Background:  Unlike HIV-1, there have been no trials addressing the choice of initial or second-line therapy for HIV-2 infection.  Thus, optimal treatment strategies for HIV-2 have not been defined. Previous studies suggest that HIV-2 is intrinsically resistant to many antiretrovirals in clinical use and that the patterns of resistance that emerge in HIV-2 differ from those  in HIV-1.

Methods:  Subjects were treated with antiretrovirals as part of a prospective, longitudinal cohort study of ARV therapy for HIV-2 infection in Senegal. HIV-2 pol genes from drug-treated subjects were genotyped for putative ARV resistance. To assess the relative sensitivities of HIV-2 strains to currently-available nucleoside reverse transcriptase inhibitors (NRTI), protease inhibitors (PI) and integrase inhibitors (INI), we developed single-cycle, cell culture-based assays that allow direct comparisons of the drug sensitivities of HIV-1 and HIV-2.  In addition, we used site-directed mutagenesis to assess the contributions of commonly-observed pol mutations to phenotypic ARV resistance.

Results:  In this study, 74 ARV-treated, HIV-2 infected Senegalese subjects were enrolled (females: 68%, median age: 46 years). Median CD4 counts and HIV-2 viral loads were 202 cells/mm³ (range: 2 to 1124) and 2.5 log10 copies/mL (range:<1.4-4.3). Subjects were followed for a mean of 396 days (range:1 to 1217 days), 7 died (9.5%; 7.2 per 100-person-years) and 23 clinical AIDS-related events were observed. RT substitutions K65R, Q151M and M184V/I and PR substitutions K7R, I54M, V62A, I82F, L90M and L99F were commonly detected in patient-derived HIV-2 pol sequences. Polymorphisms at sites that are associated with INI resistance in HIV-1 were also observed. Previously, we demonstrated that the combination of K65R, Q151M and M184V confers classwide NRTI resistance in HIV-2. Here, we show that saquinavir (EC₅₀ = 17 nM), lopinavir (EC₅₀ = 27 nM) and darunavir (EC₅₀ = 65 nM) are the most potent PI against wild-type HIV-2 and that HIV-2 is intrinsically resistant to both tipranavir and amprenavir (EC₅₀s>1000 nM). In addition, wild-type HIV-2 is highly susceptible to the integrase inhibitors raltegravir and elvitegravir. Using site-directed mutagenesis of HIV-2 PR and IN, we demonstrate the resistance profiles of individual protease and integrase mutations to available PIs and INIs respectively.

Conclusions:  Clinical outcomes of ARV-treated HIV-2 infection remain poor. Better understanding of ARV resistance will help optimize ARV regimens for HIV-2.