In addition to studies covered in prior posts (a possible second pediatric cure case, Sangamo’s gene therapy and limitations of latency-reversing agents), CROI 2014 featured a smorgasbord of presentations related to pathogenesis and cure research. Webcasts of all conference sessions are online, and many posters are now available in PDF format. Brief summaries of some notable studies are appended below, with links to webcasts and posters included where possible. Abstracts for posters that are unavailable as PDFs can be found in the final program and abstract book.
HIV Rebound in the Boston Patients
Timothy Henrich provided details on the two individuals in Boston who experienced a delayed return of HIV replication after a period during which it had been hoped they might be cured (this sobering news was covered on the blog last December). Henrich’s presentation is available via webcast. Although they were monitored frequently after interrupting antiretroviral therapy (ART), viral load reached very high levels in both cases, and this was accompanied by symptoms of acute infection. The likely explanation is that their immune systems, having been newly generated by the stem cell transplants they had received to treat cancer, had not previously encountered HIV. Consistent with this possibility, HIV-specific cellular immune responses were undetectable prior to the viral load rebound but appeared afterward. At a community workshop on cure research immediately prior to CROI, Henrich noted that the size of the HIV reservoir had been reduced at least 3 logs in these individuals and was estimated to be very small at the time of the ART interruption; 290-2900 cells in one case, 40-730 cells in the other. The results suggest the bar is set very high for approaches that aim to cure HIV infection by shrinking the latent reservoir.
Treatment of Hyperacute HIV Infection
A large crowd congregated at a poster presentation (397LB) by Hiroyu Hatano from UCSF describing extremely early ART treatment in an individual who acquired HIV infection just prior to enrolling in a pre-exposure prophylaxis (PrEP) demonstration project. The infection occurred during a short 13-day window between a final screening visit and the day Truvada PrEP was first administered. When viral load results from the sample taken at Truvada initiation became available after seven days and showed a reading of 220 copies, the individual was switched to a conventional ART regimen. The viral load seven days later was 120 copies and there was a single measurement of cell-associated HIV RNA at a level of 4.7 copies per million CD4 T cells ~32 days after infection, but all subsequent tests for HIV RNA, DNA and replication-competent virus have been negative. The results raise the possibility that ART given at such an early stage of infection may have led to an adult equivalent of the Mississippi baby case. However, the experience of the Boston patients offers a cautionary tale about the interpretation of negative virological assay results. A treatment interruption will be considered after 12 months of ART.
Three Cured Mice
Qingsheng Li displayed an interesting poster immediately adjacent to Hatano’s, but it didn't receive quite as much attention. Li’s work attempted to recapitulate the Mississippi baby outcome in the humanized mouse model, by administering ART six hours after the animals were infected with HIV. ART was maintained for two weeks and then interrupted. HIV RNA and DNA remained undetectable after interruption, even when CD8 T cells were depleted, leading the researchers to conclude that a cure had been achieved. The initial experiment (reported in the conference abstract book) involved three ART-treated mice and three controls, but the poster included data on additional groups that received ART at later timepoints after infection (12 hours and 18 hours). In these animals, HIV rebounded after ART interruption. Unfortunately the poster is not yet available on the CROI website (there is a link, but it goes to the abstract book).
HIV Integration, Latency, and CD4 T Cell Survival
At least two published studies (Josefsson et al and Imamich et al) have documented that the homeostatic proliferation of memory CD4 T cells containing integrated HIV provirus can increase the amount of HIV DNA in the body. These studies tracked copies of grossly defective virus genomes that could only multiply as a result of the proliferation of the host CD4 T cell. Homeostatic proliferation is a normal immunological mechanism that sustains memory T cells, and causes them to expand in the setting of T cell deficiency. The role of this process in maintaining the latent HIV reservoir was first described many years ago, based in part on the finding that lower CD4 T cell counts are associated with an increased rate of CD4 T cell proliferation and higher levels of HIV proviral DNA (although these levels remain extremely low compared to the overall size of the memory CD4 T cell pool). Several presentations at CROI documented the phenomenon in a different way, by demonstrating the proliferation of HIV proviruses integrated at precisely the same location in the CD4 T cell genome. Additionally, it was reported by Thor Wagner that HIV proviruses frequently integrate at very similar sites in certain CD4 T cell genes associated with the cell cycle and differentiation, leading to the suggestion that HIV integration into these genes may somehow be promoting the growth and longevity of latently infected CD4 T cells. Wagner highlighted the fact that some of the genes into which HIV appears to preferentially integrate have been identified as having roles in cancers, and this point was echoed in a poster from Stephen Hughes laboratory at the National Cancer Institute. However, some cited examples such as Bach2 are also involved in CD4 T cell differentiation (this is not mentioned in either CROI presentation), so it would be logical that genes such as this would be active in the CD4 T cells HIV infects, and the virus is already known to preferentially integrate into active genes. Additional research is needed to clarify whether there is anything sinister about these integration sites, or if they just reflect the locations HIV is most likely to have access to in the cells it is infecting. Based on the current evidence, it seems possible that these genes could simply be those that are active and available to HIV in CD4 T cells that are differentiating into a long-lived phenotype.
In an article about the research by Jon Cohen in Science, Wagner’s colleague Lisa Frenkel goes so far as to suggest that there may be a link with cancers that occur in HIV-positive people, but to me this is a wildly speculative stretch given that most of these cancers have been shown to be associated with immunodeficiency (in both HIV-positive and HIV-negative people) and inflammation. The article also could be misread as suggesting that long-term HIV treatment causes higher cancer rates, when what the literature reports is that some HIV-positive people on ART still have an elevated risk of some non-AIDS defining cancers (NADCs) compared to HIV-negative people (age is the strongest, most consistent risk factor). A review on the subject from 2012 states: “the higher risk of NADCs is primarily among males, with HIV-infected women having no higher rates of NADCs compared with the overall population.” A sex-biased cancer-promoting effect of HIV integration and latency does not appear very plausible, and T cell cancers remain rare in HIV-positive people. In sum, pending further data, I would advocate extreme skepticism about Frenkel's suggestion that there may be a connection between HIV latency and the development of cancer. Links to CROI presentations on the topic are below.
Slim Success with Gut-Targeting Anti-Activation Strategies
Among the posters at CROI were results from four different clinical trials of therapies that aimed to ameliorate immune activation by targeting the microbial translocation pathway (the leakage of bacterial products from the gut into the systemic circulation). Sevelamer (a treatment for high blood levels of phosphorus that can bind bacterial lipopolysaccharide) and meselamine (a therapy that reduces mucosal inflammation in inflammatory bowel disease) had no effect on markers of microbial translocation or inflammation. The antibiotic rifaximin was associated with very small reductions in some markers of immune activation and inflammation. The probiotic supplement Biola was reported to significantly reduce the levels of the D-dimer by about one third over eight weeks in a small study (12 probiotic recipients, seven placebo recipients and six controls) and there were trends toward reduced CRP and IL-6. Markers of microbial translocation did not change. These results await confirmation in a larger trial with a longer duration of follow up.
HIV-Specific CD4 T Cell Responses in the VISCONTI Cohort
The VISCONTI cohort comprises individuals treated very early after acquiring HIV infection (for an average of around three years) who have maintained undetectable or extremely low HIV viral loads after interrupting ART. At the last published accounting, there were 14 members, although anecdotally it has been mentioned that there may now be as many as 20. The cohort appears to lack the robust HIV-specific CD8 T cell responses observed in most elite controllers, but until CROI no data had been presented on their HIV-specific CD4 T cell responses. Assia Samri reported in a poster that HIV p24-specific CD4 T cells are detectable at “relatively high frequencies” in the cohort and have a more polyfunctional profile—but are not more abundant—compared to individuals on ART. However, the data are cross-sectional and cannot answer the question of whether the presence of these responses is a cause or effect of the persistently low viral load in the absence of ART.
Hospitalization Among Elite Controllers
Elite controllers are sometimes cited as a model of a “functional cure” of HIV infection, because they naturally maintain extremely low viral loads in the absence of ART. However, in recent years it has become apparent that most elite controllers have elevated levels of immune activation and inflammation compared to comparable HIV-negative people, and are not entirely free from the risk of disease progression. Trevor Crowell presented a poster looking at hospitalization rates among 149 elite controllers compared to 4,704 HIV-positive individuals with suppressed viral loads on ART. Elite control was associated with a significantly higher rate of hospitalization (approximately two-thirds greater), primarily for cardiovascular and pulmonary disease, compared to the cohort on ART. The data are consistent with research indicating that ART may be beneficial for at least some elite controllers, and also has implications for HIV cure research because it argues that the achievement of virological control in the absence of ART does not necessarily equate to ideal health.
Anti-PD-L1 Antibody in Macaques
Targeting the PD-1 receptor pathway has been proposed as a method for both reversing HIV latency and reinvigorating exhausted HIV-specific T cell responses. The AIDS Clinical Trials Group is on the verge of launching a clinical trial of an antibody against the PD1 ligand, PD-L1. At CROI, Stephen Mason described results of a preclinical study of the anti-PD-L1 antibody in ART-treated SIV-infected macaques. Administration was found to be safe and associated with a significant but short-term reduction in viral rebound after an ART interruption.
EraMune 02 Results
Two of the first clinical trials launched to specifically attempt therapeutic reduction of HIV reservoirs were EraMune 01 and 02, sponsored by the French Objectif Recherche Vaccins SIDA collaboration. Both investigated intensified ART, combined with IL-7 (EraMune 01) or a DNA/Ad5 prime-boost therapeutic HIV vaccine (EraMune 02). Results from study 01 were presented at a prior CROI, showing no significant HIV reservoir reduction and some evidence of an IL-7-mediated increase in HIV DNA resulting from the proliferation of latently infected cells. Results from study 02 debuted as a poster this year, also showing no significant reduction in levels of HIV DNA. However, the vaccine did successfully induce significant T cell responses against HIV Gag, Pol and Env antigens in individuals on ART.
Evidence Prednisolone Can Increase CD4 T Cell Counts and Slow Progression in the Absence of ART
In the early 1990s, there was some excitement about results from an uncontrolled, open-label study of the immune suppressant drug prednisolone in people with HIV infection. The investigators reported that administration of the drug led to sizable increases in CD4 T cell counts that had not been seen with the antiretroviral drugs that were being given as monotherapy or dual therapies at the time. The advent of three-drug ART consigned this prednisolone study to the haze of history, but at CROI there was a flashback in the form of results from a randomized clinical trial, which suggest that the prior data were not a fluke. The trial was conducted in Tanzania during a period when ART was only indicated for people with low CD4 T cell counts. Treating individuals for whom ART was not yet indicated with prednisolone was associated with a significant increase in CD4 T cell counts and a reduced risk of clinical progression. The results underscore the causative role of immune activation in the progression of HIV infection, since prednisolone acts via this pathway without directly inhibiting the virus. The researchers suggest the drug may be able to inhibit immune activation and improve immune reconstitution in immunological non-responders on ART, and perhaps could have a role as an early therapy for asymptomatic HIV infection in resource-limited settings (the latter suggestion is likely to be controversial given the understandable emphasis on trying to improve global access to ART, the efficacy of which is proven).