Professor Steven Wolinsky, Oxford Martin Visiting Fellow at the Institute for Emerging Infections, describes new research into how HIV persists in the body despite being undetectable, and considers new ways to improve the efficacy of existing treatments.
“Imagination is the only weapon in the war against reality.”
- Lewis Carroll, Alice in Wonderland
Antiretroviral drugs routinely impair virus production to levels that are undetectable in the blood within weeks of starting treatment. Once the drugs are stopped, however, HIV rebounds in the blood of treated patients with well-suppressed infection, suggesting that none of the current treatments is capable of eradicating the virus hidden within sanctuary sites in lymphoid tissue.
Intermittent virus production from reactivation of a small fraction of latently infected cells, rather than low levels of continuing viral replication within a cellular sanctuary site, is thought to drive viral rebound. This belief is behind the attempt to awaken the reservoir of latent viruses with the goal of eliminating it from the body. We now know, however, that virus evolution and trafficking between lymph nodes and blood continues in patients with undetectable levels of virus in their bloodstream.
Using a state-of-the-art deep sequencing approach to sample low frequency variants with a high degree of reliability and a novel time-calibrated statistical framework to infer the timing and direction of the key migrations of the virus within hosts, our collaborative group determined that the virus continues to replicate and evolve within a reservoir of cells in lymphoid tissue. A spatial and dynamic model explained why the appearance of drug resistance mutations is not bound to happen.
Earlier studies that did not find evidence for HIV evolution, an indication that the virus continues to replicate, sampled blood only, used low-resolution limiting dilution PCR sequencing, and employed unsuitable measures of virus evolution for the lineages under consideration.
Measures of virus evolution within a host have a critical dependence on the location of the root of the phylogenetic tree. A root that is overly distant from the lineages under consideration within a host, such as a consensus sequence from viruses found in different hosts at different times, is subject to sampling bias that has a significant effect on the time structure of the tree and evolutionary rate estimates.
We reasoned that the findings for blood are not necessarily generalizable to lymphoid tissue where the frequency of infection per cell is mostly higher and amount of antiretroviral drug is significantly lower. We found that the virus can continue to replicate and replenish the viral reservoir despite potent antiretroviral therapy.
The structure of the resulting trees shows that novel mutations arose during the sampling period, with viruses from later time points derived from viruses at earlier time points. This pattern of relationships among viral lineages necessitates not only a degree of replication competence within the viral reservoir, but also movement to new anatomic compartments. This cannot realistically happen through ordered replacement by archival sequences or selective sampling.
Our findings explain why treatment intensification fails to suppress infection and provide a new perspective on how the virus persists in the body. These findings also open up the possibilities for new ways to achieve the cellular concentrations and spatial distribution of antiretroviral drugs needed to fully suppress viral replication and preserve immune function. The modeling and testing of such innovative approaches to clearing the viral reservoir holds promise for revolutionizing HIV treatment.