High risk of active tuberculosis in HIV-infected drug users with cutaneous anergy

To determine the incidence of active tuberculosis in human immunodeficiency virus (HIV)-seropositive and HIV-seronegative drug injectors with cutaneous anergy and to examine the effectiveness of isoniazid chemoprophylaxis in preventing tuberculosis among drug injectors with positive tuberculin test results.

Prospective observational study linked to an ongoing study of HIV infection within a New York City (NY) methadone program; subjects also underwent routine intradermal tuberculin testing and multiple-antigen delayed-type hypersensitivity skin testing. The 31-month study period ended December 31, 1990.

Anergic subjects and tuberculin reactors who were HIV seropositive were compared by HIV disease status and CD4+ T-lymphocyte levels. Tuberculosis incidence was calculated for anergics (none treated with isoniazid) and for treated and untreated tuberculin reactors, by HIV serological status.

Among those seropositive for HIV, anergic subjects had more advanced HIV disease and fewer CD4+ cells (median 0.33 vs 0.56 x 10(9)/L, P less than .01) compared with tuberculin reactors, although neither clinical status nor CD4+ cell counts consistently predicted anergy. Five (7.6%) of 68 anergic subjects who were HIV seropositive and none of 52 anergic subjects who were HIV seronegative (n = 18) or of unknown (n = 34) HIV serological status developed active tuberculosis during the study period (P less than .05). The tuberculosis incidence rate among anergic subjects who were HIV seropositive was 6.6 cases per 100 person-years (95% confidence interval [Cl], 2.1 to 15.3). Of 25 HIV-seropositive tuberculin reactors who did not receive or complete 12 months of isoniazid prophylaxis, tuberculosis incidence was 9.7 cases per 100 person-years (95% Cl, 2.6 to 24.7; P = 0.56, compared with the rate among anergic HIV seropositives); there were no cases of tuberculosis in 53.4 person-years of follow-up for 27 HIV-seropositive tuberculin reactors who received 12 months of prophylaxis (rate difference between treated and untreated groups, 9.7 cases per 100 person-years, 95% Cl, 1.3 to 18.0).

Drug injectors with cutaneous anergy who are seropositive for HIV are at high risk of active tuberculosis, similar to that among untreated HIV-seropositive tuberculin reactors. A decreased incidence of active tuberculosis was seen in HIV-seropositive tuberculin reactors receiving 12 months of isoniazid chemoprophylaxis, compared with untreated or partially treated subjects. These results support the routine use of delayed-type hypersensitivity testing to accompany tuberculin testing for drug injectors with known or suspected HIV infection, and consideration of isoniazid prophylaxis for anergic as well as tuberculin-reactive subjects who are HIV seropositive, in populations with a high prevalence of coexisting HIV and Mycobacterium tuberculosis infection.