icon-    folder.gif   Conference Reports for NATAP  
  20th International AIDS Conference
July 20-25, 2014
Melbourne, Australia
Back grey_arrow_rt.gif
IAS pre-meeting "Towards a Cure" Symposium
  July 19-20, Melbourne 2014
David Margolis MD, UNC Chapel Hill
[Here is a link to the IAS HIV Cure Symposium website homepage where you will find all the "Towards a Cure" Symposium background, objectives, program, and all the presentations: abstracts, slides, and audio of each individual presentation:
Dr. Francoise Barre-Sinoussi opened the workshop to begin the end of her tenure as IAS president. She has galvanized an international effort "TOWARDS AN HIV CURE" as her signature initiative. She opened the symposium as a platform to present state-of-the-art basic science and clinical research on HIV cure research, and eulogized Joep Lange, murdered in the Maylasian Airlines attack just days before, recalling that Joep believed that a cure for HIV was possible, and that he was one of the first proponents for integrating social sciences and HIV cure research. "He would have encouraged us to go on," she said.
Jeff Lifson of the NCI Frederick laboratories gave the first keynote, discussing the role of therapeutic vaccination in HIV cure strategies. Lifson discussed in the first few years of reawakening in this field of study, clearance of persistent HIV has taken a center-stage role recently. Given the recent viral rebound in several settings where curative effect had been hoped for, Lifson remarked that attempts at eradication or long-term control would have to "deal with the very last virus" that was remaining. He held that vaccines would be a key part of this effort. It is hard to argue with him.
Lifson reported experiments carried out in his laboratory that had attempted to model the effect of the histone deacetylase inhibitor SAHA or vorinostat (VOR) that has been seen in human pilot trials (Archin et al. Nature 2012) in the SIV primate model. Macaques were infected with SIV, treated with dolutegravir/tenofovir/FTC, and then subsequently dosed with VOR at 54 mg/kg/day x 21 days, and then VOR readministered over 4 more cycles, with "rest" intervals, for a total of 84 doses. Although increases in histone acetylation were seen at various timepoints, a biomarker indicative of a SAHA exposure, and SAHA was safe, effects on persistent infection were difficult to measure. Some changes were seen in the SIV transcriptional ratio (viral RNA:viral DNA) but the results were complex, with an apparent decreased response with repeat dosing on this daily schedule. No viral clearance could be measured. Overall the results were reminiscent of those reported in man in a multidose VOR study by Archin (JID 2014) earlier this year.
In a later, related presentation, Brandon Keele (NCI) presented a new technique to genetically, persistently mark SIV DNA sequences, called ultra-deep barcoding. SIVmac239 was genetically engineered to include a molecular cassette between the vpx and vpr genes. The cassette allows for the insertion of 10 randomly generated oligonucleotides producing up to 1 million unique plasmid clones. Infectious virus is generated by transfection, producing a viral stock that is genetically identical except for the insertion. Barcoded viruses were examined in vivo in rhesus macaque for replication competence and retention of barcode over time. Real-time single genome amplification and next generation sequencing was used to quantify the number of unique viral genomes. Keele found no evidence of reduced replication competency, nor was the barcode lost or altered during culture or in vivo. Such techniques may aid future NHP studies to track the establishment and persistence of persistent viral reservoirs and the sources contributing to recrudescent viremia following release of suppressive antiretroviral therapy.
Lifson then highlighted recent studies done with Louis Picker (Oregon) in which a novel CMV-based HIV vaccine sought to overcome the limitations of conventional vaccine approaches that result in transient peak immune effects. By exploiting the evolutionarily acquired immune wisdom of CMV, Picker and coworkers hoped to induce extremely high frequency of durable CD4+ and CD8+ T cell effector memory responses. As reported in Nature and Science last year, some vaccinated monkey suppressed SIV viremia, and eventually appeared to clear SIV infection over time, at least as measured by the best assays available. Ultimately, this approach might be translated to humans, and could be used together with other cure strategies that are designed to eradicate persistent, latent HIV.
Dr. Lifson also outlined the waterfront of currently considered approaches to HIV cure: ART intensification, latency disrupting agents, immune modulators (eg. cytokines) and immunotherapies (eg. mAbs, adoptive cell therapy, therapeutic vaccination), and combinations of these approaches. He lent his support to the role of NHP models in HIV cure research for preliminary proof-of-concept, and highlighted the current limitations of the model system (adaptations of ART for NHP; drug delivery (dosage, route, compliance); bioavailability).
Richard Koup of the NIH VRC then discussed new research on killing or clearance of HIV-infected cells in lymph node germinal centers. This location is of concern as it contains a rich population of CD4 memory cells, and may represent a substantial proportion of the latent reservoir, but is an area that is off-limits for CD8 cells. Therefore targeted killing of infected cells at this site may be difficult or will have to be undertaken by other effector mechanisms. Koup suggested the use of engineered Bi-Specific antibodies. He linked VRC07, one of the VRC's broadly neutralizing Abs that recognize a broad population of HIV envelopes, to an Ab for CD3, the general T cell receptor. His findings suggested that if redirected by such Abs, CD8s could access germinal centers to clear HIV.
Melanie Ott (UCSF Gladstone) gave an overview of regulatory mechanisms which governed HIV gene expression, in particular those that play a role in viral latency. She focused particularly on modifications of proteins that regulate their activity in gene expression, be it modifications of histones (e.g. acetylation or methylation) that regulate chromatin structure and thereby nearby gene activity, or modifications of transcription factors themselves that drive gene expression. She noted that even some HIV RNAs themselves might play a regulatory role, although this has not yet been widely accepted. She reviewed modifications of the viral activator protein Tat (acetylation, methylation, ubiquitination) that can up or down regulate Tat binding and function. She described the activity of Bromodomain (BRD) factors, of recent interest due to the emergence of potent BRD inhibitor drugs in oncology, and their ability to bind or release transcription factor complexes, and inhibit or induce HIV expression, independent of the influence of the viral Tat activator. In addition to the now well-know HDAC inhibitors, she reviewed the activity of the BRD inhibitor JQ-1, who inhibition of BRD4 appears to drive the disruption of latency in model systems. JQ-1 activates HIV without the need for Tat, and has minimal cell toxicity. However when tested in patients' cells as monotherapy, JQ-1 is not very active. Further research to develop better BRD inhibitors is underway. Following this, G. Darcis from the van Lint laboratory in Belgium presented experiments showing the anti-latency activity of combinations of compounds, focusing on those that release active P-TEFb and induce NF-kB signaling. These are two pathways whose activity is needed to get the viral promoter out of the shutdown state, and leave latency. The BRD inhibitors JQ1, iBET, and iBET151 were tested in combination with the inducers prostratin, bryostatin, and ingenol in the CD4 cell line model of latency, JLat 9.2. In these experiments there did seem to be some true synergy, with a higher percentage of the cells being induced to leave latency, rather than simply the induction of more expression from the same fraction of the latent population. However, studies performed measuring HIV-1 recovery from CD8-depleted PBMCs of ART-suppressed patients were more variable and less convincing. This promising work will be further pursued.
Vicente Planelles (Utah) presented the discovery of a new family of compounds that reactivate latent HIV in central memory T cells, achieving what he called the "holy grail' in finding a source of stimulation that would lead to viral reactivation without leading to cytokine storm; cellular activation, proliferation, cytokine secretion (nb. my laboratory assisted with some of the evaluation of one of these compounds). The lead compound, called C7, is a hydroxybenzotrialzole (for you chemists). The compound class appears to potentiate gamma-C type cytokine signaling (eg. IL2), and may induce HIV in this way. C7 activity is not blocked by cyclosporine or an NFkB inhibitor, but is blocked by inhibitors of the Jak pathway or Stat 5 pathway. Planelles is beginning to explore the structural requirements for these compounds to function, and showed that C7 and some derivatives do not induce cell activation, proliferation, or CD25 or CD69 upregulation, but do induce persistent phosphorylation of STAT5. Initial studies do not show marked toxicity, but animal studies are not yet done.
Next, Margolis from UNC (that's me) discussed work done in his laboratory by two talented postdoctoral fellows, Carolina Garrido PhD and Julia Sung MD. Our group is beginning to explore how to model induction and clearance of latent HIV infection in the laboratory with the goal of creating an entirely autologous system that uses only primary cells from the patient. At issue is the need to gain an understanding of how agents may be employed either alone or in combination, and how agents that reverse latency will impact the activity of immune effector cells (that must clear infected targets). Ultimately, this must be tested in the clinic, but perhaps laboratory models can provide guidance and increase the efficiency and safety of clinical studies.
Sung studied anti-HIV CTLs that had been expanded ex vivo against HIV peptide pools. This technology is a clinically proven one, safe and effective for the treatment of viral infections such as EBV and CMV in oncology patients. Expanded CTLs can bypass an impaired anti-HIV immune response, be precisely controlled in the quantity and timing of administration, and may persist. While this high-resource technology may not be applicable worldwide, its use for proof-of-concept is feasible. We showed that HXTCs (HIV-specific, expanded cytotoxic T cells) could inhibit productively infected cells, even when cells were superinfected with autologous virus obtained from the patient's own resting CD4 T cells. A new assay termed an Ex Vivo latency clearance assay was developed, a modification of the Q-VOA quantitative viral outgrowth assay, used to measure the size of the latent reservoir. We found that HXTCs could target and kill infected cells emerging from latency in this assay, whether viral outgrowth was induced by maximal activation (PHA stimulation) of by a clinically relevant concentration of the HDAC inhibitor VOR. The presence of VOR at a concentration and time-of-exposure that was clinically tolerable achievable and induced viral outgrowth, did not blunt the activity of HXTCs. Garrido then studied an innate arm of the immune system, NK cells, in similar culture conditions and assays. Again, NK cells were able to target and clear HIV in both the actively infected cells and latently infected cells in the latency clearance assay. Although not as many experiments have been performed with VOR, it again does not appear that VOR impairs NK cell antiviral activity. A clinical trial testing HXTCs in vivo is planned.
In perhaps the 2nd most anticipated presentation of the meeting O.S. Sogaard of the Aaarhus Hospital group in Denmark presented findings from their 2nd HDAC inhibitor study, using romidepsin. They report that this Class I selective HDAC inhibitor is also safe and effectively in reversing HIV-1 latency in vivo. By giving romidepsin 5 mg/m2 IV at day 0, 7, and 14 in 6 volunteers, the group observed mild AEs but only one severe adverse event (headache) that was not felt to be related to the drug. With each dose of drug, they measured a steadily increasing level of cellular acetylation, validating drug activity in the blood. Why acetylation increased continuously but did not drop off drug was not explained --- perhaps the effect of Romi lasts a week. Measures of cell-associated HIV RNA in total CD4 cells (not resting CD4 cells as in the VOR studies) showed a very small fold-increase at the first dose, but a mean 4 to 5 fold increase in the study group at doses 2 and 3. Three of 6 patients also had low-level viremia "blips" to 20-30 plasma HIV RNA copies/ml on 1 dose, five of 6 to 40-100 c/ml on dose 2, one to 20 and one to 120 c/ml at dose 3. There was no change in HIV DNA levels overall. There were some alterations in the distribution of T cell subsets in circulation over the dosing intervals. The findings suggest that Romi may have anti-latency activity as has been shown for VOR, although it would be nice to confirm this by a rigorous assessment of cell-associated HIV RNA in circulating resting CD4 cells. Although the presence of low levels of plasma viremia might suggest that latent infection has been disrupted, it is not proof of this, and the blips could have been due to the induced circulation of effector cell populations more likely to have encountered viral particles in one fashion or another.
This was followed by the most anticipated presentation of the meeting --- the update from Debbie Persaud (Hopkins) on the unfortunate return of viremia in the Mississippi Baby. If you have not been under a rock since CROI in 2013, you know the story of the child born with no prenatal care to an HIV+ mother, with documented viremia on day 1 of life, and treated with ART hours after birth, initially with AZT/3TC/NVP and then AZT/3TC/Kaletra. On ART until ca. 12 months of age, the baby was lost to follow-up and off ART until the child re-entered care at 23 months of age. Surprisingly, viral loads were still undetectable, despite being off treatment for almost a year, and sensitive research assays have shown plasma HIV RNA repeatedly <1.6 c/ml since 26 months of age. Persaud reported that only traces of HIV DNA had been inconsistently seen in circulating PBMCs. Further the child's HIV Western blot was negative since 24 months, and no HIV-specific CD4 or CD8 responses were seen at 23 months. HIV DNA in PBMC or resting CD4s was undetectable out to 26.6 months (mean LOD <2.6 c/million), and by an outgrowth assay replication-competent HIV was not recovered from resting CD4 T cells, yielding an IUPM less than 0.154 per million at 26 months. Sixty cultures of one million CD8-depleted PBMCs were done over time since, all negative. A single copy assay with improved sensitivity (iSCA - for single copy assay with integrase PCR primers) was negative at less than 0.4 c/ml on three assays over time.
But after 27 months off ART (age 41 months), plasma viremia was again detectable twice, but only initially peaking at 16,750 c/ml. However, in the setting of declining CD4 cells, AZT/3TC/ RFV was reinitiated. Rebound HIV was genetically similar to the mother's original HIV sequences. Unusually, as viral rebound that child was found to have fully seroconverted, that is the HIV Western Blot was fully positive with multiple bands.
While disappointing, this case allows a number of important observations. The result of assays that were able to be done in this small child during ART interruption were not similar to that of an elite controller, and suggested that the residual reservoir infection might actually have been totally shut down or shut off for most of 27 months. However, rebound viremia did not look like acute infection, or like the rebound of the Boston transplant patients --- which appeared to behave like acute HIV infection in the setting of a HIV-naïve (transplanted) immune system. In this case, the child's immune system seemed to have some recollection of what HIV looked like, as the antibody response was fully developed and early viremia did not reach high, acute-infection-like, levels. One hypothesis would be that there was perhaps only a single, or a very few number of latently infected cells founded in the child, perhaps naïve cells, and that this cell or nest of cells remained quiescent for 27 months, until called upon by the immune system to respond to something. This would suggest that the reversal of latency and eradication of latently infected cells will need to be nearly complete, and should be accompanied by the provision of some sort of antiviral immune response or safety net system that persists for several years, to mop up the last few latent cells that might be left behind.
Dan Barouch (Harvard) then presented recently published work studying the effect of the administration of broadly-neutralizing Abs to rhesus magaques infected with SIV. In this model, 20 animals were infected rectally with the virulent strain SIVmac251, and then treated with TDF/FTC/Dolutegravir at varying times but as early as 3 days after infection, for 24 weeks. SIV proviral DNA was detected in lymph nodes and gut mucosa but not in PBMC when ART was started on Day 3. Early ART, as expected, reduced the burden of cellular SIV DNA. As was reported in man in acute infection (Archin PNAS 2012), pre-ART viremia during acute infection predicts the size of the viral reservoir. When ART was stopped in the (non-human primates) model, time to viral rebound correlated with the duration of viremia before start of ART.
HIV can persist despite therapy in a latent, silent infection within (primarily as far as we know) resting central memory CD4+ T cells. But the role of proliferation, that is cell division that results in one infected cell becoming two infected cells, has been proposed as a 2nd important mechanism to maintain "the reservoir" --- a grab bag term that means many different things to different investigators. But the important reservoirs are cells that contain genomes that encode for replication-competent HIV, do not express virus in the current state, but could produce infectious, replication-competent HIV if induced or activated to do so. However, the majority of HIV DNA genomes that can be identified do not fulfill this definition --- that is they contain HIV genetic material that is defective, truncated, or mutated and cannot produce a functional virus, and in many cases can express only fragments of HIV RNA. Very recently two papers were published in Science that found that many HIV DNA genomes have landed in or near a host gene that is related to cell cycle regulation. The frequency of these identical HIV DNA genomes in such places have led to the hypothesis that the HIV genomes have thereby interfered with the normal regulation of cell division, causing the host cell to proliferate and to copy the HIV sequences in the process. A key question is how many of these HIV genomes are actually a threat --- that is they can produce replication-competent HIV, and how many are mutated remnants of old HIV that will not cause disease.
Sarah Palmer's group, now at the Westmead Institute in Australia, characterized the genetic sequences of HIV found in proliferating cells of patients on effective ART. In patients that were treated for 4-12 yrs, two samples were taken 6 months apart and HIV sequences studies in the blood, GALT, and bone marrow by single cell sequencing. In one patient, 19 identical sequences were found to be expanded, but only 2 were identical to pre-ART sequences. In a 2nd patient, 27 expansions were found, only one of which was also found pretherapy, and the expansions were found only in effector memory cells, but none in central memory cells. In a 3rd patient the same pattern was seen, including one huge expansion of a sequence with a big, crippling deletion. In a 4th patient effector memory cells contained 100% of the expansions (64) found and all were hypermutants. Overall Palmer found that at least 63% of the expanded, repetitious HIV sequences found in effector memory cells encoded hypermutant, defective HIV genomes. More work will be done on this problem, but so far it appears that most of the HIV genomes that persist after proliferation may be persistent, dead-end, ineffective viral husks.
A potential clinical consequence of the presence of partially functional HIV genomes was discussed by Imanichi of Cliff Lane's laboratories at the NIAID. Imanichi showed evidence that defective proviruses found in ART-suppressed patients were still able to produce fragments of HIV RNA, and that some of those RNAs could produce antigenic HIV proteins. Some of the cells detected made no protein, some made viral capsid or viral envelope proteins, and some both. This phenomenon could drive some HIV-related immune dysfunction and inflammation, although instead this could be due to prior immune damage from HIV replication, or this residual HIV protein expression could contribute to immune dysfunction. Either way, this might be a future benefit of HIV eradication strategies that cleared cells producing HIV antigens, whether those proteins came from a replication-competent virus or not.
Remi Fromentin (VGTI Florida) presented work on surface biomarkers, hoping to identify the incredibly rare latently infected cells, which if discovered could of course lead to novel therapeutic strategies to cure HIV infection. HIV primarily persists in central memory CD4+ T cells (TCM), and the group at VGTI has reported by HIV DNA is found in transitional (TTM) and effector memory (TEM) CD4+ T cells as well. The durability of latent infection in TTM and TEM cells is controversial, demonstrated by Chomont in 2009 in a small group of patients followed for a short period of time, but now called into question by findings in the Palmer lab (above).
Fromentin hypothesized that the high levels of expression of the immune checkpoint blockers (ICBs) PD-1, LAG-3 and TIGIT would be associated with HIV persistence during ART, as they might identify memory CD4+ T cells stuck in the resting state. 20 subjects on ART for >3 years with HIV viral load < 50 cop./mL and with CD4 count >350 cells/μL were enrolled, and integrated HIV DNA in sorted memory CD4+ T cell subsets expressing PD-1, LAG-3, TIGIT or co-expressing the 3 receptors was measured. Wilcoxon test was performed to compare the associated of integrated HIV DNA and ICBs expression. PD-1 identified TCM and TTM cells enriched for integrated HIV DNA (p=0.019, p=0.004 respectively). HIV DNA+ TEM cells tended to express TIGIT (p=0.078), but did not differ from uninfected cells in PD-1 expression. LAG-3+ identified HIV DNA+ CD4+ cells (p=0.002) if TCM, TTM, and TEM were analyzed in one group together, but presumably (data not shown) for the memory subsets when analyzed as separate groups. A sensitivity analysis (ie, how many ICB+ cells were HIV DNA negative, and so what was the predictive value for having HIV DNA of a cell being ICB +) was not shown, and the frequency of replication-competent HIV in these HIVDNA + cells was not discussed. However, such markers might identify a subset of cells in which latency was enriched. In support of this, in a related presentation, Evans of the Lewin lab (Melbourne) showed that PD-1 and Tim-3, were preferentially expressed on resting CD4+ T-cells that were latently infected following co-culture with myeloid dendritic cells in vitro.
Declaring my obvious bias for a UNC colleague, Victor Garcia presented a challenging and interesting in vivo analysis of HIV replication in the myeloid compartment using the BLT humanized mouse model pioneered by his group. In this model, an immunodeficient mouse is transplanted with a human immune system systemically, generated human T and B cells and myeloid (macrophage/monocyte/dendritic) cells and most aspects of HIV replication, pathogenesis, antiviral drug response, and latency in T cells is replicated. Jenna Honeycutt in his lab re-engineered the model to lack T cells so that she could evaluate the role of macrophages/microglia HIV disease in the complete absence of T cells both in the brain and systemically. This macrophage-only-mouse (MOM) model was then infected with a variety of HIV strains known to be macrophage-tropic (CH040; HR-CSF; BaL, CH058; RHPA, ADA, THRO). CH040(4013env) was a clone isolated from an acutely infected patient that contains a macrophage-tropic envelope. Sustained replication of virus (HIV RNA +) was shown in 9 MOM with CH040. There were, however, low to undetectable levels of viral DNA in the peripheral blood cells from MoM, likely as most infected myeloid cells are in the tissues. When MOMs were given G-CSF, which mobilize myeloid cells in to the circulation, HIV DNA was seen in blood cells. ART (tenofovir, FTC, raltegravir) fully suppressed HIV replication in this model. Of interest, the decay of plasma viremia on ART had a rapid decay, faster than BLTs with T cells. This finding surprised the crowd, but supports earlier arguments in human studies that the slower "2nd phase" of the 2-phase viral decay seen in human treatment does not originate from "long-lived" myeloid cells, but from T cells of lower activation states. Analysis of HIV replication in the brain, and of latent infection in myeloid cells in this model is ongoing.