Antiretroviral Therapy for HIV: Do Subtypes Matter?

About 2 years ago, in a past issue of Clinical Infectious Diseases, Geretti et al [1] reported the virologic and immunologic response to highly active antiretroviral therapy (HAART) in drug-naive human immunodeficiency virus type 1 (HIV-1) infected patients comparing subtype B (n 5 1150) versus non-B (n 5 566) in the United Kingdom. Response (,50 copies/mL of plasma of HIV RNA at 1 year) in non-B subtypes (also known as non-B clades) was about 90% and similar, if not better, than for B subtypes at least for the most common subtypes (C, A, CRF_AF, and D). It was a retrospective cohort study. Despite the relatively high sample size adjustment for other confounding factors like ethnicity, it was difficult because the majority (86%) of persons infected with B subtype were whites as opposed to black Africans for the vast majority of non-B subtypes. One of the recommendations of Kosakowsky et al [2] in an accompanying editorial comment was to ‘‘evaluate the impact of host characteristics of the infected population on antiretroviral therapy efficacy and viral subtype.’’ In this issue of the journal, Scherrer et al [3] addressed this point using patients included in the Swiss HIV cohort study by restricting the analysis exclusively to Caucasians who received HAART from 1996 to 2009 with or without previous exposure to suboptimal therapies (monoor bitherapies with nucleoside analogs exclusively). They claimed a significantly better response among nonB subtypes driven by the improved outcome in subtypes A and CRF02_AG and, at least in part, in the subgroups exposed to unboosted protease inhibitors in the early era of HAART. The study [3] has some limitations. Like several studies in non-B subtypes, it is a retrospective unplanned cohort analysis, the subtypes were derived exclusively from the sequence analysis of the reverse transcriptase (RT) and protease (PR) genes, the definition of virological response was relatively unique for this particular study, not all patients had virological determinations at all evaluated time points, and many patients received some drugs (unboosted protease inhibitors) or combinations (zidovudine plus lamivudine; Combivir ) almost no longer in use for new patients in developed countries. The response toHAARTof nonB subtypes have already been measured in several additional retrospective cohort studies fromEurope [4, 5] and Canada [6] in adults and also in children [7]. Moreover, several recent pivotal registrational studies for new drugs like raltegravir [8], or ritonavir-boosted atazanavir [9] or darunavir [10], comparing with older drugs like efavirenz or ritonavir-boosted lopinavir have addressed the issue of response to HAART in non-B subtypes. All these large studies were randomized and multicenter, including centers from many countries where non-B subtypes are highly prevalent; as a consequence, from 10% to 50% of recruited patients were infected by non-B subtypes. The overall conclusion can be that response rate to combinations, including drugs like efavirenz, ritonavirboosted lopinavir, atazanavir, or darunavir, or to raltegravir, had been almost identical when B subtypes were compared with non-B subtypes. Why has the issue of response to HAART among non-B subtypes or even in HIV-2-infected patients become more important? The HIV pandemic (with a total of 33.3 million people living with HIV, 2.6 million people newly infected, and 1.8 million AIDS deaths in 2009) is in fact a conglomerate of several or many overlapping parallel epidemics. At least 3 independent transmission events from nonhuman primates to humans were the origin of the 3 HIV-1 groups—M (major), O (outliers), and N (new or nonmajor nonoutliers)— and have most likely occurred in the first half of the 20th century [11]. One additional transmission event might have been responsible for the HIV-2 infections. Nine subtypes have then been identified within the M group (A1-4, B, C, D, F1-2, G, H, J, and K), plus 43 circulating recombinant forms (CRFs) defined as the detection of at least 3 strains without epidemiological linkage, plus many unique recombinant forms (URFs) when less than 3 unlinked Received 1 August 2011; accepted 11 August 2011; electronically published 13 October 2011. Correspondence: José M Gatell, MD, PhD, Head, Infectious Diseases and AIDS Unit. Hospital Clinic, Barcelona, Professor of Medicine, University of Barcelona, Co-director. HIVACAT program, Villarroel, 170, 08016 Barcelona, Spain (jmgatell@clinic.ub.es). Clinical Infectious Diseases 2011;53(11):1153–5 The Author 2011. Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please e-mail: journals.permissions@ oup.com. 1058-4838/2011/5311-0018$14.00 DOI: 10.1093/cid/cir686