Perspectives on Post-Treatment Control of HIV

Several recent papers offer perspectives on the possibility of achieving post-treatment control of HIV replication by starting antiretroviral therapy (ART) during acute infection. Interest in this topic has been sharpened by reports about the VISCONTI cohort, a group of individuals in France who all started ART soon after HIV acquisition, maintained treatment for several years, then interrupted and are now controlling viral load to very low or undetectable levels. At the time of the last published update, the VISCONTI cohort included 20 individuals (identified from various trials) who have been off ART for an average of nine years.

In a paper titled "How often does treatment of primary HIV lead to post-treatment control?" in Antiviral Therapy, Janine Maenza and colleagues identify 22 individuals who started ART during acute infection (within 12 weeks of acquisition), continued for at least 11 months, then interrupted. Out of this group, only one individual with a low viral load of 2,248 copies at the time of ART initiation maintained viral load below 500 copies for three years after interruption. The viral load then rose to levels similar to baseline for the following three years. The researchers state: "given our single controller's relatively low HIV-1 RNA levels during primary infection ... we question whether he might have become a natural controller, regardless of initiation of early therapy."

A study led by Antoine Chéret, published in the Journal of Antimicrobial Chemotherapy, primarily set out to evaluate the size of the HIV reservoir in different CD4 T cell subsets in 11 participants from the OPTIPRIM clinical trial, who began ART an average of 36 days after becoming infected and maintained treatment for at least two years. But the authors note that two individuals in this study displayed control of viral load after ART interruption; in one case the level was consistently below 20 copies for 24 months, in the other there was a rebound followed by control to low but detectable levels (530 copies after 18 months off treatment). The case involving stricter containment of HIV replication showed traits similar to VISCONTI cohort members: high viral load and low CD4 T cell count prior to ART initiation, possession of the HLA B*35 allele that has been associated with faster disease progression in some studies, and low HIV DNA levels, particularly in naïve and central memory CD4 T cells. Also mirroring the VISCONTI cohort, HIV-specific CD8 T cell responses appeared weak and were not able to suppress viral replication in a laboratory assay.

A third paper by Jessica Conway and Alan Perelson attempts to mathematically model the parameters that might contribute to post-treatment control, and suggests that both the HIV reservoir size and the strength of the HIV-specific cytotoxic T-lymphocyte (CTL) response may be important. The authors explain that the model is relatively simple and that other types of immune responses weren't considered but could play a role, such as natural killer (NK) cells, CD4 T cells and antibodies. A forthcoming AIDS Clinical Trials Group (ATCG) study is mentioned that plans to evaluate whether the size of the HIV reservoir correlates with time to viral load rebound after ART interruption, as the model predicts (ACTG A5345).

An overarching and extremely important question relating to post-treatment control is not discussed in any of the papers: namely, whether the degree of HIV suppression in post-treatment controllers is equivalent to that achieved by ART when it comes to clinical outcomes. Conway and Perelson write that post-treatment control "corresponds to functional cure, sustained remission that does not require therapy," but it isn't yet known if this is true. It has become apparent in recent years that elite controllers, who show similar suppression of viral load in the absence of treatment, can ultimately lose CD4 T cells and progress to AIDS (perhaps best documented in an analysis of a large cohort of HIV-positive individuals that was published in PLoS One last year).

The majority of elite controllers also show elevated levels of immune activation, inflammatory biomarkers, and lymph node fibrosis compared to similar HIV-negative individuals. A study presented at CROI last year, and published in January in the Journal of Infectious Diseases, found that elite controllers have higher hospitalization rates compared to HIV-positive individuals on ART, primarily due to cardiovascular disease (although there are some concerns about the influence of confounding variables in this analysis, as described in an excellent AIDSMap article on the paper by Gus Cairns). Furthermore, there is evidence that ART may benefit at least some elite controllers, reviewed recently in the new open access Journal of Virus Eradication by Trevor Crowell and Hiroyu Hatano.

The fact that some post-treatment controllers appear to have weaker HIV-specific T cell responses than elite controllers might suggest that immune activation and inflammation levels are likely to be lower, but there is a lack of data addressing this issue. The main published paper on the VISCONTI cohort only compares immune activation levels to individuals on ART, reporting that they are similar, and there is no comparison to HIV-negative individuals or elite controllers (there is a comparison to a group defined as "HIV controllers" with viral loads below 400 copies, as opposed to the threshold of 50 copies generally used to define elite control). To my knowledge, no data on inflammatory biomarker levels has been presented or published, even though this was one of the questions posed to the investigator Asier Sáez-Cirión when he first described the VISCONTI cohort at a conference several years ago.

The point of raising this concern isn't to be pessimistic about post-treatment control, but rather to highlight that there needs to be greater clarity about what the term means in relation to clinical outcomes. In particular, post-treatment control should not be considered synonymous with "remission" or "functional cure;" these designations should refer to a state of health that is at least comparable to, if not superior to, that obtained by ART. At the current time, how the long-term health of post-treatment controllers compares to individuals whose viral loads are suppressed by ART is not known.

The VISCONTI researchers do attempt to highlight this uncertainty by stating that cohort participants are examples of "long-term virological remission" (emphasis added); however, as the Conway and Perelson paper shows, this distinction is not always appreciated by other scientists and is certainly lost in much of the media coverage (the BBC headline on the VISCONTI cohort was: "Early HIV drugs 'functionally cure about one in 10'").

There are reasons for optimism: anecdotally, the health of VISCONTI cohort members has been described to be good, and it is possible that they will not face the long-term risks seen in some elite controllers. Additionally, reports indicate there is a subset of elite controllers with extremely low HIV DNA levels from whom infectious HIV cannot be recovered, and gene expression analyses suggest these individuals have similar profiles to HIV-negative individuals (see abstract below from the recent Keystone Symposia on HIV cure research). While further studies are needed, this subset of elite controllers might turn out to represent a model of strict HIV control that lacks inflammatory and health consequences, thereby offering clues as to the degree of post-treatment control that equates to clinical remission.

_Richard Jefferys is the coordinator of the Michael Palm HIV Basic Science, Vaccines & Prevention Project Weblog at the Treatment Action Group (TAG). The original blog post may be <a href="http: 2015="""" tags_basic_science_vaccin="" 05="" perspectives-on-post-treatment-control-of-hiv.html"="" target="blank">viewed here.</a href="http:>