Background:  National initiatives offering NNRTI-based ART have recently commenced in Sub-Saharan Africa. In developed countries, efavirenz (EFV) is the NNRTI of choice, but in Sub-Saharan Africa, where women are the majority of those on ART, nevirapine (NVP) is widely used. Therefore, direct EFV vs NVP comparisons similar to the 2NN trial (an open-label, randomized, comparative trial of NVP- vs EFV-based HAART regimens) are urgently needed. The Tshepo Study is the first clinical trial evaluating the long-term efficacy and tolerability of ART-treated adults in Botswana.

Methods:  Completed, open-label, randomized study of 650 antiretroviral-naïve adults treated for 3 years, using a 3x2x2 factorial design comparing efficacy and tolerability among:  A:  zidovudine (ZDV)/lamivudine (3TC) vs ZDV/didanosing (ddI) vs stavudine (d4T)/3TC; B:  EFV vs NVP, and C:  community-based directly observed therapy (DOT) vs standard of care. This abstract focuses on factor B (EFV vs NVP). Baseline viral load was >55,000 and 2 CD4⁺ strata:  <201 vs 201 to 350. Intent-to-treat analyses included survival analyses, longitudinal ANOVA, and logistic regression.

Results:  At baseline there were 451 females (69.4%) and 199 males; median age 33.3 years, median baseline CD4⁺ 199 (IQR 134 to 239), plasma HIV-1 RNA 204,000 (IQR 81,000 to 447,000). Median follow-up 151 weeks (IQR 125 to 156); overall loss to follow-up 8.3%. EFV- vs NVP-treated patients had equivalent immunologic and virologic suppression rates: at 1, 2, and 3 years, respectively, median CD4⁺ increase from baseline were 138 (IQR 74 to 223), 217 (IQR 122 to 332), and 258 (IQR 146 to 387), and 92.5%, 91.7%, and 91.1% had undetectable HIV-1 RNA levels. Kaplan-Meier survival estimate was 94.0%, with no difference by NNRTI. Rates of virologic failure with genotypic resistance mutations were 4.5% NVP vs 2.5% EFV, log rank p = 0.05. No difference was found in overall virologic failure rates. Toxicities required treatment modifications in 124 patients had (25.2% NVP vs 12.9% EFV, p = 0.0001), notably NVP-related hepatoxicity (2.3%) and Stevens-Johnson Syndrome (1.8%).

Conclusions:  There were differences in virologic and tolerability outcomes, but not CD4⁺ response among randomized adults receiving NVP- vs EFV-based ART. Treatment-modifying toxicities and virologic failure with resistance were higher among those receiving NVP, stressing the need to closely monitor safety and efficacy during treatment, but overall virologic failure rates did not differ across NNRTI. Until further data are available, NVP should continue as a component of ART in Sub-Saharan Africa, where 65 to 70% of adults on ART are female, pregnancy rates are high, and EFV-related teratogenicity is a risk.