The Role of Sheep, Goats and Springbok (Antidorcus Marsupialis) As Sources of Human Infections in Namibia

Article — Artikel

Brucellae through the food chain: the role of sheep, goats and springbok (Antidorcus marsupialis) as sources of human infections in Namibia

K Magwederea,c*, A Bishia, G Tjipura-Zaireb, G Eberleb, Y Hembergera, L C Hoffmanc and F Dzivad

dead-end hosts of Brucella infections, ABSTRACT while some wildlife species can act as A confirmed case of human brucellosis motivated an investigation into the potential source potential reservoir hosts15,17,19. The patho- of infection in Namibia. Since domestic animals are principal sources of Brucella infection in genic species are B. melitensis, which pre- humans, 1692 serum samples were screened from sheep, goats and cattle from 4 presum- dominantly infects goats and sheep; ably at-risk farms and 900 springbok (Antidorcas marsupialis) serum samples from 29 mixed farming units for Brucella antibodies by the Rose-Bengal test (RBT) and positive cases B. abortus, which principally affects cattle; confirmed by complement fixation test (CFT). To assess the prevalence of human B. suis that infects swine and B. canis that 5,11,21,52,53 brucellosis, 137 abattoir employees were tested for Brucella antibodies using the standard infects dogs . Although any of tube agglutination test (STAT) and by enzyme linked immunosorbent assay (ELISA). Cattle 4 species of Brucella can cause systemic and sheep from all 4 farms were negative by RBT and CFT but 2 of the 4 farms (Ba and C) disease in humans, B. melitensis has the had 26/42 and 12/285 seropositive goats, respectively. Post mortem examination of seroposi- lowest infective dose, requiring as few as tive goats revealed no gross pathological lesions typical of brucellosis except enlarged 10 organisms to cause infection2,9,25. mesenteric and iliac lymph nodes seen in a single buck. Culture for brucellae from organs of Human brucellosis, commonly referred seropositive animals was negative. None of the wildlife sera tested positive by either RBT to as undulant fever or Malta fever, often or CFT.Interviews revealed that besides the case that prompted the investigation, a family coincides with livestock infection4,35. and another person from other farms with confirmed brucellosis shared a common history of consumption of unpasteurised goat milk, home-made goat cheese and coffee with raw The disease in humans presents as a milk and prior contact with goats, suggesting goats as the likely source of infection. All multi-systemic, acute to chronic disease 137 abattoir employees tested negative by STAT,but 3 were positive by ELISA. The 3 abattoir characterised by non-specific signs such workers were clinically normal and lacked historical connections with clinical cases. Al- as fever, headache, joint pains, musculo- though goats are often associated with B. melitensis, these studies could not explicitly impli- skeletal pains, sweating, malaise, myal- cate this species owing to cross-reactivity with B. abortus, which can also infect goats. gia, abdominal pain, lymphadenopathy, Nevertheless, these data reinforce the need for a better National Control Programme for skin rash, pneumonitis, back-ache and brucellosis in Namibia. body wasting9,21,25,36. The non-specific na- Keywords: abattoir, brucellosis, goats, Namibia, serology. ture often presents a tremendous chal- Magwedere K, Bishi A,Tjipura-Zaire G, Eberle G, Hemberger Y, Hoffman L C, Dziva F lenge in clinical diagnosis of brucellosis Brucellae through the food chain: the role of sheep, goats and springbok (Antidorcus since these signs can also occur in com- marsupialis) as sources of human infections in Namibia. Journal of the South African mon diseases or conditions like malaria, Veterinary Association (2011) 82(4): 205–212 (En.). Division of Veterinary Public Health and typhoid, rheumatic fever and pyrexia of Epidemiology, Directorate of Veterinary Services, Hospital Street, Mariental, Namibia. unknown origin21. Unlike in Tanzania25,45, the disease’s status in Namibia is not pre- cisely known as most health centres do INTRODUCTION Trade Organization (WTO) to develop not routinely test for brucellosis. Sporadic The emergence or re-emergence of minimum scientific or evidence-based cases and even small clusters are often dif- zoonotic diseases is complex and multi- international standards, guidelines and ficult to identify owing to an extremely factorial, often driven by evolving ecol- recommendations to facilitate safe trade variable incubation period (weeks to ogy, microbial adaptation, human demo- in animals and their products51.Inthe months) and a lack of pathognomonic graphics and behaviour, international case of zoonotic diseases such as brucel- clinical features or manifestations27,44. travel and trade, agricultural practices, losis, it is believed that protection of Of the zoonotic species, B. melitensis technology and industry49. The World human health can be achieved through contributes up to 70 % of human brucel- Organization for Animal Health (OIE) is control of the disease in the animal popu- losis cases worldwide14,32 with sheep, mandated under the Sanitary and Phyto- lation. The OIE/FAO (Food and Agricul- goats and camels being the main sources sanitary Agreement (SPS) of the World ture Organization)/WHO (World Health of infection9,24. The difficulty in clinically Organization) Global Early Warning diagnosing brucellosis, coupled with the aDivision of Veterinary Public Health and Epidemiology, System (GLEWS) provides for rapid noti- weakness of human health services in Directorate of Veterinary Services, Hospital Street, Mariental, Namibia. fication of major animal diseases, includ- developing countries, often contribute to bCentral Veterinary Laboratory, Ministry of Agriculture, ing zoonoses, of which B. melitensis infec- human brucellosis being inadequately Water and Forestry, Private Bag 13187, Windhoek, tion is a priority disease51. treated34. Bacteria gain entry into the Namibia. Brucellae are small, non-motile, Gram- body by ingestion, inhalation, penetra- cDepartment of Animal Sciences, Stellenbosch University, Stellenbosch, Private Bag X1, Matieland, 7602 South negative coccobacilli that are responsible tion through skin abrasions as well as Africa. for one of the most widespread zoonotic through conjunctival mucosa41. Once in dDivision of Microbiology, Institute for Animal Health, Compton, Berkshire RG20 7NN, United Kingdom. infections of medical significance world- the body, Brucella spp. survive and multi- 28,44 *Author for correspondence wide . Humans are accidentally in- ply within cells of the reticuloendothelial E-mail: [email protected] fected through direct or indirect contact system. Notably, human neutrophils Received: February 2011. Accepted: October 2011. with infected material and are invariably exhibit variable antiphagocytic activity

0038-2809 Jl S.Afr.vet.Ass. (2011) 82(4): 205–212 205 towards some Brucella strains but are vir- region triggered an investigation into the Sampling techniques tually ineffective against B. melitensis52,53. potential source of infection in this The sample size for farmed wildlife was However, once excreted, brucellae rarely area and 3 other surrounding regions in determined by the formula: n =log ~ survive for long periods in tropical envi- the foot-and-mouth disease-free zone, { )/log(1 – pmax × sensitivity), where n = ronmental conditions. Risk factors for namely Karas, Omaheke and Khomas the required sample size, p = the preva- infection in humans include handling of (Fig. 1). The main farming activities in lence (based on livestock data), ~ = the infected dead or live animals, ingestion of these regions include extensive sheep, probability (confidence level) of missing a contaminated animal products espe- goat and game rearing. These regions are diseased animal at a prevalence of up-to cially unpasteurised dairy products, poor home to 88.3 % of the 2.7 million sheep in pmax in a large population. The population handling of Rev 1 vaccine and cultures of Namibia. A single abattoir caters for size refers to the total number of spring- 2 Brucella spp . The Rev 1 vaccine is poten- animals from these 4 regions and has an bok in the 3 regions exercising commer- tially harmful to humans. Vaccinated ani- annual turnover of 219 929 sheep and cial harvesting. The sampling protocol mals should not be slaughtered for 6251 springbok (Antidorcas marsupialis). was designed to detect at least 1 positive human consumption within 3 months of Relevant background data in the study animal in the flock at a 95 % confidence vaccination. Although it is believed that regions were collected via a questionnaire limit, that is, if the disease was present at a with ordinary handling this vaccine is that included the following main catego- prevalence (p)of5%inharvested spring- harmless to man , care should neverthe- ries: i) vaccination with Rev 1, ii) animal bok when the test sensitivity is 95 % and less be taken not to accidentally inject it trace-back and trace-forward movements animals are randomly selected. In this into humans and contamination of to and from the suspected farms within a particular study on farmed wildlife, the human eyes with this vaccine should be year, iii) farms and other places visited 3 regions were defined as an epidemio- avoided. By virtue of increased contact by family members within a year, and logical unit. A total of 900 adult springbok with animals and their products, abattoir iv) sources and type of animal products from farmed flocks in 3 regions were workers are at a greater risk of brucellosis consumed. Farms presumed to be at risk sampled by systematic random sampling infection compared with other profes- where a starting and interval number was 2,45 were identified by the track back system sional groups . While notification of and designated A, B, C and D. These picked at random from 1 to 8 during com- diseases in wildlife is not a pre-condition farms held varying numbers of small mercial harvesting operations with a for imposing bans on members of the ruminants and farmed wildlife. Informa- minimum farm blood sample size of 11. WTO who implement the OIE terrestrial Only adult male and female animals were tion retrieved from the computerised code35,53, the ever-changing game farming selected for harvesting. Namibian Livestock and Traceability Sys- systems create a need to establish and For sheep and goats, the sample size tem (NAMLITS) indicated that farm A distinguish between a spill-over infection was determined by the formula: n ={1– had 2 milking Jersey cows, 130 Boer goats from domestic animals and maintenance (1 – a)1/D}{N –(D – 1)}/2 as described in and 1300 sheep. The farm serum samples of infection in wildlife species. Cannon and Roe3 where n is the required had previously been tested using RBT A single human case of undulant fever sample size, a is the error probability and declared brucellosis-free through or Malta fever was suspected in a commu- (confidence level) of observing at least annual serological testing. Farm B was nity surrounding the interface of wildlife 1 diseased animal when the disease inherited and had been subdivided into farming and domestic animals in Namibia. affects at least D/N in population, D is the portions Ba and Bb between family mem- At clinical examination, brucellosis was number of diseased animals in a popula- bers. Portion Ba held 984 sheep compris- considered a differential diagnosis and tion and N is the population size. The ing of Dorper, Damara and Persian a blood sample was collected for sero- flock size was in reference to the number logical testing at a national laboratory in breeds, 39 Boer goats and 3 Saanen dairy of animals in the group being sampled. Namibia and at the National Institute for goats, and portion Bb had 160 Boer goats The sampling protocol was designed to Communicable Diseases, Johannesburg, and 11 Saanen dairy goats. Farm C held detect at least 1 positive animal in the South Africa. The serum sample tested 285 Boer goats and an undisclosed num- flock at a 95 % confidence limit, that is, if positive for Brucella IgG and IgM antibod- ber of sheep. Farm D held 106 dairy goats; the disease is present at a prevalence of ies by enzyme linked immuno-sorbent the largest Saanen dairy goat operation in 5 % in animals of breeding age (6 months assay (ELISA), providing a definitive the region, which supplied most of the and older). If an animal was picked and diagnosis of brucellosis. This prompted milk to all 4 regions. Goat movement from not subjected to blood collection, it was the undertaking of a large serological farm Ba’s grazing camps within 1 year placed in a different holding pen. survey to provide evidence of brucellosis from the date of initial investigation to in domestic animals and springbok farm C was tracked through the Namibia Preparation of serum from domestic (Antidorcus marsupialis), a small ungulate Livestock Traceability system (NAMLITS). animals and farmed wildlife frequently found in close proximity to However, discrepancies in the actual The sample sizes of domestic ruminants farmed livestock and handled by abattoir numbers of animals were noted at some differed from farm to farm. All animals workers. The findings suggest direct or of the farms at sampling. A human case of were sampled by venipuncture of the indirect contact with goats or their unpas- brucellosis that triggered the investiga- jugular vein. On farm A, 77 sheep, teurised products as the most likely source tion resided in farm A. Commercial har- 25 goats and 2 milking cows were bled. of infection to humans, while farmed vesting of springbok was done in 3 of the On farm subdivision Ba, 39 goats and wildlife presented the least likely threat. 4 regions. 984 sheep were bled while 34 goats were To examine the trend of this infection bled on subdivision Bb. After initial posi- MATERIALS AND METHODS over previous years, data on the reported tive results of samples from farm Ba, all serological incidences of brucellosis using 285 Boer goats on farm C, all sheep and Background data, study area and Rose-Bengal test (RBT) and complement goats on farm Ba and 106 goats from farm population fixation test (CFT) in animals between D were bled and serologically tested for A human case of brucellosis in February 2004 and 2009 in Namibia were retrieved brucellosis. Blood was allowed to clot at 2009 at a commercial farm in Hardap and analysed. ambient temperature and placed in a

206 0038-2809 Tydskr.S.Afr.vet.Ver. (2011) 82(4): 205–212 Fig 1: The areas under study (Khomas, Omaheke, Hardap and Karas) in relation to foot-and-mouth disease-infected and protected areas (shaded grey). cooler box for transit to the laboratory. ture-positive animals. Sera from domestic toir, Mariental, Namibia). The sera were Serum was separated from the clot by animals and farmed wildlife were initially collected as part of the abattoir’s occupa- centrifugation (13 000 g) for 10 min, screened for brucellosis by the RBT and tional health surveillance and regulatory collected into fresh tubes and stored positive serum samples were verified monitoring of workers’ health in the meat frozen at –20 °C until required for analysis. by CFT. Reference antigens for Brucella industry. Approval to use such sera was Blood samples were collected randomly species were obtained from Onderste- obtained from the abattoir’s management from springbok at harvesting periods; poort Biological Products (Onderste- in consultation with their medical doctor April to August 2009 and July to August poort, South Africa) while the reference and the workers’ representatives. A total 2010. Typically, blood was freely collected anti-sheep red blood cell antibody (Ambo- of 137 human serum samples represent- into test tubes upon severing of both ca- ceptor) and complement were procured ing individuals involved in buying, rotid and jugular vessels, allowed to clot from Siemens Healthcare Diagnostics slaughtering and deboning small rumi- at ambient temperature and then placed Products GmbH (Erlangen, Germany). nant and game animal carcasses were in a cooler box before processing. Sera Briefly, a 2-fold serial dilution of comple- obtained. The relative proportions of each were prepared, collected and stored in a ment-inactivated animal sera and refer- age group and gender of the employees similar manner as described above for do- ence positive (Amboceptor) and negative sampled are given in Table 1. Serum mestic animals. controls were prepared in Veronal buffer samples were processed at an accredited as per the manufacturers’ instructions. An laboratory (Pathcare, Windhoek, Serological testing of domestic animals equal amount of an appropriately diluted Namibia) and stored at –20 °C prior to use. and farmed wildlife standard antigen was added to all wells Serum samples were screened by the All sera were submitted to the Central followed by complement. Complement standard tube agglutination test (STAT) Veterinary Laboratory (Windhoek) for activity was detected after addition of using B. abortus and B. melitensis antigens serological testing essentially following washed 3 % sheep red blood cells. (Linear Chemicals SL, Spain) as per the standard protocols1. In the serial testing, it manufacturers’ instructions where a titre was assumed that both the RBT and CFT Serological testing of abattoir employees of 1:80 or greater was considered positive. are 90–91.8% sensitive and 99–99.9% spe- To assess the prevalence of brucellosis in Positive samples were confirmed by en- cific at 95 % confidence in detecting humans sera were tested from abattoir zyme-linked immuno-sorbent assay Brucella antibodies in sera from cul- employees (Farmers Meat Market abat- (ELISA) using the Panbio Brucella IgG

0038-2809 Jl S.Afr.vet.Ass. (2011) 82(4): 205–212 207 Table 1: Distribution of abattoir workers tested for brucellosis on the basis of age and tive for brucellosis using the CFT. The gender. positive cases on farm Ba included all milking goats while all sheep from both Age Gender Total Positive for (years) brucellosis farms were negative. Of the 12 positive Male Female Number % goats on farm C, 6 were pregnant females and a single 6–9-month-old male goat, Number % Number % whereas none of those on farm Ba were in lamb. The 2 Jersey cows on farm A were <20 20–30 31 22.63 6 4.38 37 27.01 0 also negative by both serological tests. To 30–40 63 45.99 11 8.03 74 54.01 1 contain the disease, farms B and C were 40–50 19 13.87 2 1.46 21 15.33 2 immediately quarantined and the posi- 50–60 4 2.92 1 0.73 5 3.65 0 tive goats were isolated before being sent >6000 00 00 0for slaughter at a local non-exporting municipality abattoir. To investigate the disease status of ELISA kit (Plasmatec Laboratory Prod- gered the investigation, 3 additional fam- farms Ba and C after quarantine and ucts, Dorset, UK). ily members from the Hardap region slaughter of positive goats,, the remaining All seropositive animals were slaugh- (where farm A is located) and an unre- animals were re-tested after 3 months tered separately at a non-export local lated person from another region had and none tested positive for brucellosis, municipality abattoir at the end of the previously been treated for suspected suggesting that the infection had been slaughter week and subjected to post Malta fever. Interestingly, all humans controlled. To explain the possible cause mortem examination according to local diagnosed with brucellosis shared a of orchitis reported on farm C, the sera of health and safety guidelines for handling common history of consuming raw goat uncastrated males were tested for Brucella suspected zoonotic cases. Tissue samples milk, home-made goat cheese and coffee ovis antibodies by RBT and CFT and were collected from the liver, kidney, with raw milk and prior contact with 3 positive cases were detected. B. ovis is of spleen, lymph nodes (iliac, oropharyngeal, dairy goats. None of the farms used Rev 1 exclusive concern in rams where it causes inguinal, mesenteric and supramammary), vaccine in sheep or goats, but heifers epididymitis20,9. testicles and udder for the detection and were vaccinated with B. abortus strain 19. The annual reporting of the incidences isolation of brucellae. Where pregnant Reported clinical signs suggestive of of brucellosis in Namibia has previously animals were involved, the foetal lung and brucellosis in small ruminants of farms A focused on livestock but not farmed wild- abomasal contents, uterus, and placenta and C included abortions, orchitis, and life. With game animals increasingly membranes were sampled and processed arthritis whereas those seen in dairy goats entering the human food chain, the following standard techniques1. No abor- of farm B were abortions, retained springbok and gemsbok (Oryx gazella) tive material or vaginal discharges were placenta and arthritis. were 1st included in the reporting system obtained. Briefly, impression smears were Unpasteurised cow milk from farm A in 2009, but the incidence was zero made from a cut surface of the organs by was mainly consumed on the farm with (Table 2). To determine whether farmed 1st blotting on clean absorbent paper and small quantities occasionally sent to wildlife play a role in the transmission of then gently pressing onto a clean glass farm B. Raw goat milk was consumed at brucellosis, a total of 900 springbok slide. Smears were air-dried and heat- subdivisions of farm B at least once per samples harvested from 29 different fixed before staining by a modified acid week. In general, the substantiated links farming units were tested, but none was fast technique. Fixed smears were flooded between farms A and B were i) cow milk positive by either RBT or CFT,suggesting with dilute carbol fuchsin for 10 min, ii) individuals from farm A visiting and that this species was not responsible for rinsed in tap water, differentiated with drinking coffee with unpasteurised goat transmission of brucellosis to humans. 0.5 % acetic acid for 30 s and counter- milk on farm B and iii) a ram introduced These findings were consistent with the stained with 1 % methylene blue for 20 s. onto farm A from farm B. It was not possi- absence of clinical or pathological signs Slides were air-dried and then viewed ble to substantiate clear links between suggestive of brucellosis. It was therefore with a light microscope under oil immer- these farms (A and B) and farms C or D, considered that no further action was sion. apart from movement of people between necessary. Tissue samples were cultured on them. Nevertheless, these findings pro- agar on Farrell’s selective medium for vided some clues on the possible route of Serological testing of humans brucellae13. Briefly, samples collected at transmission of brucellosis to humans. Of the 137 serum samples screened for post mortem were homogenised in sterile Brucella antibodies, all were negative by phosphate buffered saline (PBS) using a Serological testing of domestic STAT, 3 were positive for IgG antibodies stomacher before being inoculated onto ruminants and wildlife but negative for IgM antibodies by the solid media. Aliquots of 100 m of Samples of sheep (n = 77) and goats ELISA. The distribution of tested abattoir each tissue homogenate were streaked (n =25) from farm A and farm subdivision employees on the basis of age and gender onto Brucella selective agar and the plates Bb (n = 34) were serologically negative for is shown in Table 1. The positive samples incubated aerobically in enriched CO2 brucellosis using the RBT.Initial serological were from males aged 31(1) and 40(2) with daily examination for brucellae analysis of 39 sera from farm Ba revealed years. These individuals were clinically colonies up to 7 days. 20 positive cases by CFT where a titre of healthy and received no treatment. To de- 1:8 or greater was considered positive, termine whether new cases developed, RESULTS prompting the testing of the entire flock the same 137 individuals were bled as part on that farm and the nearby farm C, to of annual health screening and were Data from questionnaire explore the extent of the problem. re-tested for Brucella antibodies by ELISA Interviews revealed that apart from the Twenty-six out of 42 goats from farm Ba and none was positive, indicating that the brucellosis positive individual who trig- and 12 of the 285 from farm C were posi- infection had cleared.

208 0038-2809 Tydskr.S.Afr.vet.Ver. (2011) 82(4): 205–212 Table 2: Number of brucellosis-positive samples from food-producing animals as determined by the Rose-Bengal test and complement fixation test from 2004 to 2009 in Namibia. Data were collected from the Central Veterinary Laboratory, Windhoek.

Year Species Total number of Number of positive % seropositive samples tested samples

2004 Dairy cows 4001 19 0.47 Other bovine 311 11 3.54 Sheep and goats 18485 19 0.10 2005 Dairy cows 2541 2 0.08 Other bovine 246 0 0.00 Sheep and goats 10191 15 0.15 2006 Dairy cows 2994 9 0.30 Other bovine 401 15 3.74 Sheep and goats 3452 21 0.61 2007 Dairy cows 1578 0 0.00 Other bovine 587 5 0.85 Sheep and goats 1486 31 2.09 2008 Dairy cows Data unavailable N/A N/A Other bovine 1910 15 0.79 Sheep and goats 13745 37 0.27 2009 Dairy cows Data unavailable N/A N/A Other bovine 1030 20 1.94 Sheep 7376 4 0.05 Goats 516 32 6.18 Springbok and gemsbok 122 0 0.00

Post mortem findings and goats as a likely leading source of this medium a priority for primary isola- bacteriology brucellosis during this period. However, tion of B. abortus from contaminated The goats were transported under a there were no corresponding data in samples like post mortem tissues. Impor- special movement permit to the abattoir humans for the same reporting period to tantly, we have previously been success- and slaughtered in accordance with local indicate a zoonotic transmission. Human ful in isolating of brucellae from organs of Health and Safety regulations relating to cases were only reported for the period seropositive goats from other regions of the handling of zoonotic agents. Not 1997 to 2003 (Fig. 3), suggesting either a Namibia using this medium, thus con- surprisingly, the majority of seropositive break in the reporting system or absence firming its quality. Although alternative goats were over 3 years old (Fig. 2). Post of cases until those triggering the present selective media for the isolation of mortem analysis and collection of tissue investigation. brucellae have been described7,23,26,43, these samples for bacteriology was undertaken have not been widely used. Non-selective by the official state veterinarian following DISCUSSION media like blood agar were not employed local instructions and precautions for The annual reporting of the incidences since brucellae are often overgrown by handling zoonotic agents. Post mortem of brucellosis in domestic animals in contaminating fungi and other fast- examination of all seropositive goats Namibia (Table 2) relies on the RBT and growing bacteria when plates are incu- revealed no typical brucellosis lesions CFT on sera from suspected cases. How- bated for 7 days. except in a single ram that had swollen ever, due to the cross-reactivity and Brucellae are usually abundant in abor- mesenteric and iliac lymph nodes. Direct potential interspecies infection by B. abor- tive material and vaginal discharges so microscopy on collected organs failed to tus and B. melitensis, it is difficult to these are highly recommended for the reveal any bacterial cells suggestive of the ascertain the actual prevalence of each demonstration of brucellae by direct presence of brucellae. Subsequent culture species. Differentiation of the species microscopy or culture methods. Although of homogenised tissues also failed to yield requires successful isolation and subse- demonstrating the presence of Brucella any Brucella colonies after 7 days of incu- quent molecular characterisation stud- organisms in tissues and fluids is central bation, suggesting the absence of an active ies42. Indeed, a multiplex PCR assay for to confirming the diagnosis of brucellosis, infection in the seropositive animals. the identification and differentiation of all serological testing remains the method of Brucella species including conventional choice for estimating the prevalence of Retrospective analysis of incidences vaccine strains now exists16 but requires the infection in either humans or animals. of brucellosis in domestic animals Brucella colonies. Such studies were It has recently been suggested that and humans impossible in our study since we were i-ELISA can be used for screening cattle To gain insights into the trend of unable to isolate Brucella from tissues of with improved specificity, but the RBT brucellosis in farm animals and humans, seropositive cases. Failure to obtain any and CFT remain confirmatory tests31. The annual records for the period 2004 to 2009 Brucella colony from seropositive could RBT and CFT are conventional tests were retrieved and analysed. Sero- have been due to the inactive infection in widely used in serological diagnosis of positive bovine cases were found to be the animals as evidenced by lack of patho- brucellosis in domestic animals where relatively high for the period under con- logical lesions or the suppressive effect their respective sensitivities have been sideration, suggesting that brucellosis is of the selective medium used. Farrell’s reported to correlate with culture-positive endemic in domestic animals. The inci- medium has been reported to be inhibi- animals8. Application of the RBT and CFT dence of brucellosis in sheep and goats tory for B. ovis and some B. abortus and serially with a simultaneous consider- was relatively low from 2004 to 2008. An B. melitensis strains29. Despite this, the ation of the test results increases the likeli- increase in the number of goat cases was selective advantage conferred by cyclo- hood of detecting infected animals39,48. In observed in 2009 (Table 2), indicating heximine and a range of antibiotics makes humans, ELISA is the gold standard test

0038-2809 Jl S.Afr.vet.Ass. (2011) 82(4): 205–212 209 Fig. 2: Cumulative frequency from different age categories of brucellosis-positive goats from 2 farms. for diagnosis of brucellosis, where its detect false positives10. However, STAT An immunocapture-agglutination test specificity is increased when IgM anti- has previously been reported to differen- (Brucellacapt) also exists37, but this is not bodies are not detected50, although a tiate between active and inactive infec- widely used. The inability to detect any conflicting report suggests its potential to tions in humans based on the titres54. positive case with STAT suggests these

Fig. 3: Number of official reports of human brucellosis in Namibia from 1997 to 2003 (OIE, Handistatus 11, http://www.oie.int/hs2/).

210 0038-2809 Tydskr.S.Afr.vet.Ver. (2011) 82(4): 205–212 were inactive infections, which was later unpasteurised goat milk or its products REFERENCES confirmed by re-testing. In general, caution like home-made cheese is a leading ante- 1. Alton G G, Jones L M, Angus R D, VergerJ M should be exercised in interpreting these cedent to human brucellosis33,38,40. 1988 Techniques for the brucellosis laboratory. serological tests, particularly in patients Where there is lack of traceability of Institute National de la Recherche Agrono- mique, Paris with chronic brucellosis, re-infections and animal movements to and from the 2. Buchanan T M, Faber L C, Feldman R A 1974 relapses, and in endemic areas where a affected farms, it is frequently impossible Brucellosis in the United States, an abattoir high proportion of the population carries to trace the origin of the infection30. In this associated disease, Part 1 53: 403–413. antibodies against brucellosis. study, lack of a coordinated approach 3. Cannon R M, Roe R T 1982 Livestock disease Although brucellosis is rarely reported during the brucellosis case investigation surveys: a field manual for veterinarians. Australian Government Publishing Service, in wildlife, cases believed to be a spillover by the Ministry of Health, State Veteri- Canberra from small ruminants have been reported nary region and Central Veterinary Office 4. Chanet V, Gourdon F, Baud O, Beytout J, in chamois (Rupicapra rupicapra) and precluded effective planning and formu- Romaszko J P 2005 Brucella melitensis in a ibex (Capra ibex) in Europe15,17,22. Farmed lation of common strategies for food married couple after a trip through Sicily. springbok were therefore tested for safety and control of zoonotic diseases. Southern Medical Journal 98: 843–844 5. Coetzer J A W,Tustin R C 2004 Brucella spp. brucellosis since these were reared in Consequently, there was a direct conflict Infections. In Coetzer J A W, Tustin R C, close proximity to small ruminants, in the implementation of simple preven- (eds) 2004 Infectious diseases of livestock (2nd which could result in a spillover of the tative measures against brucellosis in edn). Oxford University Press Southern infection. However, seroprevalences of abattoir workers that are usually exposed Africa, Cape Town: 1507–1509 23 % and 48 % have been reported in to a greater risk of infection despite the 6. Corbel M J 1997 Brucellosis: an overview. Emerging Infectious Diseases 3: 213–221 African buffalo (Syncerus caffer) popula- existence of health and safety related 7. De Miguel M J, Marín C M, Muñoz P M, tions in the Kruger National Park and handling of suspected zoonotic agents. Dieste L, Grilló M J, Blasco J M 2011 Devel- Zimbabwe respectively, where contact Although contact with infected herds opment of a selective culture medium for with domestic sheep, cattle, pigs and and/or with contaminated environmental primary isolation of the main Brucella goats was absent6,12,22. Furthermore, sources has been shown to play a major species. Journal of Clinical Microbiology 49: 1458–1463 several species of wildlife including Afri- role in the spread of brucellosis, and 8. Díaz-Aparicio E, Marín C, Alonso-Urme- can buffalo, zebra (Equus zebra), impala despite historical sporadic outbreaks of neta B, Aragón V, Pérez-Ortiz S, Pardo M, (Aepyceros melampus), waterbuck (Kobus brucellosis in humans, sheep and goats Blasco J M, Díaz R, Moriyón I 1994 Evalua- ellipsiprymnus), and hippopotamus (Hip- attributed to B. melitensis, the contribution tion of serological tests for diagnosis of popotamus amphibius) have tested positive of these and other factors to the epidemi- Brucella melitensis infection in goats. Journal of Clinical Microbiology 32: 1159–1165 for B. abortus and B. suis worldwide, with ology of brucellosis in game species in 9. Dogany M, Aygen B 2003 Human brucel- B. abortus biovar 1 being isolated in the Namibia remains unclear. Consequently, losis: an overview. International Journal of cotyledons of pregnant buffaloes19,22.In more work is required to dissect the Infectious Diseases 7: 173–182 some areas where B. abortus had been cross-species and within species trans- 10. Ertek M, Yazgi H, Özkurt Z, Ayyildiz A, eradicated in cattle populations, sheep mission dynamics of brucellosis at the Palak M 2006 Comparison of the diagnostic value of the standard tube agglutination and feral pigs have been implicated as interface of wildlife and domestic animals. test and the ELISA IgG and IgM in patients sources of B. melitensis or B. suis infection In conclusion, the data highlight the with brucellosis. Turkey Journal of Medical respectively in cattle 6,12,47. risk of Brucella transmission through Sciences 36: 159–163 It was unclear why none of the sero- goats and unpasteurised products to 11. European Commission 2006 Brucellosis in sheep and goats (Brucella melitensis). Report positive goats, including pregnant humans. It is very likely that other human of the Scientific Committee on Animal females, failed to show typical lesions of brucellosis cases were missed since the Health and Animal Welfare, Health and brucellosis on post mortem examination.It manifestations are non-specific. There- Consumer Protection Directorate General, was, however, not possible to exclude the fore, a better control programme that SANCO.C.2.AH/R23/2001. possible presence of other cross-reacting includes public awareness, vaccination of 12. Ewalt D R, Payeur J P,Rhyan J C, Geer P L 1997 Brucella suis biovar 1 in naturally pathogens. False-positive cross-reactions goats with the Rev 1 vaccine ( in which the infected cattle: a bacteriological, serological due to Yersinia enterocolitica serotype O:9 native hapten-based gel precipitation test and histological study. Journal of Veterinary 18,39 infection have been reported in cattle . would have to be used to distinguish the Diagnostic Investigation 9: 417–420 Another reason for seropositivity could serological response of infected animals 13. Farrell I D 1974 The development of a new be vaccine-induced antibodies, although from those induced in Rev 1 vaccinated selective medium for the isolation of Brucella abortus from contaminated sources. the background investigations revealed animals) and pasteurisation of milk Research in Veterinary Science 16: 280–286 that vaccination of goats with Rev 1 was should be expected to reduce the inci- 14. Feng J L 1992 Control of animal brucellosis not routinely undertaken on these farms. dences of brucellosis in humans. in China. In Proceedings, Strategies in diagno- However, the enlarged lymph nodes sis and control of brucellosis in Asia, Beijing, found in the male goat was attributed to ACKNOWLEDGEMENTS China: 10. 15. Ferroglio E, Tolari F,Bollo E, Bassano B 1998 B. ovis, suggesting the presence of multi- We wish to thank the registered com- Isolation of Brucella melitensis from and ple infections within the flock. Based on mercial game harvesting teams in alpine ibex. Journal of Wildlife Diseases 34: the historical data gathered through a Namibia for assisting in collection of 400–402 questionnaire and interviews, the animals blood samples, Dr A G Grobler, the 16. García-Yoldi D, Marín C M, De Miguel M J, testing positive for brucellosis, and a few abattoir medical doctor, for bleeding Muñoz P M, Vizmanos J L, López-Goñi I 2006 Multiplex PCR assay for the identifica- abattoir employees also testing positive, it abattoir employees, and the abattoir tion and differentiation of all Brucella is strongly suspected that goats are the management, Farmers Meat market, species and the vaccine strains Brucella most likely source of infection. Although Mariental and employee representative abortus S19 and RB51 and Brucella melitensis this could not be ascertained, it is likely committee for allowing us to use sera and Rev1. Clinical Chemistry 52: 779–781 that B. melitensis was involved, as it is other data from the abattoir. Approval to 17. Garin-Bastuji B, Oudar J, Richard Y, Gastellu J 1990 Isolation of Brucella meli- usually associated with goats, since vacci- undertake and publish this work was tensis biovar3 from a chamois (Rupicapra nation against B. abortus was routinely obtained from the Ministry of Agriculture, rupicapra) in the southern French Alps. done in cattle. The consumption of Water and Forestry. Journal of Wildlife Diseases 26: 116–118

0038-2809 Jl S.Afr.vet.Ass. (2011) 82(4): 205–212 211 18. Gerbier G, Garin-Bastuji B, Pouillot R, Véry in sub-Saharan Africa: epidemiology, con- Yazd Province. Iranian Journal of Public P,Cau C, Berr V, Dufour B, Moutou F 1997 trol and impact. Veterinary Microbiology 90: Health 31: 29–32 False positive serological reactions in bovine 111–134 42. Smits H L, Cutler S J 2004 Contributions of brucellosis: evidence of the role of Yersinia 31. McGiven J, Hendry L, Brown D, Stack J, biotechnology to the control and prevention enterocolitica serotype O:9 in a field trial. Vet- Perrett L, Mawhinney I 2008 The improved of brucellosis in Africa. African Journal of erinary Research 28: 375–383 specificity of bovine brucellosis testing in Biotechnology 3: 631–636 19. Godfroid J 2002 Brucellosis in wildlife. Great Britain. Research in Veterinary Science 43. Stack J A, Harrison M, Perrett L L 2002 Eval- Revue Scientifique et Technique, Office Interna- 84: 38–40 uation of a selective medium for Brucella tional des Epizooties 21: 277–286. 32. Memish Z 2001 Brucellosis control in Saudi isolation using natamycin. Journal of Applied 20. Gradwell D V,VanZyl F E 1975 Effectivity of Arabia: prospects and challenges. Journal of Microbiology 92: 724–728 Rev 1 vaccine in rams against Brucella ovis Chemotherapy 12: 11–17 44. Staszkiewicz J, Lewis C M, Colville J, Zervos infection. Journal of the South African Veteri- 33. Moraes P M, Vicosa G N, Yamazi A K, M, Brand J 1991 Outbreak of Brucella nary Association 46: 349–351 Ortolani M B T, Nero A 2009 Foodborne melitensis among microbiology laboratory 21. Hall W H 1991 Brucellosis, In Evans A S, pathogens and microbiological characteris- workers in a community hospital. Journal of Brachman P S (eds) Bacterial infections of hu- tics of raw milk soft cheese produced and Clinical Microbiology 29: 287–290 mans, epidemiology and control (2nd edn). on retail sale in Brazil. Foodborne Pathogens 45. Swai E S, Schoonman L 2008 Human Plenum Medical Book Co., New York: 133– and Diseases 6: 245 brucellosis: seroprevalence and risk factors 149 34. Muriuki S M, McDermott J J, Arimi S M, related to high risk occupational groups in 22. Herr S, Marshall C 1981 Brucellosis in Mugambi J T, Wamola I A 1997 Criteria for Tanga Municipality, Tanzania. Zoonosis and free-living African buffalo (Syncerus caffer): better detection of brucellosis in the Narok Public Health 20: 57–61 a serological survey. Onderstepoort Journal of district of Kenya. East African Medical Journal 46. Thimm B, Wundt W 1976 The epidemiolog- Veterinary Research 48: 133–134 74: 317–320 ical situation of brucellosis in Africa. Devel- 23. Hornsby R L, Jensen A E, Olsen S C, Thoen 35. OIE 2008 Manual of diagnostic tests and opments in Biological Standardization 31: C O 2000 Selective media for the isolation of vaccines in terrestrial animals (6th edn). 201–217 Brucella abortus strain RB51. Veterinary World Organisation for Animal Health, 47. Verger J M, Gain-Bastuji B, Grayon M, Microbiology 73: 51–60 Paris Mahé A M 1989 La brucellose bovine a 24. Kolar J 1987 Brucellosis and human health, 36. Oomen L J A 1976 Human brucellosis in Brucella melitensis en France. Annals of In Proceedings of the Fourth International Kenya. Kenya Literature Bureau: 221–224. Veterinary Research 20: 93–102 Conference on goats, Brasilia, Brazil: 505–511. 37. Orduña A, Almaraz A, Prado A, Gutiérrez 25. Kunda J S 2005 Epidemiology of human M P, García-Pascual A, Dueñas A, Cuervo 48. Waghela S, Wandera J G, Wagner G G 1980 brucellosis in northern Tanzania. PhD the- M, Abad R, Hernández B, Lorenzo B, Bratos Comparison of four serological tests in the sis, University of Edinburgh, Edinburgh, M A, Rodríguez-Torres A, 2000 Evaluation diagnosis of caprine brucellosis. Research in UK of an immunocapture-agglutination test Veterinary Science 28: 168–171 26. Kuzdas C D, Morse E V 1953 A selective (Brucellacapt) for the serodiagnosis of 49. WHO 2010 Online at: http://www.who.int/ medium for the isolation of brucellae from human brucellosis. Journal of Clinical Micro- foodsafety/zoonoses/en/ (accessed October contaminated materials. Journal of Bacteriol- biology 38: 4000–4005 2010). ogy 66: 502–504 38. Plommet M, Fensterbank R, Vassal L, 50. Wright P F, Nielsen K H 1990 Current and 27. Maichomo M W,McDermott J J, Arimi S M, Auclair J, Mocquot G 1988 Survival of future serological methods. In Adlams G Gathura P B, Mugambi T J, Muriuki S M Brucella abortus in ripened soft cheese made (ed.) Advances in brucellosis research. Texas A 2000. Study of brucellosis in a pastoral from naturally infected cow’s milk. Le Lait & M University Press, College Station: community and evaluation of the useful- 68: 115–120 305–320 ness of clinical signs and symptoms in 39. Raúl C, Mainar-Jaime P M, Muñoz J, de 51. WTO 2010 Online at: http://www.wto.org/ differentiating it from other flu-like dis- Miguel M J, Grilló, C M, Marín I Moriyón M english/tratop_e/sps_e/spsund_e.htm eases. African Journal of Health Sciences 7: I, Blasco J M 2005 Specificity dependence (accessed October 2010). 114–119 between serological tests for diagnosing 52. Young E J 1997 Brucellosis. In Emmerson A 28. Maipa A V,Alamanos Y 2005 Epidemiology bovine brucellosis in Brucella-free farms M, Hawkey P M, Gillespie S H (eds) Princi- of human brucellosis in a defined area of showing false positive serological reactions ples and practice of clinical bacteriology. Wiley, Northwestern Greece. Epidemiology and due to Yersinia enterocolitica O:9. Canadian Chichester: 337–348 Infection 133: 905–910 Veterinary Journal 46: 913–916 53. YoungE J 2000 Brucella spp. In Mandell G L, 29. Marin C M, Alabart J L, Blasco J M 1996 40. Russo G R, Pasquali P, Nenova R, Alexan- Bennett J E, Dolin R (eds) Principles and prac- Effects of antibiotics contained in two drov T, Ralchev S, Vullo V, Rezza G, tice of infectious diseases Vol. 2. Churchill Brucella selective media on the growth of Kantardjiev T 2009 Reemergence of human Livingstone, Philadelphia: 2386–2392 Brucella abortus, Brucella melitensis, and and animal brucellosis, Bulgaria. Emerging 54. Young E J 1991 Serologic diagnosis of Brucella ovis. Journal of Clinical Microbiology Infectious Diseases 15: 314–16 human brucellosis: analysis of 214 cases by 34: 426–428 41. Salari M H 2002 Seroepidemiological survey agglutination tests and review of the litera- 30. McDermott J J, Arimi S M 2002 Brucellosis of brucellosis among animal framers of ture. Review in Infectious Diseases 13: 359–372

212 0038-2809 Tydskr.S.Afr.vet.Ver. (2011) 82(4): 205–212