Durable HIV Treatment Benefit Despite Low-Level Viremia

Abstract

LEFT UNTREATED, HUMAN IMMUNODEFICIENCY VIRUS (HIV) replicates at a rapid rate, with the eventual production of billions of new virus particles per day. Given the propensity of HIV to mutate, the possibility exists that each newly produced virus contains at least 1 new mutation. Thus, from a darwinian perspective, ongoing viral replication in the presence of therapy should result in the rapid selection of drug resistance mutations and subsequent virologic rebound. These basic principles have provided the theoretical context for the “hit hard” therapeutic approach to HIV disease. According to current treatment guidelines, complete viral suppression should be the goal of therapy. Once therapy is initiated, plasma viremia, as measured by the concentration of viral RNA in plasma, should decrease to below the level of detection using the most sensitive assay available. Persistent viremia suggests ongoing viral replication and treatment failure. The findings from 2 articles in this issue of THE JOURNAL raise questions about this conceptual framework, and based on data from the 2 studies, it may be argued that “complete” viral suppression may not be a prerequisite for durable treatment benefit. Hermankova and colleagues present studies involving patients who had achieved and maintained an undetectable plasma HIV RNA level ( 50 copies/mL). Using sensitive techniques to amplify plasma viral RNA, virus was identified in plasma in 10 of 20 patients. Genotypic analysis of the isolated viral RNA revealed no evidence of new drugassociated mutations clearly related to the current regimen. The findings indicate that viral replication (defined by the release of virus into plasma) can be identified in patients apparently responding to combination therapy, but that this level of replication appears to be insufficient to select for drug resistance. Havlir and colleagues addressed a similar question in a completely different manner. Patients enrolled in a longterm treatment study were identified retrospectively and stratified into 2 groups: those whose plasma HIV RNA levels remained consistently below the level of detection ( 50 copies/mL) and those whose plasma HIV RNA levels were transiently detectable ( 50 copies/mL of RNA with a subsequent measured level of 50 copies/mL) and who were categorized as having intermittent viremia. Despite evidence of higher levels of viral replication in the latter group (although not uniformly statistically significant), both groups had similar rates of virologic failure. The occurrence of 2 consecutive HIV RNA levels greater than 200 copies/mL was considered virologic failure. Before discussing the implications of these data, it is critical to assess whether the presence of HIV RNA in plasma signifies ongoing viral replication. Hermankova and colleagues suggest low-level viremia may simply reflect the release of archived virus from long-lived cellular reservoirs into plasma and not necessarily continuous rounds of productive infection. The hypothesis that viral replication is ongoing in patients with low-level viremia is supported by several independent lines of evidence. First, genetic evolution within viral envelope sequences has been observed in a small number of treated patients with transient viremia. Second, unintegrated viral DNA and unspliced viral messenger RNA, both markers of recent cellular infection, are commonly detected in patients even after several years of effective therapy. Third, drug-specific mutations may emerge in cellular reservoirs of patients with undetectable or lowlevel plasma viremia. Fourth, patients with low-level viremia often have increasing amounts of replicationcompetent virus in cellular reservoirs. Thus, when viewed as a whole, it appears that viral replication persists in patients with low-level viremia, and that complete viral suppression is rarely achieved with current therapies. If antiretroviral therapy is only partially effective at suppressing viral replication, why were drug resistance muta-