Background:  Single-dose nevirapine (sdNVP) reduces mother to child transmission of HIV, but can select for NVP-resistant virus. Optimal ART for women with prior sdNVP exposure is unknown.

Methods:  In 7 African countries, 243 HIV-1-infected women who had taken intrapartum sdNVP at least 6 months earlier and had screening CD4 count <200 cells/mm³ were randomized to ART with tenofovir/emtricitibine (TDF/FTC) and either lopinavir/ritonavir (LPV/r) twice daily (n = 120) or NVP twice daily (n = 123). The primary endpoint was time from randomization to either death or confirmed virologic failure (plasma HIV-1 RNA <1 log₁₀ below baseline 12 weeks after treatment initiation, or ≥400 copies/mL at or after 24 weeks). The primary analysis was modified intent to treat, excluding 2 women who did not start study treatment. In October 2008, the DSMB recommended early release of study findings. Data are included through this date; results are not reported for an ongoing, parallel trial in 502 women with no prior sdNVP (per DSMB).

Results:  At entry, median values were as follows:  age 31 years, CD4 139 cells/mm³, HIV-1 RNA 5.15 log₁₀ copies/mL, time since last sdNVP use 17 months, and duration of follow-up 73 weeks. All women recalled sdNVP use; 73% had written documentation that sdNVP was provided. Significantly more women in the NVP arm (31, or 26%) than the LPV/r arm (10, or 8%) reached the primary endpoint (p = 0.0007). Of these, 36 had virologic failure (27 in NVP, 9 in LPV/r arm); 5 died without preceding virologic failure (4 in NVP, 1 in LPV/r arm). None of the 5 deaths were reported to be treatment-related. The difference between the NVP and LPV/r arms appeared to decrease with increasing time between last sdNVP exposure and ART initiation. Among women starting ART 6 to <12 months after sdNVP, 15 (37%) and 1 (3%) in the NVP and LPV/r arms, respectively, reached the primary endpoint, vs 12 (26%) and 6 (12%) when last sdNVP exposure was 12 to < 24 months prior, and 4 (12%) and 3 (10%) when last sdNVP exposure was ≥2 years prior. Baseline genotypes, run retrospectively in 239 women, showed NVP resistance in 33 (14%; K103N in 28 of 33); of these 33, 11 of 15 (73%) in the NVP arm reached the primary endpoint vs 1 of 18 (6%) in the LPV/r arm. Among the 206 women without baseline NVP resistance, 18% (NVP arm) vs 9% (LPV/r arm) reached the primary endpoint. More women (15, 12%) discontinued NVP due to adverse events than LPV/r (1, 1%).

Time to virologic failure or to death

Conclusions:  LPV/r + TDF/FTC is superior to NVP + TDF/FTC in women with prior exposure to peripartum sdNVP.