Molecular characteristics of HIV-1 subtype C and its impact on therapeutic outcome in Ethiopia

Abstract

HIV-1 is characterized by a high genetic diversity which poses several challenges and implications with regard to disease progression, drug resistance and outcome of antiretroviral therapy (ART). HIV-1 subtype C (HIV-1C) is the most rapidly expanding subtype accounting for half of the global disease and nearly all infections in Ethiopia, Southern Africa and India which are the regions with the highest burden of HIV-1 infection. Molecular characteristics of the virus in such epidemic success need to be explored to better understand this subtype.

In the thesis, we analysed plasma samples and patient data collected during 2009-2011 in a large country-wide cohort, Advanced Clinical Monitoring of ART (ACM) which was established to evaluate the longitudinal effectiveness of ART as practiced in real life in Ethiopia. The overall aim was to investigate the molecular characteristics of HIV-1C and its impact on first line ART outcome in Ethiopia. Both genotypic and phenotypic molecular techniques were employed to characterize different regions of the viral genome. In papers I and II, population sequencing (PBSS) of the V3 loop of the HIV-1 envelope from therapy naïve, patients failing therapy, as well as HIV-1C sequences from Ethiopia dated 1984-2003 was used to assess the molecular epidemiology of HIV-1C in different geographic regions and the trend of viral tropism over the last decades. We also investigated the utility of different genotypic tropism prediction tools and the impact of the predicted viral co-receptor tropism on the outcome of standard first line ART. Our results showed that the Ethiopian epidemic is still monophylogenetic, exclusively dominated by HIV-1C, CCR5 tropic viruses. Furthermore, baseline tropism had an impact on outcome of standard first line ART. While each tool predicted tropism with comparable frequency, there was yet a large discordance between the tools. We elucidated this discordance further in paper III by employing an in-house phenotypic tropism method compared with the prediction by bioinformatics tools used in paper II as well as in vitro sensitivity of HIV-1CEth strains for the co-receptor antagonist maraviroc. The results showed underestimation of R5 co-receptor usage by bioinformatics tools and effectiveness of maraviroc in HIV-1C. Expanding the exploration further to pol gene, we employed PBSS and next generation sequencing (NGS) to assess the prevalence of surveillance drug resistance mutations (sDRM) to reverse transcriptase- and protease-inhibitors as well as occurrence of DRM by NGS to the novel category of integrase strand inhibitors. The results in paper IV showed that NGS detected sDRM associated with RT- and PI- inhibitors more often than PBSS and major INSTI DRMs were found in minor viral variants. Furthermore, DRM identified before treatment was associated with a poorer treatment outcome.

In conclusion, viral tropism and drug resistance mutations at baseline have an impact on subsequent treatment outcome. Currently available genotypic tropism prediction tools need further improvement for use in HIV-1C. The Ethiopian epidemic remains uniquely dominated by R5 tropic HIV-1C since its introduction. Further investigations should be done to delineate associated molecular and epidemiological factors contributing to its uniqueness.