Objective: To assess changes in clinical presentation and outcome of HIV-associated tuberculosis (TB) before and after widespread implementation of highly active antiretroviral therapy (HAART).
Methods: We reviewed clinical charts of HIV-infected patients with culture-confirmed pulmonary TB in two referral clinical centers in Rome, Italy. The 67 patients diagnosed in 1995 to 1996 were compared with 51 patients diagnosed in 1997 to 1998. To analyze factors associated with survival we used a Cox model including antiretroviral therapy as a time-dependent covariate.
Results: Patients diagnosed in 1997 to 1998 were more likely to have TB as the first AIDS-defining illness (78% versus 58%, p <.05), to have HIV diagnosed <2 months before TB (33% vs. 7%, p <.005) and to have typical chest radiograph pattern (45% vs. 25%, p <.05), and had a higher CD4(+) count (median 105 vs. 43, p <.005). Survival at 1 year was 80% for patients diagnosed in 1997 to 1998 vs. 65% for those diagnosed in 1995 to 1996 (p by log-rank =.02). After adjusting at multivariate analysis, time period of diagnosis was not confirmed as associated with survival (hazard ratio, 1.05; 95% confidence interval, 0.39--2.81). Age, CD4+ cell count <25/mm(3), and AIDS-defining illnesses before TB diagnosis were all associated with an higher risk of death, whereas a decreased risk of death was observed in patients starting a triple combination antiretroviral therapy after TB diagnosis (hazard ratio, 0.14; 95% confidence interval, 0.03--0.57).
Conclusions: Cases of HIV-associated TB occurring in patients with advanced immunosuppression and presenting with atypical radiologic appearance tend to be relatively less common in the HAART era. HAART is a major factor in prolonging survival in these patients.