HIV-related tuberculosis due to Mycobacterium bovis.

The bovine tuberculosis eradication programmes in many of the industrially developed countries have been amongst the most effective control measures ever mounted against any bacterial disease. In Great Britain, for example, the percentage of cattle infected with the bovine tubercle bacillus, Mycobacterium bovis, dropped from around 18% in 1945 to 0.06% in 1965. The impact of such control measures on human tuberculosis of bovine origin appeared to be equally dramatic, so that many now regard this disease as a historical curiosity. In fact, the decline was not as steep as it appeared to be, owing to the failure of many laboratories to distinguish between the human and bovine tubercle bacilli, and a reluctance to notify the latter as this sometimes led to official enquiries which caused problems of patient confidentiality [1]. Although uncommon, cases of human tuberculosis due to M. bovis continue to be detected in the developed countries, but the majority of these appear to be due to reactivation of infection acquired in the days before the completion of the bovine tuberculosis eradication programmes. Thus, in South East England, very few patients of indigenous origin seen in the last 10 yrs were born after 1960, the date of completion of the eradication programme; and it is possible that the very few younger patients acquired their infection abroad [2, 3]. It therefore seemed likely that this form of human tuberculosis would become increasingly rare. With the advent of the human immunodeficiency virus (HIV) pandemic, however, this trend could be reversed. In order to comment on the possible effects of HIV infection on disease in humans due to M. bovis, it is first necessary to consider the similarities and differences between this form of tuberculosis and that caused by M. tuberculosis. Both types of tuberculosis have primary and postprimary forms, and the sites of disease reflect the route of infection. Thus, M. tuberculosis is usually inhaled and leads to primary pulmonary lesions, with occasional extrapulmonary lesions due to lymphatic and haematogenous dissemination; whilst M. bovis, which is usually acquired by consuming contaminated milk, is more likely to cause nonpulmonary lesions. Farm workers are, however, prone to primary pulmonary tuberculosis due to inhalation of infective droplets from diseased cattle. Thus, the incidence of pulmonary, relative to nonpulmonary, tuberculosis of bovine origin is higher in rural than in urban regions [4]. Reactivation of disease due to M. tuberculosis, after a period of dormancy, usually occurs in the lung where it often leads to open, smear-positive, infectious disease. There has been much argument as to whether reactivation of disease due to M. bovis occurs with the same frequency as that due to M. tuberculosis. This question is not easily resolved, as it is very difficult to determine whether a covert primary infection manifesting as tuberculin conversion is of human or bovine origin. Despite this difficulty, MAGNUS [5] used a mathematical model to demonstrate that the risk of developing late pulmonary tuberculosis following primary infection by M. tuberculosis was between two and ten times greater than that following infection by M. bovis. Nevertheless, reactivation disease due to M. bovis is encountered in regions where cattle tuberculosis is wellcontrolled, and a high proportion of such cases involve the lung. Thus, in South East England between 1977 and 1990 a total of 232 cases were diagnosed and 94 (41%) involved the lung [3]. The next most common site was the genitourinary tract (53 cases; 23%), followed by lymph node disease (39 cases; 17%). The existence of such postprimary pulmonary tuberculosis raises the question of infectivity. Human-to-cow transmission of infection has been well-documented [6]. (Although most of this transmission is by air borne infection, cattle in several herds have been infected as a result of farmers with genitourinary tuberculosis urinating on hay in the cowsheds [7]). The evidence for human-to-human transmission is less firm, and is largely anecdotal. A major problem in proving such transmission in humans is that bacteriologically positive tuberculosis only occurs in a minority of those infected, and often not until many years after infection. Thus, although many questions remain, it appears that M. bovis is less virulent than M. tuberculosis in humans, and, as a result, is less likely to proceed to postprimary infectious disease, and that human-to-human transmission leading to overt disease is a very uncommon event. If the apparent differences in virulence are the result of differing susceptibility to the host defence mechanisms, the immunosuppression induced by HIV infection could well annul these differences. The serious impact of HIV infection on human tuberculosis due to M. tuberculosis has been well-documented [8, 9]. The chance of a person infected with both pathogens of developing active tuberculosis rises from less Eur Respir J, 1994, 7, 1564–1566 DOI: 10.1183/09031936.94.07091564 Printed in UK all rights reserved Copyright ERS Journals Ltd 1994 European Respiratory Journal ISSN 0903 1936