2-YEAR FOLLOW-UP OF PERSONS WITH HIV-1-ASSOCIATED AND HIV-2-ASSOCIATED PULMONARY TUBERCULOSIS TREATED WITH SHORT-COURSE CHEMOTHERAPY IN WEST-AFRICA

Objective: To assess the response to therapy for tuberculosis using rifampicin-containing short-course chemotherapy, and to compare recurrence and mortality rates in seronegative persons and those with HIV-1, HIV-2, and dual serologic reactivity in West Africa. Methods: A cohort of 835 adult patients (167 HIV-1-positive, 143 HIV-2-positive, 243 dual-reactive, 282 HIV-negative) with smear-positive pulmonary tuberculosis was followed for 2 years under programme conditions. Standard self-administered treatment was daily rifampicin and isoniazid for 6 months, and in addition pyrazinamide during the first 2 months. Outcomes evaluated were rates of completion of therapy, cure, failure of treatment, recurrence after cure, and mortality. Results: HIV-positive patients had lower rates of completion of therapy (65-73%) than seronegative patients (79%), mainly because of increased mortality. Among patients completing therapy, failure of treatment was similarly low in HIV-positive (2%) and seronegative patients (1%). Recurrence rates after cure did not differ significantly in the 18 months of follow-up in the four serologic groups (3-7%). The respective mortality rates for HIV-1-positive, HIV-2-positive, and dually reactive patients were 20.3, 8.3, and 25.5 per 100 person-years (PY), compared with 2.2 per 100 PY among seronegatives. Conclusions: Rifampicin-containing short-course chemotherapy for pulmonary tuberculosis is associated with similar cure and recurrence rates in HIV-positive and HIV-negative persons completing 6 months of therapy. HIV-2 infection is associated with more favourable survival than HIV-1 infection or dual reactivity, even when AIDS-defining illness is already present. However, mortality is significantly increased in all seropositive groups compared with HIV-negative tuberculosis patients; thus, establishing the causes of this increased mortality is a priority.