icon-    folder.gif   Conference Reports for NATAP  
  20th Conference on Retroviruses and
Opportunistic Infections
Atlanta, GA March 3 - 6, 2013
Back grey_arrow_rt.gif
Revealing the Reservoir, and A Second Cure?
  Ronald Swanstrom PhD, UNC Chapel Hill
Session 10 "Is There Hope for HIV Eradication?" at CROI 2013 in Atlanta had a packed house and was the focus of national news
CROI: CROI Report 2013: Observations on the "Cure Agenda" - By David Margolis MD, UNC Chapel Hill and the Collaboratory of AIDS Researchers for Eradication (CARE) (03/18/13)
(http://webcasts.retroconference.org/console/player/19411?mediaType=podiumVideo). The biggest buzz came from the late breaker presentation by Deborah Persaud (Johns Hopkins Univ.) and her colleagues describing what looks to be the second person cured of HIV-1 (following the "Berlin Patient" Tim Brown). In this case an infant was born to an HIV-infected mother who had not received any antiretrovirals during the pregnancy. Within 48 hours the baby was identified as being HIV-1 RNA and DNA positive. The infant was started on antiviral therapy at 31 hours after birth. The viral load dropped in the baby over the next 4 weeks to become undetectable. At about 18 months of age there were adherence issues but when the infant was off therapy there was no rebound of virus. All standard virologic tests for viral nucleic acid (plasma RNA, cellular DNA) have been negative, with only occasional signals near the limit of detection with highly sensitive assays. The infant has not seroconverted. The interpretation is that some combination of the length of time baby was infected in utero (perhaps short), the timing of the initiation of therapy, and the developmental stage of the baby somehow worked together to allow the therapy to clear the virus without creating a residual latent reservoir. There are two important points to make about this study. First, further studies are ongoing to strengthen what are largely negative results (the hardest to prove), and if the interpretation is correct then there are (sadly) many cases of mother-to-child transmission where this protocol can be tested, with studies being planned by the NIH-funded IMPAACT network. If this interpretation is correct and can be replicated it will be an important tool for reducing the impact of mother-to-child transmission of HIV-1 and it will represent a setting where we can ask questions about how the latent reservoir is (or isn't) formed.
Much of the session covered issues related to the viral reservoir in people on successful antiretroviral therapy. In the first talk of the session Hiroyu Hatano (UCSF) examined the effect of raltegravir intensification. There is continuing discussion about whether there is ongoing viral replication in the face of potent therapy. It has been previously reported that the addition of raltegravir (an integrase inhibitor) results in an increase in 2-LTR circles, a bypoduct of viral DNA when it does not integrate, but intensification studies have also failed to reduce the low level viremia that is present in all subjects on therapy. The current study included randomization to raltegravir or placebo, and subjects were followed for 24 weeks. There were several notable observations: i) consistent with previous reports there was no decrease in the level of low level viremia with the addition of raltegravir; ii) also consistent with previous observations, there was an increase in the level of 2-LTR circles; and iii) intensification resulted in a decrease of D-dimer, a marker that is correlated with inflammation. Two other features of this study are significant, the appearance of 2-LTR circles were more likely to appear in subjects taking a PI regimen, and the appearance of 2-LTR circles was transient. Putting all of this together, it appears that there is low level replication ongoing, especially in the presence of PI inhibitors, and this can be suppressed with the addition of raltegravir. This replication is distinct from the low level viremia that is detected in the blood, which is insensitive to intensification and must have another, non-replicating source (stochastic release from latently infected cells?), but it must represent a source of persistent replication because the addition of raltegravir results in only a transient increase of 2-LTR circles, i.e. it cannot represent infection from the low level viremia pool which is always present. Finally, the decrease in D-dimer indicates that there is a biological response by the host to this replication. While the relationship between this replication and the latent reservoir is unknown, during the questions the point was made that this represents a very low level of replication and the number of infected cells that transition into the latent reservoir is likely to be low. By this reasoning the effect of addition of raltegravir on the latent reservoir would be minimal.
In the second presentation Ya-Chi Ho (Johns Hopkins Univ.) described a thorough analysis of the types of viral DNA present in resting T cells isolated from the blood of subjects on successful therapy. As expected, a large majority of this DNA was defective, with lethal levels of APOBEC3G/F hypermutations, or insertions or deletions that would be incompatible with the production of replication competent virus. However, there was a surprising 12% of the DNA that appeared to be intact and potentially able to produce virus if expressed. This level of intact viral DNA is 40-50 fold higher than the number of resting T cells that can be induced to express infectious virus after activation in cell culture. The conclusion is that the size of the latent reservoir may in fact be much larger than previously measured by virus outgrowth assays if in fact all of these intact viral genomes can be induced to express virus in vivo. Conversely, this study shows the limitation of using total DNA levels to assess changes in the latent reservoir as most of this DNA is defective and treatments to manipulate the latent reservoir may not have an effect on cells that harbor defective DNA, which is the large majority of cells that have viral DNA. Also, induction of expression of viral RNA is likely to be over-represented by induction of expression of defective viral DNA.
The next three talks examined features of cells that contribute to the latent reservoir. Maria Buzon (Ragon Institute) showed that CD4+ T memory stem cells harbor viral DNA. This DNA declines much more slowly than viral DNA in other CD4+ T cell classes, and replication competent virus can be induced from these cells when isolated from a subject on suppressive therapy. The implication is that this cell type can be infected with virus, and if the virus winds up in a latent state the self-renewal properties of this cell type will allow the viral genome to persist due to cellular replication. In the next talk, Fabio Romerio (Institute of Human Virology) identified high levels of CD2 as a surface marker for the subpopulation of cells that contain higher levels of the latent reservoir. CD2 is an adhesion molecule on T cells and natural killer cells. This observation came out of an in vitro model for latency where cells were coaxed into a resting state after infection (with the infected cells still containing detectable p24). In this model the p24+ cells had higher levels of the CD2 on the surface of the cell. This observation led to the sorting of CD4+ resting T cells from subjects on therapy into CD2high and CD2low populations, the CD2high population had higher levels of viral DNA (3-10 fold) and could be induced to express higher levels of viral RNA with activation. The CD2high phenotype may be associated with a CD4+ T cell subset that is more quiescence which may promote viral latency. The third talk on the reservoir was given by Natalia Soriano-Sarabia (UNC Chapel Hill) looking at the latent reservoir in transitional memory CD4+ T cells and gamma/delta T cells. The latter cells have a novel type of T cell receptor but generally lack CD4. It was shown that isolated cells could be infected and this infection was inhibited by an antibody to CD4, suggesting some level of CD4-dependent infection. Consistent with this inference, gamma/delta T cells that had been activated in vitro had a subpopulation that expressed CD4. When resting gamma/delta T cells were isolated from subjects on successful therapy, it was possible to induce virus outgrowth in culture, with a higher reservoir in subjects in acute infection compared to chronic infection. In comparison of central memory T cells and transitional memory T cells it was found that the former was a more robust source of inducible virus after activation in culture.
The final talk in the session about the nature of the reservoir was by Jintanat Anaworanich (the SEARCH Study Group in Thailand) who described work with a large cohort of individuals identified very early in infection and initiated on therapy. At the time of diagnosis the subjects in Fiebig stage I had total viral DNA copy numbers about 30-fold lower than those identified in Fiebig stage III, and integrated DNA was below the level of detection in about 90% of the Fiebig I subjects but only 50% of the Fiebig III subjects. Those results were obtained using cells in PBMCs, and a similar result was seen in biopsy material from the sigmoid colon. Both the Fiebig I and Fiebig III subjects had PBMC viral DNA levels at the level of elite controls by week 36 after the initiation of therapy, with the Fiebig I subjects starting at that level and the Fiebig III subjects dropping to that level. The distribution of viral DNA was greater in transitional and effector memory T cells compared to central memory T cells, again more similar to elite controllers. These results support the idea that eradication efforts may be easier in subjects who start therapy very early after infection.
The final two talks of the session dealt with attempts to manipulate the reservoir. Gregory Del Prete (SAIC Frederick) described the development of a Rhesus macaque model where SIV-infected animals can be put on suppressive therapy. This represents an important model where agents that are designed to manipulate the size of the reservoir can be tested prior to studies in humans. There will undoubtedly be discussions about how well this model represents the latent viral reservoir in humans, an area that deserved much attention. In an initial examination of this model, two different histone deactylase inhibitors (HDAC inhibitors SAHA or romidepsin) were tested ex vivo with cells from animals on therapy. Both compounds worked in inducing viral RNA production in the culture assay, as can be done with cells from humans on therapy.
The final talk of the session was a second late breaker presented by Sharon Lewin (Monash Univ., Melbourne). This study examined the effects of multiple doses of Vorinostat, a histone deacetylase inhibitor that is being explored as an agent to induce expression of latent proviruses. Single dosing with vorinostat has been shown to induce expression of viral RNA in resting memory CD4+ T cells in the blood in subjects on antiviral therapy. In this study of 20 subjects, Vorinostat was given daily for 14 days. There was variability in the pattern of response, but 18 of 20 subjects had a detectable increase - in cell-associated increase in HIV-1 RNA in at least one timepoint. There was no change in the levels of viral DNA. These studies set the stage for more intensive efforts to induce expression of the latent reservoir.