Risk factors for relapse in human immunodeficiency virus type 1 infected adults with pulmonary tuberculosis

SETTING: A study conducted by the Uganda-Case Western Reserve University Research Collaboration in Kampala, Uganda, a country with high incidence rates of tuberculosis (TB) and human immunodeficiency virus type 1 (HIV-1) infection. OBJECTIVE: To assess clinical, microbiologic and radiographic factors associated with risk for relapse in HIV- infected adults treated for initial episodes of pulmonary TB. DESIGN: Nested case-control study within a randomized prospective clinical trial comparing the safety and efficacy of thiacetazone- and rifampicin-containing regimens for TB treatment in HIV-infected adults. RESULTS: The analysis was based on 119 patients who completed therapy. Median follow-up for all subjects was 22.3 months. Ten patients relapsed a median of 12.7 months after the end of therapy; seven of these were initially treated with the thiacetazone (T)-containing regimen. Each relapse case was matched to four controls by length of follow-up after initial TB treatment. In a univariate analysis risk for relapse was associated with treatment with the T-containing regimen (OR = 4.2, P = 0.08), age ≥30 yrs (OR = 2.9, P = 0.16), and irregular compliance (OR = 3.6, P = 0.1). Baseline anergy on Mantoux tuberculin skin testing, cavitary disease, radiographic extent of disease and sputum bacillary burden, two month culture negativity, and residual cavitary disease at the end of treatment did not differ between relapses and controls. CONCLUSION: Older HIV-1 infected patients, those with poor treatment compliance, and those being treated with T-containing regimens, may be at increased risk for relapse after TB treatment and require closer post- treatment surveillance. Risk for relapse in HIV-infected adults with pulmonary TB after treatment with a nine month rifampicin-containing regimen was low (3.1 per 100 person-years observation) compared with those treated with a thiacetazone-containing regimen (10.1 per 100 person-years observation).