Estimating the Total Size of the HIV Reservoir

A paper published earlier this month in Nature Medicine by Jake Estes and colleagues addresses the difficult challenge of attempting to define the distribution and total size of the persistent HIV reservoir. The researchers employed a variety of tests to measure viral RNA and DNA -- including new imaging techniques they have developed named RNAscope and DNAscope -- in multiple tissues from untreated or ART-treated macaques (infected with SIVmac251, SIVmac239 or SHIV) as well as lymph node and rectal tissue samples from a cohort of 20 individuals on ART in Kampala, Uganda.

The main findings were that the vast majority (~99.6%) of cells expressing viral RNA -- indicating some degree of virus production -- were found in lymphoid tissues (lymph nodes, gut, spleen, and lung). ART significantly reduced viral RNA levels, but the researchers estimated a total body burden of as many as seven million cells expressing HIV RNA at any given time in tissues, despite viral load suppression to undetectable levels in the blood.

One suggested explanation for the continued viral RNA expression is ongoing replication due to poorer penetration of ART drugs into tissues (the study reports evidence of suboptimal ART levels in tissues of macaques). But the authors' note that the tests for viral RNA were unable to distinguish between ongoing active viral replication and the production of viruses by reactivated latently infected cells. ART would not be expected to block the latter process; rather, it would prevent any viruses that are produced by reactivated latently infected cells from being able to go on and infect other cells.

The total size of the replication-competent HIV reservoir that persists during ART was estimated as approximately 400 million latently infected cells, based on analyses of viral DNA levels in tissues. The researchers suggest this is broadly consistent with another recent estimate, which, when extrapolated to the whole body, gives a result of 240 million cells. The paper's abstract states a more conservative ~100 million (I'm not clear why there's a difference from the estimate provided within the paper). These numbers are larger than had originally been thought, and underscore the challenge of eliminating latent HIV infection.

The question of whether HIV replication commonly persists in tissues despite ART remains contentious, with many of the authors of this study having previously reported that it does, based on analyses of three people who had been receiving ART for six months (this prior work is cited in the current paper).

But these results were recently challenged by the research group of Mary Kearney, who found no evidence of HIV replication in children who had been on ART with suppressed viral loads for seven to nine years (Kearney's work was initially presented at CROI 2017 and has since been published in the Journal of Clinical Investigation). Kearney argued that there was a potential problem with the original claims because the Bayesian methodology for assessing HIV evolution was not appropriate for the task.

A recent paper by David Asmuth and colleagues also presents a contrary perspective, finding that ART concentrations in gut-associated lymphoid tissue (GALT) were adequate in almost all cases studied, with no evidence of residual HIV replication.

Resolving the question of whether the HIV RNA that is detectable despite ART derives from ongoing active viral replication or reactivated latently infected cells (or both) will be important, because it has implications for therapeutic approaches to reducing the reservoir.

Richard Jefferys is project director at Treatment Action Group (TAG).

[Note from This article was originally published by Treatment Action Group on Oct. 30, 2017. We have cross-posted it with their permission.]