HIV-1 drug resistance remains an important cause of failure of 2nd-line antiretroviral therapy (ART) but limited resistance data exist from resource-limited settings (RLS). We analyzed the frequency and patterns of resistance at screening for ACTG A5288: a study of 3rd-line ART after prior exposure to NRTI, NNRTI and PI in RLS.
Plasma samples from a screening visit for enrollment into A5288 had protease and reverse transcriptase genotyped in real time by DAIDS-certified laboratories in RLS. Resistance mutations and scores were determined using the Stanford algorithm (v6.2). Subtype was determined by phylogenetic analysis. Associations of drug class resistance with HIV-1 subtype, screening HIV RNA, and nadir CD4 were evaluated.
665 plasma samples were available for analysis from Asia (171, 26%), South America (138; 21%), Eastern Africa (180; 27%), Southern Africa (170, 26%) and other locations (6; <1%). Median age was 41, 48% female, median HIV RNA 4.5 log10 cps/ml and median nadir CD4 65 cells/mm3. HIV subtype C (48%), B (20%) and A1 (18%) were most common. Prior exposure to NVP (63%) was more common than EFV (56%). At time of screening, TDF (67%) and 3TC (90%) were the most commonly prescribed NRTIs with either LPV/r (55%) or ATV/r (43%); 6% had RAL exposure. High-level or intermediate resistance was common (Table): 519 candidates (78%) had resistance to one or more drugs. 137 samples (21%) had resistance to 1 drug class, 207 (31%) to 2 drug classes, and 175 (26%) to all 3 drug classes (NRTI, NNRTI, PI). However, 461 (69%) showed susceptibility to 2nd-line regimens (1 NRTI + ATV/r or LPV/r) and the majority were susceptible or had low-level resistant to DRV/r (97%) and ETR (79%). Nadir CD4 count but not screening HIV RNA was significantly associated with the number of resistant drug classes but quantitative differences were small and unlikely to be helpful clinically (Table).
Highly divergent resistance profiles were observed among study candidates being evaluated for 3rd-line ART in RLS. The majority remained susceptible to 2nd-line regimens (69%) but others had high level resistance to 2 (31%) and 3 drug classes (26%). Routine clinical parameters were not discriminatory for the extent of resistance. These results indicate that objective measures of ART adherence and access to both resistance testing and newer ARVs are needed to guide 3rd-line ART in RLS.