Background: Although an increasing proportion of HIV-1 infections in Europe are due to non-subtype-B viruses, there are few data on response and resistance patterns of these viruses to antiretroviral therapy (ART). We analysed HIV-1 virological response to ART and genotype in previously untreated HIV-infected children in the PENTA 5 trial.

Methods: 113 children with samples assayed for resistance were randomised to ZDV+3TC (32) or ZDV+ABC (39) or 3TC+ABC (42) in a partially blinded, multi-centre, comparative study. Asymptomatic children (n=45) were also randomised to NFV/placebo; symptomatic children all received NFV (n=68). HIV-1 subtype was determined from the pol gene sequence obtained in the resistance assay. Genotypic and phenotypic resistance assays were undertaken by VIRCO. Heterogeneity across subtypes was assessed using model based Wald tests.

Results: The proportion of children infected with HIV-1 subtypes A, B, C, D, F, G, H, A/E, and A/G were 15%, 41%, 16%, 9%, 5%, 2%, 1%, 5%, and 7%, respectively. There was no evidence that HIV-1 subtype was associated with virological response (either change in RNA or the proportion with RNA < 400 or 50 copies/mL at 24 and 48 weeks). Assay failure rates for genotypic and phenotypic tests were inversely related to HIV-1 RNA (genotype p=0.002 and phenotype p=0.01) and were higher in non-B than B subtypes (genotype p=0.01 and phenotype p=0.03). No differences were observed between subtype B and non-B viruses in the emergence of L90M (p=1.00) or D30M (p=0.61). M184V was detected in virtually all subtypes. However, K65R emerging in children receiving abacavir (all 3TC+ABC) was only observed in subtype B viruses (p=0.08).

Conclusion: Virological response to nucleoside analogues with or without nelfinavir was not determined by virus subtype. Assay failure rates for resistance testing may be higher with non-B-subtype viruses. The selection of the K65R mutation by abacavir may be less common in non-B viruses but this requires further studies.