1996) reports typical data in a study of 530 hospitalized patients. He found that 47.2% of patients had ESR from 30-50 mm/hr and 19.5% >50 mm/hr. I a further in a study of 100 patients he found average ESR of uncomplicated cases at 21.65 mm/hr and complicated at 31.64mm/hr. Shakin (1987) noted average ESR values in neurobrucellosis of 22 mm/hr and considered this unhelpful CRP has been measured in brucellosis by Al-Kassab (1991) , Tohme (2004) and Demiroglu (2006) and they typically found CRP to elevated in 30-73% of acute patients. These sort of hit and miss values for ESR and CRP offer vey little in the diagnosis of acute or chronic brucellosis. Williams (1982) notes that examination of peripheral blood in chronic brucellosis is usually unhelpful. He notes (1973) that leucopenia with relative lymphocytosis can occur in acute brucellosis this rare. The ESR is sometines raised in acute brucellosis but is nornal in chronic brucellosis.

Pourbagher (2006) reviewed the findings of 4 general hematological and biochemical test for 251 brucellosis patients: 92 acute, 48 subacute and 111 chronic. Unusually in this series subacute is taken as patients ill from 3-12 months from onset. The results were as follows:

• ESR (elevated): Acute, 58.7% subacute, 54.6%; chronic, 34.2%.
• Anemia: acute, 34.8%; subacute, 27%; chronic, 27%.
• CRP (elevated): acute, 38%; subacute, 22.9%; chronic, 10.8%.
• Transaminases (elevated): acute, 20.7%,; subacute, 8.3%; chronic, 2.7%.

These results show that ESR, CRP and transaminases results are lower for chronic brucellosis and offer nothing as a diagnostic tool in that case.

Simpson (1941) considered that leukopenia was likely to occur in patients with acute brucellosis. Chronic brucellosis patients could have leukopenia, normal leukocyte levels or moderate leukocytosis. He notes that the ESR is not usually high unless complications are present. In chronic brucellosis a lowered basal metabolic rarte and vascular hypotension can occur.

Young (1995) considered that routine blood tests are generally unhelpful in brucellosis

Direct Tests for Brucella.

The Brucella Culture test.

Brucella are small gram-negative coccobracilli that lack capsules, fragellae, endospores and native plasmids. They can be grown in any high quality peptone based medium supplimemted with serum. They are aerobic but may require CO² to enable growth. Hardy (1930) noted that the charactoristics of B abortus, B suis and B melitensis were similar. They are small (0.3 - 0.5 by 0.6 - 1,5 microns), nonencapsulated, nonmotile bacteria which are decolorised by Gram's method. Coccoid and bacillary forms also occur as well as oval shapes. Coccoid forms predominate in infected tissue and bacillary forms in culture.

The Brucella culture test is considered a gold standard for detecting brucellosis and a positive result is virtually certain proof of infection. However the test has a significant failure rate and is of little use in the study of chronic brucellosis. Usually blood is tested although Brucella can be cultured from a range of bodily secretions and tissue but the success rate is variable. This test works reasonably well in acute cases of B melitensis and B suis but seldom for B abortus.

Test Procedures. Tests are normally carried out on blood. Tests are only successful if the patient is bacteremic. Patients with brucellosis often have a rising temperature profile which is at or near normal during the morning but higher in the evening. In view of this, Borts (1943) considered that it was less likely that brucella could be isolated in the morning when temperature was normal than later in the day when the patients temperature had risen. Al Dahouk (2003) notes samples should be taken during the pyrexial phase when temperature is rising, ie corresponding to bacteremia. Harris (1950) writes that the ease of isolation is: B melitensis > B suis > B abortus. B abortus is difficult to isolate in general and is hardly ever isolated in the chronic phase. Success may depend on the blood sample being taken when Brucella is circulating in the blood. The chances of isolating Brucella from patients not acutely ill is low. Not infrequently Brucella can be isolated from patients where other laboratory evidence is lacking.

The most widely used method is based on the work of Castaneda (1946) , (1949) and (1961) which involves incubating 5 ml blood in a specially desighen Ruiz-Castaneda flask at 35°C for up to 6 weeks. The Ruiz-Castaneda flask uses a 2 phase system. The flask utilises a liquid media (tryptose broth) and also contains a fixed agar plate inside the bottle. During the test the agar (tryptose) plate is re-moisturised every 2 days by tilting. Commercial systems are available including frm Bactec and others. The general method is describe for cattle elsewhere.

Brucella is difficult to grow and may grow slowly. This means it is far from ideal as a diagnosic tool and other serology tests must be relied on for first intervent in accute cases. Araj et al. in the study of 173 acute brucellosis (probably B melitensis) patients found 91 with positive culture tests. These became positive at the follwing time intervals.

• Week 1 - 13%.
• Week 2 - 65%.
• Week 3 - 19%.
• Week 4 - 3%.

The culture has been developed so that it can be used to differentiate between the different types: B abortus, B suis, B melitensis etc.

Culture Test in Chronic Brucellosis: It is generally recognised that the culture test only works in a small minority of cases of chronic brucellosis. It is reasonable to suppose that for the most part the Brucella is encapsulated within localised leasions. Attempts to recover Brucella by liver biopsy and fron bone marrow are also usually unsucessful. Yhis has been reported by Evans (1939) , Foshay (1940) and Poston (1938) . Spink (1951) notes that it is unusual for Brucella to be isolated from patients whose illness has extended beyond 1 year except isolates from localized complications and occasionally the organism may be recovered from bone marrow, urine, draining sinuses and biopsied lymph nodes. Williams (1973) notes that in chronic brucellosis liver biopsy is often unhelpful.

Some workier have made sustained attempts with repeated culture testing to achieve positive results. Poston (1938) carried out extensive tests on 14 patients suspected of having chronic brucellosis and was able to isolate Brucella in 5 (B suis, 1; B melitensis, 3; B abortus, 1). In all 5 cases the agglutination titres were <1:80 (2 patients at 1:40, 2 totally negative and 1 positive at the lowest dilution.) and in all cases the durationof illness exceeded 6 months at the time of test. 1 patient gave a positive interdermal Brucella test. The difficulty of isolating B abortus is described by Spink (1947) . He carried out expensive and repeated culture tests in 100 patients, including 35 possible acute cases, strongly suspected of having brucellosis and B abortus was eventually isolated in only 10 cases. Araj (1986) found that 5% of chronic brucellosis at about 1 year from start of infection were positive. It is likely that the longer ther period between the start of infection and the test procedure the less chance there is of it working.

Failure of the Culture Test for B abortus. The Brucella culture test does not work well for B abortus. Burnet (1928) compared goats infected with B abortus and B melitensis. He found that for goats with B melitensis 17 out of 26 blood cultures were positive whereas for the goats with B abortus only 1 in 26 blood cultures was positive. Hardy (1930) found that B suis infections in man derived from hogs usually produced positive culture results whereas B abortus infections in man derived from cattle did not. Spink (1956) considers that the culture test failed in B abortus becomes the blood seldom became bacteremic. Williams (1973) considered that isolation of B abortus by culture in acute illness is usually unsucessful.

Isolation of Brucella From Localised Leisions. Brucella can also be isolated from other secretions and tissue associated with local leisions. Gauss (1950) observes that Brucella can be isolated from milk, blood, bile, urine, feces, and from organs such as splene, tonsils, lymph nodes, bone marrow, dental roots, gall bladder, uterus, fetus, vertebra. Carpenter (1932) isolated B abortus from tonsils. Harris (1943) noted that Brucella has been cultured from sputum. Martin (1961) indicates that positive results can sometime be obtained from foci of infection such as arthritic joints and cysts. A negative result is not proof of absense of brucellosis. In chronic brucellosis this test hardly ever works. Tests on bone marrow and CSF can sometimes be more successful. According to Harris (1950) Brucella can be cultured from blood, bile, feces, or urine or from spinal, synovial, pericardial, ascitic, pheuritic, or other fluids or from biopsy or operative specimens, discharge, secretion, exudate or transudate from localised infection. He notes that the chances of isolating Brucella depends of the use of specialist methods of the chance of success is slight. Corbel (1997) consideres that Brucella is rarely isolated in specimens containing competing microflora. Schirger (1960) studied 224 brucellosis patients. Of these 42 were shown to have localised chronic brucellosis by culture as follows: bone/marrow, 7; joints/spine, 10; soft tissue, 5; propatellar bursa, 3; spleen, 5; liver, 1; genitourinary tract, 7; lung, 4. He notes that some were found incidentally during surgery and that there are likely to be more related to inaccessible leisions. Patients with localised brucellosis were less likely to have marked systemic reaction but not exclusively. Patients with localised brucellosis could be asymptomatic. Some leisons failed to grow Brucella. Tissue from which brucella can be cultured include: intervertebral disc (spondylitis), femur, humerous, skull, ribs, gall-bladder, liver, spleen, popliteal aneurysm, and kidneys. He considered that suppurative arthritis was uncommon. He notes that other serology tests are low or negative in patients with localised brucellosis and suggests that the antigens in localised leisions are well incapsulated keeping the away from the general circulation. In this survey most localised brucellosis was in B suis.

Culture Test with Negative Serology Tests. In some cases patients have positive culture test results with other serology tests such as the agglutination test or even symptom free. Huddleson (1940) studied 49 patients with subclinical brucellosis infections. They were able to isolate B melitensis from 3 patients who were symptom free. It is considered that although these patients were symptom free it could not be assumed that they would remain well. Harris (1950) notes that not infrequently Brucella can be isolated from patients where other laboratory evidence is lacking. Harris describes 5 such patients:

• B suis was isolated from the blood of a farmer with a farmer with chronic symptoms and a slight temperature rise to 100°F and a agglutination titre of 1:640.
• Brucella wa isolated from the uterinr discharge of a woman who had suffered 3 spontaneous abortions at 6 months. Her symptoms were fatigue, malaise, and low grade fever with a temperature upto 99.6°F. Her blood agglutination titre was 1:160.
• A patient was ill for 4 years and B melitensis was recovered during remission after a slight exacerbation of the illness.
• B abortus was recovered from the blood of a patient with subacute infection.
• B abortus was recovered from the blood of a patient with a chronic infection.

Poston (1938) studies 14 patients with chronic brucellosis and obtained positive cultures in 5 of these blood agglutinate titres were negative in 2, present at low dilution in 1 and 1:40 in 2.

Polymerise Chain Reaction (PCR).

General Considerations.

PCR is a new type of test which has only been applied to the detection of DNA from Brucella in recent years. During the process of an infection Brucella cells breaks up and releases LPS, proteins and DNA into the blood stream. PCR is very sensitive and can detect very small amounts of DNA. It can be considered as a direct method similar to the culture test since it detects DNA directly from Brucella whereas the tests for serum antibodies are indirect methods. The DNA is chosen is unique to Brucella. At this time (2011) the test is only being used of a limited scale mainly in experimental studies but it has the potential to replace the culture test and also antibody serology tests in the diagnosis of brucellosis in humans. It can be used to detect Brucella genus as described by Bogdanovich (2004) and species as described by Newby (2003) . The test has revealed important new information about the chronic nature of brucellosis in humans. It has been shown that Brucella DNA persists in the blood of patients long after they have apparently been cured and that it is present in the blood of chronic brucelosis many years after onset of illness and in the absense of other positive serology tests. This is directly contradicary to the suugestions that chronic brucellosis is a psychological illness such as psyconeurosis and CFS or delayed convalescence which is discussed elsewhere. Just like other tests at present PCR does not work in all cases. Individuals can have both positive and negative test results suggesting intermittent release of DNA.

The test procedure involves the polymerase chain reaction to amplify and accumulate a DNA sequence the quantity of which is usually measured using a flourescent dye which interacts with the DNA. The general techniques of quantitative PCR (Q-PCR) have been described by Espy (2006) . The test is usually performed on blood sera or whole blood but other body fluids may be used such as urine demonstrated by Queipo-Ortano (2006) , CSF as demonstrated by Colmenero (2005) . and pus from a lesion as demonstrated by Colmenero (2002) . There is no standard methodology for PCR in brucellosis but quantitative real-time PCR (Q-PCR) is currently the method of choice. Tenover (2007) notes that Q-PCR improves the sensitivity, specificity and speed of detection and decreased the risk of DNA contamination of the laboratory. The accumulated amplified DNA are measured as they develop by fluorescence resonance emission transfer probe within the closed system. The sequence of DNA chosen to be used in the test detects only Brucella and not other similar bacteria. Al Dahouk (2003) reviewed the choice of DNA sequence for the test which is important because of possible cross reaction with other bacteria such as Ochrobactnum a member of the α-2 subdivision of the class Proteobacteria which on the DNA level is closely relared to Brucella

Persistence of Brucella DNA in Treated Patients.

In this section we discuss some of the interesting results which indicate that Brucella persists in patients previously thought to cured of brucellosis and in chronic patients. This work indicates that PCR can detect Brucella DNA in patients months or years after treatment including apparently cures individuals. In the earlier work such results were viewed as anomolies but more recent work has shown this is a significant effect with profond implications. Morato, et al (1999) . reported 2 asymptomatic patients with positive PCR and therefore residual DNA 6 months after treatment. Queipo-Ortuna (2005) describe a patient with positive PCR assay 6 months after treatent. Maas et al (2007) studied the brucella DNA data for 13 patients during treatment and then in a follow up approximately 4 months after completion of treatment. Although supposedly cured these patients complained of persisting symptoms. They found that brucella DNA load continued at the end of treatment and during follow up in about 62% of patients. The same result applied to both relapsing and non-relapsing patients. Some patients were intermitently positive by DNA load.

Navarro (2005) used Q-PCR to study 18 patients with acute B melitensis and monator evolution of the disease. They proved that their technique was 100% specific for B melitensis. After apparently successful treatment about 50% of patients retained a low DNA positive profile. During follow up 40% of patients, most of them asymptomatic still retained DNA bacterial load. All were given treatment and apparently recovered but of these 7 relapsed. The results show that a DNA load persisted for up to 2 years in appaently cured patients. There was no difference in relapse and non-relapsed patients. DNA load between the 2 groups. In general DNA load decreased during treatment but then persisted at low levels. 4 patients who had negative Q-PCR after treatment relapsed showing that while DNA was not present in blood it must be present elsewhere in the body such as intercellular locations. The results indicate that Brucella survives antibiotic treatment suggesting a complex interaction between antibiotic suppression, bacteria survival and host immune response.

Pappas (2008) carried out a study of 39 patients with demonstatable acute brucellosis melitenis by Q-PCR. All patients were treated successfully including 3 who relapsed. The results showed that apparently cured brucellosis patients still retained B melitensis DNA. The results for presence of Brucella DNA were as follows: 100% in acute phase before treatment; 87% after treatment; 77% 6 months after treatment; 70% at >2 years after treatment. The 3 patients who relapsed showed similar DNA profiles to the patients who did not. It is concluded that the time interval from treatment to test of about 2 years eliminates the possibility that the DNA was derives from killed Brucella killed during treatment. They conclude that the results show that Brucella has a long term presence in the human body probably in a cellular recevoir. They note that these results which strongly imply that brucellosis is an intrinsically chronic illness has profond implications for pathophysiology, diagnosis and treatment. It might be construed that the only difference between cured and chronic brucellosis is the ability of the immune system to control the illness.

The use of PCR in Chronic Brucellosis.

Castano (2009) notes that Q-PCR finds more patients with chronic brucellosis than other tests. They note that this offers new hope to a not inconsiderable group of nonfocal-diseas patients undiagnosed by classical methods complaining of nonspecific symptoms (weakness, easy fatigability, anxiety, nervousness and vague aches). Castano voices his concern for such patients, he indicates theat they wander from physician to physician without getting a diagnosis.

Castano et al (2009) used Q-PCR to study patients with a history of acute brucellosis and then chronic brucellosis and found that Brucella DNA persisted in focal illness, non-focal symptomatic illness but also asymptomatic patients persisted. This strongly suggests that they retain a low level infection. He notes that in chronic brucellosis of long duration there are no objective laboratory tests. He used quantitative real-time PCR (Q-PCR) to study 35 such patients who had well documented histories of illness with durations of 2 to 33 years. 8 patients with focal and 9 patient with non focal complaints such as fatigue, malaise, myalgia and/or arthralgia all produced positive Q-PCR results. 6 of 18 assymptomatic patients also produced positive Q-PCR results. He concluded that Q-PCR could be used to identify chronic brucellosis. 11 patients (5 focal and 6 non-focal) had treatment and 6 continued to have Brucella DNA. THere were 18 asymptomatic patients. Each patieent had 2 separate tests which were notr all positive. They concluded that Q-PCR could be used to identify chronic brucellosis. The detection of Brucella DNA in non-focal symptomatic patients suggests that Brucella is present under partial control of the immune system. They note that when heat killed Borrelia burgdorfei was injected into dogs the DNA persisted for only 3 weeks. Therefore it can be concluded that Brucella DNA in chronically ill patients must be from the degradation of live bacteria. In their results thay found that Brucella DNA was intermittently present in the blood or serum of patients and that therefore negative results do not exclude infection.

Colmenero (2002) used PCR to demonstrate the presence of Brucella DNA in pus and from a calcified granuloma in 6 cases of heptosplenic suppurative brucellosis while Brucella cultures tests on the same materials were negative.

General Conclusions.

In general these results seem to strongly indicate that PCR detects the presence Brucella in the blood of chronic assytomatic brucellosis patients and this compares with other antibody serology tests which are unreliable under the same circumstanses. As the tests have been refined the level of detection has increase. THe load of DNA persist long after treatment has ended and it may be present in symptomatic and assymptomatic patients. However Brucella DNA seems to vary in concentration perhaps according to spasmodic rease of DNA. It can be construed that the difference between chronic and assymptomatic brucellosis is determined by the ability nof the immune system to control and contain infection.

PCR used to Identify Brucella Variants.

Imaoka (2007) describe a a combinatorial polymerase chain reaction (PCR) procedure to identify four major species of the genus Brucella simultaneously involving four pairs of primers targeting the genes encoding a cell surface protein (BCSP31) and outer membrane proteins (omp2b, omp2a and omp31). This could differentiate between B abortus, B melitensis, B suis and B canis. B. abortus was identified when fragments of BCSP31 and omp2b/2a were amplified by B. abortus-specific primers and B melitensis by the amplification of fragments of BCSP31, omp2b/2a and omp31 using pair of primers B4/B5, JRF/JPR-ab and omp31. B. canis was identified by the detection of the amplicons of omp2b/2a were detected by B. canis-specific primers also BCSP31 and omp31. If specific amplifications occurred using all pairs of primers, the strain was identified as B. suis.

Serology Antibody Tests in Humans.

Apart from the culture test a range of serology tests have beeb used for the diagnosis of brucellosis in humans over a long period of time. They test for serum antibodies to Brucella. Such tests work of B abortus, B melitensis and B suis but not B canis. Principally they test for the O-side chain of LPS. Diaz (1968) and (1972) showed that surface antigens of smooth Brucella occurred as lipopolysaccharide-protein complexes containing O-polygosaccharide side chain, core oligosaccharide and lipid-A. Lopid-A is anchored in the intermediate core region. OPS which has been defined as a homopolymer of 4,6-dideoxy-4-formamide-alpha-D-mananose linked via glocosidic bonds and is immunodominant and that LPS is the antigen component that is most important in SAT, RB, and AHG tests. Moreno (1997) noted that the O-side chain which induces the major immune response in only present in smooth variants of Brucella. Rough mutants of smooth variants are usually less virulant and rough variants such as B canis and B ovis have a limited host range and ability to infect other species, Corbel (1989) . None of these tests work in all cases. The tests used most often are:

• The Serun agglutination test (SAT).
• The complement fixation test (CFT).
• The Antihuman Globulin Test (AHG) also known as the Coombs test.
• Mercaptoethanol or Dithiothreitol test (2-ME).
• The Rose Bengal Test (RB).
• The Enzyme Linked Immunosorbtion Assay (ELISA) for IgM, IgG and IgA.

Historical Perspective. It is important when considering antibodiy test for brucellosis to consider the time frame in which the tests have been used. Many references discussed in the review refer to a period when SAT was effectively the only available Brucella antibody test. The dates at which the tests were introduced was as follows:

• The SAT titre has been used from 1897 to the present day. It detects mainly IgM.
• The CFT titre was developed in the early part of the 20th century but was not used widely until after 1950. This test measures IgG.
• The AHG titre was developed in 1952 and used thereafter. This test measures mainly IgG.
• THe 2-ME or DDT titre was developed for human use in 1965. This test measures IgG.
• The iELISA was first used in humans in about 1980. It can be used to detect IgM, IgA and IgG separately of together.

The way in which these tests worked was not understood until the 1960s onwards. In general irrespective of whether the infection is cured or remains chronic IgM peaks early in the infection and then declines to low values at about 1 year. IgG appears and peaks slightly later and then declines more slowly to low or negative values. The decline of IgM and IgG may or may not related to cure. In chronic brucellosis all these tests may be negative. At the same time tests results may be unreliable for a number of other reasons which are discussed below. All these factors have made the diagnosis of brucellosis very difficult. Some of the analysis and comments of doctors struggling to understand inconsistent tests can re-interpreted Until about 1950 doctors trying to diagnose brucellosis or chronic brucellosis used a combination of tests: the culture test, SAT, the skin test and the opsonocytophagic test. All these tests are discussed in this review. Some idea of the difficulty that medical workers encountered if given by Harris (1950) who notes that the diagnosis of brucellosis, particularly in its chronic form, may be among the most difficult tasks in medicine. A clinical diagnosis can rarely be made with any degree of assurance. At best there is usually only suspicion prior to laboratory tests. The are no pathnomonic signs in either acute or chronic illness. Brucellosis can simulate other diseases so exactly that confusion is inevitable. Physical findings are usually surprisingly few. Therefore diagnosis depends on very careful study of the history of the past and present illness, symptoms and signs and multiple laboratory precedures, history of possible exposure, exclusion of other illnesses, and the possible benefits of treatment. However because of the wide exposure to Brucella a negative history of obvious exposure may not be significant. Even with the use of multiple tests the diagnisis may remain unsatisfactory. He notes that the most common mistake in the diagnosis of brucellosis is assuming a negative agglutination test at low titre rule out the illness. He considers that the following points are generally accepted:

• A positive culture test is controvertible proof of Brucella infection.
• A SAT of ≥1:80 is presumptive evidence of brucellosis.
• A skin positive skin test is is evidence of past or present infection.
• The opsonocytophagic reaction adds significancew to other laboarory tests and clinical evaluation.
• Negative test results do not eliminate a diagnosis of brucellosis.

Similar views are expressed by Simpson (1940) . He notes that although tha agglutination and skin tests all of the so called specific diagnostic tests may yield repeatedly negative results in culturely proven cases of brucellosis and that none distinguish between present and past infections. In the absence of a positive culture test the findings of positive agglutination, skin and phagocytosis must be carefully appraised in consideration of other factors. Also a diagnosis of brucellosis can be missed by to much reliance on one test or an unjustifiable diagnosis for brucellosis may be made on the basis of one test.

Far more work has bee carried out with SAT than other tests but it has also been found that the newer tests such as CFT, AHG, or IgG ELISA do not work in all cases. They test mainly for IgG which persists longer than IgM in the medium term but IgG also declines to low or negative levels inn longer term or chronic brucellosis in most cases. iELISA IgM gives similar results to SAT.

Seroconversion: Anderson (1964) showed that in a calf injected with Brucella IgM specific to Brucella appeared after 3 days and IgG (SAT) after 6 days (CFT). They showed that 2-ME and CFT gave similar results. According to Buchanan seroconversion occurs during the first week as tested by SAT. Reddin (1965) studied brucellosis by injecting 10 jail prisoner volanteers with an attenuated Brucella vaccine. He found that an early rapid responce by IgM followed by IgG.

A General View of the Tests. The SAT, RB and ELISA IgM work best in the early stages of an acute infection since they test for IgM which is the initial antibody produced from 3 about 3 days. The SAT method has been and still is the most widely used test. The procedure is relatively simple but takes at least 2 days. In acute cases the test work well but the response of individual patients is highly variable and in some cases despite obvious illness these tests are negative. The RB test at neutral pHs gives similar results to SAT and is a result can be obtained quickly but it is not so widely used. From about 7 days most patients also develop IgG and IgA antibodies. This can be tested using the CFT, AHG, 2-Me and ELISA IgG tests which measure IgG. The 2-ME and CFT tests are useful for monitoring progress of the illness in the first year but their usefulness falls off rapidly thereafter. The 2-ME test is relatively insensitive and is of relatively little use in chronic brucellosis. The CFT test is not recommended for the diagnosis of chronic brucellosis since it is less sensitive than AHG and ELISA IgG. AHG and ELISA IgG and ELISA IgA are the best test for chronic brucellosis but are also negative in many cases. This seems to be because of a decline of Brucella serum antibodies over time unrelated to illness.

The serun agglutination test (SAT).

The agglutination test was developed by Wright (1897) and . A general version of the tests is as follows: Sera at various dilutions and buffered to pH 7.2 is mixed with a Brucella antigen (usually heat or chemically killed Brucella) and the mixture incubated for 24 hours at 37°C during which time antibodies adhere to the antigen surface and cause agglutination. The opacity of the suspension is tested. The highest dilution which shows discernable aggutination (fine agglutinates with incomplete clearing of solution) is read. In non-endemimic countries 2 consecutive positive results at dilution 1:160 or a rising titre are considered indicative of illness. In endemic low to medium high counts but no symptoms are encounterd in apparently health people which makes the interpretation of the tests extremely difficult. In these circumstances a result at a dilution of 1:320 is considered positive. Generally speaking this is a cheap acurate test easy to perform and is still widely used particylatrly in poorer countries.

Factors Affecting the Usefulness of SAT.

• The test can be compromised by cross reactions caused by other bacteria with similar surface OPS. This problem applies to all agglutination tests.
• The effectiveness of this test is inhibited by blocking proteins or so called prozone effect.
• In counties where brucellosis is endemic a significant part of the population may have a positive SAT titre without obvious signs of illness. This makes the test very difficult to interpret.
• The SAT titre can be negative during acute brucellosis sometimes with a positive culture test.
• The SAT titre is negative in a large proportion of chronic Brucella infections even with a positive culture test.
• The test is slow taking 2 days to get a result.
• The usefulness of the test declines rapidly towards 1 years in most cases.
• In early work there were problems with non-standardised Brucella antigens so results could vary between laboratories. Standard antigens from a few accredited centres are now used.

Interpretation of SAT Titres.

Test results are measured in dilutions eg' 1:20. 1:40, 1:80. 1:160 and 1:320.. Over a long period of time there has been a debate as to the lowest titre which can be taken as positive proof of infeection. In countries where brucellosis is endemic (such as the USA in the 1930-60s) where a significant part of the population is or has been exposed to Brucella many people are recored as having positive SAT titres without any sign of illness. This means that lower titres are almost impossible to interpret. In these circumstances it has become common for doctors to try to interpret the results by imposing a lower limit for a positive result. The obvious dangers are that some patients with low titres will be missed. A lower limit of 1:≥160 or 1:≥320 is often used. Others consider that in countries with a low incidence of brucellosis any positive SAT titre should be considered indicative of illness even as low as 1:≥10. Seroconversion or a risihg tire are also useful indicators of infection. The SAT titre should always be used in association with another test for IgG such as; CFT, 2-ME, AHG or ELISA IgG. The following are the experiences and views of experts:

• Hardy (1930) classified SAT titres as follows: <1:40, slight clinical significance; 1:40, doubtful significance; 1:80, weakly positive; 1:160 and 1:320, positive; ≥1:640, strongly positive. But he notes that agglutinins may persist more than a year in apparently cured brucellosis while severe infections may occur with no noticablr agglutinin production.
• Evans (1938) in a review of the SAT test concluded that a titre at 1:40 was a reliable indicator of chronic brucellosis but this did not mean that a negative result excluded the disease. In one set of patients 4 of 7 cases with positive culture test gave a negative SAT.
• Simpson (1941) reviewed the use of the SAT titre and considered it to be generally a reliable test in acute brucellosis. The test becomes positive about 2 weeks from the start of the illness. A problem with the test is that results may be low of negative in about 6% of cases even with positive culture tests. Another problem is that the titre may remain at high levels years after apparent cure and individuals exposed to Brucella may have a SAT tire without signs of illness. Also results on the same patient may fluctuate widely. For this reason considerable care must be exercise in analysing the results. He considered that a SAT tire at 1:80 was arbitary and titres as low as 1:10 should be followed up. In chronic brucellosis many patients have negative or very low SAT tires.
• Griggs (1943) reviewed the use of SAT in 96 patients. 86 were negative and 10 positive. Of the 10 positive: 4 had a titre of 1:80 or higher; 3 had a titre of 1:40, 1 at 1:20 and 2 positive with unspecified titres. He considered SAT titres were less useful than ths skin or opsonocytophagic tests.
• Kolmer describes problems concerning the unreliability and lack of reproducability between SAT test results from different laboratories.
• Braude (1950) considered that a SAT titre of at least 1:80 was essential before brucellosis could be considered.
• Memish (2002) notes that in a survey of 23,000 people in Saudi Arabia 4.3% had a SAT result of 1:≥320. This means that for test purposes 1:≥320 must be taken as the positive titre level. Many people have a titre of 1:160 which may represent a previous infection or a subclinical illness. He notes that in brucellosis free countries titres as low as 1:≥40 can be considered as positive.
• Al Dahouk (2003) comments that for SAT there is no defined threshold cut off for acute brucellosis, in a non-endemic 1:80 is often considered suspicious and 1:160 as positive. In an endemic area 1:320 but previous contact with Brucella with subclinical infection makes diagnosis very difficult.
• Corbel (2006) notes that the response to SAT is individual so that some have high tires and others low titres. He noted that in a study of 238 brucellosis patients 29.2% had a titre of ≤1:80 and 3.4% had a tire of ≤1:10.
• Howe (1947) reviewed the cases of 17 laboratory workers with brucellosis and concluded that the agglutination titre was not proportional to the severity of illness and did not show consistency with remission or exacerbation.

The antihuman globulin test (AHG) or Coombs test.

THe AHG test was developed to overcome the prozone effect which affects SAT and other agglutination tests. It gives positive results when SAT is negative. This is because it measures IgG. For this reason it is considered a better test in long term or chronic brucellosis. However it does not work in all cases of chronic brucellosis particularly longer term.

The AHG or Coombs test was developed by Coombs (1945) to detect "incomplete" Rh agglutinins. It was applied to Brucella by Wilson (1952) in an attempt to disclose non-agglutinating antibodies. A typical method is described by Kerr et al (1968) . 2 sets of sera dilutions is prepared for each of the SAT and AHG titres. The antigen used is usually killed B abortus which is used at a higher concentration for the AHG titre. For the AHG titre antigen is added to the dilutions in tubes which are held at 37°C for 24 hours and then examined. Tubes with agglutination present are recorded and discarded. The other tubes are centrifuged cold at 2000 rpm for 15 minutes the precipitate retained and the supernatant discarded. The precipitate is re-suspended in buffer with thoroughly wahing and re-centrifuged 3 times to remove free human protein. The precipitate thus produced is then re-suspended in saline buffer and anti-human globulin added. The tubes are incubated for 24 hours at 37°C and then examined for agglutination as for the SAT titre.

AHG determines IgG and IgA so it does not work in the earliest stages of acute illness when only IgM is present. It is considered together with iELISA IgG and iELISA IgA the best test for chronic brucellosis although it fails to detect many cases. The main problem with this test is that it is slow and labour intensive. It usually takes at least 3 days to complete the test whereas ELISA IgG and IgA can be done in a few hours.

Kerr (1966) (1966) has shown that AHG test detects cases of chronic brucellosis missed by agglutination tests. They also found that AHG detected IgA and IgG but not IgM. Diaz (1976) showed that only IgG and IgA are active in this test. She did this by a process of elimination using anti-IgM, anti-IgG and anti-IgA.

AHG has also been found useful for detecting Brucella antibodies in cerebrosplinal fluid (CSF). Sanchez-Sousa (1990) compared the performance of SAT, RB and AHG tests on SCF and corresponding sera. They found that AHG worked in all cases but that SAT and RB only worked in only some cases. This probably relates to the concentration of IgM to IgG over time. There was a good corelation between AHG in CSF and sera but levels in CSF were lower. Therefore AHG is the appropiate test of choice. The alternative is ELISA IgG.

Ariza (2006) noted that AHG could demonstrate relapse in chronic hepatosplenic brucellosis whereas RB, SAT and Brucellacapt could not.

MacDonald et al (1967) describes a patient treated with tetracycline whose SAT, 2-ME and CFT steadily declines over 10 months but AHG remain high and who later had a relapse with joint leisons. This demonstrates the superiority of AHG over CFT.

Barrett (1953) Demonstrated that AHG cannot find all cases of chronic brucellosis. He compared the usefulness of SAT and AHG tests in chronic brucellosis. They tested 150 sera samples. If SAT was positive then AGH was also present usually in a concentration 10 times higher, but if SAT was negative, as in chronic brucellosis, then AHG was also ≤1:80. They considered that AHG was of little or no use in many cases of chronic brucellosis.

Mercaptoethanol (2-ME) or Dithiothreitol Tests (DTT).

These tests are variations of the SAT method. The 2-Mercaptoethanol (2-ME) test was first used for cattle by Anderson (1964) in an attempt to deferenciate between active and inactive brucellosis infection in cattle in the presence of low level agglutinates. It was adapted for humans by Reddin (1965) The normal SAT methodology is used except that 2-ME or DTT is added to the serum. The 2-ME reduces the disulphide bonds in IgM and therefore prevents it from agglutination. Therefore by default the test should measure IgG and IgA, however there resonably evidence to suggest that IgA is also affected so that it only measures IgG. Klein (1981) showed that Dithiothreitol (DTT) which is less tOxic and obnoxious than 2-ME could be used instead with no chamge in accuracy. 2-ME or DTT tests are used in combination with SAT to diferentiate between IgM and IgG titres. This gives a guide to the stage of the illness as described by Poole (1975) in a survey of brucellosis in north-western England.

Active and Inactive Brucellosis Infection. It has been suggested that this test can distinguish between active and inactive brucellosis. After the acute phase of illness 2-ME or DTT titres peak and then fall and it is suggested that when titres become negative then the illness is inactive. Reddin (1965) and Kerr et al (1968) note that this test measures 7S macroglobulin (IgG). Reddin suggests that the presence of 7S is an indication of active brucellosis. Buchanan (1980) studied 92 patients over a 18 month period and considered that this test could be used differentiate between active and inactive. The same methodology was used by Young to dscide between active and inactive brucellosis for 214 patients. As long as a patient had a positive 2-ME they were considered to have active brucellosis but if 2-ME became negative the illness was inactive. This concept is flawed. Problems are that results are: it gives erratic results; the test is insensitive; and test results conflict with other test methods such as AHG. Hewith (1984) point out that the 2-ME reaction is difficult to control and results vary between laboratories. Williams (1973) considered that in chronic brucellosis SAT and 2-ME tests are less sensitive test that AHG of CFT and are often negative. He did not consider that presence od IgG indicates active infection. Typical data supplied by Hemderson (1976) shows a poor results for the 2-ME test compared to AHG. Pellicer. (1988) showed that in a study of 10 relapsed patients that DTT performed no better than SAT and was a poor indicator of relapse. Smits (1999) compared four brucellosis antibody tests including SAT, DTT and AHG during the first 6 months of illness using from 247 confirmed cases of brucellosis. THe sensitivity results for SAT, SAT-DTT and AHG were as follows:

Duration of Illness	SAT	DTT	AHG
Months	%	%	%
0 - 2	79.5	67.1	39.7
2 - 4	64.4	19.1	88.1
4 - 6	76.9	13.5	92.3
>6	50.6	6.0	81.8

These results clearly show that DDT tails off rapidly and offer no specific information on the extent of infection.

The complement fixation test (CFT).

The CF test was developed by Larson in 1912 to look for evidence od Brucella infection in cattle. He also used the test to show that children (1913) and (1915) who drank raw milk developed a positive CFT titre. In early trials Carpenter (1930) , Weise (1943) and Menefee (1939) concluded that the CF test had no advcantage over agglutination tests. Michell (1930) considered that CFT might be useful in conjunction with the agglutination test particularly where agglutination tests were negative. Sasano (1931) found that clinical cases gave positive results for both SAT and CFT. The test was used too study brucellosiis in humans by Leon (1948) and Burmakin (1950) . The CFT is used to look for antibodies to Brucella mainly IgG in blood sera . The test utilizes sheep red blood cells (SRBC), anti-SRBC antibody and complement, along with specific antigen. If antibody is present in the patient's serum, then the complement is completely utilized and SRBC lysis is minimal. However, if the amount of compliment exceeds tha amount of antibody, ie. no antibodies or higher dilutions of sera, then the complement binds anti-SRBC antibody and lysis of the SRBCs ensues. SRBC are added to serum, followed by anti-SRBC and then complement. The result is taken as the highest dilution with 50% lysed blood. THe full method is described by Kerr et al (1968) . Spink and Young (1991) reviewed its use in humans and considered it no better than the agglutination test. According to Dohoo (1986) working with cattle CFT is in general a sensitive test. Corbel notes that the CF test is usaully completely negative in those not exposed to Brucella but that those with environmental exposure may have low sustained positive results. He also notes that because the test is difficult to standardise it is not suitable for laboratories with a small throughput. In part this is due to the use of labile reagents. The test has been widely used in the diagnosis of cattle and humans however it has known limitations. This test is best used in conjunction with SAT in acute brucellosis during the first year of infection. IT is of limited use in chronic brucellosis.

Problems Associated with CFT. There are a nimber of known problems associated with the CF test:

• CFT is affected by anti-compliment activity. If this is present the test cannot be read at lower dilutions. Nielsen (1996) working with cattle considered that sensitivity and specifity varied considerablely due to anti-complement activity. Lucero (1990) notes that CFT mostly identifies IgG but has severe disadvantages such as the the occurrence of anti-complement activity and general technical problems.
• CFT is effected by the prozone effect. Al Dahouk (2003) notes that CFT is affected by the prozone effect and anti-complement activity.
• Labile reagents. The use of labile reagents may lead to erratic results particularly if the tests are carried out on a small scale. Lucero (1990) noted that a severe disadvantage of CFT was the use of the use of highly labile reagents (such as complement). Spink (1951) considered CFT as technically difficult and not yeilding any more information than SAT indoubtfull cases.
• Insensitivity. CFT fixes IgG and moderate amounts of IgM but not IgA. Dahouk (2003) considered that CFT should be used only as a confirmatory test and she noted that sensitivity and specificity could vary widely. *Chronic Brucellosis: CFT is of only limited use in chronic brucellosis. AhG and ELISA igG and IgA are better. Although it is better than SAT. Kerr (1966) has shown that CFT detects cases of chronic brucellosis missed by agglutination tests.

Because of the technical problems with the test at low dilutions a lower limits of usefulness has been applied. This is usually either 1:≥8 or 1:≥10. Below this limit the test is considered difficult to interpret. Therefore in cases where a patient might have low results such as chronic brucellosis this is a considerable disadvantage associated with the test. Some worker have used lowest dilutions at 1:≥4 or 1:≥5. Pellerin (1980) showed a good corelation between CFT at 1:≥4 and ELISA IgG was possible. Marmoner (1981) also used a lower limit for CFT results at 1:≥4. Similarly Lucero (1999) considered that 1:≥5 was seropositive which means the detection of more brucellosis seropositive patients. In practice this demonstrates the limitations of the test.

Serra (2004) considered that CFT detects both IgM and IgG but should not be used in routine clinical analysis because SAT, AHG, and ELISA are more reliable.

Comparison of CFT with other Tests. CFT is less sensitive than AHG or ELISA IgG and IgA. It may be affected by anti-compliment and prozone effects and erretic results due to the use of labile reagents. It can be semi-automated to screen blood but in general is only suitable as a secondary confirmatory test. It is not suitable for chronic brucellosis. Foley et al (1968) in a survey of dairy farmers with possible chronic brucellosis abortus noted that AGH detected far mor positives than CFT. Similar results are shown by Gilbert et al (1981) . Similarly McDevitt (1969) in a review of the health of 224 veterinary surgeons found that AHG found more positives than CFT ie. 36 out of 224 had positive AHG and negative CFT. Also whereas AHG positives corelated with risk factors, CFT did not except in the first year. Poole (1975) reviewed 299 patients with signs of brucellosis. She used SAT, CFT, AHG in test work she found that wereas results for SAT and AHG could remain for several years after CFT had dropped to low levels. White (1978) shows that values for CFT are often low or negative (typically ≤4) while with the same samples positive results were obtained for IgM, IgA and IgG from RIA. Magee JT (1980) demonstated that the sensitivity of CFT was poor compared to ELISA IgG in chronic brucellosis. Heizmann (1985) notes a further problem with the test for long term illness. In a comparative study of sereology tests including 2 separate CF tests he found that CFT values from 2 different tests dropped below measureable levels within 91 to 158 days and declined more rapidly than ELISA. Similar results were found by Hewith (1984) . Dahouk (2003) noted that the usefullness of CFT declines more quickly over time than AHG. This may be because CFT fixes IgG and moderate amounts of IgM but not IgA. Gall (2004) reviewed the performance of a range of serology tests in bovine brucellosis. They noted that the performance index (PI) of CFT was lower than that of CELISA and Indirect ELISA. The Percentage coefficient of variation %CV for RB, CFT and IELSA were respectively: 14.8%, 14.1% and 9.5%. They concluded that this was due to a combination of effects: inconsistent results, complexity and labour intensity, lack of standardisation, inability to test haemolysed samples, anti-complementary results, prozoning, cross-reactions to other bacteria, They considered the overall performance of the test was a problem.

Dahouk (2003) considered that CFT should be used only as a confirmatory test and she noted that sensitivity and specificity could vary widely. She further considered that the usefullness of CFT declines more quickly than AHG.

The Rose Bengal test.

This is an agglutination test. It has not been used as widely as SAT but gives similar results. It was developed to screen cattle in about 1966. Similarly in humans it is reasonably accurate but my most is only considered suitable for screening. It is used in a plate or card form and is quick and easy to perform. A serum sample is mixed with Brucella antigen coloured with Rose Bengal Dye and gently mixed for 2 minutes. A positive test is shown by the presence of agglutinates.

Serra (2004) considered that Rose Bengall is only useful for screening in endemic areas because of poor specificity. It should not be used as a diagnostic tool especially in individuals with occupational exposure.

According to Allan (1976) in a study of bovine brucellosis due to B abortus at neutral pHs the Rose Bengal Test measures mainly IgM rather than Ig1 and Ig2.

The Indirect Enzyme Linked Immunosorbtion Assay (iELISA).

THe iELISA test has become the test of choice in developed countries. In its usual form it tests for OPS similar to agglutination test. It has a number of advantages.

• It is convenient and easy to use.
• It is available in the form of propietry kits although these can be expensive. However it is not difficult to prepare such plates in house at a relatively low cost. These can be stored for up to 1 year.
• The test does not involve labile reagents such as CFT.
• The test can be carried out in about 6 hours. This is much quicker than the SAT, CFT and AHG tests.
• The test is versatile. The ELISA IgM gives similar results to the SAT titre and the ELISA IgG similar results to AHG and CFT titres.
• iELISA can be used to test IgM, IgA and IgG separatly and also IgG subsets: IgG1, IgG2' IgG3' and IgG4.
• iELISA is probable the best test for chronic brucellosis although it does not work in all cases.

Just like SAT the iELISA IgM does not work in many cases of chronic brucellosis. iELISA IgG and IgA can be used in chronic brucellosis, but just like AHG, ELISA IgG fails in many cases. iELISA suffers from similar problems to those associated with agglutination tests including the prozone effect and cross reaction with other bacteria with similar OPS.

The ELISA test was first used for brucellosis by Carlsson (1976) in rabbit sera. Sauders (1977) used the test to investigate B abortus in cattle. The method was adapted to test for Brucella antibodies in humans by Magee JT (1980) Pellerin (1980) , Marmonier (1981) and Sippel (1982) . These early studies showed that in general ELISA proformed well and compared favourably to other tests.

A typical method of performing the iELISA test is as follows: A Brucella antigen in a buffered solution is coated onto a polystyrene microtitration plate and then the plate is washed with buffered detergent. Various antigens have been used including killed B Abortus or B mellitensis, lipopolysacharride from B abortus and B mellitesnis and Brucella surface membrane protein. Then serum samples are added at appropriate dilutions and the plate incubated for 2 hours at 37°C for 2 hours and then the plate washed. This allows antibodies to attach to the antigen. An enzyme-conjugated antihuman globulin is then added and the plate incubated for a further 2 hours at 37°C and washed. THe antihuman globulin attaches to antibodies fixed to the antigen on the plate. The antihuman globulin may be a mixture of anti-IgM + anti-IgG + anti-IgA or anti-IgM, anti-IgG and anti-IgA separately. Suitable enzymes are horse radish peroxidase or alkaline phosphatase. Finally a dye substate is then added and the plate incubated at 37°C for 45 minutes. A typical colouring agent is p-nitrophenyl phosphate. If antibodies were present in the serum then colour develops which is the assessed by optical density. If no antibodies are present in the serum there is no colour.

Comparison of iELISA to Other Tests: Gilbert et al (1981) investigated thr presence of Brucella antibodies in abattoir workers in New Zealand and found that iELISA was superior in detecting such antibodies to SAT, CFT and as good as the AHG tests. De Klerr (1985) compared ELISA (IgM +IgG + IgA) and IgM using B abortus antigen to other conventional tests and found it more sensitive than CFT. In a test of 100 occupationally exsposed workers the number of positive results were: ELISA (IgM +IgG + IgA), 20; SAT, 12; CFT, 12; ELISA IgM, 5. They also found that Rheumatoid factor(RF) caused some false positives in iELISA IgM. Memish (2001) has shown thart in the early stage acute brucellosis patients with positive culture tests that ELISA IgM works no better than SAT.

THe iELISA Test and Chronic Brucellosis: Araj (1986) studies acute or chronic ill brucellosis patients using ELISA. The 22 chronic brucellosis patients who are described as having a history of contact with Brucella of one year or more. The Positive test results for these patients were as follows: ELISA IgG, 100%. SAT, 50% (1:80) and 45% (1:160) and ELISA IgM at 35% (1:≥400). Although the results show all chronic brucellosis patiens having positive results by iELISA this probably relates to the short history of illness. SAT is not usually positive in chronic brucellosis. It is in longer term illness that these tests fail. On the basis of the same study group Araj (1988) showed that acute brucellosis sufferers had IgG, IgM, and IgA antibodies whereas chronic brucellosis sufferers had mainly IgG and IgA. A study of the subsets of IgG showed that IgG1 predominated in acute brucellosis and IgG4 in chronic brucellosis. These results are at odds with those of Kichieva (1983) . They studied IgG subclasses in 52 patients with brucellosis of 1 to 10 years and found that the concentration of IgG1, IgG2 and IgG4 were normal but IgG3 was elevated. In 10 patients given vaccine levels of IgG and IgG1 increases and IgG3 decreased. Possibly the difference in results to those of Araj relate to the duration of chronic illness.

There a widespread idea that iELISA can find all cases of chronic brucellosis which possibly derived from a misunderstanding of the work of Araj discussed above. This is not true. In practice its success rate is similar to that of the AHG test. Other workers have found that in chronic brucellosis iELISA faces similar problems to other tests. A number of workers have studied iELISA in long term chronic brucellosis sufferers and concluded that it is only partially successful. Colak et al (1992) studied 90 patients with chronic brucellosis using ELISA IgG and AHG test. All were negative by SAT. 44 patients had positive AHG results of whom 37 had positive ELISA IgG. 46 patients had negative AHG of whom 16 had positive ELISA IgG. AHG worked in 44 cases and ELISA IgG worked in 53 caes. Both AHG and ELISA IgG failed in 30 cases. Lisik D (2001) looked at 27 patients who had been diagnosed with acute brucellosis by serology and had developed chronic illness. ELISA IgG was able to diagnose 7 as positive. The author concludes tha in chronic brucellosis the humeral immunological response fades away over time and this is the reason test such as ELISA fail. Serra (2004) reviewed brucellosis serology test methods on patients in a rural area of Spain and concluded that for acute patients ELISA IgM was no better than SAT and for occupationaally exposed patients with possible new infection ELISA igG weqs no better than AHG.

Colak (1992) did a trial with 90 chronic Br patients and found AHG and ELISA IgG worked in 48.9% and 58.9% of cases respectively. Lisik (2001) noted that humeral IgG falls over time in chronic Br patients and in a test of 27 such cases found ELISA IgG worked in only 25.9%.

Elisa and Tests on CSF: Araj (1986) showed that analysis by ELISA could detect IgG, IgA and IgM in CSF at similar levels to those in their blood. The 10 patients had acute brucellosis and signs of CNS involvement. This is interesting brecause neurobrucellosis is also a problem in chronic brucellosis.

iELISA with Brucella Protein Antigen: iELISA can be used with Brucella antigens other than OPS. Brucella protein antigens are of particular interest. Araj (1989) showed that there was a good corelation between iELISA bases on whole killed B mellitensis antigen and MOMP antigen. MOMP antigen allows the detection of B carnis and B ovis which ELISA based on whole Brucella and LPS antigens do not. Fossati (1994) (1996) developed and reviewed the use of cytoplasmic protein antigen in ELISA antibody tests. Goldbalm (1991) and (1993) separated 3 fractions of protein from B abortus by centrifugation and purified a the (CYT) fraction by electrophoretic techniques. They used the (CYT) antigen to compare acute and chronic patients and also asymptomatic workers exposed to Brucella all of whom had positive results for LPS antibodies. Results for (CYT) antigen were high for acute and chronic patients and low in asymptomatic patients. They considered this a marker for active and inactive illness. Baldi (1996) compared ELISA (LPS) and ELISA (CYT) and concluded that the latter has the avantage of being free from false positives by cross-reactivity with other gram-negative bacteria. Balti (2001) notes that ElISA (CYT) is limited in usefulness in acute brucellosis compared with ELISA (LPS).

Pellerin (1980) in an in depth study of Br serology methods found ELISA produced 4.9% more positives than other tests. Saz (1987) considered that ELISA is the best serology test and detects 4% more positives than SAT, AHG, and Rose Bengal. Lucero (1999) and Gal (2004) did large scale trials of ELISA and CFT and found that ELISA was more reliable and more accurate.

Magee (1980) did a study comparing ELISA. IgG, CFT, and AHG and considered ELISA the most sensitive.

Araj (1996) compared the ELISA and AHG tests for 133 patients. Although some are said to be symptom free it is presumed that they were all possible chronic brucellosis sufferers. All were negative by AHG but ELISA found positive results for 28 patients (28.6%) mainly IgG and IgA.

Araj (1996) looked at the Brucella antibody profile of 597 people working in high risk occupations of whom 107 had possible symptoms of brucellosis. The overall serology results were: SAT, 1,7%; AHG, 15% and ELISA, 26-61% positive. Those with possible symptoms had higher values.

Ariza J et al (1992) reviewed the cases of 81 patients and showed that the persistence of high IgG levels of IgG beyond 1 year are associated with high initial serology readings. 12 months after treatment the positive results recorded for SAT, AHG, ELISA IgM, ELISA IgG and ELISA IgA were as follows: SAT, 26%; ELISA IgM, 25%; AHG, 71%; ELISA IgG, 89%, Elisa IgA, 69%. Their results show a good corelation between AHG and both ELISA IgG and IgA.

Other Less Used Sntibody Serology Tests.

THe tests used here have only had limited use in the diagnosis of human brucellosis.

Immunocapture-Aglutination Test (Brucellacapt).

The immunocapture-agglutination test a single stage proprietory test kit available from Vircell and is a new form of the agglutination test to test for brucellosis antibodies in sera. The method is as follows: Samples of diluted sera are added to wells on a microtiter plate coated with antihuman immunogolulin, then stained Brucella antigen is added followed by incubation at 37°C for 24 hours. A positive test result is shown by activation of a coating of dye on the surface of the plate. The propietor suggests that the test can detect both agglutinating and "incomplete antibodies" and therefore corelates well with the AHG test. Independant assessment suggests that it only works well for the early stages of bucellosis infection but not for chronic or relapsed brucellosis.

The usefulness of Brucellacapt has been assessed by Orduna (2000) who concluded that it could be used to give results similar to both SAT and AHG tests but was poor at low dilutions. Ariza (2006) found that whereas AHG could predict illness in subclinical brucellosis Brucellacept could not. Casanova (2003) showed that Bucellacapt is a sensitive, specific and simple test for routine work but performs poorly in chronic and relapse cases compared with AHG.

Indirect fluorescent antibody test (IFA).

THe methodology of this test is similar to ELISA. It can detect IgM, IgG and IgA separately. According to Araj (1990) IFA is as good as ELISA at detecting IgG and IgM but poor for IgA. There is a subjective element in the reading of the test. It has been superceded by ELISA.

Radioimmunoassay (RIA).

Radioimmunoassay (RIA) was developed by Chappel () as a test for brucellosis caused by B abortus in cattle. It is a sensitive test technique which can be used to differentiate between Brucella antibodies, IgM, and IgG and IgA. A microtitration plate is prepared coated with a Brucella antigen. Serum or diluted serum is then added so that any antibodies present bind to the Brucella antigen. Sufficient time is allowed then the plate is washed free of sera. Sheep anti-human IgM, IgA, or IgG are added to different wells. This attaches to the Brucella antibodies. The anti-human IgM etc is conjugated with Iodine 123 which is radioactive. The plate is washed and the radioactive level read using a Scintillation counter. White (1978) used this method and concluded that the test found more positive both acute and chronic brucellosis including CFT and 2-ME. He also showed that the IgM test gave similar results to SAT and the IgG test similar to CFT or 2-Me. He noted that positive levels of IgA were indicative of chronic brucellosis. This method has also been used by Parratt et al (1977) and Hewith (1984) to study brucellosis in humans. Using the techneque they have also shown a co-relation between the IgM positive results and SAT titres and a corelation between positive IgG results and CFT. This test is more sensitive than CFT and reveals more positives, For brucellosis the test has been superceded by iELISA (see below) which is similar and easier to perform with no radiation risk.

CELISA.

Lucero (1990) studied competitive ELISA (CELISA) using S-LPS and found it a useful method easy and quick to use. In a study of of 31 patients with brucellosis like symptoms but negative serology by RB and SAT (length of illness not given. CELISA gave positive results for all 31.

Counterimmunoelectrophoresis (CIEP).

CIEP an interesting but little used test for brucellosis. can show antibodies against S-LPS, R-LPS, or soluble protein derived from Brucella. The method is as follows: CIEP is carried out on agar slides. The sera is applied on the anode plate end and the LPS or protein antigen on the cathode plate end. Interaction of the antigen and antibodies occurs during electrophoresis which is shown by the presence of bands on the plate. In the case of Brucella soluble protein, there are no bands in the first 2 weeks of the illness but thereafter there are up to 3 bands. This is a quick test takes only 1 -2 hours. There is limited published work on this test but it may be a useful test for chronic brucellosis when the Brucella proteins antigen or R-LPS are used. One problem is that the response to R-LPS and protein antigen is much weaker than that to OPS.

Daz (1976) considered that CIEP was a rapid and sensitive method to study the serological response to Brucella of soluble proteins rather than LPS. The test showed up to 7 protein antigen bands. Mahajan (1991) studied 647 sheep with B melitensis using SAT and CIEP. SAT recorded more reactors in the early stage of the infection but CIEP was better in the later stages. Trap (1982) studied 30 experimentally infected sheep and 126 sheep with chronic brucellosis using SAT, CFT, RB, and CIEP. They found that CIEP was the best test for chronic illness for which SAT, RB and CFT were often negative.

Dipstick Assay.

This is an intersting test which could be used as a rapid test for acute brucellosis or for surveys.

Smits (1999) have developed a rapid dipstick method for acute brucellosis. A antigen prepared from B abortus 1119-2 was coated as a strip to a sheet of nitrocelloulose which was then laminated to a plastic backing and the cut in strips 2.5 mm wide. The detection involved a reagent prepared by conjugating monoclonal human IgM to colloidal dye particles (palanyl red). The test involves incubation of the dipstick with the reagent and serum for 3 hours. The amount of staining is determined by reference to a colour reference strip. The test method measues IgM and is therefore not suitable for chronic brucellosis. According to the authers, in the first 4 months of illness it works as well as SAT and is quick and easy to use. At about 6 months its sensitiviy drops to 30% presumably because of a drop in IgM levels.

Scientific Understanding of Antibody Serology Tests.

Diaz (1976) showed that IgM, IgG and IgA are active in agglutination tests. This was done by adding anti-IgM, anti-IgG and anti-IgA.

Heremans (1963) showed that CFT measured IgG and limited IgM but no IgA.

Coombs (1967) confirmed that the non agglutinating antibodies were IgG and IgA. Kerr and Payne (1967) also showed that IgG and IgA predominated with in chronic brucellosis while there was no significant level of IgM. White (1978) used radioassay to demonstate that IgM occurs in acute brucellosis and IgG in chronic brucellosis. Coughlan (1967) noted that individuals who either had chronic brucellosis or who had been repeatedly challanged by contact with Brucella had negative or weakly positive agglutination (IgM) wereas CFT and AHG tests for IgG were positive. Reddin (1965) showed that IgM is not generally present in the sera of chronically ill brucellosis patients. Kerr (1968) found that IgM readily agglutinates but is destroyed by mercaptoethanol. IgM and IgA do not fix compliment whereas IgG is not demonstrated by the agglutination test and readily fixes compliment and is also revealed by the AHG test. SAT measures mainly IgM and is good for acute brucellosis. Sippel (1982) observed that acute patients have high IgM and chronic patients high IgG. Williamson (1996) noted AHG and CFT measure mainy IgG with a little IgM and they are therefore suitable for longer term illnes.

Dahouk (2003) found the sensitivity of ELISA, AHG, and CFT were: 98%, 92%, 77% and the specificity: 98%, 98% and 87%. She indicates that the antibodies measured by the main tests were: SAT, IgM, IgG2, IgA; RB, IgM, Ig1; CFT, IgM, IgG1; and AHG, IgG1, IgG2, IgA. Buchanan considered that the effective order of usefulness of the tests in chronic brucellosis is: AHG >CFT >SAT.

According to Rice (1971) in cattle the serum of chronic shedders shows a mainly IgG1 and some IgG2 and IgM. Levieux (1974) found that IgM is the main antibody detected by SAT, Rose Bengal and CF tests. IgG1 is inactive in SAT but can be detected at pH 3.6 and by the CF test. IgG2 can be detected by SAT but not at pH 3.6 but not in the CF test. Nicoletti (1980) notes that it is very important to realise that no single test or even combinations of tests will comclusively diagnose all cases of brucellosis in cattle.

Dahouk (2003) concludes that IgM and IgG both persist for 1 year even with successful treatment. IgM disappears in the chronic phase and IgG and IgA decline to low or negative levels. This is also true of subclinical brucellosis. IgG peaks again during relapses.

Comparison of Serology Tests.

McDevitt (1969) in a survey of 224 vets exposed to B abortus and possibly suffering from chronic brucellosis concluded there was no coorelation between positive values for AHG and CF tests and symptoms. However some with positive AHG values but no symptoms could latter develop typical brucellosis symptoms with a high AHG titres.

Araj (1996) looked at the Brucella antibody profile of 597 people working in high risk occupations of whom 107 had possible symptoms of brucellosis. The overall serology results were: SAT, 1,7%; AHG, 15% and ELISA, 26-61% positive. Those with possible symptoms had higher values.

Hardy (1935) consider that Brucella can establish a focus of infection in the body and stimulate antibody production apart from manifestations of the disease. The patient is not ill but but an exacerbation can occur at any time.

Problems with Serology Tests.

Persistence of Antibodies against Brucella in Sera.

In general IgM which is measured by agglutination tests such as SAT peaks in the first few months of illness and then declines to negative or low values. This occurs in both those who are apparently cured and those who become chronically ill. IgG and IgA also decline after the initial peak in both treated and chronically ill patients. However some patients retain high levels of these antibodies even if assymptomatic or apparently sucessfully treated. An important question is whether such patients are in reality carrying a sub-acute infection as indicated by PCR tests described above. Unfortunately there are few coherent follow up studies of such patients and the contraversy of the status of chronic brucellosuis leaves a muddled picture. Here we look at factors which affect the viability of antibody serology tests.

Persistent IgM as measured by SAT: It has been widely noted that SAT titres which measures mainly IgM can persist at elevated levels for years after initial infection. Evans (1938) observed that agglutinins disappear from the blood of some patients whereas in others it can remain for several years after apparent recovery. She notes that similar results have been found by Olin (1935) and Carpenter (1929) . Evans also noted that individuals who were exposed to Brucella may develop agglutinins without notable disease. Similar results were found by Carpenter (1929) , Huddleston (1930) and (1937) , Jordon (1931) and Dooley (1932) . Similarly Buchanan 1980) notes that patients who have apparently had successful treatment still had positive SAT 18 months latter. Young (1991) noted that IgM could persist for several years despite apparent clinical recovery. It has been shown by Cooper (1992) and Gilbert et al (1981) that individuals such as veterinarians who could be continually exposed to Brucella could also retain a persistent level of IgM. Almuneef M (2002) showed that in a survey of apparently successfully treated patients in an endemic area 26.6% were still seropositive by SAT over 3 years latter.

Persistent IgG: Gazapo (1989) studied 52 acute brucellosis patients who had been initially treated with antibiotics and then their serology monitored using ELISA IgM and IgG for up to 24 months. For 29 patients after an intial peak of both IgG and IgM there was a steady drop in both to low levels over 12 months. The other 23 suffered relapses shown by a secondary peak of IgG but not IgM. From these 23 patients 3 developed localised illness and 7 gave a positive culture test during relapse. These relapsing patients were given further antibiotics but the others were not. In all these case there was again a slow overall decline in IgG levels but over a period up to 24 months. At this stage they were considered to have recovered but there is no further information.

Persistance of IgA.' Marrodan (2001) has shown that both IgG and IgA can also be agglutinating antibodies and can be responsible for positive SAT values when IgM values are low. This may explain persistently high SAT titres in people exposed to Brucella. This is also discussed by Al Dahouk (2003) who noted that tests with RIA had shown that SAT can be positive without high levels of IgM because of contributions from IgG and IgA.

Hypersensitivity. Hypersensitivity has been suggested as an alternative explanation of the cause of ongoing elevated serolgy titres and symptoms in long term and chronic brucellosis. Poole (1975) reviewed the case of veterinary surgeons repeatedly exposed to Brucella. She considered that a likely senario was that these vetinerians had contracted brucellosis and then recovered. Then subsequent high anti-Brucella serology titres symptoms with pyrexia were due hypersensitivity caused by repeated re-exposure to Brucella antigens. Henderson et al (1972) in a study of people exposed to outbreaks of B abortus showed that about 30-50% developed subclinical infections ie. they had an antibody response but not apparent symptoms. In most cases after one exposure they became seronegative after 1 year and appeared well. There is no data to determine if the illness returned latter in a chronic form. In veterinerary surgeons who were be exposed repeatedly to B abortus antibody response was 92%. IgG as shown by AHG was the main antibody in long term illness. Many developed symptoms and hypersensitivity with skin rashes on contact with infected material. In general the modes of contamination were as follows: general public, drinking raw milk; farmers, drinking raw milk and contact with contaminated material from calfing; slaughter men, contaminated carcuses; veterinarians, contaminated material mainly from calfing.

The delayed type hypersensitivity (DTH) is well documented in the study of brucellosis as used in the skin test and vaccine treatment. Also it is well documented that farmers and veterinarians who have had direct contact with infected calfing materials can develop dermatitis skin irritation and rashes. However it is extremely unlikely that hypersensitivity could explain cases of chronic brucellosis because of positive culture tests during relapses and more recently positive PCR DNA test results.

Gilbert et al (1981) [889] in a study of abattoir workers in NZ showed that some workers who were asymptomatic had antibodies against brucellosis. Workers who had worked at the plant fo up to 1 year had serum IgM but those working for longer periods IgG. They considewred that this was due to a sub-clinical infection and considered it impossible to distinguish them from patients with chronic brucellosis with a history of acute infection.

McDermitt (1994) [905] notes that in the Republic of Ireland where brucellosis remains indemic in cattle it has been shown that over 40% of blood donors in rural communities have SAT of 1:160 whereas only 14% urban dwellers have simiular titres.

False Positive Serology Tests Results due to Cross Reactions. THe main serology tests for brucellosis in humans test for antibodies to the LPS of Brucella. In this respect the O-side chain of the LPS is the immunodominant antigen. Araj (1996) notes that cross reactions can occur in serology tests. This associated with antibody response to Brucella LPS. It affects agglutination test and some iELISA tests. ElISA tests based on membrane protein are not affected. This same O-side chan is also present in other bacteria and the presence of antibodies to these can give rise to false positibe results in the brucellosis test. According to Pappas (2005) bacteria which can interfer in this way are: Francisella tularensis, Escherichia coli 0116 and 0157, salmonella urbana, Yersinia entercolitica 0:9, Vibria cholerae, Xanthomonus maltophilia and Afripia clevelandensis, This problem applies to all agglutination tests.

Ariza (2006) [902] notes that false positives can also be caused by rheumatiod factor (RF). During a relapse with an increase of IgG RF appears simulating a IgM response. They note that Mousa (1987) notes that 8.8% of patients with osteoarthricular complications in brucellosis produced RF positive results.

Antibody Response in Relapse Patients. Even if the best treatment available is used at least 10% of patients with acute brucellosis relapse.

• Pellicer et al. (1988) followed the serology of 10 patients who had relapses up to 2 years after an acute attack. The relapses often occured a few months after apparent cure. They found that IgG increased in 9 patients and IgA in 6 patients. SAT and DTT increased in 3 patients. AHG and iELISA IgG were good indicators of relapse but iELISA IgM was not. DTT which determines IgG gave poor results. 1 of the 10 cases showed no serological increase at all by any test. Relapse lasted about 3 months. Serology results had been declining in 4 of 10 patients prior to thr relapse.
• Ariza J (1992) in a further study showed that after apparent successful treatment of the 81 patients 12 patients relapsed: 4 at 1 month; 2 at 2 months; 2 at 3 months; and 1 each at 4, 6, 7 and 9 months. 2 patient relapsed twice at 7 and 15 months. Statistically there was no difference in the serology data of those who relapsed and those who did not. Of the 63 patients who did not relapse 16 had persistent serum antibodies 12 months after therapy: .Various test methods showed this as follows: SAT, 5 patients; iELISA IgM, 6 patients; AHG, 11 patients, ELISA IgG, 11 patients; ELISA IgA, 9 patients. On the basis of their serology data they could not determine if the illness was resolved or not and the patients cured. This IgG persistence does not necessarily indicate chronic illness.

Negative Serology Tests.

It is normal for SAT titres to decline to low or negative levels irrespective of whether the patient is cured. Negative SAT titres can occur in acute, assymptomatic, sub-acute or chronic forms of the illness. Problems of negative results associated with the CFT, AHG and iELISA tests are discussed under the individual tests. THe problem of negative serology tests also applies to animals is discussed elsewhere.

• Evans (1938) considered that agglutination test results are frequently negative particularly in chronic brucellosis. Similar results were found by Gould (1937) , Angler (1935) and Simpson (1934) .
• Green (1938) notes that in the vast majority of cases of chronic brucellosis that SAT is negative but that the test works well in acute illness.
• Harris (1943) and (1946) notes that SAT is negative in the majority of chronic brucellosis cases and is frequently negative in acute cases. He also notes that negative SAT titres can occur at the same time as positive Brucella culture tests. Jordon (1947) noted that the SAT titre tends to decline during the months after initial infection and usually becomes entirely negative. However in patients SAT titre remains high for up to 10 years even though there is apparent clinical recovery.
• MacDonald et al (1967) notes that in apparently successful treatment IgG decline to low or negative values but that persistent positive values were associated with relapses and arthritis.
• Evans (1938) points out that it is a common practice to assume that negative or low SAT results indicate evicence of inactive Brucella however there are a significant number of cases negative SAT in association with positive culture tests. These include studies by Burnet (1922) and (1929) , Carpenter (1929) , Gilbert (1932) , Meyer (1935) , Taylor (1931) , Huddleston (1936) and (1937) , Poston (1936) and (1936) .
• Griggs (1943) reviewed the use of the SAT titre in 96 patients. 86 were negative and 10 positive. Of the 10 positive: 4 had a titre of 1:80 or higher; 3 had atitre of 1:40, 1 at 1:20 and 2 positive with unspecified titres. He considered SAT less useful than ths skin or opsonocytophagic tests.

Negative Sreology Tests with Positive Culture Test. There are a nunber of reasons why SAT could be negative concurrent with a positive culture. In acute illness the SAT titre can be taken before seroconversion has occurred. IgM usually occurs at about 7 days after infection. Also it appears that in some patients seroconversion can be delayed. It has also been noted that antibody response to Brucella is variable from patient to patient. Corbel (2006) notes that the response to SAT is individual so that some have high tires and others low titres. He noted that in in a study of 238 brucellosis patients 29.2% had a titre of ≤1:80 and 3.4% had a tire of ≤1:10. Another possible explanation is that what appears to be an acute attack is really a relapse after a long sub-acute infection. In this case IgM could be absent but IgG might be present. A number of cases of positive culture test with negative serology tests have been documented:

• Huddleson (1936) studied 100 bacteriologically proven cases of B abortus and found that 29 had negative agglutination titre.
• Evans (1938) points out that it is a common practice to assume that negative or low SAT results indicate evicence of inactive Brucella however there are a significant number of cases negative SAT in association with positive culture tests. She refers to cases described by Burnet (1922) , Carpenter (1929) , Gilbert (1932) , Meyer (1935) , Taylor (1931) , Huddleston (1936) and (1937) , Poston (1936) .
• Harris (1946) notes that the SAT test is very frequently negative in acute illness.
• Further examples are supplied by Harris (1950) . These include: Evans (1934) , Poston (1936) and Taylor (1938) .
• Foshay (1940) notes that SAT is the best diagnosis after the culture test, but a negative SAT does not exclude illness. Non-standard methods are a problem. 6-10% of cultural positive patients have negative SAT. In many cases SAT positives only occur in the early stages of the illness but they can occur intermittently at irregular and unpredictable intervals. Agglutinins can disappear or persist for years. High titres are very unusual in chronic brucellosis and if they do occur are at low levels. In these cases critical thresholds are of no value. He concludes that the SAt test is less useful than the skin or phagocytic tests.
• Carpenter (1939) recorded several cases with positive B abortus culture but negative agglutinins.
• Castaneda (1942) reports that up to 12% of patients can have positive Brucella culture tests and negative agglutination.
• Mann (1973) notes the case of a brucellosis patient described by Payne (1972) with active brucellosis and a positive blood culture but negative serology.
• Huddleson (1940) describes an outbreak of B melitensis at a laboatory where 41 people became ill of whom 35 had positive culture tests. 49 other workers who did not have clinical symptoms were also tested and of these 3 had positive cultures. This shows that positive culture tests can occur with low or negative SAT tires but also that 3 symptomless people had positive culture tests and of these 2 had low or negative SAT titres. See Table......

SAT titre	+ve Culture	+ve Culture
Dilution	Clinical	Sub-clinical
	No Patients	No Patients
Negative	3	1
1:25	12	0
1:50	7	1
≥1:100	13	1

False Negative Serology Results due to the Protoze Effect. The effectiveness of agglutination tests is inhibited by blocking proteins or the so called prozone effect. If such blockin antibodied are present then they give rise to false negative results. The are more likely to be present in longer term illness. The tests affected are SAT, RB, CFT, and iELISA. Zimmerman (1959) and (1961) concluded that blocking antibodies were present which coated the Brucella antigen preventing aglutination and that these were the only antibodies present in long term illness. Wilkinson (1966) showed that the blocking proteins were IgG and IgA but not IgM. This coating of IgG on the antigen prevents IgM from from forming an agglutinate with the antigen and therefore this leads to false negatives results.

Zinnerman (1959) , (1959) and (1871) (1961) and Joos (1968) studied blocking antibodies and equated them to the prozone effect. They considered that blocking antibodies occurred due to prolonged immunisation and in part accounted for the failute of agglutination tests in chronic brucellosis. They recommended the AHG test in such cases. Corbel (1988) suggests that such antibodies are directed against LPS and are the IgG1 and IgG2 antibodies. Studies in cattle indicate that they are IgG....

A slightly different view is offered by Young (1991) who reviewed the prozone effect in 214 cases of brucellosis. He concluded that it was observed in 6% of sera samples especially when antibody levels were high and occurred most often at low dilutions such as 1:20 and only rarely at dilutions of >1:80 and was of little practicle importance at higher dilutions such as 1:320.

Negative Seroloy Tests and Positive Skin Test. The interdermal skin test was widely used to test for brucellosis from the 1930-60s. Like other tests there was widespread contraversy associated with its use. Problems included non standardised tests and Brucella antigens. Historically to forms of the test were used. Either the antigen was killed Brucella or protein extracted from Brucella (Brucellagen). Brucellagen induced a delayed type hypersensitive reaction while killed Brucella induced both a DTH and a humeral antibody reaction. Therefore antibody serology tests could not be carried out after theis test. The test can tell you if the patient has been in contact with Brucella but can not distinguish between past and present infection. In cattle the test has been re-evaluated in recent years. It has been found useful in countries where the incidence of brucellosis in cattle is low because it can help differentiate between cattle with possible brucellosis and cattle wth Y entero.... exposure which is the most common cross reacting bacteria in cattle. A number of authors have describe cases of positive skin tests with acute brucellosis:

• Pearson (1973) considered that patients with active brucellosis could have positive brucellin skin test sensitivity in combination with negative serology tests.
• Spink (1951) noted that 90% of patients with a SAT of 1:≥320 were bacteremic. In an indemic area of Minnesota 18.54% had demonstatable SAT titre and 2% had a titre of 1:≥160. Titres of 1:≤100 were difficult to interpret. 20% gave a positive skin test.

Serology Tests in Chronic Brucellosis.

In general all the usual antibody serology tests, that is SAT, RB, CFT, AHG, iELISA IgM and IgG which are used to diagnose acute brucellosis do no work in chronic brucellosis. Some patients give positive results and some negative. Generally the number of positive results decline as the period of illness increses. There is generally no co-relation between serology test results and symptoms of illness. Some patients have localised leisions which are easier to diagnose. The diagnosis of chronic brucellosis from clinical symptoms is also very difficult. Symptoms are generallised such as fatigue, weakness, muscle or joint pain. This creates a great problem for doctors faced with a chronic patient. Typically an average local community doctor or hospital specialist in infectious diseases will have insuffient knowledge of the subject. Such a doctor will carry out routine blood tests and specific tests for brucellosis. When these show no abnormaliuties he will assume incorrectly that brucellosis cannot be present. Having found no obvious clinical signs of illness such a doctor might give a diagnosis of chronic fatigue syndrome. The most likely doctor to reach the right conclusion is one working in a country where brucellosis is in endemic and who has long experience of the illness. Patients living in developed countries where brucellosis is now rare are probably the least likely to get to a correct diagnosis despite having superior medical facilities. Diagnosis is just part of the problem because the typical treatment used for acute brucellosis will probably fail in chronic brucellosis. So even if the doctor advocates such treatment he may assume wrongly that its failure proves that the illness cannot be brucellosis. The most important criteria for a correct diagnosis is history of exposure to Brucella. Many doctors have refused to acept that brucelosis can be present without positive serology tests. This is backed by poorly substantuated theories that brucellosis is caused by psychologic problems rather than a clinical illess often with an aggressive derogatory attitude to such patients. Regrettable these ideas which derive mainly from the USA seem to have attracted support. These are discussed elsewhere in this review. Here we show the opinions of chronic brucellosis by doctors ans scientists experienced in the field:

• Evans (1938) observes that SAT results are frequently negative in brucellosis particularly in chronic brucellosis. This also confirmed by Gould (1937) , Angler (1935) and Simpson (1934) .
• Harris (1944) notes that SAT is negative in a large proportion of chronic Brucella infections even with a positive culture test.
• Current comments (1946) quote, "There is as yet no available any easy adequate specific means of diagnosis of chronic brucellosis. The agglutination test is only significant if positive. A positive intradermaltest does not tell if the disease is active. Negative skin tests can occur with active infection.The culture test gives a positive diagnosis but the test is very difficult to perform.
• Huddleston (1949) note that in chronic brucellosis agglutination tests are negative in 80% of cases and Manchester (1942) considered that in chronic brucellosis serology results are invariably negative.
• Barrett (1953) notes that serum agglutinins are not found in many cases of chronic brucellosis. And notes that it is not unusual where serum agglutinates are absent to isolate Brucella from blood or bone marrow. He sites the work of Carpenter (1930) , Gilbert (1934) , Huddleson (1936) and (1939) , Harris (1948) , Taylor (1938) , Bower (1948) and Scarlet (1949) .
• Rotes-Querol (1956) notes that in chronic brucellosis the SAT, CF and AHG test give values that are negative of or only slightly above. Only the interdermal test gives a clear positive result.
• Zinnerman (1959) separated sera from patients with acute and chronic Br by gel electrophoresis and demonstrated that blocking proteins were absent in acute brucellosis and present in chronic brucellosis. It is noticable that in some chronic brucellosis patients levels of immunoglobulins were very low.
• McDevitt (1969) in a survey of 224 vets exposed to B abortus and possibly suffering from chronic brucellosis concluded there was no co-relation between possitive values for CFT and AHG tests and symptoms. However some with positive AHG values but no symptoms could latter develop typical brucellosis symptoms with a high AHG titres.
• Payne (1974) reports that patients can have all the signs and symptoms of Br without demonstatable antibodies and BR can never be excluded even if SAT, 2-mercatoethanol, CFT and AHG were all negative. He describes a man who was ill for 6 years with chronic brucellosis abortus who was tested 7 times with SAT, AHG, and CFT with negative results. Eventually he was diagnosed by a posative culture test.
• Thorne (1977) has says that there is no such thing as a diagnosis for chronic brucellosis and that serology results bear no relation to symptoms and signs of illness.
• Williams (1982) notes that serology tests are sometimes negative with positive culture tests.
• Kichieva (1982) studied the sera of 89 patients with chronic Br (1-10 years or longer) and concluded that the level of immunoglobulins were either low or normal.
• Young (1983) notes that chronic localised Br may be the principle form of the illness and therefore antibodies may not be demonstrated in the serum.
• Hewitt (1984) studied chronic brucellosis patients who were ill from 10 to 16 years and considered that the organism was notorious for causing relapses and producing chronic Br which may exist over a long period of time and give rise to comfusing symptoms and signs for which no laboratory analysis allowed unequivocal evidence and clinical examination must be used.
• Williams (1988) indicated that objective criteria which on all occasions confirm or exclude a diagnosis of chronic Br do not exist.
• Colak (1992) did a trial with 90 chronic Br patients and found AHG and ELISA IgG worked in 48.9% and 58.9% of cases respectively.
• Alton (1994) considered that the diagnosis of chronic brucellosis is often very unsatisfactory with negative culture and queatinable serology tests a confident diagnosis is often impossible.
• Lisik (2001) noted that humeral IgG falls over time in chronic Br patients and in a test of 27 such cases found ELISA IgG worked in only 25.9%.
• Hatami (2005) notes that in chronic brucellosis symptoms are not associated with clinical ans serological evidence of active desease.
• Cutler (2005) found that all serology tests were plagued with problems of sensitivity and specificity.
• Al-Nassir (2006) indicated that in chronic Br that serology studies and cultures are often negative and this may be caused by inadequate initial therapy and underlying localized desease.

Positive culture Test, Negative SAT.Harris (1944) notes that SAT is negative in a large proportion of chronic Brucella infections even with a positive culture test.

Hajia (2007)

Why do Antibody Serology Tests Fail in Chronic Brucellosis.

Serology tests play a very important part in the diagnosis of both animal and human brucellosis. Experience has shown them to be highly unreliable tests. Most research effort has be directed at understanding the interaction of Brucella with the process of cell mediated immunity. With regards to antibody reponse there is little research other than developing new tests. These have been useful for acute brucellosis and animal tests but have been of little use in chronic brucellosis. Therefore we can only conjecture why there is little or no serology response in animals or humans with chronic brucellosis. Although it is possible that the cur4rent tests do not detect the antibodies to Brucella that are present, it is more likely that antibodies are not present. Possible explanations are:

• Localised Leisions. There is considereable evidence in both humans and animals that Brucella with chronic brucellosis that Brucella occurs in localised leisions. In humans these have been identified either is easily detected sites such as joints or incidently in unrelated surgery. In animals they have been found by post-mortem examination. Typically they can be cysts often microscopic in size. At the same time antibody serology test give low or negative titres and it is suggested that the Brucella is insulated against the impact of the immune system.
• Mutation of Brucella from smoth to rough form. It has been shown by in vitro laboratory work and in vivo animal studies that Brucella colonies can mutate spontaneously from smooth to rough forms and vice versa. An example of the mutation of a smooth forms of Brucella into a rough form is vaccinwe strain 45/20 which was prepared by 20 passes through guinea pigs. The term smouth Brucella relates to the appearance of the bacteria under microscopic examination due to surface LPS which rough forms lack. The lack of O-polysaccharide (O-PS) means that all the usual serological antibody tests do not work for rough variants and this is a possible explanation for the gradual fall to low levels or negative values of serology tests found in chronic brucellosis. It should be noted that in the naturally occuring B canis and B ovis are also associated with negative serology tests. B canis and B ovis give rise to virulant forms of brucellosis but laboratory derived strain are usually attenuated. This was noted by Spink (1956) that rough Brucella mutants are attenuated. Work by a number of workers including Allen (1998) Freer (1996) and (1999) , Martínez de Tejada (1995) , Sola-Landa (1998) and (2002) has shown that the attenuation of rough strains may be related to increase in the antibody complement activation, the sensitivity of polycationic bacterial peptides, and altered attachment to cells. In chronic brucellosis the generallised symptoms are generally less severe than in the acute illness which might relate to the presence of less virulant rough strains of Brucella. A related area of interest is L-form bacteria which are bacteria which lack cell walls but this does not normally restrict reproduction and can revert to the normal form by growing a cell wall which includes O-PS. Change to an L-form can be induced in Brucella by treatment with antibiotics such as penicilin. Interestingly L-form bacteria are known to be more resistant to antibiotics than the parent bacteria, Hatten (1966) . Mikhaĭlov (2009) has noted the persistance of L-forms of Brucella in chronic brucellosis patients.

Test for Delayed Hypersensitivity and Cell Mediated Immunity.

The intradermal or Skin Tests

Fleischman (1918) was the first to use a skin test (cuteneous hypersensitivity test) using killed B abortus to diagnose brucellosis and Burnet (1922) did similar work with B melitensis. Simpson (1929) tested 10 proven cases of brucellosis and 25 controls by the skin test using heat killed B abortus and found that all proven brucellosis cases were positive and all controls negative. Similarly Giordano (1929) using heat killed B abortus showed severe reaction in all proven cases and negative reaction in 100 controls.

The intradermal test (skin test) elicits a delayed hypersensitive reaction in people who have been previously exposed to Brucella. A suspension of a Brucella antigen is injected into the dermal layer. In most cases this antigen is either heat killed Brucella or a suspension of nucleoprotein isolated from Brucella. The antigen can be derived from B abortus alone or from a mixture of B abortus, B melitensis and B suis. Commercial forms of necleoproteins are Brucellergin and Brucellin and these are largely free of LPS which is the main instigator of the normal humeral antibody response. There only a very limited humeral antibody response to Brucellergin and Brucellin. A DTH response to the test indicates past or present exposure to Brucella. Antibody serology tests should always be carried out before the skin test have been completed even if Brucellin is used because of the possible an induced humeral antibody response. The killed Brucella in particular can elicit severe local and systematic responses similar to the general symptoms of acute brucellosis. The test is useful in chronic brucellosis where other serology tests have failed. This is particularly the case in a protracted illness where there has been no obvious onset or cause of illness. It can also be used for epidemiological surveys. The usefulness of the test has en conpromised by confused methodology including: different antigens, dose size and interpret ation of results.

Killed Brucella vaccine v Brucellergen. The lack of unifornmly accepted methodology for skin tests has caused great problems with interpretation of results. This includes dose, choice of antigen and result interpretation. Killed Brucella vaccine gives more positives but is associated with more severe local and systematic reactions. Brucellergen affects patients less but finds less positives. Here are some examples:

• • Leavell (1931) compared intradermal tests for Brucella suspension and Brucella extract and concluded that heat killed Brucella produced better results. The main problem with the use of killed Brucella is the extreme reactions that can occur including abscess formation at the site of injection.
• Angle (1938) 163 instutional patients who had been exposed to Brucella contaminated milk, were tested with both heat-killed Brucella vaccine and Brucellergen. There were 54.6% reactors to heat-killed Brucella and only 26.9% reactors to Brucellergen.
• Harris (1950) carrieed out concurrent tests with heat-killed Brucella vaccine and Brucellergen in 48 patients. 72.9% reacted to heat-killedBrucella and only 35.4% to Brucellergen. Harris noted that the 37.5% of patients with negative reactions of Brucellergen but positive reaction in the heat-killed Brucella test had weaker reactions but he considered that the stength of the reaction has no diagnostic significance.
• McCullough (1949) considered the use of brucellergin most useful. A positive test shows that the idividual has had contact with Brucella and become sensitised. A small proportion of patients do no show a response. Thev skin test is often negative early in the early stages of the disease. There is no relation between the severity of the response and the likelihood of active infection. In a doubtful case a negative skin test makes brucellosis relatively unlikely.
• Goldstein (1934) compared the usefulness skin tests for brucellosis using whole heat killed B abortus and a fat free ground antigen in saline solution. 253 patients were tested with heat killed B abortus and 10.3% gave positive results. 92 were treated with fat-free ground Brucella antigen and 9.6% were found to be positive. Heat killed Brucella gave much more severe positive reaction. This typically included: Erythema around the injection site starting at 12 hours and increasing up to 72 hours. Steaks of lymphatitis, a central hard nodule of 1-3cms diameter, erythematous area of 4-12 cm diameter lasting up to 5 days. Possibly enlarged epitrochlear or auxillary lymph nodes and possible moderate to severe fever and general malaise. The negative reaction showed a small red nodule surrounded by an area of erythema which faded in 48 hours. Among the patients given fat-free antigen, the reaction was milder only a few developed fever and had malaise and necrosis at the site of injection only occurred in 4 patients. In the negative results only a few developed erythema which faded within 36 hours. Goldstein showed that the heat killed Brucella produced a marked agglutination titre response whereas the fat-free antigen produced only a very limited agglutination response. This was confirmed by trials using guinea pigs. Similar results had been found by Levin (1930) and Giodano (1929) . It was considered that the fat-free antigen was superior.

Critera for a Positive Skin Test. According to Harris (1950) positive skin test can occur in the following circumstances:

• In active brucellosis in association with other positive serology tests and typical symptoms.
• In cured patients often long after the illness has gone.
• In people who have contact with Brucella but who have not developed illness and who apparently have natural immunity.
• In people who have contact with Brucella and retained a subacute infection likely to develop into chronic brucellosis.
• In people who have had acute or subacute brucellosis and developed chronic brucellosis.
• In people who have previously received skin tests or vaccine treatment.

Harris (1950) notes that positive skin tests can be associated with negative antibody serology tests and negative opsonocytophagic reation but positive culture tests.

False negative reactions. According to Harris (1950) young children produce less positive skin tests than similar older children. Less positive skin tests may also occur in older people. He also notes that vaccine treatment can lead to desensitisation associated with improved health. He has found that negative skin tests can occur in the presence of positive culture tests but concludes that whereas negative agglutination tests are common in such circumstances negative skin tests only occur in 5% of cases of brucellosis demonstrated by other means. He considers that reports of failure in this test is mainly due to choice of the wrong antigen and incorrect criteria for evaluating the test.

False positive serology tests. Some people who were previously negative for agglutination tests give positive values after the skin test.

• Kirby (1942) studied 50 apparently healthy persons with negative agglutination titres. 25 were tested with Brucellergen (0.1 cc at 1:12000) and 25 with Brucella vaccine (Lederke's mixed B abortus and B suis(0.1 cc). Of 25 tested with Brucellergen 11 had positive skin tests while of 25 tested with Brucella vaccine 4 had positive skin tests. 12 out of 15 of the persons with positive skin test and 10 of the persons with negative skin tests developed agglutination titres. All titres were ≤1:80. Agglutinates persisted for 4 weeks in only one patient. Most probably the persons with positive skin tests had had previous direct contact with Brucella an possibly had subacute infections.
• Harris (1950) notes that where agglutinations are low before skin testing they are often much higher afterwards due compared to those persons with negative agglutination titres before skin testing. In consideration of modern knowledge we might consider this an amnestic antibody response. Harris describes work by Criscuola which shows the unpredictability of agglutination titres in response to postive results of skin tests with heat-killed Brucella. He divides the patients into groups:
• Group 1: A SAT titre of 1:500 appeared 8 days and disappeared by 60 - 90 days.
• Group 2: A SAT titre of 1:500 appeared at 8 days but remained at 1:80 - 1:100 at 60 days.
• Group 3: A SAT titre of 1:500 appeared at 4 days and disappeared by 60 day - 90 days.
• Group 4: A SAT titre of 1:80 appeared 8 days and disappeared by 60-90 days.
• Group 5: A SAT titre of 1:80 appeared in 8 days but was increased by the 13th day.
• Group 6: Negative SAT at all times.

He also looked at the response of patients with negative skin tests, Also divided into groups:

• Group 1: A SAT of 1:80 appeared at 8 days and disappeared by 13 days.
• Group 2: A SAT titre of 1:500 appeared at 8 days but disappeared by 60 days.
• Group 3: Negative SAT titre at any time.

A typical positive result is a raised or swollen red spot after about 48 hours. The skin test does not identify all proven cases of brucellosis, either acute or chronic.

'Dose Size. Leon (1947) considered that the severe reaction to the skin test reported by earlier workers was due in part to the large number of bacteria that they had injected. He compared standard injections of formalen killed B abotus and B melitensis at dose injections of 10,000, 100,000 and 1,000,000 bacteria. He compared people with a history of brucellosis melitensis and normal individuals apparently free of brucellosis. It is accepted that some normal individuals could respond to Brucella because of previous unrecorded contact. The allergic reaction started about 1 month after the commencement of the illness and increased in intensity with time up to 5 years or more. The reaction was slightly stronger to B melitensis than B abortus. He noted an unusual effect. If repeated injections were started within 6 months of the start of the illness then the allergic response increased with time but if they were started after 6 months then the patient gradually became desensitised even if larger doses were used. This was associated with an improvement in health.

Dose.	Normal	Brucellosis
10,000 organisms	6.4 %	41.6 %
100,00 organism	8,5 %	88.8 %
1000,000 organisms	26.1 %	100.0 %

Dustin (1940) notes that the problems associated with the intradermal skin test relate to the use of different antigens (heat killeed bacteria, filtrate of ground bacteria cells and lipid free antigen (eg brucellergen) and different concentrations used in testing. Nevertheless the test gives useful results. They note that the greater the amount of contact with infected animals the more positive results are obtained. in their work they used a dilutes fat-free aqueous solution derived from equivalent proportions of B abortus, B suis and B melitensis. They used 3 dilutions at 250,00, 25,000 and 2.500 bacteria per cc. A discoloured area of 5 mm was considered positive. Strong concentrations of the antigen increased the agglutination titre but for only a few weeks.

Tests using Killed Brucella Harris (1943) , (1946) , (1949) and (1950) prefered a vacinne based on heat killed B abortus rather than mixed vaccine or Brucellergin and he considered that it found more positives. In a study of 48 patients heat-killed B abortus found positives in 72.9% of cases and brucellergin in 35.4%. He considers that a mixed vaccine causes too violent a reaction. He allowed 10 days should for the development of a positive result. Usually a skin reaction becomes apparent by the 5th to 7th day and persist for 10 days. A faint blush indicated a negative result. A positive result varies from a visible reddened nodule at the site of injection to marked reddening and/or and marked sysemic reaction. A positive reaction shows a reaction to Brucella acquired through infection but did not demonstrate if the infection is active. The skin test creates an antibody reaction so serology test will not work after it. He notes that a skin test using heat-killed B abortus could cause a degree of necrosis.

Classifications of Skin Test Reactions. A number of brucellosis workers have described the reaction to the skin test:

• Harris (1950) described a classification of reactions to the skin test as follow:
• A weakly positive reaction (+): A reddened indurated area of about 5mm on the 4th to 8th day.
• A positive reation (++): An area of reddening and indutation of >5mm possibly with surrounding mild erythema.
• A strongly positive Reaction (+++): An area of redness and marked induration of >10mm with surrounding erythema.
• A viotent positive reaction (++++): A marked swollen and indurated area with an angry erythema. This is often accompanied by lymphadenopathy, axillary adenitis, systemic reaction and occasional central necrosis.

He notes that all positive skin significance, ie., that at some time the patient has been in sensitised to Brucella presumably by infection. Low grade reactions occur which cannot be classified as positive or negative but should be monitored for several weeks. If the result remains for more than 2 weeks or increase in prominence they should be re-classified as positive. Favorite (1935) concluded that a positive reaction tended to persist for several weeks whereas the negative reaction ususally disappeared in about 5 - 7 days. Harris notes that the reaction to the skin test using heat-killed Brucella can be delayed. The result can be negative or equivical at 4 days but can show reddening and papule formation of ≥5 mm by the 8th day. Hagebusch (1941) that the development of a reaction to the skin test could take up to 10 days. Harris notes the work of Chrisuolo who studied reactions to the heat-killed Brucella (0.1 cc) skin test in 337 people:

• Reaction greater on the 4th day than 8th day at 44.2%
• Reaction greater on 8th day than on the 4th dat ay 22.2%.
• Reaction negative on 4th day, positive on 8th day at 18.4%.

Harris (1950) notes that

• Spink (1962) classified the skin test reactions as follows:
• O: Erythema <1 cm.
• +: Redness and swelling of 1 cm.
• ++: Redness and swelling of 2 cm.
• +++: Marked erythema and induration (possibly with lymphadenopathy).

• Green (1938) describes the skin test: a small quantity of killed Brucella is injected intradermally into the forearm. The reaction begins in 24 hours or less. A rapidly spreading red area appears with an indurated centre usually with edema. There is associated heat, pain and itching and sometimes sloughing of superficial tissue. He suggests that a reaction persisting for 3 days can be taken as positive. Other reactions by patient are variable: 25% show no reaction; 10% feel better; 65% have an intensification of symptoms. Rarely they have severe constitutional symptoms.

Tests using Brucellergen and Brucellin. Griggs (1943) reviewed the intradermal skin test on 99 patients and considered it more useful than SAT. He used Brucellergen at dilutions of 1:12,000, 1:48,000 and 1:120,000 and vaccine. Brucellergen at 1:12,000 is milder in effect than vaccine. Using a combination of these tests his results were as follows: Positive, 97; negative, 2. He notes that a negative test does not rule out the disease if other evidence supports the diagnosis and other data is needed to establish active infection.. Skin necrosis was a problem and occured in 11 patients (See vaccine treatment). A positive is shown by a raised or hardened or raised spot or a flattened red spot persisting for many daysn or a ecchymosic spot.

Foshay (1940) noted that vaccines or brucellergen were used of the Brucella skin test but he preferred brucellergen as described by Huddleston which gives fewer false positives and less severe local or general reactions. Skin tests are positive in the great majority of proven cases but it has been found negative in a small number of culture positiver cases. It is also positive where there has been previous contact. Skin sensitivity can be acquired in exposed subjects who have never been ill. It cannot distinguish between past and present disease. It can disappear after recovery from illness but in most cases it persists for many years afterwards. Also Foshay (1938) considered that the skin test could not be used to assess the state of recovery of a brucellosis patient and desensitation which can occur during vaccine treatment did not equate to cure.

Chronic Brucellosis. Manchester (1942) considered that the skin test could be used to indicate chronic brucellosis in circumstances of negative agglutination test although he recognised the problems.

Problems with the Skin Test. Goldstein (1934) found that the use of a heat-killed suispension of Brucella for skin testing stimulated positive SAT titres in most subjects but Goldstein noted that if a fat free antigen was used there was no response. Braude (1948) noted that a skin test with Brucella antigen lead to a rise in SAT titres. Harding (1949) found that 20% of students previously with negative aggluination titres developed positive tires after a skin test. Elton (1948) considered that the increase in SAT titre was more dependant on the amount and type of antigen used. Harris (1949) notes that the skin test must be carried out after the agglutination antibody test because the skin test will evoke a false positive result and is therefore would be compromised. This also applies to the opsonocytophagic test. Barrett (1953) examine 14 health volunteers using a bucellin skin test. All were tested by SAT and AHG before the skin test and only one had a positive reaction. The skin test produced 2 reactors at 48 hours. One volunteer with a previous positive SAT and AHG was positive again. All the others were negative for skin test and SAT. The tests were repeated over 4 months with no change except for the 2 skin test reactors who showed strong skin test reaction and high SAT and AHG. He also noted that in patients with active brucellosis there was a usually but not always a sharp rise in SAT titres.

Evans (1938) Heathman (1934) considered the skin test relatively unreliable. Spink (1945) found examples of negative skin tests in acute brucellosis and Foshay (1940) , McBride (1934) and Debois (1930) found examples of negative skin tests in chronic brucellosis.

Evans (1938) notes that a positive skin test denotes a past or present infection. She notes that people in contact with infected animals or food could develop a state of hypersensitivity without symptoms. This has been verified by Sorge (1925) , Bastia (1928) , Dubois (1930 and 1931) , Debois (1931) , Huddleston (1930) , Thomsen (1931) , Dooley (1932) Johns, (1932) , McBryde (1934) , Molinelli EA. (1934) , Olin (1935) and Binder (1937) . Evans (1938) also notes that in some cases the intradermal test has prove negative in some patient with positive culture tests. This has been shown by Debois (1931) , Taylor (1935) , Poston (1936) , Vendel (1926) and Robinson .

Epidemiological Surveys. Skin tests can and have been used for epideniological survey. Recently the test has been adopted to screen for cattle with exposure to brucellosis in countries with a low incidence of the disease. The test can also differentiate between infection due to brucellosis and other cross-reacting bacteria. The following are some examples of surveys in humans carried out with the skin test:

• Gersh (1938) carried out skin tests on 491 hospital patients and found 12% gave positive results whereas only 1.2% of 5000 patients gave a positive agglutination test.
• Foshay (1940) using Brucellergen compared the usefulness of the skin test in group of patients in an endemic area of California and Cincinnati. The Californian group was mostly comprised of possible chronic brucellosis patients. The Cinicinnati group were normal individuals with limited possible contact with Brucella. Results showed 91% reaction in the Californian group and 8% in the Cincinnati group indicating that the tests give viable results.
• Griggs (1943) notes that after the use of Brucella antigen vaccine patients who previously had negative SAT result became positive. Of 60 patients with with negative SAT, afterwards: 29 became positive (Of these 5 were at 1:320 and 17 at ≥1:80) and 31 remained negative).
• Krakauer (1951) studied 34 patients probably suffering from brucellosis of whom 32 reacted positively to the brucellergen skin test showed a significant rise in SAT titre 8-22 days later. Whereas 85 controls showed no reaction to the skin test and low SAT titres.
• Davis (1957) <ef>http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1962762 Davies JE. Chronic brucellosis in general practice. Brit Med J. Nov 1957. 1082-87.</ref> studied chronic brucellosis in an area of UK and found that 21% of 126 patients in hospitals but with no history of brucellosis had positive skin tests. This corresponded with a concurrent survey by Dalrymple-Champneys (1950) which showed that at that time in the UK upto 20% of cattle were infected by B abortus. However he noted that a skin test often resulted in a subsequent positive SAT test.

Harris (1950) considered that a positive skin test combined with a SAT of 1:40 and some degree of phagocytosis was a tentive diagnosis for chronic brucellosis. He points out however that SAT in chronic brucellosis is usually negative

Leukocyte migration inhibition (LMI).

LMI is an alternative method of method of measuring Cell Mediated Immunity and therefore delayed hypersensitivity to the Interdermal Skin Test. Whereas the antigen in the skin test can sensitises the host to further tests and possible induces a humeral antibody reaction so that it can only be used once the LMI can be repeated several times and can be used to monitor a fluctuation condition. A test capillary tube migration system using peripheral blood was described by George (1962) and developed by David (1964) . THe system was further modified by Soborg (1967) (1970) and (1971) for use with human leukocytes. The method was used to measure developed an method of testing migration rate of human lymphocyte in media containg Brucella antigens. They measured LMI using B abortus antigen test and reported a good corelation with the intradermal sensitivity to brucellin but there was no corelation to humeral hypersensitivity as represented by the agglutination titre. They also showed that the test could differential between hypersensitivity due to Brucella antigen and Tuberculin. The result is determined as a ratio of the areas of migration circle to that without the antigen. In order for the test to work it requires the presence of both lymphocytes and and granulocytes. The lymphocytes must be sensitised but not the granulocytes. It is considered that when this combination is exposed to antigen that a soluble fraction is released which apparently interacts with granulocytes to induce migration inhibition.

In brief the method used by Soborg (1967) method for LMI was as follows: Venous blood was collected and mixed with heperin and allowed to sediment at 37°C for 1 hour and the plasma collected excluding erythrocyte. The white blood cells were washed and centrifuged (1000 rpm) 3 times the the white cells wre resuspended in Eagle medium. THe suspension was drawn into a 1.4 mm diameter capillary tube which was then sealed by melting at one end and centrifuhed at 3000 rpm for 10 minutes. The tube was cut just below the liquid-solid interface and the tube inverted (cell portion) in a fixed position at the bottom of a circular 1 ml tissue culture chamber which was then sealed and was left at 37°C for 24 hours and the area of the diffusion circle measured. The migration index was determined as the ratio of migration through the culture media with and without antigen present. Using this method he showed that 15 brucellosis patients with positive skin tests were compared with 15 controls. The average migration index for the brucellosis patients was 0.59 ± 0.07 and for controls 0.94 ± 0.08.

This is an intersting test which has been studied by a number of workers. There seems to be a good corelation between the results of this test and the inerdermal skin test but not with serology test results. The test could be particularly useful to identify cases of sub-acute or chronic brucellosis and to monitor treatment in such cases. In the latter case it is concluded that in chronic brucellosis patients have a poorer response (lower values) to LMI than patients with acute brucellosis. Treatment is considered successful if a reduction in LMI occurs. The following is a review of the use of the test:

• Soberg (1967) used a inhibition migration techneque to compare 2 groups: one group were brucellin-position and the other negative. There was a close corelation between positive migration inhibition and positive brucellin testresults but there was no corelation to serology tests.

• Soborg (1970) (1971) used the leukocyte migration inhibition (by B abortus antigen) test and reported a corelation between skin sensitivity to brucellin and leucocyte inhibition.

• Mann (1973) used leucocyte migration inhibition to assess patients with possible brucellosis. In their opinion the test was not an objective test for determining activity of infection but was useful to give a possible diagnosis of brucellosis in cases where antibody serology tests were absent or low. One problem with this work is that the designation of whether the patients were suffering with brucellosis was done on a subjective assessment basis by doctors.

• Nyerges (1975) compared 18 patients with B abortus to 19 controls using skin test, antibody serology tests and LMI. There was a fairly good corelation between LMI and the skin test but no corelation with serology tests, the severity of the skin test reaction or the clinical symptoms. Nearly identical results were obtained for skin tests and LMI in subacute and chronic brucellosis. It is considered a good alternative to the skin test.

• Raptopoulou-Gigi (1980) used the LMI test to evaluate the usefulness of levamisole to as an agent to counter anergy in chronic brucellosis. LMI was used to monitor the treatment which was apparently sucessful. LMI decreased to normal levels as a consequence of treatment.

• Rementsova (1983) used LMI to diagnose 201 patients with chronic brucellosis. He considered that it could be used a diagnostic criterium for chronic brucellosis especially in seronegative cases of the disease.

• Boura (1984) . (1989) and (1999) has used the test to monitor treatment of patients with chronic brucellosis. This shows its obvious advantage over the skin test.

Brucella ovis seems to be almost exclusive to sheep and evidence of human infection is limited, Meyer (1982) .

In Vitro ELISA INF-γ Test.

Weynants (1995) developed an in vitro blood INF-γ test which in trials using cattle gave similar results to delayed type hypersensitivity (DTH) which was performed as an interdermal injection with Brucellergene. The test depends on a cell-mediated immunity type reaction stimulating the production of INF-γ. Because the test is carried out in vitro there is no chance of the recipient becoming sensitised as happens with the interdermal test. Heparinised whole blood in buffered saline was cultured with Brucellin, a mixture of cytoplasmic proteins from Brucella melitensis. The culture was held for 18 - 24 hrs then the supernatant harvested and tested with a proprietary INF-γ ELISA. 10 brucellosis free heifers were experimentally infected with B abortus s544, tested regularly during 80 days when they were slaughtered and their organs tested by bacteriological cultue. THe level of infective dose was chosen to give minimal seroconversion at the limit of approved serology tests. All 10 heifers were tested by the ELISA INF-γ test before infection and gave low simulation indices (SI). These were taken as a guide for non-infections. The animals were also tested by standard serology tests including CFT, RB and ELISA. They were also tested by DHT. A second trial was carried out on 23 brucellosis free cattle in fixed stalls where 1 animal became infected with B abortus 3 and aborted. The ELISA INF-γ test was used to monator the spread to other animals together with DTH, approved serological tests as well as bacteriological data at slaughter. In both trials the IFN-gamma assay detected infection in more animals than any combination of the serological tests, and it detected infection earlier than these tests. The DTH test is a very useful test but has the disadvantage that it is time consuming taking up to 3 days. The new test is more rapid ~24 hrs and more convenient. Weynants also considered that ELISA INF-γ can differentiate between infected animals and those showing false-positive serological reactions in a similar way to DTH such as from bacteria with similar LPS to Brucella.

Dobracki (1991) used a INF-γ test to test human chronic brucellosis patients. Whole blood was stimulated with non specific and Brucella specific antigens. The amount of INF-γ stimulation was he same in both controls and chronic brucellosis patients indicating that the INF-γ system worked correctly in the chronic brucellosis patients. Specific Brucella antigens produced produced a 4 fold increase in INF-γ. The results show that in both active and inactive chronic brucellosis delayed hypersensitivity is still immunologically active. The test is useful in long term latent brucellosis.

Tests for Phagocyte Anergy.

Opsonocytic Test

Tho opsonocytic test measures the degree of resistance by phagocytes (white blood cells) to live brucella in the blood of patients. A result is obtained using simple index devised by Foshay is used. According to the method of Gersild (1941) patient's serum in saline solution is mixed with blood in sodium citrate buffer and live brucella culture added. THe mixture is incubated for 30 minutes at 37°C. The amount of phagocytosis is measured by observing the by counting the number of bacteria engulfed by individual leucocytes is counted according to the predetermined index. Effectively in the absense of brucellosis past or present phagocytosis does not occur. Heizmann et al. noted that the test was time consuming and could not distinguish between brucella and y enterocolitica and considered it obsolete. The use of live brucella. makes it a danger to laboratory workers. In general there are technical difficults with the test and it is no longer used.

The value of the test for diagnosis has been confirmed by: Keller (1936) Evans (1937) Foshay (1937) , but the reliability was questioned by Evans (1937) and Tovar (1944) .

McCullough (1949) consideres that the opsonocytophage test does probably does reflect bodily response to Brucella infection but it is difficult to do and interpretation of the results is difficult.

Harris (1943) , (1944) and (1946) is an advocate of the test and notes that it is essential that blood is tested wihin 3 hours for good results. He noted that if SAT, skin test are negative then the test is of no value. A positive skin test plus a low phagocytic index indicates the patient has not recovered whereas a positive skin test plus high phagocytic index indicates recovery however this recovery may not be permanent. The test is not negative or positive but only suggestive. A problem is that high values may occur in the presence of localised brucellosis. There is definite merit to this idea which is effectively a test for anerginicity.

Harris (1949) considered that the opsonocytophagic test worked well if performed correctly. This included: use of blood less than 3 hours old; the correct concentration of citric buffer; the use of pure virulent Brucella; and meticulous technique.

Huddleston (1934) Keller (1936) and Gould (1937) tested th opsonocytophagic reaction and concluded that a strong positive reaction indicates immunity and a weak reaction indicated infection.

Evans (1938) reviewed trials of the opsonocytophagic test and considered it unreliable although her approach was different from Huddleston's. She noted that both SAT and the osponocytophagic test varied over time.

Griggs (1943) notes that this techneque is subject to much error; human , time and methosd variation neverthe less he considered a useful test if done properly and measured resistance to the illness. He recorded phagecyte the index as a percentage. The results for the phagecyte index in 89 patients were as follows: ≤10, 42 cases; 11-30, 30 cases; 31-50, 14 cases; 51-90, 3 cases. The average index value was 20 and of these 64 patients who recieved vaccine treatment after which phagocyte index increased to 61-99 in 72% of patients. He considered that both the initial low index values and the increase in values after vaccination were useful diagnostic aids.

Foshay (1940) studied the phagocytic test in 3000 individual month by month over several years and considered this partially confirmed its generally useful using the methodology of Huddleston (1933) . The method depends on the skin test. Unusual results included high phagocytic levels in patients with uninterrupted disease, results fluctuationg from very high to very low in asymptomatic recovered patients. Some apparently normal individuals had violent skin reactions, negative SAT and low phagocytic titres indicating possible latent infection. Possibly in consideration of more modern knowledge these results are not so suprising.

Harris (1950) describes a case which illustrates the usefulness of the phragocytic test for monatoring a patients recovery from brucellosis. A patient with symptoms of chronic brucellosis, ie., fatigue, sweating, myalgia, weight loss, shifting arthritia, and headache, no physical findings except slight swealling of one knee joint and low grade fever. SAT was at 1:40. There was mild leukopenia and ESR was at 36mm (Westergen). Culture tests and animal inoculation on blood, urine, and joint synovial fluid were all negative. THe skin test performed after the other tests was positive. The results of the phagocytic test were as follows:

• Stage 1. This phagocytic test carried out prior to skin test and shows modest resistance. In conjunction with the positive skin test and the lo SAT this and the persistant symptoms suggested chronic brucellosis.
• Stage 2. This phagocytic test was taken after a course of 'B abortus vaccine treatment of one month. The increase in phagoctosis indicates increased resistance to Brucella. At this stage the patient felt better with lessening of sweats, arthralgia, low grade fever and fatique and reduce knee joint swelling. ESR was down at 22mm.
• Stage 3. This phagocytic test was carried out after a further 5 weeks of treatment. At this stage the patient considered himself well. The increased resistance showed by the pagocytic test corresponded to clinical inprovement. ESR was down to 14mm. Treatment was discontinued.
• Stage 4. After a further 6 months the patient suffered a relapse with the return of symptoms ans signs including a slight temperature rise, and leukopenia. This is mirrored by the drop in the phagocytic index showing a lessening of Brucella resistance.
• Stage 5. Vaccine treatment was resumed for a further 3 months during which time the patient became symptomless. THe phagocytic index again increased showing strong Brucella resistance. The patient remained well for 6 years and in general the phagocytic index remain high. On one occasion a drop in th phagocytic index was associated with a slight relapse. It was not assumed that because the patient was well that he was cured.

Treatment	Phagocytosis	Phagocytosis	Phagocytosis	Phagocytosis	Index
Stage	Marked	Moderate	Slight	None	Value
	Cells	Cells	Cells	Cells
Stage 1	0	5	9	11	24
Stage 2	5	9	11	o	58
Stage 3	16	9	0	0	91
Stage 4	0	10	14	1	44
Stage 5	19	6	0	0	94

N1

The opsonocytophagic test was first reported by Metchnikoff (1901) and further developed by Hektoen (1906) , Alexandrini (1932), Shattock (1908) , Vetch (1908) and Wright (1904) . the test depends on the ability of white blood cells (neutrophils) to phagocytise Brucella organisms. The test requires the use of a freshly prepared a virulent laboratory strain of smooth B abortus and a fresh sample of blood (not more than 3 hours old) in sodium citrate solution. The technique is difficult but can give interesting results if performed correctly. The 2 solutions are conbined and incubated for 30 minutes at 37.5°C then a sample is transfered to a slide is prepared which is dried and stained with a suitable dye, eg., carbol toluidin blue. The slide is examined and 25 neutroplils are chosen at various parts of the slide and the number of Brucella phagocytised in each. The resulkts for each cell are classified as follows:

• Marked phagoctosis - ≥41 Brucella within the cell.
• Moderate phagocytosis - ≥21≤40 Brucella within the cell.
• Slight phagocytosis - ≥1≤20 Brucella within the cell.
• No phagocytosis - 0 Brucella within the cell.

Phagocytosis text	Brucella engulfed per Cell	Harris's Calculation
Marked	≥41	N1 x 4
Moderate	≥21≤40	N2 x 3
Slight	≥1≤20	N3 x 1
None	0	N4 x 0

Fosley (1945) developed a nomogram for the conversion of bacteria counts into a phagocytic index numbers with index values from 0 to 100. An easier to use method which gives the same results is discussed by Harris (1950) and is shown in the table. The index value is the sum: N₁ x 4 + N₂ x 3 + N₃ x 1 + N₄ x 0.

Harris considered the test as largely quantitative showing the degree of resistance to Brucella infection and should not be seen as positive or negative. to wsdtablish a diagnosis. There is no definite corretation between active brucellosis and the phagocytic index. An patient with acute or chonic brucellosis patient may have may have a high or low index depending on his immunological response. A low index in a chronically ill patient indicates the suppressed state of the immune response and absense of a phagocytosis does not rule ourt infection. The phagocytic index is best used as an aid to dignosis where illness is suspected on the basis of other laboratory and clinical evidence. It is best used to monator a patients response to treatment in association with evidence of clinical inprovement.

A high phagocytic index does not necessarily indicate a lasting recovery only increased resistance to Brucella inrelation to re-infection or exacerbation of a latent infection. Children appear to get well even though they have a low phagocytic index. In a minority of patients an increase in the phagocytic index does not corres-pond with clinical inprovement possibly due to localised infection.

An interesting observation is that other diseases such as Tularemia and vaccines such as cholera vaccine, Eiscle (1947) can increase the phagocytic index for Brucella. A high phagocytic index in an apparently healthy person may indicate an unsuspected or latent infection. This is of possible therapeutic value. A low or negative phargocytic index may indicate susceptibility to Brucella infection.

Harris notes that virulent strains of Brucella are less susceptible to pathocytosis that less virulent strains of the same species.

Passive hemagglutination test.

Renoux (1975) and (1976)

Symptoms and Clinical Diagnosis of Chronic Brucellosis.

General considerations.

Acute Brucellosis.

The most common symptoms of acute brucellosis are: fever, chills, sweats, muscular and joint pain, headaches, cough/dispnea. abdomin pain, general malais and fatigue. Brucellosis may affect virtually any organ of the body. And foci of infection an occur in the liver, spleen, kidneys, testis, ovaries, brain, bone marrow and lungs. Occasional fatal infections usually involve the heart. Manifestations of the illness are protean and there is no specific symptoms by which a charactorisic diagnosis may be obtained. Observable localised complications are most likely to involve the liver, spleen, and joints.

The first description of chronic brucellosis was made by Marston . He called it Mediterannean remittent or gastric remittent fever. He described the symptoms as,"a fever charactorised by the following symptoms and course' a preliminary stage of subacute dyspepsia, anorexia, nausea, headache, feeling of weakness, lassitude, inaptitude for exertion, mental or physical, chills, muscular pain, lastly a fever having a long course, 3 to 10 weeks marked by irregular exacerbations and remissions, great derangements of the assinilative organs, tenderness in the epigastric region, and splenic enlargement. It is prone to relapse, has a protracted convalescence and is frequently marked by rheumatism." The illness was studied in detail by Hughs (1897). He wrote that the illness is an endemic pyrexial disease having a long and indefineable duration, an irregular course with an almost invariable tendency to undulatory pyrexial relapses, charactorised by comstipation, perspiration, neuraglia accompanied by swelling effusions of the joints and other rheumatoid symptoms Various attempts have been made to charactorise chronic brucellosis. He considered that the illness had four forms: malignant, undulatory, intermittent and ambulatory.

Al Dahouk (2003) notes that in general there is little variation between the incidence of symptoms between acute, subacute and chronic brucellosis.She describes the main focal disease as: sacroiliitis, spondylitis, arthritis, endocarditis, meningitis, encephalitis, myelitis, epididymo-orchitis and organ abscesses. Musculoskeletal involvement is common (40%)including: sacreoiliitis, perpheral joint arthritis, spondylitis, osteomyelitis and buritis are described.

Evans (1954) notes that while acute brucellosis can usually be diagnosed readily but milder forms of the illness are left undiagnosed. Initial symptoms may be only a few days of slight fever but cultures may be obtained. Hardy (1930) noted that 0nly 25% of patients were ambulant and physical examination yielded nothing and the only consistent finding was weakness. Martin (1931) found that only 3 of 20 workers with positive serology and symptoms of brucellosis had been identified.

Schirger (1960) in a study of 224 of mixed illness period (<1 year, 113; >1 year, 44; indeterminate, 67. His confirmed brucellosis patients recorded the typical symptoms of serology proved patients and culture proved localised brucellosis patients as follows: fever, 92.6% and 33.3%; chills, 51.9 and 14.3%; sweats, 35.8 and 14.3%; fatigue, 48.1% and 21.4%; weakness, 11.1% and 19,0%; loss of weight, 23.4% and 16.7%; splenomegaly, 14.8 and 16.7%; backache, 8.4% and 2.4%; adenopathy, 18.5% and 16.7%; hepatomegaly, 7.4% and 11.9%; neck rigidity, 1.2% and 0%.

Bierring (1929) analysed 150 brucellosis patients in Iowa in terms of clinial responce. In some cases the onset was charactorised by rigors, chills, profuse sweating, muscular and/or joint pain, loss of weight, and persistent fever. More often it is charactorised by a gradual and insiduous onset with the develpoment of noticable weakness, tiredness particularly noticable in the afternoons, headaches, backache, loss of apetite, digestive distress, constipation. After a few days the patient may notice the onset of chills, slight rigors, and a afternoon temperature rise. At some stage this may be followed by sudden onset of the severe the severe fever symptoms discussed above.

Hardy (1930) noted that becase of the protean symptoms of brucellosis it could be confused with other illnesses. These included: typhoid, malaria, tuberculosis, influenza, acute rheumatic fever, subacute bacterial endocarditis, bronchitis, pyelitis, appendicitis, cholecystitis and tularemia. Often a diagnosis of brucellosis is arrived at after elimination of these other diseases.

Corbel (2006) in a review of 500 patients with acute brucellosis the following appearances of illness: weigh loss, 326; constapation, 234; diarrhoea, 34; cough, 122; rash, 72; ill appearance, 127; pallor, 110; lymphademopathy, 160; splenomegaly; hepathomegaly, 97; jaundice, 6; cardiomurmur, 17.

Spink (1954) considered that endotoxin shed by Brucella was responcible for the inflamatory manifestations in acute brucellosis such as fever. He prepared endotoxin from smooth strains of Brucella and determines its effect on mice and found it to be highly toxic. He found that the endotoxin from B abortus, B suis and B melitensis were of equal virulance and concluded that the greater inflamatory effect of B melitensis was because it was more invasive and produced more bacteria.. Antibiotics did not alleviate the toxic effect but the cortisone (anti-inflamitary) did. He noted that administration of antibiotics often induced an imflamitory responce due the release of endotoxin. He suggested that in mice endotoxin negatively impacted on the adrenal function.

                       Table 1.

Symptom	Hardy 1930	Dalrympole 1960.	Spink 1956	Buchanan 1974	Lulu 1988	Colmenero 1995	Mantur 2007
Sweating %	83	78.7	70.8	93.0	89.0	84.9	3.6
Weakness %	100	76.1	92.5	93.0	75.0	1.3	NR
Fever %	100	NR	97.5	95.0	100	98.1	77.8
Malaise %	NR	71.8	NR	95.0	NR	NR	NR
Chills %	79	35,4	79.3	86.0	69.0	95.0	NR
Anorexia %	74	60.0	72.7	52.0	42.0	NR	3.2
Aches and Pain %	42	54.8	67.7	68	NR	NR	NR
Weight loss %	90	2.5	NR	50.0	13.0	NR	0.8
Abdominal pain %	32	12.1	10.0	NR	16.0	NR	3.2
Constipation %	54	39.3	15.7	NR	8.0	NR	NR
Diarrhea %	NR	5.4	10.7	NR	3.0	NR	NR
Nausea & vomiting %	NR	3.5	5.7	NR	18.0	NR	3.2
Dysuria %	NR	NR	10.7	5.0	5.0	NR	NR
Cough %	45	21.7	21.4	23.9	17.0	NR	3.3
Sore Throat %	NR	17.2	NR	NR	NR	NR	NR
Dyspnea	NR	2.1	NR	NR	1.0	NR	NR
Arthralgia %	32	13.9	37.1	17.0	70.0	73.0	21
Myalgia %	NR	NR	NR	28	38.7	NR
Backache %	48	NR	40.0	23.6	58.0	NR	14
Arthritis %	NR	NR	NR	NR	22.0	NR	NR
Testicular Pain %	NR	1.1	NR	6.0	6.0	4.5	2
Headache %	62	64.8	64.4	44.0	39.0	NR	2.2
Depression %	NR	3.3	34.7	NR	1.0	NR	NR
Nervousness %	50	NR	43.0	NR	0	NR	NR
Insomnia %	34	2.7	36.3	NR	3.0	NR	NR
Irritability %	NR	NR	1.4	NR	NR	NR	NR
Neck pain %	29	NR	33.0	NR	1.2	NR	Nr
Visual disturbance %	NR	5.4	7.4	2.0	NR	NR	NR
Dizziness %	NR	NR	NR	2.0	NR	NR	NR
Tinnitus %	NR	NR	NR	1.0	NR	NR	NR
Nose bleed %	NR	8.3	NR	NR	NR	2.3	NR
Skin rash %	Nr	8.4	3.3	NR	NR	3.4	NR
Splenomegaly %	33	22.2	40.5	10.0	19.0	22.3	17.2
Lymphadenopathy %	NR	NR	38.8	14.0	9.0	NR	3.1
Hepatomegaly %	NR	6.0	20.6	38.1	8.0	NR	10.0
Cardiac abnrm %	Nr	NR	5.7	NR	NR	21.1	NR

THe data in Table 1 is from the following sources: Hardy (1930) reviewed the cases of 300 brucellosis patients in Iowa, USA (mainly B abortus and B suis. Dalrymple-Champney (1960) described the symptoms of brucellosis, mainly caused by B abortus in the UK. Spink (1956) reports signs ans symptoms of brucellosis in 121 patients in Minnesota, USA (mainly B abortus). Buchanan (1974) investigated brucellosis in 100 abbattoir works in Iowa, USA mainly caused by B suis. Lulu (1988) reviewed 400 cases of brucellosis in Kuwait. Colmenero (1996) reviewed 539 cases of acute brucellosis mellitensis in Spain, including signs and symptoms. Mantur (2007) describes the signs ans symptoms of illness in 740 patients with brucellosis (mainly B melitensis in India.

There have been many useful studies of the general presentations of brucellosis including Ariza et al , Bosilkovski et al , Colmenero JD, et al, Geyik MF et al , Lulu AR et al .

Hardy (1930) reviewed these as they applied to B suis and B abortus as they appled in Iowa. he noted that more severe forms of the illness occurred in occurred in B melitensis.

• The intermitent type was most copmmon in his patients. The charactoristics were: insidious onset, increasing weariness during day, general aching, headaches, lack of appetite, evening chills, moderate insomnia, possible fever, backache, stiffness, neck and joint pain, constipation, possible loss of weight, possible cough sometimes persistent, the development of drenching night sweats, limited occurancew of physical abnormalities which might include anemia, weakness, loss of weight, palpable spleen, tender abdomin and slight temperature rise in evening. The severity of the illness was variable and the duration was up to 4 months and required bed rest.

• The ambulatory type was similar to the intermittent type but less severe. The symptoms were the same but less pronounced including weariness. Occasional physical abnormalities were a palpable spleen and evening temperature rise. Duration was up to 4 months. The patients could often continue work.

• The undulant type which is charactorised by relapses and a wave like temperature curve was rare in this series and is more common in B melitensis. The onset was charactorised by weakness, general aching, headaches, and anorexia and was often confused with flue. A series of relapses occurred with development of further symptoms of insomnia, constipation and night sweats.The temperature increased in a stepwise manner dayu by day. Morning remission of symptoms did not occur. Either this process was repeated one or more times but usually there was a decrease in intensity and duration over time.

• The malignant type was rare in B suis and B abortus of this series. The course of the illness was a sudden acute onset with a very high and sustained temperature, general prostration, severe headache, backache, marked anorexia and true rigors often leading to death. It shoulkd be noted that overlap between these forms occurred. THe sdymptoms noted in Hardy's patients were as follows: weakness, 100%; sweating, 84%; chils. 77%; rigors, 37%; general aches, 42%; headache, 62%; bachache, 47%; neck pain, 29%; joint pain, 32%; anorexia, 74%; constipation, 54%. Signs of illness were: fever, 100%; loss of weight, 90%; palpable spleen, 33%; abdominal tenterness, 20%; leucopenia, 60%. Weakness was the most pcommon and persistent symptom. In milder cases it only occurred in the afternoon onwards. Sweating occurred during sleep and was generally of shoret duration but could be prolonged and drenching. Chills were associated with temperature rise. General aches are absent in bed but seemed to be aggravated by exercise and are described as similar to muscular sourness due to over-exerxise of that occurring with flu. Headaches are described as associated with fever and more severe in the evenings, frontal and possibly associated with pain in the eyes. Back pain mainly affects the lumber region and is aggravated by exercise and can be persistent. Pain in the back of the neck presented as stiffness and could be severe and the first indicator of illness. In intense cases this could lead to menigitis. Joint pain was usually mild. Abdomin pain could be continuous or cramping. It often occurred in the epigastrium of the right lower quadrant and could be confused with other conditios (eg appendisis). Nausea and vomiting could occur. Lack of appetite occurred in the afternoon associated with fever and anorexia.

The following typical example is that by Irmali in 1959 Types of chronic Br are:

• Insiduous course. It suggested by Young that up to 50% of cases start in an insiduous way. This can mean Combined with confusing symptoms this can mean that the early opportunity to diagnose the illness by serology tests may be missed.
• One or more recurrent attack following an acute attack. In this case there is a reasonable opportunity to get a positive result from serology tests.
• Deseased with localised involvement. In these cases there may be an obvious foci of infrection such as a cyst or swollen joint. In some cases recovery of infected body fluid of tissue may give a positive brucella culture or antibody test.
• Desease with typical symptoms and history but no obvious physical findings and low or negative serology. There are no physical findings in 85% of cases.

Chronic Brucellosis.

The symptoms of chronic brucellosis are similar but less defined. General recorded symptoms include: malaise, excess sweating (at night and on exercise), irritability, depression, rheumatism, arthritis/joint pain, lower back pain, and buritis or synovitis, heachache, tinnitus, dyspepsia, lassitude, insomnia, vague pains, and sexual impotency.

Symptoms	Calder 1939	Manchester 1941	Darley 1947	McDevitt 1973	Pourbagher 2005
Sweating %	49	21.1	23	19.5	45.9
Weakness %	89	50.0	80	14.6	57.7
Fever %	85	13.4	34	-	27.0
Chills %	51	26.3	13	-	-
Malaise %	-	-	-	20.7	-
Aches & Pains %	66	60.5	22	-	96.4
Abdominal Pain %	48	-	12	12.2	23.4
Constipation %	57	18.4	19	4.9	-
Diarrhea %	-	-	5	6	5.4
Indigestion %	-	44.7	30	-	-
Nausea %	-	-	-	-	20.7
Dysuria %	-	-	13	-	-
Back Pain %	53	-	24	25.6	63.1
Neck Pain %	57	-	7	-	-
Arthritis %	-	-	-	11.0	-
Rheumatism	49	-	-	15.9	-
Headache %	69	26.3	47	17	47.7
Depression %	-	-	23	13	9.9
Insomnia %	54	-	23	11.0	-
Anxiety %	-	-	19	-	26.1
Irritability %	-	-	26	22.0	-
Nasal congest %	-	-	31	-	-
Paresthesia %	-	-	12	-	-
Dizziness %	18	-	24	-	-
Palpitations %	-	26.3	20	-	26.1
Chest Pain %	-	26.3	11	-	14.4

The sources of data in the table are as follows: Calder (1939) reviewed the symptoms of 550 patients with chronic brucellosis (including soime with acute brucellosis).Darley (1973) corelated data for 74 chronic brucellosis patients whose cases were proven by Brucella antigen skin tests. McDevitt (1973) corelated symptoms of chronic brucellosis (mainly B abortus) in 82 vetinerians in contact with infected cattle for the UK. Manchester (1942) carried out a survey of the most common symptoms of chronic brucellosisin in 38 patients. Pourbagher (2005) studied 251 patients with brucellosis in Turkey including 111 patients with chronic brucellosis.

Hardy (1930) attempted a follow up of his patients. Although they had recovered from acute brucellosis a great many complained of prolonged debilitating illness. Symtoms included weakness, easy fatigue, fever, stiffnessor pain in muscles or joints, headache, backache, general aching,anorexia, palpitations and sweating. Some had relapses less severe than the first acute attack.

It will be shown in a later that chronic Br can seldom be diagnosed by laboratory serology tests which frequently are of little use. This means that physicians must rely on the patients pasted history of illness and their existing symptoms. Araj (1997) considered that in the absence of usable serology data chronic Br can be diagnosed on the basis of previous illness and current symptoms. Hewitt (1984) noted that the organism is notorious for causing relapses and produces chronic Br which may exist over a long period of time and give rise to confusing symptoms and that no laboratory analysis allows unequivical statement and clinical examination must be used. Young (1989) notes that in the diagnosis of Br there is no substitue for a detailed clinical history, Physicians must listen to patients.

Henderson (1975) in a review of brucellosis in vetinerary surgeons with chronic brucellosis abortus considered that the soundest way of assessing ill health which had possibly been caused by brucellosis seemed to be thorough clinical examination and disregard of serology findings. Interpretating results by the usual serology tests in the absence of a clinical examination is probably of doubtful value and may even be misleading. This was on the basis of detailed clinical and serological examination of 46 vets.

Foley (1968) reviewed the health of 80 veterinary surgeons in SW Eire who were exposed to B abortus and found that 72.5% had symptoms consistent with brucellosis: erythematous rash from cattle midwifery contact, 37 cases; lower back pain, 22 cases; malaise, 22; sweating, 21 cases; irritability, 20 cases; weakness, 19 cases; rheumatism and arthritis, 18 cases; depression, 14 cases; headaches, 13 cases. Most had elevated AGH test results. Most considered themselves as suffering from vague and unpleasant ill-health.

Gauss (1950) reviewed chronic illness in 20 patients and considered gastrointestinal a main symptom. Other corelated symptoms were: fatigue, 8; weakness, 8; low grade fever, 2; dizziness/giddiness, 6; restlessness, 2; irritability, 3; dry throat, 1; muscle pain weakness, 4; joint pain, 7; palpitations, 7; tachycardia, 1; headaches, 7; insomnia, 2; weight loss, 2; backache, 4; chest pain, 1; cheast pain, 1.

Johnson (1954) notes that the main form of chronic brucellosis is reticulo-endothelial but other organs are involved. these include: ophthalmic changes, neuritis, arthralgia and myalgia, central nervious system problems, Typically iis a chronic inflammatory process or similar to tuberculosis.

Forbus ((1943) referred to gross and microscopic leision formation in chronic brucellosis as follows: "The basic reaction is a progressive proliferation of the large mononeuclear cells of the reticulocendothelial system accompanied by the exudation of finrin and sometimes by hemorrhage. This is followed by coagulative necrosis and finally by the proliferation of fibroblasts or the formation of a dense scar composed of reticulum."

Descriptions of Chronic Brucellosis.

• Simpson (1930) reviewed the symptoms of 90 patients in Dayton, Ohio. He notes that during febrile episodes the pulse rate can either be is disproprtionally slow or elevated. Nocturnal febrile episodes are also marked by restlessness and insomnia. He notes that unlike typhoid patients remain clear thinking. His observation is that the tongue is often coated with a yellow fur possibly with a fetid smell. He notes that of the 90 patients he studied 50% did not appear ill. The spleen was palpible in 30. He also noted that a similar sweetish fetid odor was associated with drenching sweats.

• Evans (1937) notes that chronic brucellosis is very difficult to diagnose Typically physical signs of illness, temperature curve, and postive SAT are negative throughout the illness. She had contact with many patients and notes that they typically have a long illness with no organic abnormalities and are diagnosed first with neurasthenia and finally with a brucellosis. She notes the case of a patient who was assessed in 3 hospitals and a leading teaching hospital during a year of illness none of them diagnosed the illness. During a separate operation Italic textBrucella was cultured from a focus of infection.

• Simpson (1941) notes that whereas acute brucellosis is relatively easy to diagnose the protean manifestations of chronic brucellosis make its clinical diagnosis very difficult. On the basis of studies by Evans (1934) (1938) , Poston, Angle (1939) (1942) , Scoville, Thames (1935) (1935) (1939) , Calder (1939) , Hamman, Harris (1941) and Wells he asserts that chronic brucellosis can exist without: a significant temperature curve, physical signs of illness, positive SAT titre and/or skin tests for long periods of time. Also many patients do not go through a acute febrile illness; in his experience this was less than 10% (most of his patients suffered from B suis and B abortus - the milder forms. Nevertheless chronic brucellosis is a debilitating illness that often prevents the patient from working. He considered at the time of writing that there was a significant epidemic of brucellosis in the USA. He notes that this lack of evidence for brucellosis has almost inevidently lead some doctors to diagnose neurasthenia or psychoneurosis. He notes that McGinty has listed 150 different manifestations of chronic brucellosis but considered that there are 3 cardinal features in most cases: weakness, low-grade fever andlack of objective physical findings. Mild fever may come and go. It may be present or absent for weeks or months but it can also be absent for months but sudden exacerbations may occur with either accentuation of previous symptoms or new symptoms affecting the respiratory, cardiovascular, genito-urinsary, gastrointestinal, skeleton or nervous systems. Possible complications are: pneumonia,endocarditis, orchitis, epididymitis, prostatitis, oophoritis, cholecystitis, hydarthrosis, arthritis, spondylitis, osteomyelitis, ocular complications, or meningoencephalitis. Although these complications are usually associated with acute brucellosis he considers that they are more often likely to occur in the chronic phase of the illness. Stuies of chronic brucellosis indicate that symptoms related to the central nervous system are almost universal in chronic brucellosis. He notes that endotoxins released by Brucella have a toxic impact on the nervous system. He also notes that Brucella has been recovered from the blood, urine, bile andextirpated tissus of patients with long term (years) low-grade fever. He considered that the death rate in brucellosis was about 2% associated with a overwhelming acute attacks but he noted that it could occur after a relapse at any stage of the illness probable related to localised infection including the meninges, brain, heart valve or lungs.

• Griggs (1943) studied 100 patients with chronic brucellosis from 1 to 57 years duration, mostly 3-20 years. He considered that there were 2 main varients: undulant course (76%)and remittent course (22%). Typical illness history includes: long periods of mild illness with frequent more intense attacks in which the diagnoses are puzzling, varied and atypical; persistent low grade fever; flu-like symptoms once or several times per year. Freequently a wrong diagnosis of neurosis has been made. Typical symptoms are: ease of fatigue, lack of energy, weakness and lethargy affecting the chronic course and attacks; aching or painful muscles, nerves, joints and bones; headache often with light-headedness during attacks; catarrh; digestive system disorders, usually with excess gas; fever with slight temperature rise to 99°F. Less common symptoms include: tachycardia, palpitations, excess perspiration, and epistaxis. He notes that patients find it difficult to relate their symptoms because apart from about 3 more significant symptoms they have many more minor symptoms the relating of which makes him/her appear to be neurotic. Such symptoms are: are varied, often seemingly unconnected and inconsistent and they are usually of long duration.

• Darley (1947) attempted to classify chronic brucellosis cases using the interdermal skin test (IDST) which he calls sensitizatio.. He noted that clinically in proved chronic cases that subjective findings were multiple and variable but that objective symptoms were conspicuous by their absence. Also that culture tests were rarely positive in chronic brucellosis and the serum agglutination test is of little value but the IDST is most useful in chronic brucellosis because it gives a higher level of positives. The disadvantages of the test are: it does not distinguish between past and present infection, occasional produces false positives and can give negative results while SAT results are positive (in this series, 7.26%). From a large group tested by IDST 74 were selected on the basis of positive IDST and typical symptoms and compared to controls. Darley demonstated that there was no corelation between other allergic conditions (such as hay fever) or psychoneurosis (?) and positive Brucella IDST results but he showed a statistical corelation of incidence of atypical rheumatoid complaints (joint pain, swelling, etc) to positive IDST results. In the 74 patients taken as having chronic brucellosis, the most typical symptoms were fatigue, muscle and joint aches, non-migraine headaches, digestive complaints, and low-grade fever of which muscle and joint pain, fever, and sweating were considered statistically significant compared to (so called) IDST negative psychoneurotic patients. He concluded that positive IDST coccured far more often in chronically ill psatients than in the well and this seemed to indicate that Brucella infection was frequently involved in chronic ill health.

• Gauss (1950) notes that chronic brucellosis is difficult to diagnose. there is no pathognomonic signs while clinical signs may be variable and bizarre. He notes the similar diagnosis of the dihonourable neurasthenia. Harris (1937) notes the simularity to tuberculosis: fatigue, loss of weight, seats, cough, slight elevation of temperature, mild chest pains. Gaus notes that others described the disease as simulating neurasthenia, psychoneurosis, bronchial asthma, early tuberculosis, influenza, anemia, syphilis, peptic ulcer, cholecystitis, appendicitis, ulcerative colitis, arthritis, rheumatism, typhoid, endocarditis, coronary disease, functional heart disorder, bachache, and skin disorders. He concludes that brucellosis has a low mortality but a high morbidity with many distressing symptoms. He notes muted response of the human body the simularity to tuberculosis and suggests that the term "cure" should be replaced with "quiescent" and "arrested".

• Barrett (1953) noted the difficulties faced by chronic brucellosis sufferers. Patients usually look well but complain of relatively severe symptons. Their illness is then often construed as psychosomatic and they are often treated as neurotic, some are even referred to psychiatric clinics. He cites the case of a 55 year old woman ill for 10 years who had previously been cheerful and energetic. Her doctor described her as neurotic and a hypochodriac. Her symptoms were: night sweats, fatique, diffuae rheumatism, constipaation, dizziness, minor syncopal attacks, and paraesthesiae of the hands.5 episodes of haematuria with associated dysuria and micturition up to 10 times per day without apparent evidence of urinary infection or related problems. The only physical abnormality was a slightly tender and moderately swollen liver. SAT was negative but a skin test was strongly possitive for Bruceella. The patient responded well to treatment.He also described the illness of a 45 year old verinarian with a skin infection due to chronic brucellosis for 15 years. He was first diganosed by SAT tire. He developed roughened skin and warty excrescence of the hands. A biopsy of the infected area showed an area of hyperplastic squamous epithelium overlating a dermis markedly infiltrated with chronic inflammatory cells some of which had the form of non-caseating granulomata. He gradually developed a general dermitis and increase severity of symptoms including lassitude and malaise. He was treated with tetracycline and vaccine and initially this cause a severe exaccebation of his symptoms however he the gradually inproved in health including resolution of the skin condition but he had a limiterd relapse after 4 months.

Barrett (1953) studied 25 patients with chronic brucellosis (most likely B abortus) and an average duration of illness of 6,5 years. Their described general symptoms were: sweating, 24; lassitude, 17; chills or rigors, 13; joint pain, 9; diffuse rheumatism, 23; abdominal pain, 11; hepatic enlargement, 15; splenic enlargement, 9. Of the 25 patients 11 had low or negative SAT titres. 12 patients were tested by liver biopsy for histological appearance. Of these 10 had muliple granulomata and 2 focal infiltration. In each case there was an abnormal histological appearance. Of these 10 had muliple granulomata and 2 focal infiltration. Of the patients with multiple granuloma 5 had low or negative SAT titres. The most charactorisic leision was a clearly delineated nodular granuloma (resembling a non-caseating miliary tubercie) found anywhere within the liver parenchyma. The size of the leisions was 100-180μ. Other leisions were diffuse. THe leisions were comprised mainly of epithelioid cell type although eosinophils, lymphpcytes, plasma cells and occasion and sometimes polymorphonuclear and giant cell were found. There was a surrounding was a fibrous network. Although Brucella could not be grown from the samples it was considered that the was due to Brucella infection.

• Foggitt (1953) notes that the symptoms of brucellosis vary greatly in severity. He suggests that rheumatic pains, back cheat or limb stiffness, increased sweating especially at night, fatiguability, pain or swelling in knees, ancles and other joints, headaches, chills, fever, gastro-interal symptoms such as constipation or appetite. Symptoms similar to neuroses occur, Patients with chronic brucellosis do not usually appear ill ever with severe symptoms. It is essential that doctors ask relevent questions since the patients own description may be vague. Severe symptoms can simulate other illnesses such as: angina pectoris (possibly due to sensory nerve involvement), rheumatic arthritis, appendicitis, cholecystitis, sciatica,and lumbago.

• Foggitt (1953) was a sufferer of chronic brucellosis and offers a useful insight into symptoms. He notes that background symptoms of chronic brucellosis are: muscle weakness and easy fatiguability and general lassitude; muscilar and joint pain; sweats and chills; abdominal symptoms chiefly of pain and flatulance; neurological symptoms. The general lassitute associated with the illness often leads to lack of exercise and therefore weight gain. Patients often look well yet complain of ilness which seems out of proportion to their appearance. On the basis of this they are often labelled neurotic or neurasthenic. A typical symptom is general stiffness particularly in the morning. This like the after effects of a hard days work the privious day, ie similar to a well person who has over-exercised. He observes that muscular weakness is present at some times but not at others and comes on suddenly and it is usually worst in the evening but bears no relation to the amount of work done. Individual muscles or sets tire more rapidly than in a health person. Patients find standing for any length of time difficult. Muscle and joint pain are usually transient. Sweating was less pronounced in chronic brucellosis but increased sweating was associated with bouts of illness.

• McDevitt (1973) in a review considered that the most common symptoms of chronic Br are malaise, sweating, irritability, depression, rheumatism, arthritis and backache. Harami (2005) notes that typical symptoms are persistent fatigue, malaise and insomnia, recurrent depression, vague pains and inertia and sexual impotency. Typically recurrent flu like lassitude, pain and sweats and frontal and retro headaches. Low back pain without spondylitis, limb pain or stiffness usually muscular, sweats usually at night with normal temperature. According to McDermitt (1973) typical symptoms of chronic brucellosis are: sweating, lassitude, malaise, irritability, depression, insomnia, rhematism, arthritis and bachache. Williams (1982) considered that typical symptoms of chronic br are: headaches, lassitude, low back pain with no X-ray findings, limb pain (often muscular), sweating on effort and at night, anxiety, depression.

• Williams (1973) gives an excellent account of the symptoms of chronic brucellosis mainly relating to B abortus@. The onset of chronic brucellosis may be insidious or it may follow an acute attack which was untreated or inadequately treated. It may take the forn of an acute relapse long after a supposed cure. Typical symptoms are flu-like attacks, lassitude, headache, pain, and sweats. Lassitude can be mild or diablingly severe. Sufferers who continue to work may rneed to rest more often than before and even sleep in the afternoon. Low back pain without X-ray evidence is common. Headaches are frontal or retro-orbital. Limb pain is more often muscular than articular and are more proximal than distal. Sweats occur on physical effort and at night and the frequency is an indication of the severity of illness. Suicidal depression is rare in farmers can be affected by related financial problems. Momentary dizziness may be associated with bouts of sweating and fatigue but may occur alone in a strange and frightening way unrelated to other symptoms. The patients appearance may reflect illness often with weight loss or he may look normally well. Moderate splenomegaly occurs in about 10% of patients but is not related to severity of illness.

• Manchester (1942) has recorded a typicalm bout of recurrent illness in chronic brucellosis. This starts either with an insidious increase in fatigue, muscle soreness and joint pain or an abrupt onset. The second phase involves chills, malaise, headache, backache, fever, anorexia, sweating, night sweats, and sometimes gastrointeral complaints, joint pain, muscle soreness. Rhinorrhea, cough and sore throat can also occur during this phase. The third phase is gradual convalescence. His patients called this the "Grippe". He carried out a survey of the most common symptoms in 38 chronic brucellosis. These were: fatigue, 19; temperature rise, 10; headaches, 10; recurrent attacks, 10; night sweats, 8; cough, 4; sore throat, 5; chest pain (angina pectoris0, 10; coronary occlusion, 4; joint/muscle pain, 23; indigestion, 17; constipation, 7.

• Pourbagher (2005) studied 251 patients with brucellosis in Turkey. Of these 111 patients had chronic brucellosis. The general clinical symptoms reported for these patients were as follows: arthralgia, 107; fatigue, 64; back pain, 70; fever, 30; sweating, 51; headache, 53; palpitations, 29; appetite loss, 23; nausea, 23; abdominal pain, 26; anxiety, 29; weight loss, 13; cutaneous; chest pain, 16; diarrhea, 6; depression, 11; cough, 14.

Corbel (2006) notes that susceptibility to brucellosis in humane depends on a number of effects: the response of the immune system, route of infection, size of inoculum and the species of Brucella. Although 'B melitensis and B suis generally cause more severe illness but the illness in B abortus and B canis can sometimes also be severe. Possible modes of direct infection are: throuhg cuts in skin, via the conjuncival sac, aeroso; inhalation, injestion of contaminated food (milk etc), blood transfusion, tissue transplant, sexual transmission. The latter 3 are rare.

Physical Signs of Illness.

General.

Hardy (1930) recorded signs of illness in 300 acute B abortus and B suis patients. He noted that in most cases the patients seen in bed did not look or seem physically or mentally ill but they could be pale and look tired. Possible signes of illness were a coated tongue, Io% had rales indicative of bronchitis, abdominal tenderness either difuse or localised in the right upper or lower quadrant or less often the left upper, 30% had a palpable spleen without marked enlargement, rose spot skin rash, muscle soreness. He notes that most likely no physical signs of illness are present except fever. He also noted that in brucellosis pulse is not directly related to ever and usually remained in the normal range. Some patients had a slow pulse similar to typhoid or occasionally a unduly rapid pulse occurred.

Barrett (1953) noted that physical examination was never diagnostic in chronic brucellosis. In this series careful palpitation of the liver indicated slight enlasrgement in16 0f 25 patients.

Simpson (1930) notes that brucellosis can be easily confused with typhoid fever, malaria, tuberculosis and influenza.

Young (1991) notes that onset is insieous in about 50 of cases. Typical symptoms are fever, sweats, anorexia, fatigue, weight loss and depression. There is a paucity of physical findings, of which the most usual are fever, mild lymphadenopathy and occasional heptosplenomegaly.

Temperature, Fever, Sweating.

Temperature rise is a typical symptom of acute brucellosis as noted by Mousa ) but in chronic brucellosis fevrile temperature rises are usually absent however chills and sweats are typical symptoms. Hatara (2005) notes that in chronic brucellosis sweats occur usually on effort or at night but with normal temperatue. Williams noted that hyperopyrexia is rare and Rust notes that lipopolysaccarides from brucella are less pyrogenic than other virulent gram negative organisms. Gotazza (1998) notes that patients with chronic brucellosis often have a sense of fever which generally cannot be determined. Begger (2005) notes that in acute and chronic brucellosis patients fever occurs in 37% and 20% of cases respectfully. Anernathy and Spink studied the effect of iv injections of brucella endotoxin (lipopolysacchride) on acute and inactive (presumably subclinical and chronic) and controls by measuring the rise in temperature. Typical effects were fever with temperature rise, headaches, myalgia and chills and fatigue which are typical symptoms of the illness. High doses of endotoxin produced response in all patients. At low doses produced significant response in acute patients slight response in inactive patients and less in controls. Repeat treatment lead to tolerance of the endotoxin and reduced response. This seems to explain the lact of temperature reponse in chronic brucellosis.

Evans (1954) notes that although brucellosis is alternativerly called undulent fever becouse of flucuating temperates, but this only applies in acute illnes. In chronic brucellosis the temperature is usually normal. Intermittent rises may occur, often months apart, and temperature rise is limited to 101°F. In such circumstances a temperature rise in late afternoon or evening is most likely.

Manchester (1942) noted that in some chronic brucellosis patients a small temperature rise to 99-100°F could occur in the afternoon.

Craig (1903 and 1906) noted that the low, uncharactoristic temperature curve in chronic brucellosis was a hinderence to diagnosis and could cause misdiagnosis.

Griggs (1943) notes that active brucellosis can occur without fever. However he suggests that if great care is taken then slight temperature rises can be detected in some cases of chronic brucellosis. This involved the use of certified thermometers, a 5 minute test duration, determining the best time of day for the test and a week of tests, Using these methods he tested 97 patients. Of these: 74 had fever and 23 did not. In the patients with temperature rise the recorded levels were: ≤99°F. 13 patients; 99-100°F, 39 patients; >100°F, 6 patients; unspecified or occasional fever, 16 patients. On the basis of this study of 100 patients he concluded that apart from temperature, absense of physical signs of illness is typical of chronic brucellosis. In this 100 patients: an enlarged liver in 1 case and 2 cases of enlarged spleen, 5 patients showed desructive boneleisons by X-ray.

Hardy (1930) as a part of a study of 300 patients with acute in Iowa reviewed the temperature in terms of the classificatons: malignant type, undulant type, intermittent type and ambulatory type. In the malignant type the temperature can rise steadily up to 106°F and can be associated with death. Such case are relatively rare. In the undulant form, the temperature rises and falls in a fairly regular cycle with a temperature rise to about 102 - 103°F. Usually the frequency and extent of temperature diminishes over time. The intermittent type is similar to the undulant type but temperature are less often. In the ambulant type, temperasture rise is less and typical cycles are not present. It is suggested that over-exersion can trigger cycles of temperature rises.

Barrett (1953) Barrett GM, Rickard AG. Chronic Bruellosis. Quart J Med. Jan 1953. 22(85). 23-42.</ref> noted that during the hospitalisation of 11 patients with chronic brucellosis a low grade pyrexia occurred but temperatures did nort exceed 37.5°C.

Localized and Focal Infection.

Localised brucellosis in general.

Brucellosis can occur as a localised or focal illness with or without general symptoms. Almost anyy part of the body may be infected in this way. Such localised illness my be obvious such at a joint of difficult to identify such as in an internal organ. Sometimes localised brucellosis occurs without any positive serology or similar tests. This is particularly true in chronic brucellosis. This discussed below under tests. For example Jordon (1947) describes 2 cases of localised brucellosis. He describes a boy of 9 suffering from brucellosis with repeated bouts of tonsilitis who recovered after tonsillectomy and a women who recovered from brucellosis after removal of a ovarian cyst from which B suis was recovered.

Skin and Muscular invovement in Brucellosis.

Rotes-Querol (1956) reviewed 174 patients with B melitensis in Spain. He noted that skin and muscular involvement was limited. His findings were: myalgia, 11 cases, suncutaneous nodule, i case; celluialgia, 1 case; suppurative cellulitis, 1 case; buritis, 3 cases; tendinous synovitis in fingers, 1 case; primative Brucella myositis, o cases; cellulitis, 0 cases. The subcutaneous nodule was on the tibia was shown by biopsy to be young connective tissue and many round brucellar granula like cells. 2 cases of buritis involved the olecranan and 1 the trchanter and in all cases were surrounded by an area of cellulitis. A biopsy showed a diffuse inflamation with microscopic abscenses. He considered that the interaction of B melitenis with sofr tissue could be subdiveided as: myalgias, inflamation of tissue with the formation of granulation tissue and formation of micro-absesses (possible suppurative) such as in the tissue of bursas or cellulits.

Hardy (1930) noted that skin manifestations included scattered maculae and possibly red spots.

Corbel (2006) notes that a variety of skin coditions including: rashes, nodules, papules, erythema nodosum, petechiae and purpura. Cutaneous ulcers, abscesses ans suppurative lymphangitis occurs mainlyy in B suis.

Simpson (1930) reviewed symptoms of skin in volvement in 90 cases of brucellosis (mainly B abortus. 10 patients had skin leisions. Generally these were small macules but in one casethere was a maculo-papular eruption. In 3 cases the leisions occured on the abdimin as roseola similar to typhoid fever.

Bursitis.

Bursitis caused by Brucella is a significant but less common symptom of brucellosis. Kennedy (1904) was the first to recognise bursitis and synovitis as a symptom of brucellosis. In the case of a subdeltoid bursitis Brucella was isollated from the bursal aspiratations. Harris (1941) described 3 cases.

Traboulsi (2007) review the case of a woman with a left knee swelling of a years duration. She had a history of chronic lower back pain but no other symptoms. An aspirate from the knee cultured B melitensis. The size of the bursa was reduced by antibiotic treatment. Other cases have been discussed in the literature. Three cases are described by Johnson (1954) , 1 case affecting the knee by McDermitt (1994) and 1 case involving the shoulder by Guiral (1999) . Mousa (1987) noted that of 169 patients with osteoarthritic complications of brucellosis 1.2% had tendinitis and bursitis. Schirger (1960) notes that prepatellar burtitis is not uncommon. It is most probable that bursitis is more common in chronic brucellosis.

Baker (1928) reports a patient with intermittant hydrarthrosis and general debilitating illness for 9 months. Brucella was cultured from the liquid from both knees. SAT was negative.

Davis (1957) studies 11 patients with chronic brucellosis and found 1 with bilateral hydrarthrosis of the knee joints without destructive changes.

Johnson (1954) reviewed 4 cases of bursitis caused by Brucella. They considered bursitis as a locallised form of the illness.

• Case 1 was a sixty-three year old farmer with a swollen left knee of 6 years duration. The bursa had appeared without trauma, pain, local heat or redness. Repeated aspiration of the bursa produced a greyish fluid. 4 months prior to treatment he had a flu-like illness after which the swelling increased with soreness and local heat. Movement of the knee was not affected. SAT was 1:1600. X-rayss indicated a soft anterior swelling but the knee joint was normal. The bursa wasa excised and B abortus was recovered from the operation tissue. He was treated with antibiotics. The cross-section of the bursa indicated a thick wall and fibrin filled cavity.
• Case 2 was a sixty year old farmer who had had an arterior swelling on the right knee for 8 months. The bursa had developed without pain without redness or heat. The busra had been aspirated 6 times with a yellow liquid removed. He had suffered from a productive cough for 2 years. Movement of the knee was not affected. The bursa was excised and B abortus was cultures fron the tissue. SAT was 1:400. The patient was treared with antibiotics.
• Case 3 was a 40 year old farmer with a bursa on the left knee of 14 years duration which had started with pain in the knee. The knee had been aspirated on several occasions. Over a period of time the mass decreased in flucuance. The function of the knee was normal. X-ray indicated the knee bone was normal. THe bursa was excised. The fibrotic wall was densely fibrotic and inflamed with 0.4 cm wall thickness. The cavity was filled with fibrin. A culture for B abortus on the excised tissue was positive. Blood culture was nrgative.
• Case 4 was a 36 year old farmer with a swelling on the right knee of 3 years duration. Initially the swelling was intermittent but then latter permanent. Multiple aspiration was ineffective with the swelling rapidly returning. There was no pain or reduced knee movement. There had been previous pyelonephritis illness but SAT was negative. X-ray indicated the knee was normal. The bursa was removed. The sac was filled with yellowish fluid and finrin. Culture of the specimen removed was positive for B abortus . The SAT was positive at 1:40.

Johnson noted tat fibrosis was present in all these cases and that this was most likely inpenetrable to antibiotics therefore surgical removal is advisable. They also note that whereas aspitate often gives a negative Brucella culture a test on tissue from the same site is often positive.

Foggitt (1953) noted that in chronic brucellosis joints are often painful and swollen and that buritis can occur. He refers to a case of "housemaids knee" which proved on excision and testing to be caused by Brucella.

McDermitt (1994) describe the case of a 39 year old man with an infected skin leison on his left knee of 7 years duration. At initial contact the leision was treated with flucloxacillin but the leison recurred in 3 months with associated fever and inguinal adenopathy. Brucellosis was tentitively diagnosed on the basis of a AHG teat of 1:5120 however the nature of the skin leision was not determined. After 3 a further 3 years the a discharging sinus was formed. associated with granulated tissue with adjacent chronic inflammation and necrosis. A diagnosis of an inflamed bursa was given. After a further 4 years a new purulent discharging sinus was present with redness, swelling and tenderness. There was no limitation to knee movement throughout the illness. The patient had a SAT of 1:1280 and AHG test of 1:5129. The bursa was surgically drained and resectioned. Culture of resectioned tissue yeilded B abortus biotype 1. THe purulent discharge failed to culture Brucella. The bursa consisted of a chronically inflamed hyperplastic synovium with surface fibrinous exudate with poorly formed necrotising granulomata in the surrounding connective tissue.

Pourbagher (2005) used sonography to determined the extent of buritis in 251 brucellosis Turkish patients of whom 212 patients had arthalgia. 13 (5.2) patients had buritis. The joints affected were: suprapatellar, 6 patients; olecranon bursa, 3 patients; subacromial bursa, 3 patients; lateral malleolus, 1 patient. Culture of synovial fluid aspirated from all sites was negative.

Osteoarthritic symptoms

General.

McDermitt (1994) notes that arthritis occurs in 10-25% of patients and may predede, accompany or follow systemic infection. The frequency of manifestations is: sacroilitis> peripheral arthritis (usually monoarthritis affecting the knee, hip, ankle and shoulder)> mixed forms> sponylitis (the most frequent destructivw form). Arthritis is associated with acute, undulant and chronic forms. Spondylitis is strongly associated with chronic illness and tends to occur in elderly patients. Sacroilitis and perpheral arthritis are more frequently acute or subacute in children and young adults.

Rotes-Querol (1956) reviewed 174 patients with B melitensis in Spain. Of the 174 patients 148 (85%) had indications of osteo-arthicular involvement. He noted that others: Cantaloube (1911), Ruiz-Castaneda (1954) , Betoulieres (1948) and Rimbaud (1947) had found similar results. He noted that these were higher percentages than those found by Hardy (1929) and (1937) and Harris (1950) for B suis. Likewise results by Hardy (1929) , Dalrymple-Champneys (1935) and Sylvest (1951) showed much lower results in B abortus.

Gotuzza, (1982) notes that in a study of 304 patients that osteoarthicular is the most common complication in brucellosis affecting 1 in 3 patients. Peripheral joints are most often affected typically the knee or hip. Williams (1982) noted that in Br suppurative arthritus affects typically the weight bearing joints and buritis or synovitis is the likely cause of a painful knee. Obvious sighns of illness are uncommon in chronic brucellosis.

Manchester (1942) noted that in a study of 38 patients with chronic brucellosis the main joint manifestations in chronic brucellosis were arthralgia, joint pain, stiffness and sometime moderate swelling witout redness. In a further analysis of 22 chronic brucellosis patients with rheumatic manifestation the main manifestations were: muscles, 11; hands, 14; wrists, 12; elbows, 7; shoulders, 13; back, 14; hips, 7; knee, 18; ancles, 8 and feet, 4. The main joint complaints were: soreness, 14; stiffness, 13; aching, 9; pain, 2; lumbago, 4; sciatica, 2; swelling, 3; repeated swelling, 1; and muscular soreness, 11

Mousa (1987) in a review of 169 patients with osteoarthicular complication in brucellosis noted that the instance of arthritis was as follows: hip, 53%; knees, 36%; sacroiliacs, 20%; ankles, 15%; elbows, 5.3%; shoulders, 5%; wrists, 3.5%; and sternoclavicular, 1.8%. Spondylitis, 6%; osteomyelitis, 2.4%; and tendinitis and bursitis, 1.2%.

Gotuzzo (1987) studies 39 families with 232 members subject to B melitensis infections. of these 118 were diagnosed as infected. The most likely cause of illness was eating contaminated food. Arthriris of the sacrioliac and peripheral joints is more commn in younger individuals and spinal invovment (spondylitis) is more common in the elderly and in chronic brucellosis. Women were more susceptible to severe illness and this included articular involvement at 34.5 compared to men at 8.1%. The rate was less in children at 13.6% compared with ~27% in older family members. The proortion of 22 family members with articular invovement was: periferal monoarthritis, 9; peripheral polyarthritis, 3; sacroilitis, 5; spondylitis, 2; mixed arthritis, 1; extra-articular rheumatism, 4. Gottuzo (1982) reviewed 304 cases of hospitalised bucellosis melitensis (varient 1) patients most of whom had acute illness. The most frequent symptoms were fever and mailaise, 80% and arthralgias, 42%. Arthritis was present in 33,8%. The age profile of the patients was: <15, 23.3%; 15-35, 47.6%; 35-55. 17.4%; and >55, 11.5%. The incidence of arthritic involvement according to age was: <15, 29.5%; 15-35, 35.2%; 35-55. 33.9%; and >55, 39.1%. The incidence of the 4 main articular involvements were: sacroilitis, 46.6%; peripheral arthritis, 38.8%; mixed (sacroilitis and peripheral arthritis), 7.8%; sponylitis, 6.8%. Sacroilitis was most frequent in the 15-35 age group, peripheral in the <15 and 15-35 age groups, mixed in the 15-55 age group and spondylitis in >55 age group. Spondylitis occurred in 4 older patients with chronic brucellosis. Synovial fluid was tested in 11 patients and 3 were found positive by culture test. These recovered more slowly during antibiotic treatment. Sacroilitis usually occurs in the acute and undulant forms of the ilness is usually unilateral, non-destructive and resolves promply with antibiotic treatment. Peripheral arthritis usually occurs in the acute and undulant forms of the illness involving one of the leg joints (mainly the hip and knee). IT is usually non-destructive but may become so without adequate antibiotic treatment. Spondylitis occurred most often in chronic brucellosis and mainly affects the limber spine. It usually affects multiple vertibra and disc space. X-ray findings are similar to tuberculous athritis of the spine (Pott's Disease). Unlike Pott's disease, the vertibral anterior and upper margins are eroded with step deformity or rounding (Pon's sign) together with proliferative and reparative changes. Paravevertibral abscesses wich are common in tuberculosis are only occasional seen in brucellosis. Alarcon (1987) considered that musculoskeletal involvement can occur in up to 85% of brucellosis and is worst with B melitensis. This is based o the findings of Feiz (1978) and Potes-Querol (1957) . Peripheral arthritis is the most common manifestation and can affect 45% of patients mainly in the lower extremities joints. Associated problems are buritis, tendinitis, enthesopathies and osteomyelitis (uncommon due to use of antibiotics). Brucella was recovered from 60% of synovial fluid in peripheral brucellosis but this cannot be used as a guidwe to active arthritis as in other bacterial arthritis. The most likely explanation is the intracellular nature of brucellosis.

Al-Shabed (1994) reviewed 334 patients with proven systematic brucellosis and of the musculosketetal system ie back pain with or without neurologic deficit and joint pain from a consideration of imaging techneques: normal X-ray, skeletal scintigraphy, CT and MRI. He notes that the spine and large joints are frequently affected. Radiological evidence was established in 224 patients of whomw: 118 had spinal involvement; 81 had extra-spinal leisions and 25 had both. Pain, localised tenderness or radiculopathy and myelopathy are initial symptoms of brucellar spondylitis and pain tenderness, and swelling are features of Brucella arthritis and the less common osteomyelitis and myositis.

Young (1991) notes that joint involvement is the most common complication in brucellosis and can occur in up to 40% of cases.of mainly acute cases. The most likely forms of the diseases are: sacroilitis, peripheral joint arthritis, spondylytis, osteomomyelitis and bursitis. Spondulytis occurs mostly in the elderly and is shown by widening of the joint spaces or blurring of the articular margin can be seen but in the early stages standard X-ray technology shows nothing and CT scans show >90% of cases.

Spinal Brucellosis.

Lifeso (1985) considered a group of 70 brucellosis melitensis patients of which 21 developed spinal problems (spondylitis). Their average age was 53 years. Lifeso notes that the disease is difficult to diagnose and can be confused with spinal turculosis. The diagnosis was based on a minimum brucellosis antibody titre of 1:80, X-ray evidence of spinal involvrment and clinical response to treatment. 9 patients showed neurological involvement which was always due to extradural inflamitary mass. One patient had paraplegia at T3 and another tetraplegia at C4 but both were able to walk after anterior spinal decompression. The other seven had partial neurological impairment on presentation. Four required surgical intervention of whom 3 recovered successfully but one retained a spastic walk. The other 3 patients recovered through medical treatment. 9 patients had no X-ray abnormalities or only subtle changes recognised retrospectively. These included: slight retropharyngeal soft tissue swelling, 2 patients; slight changes in the sacroiliac joints (similar to early seronegative arthritis), 2 patients; disc space narrowing, osteophyte formation and sclerosis (initially attributed to osteoarthritis), 2 patients; no X-ray abnormalities, 3 patients. 12 patients had recognisable leisions varying from space narrowing with focal vertebral body endpoint erosion to severe contiguous leisions involving up to 4 vertebra and related disc spaces. 13 leisions occured in the lumber spine; 8 in the thoraic spine; 6 in the cervical spine; and 2 in the sacroiliac joint. 12 patients had a single leision; 4 patients had a single leision with multiple contiguous levels; 4 patients had multiple single level leisions; and 4 paitients had multiple leisions with multiple contiguous levels. CT scans provided better visualisation of the spinal canal and adjacent tissue. In comparison with tuberculous spondylitis paravertebral soft tissue masses were smaaler and more localised and psoas abscesses were absent. A variety of antibiotics was used for treatment but it was found that duration opf treatment was most important. 11 of 20 patients given 6 weeks treatment relapsed and 8 of 11 patients who relapsed following the initial treatment had less than 3 months chemotherapy and 6 of 9 patients given treatment for longer than 3 months were cured. After 1 or 2 courses of treatments 14 of 20 patients were cured the others were symptom free. 10 of 21 patients requireed surgery (surgical decompression, needle aspiration and laminectomy). Healing was associated with bony ankylosis in 59% of leisions. Persistent disc space narrowing with remineralisation of previously damaged bone.

In conclusion Lifeso notes that the earliest X-ray signs of spinal brucellosis are non-specific and may appear months after the onset of symptoms. The earliest signs are: miminal rarefication of the contiguous vertibral end plates, disc space narrowing, a small tissue abscess and occasional anterior vertebral body erosion. Often an elevated level of serum antibodies to Brucella is the only direct indicator. Progression of the illnes is associated with vertebra; body destruction with sclerosis at the margins of the infected area. Eventually the spine stabalises with disc space narrowing, marked sclerosis and osteophytes anterior and lateral to the involved disc space. Bone alkylosis occurred in most cases.

Al-Shabed (1994) notes that the spine is the most frequent site of musculosketetal brucellosis and this is one of the most serious complications. He points out that localized back pain or neurologic deficit may be the only manifestation of the illness. As well as this serology tests may be negative or equivical. In such cases response to antibiotic treatment may be the best guide to the illness. Brucellosis spondylitis is unusual because it can occur in a focal or diffuse illness.

• Focal Infection.In the focal form osteomyelitis is located in the anterior aspect of an end plate where there is a rich blood supply. If the area of infection is small, the organism has limited virulence or if the host has relative immunity only a focal ares of bone destruction develops at the diskovertebral junction. The lumber spine and particularly L-4 are most likely to be areas affected by brucellosis spondylitis. Ths is the earliest leision of spinal brucellosis that can be imaged and is invariable associated with bone sclerosis, anterior osteophyte formation (parrot beak). The disk paraspinal soft tissue and spinal canal are generally not affected. These changes may be present due to current or previous infection in endemic countries. Active illness should be diagnosed with history, clinical featutres and high or increasing antibody titres (may be equival). Also active osteitis can be depicted using scintigrapgrasphy. Without successful treatment reinfection may occur or progression to a diffuse form can occur. Children are not usually affected and the average age of onset is 40 years.
• Diffuse infection.In the diffuse form the infection can affect the entire vertebral body, disk, paraspinal soft tissue, neighbouring vertebrae and epidural space can be infected. This a slow process which can be be difficult to differentiate from othe degenerative disease, This type of infective spondylitis can best be documented using MRI or scintigraphy. features of the diffuse form of brucellosis as depicted by Ct scan are: the presence of bone destruction associated with bone sclerosis, obliteration of paraspinal muscle-fat planes representing edema, absrence of detectable paraspinal abscesses, and minimal or no epidural extension. Disk gas in small amounts occurs in ~30% of cases probably due to the slow destruction of the disk due to vertebral osteomyelitis. Unlike tuberculosis spondylitis the vertebral body morphology features are invariably preserved. Other factors which differentiate this infection from tuberculosis are that it is most likely to occur in the lumber region with associated disk gas, moderate epidural estension and invariably bone sclerosis. , osteomyelitis affects the entire vertebral end plates or the whole wertebral body. Spread of the infection occurs via the ligaments and vascular communications to involve adjacent disks and vertebra. When bone softening and destruction occurs from this process herniation of disk material into the osseous end plate and spread of infection into neighbouring soft tissue can occur. Imaging technology is insuffient to differentiate between brucellosis and tuberculosis. For tuberculosis this requies isolation of the tubercle mycrobacteria or the presence of caseating granulomas or the isolation of Brucella, positive serum Brucella antibody test or positive response to antibiotics.

Rotes-Querol (1956) reviewed spinal cpomplications in 174 cases of B melitensis. Sacroilitis occurred in 65 cases and more often in men and in younger patients. It is usually an early acute illness symptom and was: unilateral, 36 cases; and bilateral, 29 cases. The intensity of illness is variable from mild arthralgia with lameness to of a lower limb lasting a few days to acute arthritis and prostration. The illness lasts up tp 3 months and could be cured without clinical intervention. X-ray at between 10 and 15 days could showed alterastion to the periarticular osseous density (osteoporosis or osteosclerosis) or occasionally articular fusion. Only rarely do such problems persist. Spondylitis occurred in 93 cases (1 vertibral region, 43 cases; several verrtibral regions, 50 cases) and was more frequent in men than women. It was more common in elderly patients. The distribution of illness was: Lumer, 79, dorsal, 22; and cervical, 17 cases. He indicates 3 types: acute, patient prostrate for 2 12 weeks; subacute, patient walks with difficulty - duration of 3 months to at least 12 months; residual, vertibral pain lasting many years. Complications include: vertibral absesse, 21 cases; cervical meningomyelitis, 5 cases; pure melitis, 2 cases; meningeal radiculitis, 5 cases; latent meningeal reactions, 10 out of 14 cases in which lumber puncture was performed. In more serious cases X-ray shows charactoristic destructive leisions of the epiphyseal angles and erosion of the front and lateral surface of the vertibral bodies and in some case intraspongy geodes are visible. The related discs are reduced in height and contiguos disks may be swollen. Destruction is rapid but is limited by osseous reconstruction (osseous sclerosis, osseous bridging and proliferation of the periosteum). Rotes-Querol (1956) also reviews the formation of leisions caused by B melitenis in spondylitis. He divids these into macroscopic and histological leisions. Macroscopic leisions can be destructive leisions, reconstructive leisions and degenerative leisions. Destructive leisions are formed by destructive granulative tissue starting in the bone marrow and destroying the surrounding osseous trabeculae and invading and destroying the intervertibral disc. This occurs in the acute illness. Reconstructive leisions develop in the sub-acute phase of spondylitis and continue in the residual phase after the end of the inflamitary activity. There is an increase in the number and thickness of the osseous trabesculae surrounding the itraspongy geodes, thickening of vertibral bodies and the formation of osseous bridges. Degenerative leisions may occur after the inflamitory phase or before the onset of spondylitis. Histological leisions involve the formation of granulation tissue due to interstitial oedema and monocyte infiltation and can produce granulomas similar to those seen in other brucella affected organs. They consist of round monocytes with many eosinophil cytoplasm and large round nuclei. where granulastion tissue is dense patial or total reabsorption of the bone trabesulae can occur. Nodules of granulation tissue can be formed with central necrosis which may becone caseous or suppurative. Bone destruction occurs over a 3-4 month period and is accompaied by a proicess of limitation and reconstruction. Connective tissue proliferates around the brucellar nodule, new osseous trabecular form and increase in thickness, new bone forms from the vertebral peristerum and osseous bridges form. Destruction of the intervertebral disks is caused by the granulation tissue by direct erosion of the cartilage and by holes produced in the disk by pre-existing intraspongy hernias. X-ray imaging is ineffective at determining these leisions.

Pourbagher (2005) noted that 212 of 251 Turkish patients had arthralgia as a symptom and 114 were found to have osteoarticular involvement. Sacroilitis occured in 71 patients of which 48 had unilateral involvement. 71 patients were in the acute phase and 25 in the chronic phase. Spondylodiskitis was revealed by scintiraphy and MRI (the method of choice). 10 patients wre acute and 4 were chronic. Spondylodiskitis was distribution was: lumber spine, 15; thoracic spine, 6 patients; cervical spine, 1 patient. The destructive changes are similar to tuberculosis and therefore require serological back-up.Interesting images are available, mainmly MRI.

Extra-spinal Bone and Joint Brucellosis.

Al-Shabed (1994) reviewed 334 patients with proven systematic brucellosis and of the musculosketetal system ie back pain with or without neurologic deficit and joint pain from a consideration of imaging techneques: normal X-ray, skeletal scintigraphy, CT and MRI. 81 had extrea-spinal leisions as follows: osteomyelitis of long bone, 5; myositis, 4; arthritis (synovial), 66; arthritis (destructive), 6. For 77 patients with joint involvement, the distribution was: knee, 25; sacroiliac joint, 23; shoulder, 11; hip, 6; ankle, 2; elbow, 1. Brucella organisms locate in the synovitis with resultant effusion. Such effusive liquid is generally sterile. A similar process can occur in muscles and tendones. Rarely location in bone marrow of long bones can cause osteomyelitis or destructive arthritis. Joints. The knee and sacroiliac joints are most likely to be involved. The greater majority of infected joints show increases uptake of tracer by bone scintigram, increased synovial fluid with normal bone marrow on MRI and normal X-ray. The aspirate from such joints is usually sterile and treatment with anti-Brucella medication offers relief of symptoms with resolution of effusion but no alteration in X-ray findings. Destructive arthritis is rare and treatmebt may not stop further bone destruction presumably due to cartilage damage. Radiographic features are similar to tuberculosis and septic arthritis. Bones. Brucellar osteomyelitis of the extraspinal bones is rare. Such leisions are best detected by scintigraphy and MRI. Surgical intervention may be necessary but antibiotic treatment can be effective. Muscles. Brucellar myositis is rare and is best indentified by MRI and CT. Antibiotic treatment usually works in such cases.

O'Donnoghue (1933) reports a case of septic arthritis of the hip caused by B melitensis. A 12 year old girl living on a farm in Nebraska became ill with chills, malaise, and headaches. SA test for brucellosis was positive at 1:160. This was before antibiotics and she was treated with intravenous mercurochrome (which probably had little positive effect). After 10 weeks she had inproved somewhat but 2 weeks later repalsed with severe pain in thee right hip with which she had limited movement. The gland in her right groin was markedly enlarged and tender. Her spleen was slightly enlarged. The hip was drained surgically and a large amount of pus was released. Culture of the pus grow B melitensisd, X rays showed a solid boney ankylosis of the hip. At this time brucellosis was widespead in this region of the USA and he suggests that there might but not necessarily as severe. be many more similar misdiagnosed hip problems. He notes the arthralgic symptoms are present in most cases of brucellosis.

Coventry (1949) describe septic arthritis of the hip caused by B suis. A 22 year old farmer had first been ill surgically treated for a gangrenous appendix (it should be noted that brucellosis can mimic appendisitis) however he retained intermittent pain in his groin which transfered to his right thigh after 10 weeks. At this time he had a temperature of 99,6°F and slightly enlarged cervical, axillary and inguinal lymph glands. The hip was extremely painful with obvious quadriceps muscle atropy with tenderness of the thigh. SAT for Brucella antibodies was 1:800 dropping to 1:80 over 5 months. Brucella culture of blood was tried on a number of occasions but was always negative. Brucella cutlure of an aspirate from the joint was also negative. X-ray indicated osteopotosis of the head of the right femir and some destructive arthritis suggestive of tuberculosis. Examination of the femir head showed early severe damage to the cartilage and a pale boggy sinovial membrane. Excised samples of these were positive for B suis by culture. He was treated with antibiotics and made a good recovery including good hip movement and no pain. X-ray indicated showed residual slight bone atrophy.

Rotes-Querol (1956) reviewed peripheral joint involvement in 174 patients with B melitensis and found that osteophytes occurred in 86 patients. He classified the patients in 3 catigories:

• Arthralgia. occurred as moderate pain and stiffness in 53 (30%) of patients.It was most common in the early stage of the illness and was associated with febrile periods of illness but not exclusively. It can affected 1 or more than 1 joint but generally affects a particular joint only for a few days.
• Mild or Fugative arthritis occured in 39 (22%) of patients usually after arthalgia in the 2-8th month period. It considered of articular effusion with pain, synovial swelling and local warmth generally without reddening of the skin. The synovial fluid is sero-fibrinous and Brucella may be cultured from it. It lasts for 1 to 8 weeks and leaves no X-ray, anatomiical changes or functional sequelae. The most frequent sites were shoulder, knees, ankles, elbows, hips, the interphalanges of hands and wrists and the metastarsal-phalanges of the big toe. Shoulder periarthritis occurred in 9 cases of 18 cases with osteophytes with pain and limited movement for 2 to 4 months.
• Severe or destructive arthritis affected 8 patients in this series: hip, 6; knee, 1 case; big toe (metastarsal-phalangeal joint), 1 case. This can occur in acute or subacute illness lasting 3 to 8 months. X-ray changes are those of typical infective arthritis. The articular fluid maybecome suppurative but normally fistulised to the surface. Serious anatomical or functional sequelae are likely to occur including destruction of the articular cartilage.

Rotes-Querol (1956) also reviewed non-skeletal bone marrow involvement in B melitensis (osteomyelitis) and concluded that it is a rare complication. In his series of 174 cases there were only 2 cases of costal osteitis. In a review of 27 collected examples this affected: 1 site, 17 cases; and more than 1 site, 10 cases. The sites most affected were: the anterior surface of the ribs, 17 cases; the femar, 7 cases, the cranium, 4 casaes; the tibia, 3; the humerous, 2 cases; the iliac, 1 case; the ischium, 1 casse; coracoid process. These were involved in 2 ways: acute accommanied by fever and chronic in which the localised leisions are evolve slowly over a long period. He notes that the usual serology tests are not useful in such cases, only the skin test gives a significant positive response.

General Diagnosis of chronic brucellosis.

Barrett (1953) notes that the diagnosis of chronic brucellosis ('B abortus) must be established by careful analysis of the patient since serology and culture tests are often of little value.

• Clinical history and examination of the patient are of great importance.
• Too much faith in bacteriological examination.
• A long standing history of nught sweats and diffuse or localised pain despite the patient appearing normal.
• Occupation such as farmer.
• Clinical examination may reveil enlargement of the liver or spleen.
• Lymphadenopathy is not common.
• ESR and blood count are usually normal.
• Blood test then intrademal test carried out. If skin test and serology tests are all negative this suggests brucellosis is absent. They have no examples of demonstratable SAT titre with negative skin test but the reverse is common. A sharp rise in SAT after about 2 weeks is considered proof of active infection in association with an appropriate clinical history. patients with a slight SAT titre rise should be considered for liver biopsy for condfirmatory proof.

Simpson (1930) in a study of acute brucellosis patients found that agglutinins generall developed between 5 and 14 days from onset although delays upto 4 weeks are common, therefore a test at 7 to 10 days is optimum.In some patients although apparently recovered agglutinins continue to be present for months or even years while in others they disappear.m He notes that Carpenter (personal communication) has found that 6% of patients with positive Brucella culture test have negative a SAT titre.

Q1.

Scarlett (1948) concludes that on the basis of his own 25 years study of brucellosis including 100 with chronic brucellosis that it is a chronic illness of which acute cases form only a small part. The idea that the frequently described decline in SAT titre means that it is it is a self limiting illness is erroneous and does not demonstrate the true nature of the illness. In the chronic illness the epidemiological history is often obscure, course of the illness either intermittent or continuous and the symptoms protean. A psychological component was often present, fever may not be evident and there may be no obvious physical signs of illness. Laboratory tests are inadequate. He is in no doubt that at the time of writing chronic brucellosis is a widespread illness. He gives what might be a typical history of illness. An individual has a mild infection probably with a later flare-up or obscure febril episodes. The patient then hass a number of years of ill healthn without anadequate diagnosis and is partially of severely debilitated and often unable to work. Scarlett notes that of 4105 cattle tested in Alberta, Canada in 1948 12% had a SAT titre of ≥1:100. In general he concludes that in endemic areas about 10% of people have been infected and that these can have an active form of the illness in some degree and there may be a latent period of symptomless illness lasting many years followed by a sudden onset of illness. Also patients who have seemingly recovered completely fron an acute attack may have a sudden relapse years later.

Scarlett studied 100 patients with chronic brucellosis w2hose main symptoms are shown in Table .... (transcribed from a chart). As well as fatigue patients also suffered periods of extreme fatigue which they termed "black-outs". Aches and pains affected both muscles and joints as sourness and stiffness. The symptoms of nervousness were irritability, emotional instability and a curious mental depression. Gastro-intestinal distrubance involved mainly epigastric burning, flatulance and constipation. In general there were no physical symptoms except a pallor in 10 patients and splemomegaly in 3 cases.

Symptom	Percentage
Fatigue or weakness	100
Aches and Pains	93,6
Fever	79.8
Nervousness	72.3
Headaches	78.7
Gastro-intestinal	59.7
Rheumatism	58.5
Weigh Loss	53.9
Periods of exreme fatigue	39.4
Sweating	29.8
Cough	17.0

Scarlett considered that the greatest obstacle to cure in chronic brucellosis is the tendency for Brucella to exist in foci of infection in a low-grade form of the illness. Typical sites for foci include the biliary tract and retroperitoneal and other lymph nodes.

He gives some examples: He notes 4 cases of chronic brucellosis where Brucella has been recovered from the gall bladder and remission has been achieved by cholecystectomy. Similarly in some cases where the spleen was the foci of infection the removal of the spleen has also inproved health in the same way.He describes one case: A 46 year old man suffered with symptoms of epigastric discomfort, indigestion and flatulance. Also he suffered from influenza type attacks, and periodic evening fevers. He had aBrucella SAT titre of 1:1280 and a strongly positive interdermal test. A cholecystectomy revealed a that the gall bladder was grossly thickened showed granuomatous tissue with pus present. A sample of this material gave a positive culture for B abortus but a bile sample was negative. The patients health inproved after the operation.

He notes that joint manifestations are a common form of localised brucellosis. Of the 100 patients under study 10 had atypical arthritis with tender and swollen joints. He describes a man od 45 who suffered from marked stiffness and soreness in most of his joints particularly the hands, knees and back for 4 years. Treatment for arthritis failed. His knees and hands were slightly swollen and he had an SAT titre of 1:160 and a strongly positive skin test. After 4 monts of vaccine treatment we improved and could work again.

He describes a case of a 27 year old women who suffered for years from extreme fatigue, headaches, widespread aches and pains. She was prostrate, eotional and depressed. She was considered by her doctors to bea hopelessly unstable neurotic individual. Tests for brucellosis indicated a SAT tire of 1:40 and a strongly positive skin test. After 6 months treatment withvaccineshe made a remarkable recovery.

He dscribes a woman of 24 ill for 3 years with weakness and extreme lethargy, generasl aches and pains and weight loss who was prostratew and unable to work. She had no physical signs of illness. Her SAT titre was negative and her skin test moderately positive. After 3 months vaccine treatment she was sufficiently well to return to work although she suffered from occasional flare-ups.

In general he considered that all laboratory tests for chronic brucelklosis have serious limitations. All tests must be considered in relation to clinical findingsd. Individual tests are useless. The results of laboratory tests for the 100 chronic brucellosis patients was as follows:

• Interdermal test: total, 96; strongly -ve, 25; moderately positive, 68; negative, 3. 200,00 organisms of a vaccine containing mixed killed B suis and B abortus was used. He notes that this test is generally considered to give 92% positive results in proven cases. This test cannot distinguish between past and present disease or active and inactive infection so it cannot be used to make a diagnosis on its own.
• SAT titre: Total, 91; negative, 66, 1:29, 2; 1:40, 6; 1:80, 10; ≥1:160, 7 casaes. This is 72% negative and according to Scarlett is typical of what other workers have found. Therefore in chronic brucellosis in particular SAT is not a reliable diagnostic tool. Also although SAT titres may be negative it is known that positive titres may appear intermittently at uncertain intervals in such cases.
• Phagocytic test: Total, 91; moderate, 11; slight, 5; negative, 75. The phagocytic test is only of use in conjunction with other tests. It is best used to monator treatmen where a progressive increase in the phagocytic index is usually commensurate with clinical improvement.
• General blood tests: He concludes that tests for relative lymphocytosis, leukopenia and macrocytic anemia were of no diagnostic value.
• Sedimentation rate: He notes that in his own work and thsat of others sedimentation rate remains normal throughout the course of the illness.

In 1948 Scarlett considered vaccine therapy was the best available treatment for chronic brucellosis using nitrous acid killed B suis at a concentration avoiding strong reactions. He preferred an intramuscular rather than a subcutaneous injection. Usually 9 to 12 moths treatment is required with gradually increasing dosages. Injections of vaccinew are given daily at the start and then twice weekly as a maintenance dose once there is a marked inprovement. Prolonged treatment is necessary to reduce the incidence of relapse. Relief from symptoms starts at from 6 to 12 weeks. The criteria for success are: disappearance of symptoms including fever, weight gain, increased ability to perform work and an increase in the phafocytic index. The treatment is time consuming and tedious neverthe less patients responded well to because of the improvement to their health. The results pf treatment are shown in Table .... 75% of cases can be expected to show complete arrest of chronic brucellosis using vaccine treatment however the disease is not necessarily cured and reoccurrance can take place years later.

Response	Patients
	Number
Excellent	8
Very good	29
Good	31
Improvement	18
Slight	4
Death	1

Gastrointestinal and heptobillary sytems.

General symptoms.

Madkour (1989) reviewed the symptoms of gastrointestinal brucellosis in 425 patients and found them as follows: Loss of appetite, 77.9%; Loss of wight, 64.2%; nausea and vomiting, 5.4%; constipation, 44.2%; abdominal pain, 44.9%; diarrhea, 7.0%; jaundice, 1.2%; heatomegaly, 16.7%; splenomegaly, 17.2%; pancreatitis, 0.2%; raised liver enzymes, 23%; raised serum bilrubin, 2.8%. Complications include: hepatitis, cholecystis (with or without gall stones), suppurative granulomas in th spleen or liver and pancreatitis. He notes that there are usually no specific symptoms indicating liver involvement. The liver and spleen are involved to some extent in virtually all cases. Young (1995) indicated that heptobillary complications probably occur in the majoritory brucellosis cases and are suggestive of viral hepititus. Mousa (1988) did a study of 115 brucellosis patients and found that 34 had gastrointestinal synptoms of whom 17 had dyspepsia. Endoscopic investigation indicated that 6 had erosive gastritis. Aziz (2005) notes that gastrointestinal manifestations may be the only presenting features. Mild complaints include: anorexia, vomiting, diarrhea, consipation. The most observed liver abnormalities are raised transaminases and alkaline phosphatase, The incidence of symptoms and complications are: splenomegaly, 29-57%; anorexia, 22-45%; vomiting, 7-18%; abominal pain, 9-17%; abominal tenderness, 15%; constipation, 2-12%; diarrhoea, 3-6%; colitis, rare; pancreatitis, rare; cholecystitis, rare; peritonitis, ascites and intestinal obstruction, very rare.

Hardy (1930) notes that fever, abdominal pain and localised tenderness in brucellosis may mislead a a physician to a dignosis of chronic or subacute appendicitis or cholecystitis and apply inappropriate surgery. Simpson (1930) has recorded 12 such appendectomies and 2 cholecystitis.

Simpson (1941) in a review of his own work and that of others in the |USA considers that the main gastroenteral complains are anorexia which parallel the severity of the infection. Abdominal pain is common and often misinterpretted. Diarrhea is less common.

Corbel (2006) notes that as well as systemic symptoms B melitensis can have a typhoid like charactor which is probable related to the most frequent route of infection via food. He notes that gastroenteral compaints also occur. These include: nausea, vomiting and abdominal pain. Rare symptoms are: ileitis.colitis and spontaneous peritonitis.

Simpson (1930) notes that abdomin pain due to brucellosis has frequently been misinterpretted as appendicitis or cholecystitis.

Young (1995) notes that brucellosis can be a foodborne infection and that associated symptoms are: anorexia, weight loss, nausea, vomiting and abdominasl discomfort and occasionally: ileitis, colitis, and spontaneous peritinitis caused by B melitensis.

Young (1995) notes that the liver is involved in the majority of cases of brucellosis. Tests results are mildly abnormal. Sometimes transaminase levels are elevated suggestive of hepatitus as noted by Randolf (1994) and Losurdo (1994) . Cohen (1957) notes that Brucella can be isolated from liver tissue where there is no sign or symptoms of liver disease or laboratory evidence of liver dysfunction. The style of infection in the liver depends of the species of Brucella. Spink (1949) notes that for B abortus the charactoristic lesions are noncaseating epithelioid granulomas. Young (1995) observes that in B melitensis a range of lesions occur including: small aggregations of mononeuclear cells are found in portal triads or surroundingfoci of hepatocellular necrosis, larger aggregates of imflammatory cells can obscure portal structures and extended into the parenchyma resembling active hepatits, aggregation of cells including histiocyte may formstructues like loose granulomas, and rarely small true epithelioid granulomas. In Brucella suis hepatic abscesses and chronic suppurative lesions commonly occur. Williams (1982) describes 2 examples.

Leon (1953) studied diabetes in 150 proven brucellosis cases. He found that 67.7% had diabetes and 61.4% had liver deficiency. In a control group of 36 other diseases (syphilis, streptococcus endocarditis and amebiasis) 13.8% were hyperglycemic and 50% had liver deficiency. Pancreativc abnormalities have been noted by Hughes (1897) , Rothenberg (1933) and Harris (1950) . Leon (loc. cit.) concluded that the diabetes cases were due to liver deficiency (69.1) and pancreatitis (29.9%).

Chronic brucellosis.

It is interesting to note that in the first description of brucellosis by Marston (1851) emphasises the gastric nature of the illness and calls it "gastric intermittent fever. He mentions the symptoms as subacute dyspepsia, anorexia, nausea, great derangement of the assimilative organs, tenderness of the epigastric region. He mentions epigastric pain similar to the symptoms of an ulcer. He wrote from his own personal experience. Harris (1943) also notes that brucellosis patients may have epigastric discomfort and related symptoms suggestive of an ulcer.

McDermitt (1973) notes that gastritris is a typical symptom of chronic Br. Possible gastrointestinal symptoms in chronic brucellosis are abdomin pains, swollen abdomin, dispepsia, gastritis and diarrhoea.

Gauss (1950) working in Colorado, USA bat that time an endemic area notes that a significant feature of the symptoms of chronic brucellosis is a gastrointestinal syndrome which can dominate the clinical picture. He reviewed 20 cases of chronic brucellosis and concluded that gastrointestinal effects were significant symptoms. Patients likely to have these symptoms included: those with direct contact to Brucella, those given the "psycho" diagnosis of neurasthenia and psychoneurosis, those with unexplaines gastro disorders and those with unexplained minor temperature rises. Symptoms of the 20 patients included: gastric irritation, irritable colon syndrome and biliary dyspepsia. The main symptoms were: epigastric distress, 6; rawness feelings in andomen, 3; diffuse abdominal distress, 9; bloating, 11; eructation, 8; constipation, 12; acid indigestion, 3; food intolerance (fried food / onions); nausea, 4; flatulance, 4; irritable bowel, 1; heartburn, 2.

Harris (1944) considered that abdominal distension of otherwise ocscure origin is often indicative of chronic brucellosis.

Simpson (1941) observed that hypochlorhydria or achlorhydria are common features of chronic brucellosis.

An interesting example of inappropriate treatment relating to the difficulty of diagnosing brucellosis is described by Wallis (1957) . He describes a girl of 12 who prior to diagnosis with brucellosis had received the following unnecessary procedures: appenicectomy, intravenous pyelography, sigmoidoscope, a diagnosis of psychoneurosis, and long hospital stays.

Chronic Hepatosplenic Suppourative Brucellosis.

Granuloma of the liver and spleen tend to form during acute attack but gradually disappear with successful antibiotic treatment. Suppurative abscesses may also form. In a minority of cases these leisions can become dormant sometimes including calcification but can then re-emerge to cause serious illness many years later. During this dormant time the patient's illness may be chronic or asymptomatic. Madkour (1989) suggests that the typical dormant period is 10-25 years, that lack of early antibiotic treatment is a contributary factor, and that B suis is the most likely but not exclusive cause. Possible associated symptoms are: epitaxis, gastrointestinal bleeding, ecchymosis, thromocytopenia, and anaemia. During the relapse Brucella can be culltured from the granuloma or suppurative lesion. Such cases have been described by Spink (1964) Yow (1961) and Schirger (1959) . Ariza (2006) report a man ill with brucellosis and apparently cured and living with no apparent risk factors for brucellosis suddenly taken ill after 40 years with high fever, chills, hypertension, peripheral hypoperfussion, scleral icterus and painful heparic enlargement, and unknown source os septicemia. Brucella test were: RB, -ve; SAT 1:40; culture, -ve but seroconversion was later noted and brucellosis confirmed.CT indicated a Brucella granuloma which was resectioned. The patient received 5 months of treatment and then returned to health for 4 years (end of study). He also describes the case of a woman was diagnosed and apparently cured and remained well for 20 years. During this period she had no risk factors for brucellosis. Then she suffered a relapse with high fever, chills, sweating, arthralgia and weakness, hepatomegaly, splenomegaly and right sided pleural effusion.Tests for brucellosi were negative. After 4 weeks, brucellosis was confirmed by SAT and AHG tests. CT revealed a reactivated Brucella liver granuloma. She recovered unwith antibiotics but became ill again 7 years later after a surgical hepatic resection and antibiotic treatment.

Colmenero (2002) reviewd 7 patients with hepatosplenic Brucella abscesses. CT showed large poorly defines leisions with heterogeneous attenuation and thick central calcification surrounded by hypointense areas. In all cases there were granulomas with central necrosis, polymorphic infiltrates, feww giant cells and peripheral fibrosis. The treatment of choice was surgical intervention necessary to garantee removal of the central calcified nucleus responcible for the persistent infection. Antibiotic therapy plus percutaneous drainage did not work.

Brucellosis of the nervous system.

Brucellosis of the central nervous system.

Many of the typical symptoms of chonic brucellosis such as depression, headaches, tinnitus, insomnia, irritability, anxiety and lower back pain point to neurological problems. This situation taken with lack of serology results has lead to a range of diagnosis including neurasthenia and CFS. Williams notes that some erronious labels applied to neurobrucellosis include; psychoneurosis, delusional, psychosis, schizophrenia, tuberculosis, intracranial tumour and cerebrovascular insufficiency. Buchanan (1974) indicates that headaches, dizziness and tinnitus are related to CNS involvement in Br. Araj (1988) noted that CNS complications occur in up to 3-5% 0f accur cases . Lulo (1988) note that neurological problems affect 6% of acute patients and 28.5% of chronic patients. Of 42 chronic patients 9 had 8th cranial nerve damage. Seidel noted that 13% of Br patients had neurobrucellosis. The most common symptoms are: meningocephalitis, polrodeculoneuritis and cranial nerve damage. Fincham (1963) described the typical simptoms of neurobrucellosis as increasingly troubling headache, extreme headaches, blackouts, tinnitus, fever, shaking chills, drowsiness and dizziness. In serious cases seizures and unconsciousness can occur. Kochar (2005) says that neurobrucellosis is an important manifeatation of the illness and encompasses neurlogical and psycological presentations. Menigocephalitis is most common in acute Br. The chronic form can either be peripheral (proximal polyradiculoneuropathy) and/or CNS (myellitis, cerellus deficit or cranial nerve damage). This illness is often secondary to to focal illness such as in the bone marrow. Cutler (2005) noted that neurological complications in brucellosis are associated with anxiety, amnesia, delusions, hallucination, delirium, phobias, irritability and prolonged occuital headaches. Izadi (2000) ) considered that 8^th nerve damage is a charactoristic symptom of Br and that typical symptoms are: hearing damage, tinnitus, vertigo and dizziness. Shakir (1987) says that 8^th nerve damage and lower back pain are typical symptoms. He notes that ELISA test on CSF can identify antibodies. Spink says of neurobrucellosis that it has a spectrum ranging from acute to chronic with minimal local disease to massive widespread NS disease. It can be latent, subtle or persistent. Typical manifestation will not be found. Young (1995) noted that Br meningitis is can be acute or chronic and is the most common symptom. Typical symptoms are dizziness, inability to concentrate and sexual problems. He notes that this meningitis can be a serious complication. Solero (1998) says that although neurobrucellosis is curable permanent nerve damage may occur. Rust(2006) notes that neuro Br mostly occurs in the chronic phase of the illness. It usually develops in the wake of a latent period after an initial bout of acute Br. Its onset is usually abrupt. In endemic countries Br may result in permanent neurological damage.

Sanchez-Sousa (1990) notes that the reported incidence of neurobrucellosis varies fro 0.5% to 25% according to the criteria use to define it.

Hardy (1930) from a study of about 330 patients noted that the most common symptom during acute brucellosis was varying degrees of insomnia. Other symptoms were: restlessness, irritability, undue apprehension. Delirium and coma were rare.

Simpson (1941) notes that features of brucellosis are restlessness and insomnia associated with noctural febrile exacerbations. Brucellosis can affect the brain, spinal corde or meninges. This can occur in the early acute phase of the illness with signs of encephalitis, myelitis or meningitis however it is more commonly observed as a delayed complication in chronic brucellosis. He considered that according to the work of Rogers () , meningeal involve is the predominant effect and encephalitis or myelitis are secondary effects. The symptoms of neurobrucellosis may be the only manifestation of the disease. The symptoms may vary widely dependant on the extent of meningeal invasion and the brain. These include: severe headaches, vertigo, diplopia, nuchal ridigity, aphasia, psychic disturbance, and various forms of paralysis. The spinal fluid be increased in pressure and show pleocytosis, increased albumin and decreased globulin and sugar. Diagnosis of Brucella by cullture is desirable.

Corbel (2006) considered that in Brucella menigitis brain scans using CT or MRI are usually normal however it can detect space-occupying leisions, the integrity of the epidural space, and basal ganglia calcification.

Pappas (2007) carried out a review of neurobrucellosis. He notes that neurobrucellosis is a complication which is likely to be associated with delayed diagnosis and treatment. Such delays may occur in both endemic countries (developing) where medical facilities are limited and in non-endemic countries (developed) due to low index of suspician. The general opinion suggests an incidence of 3-5% however he notes that many patients who are considered successfully treated for brucellosis continue to exhibit neuropsychiatric symptoms or chronic fatigue like symptoms despite microbiological evidence of disease relapse. He notes that the nature of this pathology remains obscure although there is persuasive evidence that apparently healthy persons can retain Brucella in their system years after treatment. Pappas considered that the most frequent forms of neurobrucellosis are: meningitis/ encephalitis/ meningoencephalitis, myelitis/myeloencephalitis, radiculopathy, vascular syndromes (mycotic aneurysms), cranial nerve involvement, cereballar syndromes, peripheral neuropathy, brain/intramedullary abscess. Less common are: brain/intramedullary granuloma, demyelinizatio syndrome, isolated papilledema, isolated hearing loss, isolated seizures, pseudotumor syndrome, diabetes insipidus.

Ocular Manifestations in Brucellosis.

Brucellosis can affect the eye and optical system including the external ocular muscles, the uveal tract, the retina and the optic nerve. Disorders can be purely ophthalmopathilogical or neuro-ophthalmological. There are many recorded manifestations. The most common are: uveitis, choroiditis, optic disc and retina neuritis and optic disc and retinal hemorrhaging/edema. Other reported forms include: dacryoadenitis, episcleritis, nummular ketatitis, endophthalmitis/iritis, iridocyclitis and anterior cyctitis, conjunctivitis, clouded vitreous humour, cornea damage/ulceration, iritis, retinal leisions, chororetinitis, retrobublar neuritis, retina detachment, fundus involvement, cataract, glaucoma, and strabismus. In acute brucellosis the most common ocular effect is uveitis (mainly anterior uveitis) and in chronic illness choroiditis. The most frequent symptoms described by patients are: double/disturbed vision, diminished vision, red/sore/painful eye, blurred vision, photophobia, and conjunctovitis. One or both eyes masy be affected. Ocular brucellosis is more common in chronic brucellosis. In severe form particularly if untreated it can cause blindness or partial blindness of one or both eyes. It more often occurs in the chronic phase of the illness. In chronic brucellosis the ocular manifestations are likely to be accompanied by the usual symptoms: weakness, fatigue, chills, sweating, headaches, etc. The balance of evidence suggests that symptoms are worse in B melitensis and B suis than in B abortus.

The eye may be infected directly by Brucella via the conjuctiva such as by aerisol spays or physical transfer to the eye. Van Rooyen (1091) records the case of a vetinerian accidently sprayed in the eyes by B abortus vaccine who then developed keratoconjunctivitis in both eyes. After 2 months he had a positive brucellosis antibody test and was treated with antibiotics unsuccessfully in the case of the eye. Subsequently the right eye deteriorated with a corneal abscess, uveitis, cataract and finally panophthalmitis so and the eye was enucleated.

Molinelli (1947) studied 95 patients with brucellosis. Of these 41 had ocular manifestations and only the fundus was affected in 20. Cremona (1951) examined 160 B abortus and B suis patients and found eye leisions in 89 and of these 76 had venous conjestion of the fundus oculi and 33 conjunctivitis. Dalrymple-Champneys (1954) reviewed experience with 1134 patients (mainly B abortus) and found transient visual problems: (nystagmus, diplopia, difficulty in focusing, etc), 68 cases; conjunctivitis, 3 cases; eye ache, 7 cases. One patient developed blindness in one eye and another partial blindness due to retinal hemorrhaging.

Green (1938) reviews 33 cases of brrucellosis with ocular involvement. These case were proven by epidemilogical, illness history and serology tests. THe main symtoms described by patients were: diminished vision, 8; double vision, 6; eye pain, 4; blurred vision, 2; conjunctivitis, 2; red inflamed eye, 2; photophobia, disturbed vision, lactimation narrowed vision field and . The main ophthalmological findinds were: papilledema, 5; optic nerve atraphy, 4; retinitis 4; retinal hemorrhage, 4; optic disc neuritis, 3; vitrous humor clouded, 3; retinal edema, 2; optical disc hemorrhage, 2; optic disc leisions, choked optic disc, 2; optical disc edema, 2; fundus hemorrhage, 2; conjested optical disc, 2; chorioditis, 2; cornea damage/ulcer, 2; keratitis, 2; retinal vein tortuosity, 2; retinal leisions, iritis, iridochoroiditis, chororetinitis, optic nerve/ disc edema, retrobulbar neuritis, blocked pupil, macular discolouration, iridocyclotis and anterior uveitis, 2; strabismus. At this times vaccination was the main treatment, Results were variable: complete cure, dinished vision, partila blindness, 5; and blindness, 5.

Bekir (2000) describes the case of a boy of 16 with unilateral dacyoadenitis (enlarged lacrimal gland). This was proved to caused by brucellosis mellitensis by SAT at 1:640 and culture test on lacrimal gland aspirate.

Rolando (2008) carriee out a review of of ocular manifestations in 1551 paients with brucellosis melitensis diagnosed on the basis of clinical findings, SAT or culture of B melitensis. Of 965 patients with acute brucellosis 7 were affected. Of 570 patients with chronic brucellosis 45 were affected. Uveitis occured in 43 of the 52 patients with posterior uveitis occurring in 21 cases. Panuveitis occurred in 9 patients of whom 8 were were designated as legally blind of whom 5 had no light perception. Possible alternative diagnosis are tuberculosis and toxoplasmosis. Rolando corelates possible presentations of ureitis. These are anterior uveitis: iritis, iridocyclitis and anterior cyctitis; intermediate uveitis including pars planitis, posterior cyclitis and haylitis; posterior uveitis including choriditis, chorioretinis, retinis and neuroretinitis.; and panuveitis including inflamation of the whole ureal tract. Possible complications were: cataracts, glaucoma, macrulopathy, viteal atleration, phthisis bulbi, opticarrophy, neurovascular retinal membrane, tractional retinal detachment, Behcet syndrome, Harada syndrome (posterior uveal inflamation with retinal exudations and meningeal irritation) and Vogt-Koyanagi-Harada syndrome (anterior segmenrt inflamation associated with vitiligo, poliosis disacosia and meningeal irritation. In Rolando's 36 patients the incidence of complications and sequelae were as follows: Cataracts, 18; vitreal alteration, 11; phthisis bulbi, 6; maculopathy, 6; glaucoma, 5; neovascular retinal membrane, 3; Harada syndrome, 3; optic atrophy, 1; tractional retinal detachment, 1; Behcat syndrome, 1; Voogt-Koyanagi-Harada syndrome,1; none, 16.

Puig Sonales (1953) noted that in consideration of the protean manefectations of brucellosis it is unlikely that the eye with its variety of tissues and rich circulatory system would escape infection. He carried out an in depth review of ocular brucellosis. He records that ocular brucellosis was first demonstrated by Fabyan (1912) as discussed by Barky (1939) in guinea pigs which developed sclera, choroid and lactimal inflammatory infultrations. He also found that the eye could be directly affected by Brucella. Orloff (1928) reviewed other examples of ocular brucellosis in animals. Guinea pigs developed corneal infiltrations and cataracts 4 months after initial infection. Goats infected with Brucella developed corneal infitration and optical disc edema. Infected bulls developed optic atrophy and relapsing bilateral uveitus leading to cataract or retinal detachment. Lemaire (1924) as recorded by Greene (1938) first described ocular brucellosis in a human who suffered from bilateral optic neuritis and ocular paralysis associated with brucellosis meningitis. Puig Sonales paper contains a list of recorded cases of ocular brucellosis from the literature which shows the variety of disorders. Puig Sonales reviews 413 cases of his own in Mexico all confirmed as suffering with brucellosis of which 60 patients associated oculat disorders. 12 patients had purely ophthalmopathies and 48 neuro-ophthalmopathic disorders. The number of patients with purely ophthalmopathies were: Punctate keratitis, 1; tenonitis, 1; retinopathy (macular retinopathy, 1; retinal phlebitis, 1.), 2; uveitis (iridocyclitis,6; bilateral uveitis, 1; hypertensive uveitis, 1), 8. The number of patients with neuro-ophthalmopathic disorders were: papillary conjestion, 22; papillitis, 8; papilledema, 2; retrobulbar optic neuritis, 5; optic disc atrophy, 3; postneuritis, 3; optic neuromyelitis, 1;optochiasmatic arachnoiditis, 1 case; ocularotary paralysis, 3 cases; ocularotary paralysis combined with papilledema, 1 cases. The duration from initial illness to delopment of ocular illness was: <6 months, 24; 6-12 months, 6; 12-60 months, 22, >60 months, 7; unknown, 1.

noted that cases of ocular brucellosis included: uveitis, neuritis, optic neuritis, papilloedema, keratitis and other diverse ocular and neurolical presentations.

Abd Elrazak (1991) describe a 13 year old girl with proven brucellosis and blindness and bilateral papiledema. CT scan showed irregular thickening of optic nerve. She recovered completely after antibiotic and high dose corticosteroid treatment. Cavallaro (1990) describes a 14 year old boy with bilateral papilledema. He recovered after antibiotic treatment without steroids or other anti-inflamitary drugs.

Gungur (2001) describe 3 cases of episceritis caused by proven brucellosis. In all cases the condition was associated with other symptoms typical of brucellosis including: fever, night sweats, arthritis, and lower back pain. Typically they had 'red eye'/hyperamia manifested as engorgement of the temperal quadrant of the bulbar conjunctiva also affecting the adjacent corneal limbus and cornea. Visual acuity was not affected. Steriod treatment was ineffective and lead to recuurence of the condition. Long treat antibiotic treatment (docycyclin anf riframpicin) for greater than 3 months was effective.

Ocular Involvement in Chronic Brucellosis

General informtion indicates that ocular brucellosis occurs mainly in the chronic phase of the illness. Puig Sonales (1953) considered that uveitis occurs most frequently in the chronic and subacute illness. Gungur (2002) reviewed ocular manifestations in 147 proven brucellosis patients and found that 38 had ocular complications: conjunctivitis, 26; anterior uveitis, 6; posterior uveitis, 1; dacryoadenitis, 2; episcleritis, 3. 3 patients with uvetis also had brucellosis spondylitis. Concluded that ophthalogical examination of brucellosis should be performed routinely. Occular manifestations were more common in chronic patients (73%). Wood (1953) noted cases of ocular involvement in the chronic illness. Rolando (1985) noted that of 15 patients with ocular involvement 11 had chronic brucellosis. Harris (1945) notes that ocular manifestations occur in both acute and chronic brucellosis. In the latter case this may lead to mis diagnosis.

Greene (1938) in a review gave examples of ocular brucellosis occurring in both acute and chronic brucellosis. He describes the case of ocular brucellosis reported by De Jong (1936) of a 27 year old farmer exposed to B abortus. He initially had typical symptoms of brucellosis: weakness, fatigue, frontal headaches, general aches and pains, constipation, nervousness, anorexia, urinary frequency and nocturia. His ilnes was confirmed by SAT and culture of Brucella from spinal fluid. 8 months after the onset he developed headaches, tinnitus, vertigo, nausea, and vomiting. At the same time he developed neuroretinitis, weakness of ocular motion, diplopia and retinal hemorrhages. Some improvement was achieved by vaccine treatment.

Sungur (2009) reviewed 132 patients with proven brucellosis and typical symptoms. From these he found 44 eyes of 28 patients (18 acute patients, 26%) affected. The types of illnes in these 44 eyes was as follows: anterior uveitis, 18 (41%); choroiditis, 14 (32%); panuveitis, 4 (9%); papilledema (secondary to Brucella meningitis), 4 (9%); retinal hemorrhage, 4 (9%). All the acute patients had anterio uveitis whereas patients with panuveitis, choroiditis, retinal hemorrhage and papilledema had chronic brucellosis.

Rabinowitz (2005) discusses tha case of a 39 year old man with short term onset of trouble with both eyes but no other known illness. These for syphilis, tuberculosis and antineuclear antibody were negative. A history of drinking raw goat's milk suggested chronic brucellosis. This was confirmed by a SAT of 1:160. The right eye showed hyperemic optic disc with serous retinal detachment. The associated macula showed creamy wtiish yellow choroidic leisions. Similar effects were present in the left eye. He was treated with antibiotics plus methasone phosphate drops and oral prednisone with significant inprovement. Informative imagery is provided.

Tunc (2004) refers to a patient with chronic brucellosis previously confirmed by a Brucella antibody test who developed bilateral vision loss. There was optic disc swelling, optic neuropathy and significant retinal detachment in both eyes.

Foggitt (1953) describes a 33 year old man with impaired vision in one eye wich started 2 months previously and suddenly worsened. The patient had fairly typical symptoms of chronic brucellosis and a positive Brucella skin test. The eye was slightly iritated and conjested. Examination showed hazy vitreous and some anterior choroiditis. The diagnosis was anterior uveitis. One year later he developed periphlebitis of the other eye with hazy vitreous and tortuous retinal veins. He was treated with antibiotics which was fairly successful although he retained low grade iritis and some retinal atropy in one eye.

Tabbara (1999) describes 34 year old Saudi women with an nine year history of left eye pain, eye redness, lacrimation and headaches. The right eye was unaffected throughout. During this time her eye was treated with cortocosteroids. She also had an episode of arthritis of the knees nd interphalangeal joints. She also suffered episodes of fever. She had previously drank unpasseurised milk. Chronic brucellosis mellitensis was proved by a blood titre of 1:1644 and by culture from an aspirate of a large left paravertiral mass near T9. Examination of the eye revieled anterior granulomous uveitis, vitreous inflamatory cells and posterior choroiditis. The sclera was swollen, discoloured and tender. The patients health improved dramatically after antibiotic treatment however she retained some lens opacity.

Levy (2005) reports on a 6 year old girl who had drunk unpasseurised milk and developed proven (SAT 1:160 and CSF culture positive) chronic brucellosis. Her symptoms were headache, fever, vomiting and typical menigitis symptoms. Eye examination showed optic disc edema/hemorrhaging, Her condition gradually improved with antibiotic treament so that papillrdema had gone after 1 month. He notes that papilledema is a frequent clinical feature but rarely leads to blindness.

Woods (1946) (1949) (1951) studied 200 cases of of uveitis and found that the most likely cause was: tuberculosis, 41%; syphilis, 14% and brucellosis 7,5%.Wood described torpid relapsing granulomatous iritis slowly invading the whole uveal tract and eventually ending in phthisis bulbi. He further defines brucellar uveitis as as a chronic granulomatous process with relapses and remissions. He describes charactoristic traits of brucellar choroiditis as a slightly swollen inflammatory foci without a noticable edematous reaction in the neighbouring retina with a chronic evolution. Of his 15 cases of ocular brucellosis, 8 suffered granulomatous uveitis and thiuckening of the iris, wide synechias and abundant keratic precipitates. The other 7 were free of granulomas, lacked iris thickening and had no nodules. 5 had generalised choroiditis whereas 3 had lo=calised inflammatory foci. 2 had corneal lesions (ketatitis nummullaris and vascular neoformation) 1 also had evidence of papilledema. 12 of 15 patients had bilateral infection and 11 of 15 had repeated relapses.

Puig Sonales (1953) `reviewed disorders of the optic nerve associated with B melitensis from his own work. These effects are mainly related to chronic brucellosis; Early ocular involvemet including papillary congestion, papillitis, papilledema or retrobulbar optic neuritis usually occurs at about 17 months after the start of the illness and simple or postneuritic atrophy of the disc after about 68 months. He considered that the infection of the optic nerve could occur during the early stages of the illness during septicemia from the blood stream to the neuroaxis and also that the organism could remain hidden in the tissue during a subacute or chronic phase of illness. Latter infection of the neuraxis by Brucella could be from bood or lymph relating to existing old spondylitis or from the perineural cellular tissue or directly through nerve tissue. Puig Sonales reviewed the charactoristics of disorders of the optic nerve which occurred in 44 of 413 patients in his study. Papillary conjestion is defined as changes in colour (redness of the disc with edge blurring) with demonststratabe blind spot enlargement and with vein engorgement. Papillitis represents a progression of symptoms with (mostly) periferal diminution of vision and disc edema. Retrobulbar optic neuritis could be uni or bilateral with loss of vision, peripherally, centrally, both or complete loss of vision. 2nd cranial nerve involvement could occur. Papilledema was associated in some cases with 6th cranial nerve involvement. The general impact of Brucella on the 2nd cranial nerve were: Campimetric evaluation: enlargement of the blind spot, irregular or concentric peripheral depression of the field occasional virtually complete hemianopia generally sparing of central vision. Functional effects were: a normal fundus, papillary congestion, papillitis, simple or postneuritic artrophy of the disc. In less severe cases the fundud may recover its normal aspect in severe cases there is descending atrophy of the disk.

He considered perineuritis of the 2nd cranial nerve as secondary to inflammation of meningues in the chiasmic cistern. Michel (1938) considered that nervous system oinvolvement at a variety of severity was present occurred almost invariably in brucellosis subsequent to meningeal infection. Roger (1939) found slight inflammatory products in spinal fluid in 10% of brucellosis patients. Brucellar meningitis affects the cisterna magna demonstrated by the involvement of the 9th cranial nerve. Surgery has demonstrated adhesive and cystic arachnoiditis.

Brucellosis of the hearing system.

Gladysz (1973) carried out audiological tests on 147 chronic brucellosis patients in Poland Of these 52 patients (33%) had resultant hearing loss. Usually hearing loss occurred early in the disease. In otherwise asymptomatic patients it was a possible clue to th illness.

Muszynski (1975) examined 105 chronic brucellosis patients for hearing loss. Tinnitus was very common and 16 suffered from recurrent attacks of vertigo. 76 suffered from hearing loss (12 with no direct proof of being brucellosis). In these patients VII nerve damage was the main effect. Of these 64 patients: 26 had mixed hearing loss, 34 moderate hearing loss, and 4 with severe perceptive hearing losses. Of 67 patients examined 16 showed diminished sensitivity of the labyrinths to caloric stimulation. A possible explanation for this damage to the inner ear is that Brucella toxin on the cochlear nerve.

Osuch (1975) studied 120 chronic brucellosis patients using audiological tests. Of these 120 patients 93 had hearing loss which could be attributed to chronic brucellosis. 11 patients were shown to have progressive hearing loss. In the other 82 patients treatment stopped the progress of hearing loss and in 13 some improvement was noticed.

Jezyna (1977) reviews the case of 150 patients with chronic brucellosis. Perceptive hearing loss was found in 72 patients and conductive hearing loss in 11 patients. 7 of these had moderate hearing loss and 4 severe hearing loss.

Elidan (1985) describe the case of a 40 year old female with no history of ear infection or signs of acute otitis who contracted brucellosis during her work. The illness was diagnosed 2.5 months after the symptoms developed by B melitensis culture test and SAT at 1:200.. The patient improved but then suffered a relapse after 11 months with a further positive B melitensis culture test. At this time the patient suffered progressive hearing loss in the right ear withou tinnitus of dizziness. Otoscopic examination showed an opacified ear drum but with a normal nasopharnx. A yellow sterile transparent fluid was recovered by paracentesis. The hearing loss was of a mixed type with combined damage to the middle and inner ear by a chronic fibrotic process. Further treatment with tetracycline and streptomycin improved the general health of the patient. The hearing was not recovered but was static. The general otoscopic picture was of chronic adhesive process with marked limitation of tympanic membrane movement.

Elidan notes that in brucellosis there may be considerable hearing loss either during the acute phase or the later chronic phase of the disease. Damage is usually to the inner year and/or the 8th nerve with perceptive type of hearing loss, tinnitus, dizziness and reduce response to caloric irrigation. Less common conductive loss may be due Brucella invasion of the middle ear with granuloma formation, scarring and resultant chronic adhesive process. He notes that otolaryngologists should be aware that these type of symptoms can be due to brucellosis because early intervention can prevent irreversible ear damage.

Respiratory Tract and Heart Disease.

Respiratory Tract Disease.

Brucellosis can effect the upper respitary tract in both acute and chronic illness. Typical symptoms are: sore throat, pharyngitis, laryngitis, coated tongue, hoarseness and epistaxis. Together with other symptoms this can be confused with influenza. The tonsils can be infected with Bruceela.

A cough and dyspnea are common symptoms of brucellosis which can occur at any stage of the illness. Bruce (1889) noted that in patients with Brucellosis mellitensis at least 50% of patients developed a cough. Hardy (1930) observered that 33% of patients had a cough. Mousa (1988) in a study of 379 patients with br found that 87 (23%) had respitary symptoms including dry cough, dyspnea, pleurisy and flu like illness. Of this group 54 patients had normal x-rays. Of those with abnormal x-rays: 1 patient had a solitary granuloma, 5 patients had pneumonia patches, 15 had general haziness, 9 patients had localised haziness, and 2 had pleural effusions. Respitary illness was common symptom of (neglected or chronic) brucellosis before the advent of antibiotics. Garcia-Rodriguez (1988) in a review of pulminary brucellosis notes that the use of antibiotics has reduced the incidence of the illness. Prior to this the level was 30%. He indicated the clinical manifestations are highly variable and include tracheal ulceration and pulmonary abscesses. The most common manifestation is coughing without evidence of x-ray change and examination of sputum shows nothing. He concludes that pulmonary brucellosis is usually the result of untreated chronic brucellosis. Madkour (1989) notes that a cough is generally a relatively mild symptom of brucelllosis and is probably under reported and that prior to the use of antibiotics respitory manifestations of brucellosis were common but are now rarer. A dry cough is most common symptom and in the early days occured in 58-65% of patients. He notes that the more serious manifestations of pulminary brucellosis are: haemoptysis, peeurisy, pleural effusions, empyema, progressive pulminary fibrosis, hilar lymphadenopathy, solitary or multiple nodules or military shadowing in the lung, parenchyma, dyspnea, calcified nodules in the lung and ascesss formation. Young (1995) notes that in pulmonary brucellosis occurs in the form of a cough and dyspenia in 15% of cases. The related possible forms of br are: bronchitis, pneumonia, pleurisy, nodules, lung abscesses, empyema and hilar and paratracheal lymphodenopathy. Colmenero (1996) in a study of 530 patients noted that a dry scarcely producive cough was the most common symptom. Kologlu (2006) in a study of 138 br patient (acute, 79; sub-acute, 28; and chronic, 36) found that 21% had pulminary complications. Al-Nassir (2006) noted that radiographic findings are typically absent, even in patients with prominent respiratory symptoms.

2 examples of chronic brucellosis with severe symptoms of pulmonary involvemant are described by Harvey (1948) :

• A 43 year old woman who developed a chronic cough which lasted intermittently for 1 year. Her only other symptom was occasional night sweats. X-rays showed some hilar, root branch and central bronchial thickening which , chest pain and palpitations on exerting effort and she was given a diagnosis of TB. She stayed at a sanitorium but not positive signs of TB were found. After a further year she retained the cough but also developed fever of up to 102°F. There was no dyspnea or hemoptsis. An X-ray showed changes suggestive of chronic tuberculous type of bronchopnemonia infiltrations but SAT indicated brucellosis. Opsonocytophage index and skin test were also positive but Brucella culture was negative.
• A woman of 18 is first ill in 1940 with an episode of pleurisy effusion of unknown cause. A year later she develpoed Bell's palsylasting 6 months. Then in 1942 she developed painfull lymphadenopathy persisting for a year with fever and developing respiratory symptoms including hilar lymphadenophathy with rising temperatures up to 101°F. Other symptoms were malaise, fatigue, anorexia, insomnia and muscle soreness. Also a persistant cough with yellowish sputum lasting 3 months with chest pain. The cough gradually decreased but the chest pain remained. In the next 3 years of the study her symptoms were: dyspnea on exertion, fatigue, palpitations and perspiration during light exercise, nausea sometimes with vomiting, loss of apetite a severe pain in joints. There were 3 psitive SAT tests at 1:80 and a positive opsonocytpphagic test but skin tests were negative.

Heart Disease.

Heart disease in brucellosis is widely described as occurring in about 2% of cases during acute attacks as endocartitis with vegetative degeneration of the aortic valve often leading to death usually by congestive heart failure. However there is evidence that it can also occur in a non-fatal form both in acute and chronic illness. A number of different manifestations have been noted. Bassett-Smith (1906) considered that Mediterranean fever had a specific effect on the heart. He noted that of 42 service personnel ill with brucellosis and then invalided out of military service in 11 this was due to organic heart disease. There were also others less ill with heart murmurs. Eyres (1908) in a review of 1000 patients found endocarditis and murmurs in 5% with brucellosis. Posteli (1943) studied 75 patients and concluded that there was cardiovascular involvement in 53. The main were hypotension and heart murmurs with cardiac enlargement. Moeschlin (1943) studied 34 patients by electrocardiography and found changes interpreted as myocardial damage in 11 in some caes with heart murmur. Amuchastegui (1948) studies 61 adults and 55 children with brucellosis 27 adults had signs of valvular leisons most interpeted as mitral insufficiency. 38 adults had evidence of myocardial disease. Valvular murmurs were noted in 22 children and myocardia in 31. Abnormalities disapeared over the 2 years after recovery. Maldonado-Allende (1948) in a study of 428 cases of brucellosis noted that 47 had cardiovascular involvement. 6 showed bundle brasncjh block. Panuccio (1957) studied 137 hospitalised patients with brucellosis and found 22 with evidence of mycocardia. Examples of subacute endocarditis have been described by Linn (1966) and Kast (1972) .

THe above involve patients who developped chronic heart effects during an acute attack. Perry (1958) corelated cases of patients with chronic brucellosis who developed associated heart disease.

• A woman of 64 had acute brucellosis with no apparent heart involvement and after apparent recovery was again ill 7 years later this time with with cyanois, edema and systolic murmur with possible calcific aortic stenosis. After a further 4 years the symptoms were gone.
• A woman of 26 had confirmed acute brucellosis after drinking raw milk. The illness became chronic. After 4 years of recurrent episodes with pain over the heart (precordial) and dyspnea. Examination showed slight heart enlargement and both systolic and diastolic murmur which reduced in severity over 8 years but not entirely. Rheumatic fever was ruled out.
• A boy of 15 who worked on farms developed a brucellosis type illness and was the chronically ill with joint pain and lower back pain but no heart problems for 52 years when he had a significant relapse with chills, fever, joint pain and an aortic systilic murmur. After another 5 years he had recurrent episodes of dyspnea, substernal pain including right arm and adema and high temperature with SAT at 1:160 and ECG showed complete right bundle block. The patient subsequently died.
• A man aged 43 developed brucellosis after drinking brucellosis contaminated milk. His symptoms included laryngitis, slight cough, pain in neck: upper mediastinum, shoulderes, chest and elbow with slight fever malaise and fatigue. There was a strongly positive skin test and a weakly positive SAT (1:80). This was followed by 5 years chronic illness. After this he suffered from dizziness with blurred vision and blackouts and slow pulse rate. ECG showed transient heart blocking. Examination showed aortic stenosis and a systolic murmer. During the next 7 years his health improved generally but the systolic murmur remained.
• A vetinarian aged 32 working with infected cattle and pigs developed a febrile illness diagnosed as brucellosis but not confirmed. He then had ongoing symptoms of joint pains and night sweats for 14 years, then generally well for 8 years when he was suddenly takeen ill with trembling, weakness and nausia and heart problems. 7 years later he had sudden precordial and left shoulder pain. Examination revealed an aortic systolic murmur, calcification of the aortic valve and left ventricular enlargement together with fatigue, dyspnea, substernal pain radiating down both arms. The patients health remained unchanged until he died 7 years later with evidence of severe myocardial failure.

Perry also reviewed 5 other similar case with histories typical of brucellosis and periods of chronic or subacute illness before sudden relapse with typical symptoms of brucellosis and also heart problems. Heart problems included enlargement and systolic and diastolic murmurs. Autopies indicated valvullar leisions and calcific aortic stenosis which is similar to fatal brucellosis heart disease. He notes that it is probable that mild heart infections probably return to normal.

Manchester (1942) reviewed 38 chronic brucellosis sufferers and found 10 with coronary related disorders. No valvular disease such as aortic stenosis was observed to account for angina symptoms. He considered that these symptoms corelated with bouts of illness and that temporal arteritis might be involved. Palpitations and extrasystoles were the most frequent complants.

Palpitations are also reported by Williams (1982) and Hatami and they both suggest that this involves ectopic dysrhythmia. Al Nassir notes that fever in chronic brucellosis is associated with a relative Bradycardia which can be associated with fatigue, dizziness, palpitations and shortness of breath.

Gauss (1950) notes that in chronic brucellosis heart symptoms are common manifested as tachycardia, palpitations and arhythmia.

Madkour (1989) noted that chances of survival have been increased by the use of suitable antibiotics and surgical procedures including valve replacement. Onset of this complication may occur at any time during the illness. It may be preluded by arrythmias such as supraventricaular tachycardia, artrial fibrillation or atrial or ventricular ectopic beats due to mycardia. Possible complications in brucellosis include entrocarditis, myocarditis, pericarditis, mycotic aneurysm and deep venous thrombosis. The main symptoms indicative of Brucella endocarditis are: dyspnea, orthopnoea, oedema, epistaxis, pupuric rash of skin or buccal mucosa, and embolic manifestations, neurological deficits and blood in urine.

Brucellosis of the Genito-Urinary Systems.

Hardy (1930) considering about 300 cases of B suis and B abortus noted that orchitis, mastitis, and abortion could occur. In this series 5% of males but were mild in one third of cases. This could occur at nay stage of the illness. 3% of women developed bilateral mastitis. Neither were lactating and in both cases the condition resolved within 2 weeks. Some cases of abortion have been noted. Kriistensen (1929) recovered B abortus from the placenta of a women who aborted. Simpson (1930) reported 5 abortions in women with positive SAT titres. They had histories of brucellosis type illness. Other conditions noted by Hardy were pyelitis, cystitis, nephritis, seminal-vesiculitis, and prostatitis. Not uncommon signs of illness are: frequency, painful micturition and pus in thr urine. Such infections of the genito-urinary tract are generally late onset localised infections.

Simpson (1940) notes that in cattle there is a predilection for infection of the genital tract, hence the name contagious abortion and he considers that the sametype of localisation can occur in humans: painful swelling of the testes, acute epididymitis, orchitis, prostritis and seminal vesiculitis. He also considered that abortion can occur in women living on farms and in contact with infected animals or using infected dairy products. Calder (1939) (1939) found a history of one or more miscarrages in 32% of married women, miscarrage followed by sterility, or up to 5 or 6 miscarrages. Simpson notes that Carpenter ( personal communication) cultured B abortus from an 4th month aborted human fetus. Frei () isolated Brucella from the vaginal discharge of a woman who had previously aborted.

De Forest (1917) described cases of human abortionon farms affected with B abortus.

Corbel (2006) notes that the most common genitourinary complications in men are epididymitis and orchitis which can mimic testicular cancer or tuberculosis. It has shown that banked spermazoa may be contaminated. For women infection during pregnancy can cause abortion and fetal infectioon. Rarely transfer of infection via human milk has been reported.

Simpson (1930) studied 90 patients witha cute brucellosis (mainly B abortus) and concluded that the disease of the genital tract in humans beings was similar to that in animals, cattle and guinea pigs. The perdominate feature was painful swelling of the testes. He notes that there is some evidence that the disease can be transmitted from bulls to cows via seminal fluid. 16 patients were affected in 3 of which there was evidence of seminal vesiculitis, prostrastitis, epididymitis and ochitis in 3 cases.

Simpson (1930) reported 5 cases of women who repeatedly aborted with no evidence of syphiliswith seological evidence of B abortus and a history of febrile illness of 3 to 6 years duration and drinking raw milk.

Simpson (1930) notes tha case of a woman with undulant fever from which B abortus was cultued from a tubo-ovarian abscess which developed as a late complication.

Young (1983) describes a 26 year old woman who apparently contracted brucellosis during laboratory work. Initially she had suffered from a flue like illness withfever, back pain, and chills which was not recognised as brucellosis which appeared to gradually resolve. During a subsequent pregnancy she developed vaginal bleeding, fever, a temperature of 104°F, back pain, chills and fever. A diagnosis of fetal membrane rupture wand chorioamnionitis was made. Despite treatment with clindamycin and gentamicin the women was therapeutically aborted to guard the patients life.. Subsequent blood and placenta tissue cultures were positive for B melitensis.

Examples of the isolation of Brucella from human placental and/or fetal tissue have been described by Criscuolo (1954) , Carpenter (1931) and Janboo (1939) . cases.

Eyres (1904) and Williams (1973) are sceptical as to the link between brucellosis and abortion.

Criscuolo (1954) reported levels of spontaneous abortion in pregnant women at about 10%.

Sarram (1974) noted that 11.6% of 51 pregnant women with brucellosis melitensis aborted at ≥12<24 weeks.

Figueroa (1995) describes 4 cases of brucellosis during pregnancy. All cases were treated with riframpicin and all pregnancies were successfully full term.

Poole (1972) reports tha case of a spontaneous abortion in an asymptomatic patient. The case is unusual because the patient was apparently in good health but had persistent high serology test results: eg. SAT at 1:10240; 2-ME at 320; AHG at 1:20480. In consideration of the fact that some individuals have such high Brucella serology results without apparent illness and considering possible adverse side effects, antibiotics were not proscribed. Subsequently the woman developed vaginal bleeding and aborted. B abortus was cultured from the amniotic fluid. It is suggested that the chorioamniotic tissues are infected in such cases. It is concluded that antibiotics are essential in such cases.

Hackmon (1998) notes that there is a higher incidence of abortion, premature rupture of membranes, and oreterm delivery in animals. They present 7 cases of brucellosis in pregnancy. In 1 case there was preterm premature membrane rupture and early delivery, in 2 cases there was early delivery with clinical chorioamnionitis, in 1 case there was full term delivery but preterm premature rupture of membrane and intra-uterine growth retardation and in 2 cases there was postpartum endometritis but full term delivery. 1 case delivered normally.

Makhseed (1998) studied the relationship between Brucella infection and reproductive failure expressed as abortion, intrauterine fetal death, and preterm delivery in human women in Kuwait. The studied pregnant women using ELISA, agglutination tests, and Brucella culture test: 227 women were preterm delivery, 51 were intrauterine deaths, 29 were spontaneous abortion and 39 were full term normal deliveries. Higher titres serology tests were associated with higher rates of preterm, abortion, intrauterine fetal death. Acute or chronic brucellosis was found in: 8% of preterm; 10% on intrauterine fetal death; 7% of abortions and none in the control group. Culture tests failed to find Brucella from placenta tissue of any patient. The results strongly indicate a possdible link between Brucella infection and human birth problems.

Khan (2001) review 92 pregnant women with acute brucellosis in Saudi Arabia. The rate of spontaneous abortion was 43% (≤ 24 weeks gestation), the rate of intrauterine death was 2% (>24 weeks gestation). Cotrimoxazole or trimoxazole plus riframpicin gave good protection. However if vaginal breeding even if antibiotic treatment was used this usually lead to spontaneous abortion. This study shows a strong link between brucellosis and spontaneous abortion. 4 of 7 infants tested had agglutination titres of ≥1:320 and the only infant whose blodd was tested by Brucella culture was positive. Spink (1956) points that during the bacteremic phase of brucellosis, Brucella can induce human abortion but not necessarily any more often than other bacterial infections. Khan reviewed the literature concerning other bacterial infections including Campylabacter, Salmonella, E coli and concluded on the basis of limited data that instances of abortion was much less in these illnesses and they concluded thatBrucella was much more likely to cause abortions irrespective of the severity of the infection. They considered that antibiotic treatment was essential to prevent abortion in pregnant women with brucellosis. On the basis of this relately limited study they concluded that cotrimoxazole was riframpicin were safe in such cases.

Lulu (1988) reported the cases of 35 women at ≤12 weeks of prenancy of whom 31% aborted.

Madkour (1989) found a 40% abortion rate amung 30 pregnant women with brucellosis.

Harris (1954) notes that in a review of 372 patients relative impotency is a frequent complaint in chronic brucellosis.

Opposition to the Chronic Brucellosis Diagnosis.

Brucellosos may start in an acute or incidious way. Chronic brucellosis may develop from either. An incidious attack may not even be successfully diagnosed. Chronic illness is more likely if antibiotic therapy is not used although such therapy often fails. Many patients have a history of incomplete recovery with ongoing symptoms. In many such cases during the course of this illness, which may last for a lifetime, the Brucella serology antibody titre will decline and become negative and the results of other blood tests are generally normal. It is also noted elsewhere in this article that most such patients generally do not appear ill and have no physical signs of illness but still complain of debilitating symptoms. These symptoms are generally some of the same symptoms that they suffered during the the acute illness but in a milder form. These patients are often divided up in a crude way based on serology tests despite the knowledge that such tests are know to be flawed.

• Group 1. These patients have symptoms of chronic brucellosis and retain elevated Brucella antibody titres and are judged to still have brucellosis. Some other patients have elevated tires but are asymptomatic.
• Group 2. These patients have negative Brucella serology but demonstrable residual foci of infection which can be shown to be caused by Brucella and are obviously considered to be suffering with brucellos. Some of this group are asymptomatic.
• Group 3. These patients who are the majority have the general symptoms associated with chronic brucellosis but negative Brucella serology test with no obvious signs of illness. They may have also been subjected to further unsuccessful treatment with antibiotics. The situation is also clouded by the incorrect assumption that newer serology tests such as AHG or ELISA can detect all cases of chronic brucellosis. All this makes diagnosis very difficult. Such patients may either be given a diagnosis of chronic brucellosis on the basis of common sense or be given a range of psychiatric derived diagnoses. The diagnosis that they recieve will therefore depend entirely on opinion of their doctor who may not understand brucellosis..

Group 3. From a consideration of the history of the study of brucellosis it can be seen that most expert doctors and scientists support the chronic brucellosis diagnosis despite these difficulties of diagnosis. This is an obvious commonsense diagnosis mainly based on the similarity and continuity of symptoms of illness from acute to chronic illness. There are also a small but significant number of cases of such patients who fit the above criteria but who have later been proved to have brucellosis either by positive Brucella culture tests or discovery of undiagnosed foci of infection. These issues are discussed in other parts of this report. The alternative view is put by a vociferious minority of doctors and psychistrists and accepted by those with limited knowledge of the subject. Three groups have pressed the psychiatric diagnosis lead by: WW Spink, LE Cluff et al., and EJ Young. They consider that instead of chronic brucellosis patients develop a purely psychological or emotional illness. Such alternative diagnoses include: neuroasthenia, phychoneurosis, delayed convalescense and chronic fatigue sydrome (CFS). This has created a confused situation which where there are no clear rationale on how to respond to chronic brucellosis patients. Associated with these diagnoses is a derogatory connotation that chronic brucellosis patients are mallingerers and ne'er do wells who are looking for compensation rather than do an honest days work. This whole situation is extremely destructive to patient welfare and doctor patient relationships. In the USA such alternative diagnosis have been used to attack chronic brucellosis patients to prevent them getting industrial compensation. Here we discuss the merits and otherwise of these advocated alternative theories.

Neurasthenia and Psychoneurosis v Chronic Brucellosis.

Neurasthenia was first described by Beard (1869) . The main symptoms which have been ascribed to neurasthenia are: fatigue, headaches, anxiety, impotency, neuralgie and depression, dizziness, dyspepsia, flatulence and intracranial pressure. This is remarkably similar to the much quoted generallised symptoms for chronic brucellosis. Neurasthenia was widely diagnosed in the early part of the 20th century but then declined out of use. It was originally considered as a physical illness but later was interpreted as a psychiatric disorder and became synonymous with psychoneurosis. Both were used as alternative diagnosis for chronic brucellosis. Psychoneurosis depends on psychiatric analysis. Many of the techniques used in such studies are now considered subjective and unscientific. These psychiatric techniques were used to attempt to prove patients who were considered by others to have chronic brucellosis were not physically ill but had psychiatrical or behavioural disorders. It was also strongly implied they were really mallingerers. Gauss (1950) noted that many chronic brucellosis patients were given the "psycho" or "neuro" diagnosis of neurasthenia with its connutation as a dishonourable diagnosis.

There is a close resemblence between the symptoms of neuasthenia and chronic brucellosis and it is reasonablr to postulate that many if not most of these cases of neurasthenia were really chronic brucellosis. It is also interesting to note that some cases of neurasthenia recorded in the early 20th century are described as starting with a flue like illness similar to an acute attack or relapse caused by brucellosis. Such cases are described by Kraepelin (1902) , Dutil (1903) , Savill (1906) and Oppenheim (1911) . There is little doubt that human activity involving the gathering of herds of domestic animals faciltated the spread of brucellosis in both animals and humans. Humans can also be infected by the handling of meat and the consumption of contaminated milk products. It is therefore likely that chronic brucellosis have been endemic in humans from their early history. There are many historical acoounts of a brucellosis-like feverish illness in humans and abortions in domestic animals from Hyppocrates (c400 BC) onwards. It is known that one infected cow or goat can infect many animals and humans. Batching of milk can add to the problem of human infection. Eventually this problem with milk was overcome by pasteurisation. The history of brucellosids in the USA from about 1910 is well documented. Prior to the introduction of pasteurisation of milk at the instigation of Alice Evans and against huge oppposition there is little doubt that a large proportion of the population was in direct contact with B abortus through infected milk and it was inevitable that many of these must have become infected. This was before the advent of antibiotics and a significant number (abou 30%) would have gone on to get chronic brucellos. Evidence suggests that many were misdiagnosed with neurasthenia. A number of research workers have given some idea of the scale of the problem in the USA. Melvin (1911) reported that 19% of distributed milk from dairies was contaminated with B abortus. Huddleston (1937) estimated on the basis of a survey in Michigan, USA that up to 7% of the population was infected with Brucella. Evans (1947) and (1954) considered that 6-10% of cattle in the USA at the time of report secreted B abortus in their milk and a considerably larger percentage of cattle carried the disease. With the introduction of gradual introduction of milk pasteurisation the incidence of human brucellosis declined including chronic brucellosis. This is also about the time of the decline in the diagnosis of neurasthenia. After this time brucellosis in humans was mainly associated with those in contact with infected animals, such as veterinarians, farmers and meat process workers.

A Typical Diagnosis of Psychoneurosis/Neurasthenia from the 1930s.

Psychoneurosis. A general description of psychoneurosis is given by Strecker (1935) . Psychoneurosis is described as half-way between psychosis and mental normality with partial personality altering but with a normal grasp of reality. The subject group of patients are those who complain of physical symptoms of illness but for whom a thorough physical medical examination reveals nothing. It is considered that psychiatric techniques including psychoanalysis can be used to prove that the illness is not physical by rather related to the subconscious. The patient develops a suppressed mental complex based on an accumulation of significant past experiences where there is a mental conflict due to the divergent tendencies of the mind leading to the manifestion of physical symptoms of illness. There are conidered to be three main problems areas of conflict: failure to measure up to self ideal or ego and sexual and social problems and failure in sexual and social interactions. Some typical problems which are described are: guilt due sexual affairs, excessive mastibation, shame due to mastibation, sexual inadequacy, shame due to homosexuality, fear of rejection, mother fixation and fear of failure at work. It considered that such psychosomatic illness may be triggered by an illness or trauma and may continue after the organic illness has been resolved. According to their own survey it is suggested that the predisposition to psychoneuroses are: unfavourable home life, 70%; weak constitution, 85%; chronic disease, 3-12%; sex conflicts, 3-12%; financial difficulty, 3-12%; restricted outlets, 3-12%; and mental defect, 3-12%. The most common precipitating factors are: sexual problems and relationships, 22%; accident, 13%; marital crisis, 12%; financial crisis, 11%; operation, 12%; death or illness in the family, 9%. Types of psychoneuroses described include: hysteria, traumatic neuroses. compulsive neurosis, neurasthenia, and anxiety neuroses. Neurasthenia is of most interest to us, for which a wide spectrum of symptoms is described:

• General: fatigue, loss of weight, inability to concentrate, uncertain memory, fear of instability, inferiority complex, irritability, depression, phobias, anxiety.
• Alimentary: capritious appetite, anorexia, indigestion, distention, eructation, nausea, vomiting, consipatio, diarrhea, mucous colitis.
• Circulatory: cardiac discomfort, tachycardia, palpitations, pseudo-angina sensation, heart irregularity.
• Vasomotor: pallor, blushing, sweating, coldness, heat andnumerous other phenomena.
• Genito-urinary: Impotency, nocturnal emissions,dysmenorrhea, dyspareunia, frequency of micturition, increased urinary output, "loose kidney", etc.
• Respiratory system: frequent colds, shortness of breath, sometimes hastened breathing rate and shallow breathing.
• Nervous system: peculiliar sensations in head or body, cranial pressure, headaches (especially occipital), painful uncomfortable sensation in abdomen, rectum, breasts, etc, giddiness, dizziness, insomnia, photophobia, tinnitus.
• Widespread symptoms include: backache and general aches and pains.

A Neurasthenia Diagnosis by Strecker. The illness of a 37 year old women is discussed. She was first ill as a child of 11 with severe nasal catarrh, tonsilitis, and stomach and intestinal problems with discomfort associated with food consumption diagnosised as possible tuberulosis. During the subsequent period she never felt well and had gastro discomfort. At the age of 37 she had a sudden onset of a more severe illness which lasted for more than 3 months. The ongoning symptoms were: fatigue, anxiety, depression, weight loss, severe mainly retro-orbital headaches with some distortion of sight, gastro problems (including poor appetite, nausea, bloating, left-sided abdominal pain), severe back pain, nasal congestion, and urinary frequency. She spent some time resting in hospital during which time she gradually recovered. The extent of residual illness is not give but she was able to resume work. It is construed that although the woman has some moderate physical ailments such as catarrh her problem was of psychosomatic origin. The main points of this were as follows:

• An unsatisfactory childhood with a unemotional mother and disturptive siblings.
• A fixation on her gastrointestinal problems.
• She was embarrassed due to an oral odor caused by the nasal catarrh which caused her to avoid close contact with people.
• An inferiority complex based on a failure to fulfill her potential. A subconscious conviction that she was a failure.
• Disappointment concerning her position in life and failure to find a suitable husnad leading her to live a isolated life rather than social interaction. She had been academically gifted at school but was forced to leave school to support the family.
• An association of a small ear operation and nervous fatigue leading to the psychosomatic illness of headaches, gastrointestrinal probems, etc.
• The general diagnosis was as follows: The development of a vicious somatopsychic circle - physical and environmental shortcomings in childhood giving rise to a sense of inferiority, this in turn determining an unconscious or even partly conscious wish to escape from an undesirable situation. Failing of adequate and real personal and social compensation it still urges some method of escape, without too compromising realisation of actual motives. At length, it finds such an escape in the development of neurasthenic symptoms, which not only offer a break from detested environmemntal reality and from dissatisfaction and responsibility but, also, leave intact to some extent at least, a measure of conscious self-respect. In neurasthenia the tension and anxiety remains but the energy is focussed away from the real and disturbing issues and fixed on the body.

Comment. It is difficult to take this diagnosis seriously. The dismissal of the physical illness and it replacement with a sujective psychiatric analysis is laughable. We can make a very good case that this woman was suffering from chronic brucellosis. Tests for brucellosis would be unlikey to discover the disease. The symptoms are typical of brucellosis. At that time in the USA raw milk contaminated with Brucella was widely drank by the general population and their was undoubtedly an epidemic of brucellosis. It is pobable that this woman contracted brucellosis as a child and that this developed into a chronic infection. The episode of illness at age 37 represents a typical relapse followed by a gradual recovery probably back to a chronic state of health.

Spink's Unsympathetic Review of Chronic Brucellosis.

There is little doubt that Spink was a good scientist who did excellent work but he is extremely unsympothetic to patients with chronic bruceelosis. Spink (1951) noted that chronic brucellosis sufferers had negative serology, ill defined symptoms often similar to neurasthenia, psychoneurosis and anxiety sometimes with a low grade fever ≤100°F and concluded that many of them had emotional unstability and psychoneuroses. On the basis of this view he review 65 patients whose illness had begun with a verified acute illness but had received no treatment. 41 cases were proved by Brucella culture tests and the others by epidermiologic data, clinical features, hematologic data and positive agglutination tests. Of these 65 cases: 10 recovered completely within 3 months; 25 had subacute illness with symptoms lasting up to 12 months and then apparent recovered; and 30 were ill for longer than 12 months, from 2 to 7 years at the time of testing. One had chronic illness for 13 years. These he divided into 3 groups: 5 he considered had relapsing brucellosis without localised disease based on clinical and laboratory data and he notes that they apparently recovered within 3 years; 12 patients he designated as having chronic brucellosis due to localised complications often with relapses; the remaining 13 patients complained of ill health without objective evidence of active disease including negatice Brucella serology. Unsympathetically he described this group of patients as with monotonous regularity complaining of weakness, vague aches and pains, easy fatigability, nervousness and mental depression. He subjected these 13 patients to psychiatric analysis and found that 7 were maladjusted emotionally and he designated them as severe psychoneurosis or psychopathic personalities. 2 patients were designated as emotionally unstable due to alcoholism and 2 were judged emotionally unstable because the failed to adjust to military life prior to brucellosis infection. He considered that some patients had clear-cut emotional problems prior to brucellosis illness and in other brucellosis provoked emotional instability. He suggests that incidence of chronic brucellosis was linked to the availability of compensation payments. 4 patients (31%) were appaently normal and well adjusted individuals and their illness could not be explained. A second study of 61 patients who recieved antibiotic therapy lead to 24 chronic cases of whom 12 had symptoms of illness without objective signs of illness.

Magoffin and Spink (1950) discussed 48 patients with culturally proven brucellosis with recovery times from i month to 6 years. 13 were said to recover within 6 months, 24 within 1 year, 32 within 2 years and 8 still not recovered at ≥5 years. Patients were divided into "active desease" which was defined as the presence demonstratable fever and/or bacteremia and/or incapacitating severe symptoms and convalscent patients who had symptoms of weakness, nervousness, easy fatigue, headaches and generalised aches. The main criteria for this classification was the agglutination test once this had fallen to low or negative titres patients were considered to have a postinfection syndrome because they wre unable to meet the normal demands of convalescence.

Critique of Spink's Work.

• Spink assesses the charactor of his patients by psychiatric tests. These are not specified but such tests carried out at this time were often biased and unscientific.
• The type of neurological problems which he associates with chronic brucellosis are also present in proven acute brucellosis (See: ).
• Spink describes a man with chronic brucellosis but designated as localised brucellosis. He initially had proven brucellosis demonstrable osteomyelitis which was drained surgically and thereafter he remained chronically ill for 11 years. His Brucella serology titre was initially 1:400 dropping to ≤1:40. Later in his illness he suffered abdomen pain and was subjected to exploratory operation which revealed both a pericholecystic abscess and heptic abscesses due to B suis all of which were drained surgically. Subsequently he apparently recovered. During his illness he was described as being unstable emotionally, an alcoholic, addicted to morphine and suffering periodic depression. If he has been assess prior to the exploratory operation would he have been designated as an emotionally unstable drunk and drug addict. He probably too morphine for pain relief. How many more of these cases had undiscovered internal foci of infection?
• Evans' Criticism of Spink's Work. Evans (1961) comments on the general negative attitude of Spink to his patients. She points out that Spink the scientist is willing to to accept the possibility of chronic brucellosis but Spink the doctor is antagonistic to them. She notes the following. Spink (1950) complains that chronic brucellosis patients are always complaining without showing satisfactory evidence of disease. He writes, "these tense and neurotic patients use brucellosis as a crutch." Spink (1960) concludes that, "for the most part brucellosis is an occupational disease and the interpretation of chronic illness can be the result of disability insurance." On this point Evans notes that the description of the symptoms of chronic brucellosis were the same before the introduction of compensation as after it became available. Spink (1960) also recommends that patients with acute brucellosis should be told that they have a self limiting illness to prevent them from mallingering.

Delayed convalescence.

As far as Brucellosis is concerned the term delayed convcalescence is derive entirely from the work of Cluff (1959) , (1991) . This is another theory that relates brucellosis to psychological problems rather than physical illness. The study involves 60 governmant laboratory workers who contracted B suis andB melitensis during their work. Of these 16 of the workers developed chronic brucellosis and remained ill for at least 4 to 6 years. The study included these 16 plus 8 recovered patients. 8 are termed acute-recovered, 6 chronic-recovered and 10 chronic-symptomatic. These 24 were subject to clinical and physical tests such as: EEG, neurological tests, psychological tests to evaluate intellectual functioning, tests for persistent of Brucella infection and tests for physical abnormality. Cluff concluded from these tests that they had no physical illness. They noted that there was no difference between acute-recovered and chronic-symptomatic. They observed that the symptoms of chronic brucellosis were similar to neurosis: fatigue, headache, myalgia, arthralgia, "nervousness" and depression. They then subjected the 24 to psychological and psychiatric examination. On the basis of their study they concluded that chronic brucellosis was related to emotional disturbance especially depression and that this not just secondary to the impact of the acute illness but is related to pre-illness personality effects. They conclude that the continuation of the chronic illness is related to psychological factors. The called this delayed convalescence.

Evans (1961) comments on this diagnosis of government doctors which effectively designated the 16 chronic brucellosis as malingerers because they had been suffering with psychoneurosis prior to getting brucellosis. Evans considered this persecution. She also wrote "because the clinical picture is vague, chronic brucellosis cases have often been thrown onto the clinical scrapheap of neurasthenia. This stigmatization adds to to the physical ills and mental anguish."

Young's opposition to the Chronic Brucellosis Diagnosis.

Young (1991) studied 214 cases of posible brucellosis on the basis of serology tests. The main tests used to assess the patients was based on the work of Bucheanan (1980) who used a combination of the SAT and 2-ME tests to differentiate between active and inactive brucellosis. They construed that if SAT and 2-ME were both elevated then brucellosis was active but SAT tires but if SAT remained elevated but 2-ME declines to low or negative values then brucellosis is inactive. In Young's study the patients were divided into 4 groups. 108 cases were in Group 1 who had negative SAR and 2-ME serology and were considered not to have brucellosis despite a previous history of illness or a history of exposure to Brucella including 37 vetinarians and also having general non-specific complaints and fever. Young notes an exception to this diagnosis. One patient with such a diagnosis later developed positive SAT and 2-ME titres and a positive B melitensis culture test and subsequently died. There were 57 patients in Group 2 who had 87.7% had positive SAT titres and 96.5% had 2-ME titres and were considered to have active brucellosis. These patients also had a history of contact with Brucella, ie. vetinarians, abattoir workers, farmers and laboratory workers. They had fever and variable symptoms. There were 37 patints in Group 3. In this group agglutination titres were mainly positive but lower than in Group 2. For SAT 75.6% had titres of ≤1:80 and 24.4% had titres of ≥1:160 and for 2-ME 83.8% had tires of ≤1:40 and 16.2% had tires of ≥1:80. Titres declined during follow-up of up to 31 months. These patients are reported as not feverish but having non-specific symptoms such as fatigue, backache and joint pain. They are described as having no particular history of prior infection and having less contact with animals than Group 2 patients although 45.9% are vetinarians. On this basis thay are designated as brucellosis inactive. There were 12 patients in Group 4 with positive agglutination but lack of historical or clinical information. 8 had SAT of ≤1:80 and 4 of titres ≥1:160. 2-ME was positive in only 2 patients. These patients were designated as unclassified.

Young (1990) he writes that in some cases workers compensation or litigation appears to be the major stimulus for the prolonged complaints.

A Critigue of Young's Work

• Young is antagonistic to the diagnosis of chronic brucellosis. He highlights the work of those who suggests alternative diagnosis to chronic brucellosis including Cluff (1959) and Spink (1951) but denegrates or ignores contrary opinions. Criticism of the work of Spink and Cluff is given above.
• He is also antagonistic to chronic fatigue sufferers. In the above paper he notes that if patients have negative serology tests and no signs of a persistent focus of infection n they appear to suffer from psychoneurosis and a diagnosis of chronic brucellosis only serves to perpetuate a myth. Similarly Young (1995) writes that 66% of patients with chronic complaints (brucellosis) are without discernable localised brucellosis and objective signs of infection suffer from a form of psychoneurosis and he considered that the availability of workers compensation was a stimulus for continued complaints of ill health. Also Young (1989) he writes that such patients with chronic complaints (brucellosis) suffer from emotional disturbance which they will not accept and require psychotherapy. Again he notes that workman's compensation or litigation appears to be the major stimulus for prolonged complaints.
• Young's designations of brucellosis as active and inactive based on the use of SAT and 2-ME lacks credibility. Other workers have found that the 2-ME test is unreliable and insensitive and because of this falls off rapidly over time compared with other tests such as AHG. It is probable that 2-ME only measures part of the IgG but not IgM or IgA.
• Most other workers have concluded that SAT titres fall to low or negative values over time, usually around about 1 year from initial infection but remains elevated in a minoritory of cases and consider that it is usually negative in chronic brucellosisd. Young does not believe in chronic brucellosis so he disputes this. He considered that all these other workers had negative serology tests because they did not use standard antigens in their tests. He quotes the work of Spink (1956) "the agglutination test is a reliable procedure in approaching the diagnosis of diagnosis of brucellosis. When properly standardised brucella antigen is used the great majority of patients with bacteriological proved disease will have a titre of 1:160 or 1:320 or above." General information indicates that the use of standard antigens has not altered the results and it is difficult to take Young's opiunion seriously. The general view is that SAT titres rise and then falls over time it mainly tests IgM which peaks and the falls early in the course of brucellosis.
• Young suggests that exception where negative serology test are combined with positive culture tests or unsuspected localised Brucella lesions are rare cases which can be discounted. Also they do no fit with his ideas. His position is undermined by similar results in animals found post mortum and recent papers detailing positive PCR Brucella results in chronic and convalescent brucellosis patients.

A Group 3 cases described as agglutination positive and likely inactive brucellosis. In September 1984 a 28 year old abattoir worker was diagnosed with brucellosis on the basis of typical symptoms and positine SAT titres of 1:320. He received treatment of tetracycline for 4 weeks with some improvement but residual symptoms of fatigue, malaise, lethargy, and decrease libido continued. In November 1984 he received further treatment of 3 weeks tetracycline and 2 weeks streptomycin. This did not produce any improvement. In December 1984 he receive a further 8 weeks treatment with tetracycline which also failed to alleviate the symptoms. By March 1985 his SAT had declined to 1:80 but the patients illness continued and he could not work. In August 1985 his SAT was 1:80 and 2-Me 1:40. No further treatment was given. In March 1986 he applied for brucellosis compensation and disability payment. He is described as health looking with no disabilities. He is denied compensation and offered job retraining and psychological counselling. By Sepember 1986 his SAT was 1:20 an 2-ME was negative.

Analysis of this Case. The initial treatment is a mono-therapy and is too short. These type of treatments have a history of failure and relapse. The second treatment is suitable but also too short. The treatment should have been continued until the patient was completely free of symptoms. Early effective treatment are essential for success in the treatment of brucellosis otherwise complications and chronic illness can ensue. Young implies that this patient is a mallingerer feigning symptoms to get compensation but would he of planned it from the early stages of his illness. No detailed account of the man's symptoms is given. Suggesting that these are considered unimportant. The clinician states that the patient looked well impling mallingering. Williams notes that chronic brucellosis patients may look deceptively well.

Corbel (2006)

Corbel (2006) describes chronic brucellosis as existing in three forms: relapsing, chronic localized infection and delayed convalescence. Corbel's postulation appears to drawn from the work of Spink, Cluff, Young (See above). It is interesting to note that the illness of Alice Evans could be fitted into all 3 catagories (See below).

Criticism of Corbel's Work. Criticism of the work of Spink, Cluff and Young has already been discussed. Some specific points are:

• He defines delayed convalescence as a persistence of symptoms without objective signs of infection such as fever. This is wrong becsause many chronic brucellosis suffers have a low grade fever with a sightly elevated temperature and related chills and sweats.
• Corbel states that the delayed convalescence diagnosis is based on psychological studies. In fact psycological and psychiatric studies were used. We have to question the viability of such studies in consideration of modern ethical standards. For examople Cluff compares chronic brucellosis sufferers to 2 boys who both fall down and hurt their knees. On boy stoically accepts the pain but the other runs crying to his mother. Is this science?
• This document is issued in the name of the World Health Organisation. These psychiatric based theories are extremely dubious and not generally accepted. Why was a balanced view not offered?

Chronic Fatigue Syndrome (CFS).

There is no such thing as a positive diagnosis for CFS. The diagnosis is based on three factors: elimanation of other possible illnesses or diseases with similar symptoms; normal test results for a range of blood and urine tests and a typical range of symptoms. Amung alternative illnesses which could give similar symptoms are: hypothyroidism, sleep apnoea, nacrolepepsy, unresolved illnesses such as hepatisis B and C, mental disorders such as severe depression, bipolar disorder, schizophrenia, dementia and anorexia, and alcohol or substance abuse. The general blood and urine tests used are: blood counts, ESR, CRP, alanine aminotransferase, total protein, albumin, globulin, alkaline phosphatase, calcium, phosphate, glucose, blood urine nitrogen, electrolytes, creatinine, thyroid stimulating hormone, and urinalysis. A typical range of accepted symptoms includes: impaired memory and concentration, sore throat, tender cervical or antillary lymph nodes, myalgia, joint pain, headaches, unrefreshed sleep and post-exertion malais. A range of other symptoms is often included. Attempts to link CFS to diseases such as mononeucleosis have failed. All patients who have this negative diagnosis are told they have CFS. CFS as a diagnosis is rejected by many doctors and disliked by patients. Some doctors equate it with mental or psycological illness and just like neurasthenia and psychoneurosis it connutations of mallingering. It is noted elsewhere in thia report that chronic brucellosis patient often have negative serology results for Brucella and normal blood tests, no other illnesses and symptoms somewhat like CFS so they can easily fall into the CFS trap. The criteria of diagnosis given above cannot differentiate between the 2 illness and few doctors understand this situation. It is therefore likely that there are many chronic brucellosis patients who have been given CFS as a diagnosis. This shows a serious fault in the CFS diagnosis.

The alternative view has been widely circulated particularly in internet articles that chronic brucellosis is really CFS although there is no evidence to back this up. Such a thesis has been proposed by Straus (1991) . He reviewed a range of illnesses including: neurasthenia, DuCostal's (effect) syndrome, chronic brucellosis, hypoglycemia, total allergy syndrome, chronic candidiasis and chronic Epstein-Barr Virus infection and concludes that they are all manifestations of CFS. In the case of chronic brucellosis he briefly discusses the work of Evans which he missunderstand and misinterprets and then looks at the work of Spink (1951) and Cluff (1959) which are both analysed above in more detail and states that the work of Spink and Cluff has ended speculation about chronic brucellosis because they have proved that infection is absent and therefore he concludes that the continuing illness must be chronic fatigue. He also states that on the basis of his analysis that chronic fatigue can be considered free from consideration of microbial or pathogenetic hypothesis. This is a shallow piece of work because he seems unaware of the general contraversy and problems associated with the diagnosis of brucellosis most of which predates his paper including information supplied in this study. Nor is he aware of the faults in the work of Spink and Cluff which we have also discussed. As far as chronic brucellosis is concerned his ideas are wrong. Unfortunately for chronic brucellosis sufferers these type of ideas have been widely quoted and have been accepted by physicians with limited knowledge of brucellosis.

Previously we have noted that many patient diagnosed with neurasthenia must really have had chronic brucellosis. Here we have noted that the standard method of diagnosing CFS cannot distinguish it from chronic brucellosis. In Brucella endemic countries this is likely to be a significant problem. In developed countries where brucellosis has been effectively eliminated the main source of infection is travel to endemic countries and consumption of imported contaminated foods often sent by family friends. Brucellosis is poorly manage in developed countries because of the limited experience of doctors so chronic brucellosis is more likely to develop ans subsequently be designated as CFS. This may be particlularly the cases of brucellosis occurs as a low grade insidious infection likely to be missed in its early stages and therefore have negative serology tests.

As a conjecture we might also consider if significant numbers of CFS patients have chronic brucellosis by human to human contact. In cattle it has been shown that cows with chronic brucellosis pass the disease on to their calves. In both cases serology testa are generally negative and illness is not apparent. Cattle cannot of course describe chronic symptoms. Could this also happen in humans? There is little doubt that sub-acute and chronic brucellosis was widespread in the 1930-70 in the USA and Northern Europe before milk pasteurisation and cattle herd management to eliminate brucellosis. Could women infected in this way have unknowingly passed the disease to their children who when typical general symptoms occurred have been subsequently diagnosed as having CFS?

It is interesting to note that Evengard (1999) in a review of chronic fatigue syndrome suggests that neurasthenia was really CFS and the decline of the in the frequency of the diagnosis he ascribes to alternative increasing sophistication of psychiatric diagnosis such as obsesssive-comp[lusion disorder, anxiety, neurosis and hysteria. We would ascribe the decline to the increased use of pasturised milk and the consequential decreased contact with B abortus. Evengard quotes the work of LLoyd (1993) who considered that CFS often follows a recognized or presumed infection and the disorder may therefore result from a disordered immune response to a precipitating infection or antigenic challenge. Such agents of disease include bacteria (Brucella), richettisial organisms and protozoal infection (toxoplasmosis and giardiasis). The imputation is that the chronic illness that often develops after a Brucella infection is not chronic brucellosis but rather CFS as a sequella.

Afari (2003) discusses clinical presentation of CFS. He notes that patients often report excellent pre-illness health, MacDonald (1996) with an abrupt flue-like illness onset, Salit (1997) and Schluederberg (1992) . After this illness patients have chronic fatigue syndrome, anorexia, nausea, drenching nights sweats, dizziness and intolerance to alcohol and pharmaceuticals that affect the CNS, . Patients suffer social depravity, 0.33% are unable to work and 0.33% can only work part-time. long term studies indicate that 17-64% of patients with CFS improve but only 10% recover and 1-20% get worse, Vercoulen (1996) Ray (1997) and Bonner (1994) . Factors predicting a poorer outcome, Joyce (1997) include: older age, longer illness duration, fatigue severity, comorbid phyhiatric illness, and a physical attribution. Children and children seem to recover more quickly.

Buchwald (1995) in a survey on CFS by questionaires of 4000 people in Washington State, USA notes that 0.07-0.2% of people in western counties suffer from CFS.

Hagebusch (1941) in a review concluded that there was a significant incidence of brucellosis in children but that the illness was milder than in adults with non-specific symptoms and apparent spontaneous recovery so that the illness was often missed.

Saddler (1960) notes that because of the confused relationship of symptomatology to infection and inadequacy of diagnosic methods the number of proven cases of brucellosis may only be a small proportion of the total number with many subclinical and undiagnosed cases. Stoenner (1949) suggested a ratio of 1:26 and Magoffin (1949) 1:5. He notes that there is widespread reference in the medical literature to asymptomatic "normal" individuals with either positive Brucella agglutination results or positive Brucella cullture tests. He also suggests that low dose infection by Brucella may result in low grade illness which goes unrecognised.Also Humphreys (undated) points out that 12% of proven cases of brucellosis have no obvious epidemalogical cause. He notes that B abortus var 19 vaccine used to suppress brucellosis in cattle is virulant to humans and causes chronic illness and has been found in milk.= of vaccinated animals.

• Blood Transfusion. This a potential route of infection in humans. As little as 5 organisms of B suis or B melitensis injected into guinea pigs can cause disease. Screening methods depend on antibody serology tests which are less than efficient. (See )

• Milk Cows, goats and sheep which are infected with Brucella secret it in there milk. ch The consumption of contaminated milk and milk products by humans is a main source of illness. Such disease in both humans and animals can go undetected. Raw milk is still widely drasnk even in developed countries. (see )

• Meat contaminated with Brucella has been recognised a significant cause of brucellosis. Those most likely to be affected are abattoir workers. Others at risk are butchers, restaurant workers and those handling meat in the home. Brucella can persist in meats for weeks. Lymph nodes which are present in meat and meat products can be highly contaminated. Infection is most likely through abraded skin or trhrough areisols. (see )

• Live attenuated Brucella vaccine. Live attenuated vaccines have been used to control and erradicate brucellosis in cattle, sheep and goats. Mainly B abortus Strain 19 in cattle and B melitensis strain Rev 1 in sheep and goats. These vaccines cause acute and chronic brucellosis in humans. The vaccine can be present in both milk and meat. (See ) Luchsinger (1963) showed that calves vaccinated with B abortus strain 19 could develop a persistent infection.

• Brucella canis. Brucella canis affects mainly dogs but can be transmitted to human because of the close association of dogs with humans. B canis is often a mild illness in humans and is disfficukt to detect. It is likely that many cases go undetected. The illness is chronic in dogs and may be so in humans. (See )

• Vertical transmission. from mother to offspring has been proved to occur and then persist in the off springs in guinea pigs, mice, goats, sheep, mink, cattle and humans (See ). In cattle this can occur in chronically ill cows (carriers). THe disease in such calves is almost undetectable but Brucella Can be secreted in milk later in life. Brucellosis in children is often mild and can go undetected.

• Wrong Diagnosis. Patients who have had actute brucellosis may subsequently develop chronic brucellosis and because of negative serology data be misdiagnosed as having CFS.

• Wrong Diagnosis Patients who have had no acute phase of illness and without a diagnosis may develop chronic brucellosis without any diagnosis.

• Misuse of antibiotics. During the 1950-90 antibiotics were commonly used for any form of illness. This could have partially cured brucellois leading to a chronic form of the illness.

Criticism of Psychiatric Alternate Diagnoses.

Evans's Support for the Chronic Bruccellosis Diagnosis.

Alice Evans was a famous contributor to the scientific understanding of brucellosis and she was strongly opposed to the neurasthenia and psychoneurosis diagnoses in chronic brucellosis. She herself suffered from chronic brucellosis and was given the neurasthenia diagnosis. She therefore had a very clear understanding of the issues involved. Evans (1947) describes her struggles against the neurasthenia diagnosis before she was proved to have brucellosis by chance. She caught B melitensis although it was undiagnosed and had a relatively mild illness for 9 months. This was designated by doctors as neurasthenia. At the end of this period she had an acute attack (a relapse) with a positive Brucella culture. Thereafter she suffered 5 years of poor health during which she was often incapacitated. During this time doctors at 4 different hospitals again diagnosed her as having neurasthenia. At the end of this 5 years she then had a gall bladder operation which showed lesions found to contain Brucella by culture. She was ill for at least 20 years with a variable illness with periods of severe incapitation and periods of partial or complete almost complete recovery . She makes it clear that these false neurasthenis diagnoses had a profoundly negative impact on her life. She says that this second positive Brucella culture test thus accidently relieved of the stigma of imaginary illness and she considered that such misundestandings try the patient almost beyond indurance. 2 similar examples are provided by Schinger (1960) . In one case a Brucella granuloma was detected in an entirely unrelated renal operation and in another the only evidence of abnormality was an asymptomatic nodule detected on an X-ray.

• Evans (1961) reviewing the papers by Spink and Cluff writes that they apply the verdict of guilty by feigned illness unless proved innocent, or a diagnosis of neurashenia continues to be the fate of sufferers of chronic brucellosis. She notes that a diagnosis of neurasthenia is seen to imply malingerer and that chronic brucellosis sufferers analysed in this way are treated as imposters trying to get compensation to which they are not entitled.

• Evans (1961) points out that neurological problems are associated with acute and chronic brucellosis citing the work of Roger (1938) who indicates that central and peripheral nervous system involvement can dominate the clinical picture.

Harris and Chronic Brucellosis Diagnosis V Psychoneurosis.

Harris is a strong supporter of patients with chronic brucellosis even if it occurs without positive serological and interdermal Brucella tests and recognises the difficulties such patients have in getting a correct diagnosis. However at the same time he is also supports the diagnosis of psychoneurosis and the related psychological/psychiatric techniques. He acknowledges that it is very difficult to diferentiate between chronic brucellosis and psychoneurosis because chronic brucellosis has protean symptoms and that some of these closely resembled psychoneurosis. He concludes that many chronic brucellosis suffers have had their illness wrongly designated as neurasthenia/psychoneurosis. Harris (1955) studied 370 patients with chronic brucellosis. He concluded that in most cases the known symptoms of acute and chronic brucellosis were so similar to those projected for neuroasthenia that it was not possible to differentiate between them. However he concludes that by painstaking work mainly using psychological/psychiatric techneques he make a tentative judgement between them. He appears to be trying to please two groups with diametrically opposing views. His methodology which is discussed in detail is crude, unreliable and unscientific.

Harris (1950) describes the difficlty facing the doctor in differentiating between chronic brucellosis and psychoneurosis as follows:

• The diagnosis of brucellosis cannot easily be be established or ruled out.
• Chronic brucellosis often gives a textbook picture of neurasthenia.
• Chronic brucellosis and related or unrelated psychogenic illnes can coexist.
• Emotional disturbance may be precipiated or aggravated by chronic brucellosis and such disturbances may prevent recovery the illness.
• If brucellosis and psychological illness co-exist then it is likely that a physician would diagnose brucellosis and a psychiatrist will diagnose psychoneurosis.
• A patient recovered (?) from brucellosisbut with residual symptoms may attribute all symptoms to chronic brucellosis and are unwilling to accept a diagnosis of psychoneurosis.

Support for Chronic Brucellosis.

Harris (1943) notes that neuritis is the most frequent erroneous diagnosis made in chronic, afebrile brucellosis or in patients with low grade fever and complaints such as fatigue, weight loss, joint and muscle pain, headache, mental confusion, backache and the like. Often patients are seen by doctors who offer this diagnosis while failing to do appropriate tests and are only diagnosed when they develop objective symptom or have a significant relapse.

Harris (1954) and (194) describes a case where a female patient is given a diagnosis of psychoneurosis but is subsequently shown to have chronic brucellosis. An ill Navy WAVE was initially diagnoseed as having brucellosis on the basis of a positive SA test with a titre of 1:160. She remained ill with many symptoms including: fatigue, myalgia, arthralgia, pelvic pain, weight loss, and low grade fever for three years at which time her SAT test were very low or negative. Skin test was positive. Cultures test were also negative. She was seen by various Veterans Hospitals and was given diagnosis of psychoneurosis or mallingerer based on analysis of her behaviour. Five years from her first visit B melitensis was isolated from her blood. As the illness worsened she suffered from psychotic episodes, manic behaviour and paraniod ideas, a fever and skin nodules consistent with her illness. Response to antibiotics was temporary.

Harris's Psychological/psychiatric Techneques.

In consideration of his opinion Harris developed the following stategy to differentiate between suffers of brucellosis and thos with psychological problems. The methodology was as follws:

• Interview relating to current complaints, history using sodiumamytal as a true serum.
• Tests to determine the Cornell Selective Index (CMI).
• Projective psychological studies carried out by expert psychologists including: the Rorschach test, figure drawing, story association with figures, hand writting analysisthematic, apperception tests and sometimes intelligence tests (Belleevue-Weedsler).
• Evaluation of psyochogical, clinical and laboratory data and an attempt to choose between sonatic and psychological problems or determin which is most important.
• Treatment then depands of decision reached:

• Brucellosis not complicated by psychological problems. Vaccine treatment.
• Psychological problems. Treated by support measures and possibly reference to a psychiatrist.
• Long term follow-up of both psychological and laboratoy tests as appropriate.

Harris quotes the work of Bowman and Ruesch (1946) who make a link between chronic disease and psychological invalidism. It is suggested that patients equally ill with the same disease react in different ways, some recover while others becomes invalids. It is considered that if a patient desire a quick and complete recovery this will happen but in other patients the symptoms continue because of a subconscious need to dominate or punish others, to gain presige, to evade unpleasant duties, to obtain compensation or to achieve some other desired goal. Therefore chronic disease is a psychosomatic problem based on a pre-morbid personality. Thus during convalescence complaints and symptoms are in part the result of the individual's deliberate selection and re-enforcement of peripheral stimuli and the concurrence of psychological invalidism and chronic illness may prolong or postpone recovery indefinately. Ruesch notes that chronically ill patients who are perpetually referred to doctors (such as with chronic brucellosis patients) go to the doctor to be treated but few want to get well rather they look for reassurance in relation to fears and anxieties, they enjoy the association with the "glamoorous" doctor, or are hypochondriacs.

The Cornell Medical Index (CMI). The usefulness of CMI has been reveiwed by to Abramson (1966) and (1965) . It comprises 195 simply worded questions with yes/no answers of which 51 deal with mood and feeling patterns. The distribution of questions is: eye and ear, 9; respitary system, 18; cardiovascular system, 13; digestive tract, 23, musculoskeletal system, 8; skin, 7; nervous system, 11; fatiguability, 7; frequency of illness, 9; miscellaneous diseases, 15; habits, 6; inadequacy, 12; depression, 6; anxiety, 9; sensitivity, 6. Diagnostic interpretation of the test results are dependant on the doctor's own knowledge and opinions. A typical way to use the test is a tally of yes scores. A tally of ≥30 is considered abnormal for the whole test and ≥10 fot the mood and feelings questions alone. According to Brodman (1952) it can be used to identify hostital patients with emotional problems. Abramson (1966) correlated data for different groups of people using the ≥30 cut-off. The number of people showing ≥30 results were as follows: patients with a history of emotional disturbance > hospital patients with no history of psychiatric illness > healthy people. The CMI is not of great value in the study of specfic disorders or of general somaric health. It is best used as an epidermiological tool rather than an aid to diagnosis. A significant limitation is that results vary according to culture and age, Brodman (1953) . In consideration of Harris's use of the test to evaluate chronic brucellosis patients we note the following: Common sense suggests that any chronically ill patient with ongoing debilitating symptoms would suffer some degree of anxiety and depression. Also anxiety, depression and irritability have widely recorded as symptoms of brucellosis in all forms of the illness. Positive results from CMI are therefore effectively inevitable.

Projective Test are derived from Freudian Psychoanalysis and proport to find hidden emotional conflict. THe tests used by Harris include the Rorschach test, figure drawing, story association with figures, hand writing analysisn and thematic apperception. They have individually and collectively been widely critised because of they are lacking in scientific merit, they are unreliable in use and they rely heavily on subjective clinical judgement.

Comments on Harris Work.

Harris is well meaning but misguided. His attempts to differentiate between chronic brucellosis do more harm than good. The techneques are crude and unscientific and could not work. It is noted from his various papers and book that patients usually reject the psychiatric diagnosis and remain convinced that they have chronic brucellosis. Who could blame them.

General Comments.

The central dogma that supports alternative psychiatric diagnoses for chronic brucellosis patients is that they cannot have brucellosis because they have negative brucellosis serology tests. A secondary argument indicates rthat they cannot be ill because they have no obvious clinical or physical signs of illness. Throughout this text we have given information that supports the chronic brucellosis diagnosis. The following are some key points:

• Negative serology tests. In patients with acute brucellosis serum antibodies test reach a peak and then decline towards negative levels at about or after 1 year irrespective of if they are cured or not. Carpener notes that false-negative results occur in brucellosis depending on the antibody response which relates to thew stage of the illness. Carpenter (1975) and (1979) . In this situation some patients with negative antibody serology tests have positive Brucella culture tests which prove their illness (See Text ). Such cases are relatively rare buts significant. There are also similar cases of chronic brucellosis in animals with negative serology results but positive Brucella culture tests carried out post mortem. There is also recent evidence that patients with chronic brucellosis or even apparently cured patients retain specific Brucella DNA yearts after their first illness. It is also of note that patients with proven brucellosis have the similar symptoms to those otherwise designated.
• Localised brucellosis. This means that Brucella infection occurs at a focal area. This has been shown to occurs in chronic brucellosis. In some cases the foci is obvious such as an infected knee but otheres it is internal and difficult to find. In some cases they have only been found by chance. Often such patients have typical symptoms of illness but negative Brucella serology data but Brucella can sometimes be extracted from the infection focus (See ). Young suggests that these examples can be dismissed because they are rare. One of the reasons they are rare is that their discovery depends on surgery which is not usually an acceptible procedure in such cases. However Brucella has been recovered from the internal organs such as the spleen in animals with chronic brucellosis.
• Lack of evidence of illness. It has generally been accepted that chronic brucellosis can exist without obvious signs of illness. Amongst those proffering this view are: Dalrymple-Champneys (1960) , Janbon (1953) , Castaneda (1953) . Others are given in the text (See )
• Scientific studies in humans, mice, cattle amd other animals have shown that Brucella can impair multiple functions of the immune system ensuring its own survival as a chronic illness. This many involves establishing itself as a intracellular parisite where it can survive and multiply as a chronic illness.
• Depression. Neurological involvement in brucellosis is the norm including the CNS. Associated with this are neurological symptoms including depression. This has been widely reported and occurs in all phases of the illness; acute, sub-acute and chronic. Therefore using psychiatric analysis to prove depression is a self fulfilling prophecy. Clearly patients with depression will give negative answers about their lives past and present.
• Persecution. All these psychiatric diagnoses for chronic brucellosis are associated with denigration of the patients charactor and behaviour. One reason for this is that such techniques were used to try and prevent chronic brucellosis patients getting industrial compensation. In these discussed medical papers they are accused of being dishonest, hypocondriac, neurotic, mallingerers, weak charactors, drunks, and psychopaths.

Schirger (1960) noted in a study of 224 patients at the Mayo from 1940 to 1958 that he did not consider chronic brucellosis because he considered the condition emotional rather than brucellar disease.

Hypersensitive Allergic Reaction.

It has been suggested that people who are exposed to repeated challange by Brucella may develop a allergic reaction which simulates the symptoms of brucellosis and that this is the cause of chroonic brucellosis. Such allergic reactions have been observed in vetinarians repeatedly in contact with contaminated birth material. However in Western countries where people have hardly any contact with Brucella chronic brucellosis occurs.

Huddleson (1930) studied 49 veterinarians and found that 28 (57%) had positive SAT titres (≥1:50). 3% had a history of illness. Many with positive SAT developed skin rashes on their arms when they removed retained placenta fron aborting cows. They also developed a general brucellosis type malaise. A skin test with the same antegen produced the a similar result. THis seems to suggest that they had suffered a subacute brucellosis infection and that they had acquired immunity due to frequent challenge by Brucella. The response was described as an allergic reaction.

Treatment of Chronic Brucelosis.

Limitations

In this review of the treatment of chronic brucellosis children of <8 years and pregnant women is excluded. In such cases some antibiotics cannot be used due to adverse effects. It is generally considered that children seldom develop chronic brucellosis.

History of treatment.

There was no effective treatment for brucellosis until the introduction of antibiotics. Some of these earlier treatments were: quinine proposed by Bruce in 1889, mercurochrome used by Simpson et al in 1929, and sulphamilamide used by Smith, et al. in 1939. Various forms of vacination were used tried with different opinions on their usefulness. This is discussed below. Penicillin, the first antibiotic, was found by Haden to be ineffective in brucellosis. Streptomycin was first used in 1947 by Pulaski (1947) with variable but limited sucess. Pulaski and also Eisele (1947) obtained good results using a combination of streptomycin and sulfadiazine but alough patients recovered there was a high relapse rate. The first tetracycline, chlorotetracycline (aureomycin) was used by Spink (1948) to treat 24 patients with significant success. The combination therapy of chlorotetracycline and dihydrostreptomycin was found to be successful by Heilman (1949) in the control of brucellosis in experimentally infected mice. Herrell (1949) and (1952) used this combination with considerable success in humans and this laid the foundations of modern antibiotic treatments. A problem with hydrostreptomycin is delayed hearing loss as described by Allison (1949) therefore streptomycin was preferred.

Vaccine Treatment.

Vaccine have been used in Brucella infection in two ways. Vaccines have been developed to protect animals from infection and in humans vaccines treatment has been developed to treat chronic brucellosis.

Vaccine treatment in humans.

Attenuasted vaccines successful for animals have proved dangerous to humans. Veterinarians who have acidently injected themselves with such vaccine have developed both acute and chronic illness. At this time there is no anti-Brucella vaccine for humans. However vaccine derived from killed Brucella have been used as a treatment. In this case the vaccine is used as a treatment to stimulate the bodies own defences.

Vaccine treatment was the best treatment for acute and chronic brucellosis prior to the use of antibiotics. It was used in the 1930-50s. It is similar to the inradermal skin test. Different methodology in its application have it difficult to assess. There is little doubt that it could inprove the health of some patients but there is some doubt about the perminence of such improvements.

The first attempt to use dead brucella as a vacinne was by Wright . In 1929 Simpson treated 29 patients with a course of 3 injections of a vacinne of killed Br abortus A rise of temperature and exacerbationof symptoms and they considered that the treatment inducing favourable results. De Jong (1936) treated a patient with Br melitensis meningitis with a vacinne derived fron B suis ans melitensis but saw no evidence that it worked. The use of vaccines was advocated bt Harris . Brucella abortus vaccine was used. The idea was to stilulate the activity of the immune system and was measure using the opsonocytophagic test. The viability of te tratment was disputed.

Griggs (1943) notes that patient response to vaccine treatment is difficult to assess but a clear can be obtained even if complete cure is not achieved. He treated 91 patients and found: 83 patients improved and 8 patients unimproved. Of the 83 inproved patients he subjectively assess their improvement as follows; uNIMPROVED (Neither better or worse overall. Some cases were alternatively better and much worse), 8; Fair (Fair (Some definite improvement but no complete relief of symptoms for significant periods), 7; good (Better in every way, upto 80% improved), 37; very good (Marked improvement or loss of disability or relatively long remission of symptoms), 24; excellent (apparent rapid recovery for a number of months with the possibility of a permanent recovery), 15. A problem with the treatment is skin leisons (necrosis) which are more likely with killed Brucella vaccine and higher concentrations of Brucellergen. Such leisons heal but increase sensitivity and make further vaccine theraoy very difficult. Also increased sensitivity is concomitant with increased severity of symptoms. This can partially be negated by lower concentration of the vaccine. In this series necrosis occured in 10 cases and near-necrosis in 9 cases. Griggs (1948) reviewed the effectiveness of vaccine treatment in 133 chronic brucellosis patients after 1 to 6 years (most from 2 to 5 years) based on questionaires. This showed: improved, 112; unimproved, 18 and worse, 3. Of the 18 patients with no improvement: 5 were hyposensitive and discontinued treatment; 8 had a less certain diagnosis and complications; 3 had a probable diagnosis and complications. Of the 56 patients with a definite diagnosis of chronic brucellosis, only 6.6% were unimproved and of patients with a probable diagnosis (small doubt), 17% were unimproved. The views of the patients on their own illness were: complete recovery, 13.8%; very good result with marked improvement, 27.6%; good result with general inprovement (60% improvement), 28.4%; fair result with some temporary or incomplete improvement, 13%. The study showed that the most important adverse factors to improvement were: age, 89% of the recovered group were under 50; duration of illness, ineradicable foci of illness or complications and hyposensitivity to the vaccine. The average length of vaccine treatment was 1.5 years. He offers an interesting analysis of the symptoms most likely to be improved by treatment: weakness fatigue, 49; athralgia, myalgia and neuralgia, 42; cheat pain, 3; nervousness, irrtability depression, 16; insomnia, 4; fever, 14; general health, 14; intestinal distress, indigestion, diarhea, colitis, consitpation, peotic ulcer symptoms, 14; headache, 9; dizziness, 5; palpitations and related, 9; flu-like symptoms, 7; urinary infection, cystitis, pelvic pain, 6; night sweats, 4; weight loss, 3. General fatigue and rheumatic symptoms are the most likely to be improved and intestinal, and headaches and dizziness are more stubborn. Relapse was reprted in about 50% of patients but could be reversed by fyrther treatment. He notes that chronic brucellosis is similar to tuberculosis, syphilis and leprosy in that it cannot be cured easily. once established it it resistant to antibiotics but vaccination helpds the host to develop their own resistance.

Spink (1953) notes that according to the work of Janbon a combination of antibiotics and vaccine therapy is best in chronic brucellosis. The vaccine treatment depends on inducing a febril state. About 4 to 6 injections are given several days apart and seem to manifest considerable improvement.

Leon (1947) used repeated injectins of formalen killed B abortus and B melitensis at dose injections of 10,000, 100,000 and 1,000,000 bacteria started no earlier than 6 months from the onset of the illness as a means of treatment of brucellosis. Repeated doses of increasing bacterial concentration were associated with desensitisation and this was associated with recovery in about 5 months. 15 patients are discussed. 5 were desensitised and recovered when 16 to 19 injections were given. 7 received 8 to 18 injections and recovered without full desensitisation. 3 improved but did not complete treatment. 20 patients who started treatment before 6 months and therefore had progressive allergy all improved.

Comment.

With the advent of antibiotic treatment, vaccine tretment has fallen out of use, Although antibiotic treatment works reasonable well in acute brucellosis it is of more limited value in chronic brucellosis. In this case vaccine treatment could still offer a viable alternative. Se also boura below.

Antibiotic based treatments

The antibiotic treatments used in acute, subacute and chronic brucellosis are generally the same. The scientific considerations for these choices are reviews below. Scientific support for the use of antibiotics for treatment of brucellosis.

Brucella is a gram negative, non spore forming, facilitive, intracellular bacterium. Therefore there are a range of antibiotics which can potentially be used in its treatment. The most important of these are: tetracyclines, aminoglycosides, co-trimoxazole, riframpicin, and quinolones. In order for such therapies to succeed the antibiotics must which can penetrate macrophages or other calls to give greater intracellular concentration is important. In the treament of neurobrucellosis antibiotics must be able to cross the blood brain barrier. Early trials with monotherapies were relatively unsucessful particularly if treatment was delayed. The acute illness is best treated early to prevent complications and relapses. For this reason combination treatments of 2 or 3 antibiotics are used which have been found to be more effective.

The effectiveness of antibiotic treatments against Brucella can be assessed in 2 ways: clinical trial studies with human patients and animals and laboratory tests.

• Clinical trial studies with patients and animals. Such trial are reviewed in detail by Hall and others trials are indicated below.
• The minimum inhibition concentration (MIC) is tested in the laboratory. The antibiotic is exposed to the bacterium either in a broth or on agar plates . The lowest concentration of antibiotic which inhibits the growth of the bacteria is the MIC and measured in μg/ml. Lopez-Merino et al measured MIC(90) against 97 strains of brucella for a range of antibiotics: tetracyclines at 0.5μg; fluoroquinolines at 0.5μg; rifampicin at 2.0μg; streptomycin at 4/0μg; co-trimoxazole at 8.0μg. mIt should be noted that good results in these tests do not necessarily corespond to clinical sucess. Factors such as: antibiotic tissue concentration, penetration of macrophages, half life, and ability to cross the blood/brain barrier are also important together with adverse toxic effects.

These trials have shown: aminoglycosides such as streptomycin, gentamicin and netilmicin; tetracyclines such as tetracycline and doxycycline; macrolides such as erythromycin; quinolines such as ciprofloxacin and ofloxacin; co-trimoxazole (TMP-SMZ); riframpicin, aminopenicillins and cephalosporins . A comprehensive review of the use of these antibiotics and others singly or in combination is provided by Hall in 1990 and Madkour .

Garcia-Carrillo (1990) notes that because it is an intracellular parasite it is well protected against antibodies and antibiotics. In vitro studies do not necessarily indicate the eff4ectiveness of treatment. Tetracyclines can work in simple cases but in chronic illness in man long tewrm treatment is necessary.

Mono Antibiotic Therapy.

Although Hall suggested that some mono antibiotic therapies could work, particularly tetracycline, they are no longer used because of the unacceptible relapse rate. Young reviewed treatmentr histories and found that failure rate with tetracycline was up to 30%. Mousa, et al. in a study of 100 patients saw a failure rate of 51%.

Combination Therapy.

These are a number of combined therapies which have been studied and considered to give good results in the treatment of brucellosis:

• • Doxycycline and aminoglycosides (streptomycin or gentamicin).
  • Doxycycline and riframpicin.
  • Quinolones (Ciprofloxacin or ofloxicin) and riframpicin.
  • Doxycyline, riframpicin and gentomicin (or streptomicin).
  • Doxycycline, riframpicin and co-trimoxazole.

WHO recommended treatment for Brucellosis

• The treatment of brucellosis recommended by FAO/WHO in 1971 was tetracycline and streptomycin and in 1986 the FAO/WHO was doxycycline 200 mg daily plus streptomicin 1g daily for 2-3 weeks or doxycycline at 200 mg per day plus riframpicin at 600-900mg per day both for at least 6 weeks.

In 2006 Ariza et al and a group of leading experts in brucellosis reviewed available information and concluded that first line treatments for for brucellosis remains that recommended by the WHO. Doxycycline 600-900 mg daily for six weeks and plus gentamicin 5 mg/Kg daily for 7 days was also recommended. Cifroxacin 500 mg twice dayly for 6 weeks or ofloxacin 400mg twice daily for 6 weeks in combination with doxycycline 200mg per day were second line treatments for non complicated brucellosis. It was proposed that streptomicin should be kept for TB infection.

Doxycycline and streptomycin

Doxycycline and streptomycin is still consider one of the best treatments for brucellosis Cisneros et al recommended this treatment. Castillo et al in a study of 111 patients considers doxycycline/streptomycin slightly more effective with less relapses but doxycycline/riframpicin more convenient to patients and doctors. A typical minimum treatment regime suggested by Acocella et al of oral doxycycline 200mg/day plus im streptomycin 1g/day for 21 days. The recovery rate without relapse for this treatment is about 85%. A synergistic effect of tetracycline and streptomycin is reported by Heilman and Richardson et al . Solera et al considered doxycycline/gentomicin an effective treatment for brucellosis. Solera et al considered that doxycycline for 45 days plus streptomycin for 14 days is the treatment of choice. Gentomicin or netilmicin for 7 days can be used instead of streptomycin.

Doxycycline is a semi-synthetic derivative of oxytetrocycline . It has the lowest MIC of the tetracyclines against brucella. It is more lipid-soluble than the other tetracyclines giving higher intracellular penetration and tissue distribution and concentration. It is adsorbed via the duodenum which is considered advantageous. It has a longer half life at 20 hrs compared to tetracycline at 7 hrs. It is less nephrotoxic than tetracycline. Side-effects include: gastroenteral upsets, photosensitivity of skin, and Jarisch-Herxheimer reaction can occur. There are is recorded instances of brucella resistance. In other bacteria acquired resistance has a ribosomal or a plasmidic mechanism.

Steptomycin is bactericidal and interfers with bacterial protein synthesis. It is given intremuscularly for brucellosis and diffuses rapidly into body fluids and tissue but does not cross the blood brain barrier. Its side-effects are possible nephrotoxicity, vestibular disturbance including vertigo and occasional deafness which is dosage dependant. Rare side effects are drug allergy, agranulocytosis and aplastic anaemia.

The action of gentamicin is similar to that of streptomicin. It also has similar side effects. High serum levels can ototoxic and it is more nephrotoxic than streptomicin.

Doxycycline and riframpicin.

• According to Williams in 1988 the recommended treatment for brucellosis is 600-800 mg daily plus doxyclycline of 200 mg taken together in the morning for at least 6 weeks with a longer treatment if recovery is slow. De Rycke ) in 1980 and Farrell et al in 1990 reported a synergistic effect between tetracyclines and riframpicin. Acocella et al in a study of 143 patients found that doxycline/riframpicin was just as effective as doxycycline/ streptomycin. Ariza et al in 1985 found that doxycycline/ streptomycin and doxycycline /riframpicin were similar in effectiveness except that in the latter case more patients relapsed if they had spondylitis. Solera et al compared treatments of 194 patients with acute brucellosis using doxycycline-riframpicin or doxycycline-streptomycin and considered the latter more effective with fewer failed treatments. Bertrand considered doxycycline/riframpicin the best treatment with the best effecacy/safety ratio. If osteoarticular infection were present 3 to 6 weeks of treatment is necessary with 2-3 weeks of streptomycin at the start. He considere that if the nervous system was affected 3-6 months of riframpicin and co-trimoxazole (320-1600 mg/day.) was required.

Riframpicin is a bactericidal drug which inhibits bacterial RNA synthesis. It is taken orrally. It distributes well in body fluids and tissue and crosses the blood/brain barrier well Side effects include: flushes, itching. Hepitisis can occur if there is pre-existing liver damage. Godeau et al. considered that teracyclines are synergistic with riframpicin but co-trimoxazole is an antagonist.

Monocycline is a possible alternative to doxycycline. Cascio et al and reviewed the use of minocycline/ riframipicin for 3 weeks in 239 cases and concluded that cure was good with only 2% relapse rate. 29.8% had moderate side-effects. There are only limited trials of this treatment. Grasso also reviewed its use.

Ciprofloxacin (or ofloxicin) and riframpicin or doxycyclin.

A typical treatment regeme for ciprofloxacin/riframpicin suggested by Agalar et al . This was riframpicin at 600 mg/day and ciprofloxacin at 1000mgg/day for at least 30 days. In a comparative study of forty patients against doxycycline/riframpicin they concluded that ciprofloxiacin/riframpicin was as effective and quicker acting. Akova et el treated 61 acute brucellosis patients with either ofloxacin-riframpicin or doxycycline-riframpicin for six weeks and concuded that the 2 treatments were of similar effectiveness. The former was tolerated better with fewer side effects. Karabay compared treatment of doxycyclin 200mg-riframpicin at 600mg per day for 45 days and ofloxicin at 400 mg per day /riframpicin at 600 mg per day for 30 days and concluded that they were similarly effective. the latter treatment worked quicker and reduced fever faster. Saltoghu, et al compared doxycycline at 200mg - riframpicin at 600 mg per day with ofloxacin at 2 x 200mg per day - riframpicin at 600 mg per day. Both treatments were for 45 days and he found similar effectiveness.

Small scale trials of doxycyclin and ciprofloxicin by Aygen and Pappas have shown promise particularly in the treatment of osteobrucellosis.

In a recent review of the use of quinolones to treat brucellosis by Pappas (2006) he notes that the small trials carried out to date are insufficient to recommend widespread use of quinolones at this time.. A large scale trial is needed. He suggests thatn they may be The development of resistance could be significant.

Ciprofloxicin al-Sibai, MB He also notes the development of resistance to ciprofloxacin. The MIC ) at 0.25-0.5 μg/ml. Duisenova in a study of 105 patients (51, acute;19, subacute; 35, chronic.) considered that ciprofloxacin (500 mg bid for 14 to 20 days) was more clinically and immunologically effective than treatment with doxycycline/riframpicin.

Doxycycline, riframpicin and co-trimoxazole.

• Doxycycline, riframpicin and co-trimoxazole is recommended as a treatment for neurobrucellosis by alDeeb, .

Co-trimoxazole is a combination of trimethoprim (TMP) and sulphamethoxazole (SMZ) at 1:5. SMZ inhibits the use of aminobenzoic acid by Brucella and TMP interferes with the enzyme, dihydrofolate reductase. Their action together is syergistic . Typical dose is TMP-SMZ at 160/1600 mg taken orrally. Distribution in tissue and body fluids is good. Serum half life is ~12 hours. It achieves penetrates the blood brain barrier in effective concentration. Side effects are: gastroenteral upsets, possible nephrotoxicity with pre-existing renal problems. neutropenia, thrombocytopenia, aplastic anaemia and megaloblastic change occur rarely. Also skin rashes.

Antibiotic Treatment in Pregnant Women and Children under 8 Years.=

Corbel (2006) reviewed the use of antibiotics in preganant women and children. Tetracyclins are normall not used because they stain the teeth of fetuses and small children and possible drug-induced fatty necrosis of the liver and pancreatitis in the pregnant women. The teragenic consequences on the fetus of riframpicin, co-trimoxazole and fluoroquinolins are unknown. Streptomycin may be toxic to the fetus but gentomycin less so. Treatments which have been tried are monotreatements using co-trimoxazole or riframpicin.

Treatment of Localised Brucellosis

Pappas (2007) has reviewed the available studies in the treatment of neurobrucellosis. The ability of antibiotics to cross the blood brain barrier is crucial. Antibiotics which cross theBBB include doxycycline (doxycycline is more lipophilic than most other tetracyclines), riframpicin, and co-trimoxazole. This combination is a suitable first phase treatment for neurobrucellosis. Streptomycin and other aminoglycosides do not normsally cross the BBB but this may happen to some extent during inflamation. Anti-inflaminary agents such as corticosteroids would negate this effect. He concludes that streptomycin should not be used not only because of the above reason but also because its neurotoxicity in the presemce of cranial nerve damage. Ceftriaxone (and other cephalosporins) are effective against Brucella in in vitro studies and reaches high concentration in CSF. Results of trials of its use are variable and difficult to unterpret with a possible 28% failure rate. Therefore Pappas considers that it should be used as a second phase treatment. Tigecycline a lipophillic glycylcyline structurally similar to tetracyclines may have considerable potential. Surgical intervention may be necessary in cases such as intracranial abscesses. There is no specific optimal period of treatment although 3 to 6 months are suggested. In the case of new acute illness a shorter period of treatment can produce good recovery without sequelae whereas if treatment is delayed for months recovery may be only partial with typical residual chronic symptoms may occur event if treatment is extended up to 18 months. Residual deficit such as hearing loss may not improve.

Duration of treatment

Chronic brucellosis is more difficult to cure than the acute illnes. This is because of the development of deep seated foci of infection in chronic brucellosis. Williams (1982) notes that even in acute brucellosis prolonged treatments may be necessary. He recommended a typical treatment of 6 weeks but considers that 12 weeks or more may be necessary and that chronic illness is more difficult to cure. Moreno-Lafont et al (1995) indicated that internal locations of br in the reculoendolial system may cause even high concentrations of antibiotics to be ineffective with a high rate of failure and relapse. The normal procedure is to repeat treament one or more times. Geyik (2002) noted that in the case of deep seated foci infections treatment exceeding 12 weeks are necessary. Hatami (2005) consideres that chronic br patients present a difficult therapeutic problem needing antibiotic treatment from 2-6 months.

Williams (1973) notes that brucellosis is an intracellular parisite relatively inaccessable to antibiotics necessitating long treatment. Occasionally in chronic brucellosis the response to prolonged treatment is poor.

Failure of antibiotic treatment.

There is anecdotal evidence that some patiens with chronic brucellosis fail to recover when treated with antibiotics. Because of the confusion associated with the diagnosis this is not well recorded. The reason for this failure has not been determined. Possible considerations are inaccessability of brucella organisms to antibiotics and antibiotic resistance.

• In order to kill brucella antibiotics must be present in sufficient concentration to kill the organism. Although the concentration may be sufficient in blood and other bodily fluids this may not be the case if the bacteria is protected ) in cysts, mini-cysts or granulomas. In neurobrucellosis the concentration of antibiotics must be sufficient to kill brucella.
• Evidence has been found for antibiotic resistance by brucella . This includes streptomycin, ciprpfloxicin, anf rifampicin. It is almost impossible to measure antibiotic resistance in chronic brucellosis because brucella cannot be recovered by culture in these cases. It is noted that there is a difference in resistance induced in the laboratory and that in the body

Spink (1950) observed by experimentation that intracellular Brucella are protected from the lethal action of antibiotics such as streptomycin an d tetracycline. He concluded thatt even with the best therapy Brucella may persist in the tissues for long periods. He conjectured that this could explain the chronic clinical course and frequent exacerbations of illness in some patients.

Castano (2009) reviews the difficultis of successful treatment in chronic brucellosis for 11 patients (5 focal disease patients and 6 non-focal symtpomatic). All recived 1 to 4 courses of antibiotics: doxycycline, 100 mg/12 hrs; or doxycycline in combination with riframpicin, 900 mg/24 hrs, gentamicin, 240mg/24hrs, streptomycin, 1g/24hrs, or cotrimoxozole (trimethoprim, 160mg/12 hrs and sulfamethoxazole 800mg/12hr. 5 had transitory resolution of symptoms, 2 had long term improvement and the rest had no benefit. Treatment time was 45 to 495 days (mean 191 days). They noted that Brucella DNA persisted in 80% of patients after treatment.

Davis (1957) notes from his experience in the use of tetracycline/streptomicin in the treatment of chronic brucellosis that it is not possible to talk of cure in a disease that is notorious in its tendency to relapse therefore the results of treatment should be regarded as improvements to a point where that patient can resume normal activities.

Spink (1960) described a murine experiment in which despite long term treatment with tetracycline Brucella were recovered from the spleen of the animals long after completion of the treatment.

L-Forms.

De Rautlin de la Roy (1986) demonstrated the resistance of B melitensis after treatment with riframpicin and doxycycline.

Alternative Treatments

Only limited work has been done on alternative treatments of brucellosis. Almost exclusively these relate to overcoming anergy of the immune system.

Treatment with Vaccine.

The first attempt to use dead brucella as a vacinne was by Wright . In 1929 Simpson treated 29 patients with a course of 3 injections of a vacinne of killed Br abortus A rise of temperature and exacerbationof symptoms and they considered that the treatment inducing favourable results. De Jong (1936) treated a patient with Br melitensis meningitis with a vacinne derived fron B suis ans melitensis but saw no evidence that it worked. Vaccine treatments were used in the 1940s before the introduction of antibiotics.. The use of vaccines was advocated bt Harris . Brucella abortus vaccine was used. The idea was to stilulate the activity of the immune system and was measure using the opsonocytophagic test. The viability of te tratment was disputed. It fell out of use with the introduction of viable antibiotics.

Spink (1962) carried out a trial in which 32 men volunteers were injected with a solution of 2.5 x 10⁸ viable organisms of an attenuated vacinne either B abortus strain 19-BA or B melitensis strain Rev 1. The trial was afailure 3 in the B abortus and 11 iof 16 in the B melitensis group developed acute brucellosis. However the trial does provide useful data. From the injection of the vaccine tothe start of illness took 3 to 23 days (the majority at 7 to 14 days) and sera agglutinin response occurred between 1 and 4 weeks after injection but in a minority of cases it remained low or negative.

Treatment with Levamisole

Levamisole is an antihelminthic agent with immune stimulating properties with possible improvement of anergy to brucella. Levamisole has been used in a number of studies to boost the immunee sytem:

• In such a study Thorne 50 patients with chronic brucellosis were treated with levamisole. He noted that at first the patients developed acute exacebated symtoms of brucellosis but then bcame symptom free. Recovery was associated with an increase of T-lynphocites levels. However over the course of a year the patients were subject to relapses in which T-lymphocite level dropped. Reintroduction of levamisole reversed the process.
• Renoux (1977) did a similar study on chronic brucellosis patient for whom tetracyclin had not worked. THe obseved course was that symptoms brcame worse for about seven days after 10-12 days, followed by immuno-restoration and remission of symptoms and blastogenic response. Patients remained well during 1 year of follow-up.
• Raptopoulou-Gigi (1980) described the treatment of 10 patientsd with chronic brucellosis. The treatment was 150 mg of levamisole on three consective days in each week for 6 months. 9 of 10 patients were symptom free after a 6 month follow up period.
• Boura (1984) showed in an in vitro experiment that levamisole improved the monocyte phagocytosis in peripheral blood taken from chronic brucellosis patient but not up to normal levels. In a separate experiment 11 chronic brucellosis patients were treated for six months with levamisole (150mg per os daily for three concecutive days each week). At the end of the study 6 patients were symptom free and 5 were greatly inproved. T lymphocyte numbers and monocyte phagocytosis returned to normal. This strongly suggests that the use of levamisole had been very useful.
• Mukovozova (1988) noted that in a trial with 97 patients the impact during a [eriod of exacerbation and follow up was very noticable. The immunomodulating effect of levamisole was shown by an increase in circulating lymphacyte subpopulations.
• Irmak (2003) 71 chronic brucellosis patients were given rifampicin, 600mg/day + doxycycline, 200 mg/day and of these 36 were given levamisile 80 mg every other day all for 6 weeks. On a statistical basis the group given levamisole had a much better outcome in terms of reduction of symptoms.
• Dizer (2005) studied the usefullness of levamisole in the treatment of chronic brucellosis patients monatoring the levels of lymphocyctes: CD19, CD3 HLA DR, CD16-56, AND CD4/CD8 and found no effect. No mention of patient health is given.

• In consideration of all these studies levamisole offers considerable potential with or without combination with antibiotics.

Treatment with Interferon Alpha (INF-α).

Printzis (1994) carried out in vitro trials of INF-α 2b or levamisole or antibiotic therapy on 32 anergic patients with chronic brucellosis. It was considered that better results were obtained with INF-α than levamisole. Conventionasl treatment was the least effective.

Treatment with Klebsiella Pneumoniae Extract (KPE)

Bacterial extracts have been shown in animal studies to enhance antibody response to antigens and to induce B-cell proliferation. In humans, amung other benefits,it has been shown to enhance monocyte phagocytosis and babtericidal activity in chronic bronchitis . Treatment for 3 months: 1st month, 2mmg KPE for 8 days out of 31; 2nd and 3rd months, 1mg KPE for 8 days out of 31. In the study 15 out of 19 patients with chronic brucellosis improved. Phagocytosis reached normal levels. Chemotaxis and delayed hypersensitivity (skin test) showed inprovement. This is an intersting study but more work is required.

Treatment with Ascorbic Acid

This is another study by Boura (1989) of 14 patients with chronic bruellosis. This study records that in previous studies it is suggested that ascorbic acid can stimulate lymphagocyte transformation, polymorphoneuclear leucocyte motility, improve neutrophil function and enhance migration and chemotaxic of human leucocytes. The patients were given 1000 mg of ascorbic acid oraly per day for 30 days. Prior to the trial all the patients had deficient levels of ascorbic acid in peripheral blood and monocytes. After treatment ascorbic acid in peripheral blood increased by not in monocytes. This study shows that random migration against casein and brrucella is impaired and the decline in monocyte function affecting both phagocytic and chemactic responsed. Both random migration and chemotaxis were restored by the treatment. This treatment does not fully restore the disabled monocyte-macrocyte system however there may be sufficient imflammatory responsw to acheive some effect.

Treatment with Immunofan

According to Maltcev and Lebedev Immunofan is synthetic hexapeptide (Arg-alphaAsp-Lys-Val-try-Arg) with molecular weight 836 based on the thymil hormone Thymopoietin. Maltcev reports that immunofan is a powerful immunoregulator which restores cellular and humeral immunity. It stimulates the processes of T-lymphocyte maturation, restores the balance of T-lymphocyte sub-populations and intensifies the activities of killer cells and neutrophils. It increases the synthesis of IgG, IgM and IgA. Where there is decreased lymphocyte activity it stimulates the formation of IL-2. It increases antibacterial activity, shortens clinical symptoms. Limited trials have been carried out in Russia (1989)on patients with chronic brucellosis apparently with good results but the author has no access to the results. The dose for chronic brucellosis is 1-2 mcg/kg administered subcutaneously or intramuscularly in 3 days for a month. Side effects are reported to be quite limited. Further information is provided by Suleimaranov and Pokrovsky .

Miscellaneous.

Goldstein (1934) notes that individuals exposed to Brucella can develop significant agglutination titres without aparent clinical symptoms. Also individuals with positive Brucella skin tests but negative agglutination titre could have obscure past histories of indiagnosed febrile illnesses, vague gastro-intestinal disorders and unusual types of arthritis.

Wallis (1959)