By Richard Jefferys
At the recent International AIDS Conference in Washington, D.C. (AIDS 2012), the research effort to develop a cure for HIV infection attained a higher profile than it ever has in the past. At a press conference on July 19, the International AIDS Society (IAS) officially launched its Global Strategy “Towards an HIV Cure,” in conjunction with a two-day symposium that immediately preceded the main conference. The IAS strategy is chaired by Françoise Barré-Sinoussi and Steven Deeks and involves a multiplicity of stakeholders, including TAG. Details are available free online in a document titled “Full Recommendations, 1st Edition, July 2012.” A shorter summary and commentary have been published in the journals Nature Reviews Immunology and Nature, respectively. In essence, the strategy is a scientific review of the obstacles to curing HIV (as they are currently perceived) and possible approaches to overcoming them. Among the goals is to enhance collaboration among stakeholders and attract new sources of funding to support cure research.
Presentations and video from the IAS symposium, held on July 20 and 21, will be available online soon. Because abstracts were selected from submissions to AIDS 2012, many of the studies discussed at the symposium—including those by Fabio Romerio and Timothy Henrich, described below—were subsequently presented at the conference and can already be viewed on the website.
The VISCONTI Cohort
Among the most publicized data presented at the IAS symposium related to a French cohort (dubbed the VISCONTI cohort) comprising 14 individuals treated within ten weeks of infection who, after an average of around three years on antiretroviral therapy (ART), interrupted treatment and have subsequently maintained control of viral load to less than 50 copies/mL for an extended period (a median of 6.6 years; range: 4–9.5 years). This study was the focus of an overview talk by Asier Sáez-Cirión and an abstract presentation by his colleague Christine Bacchus. Preliminary results have been published and presented before: in a letter to the journal AIDS in 2010, five cohort members controlling viral load off ART for a median of 6.25 years—designated posttreatment controllers (PTCs)—were described (out of a total of 32). At the Conference on Retroviruses and Opportunistic Infections (CROI) in 2011, a poster presentation reported that the number of PTCs had increased to ten (median duration of control: six years).
Sáez-Cirión updated these results with the information that a total of 14 PTCs have now been identified. Notably, four of these individuals are showing declines in HIV reservoirs (as measured by HIV DNA levels) over time. Sáez-Cirión highlighted a number of unusual features of this cohort that set them apart from elite controllers (individuals who control viral load to undetectable levels in the absence of any treatment). Most importantly, they lack the favorable immune response (HLA) genes that are consistently associated with elite control: HLA B*57 and B*27. Instead, around half the PTCs possess the HLA B*35 allele, which in untreated HIV infection is associated with a significantly increased risk of rapid disease progression. In addition, HIV-specific CD8 T-cell responses are of lower magnitude than those typically seen in elite controllers, and levels of immune activation and HIV DNA are also low.
Sáez-Cirión attempted to assess how frequently the PTC phenomenon occurs after primary infection treatment. In a preliminary look at the French Hospital Database on HIV, 756 individuals were identified who started ART within six months of infection and continued for at least a year. A subset of 74 eventually interrupted ART and, of these, 15.7% maintained undetectable viral load for a minimum of two years. Sáez-Cirión also cited a study by Cécile Goujard and colleagues (published shortly after the meeting in the journal Antiviral Therapy); in this case, out of 164 participants, 8.5% maintained viral loads below detection for two years after interruption, and 7.2% at three years. In contrast, Sáez-Cirión noted that an analysis of 34,317 HIV-positive individuals in France identified only 81 elite controllers, putting the estimated proportion of individuals likely to attain control of viral load in the absence of any ART at around 0.24%.
The duration of viral load control in the VISCONTI cohort PTCs makes them unusual, and they are receiving attention as a possible model of a “functional cure” in which HIV is suppressed (rather than eradicated) without treatment. But there are many unanswered questions and apparent contradictions with other studies that need to be addressed and resolved. Sáez-Cirión noted that all but one of the 14 individuals had symptomatic primary infection, high viral loads, and low CD4 counts at the time of initiating ART—as he put it, their primary infection appeared “tougher” than is typical. Yet in the Goujard study he cited, the factors associated with becoming a PTC were just the opposite: high CD4 counts and low viral loads (in addition to female sex). An independent analysis of the frequency of PTCs—which the authors acknowledge was prompted by the VISCONTI data—was published in the Archives of Internal Medicine on July 23. A total of 259 individuals from the multicountry CASCADE cohort were identified who received ART within three months of infection. The probability of maintaining PTC status 24 months after ART interruption in this analysis was 5.5%, and the characteristics of these 11 individuals did not differ from those of the overall study population. Sáez-Cirión was questioned at the symposium regarding levels of inflammatory biomarkers and any clinical events in the VISCONTI cohort PTCs; he responded that these analyses are ongoing and incomplete. In terms of CD4 counts, Sáez-Cirión stated that only one of the 14 is showing a decline over time. The reason for the apparent overrepresentation of HLA B*35 is as yet unclear; when quizzed on the issue, Saéz-Cirión suggested that possession of this allele may have explained the high prevalence of symptoms in the cohort which, in turn, prompted them to start ART early. However, previously published studies have not reported an association between HLA B*35 and primary infection symptoms. Further complicating the question is the existence of HLA B*35 subsets named Px and Py, and only people possessing HLA B*35 Px have been reported to experience rapid HIV disease progression: the distribution of Px versus Py in the VISCONTI cohort PTCs is as yet unknown.
Replication-Competent HIV Reservoirs May Be Underestimated
Latency expert Robert Siliciano presented data suggesting that the amount of replication-competent HIV DNA in people on ART may be greater than has been thought. The most commonly cited estimate posits that only around one out of every 100 latently infected resting CD4 T cells harbors replication-competent virus (the reason for the difference being that the majority of the viral DNA is mutated in ways that render it nonfunctional). Siliciano looked at 179 samples of CD4 T cells containing latent HIV that could not be induced to replicate by the standard method of PHA stimulation (PHA or phytohemagglutinin is a substance that triggers CD4 T cell division and thereby activates latent HIV). He found that while the majority of the HIV DNA proviruses were indeed defective (due to lethal hypermutation, deletions, and other alterations), an average of 16.8% (range: 6–36%) were fully intact. Siliciano’s laboratory has cloned these intact sequences and confirmed that the viruses are able to replicate. The results imply that the size of the replication-competent HIV reservoir in people on ART may be 50-fold larger than in prior estimates. However, among the questions that remain to be answered are whether these viruses can be induced to replicate in vivo, and if approaches other than PHA stimulation might be able to coax them out of hiding.
High CD2 Expression as a Marker for Latently Infected CD4 T Cells
The laboratory of Fabio Romerio at the Institute for Human Virology in Baltimore has developed an in vitro model of latent HIV infection that attempts to closely mimic the in vivo situation using primary CD4 T cells (as opposed to immortalized cell lines, which may not accurately reflect the biology of cells in the body). In a presentation at the symposium, Romerio described the use of this model to identify cell surface markers that may be preferentially expressed by latently infected cells. This is an important goal for cure research, because it could facilitate the targeting of anti-latency approaches to the cells most likely to be infected. The lead candidate that emerged from Romerio’s work is CD2, which was expressed at higher levels on infected versus uninfected CD4 T cells. Romerio is collaborating with Nicolas Chomont from the Vaccine & Gene Therapy Institute of Florida to investigate whether these results are also reflected in vivo: a preliminary study involving six individuals on long-term ART found that, in all cases, HIV DNA was more commonly present in CD4 T cells expressing high levels of CD2.
T Memory Stem Cells as an HIV Reservoir
One of the challenges for anyone attempting to follow the immunology literature is the ever-expanding panoply of cell subsets that are being described. The most recent addition to the T-cell family is the T memory stem cell or TSCM for short. According to a study published in Nature Medicine last year, TSCM represent the least differentiated form of memory T cells; they are said to have stem cell-like properties because they can proliferate extensively and give rise to the many more specialized types of memory T cell that exist (central memory cells, transitional memory cells, effector memory cells and effector cells). María José Buzón from the Ragon Institute of MGH, MIT and Harvard gave a presentation at the symposium on the contribution of TSCM to the HIV reservoir in individuals on ART and elite controllers. Buzón showed that while TSCM only make up a small proportion of circulating T cells (0.5–1%), HIV DNA could be detected in TSCM and they accounted for around 17% of the total detectable reservoir. The number of latently infected TSCM appeared stable and Buzón’s preliminary data hinted that their relative contribution to the HIV reservoir might increase over time in individuals on ART.
Absence of a Detectable HIV Reservoir after Allogeneic Stem Cell Transplantation
Timothy Henrich from Brigham and Women’s Hospital and Harvard Medical School presented two case reports relating to individuals with HIV who received allogeneic stem cell transplants for the treatment of cancer. Strikingly, replication-competent virus can no longer be detected in either individual after around 2 and 3.5 years of follow-up, respectively, and their anti-HIV antibody levels are declining. However, both remain on ART, so it is not known if they are cured (the extensive media coverage Henrich’s talk received was not always clear on this point). Unlike Timothy Brown, the “Berlin Patient,” who received a stem cell transplant from a donor homozygous for the CCR5delta32 mutation, the transplants in both these cases came from individuals with normal CCR5 expression. Further study may thus help reveal whether the CCR5delta32 mutation was necessary for achieving a cure in Brown, or if the transplant itself and associated factors—such as graft-versus-host disease, where the new immune system that develops from the transplant attacks the older host cells—can be sufficient. Plans are under way to conduct analytical ART interruptions to assess if a cure has been achieved in these individuals.