Dr Richard Bellamy alludes to the important fact, frequently ignored by immunologists, geneticists and epidemiologists, that tuberculosis has several different clinical forms.1Physicians have emphasised the difference between primary tuberculosis, which is comparable to Lurie’s susceptible rabbits with disseminated disease, and post primary tuberculosis, best characterised by smear positive pulmonary tuberculosis and Lurie’s “resistant” rabbits. HLA associations with tuberculosis have indeed been inconsistent when all forms of tuberculosis are included. However, the HLA association with DR2, and particularly with its subtype DR15 in linkage disequilibrium with DQ5, was found only in patients with smear positive pulmonary tuberculosis.2 3 These observations have been refined using DNA based HLA typing and have confirmed a link with the genes DRB1*1501 and DQB1*0502.4Antibody levels to epitopes of the 38kDa antigen of Mycobacterium tuberculosis restricted antigens were higher, suggesting an enhanced immune responsiveness in those with HLA-DR15.3The relative importance of the genes involved in susceptibility can be assessed by the gene frequency, but also by the attributable risk—that is, how much of the disease can be attributed to the presence or absence of a particular gene (34% with 95% confidence intervals of 16 to 43% were suggested for DR15 in one population3).

The Lurie experiment suggests that a comparison between patients with different forms of tuberculosis, matched by ethnic origin, may be valuable in identifying candidate genes for susceptibility to tuberculosis. Since smear positive pulmonary tuberculosis is responsible for transmission of the disease, an understanding of its pathogenesis will be especially important in finding new ways to control tuberculosis.