The 2016 Conference on Retroviruses and Opportunistic Infections (CROI) took place in Boston from February 22-25. CROI deserves kudos for pioneering comprehensive webcasting, and all sessions are available for viewing online. Results from several significant cure-related clinical trials were debuted during the meeting (links to the webcasts are in parentheses):
Ole Søgaard from Aarhus University in Denmark presented findings from a trial that combined the HDAC inhibitor romidepsin, a latency-reversing agent, with the therapeutic vaccine Vacc-4x (webcast). The rationale for this combined “kick & kill” approach is that romidepsin can cause latently infected cells to produce HIV proteins, potentially allowing these cells to be recognized and killed by HIV-specific T cell responses that have been induced or boosted by the vaccine. A series of immunizations with Vacc-4x and GM-CSF adjuvant were given first, followed by three infusions of romidepsin. Søgaard reported that romidepsin administration led to significant increases in HIV RNA, consistent with a latency-reversing effect, after which there was a significant decline in levels of total HIV DNA averaging 39.7%, but only a slight and non-significant drop in levels of integrated HIV DNA (these are two surrogate measures of the size of the HIV reservoir). Virus outgrowth was quantified in six of 17 participants who showed detectable levels at baseline, and all six showed a significant decline of around 38%. However, despite this evidence of some diminution in the size of the HIV reservoir, no significant delay in HIV viral rebound was observed when ART was temporarily interrupted in the final phase of the trial. Søgaard concluded that the data support the idea of combining latency-reversing agents with therapeutic vaccines, but improvements are needed to enhance the magnitude of the effect.
Joe Eron from the University of North Carolina discussed the ever first clinical trial of an antibody targeting the PD-1 pathway in people with HIV (webcast). PD-1 is a molecule that can become persistently upregulated on T cells that have become exhausted and dysfunctional, and antibodies that block the interaction between PD-1 and the ligands it binds to (PD-L1 and PD-L2) have been shown to restore T cell function. Two antibodies against PD-1 have been FDA-approved for the treatment of cancers due to their ability to enhance cancer-specific immunity and promote clinically significant tumor shrinkage. Additionally, CD4 cells latently infected with HIV commonly express PD-1, and antibodies against PD-1 have been shown to reverse viral latency in laboratory experiments.
The trial described by Eron was conducted by the ACTG and involved an antibody owned by Bristol-Myers Squibb that targets PD-L1. The original intent was to study single infusions of various doses in people on suppressive ART, however only the lowest dose (0.3mg per kg) was administered due to an unexpected concern about the potential for retinal toxicity that emerged from animal experiments. No evidence of similar toxicity was observed in the six individuals who received the anti-PD-L1 antibody. However, one person developed autoimmune pituitary insufficiency nine months after the infusion, a serious concern because autoimmunity is a known risk associated with targeting the PD1 pathway.
Of the six anti-PD-L1 antibody recipients, two showed distinct evidence of increased Gag-specific CD8 T cell responses (measured both by interferon gamma production and expression of CD107a, a marker of cytotoxicity) but the overall average change compared to a control group of two placebo recipients did not reach statistical significance. There was also no significant change in HIV RNA levels measured by a single-copy assay, however one individual did show a 10-fold drop in cell-associated HIV RNA and Eron noted that this was the person who experienced the greatest increase in Gag-specific CD8 T cell responses. In the question & answer period after the presentation, Eron also mentioned that this individual had the lowest CD4 T cell count and highest level of PD-1 expression on T cells at baseline (consistent with prior reports that PD-1 expression progressively increases during disease progression).
The anti-PD-L1 antibody is not going to be studied further, but the anti-PD-1 antibody pembrolizumab (which is FDA-approved as a cancer therapy) is being evaluated in people with HIV and refractory cancers in a new clinical trial. The safety concern relating to autoimmunity makes it unclear if it will ever be possible to target the PD-1 pathway in people with HIV who do not have concomitant cancers - one possibility might be to try and restrict the activity of the anti-PD1 antibody to just HIV-specific T cells, if there is any biologically feasible way of doing so.
Katherine Bar from the University of Pennsylvania described results from a trial involving three infusions of the broadly neutralizing antibody (bNAb) VRC01, given before and after an interruption of ART in order to assess if viral load rebound would be delayed (webcast). The antibody was safe and well tolerated but did not prevent viral load rebound. There was evidence of a slight delay compared to historical controls, with more VRC01 recipients maintaining viral load suppression four weeks after interrupting ART, but the difference had disappeared after eight weeks. HIV samples from some participants displayed resistance to VRC01. Bar highlighted the need to better understand the relationship between HIV neutralization measured in vitro and antibody potency in vivo, and noted that combinations of different bNAbs will likely be required to improve results. Another somewhat similar trial conducted by Tae-Wook Chun at the National Institute of Allergy and Infectious Diseases was presented at CROI as a poster, and reported broadly consistent findings.
After Bar’s talk, Michel Nussenzweig from Rockefeller University commented that antibodies more potent than VRC01 may perform better, citing unpublished data from a trial of the bNAb 3BNC117 that is being led by his colleague Marina Caskey. In that trial, Nussenzweig said, viral load rebound was delayed by an average of 6.5 weeks after an ART interruption, with 30% of participants maintaining suppression for over nine weeks. 3BNC117 is one of several more potent bNAbs discovered after VRC01, so this offers some hope that superior results are achievable, particularly with combinations. For cure research, the ultimate aim is to test whether these bNAbs can help promote clearance of HIV-infected cells via mechanisms such as antibody-mediated cellular cytotoxicity (ADCC).
In the pre-clinical realm, Gilead caused a splash with data from a new study of their TLR7 agonist GS9620 in macaques. The results were presented by James Whitney (webcast), who reported at last year’s CROI that a TLR7 agonist appeared to induce virus production by latently infected cells in SIV-infected macaques on ART. The new study tested lower doses of two TLR7 agonists, GS-986 and GS9620 (the latter compound is already in clinical development for hepatitis B); the aim of using lower doses was to minimize induction of alpha interferon and associated toxicities. Evidence of latency reversal was observed in the form of SIV RNA increases after dosing, and two of nine macaques have maintained undetectable viral loads for 3-4 months after an ART interruption (no delay in viral load rebound was seen in the study presented last year). Follow up of these animals is ongoing. Whitney stated that GS9620 is now entering a phase Ib trial in HIV-positive people on ART, but Gilead have not registered the trial in clinicaltrials.gov so information on enrollment criteria and locations is not available.
Morgane Gossez from the University of Oxford reported on an analysis of the SPARTAC trial comparing the frequency of post-treatment control of viral load in 22 participants in Africa and 44 in the UK (webcast). The SPARTAC trial design was described on the blog recently. Five of the African individuals have maintained viral load below 400 copies for over 3.5 years of follow-up, whereas all of those from the UK experienced viral load rebounds. Further review of records indicated that two of the five post-treatment controllers had undetectable viral loads at the time of ART initiation, suggesting they may have been elite controllers, but that was not the case for the remaining three. Additional studies are being conducted to look for factors associated with this outcome. Gossez noted that the biomarkers previously reported to be associated with delayed viral load rebound in SPARTAC did not show significance in the subset of African participants, however, in response to a question, Gossez acknowledged that this may have been due to the small sample size.
The potential role of gene therapy in HIV cure research was addressed at CROI in a plenary talk by Paula Cannon from the University of Southern California (webcast). Cannon reviewed the various technologies that are now available to manipulate both host and HIV genes, and cited evidence from ongoing trials of Sangamo's gene therapy that positive effects may be achievable in people. Noting that gene therapy is sometimes viewed as too impractical to be pursued, Cannon made a strong case that it should be viewed as an important element of the cure research effort.
Lastly, in a poster presentation with echoes of the first report on Timothy Brown at CROI in 2008, a group of German researchers described the case of an HIV-positive individual who received a stem cell transplant from a CCR5delta32 homozygote donor as part of a series of treatments for cancer (acute myeloid leukemia). The individual experienced two relapses but ultimately the cancer went into remission in 2013. All tests for HIV DNA have since been negative in peripheral blood, rectal tissue and bone marrow, and HIV-specific antibody responses measured by Western blot are waning. Importantly, the individual remains on ART and researchers plan to search additional tissues for evidence of HIV before considering interrupting treatment. To my knowledge, this is only the second report of a successful stem cell transplant from a CCR5delta32 homozygote donor in a person with HIV (the first being Timothy Brown) - although it has been tried in other cases, these individuals died either due to the underlying cancer or complications from the procedure. The researchers are hoping that, like Brown, this individual may be cured of HIV, but it remains to be seen whether this hope will be borne out.