Purpose: To compare long-term virologic, immunologic, and clinical outcomes in antiretroviral-naïve persons starting efavirenz (EFV)- versus nevirapine (NVP)-based regimens.
Method: The FIRST study randomized patients into three strategy arms: PI+NRTI, NNRTI+NRTI, and PI+NNRTI+NRTI. NNRTI was determined by optional randomization (NVP or EFV) or by choice. For this randomized substudy, the primary endpoint was HIV RNA >50 copies/mL after 8 months or death. Genotypic resistance testing was done at virologic failure (VF; HIV RNA >1,000 copies/mL at or after 4 months).
Results: 228 persons (111 randomized to EFV, 117 to NVP) were followed for median 5 years. Rates per 100 person years for the primary endpoint were 41.2 (EFV) and 42.8 (NVP; p = .59). The percent of persons with HIV RNA <50 copies/mL was similar throughout follow-up (p = .24), as were average increases in CD4+ cells (p = .30). 423 persons declining the substudy chose EFV; 264 chose NVP. There were 915 persons in the combined cohort (randomized and choice). In the combined cohort, the risk of VF and VF with any NNRTI or NRTI resistance or resistance of any class was significantly less for EFV compared to NVP.
Conclusions: EFV-based regimens as initial therapy resulted in a lower risk of VF and VF with resistance than did NVP-based regimens, although immunologic and clinical outcomes were similar.