1253
PUBLIC HEALTH
Mycobacterium bovis infection in human beings in contact with elk (Cervus elaphus) in Alberta, Canada
Human infection with Mycobacterium bovis is rare in developed countries because of milk pasteurisation and the slaughter of infected cattle. An epizootic of M bovis infection in domesticated elk (Cervus elaphus) in Alberta, Canada, which started in April, 1990, prompted us to seek human contacts of elk herds. There were 446 identified contacts, in 394 of whom tuberculin skin tests were done. Of 81 contacts who were skin-test positive, 50 had been in contact with culture-positive animals. 6 of 106 subjects tested a second time became tuberculin positive. 1 case of active M bovis infection was diagnosed by sputum culture. The mode of transmission of M bovis from these farm animals to man is likely to be aerosol isation of infected particles.
Because of the apparent susceptibility of farmed Cervidae (deer) to M bovis infection, and the evidence of spread to man, control measures to prevent human infection should be developed. Introduction
Mycobacteriwn bovis infection is rarely reported in man in developed countries, but before pasteurisation of milk and slaughter of infected cattle, M bovis accounted for for30 % of human tuberculosis cases in the USA. In Alberta, Canada, between 1982 and 1989, there were 1493 human isolates of M tuberculosis and 1 of M bovis.2 The isolation of M bovis from a domesticated elk (Cervus elaphus var canadensis, known as red deer in Europe) in Alberta prompted us to follow up human contacts.
Methods Index animal case In April, 1990, a veterinary surgeon was asked to examine a sick elk. The animal was part of a herd of 150 on one of the 110 registered elk farms in Alberta. A retropharyngeal abscess was noted and was treated with streptomycin injections for 1 week with no improvement. Over the next 6 weeks two drainage procedures were done in a wind shelter. Penicillin and sodium iodide were given for suspected actinomycosis. In July, the lesion had extended to the cervical area. Excision was unsuccessful and the animal was killed. At necropsy, tuberculosis was suspected on the basis of extensive caseating and purulent lesions of cervical, thoracic, and mesenteric lymph nodes and of the lungs, liver, and spleen. Granulomas with numerous acid-fast organisms were confirmed by the Agriculture Alberta laboratory. The isolation of M bovis in November, 1990, was reported to the Tuberculosis Services, Alberta Health.
Contact follow-up The owner of the farm from which the sick animal came and his family had tuberculin skin tests and were found to be negative. The veterinary surgeon and his assistant had positive reactions to five
tuberculin units of purified protein derivative. Both individuals symptom free and had normal chest radiographs. Sputum from the veterinary surgeon (one of three specimens submitted up to December, 1990) yielded a mycobacterial growth in January, 1991. Treatment was started with isoniazid, rifampicin, and pyrazinamide. The isolate was confirmed as M bovis in March, 1991. The organism was sensitive to isoniazid, rifampicin, ethambutol, and streptomycin, but resistant to pyrazinamide. Domesticated elk herds in Alberta were investigated for evidence of M bovis infection. 69 herds were quarantined because they contained animals purchased from the index herd. 32 of these herds contained animals with positive tuberculin skin tests and 7 contained animals with lesions that grew M bovis. 2 herds were slaughtered and disposed of at a single rendering plant where skinning preceded grinding of the carcasses. The end product, a chicken feed, was heated to 120°C for 15 min before dehydration. Individuals who had handled animals since April,1990, including private veterinary surgeons, farmers, and their employees, Agriculture Canada field and plant inspectors, Agriculture Alberta veterinary surgeons and technicians, and rendering plant staff, were notified of the potential risk of infection and asked to assess their degree of exposure and report for testing at their local health centre. Subsequently, the handling of animal hides was shown to involve some risk of exposure while the product was being off-loaded and spread out for grading and salting. were
Results
By July 31, 1991, 446 contacts had been identified, of whom 394 were assessed. Tuberculin skin testing was negative in 313 individuals (2 with a history of BCG [bacille Calmette-Guerin] vaccination). 33 had a history of previous positive tests (11with BCG vaccination) and 48 were positive for the first time (20 with BCG vaccination). The overall initial rate of positive tuberculin tests among contacts was 21% (81 of 394). This rate is higher than the estimated 10% rate of positivity for the tuberculin skin test in human beings in Alberta. The highest rate of previous and newly positive skin tests was found among those who had had contact with herds containing culture-positive animals (50 of 182) (table I). To date, there have been 6 conversions from a negative to a positive result among 106 subjects who have had a second tuberculin test. When results of the first tuberculin test were sorted by contacts’ occupation, 8 (27%) of 30 veterinary surgeons dealing with tuberculin-positive herds were tuberculin positive, whereas only 1 (5%) of 20 veterinary surgeons dealing with tuberculin-negative herds was positive (table II). Tanning-plant workers had the highest rate of positive tuberculin tests (23 of 36,64%). Most of these workers were born outside Canada in countries (Portugal and Vietnam) where tuberculosis rates are substantially higher than in Canada. All 6 individuals who were tuberculin positive on ADDRESS: Division of Infectious Diseases, Department of Medicine, 2E4 11 Walter MacKenzie Health Sciences Centre, University of Alberta, Edmonton, Canada T6G 2J3 (A. Fanning, FRCPC, S. Edwards, BSc). Correspondence to Dr Anne Fanning.
1254
TABLE I-RESULTS OF TUBERCULIN TESTS BY JULY 31,1991, IN SUBJECTS REPORTING CONTACT WITH ELK HERDS
Data
are no
of subjects unless otherwise stated;
[]
=
no
of subjects with known history of BCG vaccination.
TABLE II-RESULTS OF THE FIRST TUBERCULIN TEST BY JULY 31,1991, BY OCCUPATION OF SUBJECTS REPORTING CONTACT WITH ELK HERDS
Data are given as no of subjects; [ ] no of subjects with known history of Subjects whose type of contact is unknown are excluded from this table
BCG vaccination-
second testing had been in contact with culture-positive herds. These 6 individuals were 2 workers from the rendering plant-Iwho unloaded carcasses and a second who skinned them-a field inspector dealing with a heavily infected herd, 2 laboratory technicians carrying out necropsies, and a meat inspector seconded to the rendering plant. All tuberculin-positive individuals were advised to have a medical assessment, including chest radiography and sputum culture. As a result of assessment, 9 individuals require no further follow-up, and isoniazid prophylaxis was recommended in 42, of whom 33 accepted. 12 subjects will be followed up at 6, 18, and 30 months with chest radiography and sputum culture. Apart from the index veterinary surgeon, no contacts have yielded positive sputum cultures.
Discussion The best known means of transmission of M bovis to man is by ingestion of infected milk, but there is also clear evidence of aerosol dissemination.3 Before pasteurisation, when milkborne M bovis infections greatly outnumbered those from droplet spread, most lesions were extrapulmonary (gastrointestinal, lymph node, renal). At that time, M bovis accounted for 5% of tuberculosis cases in the UK. In south-east England, 14% of human mycobacterial isolates between 1977 and 1984 were M bovis ;4 60% of lesions were extrapulmonary. In Ontario, Canada, 05% of all human tuberculosis cases between 1964 and 1970 were due to M bovis;5 58% of lesions were extrapulmonary. Of the 13 patients from Ontario who were born in Canada, 10 were possibly infected in their youth.
possibility of present day infection with M bovis be ignored. Among abattoir workers in South Australia, 4 cases of M bovis infection from four abattoirs were reported within 2 years in the 1980s, and 54% of workers were found to be tuberculin positive.6 A fifth case was identified upon further investigation. 4 of the 5 patients had lung lesions and 1 had renal tuberculosis. There was evidence of intrafamilial spread from 1 of the patients. In Queensland, Australia, 67 of the 87 patients with M bovis infection reported in the past 35 years had pulmonary The
cannot
infection and all but 29 had substantial contact with cattle or known cases of bovine tuberculosis.7 In Alberta, human tuberculosis due to M tuberculosis is uncommon-ie, 6 4cases/100 000 population in 1990. Rates are highest in the elderly, in aboriginals (native North Americans and Inuit; 40/100 000), and in people born outside Canada. BCG vaccine is given to aboriginal children at birth. M tuberculosis is almost invariably spread by aerosolised droplets from the cough of a person with a large productive pulmonary lesion. Adult reactivated tuberculosis is predominantly a pulmonary disease (80%), but lesions may affect any organ system. People identified as recently or previously infected and at high risk of reactivation can be offered isoniazid prophylaxis for 9-12 months, which yields 80-90% protection against late reactivation. The high rate of initial tuberculin reactivity and our finding of the conversion of 6 subjects to a positive tuberculin test strongly suggest human infection. Although the veterinary surgeon with active tuberculosis was not ill, the culture of M bovis from his sputum confirms ,
1255
transmission. M bovis infection was reported in an elk herd in South Dakota in 1980;H 3 people in contact with the herd converted to a positive tuberculin skin test. We believe that M bovis may have been transmitted by aerosolisation of the organism, either by the cough of infected elk or by creation of infectious particles when the rendering plant and necropsy laboratory were hosed down. Although the veterinary surgeon worked out-of-doors, where dilution of aerosols would be expected, he became infected. The risk of M bovis infection proceeding to active disease in humans is not known. The tuberculosis control measures reported here were instituted in the belief that M bauis and M tuberculosis are similar with respect to pathugenic mechanism, risk of development of disease, and opportunity for intervention. In view of the increase in number of domesticated herds of Cervidae (deer) and evidence of spread of M bovis from animal to man, as well as from animal to animal, there is a need to develop control measures for the cervid industry to reduce to a minimum risk of human infection. People at risk are handlers of sick animals and handlers of carcasses. Screening and regular follow up of workers at risk are essential until the disease is eradicated from this animal population. The organism is usually sensitive to isoniazid,
rifampicin, and ethambutol, but not to pyrazinamide. Although isoniazid prophylaxis seems to be appropriate for individuals who are infected but do not have active disease, the degree of protection offered by prophylaxis has not been studied for M bcrais infection. We thank Ur Gerald Hauer, and the staff of Tuberculosis Services, Alberta Health, Agriculture Canada, and Agiicultural Alberta tor the collection of data.
REFERENCES 1. Karlson AG, Carr DT. Tuberculosis caused by Mycobacterium bovis. Ann Intern Mcd 1970; 73: 979-83. 2. Fanning A, Kanik G, Mah M, Chomyc S, Flowerdew G. Mycobacterial isolates: distribution of species in Alber ta. Am Rev Respir Dis 1991; 143:
A284. 3. Collins CH, Grange JM. Zoonotic implication of Mycobacterium bovis infection. Ir Vet J 1957; 41: 363-66. 4. Yates MD, Grange JM, Collins CH. The nature of mycobacterial disease in south-east England, 1977-84. J Epidemiol Commun Health 1986; 40: 295-300. 5. Wigle WD, Ashley MJ, Kellough EM, Caseus M. Bovine tuberculosis in humans in Ontano. Am Rev Respir Dis 1973; 106: 528-34. 6. Robinson P, Morris D, Antic R Mycobacterium bovis as an occupational hazard in abattoir workeis. Aust N Z J Med 1988; 18: 701-03 7. Georghion P, Patel AM, Konstantinos A. Mycobacterium bovis as an occupational hazard in abattoir worker s. Aust N Z J Med 1989; 19: 409. 8. Stumpff CD. Epidemiological study of an outbreak of bovine tuberculosis in confined elk herds. Proc US Animal Health Assoc 1982; 86: 524-27.
VIEWPOINT A temporary chef de
requirement for surgeons who are training in English-speaking countries to spend long periods abroad The
in relation to local factors such as the and timing of jobs, requirements for special availability with reintegration at the end of a difficulties skills, and the sabbatical, availability of funds and fellowships. The choice of venue for such placements is also subject to many influences, not the least of which is language; as a result, surgical training in Europe remains a largely unknown quantity. There are striking differences between the French surgical system, for instance, and those in Australia or the UK. These contrasts offer a unique oppci Lunit.y to further one’s surgical experience. If an a
PUBLIC HEALTH
Mycobacterium bovis infection in human beings in contact with elk (Cervus elaphus) in Alberta, Canada
Human infection with Mycobacterium bovis is rare in developed countries because of milk pasteurisation and the slaughter of infected cattle. An epizootic of M bovis infection in domesticated elk (Cervus elaphus) in Alberta, Canada, which started in April, 1990, prompted us to seek human contacts of elk herds. There were 446 identified contacts, in 394 of whom tuberculin skin tests were done. Of 81 contacts who were skin-test positive, 50 had been in contact with culture-positive animals. 6 of 106 subjects tested a second time became tuberculin positive. 1 case of active M bovis infection was diagnosed by sputum culture. The mode of transmission of M bovis from these farm animals to man is likely to be aerosol isation of infected particles.
Because of the apparent susceptibility of farmed Cervidae (deer) to M bovis infection, and the evidence of spread to man, control measures to prevent human infection should be developed. Introduction
Mycobacteriwn bovis infection is rarely reported in man in developed countries, but before pasteurisation of milk and slaughter of infected cattle, M bovis accounted for for30 % of human tuberculosis cases in the USA. In Alberta, Canada, between 1982 and 1989, there were 1493 human isolates of M tuberculosis and 1 of M bovis.2 The isolation of M bovis from a domesticated elk (Cervus elaphus var canadensis, known as red deer in Europe) in Alberta prompted us to follow up human contacts.
Methods Index animal case In April, 1990, a veterinary surgeon was asked to examine a sick elk. The animal was part of a herd of 150 on one of the 110 registered elk farms in Alberta. A retropharyngeal abscess was noted and was treated with streptomycin injections for 1 week with no improvement. Over the next 6 weeks two drainage procedures were done in a wind shelter. Penicillin and sodium iodide were given for suspected actinomycosis. In July, the lesion had extended to the cervical area. Excision was unsuccessful and the animal was killed. At necropsy, tuberculosis was suspected on the basis of extensive caseating and purulent lesions of cervical, thoracic, and mesenteric lymph nodes and of the lungs, liver, and spleen. Granulomas with numerous acid-fast organisms were confirmed by the Agriculture Alberta laboratory. The isolation of M bovis in November, 1990, was reported to the Tuberculosis Services, Alberta Health.
Contact follow-up The owner of the farm from which the sick animal came and his family had tuberculin skin tests and were found to be negative. The veterinary surgeon and his assistant had positive reactions to five
tuberculin units of purified protein derivative. Both individuals symptom free and had normal chest radiographs. Sputum from the veterinary surgeon (one of three specimens submitted up to December, 1990) yielded a mycobacterial growth in January, 1991. Treatment was started with isoniazid, rifampicin, and pyrazinamide. The isolate was confirmed as M bovis in March, 1991. The organism was sensitive to isoniazid, rifampicin, ethambutol, and streptomycin, but resistant to pyrazinamide. Domesticated elk herds in Alberta were investigated for evidence of M bovis infection. 69 herds were quarantined because they contained animals purchased from the index herd. 32 of these herds contained animals with positive tuberculin skin tests and 7 contained animals with lesions that grew M bovis. 2 herds were slaughtered and disposed of at a single rendering plant where skinning preceded grinding of the carcasses. The end product, a chicken feed, was heated to 120°C for 15 min before dehydration. Individuals who had handled animals since April,1990, including private veterinary surgeons, farmers, and their employees, Agriculture Canada field and plant inspectors, Agriculture Alberta veterinary surgeons and technicians, and rendering plant staff, were notified of the potential risk of infection and asked to assess their degree of exposure and report for testing at their local health centre. Subsequently, the handling of animal hides was shown to involve some risk of exposure while the product was being off-loaded and spread out for grading and salting. were
Results
By July 31, 1991, 446 contacts had been identified, of whom 394 were assessed. Tuberculin skin testing was negative in 313 individuals (2 with a history of BCG [bacille Calmette-Guerin] vaccination). 33 had a history of previous positive tests (11with BCG vaccination) and 48 were positive for the first time (20 with BCG vaccination). The overall initial rate of positive tuberculin tests among contacts was 21% (81 of 394). This rate is higher than the estimated 10% rate of positivity for the tuberculin skin test in human beings in Alberta. The highest rate of previous and newly positive skin tests was found among those who had had contact with herds containing culture-positive animals (50 of 182) (table I). To date, there have been 6 conversions from a negative to a positive result among 106 subjects who have had a second tuberculin test. When results of the first tuberculin test were sorted by contacts’ occupation, 8 (27%) of 30 veterinary surgeons dealing with tuberculin-positive herds were tuberculin positive, whereas only 1 (5%) of 20 veterinary surgeons dealing with tuberculin-negative herds was positive (table II). Tanning-plant workers had the highest rate of positive tuberculin tests (23 of 36,64%). Most of these workers were born outside Canada in countries (Portugal and Vietnam) where tuberculosis rates are substantially higher than in Canada. All 6 individuals who were tuberculin positive on ADDRESS: Division of Infectious Diseases, Department of Medicine, 2E4 11 Walter MacKenzie Health Sciences Centre, University of Alberta, Edmonton, Canada T6G 2J3 (A. Fanning, FRCPC, S. Edwards, BSc). Correspondence to Dr Anne Fanning.
1254
TABLE I-RESULTS OF TUBERCULIN TESTS BY JULY 31,1991, IN SUBJECTS REPORTING CONTACT WITH ELK HERDS
Data
are no
of subjects unless otherwise stated;
[]
=
no
of subjects with known history of BCG vaccination.
TABLE II-RESULTS OF THE FIRST TUBERCULIN TEST BY JULY 31,1991, BY OCCUPATION OF SUBJECTS REPORTING CONTACT WITH ELK HERDS
Data are given as no of subjects; [ ] no of subjects with known history of Subjects whose type of contact is unknown are excluded from this table
BCG vaccination-
second testing had been in contact with culture-positive herds. These 6 individuals were 2 workers from the rendering plant-Iwho unloaded carcasses and a second who skinned them-a field inspector dealing with a heavily infected herd, 2 laboratory technicians carrying out necropsies, and a meat inspector seconded to the rendering plant. All tuberculin-positive individuals were advised to have a medical assessment, including chest radiography and sputum culture. As a result of assessment, 9 individuals require no further follow-up, and isoniazid prophylaxis was recommended in 42, of whom 33 accepted. 12 subjects will be followed up at 6, 18, and 30 months with chest radiography and sputum culture. Apart from the index veterinary surgeon, no contacts have yielded positive sputum cultures.
Discussion The best known means of transmission of M bovis to man is by ingestion of infected milk, but there is also clear evidence of aerosol dissemination.3 Before pasteurisation, when milkborne M bovis infections greatly outnumbered those from droplet spread, most lesions were extrapulmonary (gastrointestinal, lymph node, renal). At that time, M bovis accounted for 5% of tuberculosis cases in the UK. In south-east England, 14% of human mycobacterial isolates between 1977 and 1984 were M bovis ;4 60% of lesions were extrapulmonary. In Ontario, Canada, 05% of all human tuberculosis cases between 1964 and 1970 were due to M bovis;5 58% of lesions were extrapulmonary. Of the 13 patients from Ontario who were born in Canada, 10 were possibly infected in their youth.
possibility of present day infection with M bovis be ignored. Among abattoir workers in South Australia, 4 cases of M bovis infection from four abattoirs were reported within 2 years in the 1980s, and 54% of workers were found to be tuberculin positive.6 A fifth case was identified upon further investigation. 4 of the 5 patients had lung lesions and 1 had renal tuberculosis. There was evidence of intrafamilial spread from 1 of the patients. In Queensland, Australia, 67 of the 87 patients with M bovis infection reported in the past 35 years had pulmonary The
cannot
infection and all but 29 had substantial contact with cattle or known cases of bovine tuberculosis.7 In Alberta, human tuberculosis due to M tuberculosis is uncommon-ie, 6 4cases/100 000 population in 1990. Rates are highest in the elderly, in aboriginals (native North Americans and Inuit; 40/100 000), and in people born outside Canada. BCG vaccine is given to aboriginal children at birth. M tuberculosis is almost invariably spread by aerosolised droplets from the cough of a person with a large productive pulmonary lesion. Adult reactivated tuberculosis is predominantly a pulmonary disease (80%), but lesions may affect any organ system. People identified as recently or previously infected and at high risk of reactivation can be offered isoniazid prophylaxis for 9-12 months, which yields 80-90% protection against late reactivation. The high rate of initial tuberculin reactivity and our finding of the conversion of 6 subjects to a positive tuberculin test strongly suggest human infection. Although the veterinary surgeon with active tuberculosis was not ill, the culture of M bovis from his sputum confirms ,
1255
transmission. M bovis infection was reported in an elk herd in South Dakota in 1980;H 3 people in contact with the herd converted to a positive tuberculin skin test. We believe that M bovis may have been transmitted by aerosolisation of the organism, either by the cough of infected elk or by creation of infectious particles when the rendering plant and necropsy laboratory were hosed down. Although the veterinary surgeon worked out-of-doors, where dilution of aerosols would be expected, he became infected. The risk of M bovis infection proceeding to active disease in humans is not known. The tuberculosis control measures reported here were instituted in the belief that M bauis and M tuberculosis are similar with respect to pathugenic mechanism, risk of development of disease, and opportunity for intervention. In view of the increase in number of domesticated herds of Cervidae (deer) and evidence of spread of M bovis from animal to man, as well as from animal to animal, there is a need to develop control measures for the cervid industry to reduce to a minimum risk of human infection. People at risk are handlers of sick animals and handlers of carcasses. Screening and regular follow up of workers at risk are essential until the disease is eradicated from this animal population. The organism is usually sensitive to isoniazid,
rifampicin, and ethambutol, but not to pyrazinamide. Although isoniazid prophylaxis seems to be appropriate for individuals who are infected but do not have active disease, the degree of protection offered by prophylaxis has not been studied for M bcrais infection. We thank Ur Gerald Hauer, and the staff of Tuberculosis Services, Alberta Health, Agriculture Canada, and Agiicultural Alberta tor the collection of data.
REFERENCES 1. Karlson AG, Carr DT. Tuberculosis caused by Mycobacterium bovis. Ann Intern Mcd 1970; 73: 979-83. 2. Fanning A, Kanik G, Mah M, Chomyc S, Flowerdew G. Mycobacterial isolates: distribution of species in Alber ta. Am Rev Respir Dis 1991; 143:
A284. 3. Collins CH, Grange JM. Zoonotic implication of Mycobacterium bovis infection. Ir Vet J 1957; 41: 363-66. 4. Yates MD, Grange JM, Collins CH. The nature of mycobacterial disease in south-east England, 1977-84. J Epidemiol Commun Health 1986; 40: 295-300. 5. Wigle WD, Ashley MJ, Kellough EM, Caseus M. Bovine tuberculosis in humans in Ontano. Am Rev Respir Dis 1973; 106: 528-34. 6. Robinson P, Morris D, Antic R Mycobacterium bovis as an occupational hazard in abattoir workeis. Aust N Z J Med 1988; 18: 701-03 7. Georghion P, Patel AM, Konstantinos A. Mycobacterium bovis as an occupational hazard in abattoir worker s. Aust N Z J Med 1989; 19: 409. 8. Stumpff CD. Epidemiological study of an outbreak of bovine tuberculosis in confined elk herds. Proc US Animal Health Assoc 1982; 86: 524-27.
VIEWPOINT A temporary chef de
requirement for surgeons who are training in English-speaking countries to spend long periods abroad The
in relation to local factors such as the and timing of jobs, requirements for special availability with reintegration at the end of a difficulties skills, and the sabbatical, availability of funds and fellowships. The choice of venue for such placements is also subject to many influences, not the least of which is language; as a result, surgical training in Europe remains a largely unknown quantity. There are striking differences between the French surgical system, for instance, and those in Australia or the UK. These contrasts offer a unique oppci Lunit.y to further one’s surgical experience. If an a
