Background:  Nevirapine (NVP) is the most widely used NNRTI because it is cheap and available in fixed-drug combinations. However, it has drug interactions with rifampicin (RFM) and it is not clear whether they can be used efficaciously and safely together.

Methods:  We conducted a randomized open-label clinical trial to assess the efficacy and safety of 2 once-daily ART regimens in the treatment of HIV co-infected patients with TB. The trial was conducted at the Tuberculosis Research Centre, Chennai, India, between May 2006 and June 2008. HIV-infected patients with pulmonary or extra-pulmonary TB were enrolled and treated with a standard short-course anti-TB regimen (2EHRZ₃/4RH₃). At the end of the 2-month intensive phase, patients were randomized to receive either efavirenz (EFV; 600 mg) or NVP (400 mg, after a 14-day phase with 200 mg) along with didanosine (ddI; 250/400 mg) and lamivudine (3TC; 300 mg), all given once-daily in the morning. Treatment was directly observed on 3 days of the week for 24 weeks. Sputum smear and mycobacterial culture was performed every month. CD4 and viral load were counted at baseline, 4 weeks, 16 weeks, and 24 weeks of ART and liver function was monitored every 2 weeks for the first 2 months. A viral load >400 copies/mL or death at 24 weeks was considered an unfavorable response. Intent-to-treat analysis was used and the proportion of patients with failure compared using c² tests.

Results:  We randomized 127 HIV⁺ patients (99 males) with TB at a mean age of 35.5±7.6 (SD) years and mean weight of 42.3±8.0 kg to the EFV or NVP regimen. Median CD4 count was 84 cells/mm³ and median viral load was 310,000 copies/mL at baseline. Of the total, 94 patients had pulmonary TB, while 31 had extra-pulmonary disease. TB treatment was successful for 92% of patients. By intent-to-treat analysis, at 24 weeks of ART, 50 of 59 in the EFV arm and 38 of 57 patients in the NVP arm had a viral load of <400 copies/mL, p = 0.038. There were 5 deaths and 11 virologic failures in the NVP arm while there were no deaths and 5 failures in the EFV arm. There were 5 severe adverse events in the EFV and 6 in the NVP arm. The Data and Safety Monitoring Board (DSMB) recommended withholding intake into the NVP arm in December 2007 and stopping intake to the study in June 2008.

Conclusions:  Compared with ddI/3TC/EFV, a regimen of once-daily ddI/3TC/NVP was inferior with more frequent virologic failure and death. Once-daily NVP should not be used in patients initiating ART while on anti-TB therapy.