icon-    folder.gif   Conference Reports for NATAP  
  18th CROI
Conference on Retroviruses
and Opportunistic Infections
Boston, MA
February 27 - March 2, 2011
Back grey_arrow_rt.gif
CROI 2011 Immunology Report
  David Margolis, MD, University of North Carolina
The annual Conference on Retroviruses and Opportunistic Infections (CROI) brings news of groundbreaking advances in some years, while other years are anticlimactic. This year's CROI, held in Boston, may have had more attendees than ever before, and while there was not a major event that promised to immediately restructure clinical practice, many significant events and presentations took place. Some of these may eventually affect clinical practice, and others promise to be the focus of intense efforts over years to come.
TOPICS in this report:
- Tenofovir PrEP, oral, vaginal, topical, rectal
- Protective Vaccine
- "Cure", Eradication
- Gene Therapy - Zinc Fingers & challenges to its successful use
- Intensification of ART
Part I: Advances in Prevention of HIV Infection:

The past year or so has yielded a hopeful troika of major studies that have advanced the field of HIV prevention. CAPRISA 004 showed that the use of a vaginal tenofovir gel before and after sex could result in a 39% reduction in new HIV infections in African women. iPrEx found that Truvada (the fixed dose combination tablet containing tenofovir and FTC or emtricitabine) taken on a daily basis reduces the risk of HIV infection by 43.8% in men who have sex with men. And a little over a year ago, a HIV vaccine trial in Thailand, known as RV144, tested a "prime-then-boost" combination of two vaccines: ALVAC HIV vaccine (the prime) and AIDSVAX B/E vaccine (the boost). The vaccine combination was based on HIV strains circulating in Thailand, and lowered the rate of HIV infection by 31.2% compared to placebo based on the modified intent-to-treat analysis (new infections n=51 vs. n=74, respectively in more than 16,000 volunteers; p=0.04).
Although we understand well how the use of antiretroviral before HIV exposure could prevent HIV infection, how to use the results of CAPRISA and iPrEx pre-exposure prophylaxis (PrEP) studies in a practical way in the current treatment environment in both the developed and developing world is unclear. And how RV144 induced its modest protective effect, and how its results direct us to move the development of a highly effective HIV vaccine forward, is also still unclear.
Jonathan Mermin's plenary talk (Abstr. 19; view here) reflected broadly on these and similar challenges. As is well known, there are more than 1.1 million HIV+ people in the US, with this number increasing yearly due to the roughly 56, 000 new infections every year. While numerous behavioral interventions have been shown to decrease HIV risk behavior, acquisition, or transmission, the recent advances now begin to arm us with biomedical interventions (PrEP drugs, microbicides, and perhaps some day a vaccine) that can be used. Mermin pointed to the challenges posed by large disparities in HIV acquisition and healthcare access that exist across sub-populations of race, ethnicity, and income, and to the continuing increase in HIV incidence among gay and bisexual men.
Mermin pointed out that faster action in prevention will have a bigger impact than if efforts are delayed while the epidemic grows further. Prevention of transmission by HIV-positives, through ADAP and the administration of ART, risk reduction sessions, and case management can have a very large impact on reduction of transmission even when focused on a relatively small number of patients. Additionally newly validated PrEP approaches, needle exchange, and other prevention efforts in HIV-negatives could be focused on high-risk groups for maximal cost-effectiveness. However, before choosing to make trade-offs, such as deciding to pay for ADAP rather than rolling out PrEP, these choices should be modeled and evaluated carefully. To significantly decrease HIV incidence, the rate of new infections will have to be reduced from 56,000 per year to 16,000 per year. This is a long way to go, and there is an obvious tension between the need to do something as quickly as possible, and make the best choices based on limited information.
Connie Celum gave a nice overview of ART as PrEP (Abstr. 120; view here). She reviewed prior studies, including CAPRISA and iPrEx, and discussed the obvious potential benefits of PrEP, and the concern that thus far PrEP uses drugs that are also the cornerstone of current first-line therapy. She also highlighted microbicides that were farther back in the pipeline, including drugs other than HIV nucleoside reverse transcriptase inhibitors, and long-acting delivery systems such as vaginal rings.
Largely, Celum emphasized the need for "bridging data" to inform the best use of new prevention tools as they are developed and validated. To obtain this, we will need demonstration projects, clinical trials, mathematical modeling, and cost analyses so that PrEP can be delivered effectively. PrEP should be implemented in a multi-component, evidence-based manner, targeted to reach the most people who are at high risk, and will ultimately have to include an effective HIV vaccine, if available. Most critically, and like so many things in HIV care, the development of PrEP cannot be a zero-sum game: gains in implementing PrEP must be implemented without taking away from efforts to expand access to ART. The CDC has issued interim guidelines for the use of PrEP (view here) which emphasize targeting men who have sex with men (MSM) at high risk for HIV acquisition, delivery as part of a comprehensive set of prevention services, and accompanied by monitoring of HIV status, side effects, adherence, and risk behaviors at regular intervals. Few clinical care sites are thus far prepared to provide this, or to pay for it.
The iPrEx study of oral tenofovir as PrEP: An entire morning session was dedicated to a detailed overview of the results of the iPrEx study, and its many elegant and detailed nested substudies. Virtually every question about the use of Truvada as PrEP in MSM that could be asked and answered in a single large study was here on display (View iPrEx session). Bob Grant (abstr. 92) reviewed the overall results of the study that he designed and led (N Engl J Med 2010; 363:2587-99), and reported that the revised efficacy of Truvada to prevent HIV infection was 42% (83 infection in placebo arm, 42 in Truvada arm). All infections in the active arm of the study were associated with undetectable (91%) or low (9%) PrEP drug levels in blood, indicating no recent PrEP use. Grant then went on to discuss potential "collateral harms" that could result from the use of Truvada to prevent HIV infection, including drug toxicity, infection with Truvada-resistant HIV, and increased risky behavior. In this study that included extensive monitoring and counseling, these negative effects were not strikingly common. Grant made the argument that there could be many collateral benefits of PrEP as well, including expanded HIV testing, care, and prevention teaching, that might let the supply of ART catch up to the demand for it, by reducing the spread of infection.
Liu (abstr. 93) and Mulligan (abstr. 94LB) then outlined bone mineral density (BMD) loss in MSM taking Truvada for PrEP in San Francisco and internationally. The surprising finding here was that in SF four times more MSM were found to enter the study with low BMD. Similarly, in the international study, at baseline 36% had low bone mineral density in the spine and 18% in the hip. This may be an issue specific to these cohorts, or may suggest that "normal" Z scores need to be redefined. Overall as with other studies of tenofovir use there was a small but statistically significant decline in BMD at the total hip and femoral neck, without any increase in the frequency of fracture observed. Although no major problems with BMD have been seen so far, if Truvada PrEP is adopted and used for prolonged periods of time, extended studies of this issue will be needed. Amico (abstr. 95LB) reported findings of adherence monitoring and drug level measurements within these studies. Overall, self-report and pill count-based adherence were associated with drug detection, and self-report of missing 50% or more of doses associated with low or no drug levels detected. Refill-based (pharmacy) assessment were similarly predictive. These findings were generally reminiscent of ART adherence studies. Anderson (abstr. 96LB) reported the measurement of intracellular active tenofovir and FTC in frozen blood cell samples. FTC-TP and TFV-DP were detectable in 100% of the viably cryopreserved PBMC from those with suppressed HIV RNA, and so this assay seems well-suited to demonstrating adherence, but as drug could be detected some of the time in patients who appeared to have intermittent adherence it seemed that single assays would not be very useful, and serial assays would be needed to measure or monitor drug adherence.
Teri Liegler (abstr. 97LB) studied minor viral sequence variants that might convey drug resistance in iPrEx patients using allele-specific PCR designed to detect K65R and K70E (TDF), and M184V/I (FTC) mutations. In 100 infections examined that occurred on study, none showed FTC or TDF resistance by standard population sequencing. 91 samples were analyzed for minor variant drug resistance by allele-specific PCR, but these were only found in patients who were inadvertently enrolled in the study, despite the fact that they were infected prior to entry.
Other PrEP data: Several other approaches to PrEP were presented at CROI. Peter Anton reported the results of a small study (abstr. 34LB) that showed that rectal dosing with the vaginal formulation of 1% tenofovir gel in MSM was neither entirely safe nor fully acceptable volunteers. Tissue culture studies also suggested that a single use of tenfovir oral or rectally these might not be effective at preventing HIV infection due to anal sex. This indicated a need for better rectal-specific formulation of microbicides. Craig Hendrix (abAbstr. 35LB) presented a pharmacokinetic study comparing oral tenofovir, vaginal tenofovir 1% gel, or both. Acceptability of the single or combined regimens was reported as good, although oral dosing was preferred. No drug was detected in plasma following vaginal dosing, and higher concentrations were found in vaginal tissue and cervicovaginal lavage fluid following vaginal dosing. There was no advantage seen (higher concentrations in any compartment) to dual oral and vaginal dosing. Ultimately, a clinical trial with HIV infection endpoints will be required to discern whether it is more effective to have high vaginal tissue cervicovaginal fluid PrEP drug levels (vaginal dosing), or higher plasma levels (oral dosing).
Morris (abstr. 990) found that daily vaginal administration of 50 mg of the anti-HIV-1 monoclonal antibodies C2F5, C4E10, and C2G12 to healthy women for 12 days was safe and well tolerated. Concentrations measured appeared to be sufficient to block HIV transmission for many hours post-administration. Passive delivered monoclonal antibodies could be a novel and effective microbicide approach. Finally, in a poster presentation (abstr. 1001) the International Partnership for Microbicides of Silver Spring presented continuing work to develop and test a vaginal microbicide ring continaing the potent NNRTI, Dapivirine (TMC 120). The goal is a formulation of vaginal rings for monthly or longer use. Systemic exposure to Dapivirine was low, and peak concentrations were reached in vaginal fluids 24 hours after insertion. Multiple DPV vaginal rings were well tolerated with no safety concerns in healthy HIV- women, and pharmacokinetic data supported use over a 35-day period.
Making a Protective Vaccine: John Mascola of the NIH Vaccine research Center discussed recent studies (abstr. 62) of the isolation and characterization of broadly neutralizing HIV monoclonal antibodies (bn-mAbs). Several such bn-mAbs were described in the 1990's, but since 2008 three new important bn-mAbs to HIV have been described. Recent technical advances in isolating and growing antibody-producing B cells from HIV-1-infected donors have allowed a dramatic increase in the isolation of new bn-mAbs. These monoclonal antibodies can direct us to target specific structural features on the HIV envelope in vaccine development, and perhaps synthetic bn-mAbs could be used to prevent mother-to-child transmission, or given directly as a microbicide. Laurent Verkoczy (abstr. 63) followed with as discussion of another group of bn-mAbs that have been uncovered by the CHAVI group. These Abs have some reactivity against human self antigens, and so the B cells that make them may be naturally deleted --- perhaps explaining the rarity of bn-mAbs. Using novel techniques to detect Abs derived from bone marrow B cell precursors, she showed evidence that bn-mAbs are encoded in the precursor cells. Therefore, techniques to break the natural restrictions that prevent the maturation of Abs against host might allow these bn-mAbs against HIV to develop. Obviously the challenge would be to do this without allowing autoimmune disease to develop.
Jerome Kim, leader of the Thai vaccine trial, presented the work of his team and a large group of collaborators seeking to understand exactly how the vaccine used in the trial offered the marginal protection observed in the trial, to guide future improvements in vaccine design (abstr. 65). Efforts thus far have focused on a) the HIV envelope gp120 antigen used in the trial, and b) the gp120-specific antibodies that were induced by vaccination. In the industrial process used to synthesize the vaccine antigen, the antigen was "tagged" with a protein derived from herpes virus, the gD peptide, to simplify the production and purification of the antigen.
It appears that, quite by accident, this addition to the natural gp120 evoked enhanced expression of functional gp120 epitopes, and enhanced binding of antibodies to the vaccine antigen. The herpes protein seemed to, in general, augment the immune response to the HIV gp120 antigen. Many complex further studies are planned to help to understand how the Thai trial vaccine seemed to succeed, and to help design better 2nd generation vaccines.
Part II: The "Cure" of HIV Infection:
This word, enough to elicit criticism or ridicule only a few years ago, was spoke aloud at a shocking number of sessions and presentations. Clearly the game has changed, most likely due to the gradual realization that we cannot treat all the patients who need it chronically for as long as treatment is needed, and the example of the "Berlin patient" that demonstrated that cure --- if in an extreme situation --- was conceivable. However, to discuss cure or eradication, we should first define what we mean. When people are HIV-infected, they have HIV particles in their bodies, produced by infected cells, containing HIV RNA, and these viruses are able to infect new cells, new people, and cause disease either directly by killing cells or by causing inflammation. When HIV infects a cell, it is copied into DNA and integrates into the host cell, essentially becoming a new gene in the genome of the infected person. Many of these viral genomes are defective and cannot make a virus, but even the defective ones might make HIV proteins, cause inflammation, and induce some sort of disease.
So I think that when most people talk about a "cure" they mean that they hope for a state in which they take no drugs, have a healthy immune system, are not infectious to anyone (including their children), and will not get any illness related having had HIV in their body --- either an immune deficiency or an inflammatory problem. To achieve this may require the removal, destruction, or inactivation of all the virus and functional viral genomes in a person. Perhaps some viral DNA genomes might be left behind, as long as they were not doing anything that cause any damage or inflammation. This is a pretty tall order, but I think it is what we are talking about when we talk about "cure" or "eradication."
Very recently an alternate goal has sprung up, called "functional cure." To me, to be worthwhile, this has to be pretty close to actual cure, but as I understand it "functional cure" would be a state in which there was still HIV in your system, but it was highly controlled by the immune system or some other phenomena without the use of medications. You would have a normal healthy immune system, and no "inflammation problems." There would be no chance, or very little chance, that you could transmit infection. This seems to me to be also a tall order, as we know that even natural elite controllers of HIV infection may transmit HIV, and such patients tend to have elevated levels of inflammatory markers, and do not have totally normal immune systems.
Jon Karn (abstr. 163) laid out the molecular mechanisms through which the HIV genome can persist silently within rare resting CD4+ T cells, the major latent reservoir [(at 32:41 here) in session called "Obstacles to A Cure"]. These mechanisms define the targets of drugs or therapeutic strategies that seek to eradicate infection by flushing the latent virus out of quiescence. Karn described how, rarely, the integrated virus can fail to produce enough of its own activator protein, Tat, and then lapse into the silent state. This failure can arise from reduced levels of viral mRNA expression, induced by changes in chromatin within which the virus is integrated. This silenced state, once established, is enforced by several factors such as a cellular protein complex called NELF (negative elongation factor), and insufficient availability of another cellular factor complex called P-TEFb (positive transcription elongation factor-b).
Karn concluded that therapies must be developed to both remove restrictive chromatin markings and activate P-TEFb to allow purging of latent viral genomes.
Mutating people: gene therapy to create HIV-resistant cells: The only person known to be cured of HIV infection is the "Berlin patient," an HIV-infected individual with leukemia who received 2 bone marrow transplants from a CCR5∇32 homozygous donor. The donor's cells carried the mutation in the CCR5 HIV co-receptor that naturally makes cells uninfectable by HIV, or at least by the dominant HIV strains that must use the CCR5 receptor. It should be mentioned that the role of the CCR5∇32 cells in the cure of this patient is unclear, as is the contribution of other factors in the case such as T-cell ablative chemotherapy and graft-vs-host disease. However, to attempt to recapitulate this success but avoid a bone marrow transplant, efforts are ongoing to alter the genes that encode the HIV co-receptors CXCR4 and CCR5. Following this "genetic surgery," cells from a patient could be returned to the body, and might result in a cure or a functional cure.
Carl June (abstr. 165) has been working on gene therapy for more than 20 years. He provided and overview of studies [(at 1:28:45 here) in "Obstacles To Cure" session]. that were also reported in more detail by Lalezari (abstr. 46) and Wilen (abstr. 47). The genetic surgery in question involves the use of a zinc-finger nuclease developed by Sangamo Biosciences. This engineered enzyme has a DNA-binding motif called a zinc finger that can be designed to recognize and bind a specific DNA sequence, in this case part of the normal human CCR5 gene. The zinc-finger is made as a fusion protein connected to a DNA cutting enzyme, a nuclease. This enzyme is delivered in the laboratory to CD4 cells taken from the patient by an engineered adenovirus, a relative of a cold virus that has also been used to deliver genetic vaccines for HIV. When the enzyme gets into CD4 cells, it cuts a tiny chunk out of the CCR5 gene, and when the natural DNA repair mechanisms of the cell repair the cut, extra DNA used to repair the damage leaves a mutant sequence in the CCR5 gene, encoding a now-dysfunctional receptor.
The ongoing phase I clinical trial has 2 cohorts of patients: cohort 1 includes patients who have failed 2 or more HAART regimens and remain viremic; cohort 2 includes patients doing well on a stable ARV medication, who are willing to undergo a single structured treatment interruption. June reported on 14 patients (9 evaluable at the time of presentation) treated at UPenn so far. There were no serious adverse events. As the mass indusion of modified T cells, there were increases of 100-200/μl CD4 cells in most patients that seemed to persist for up to a year, and a persistent detection of a stable number of modified CD4 cells with mutant CCR5 genes. The numbers of these modified cells in the rectal mucosa of one patient clearly increased, and in two others were stable after an initial rise after transplant. Overall, the stability of these cells was hopeful in one sense, but their stable number left the concern that these cells were not able to expand any further. No immune function data for these cells was shown. Two of the patients did undergo a treatment interruption, as planned, and viremia rebounded. These patients went back on ART.
Lalezari (abstr. 46) reported an open-label, single-arm study of 9 aviremic HIV subjects on ART with CD4 counts of 200 to 500 cells/mm3. Three patients each received 1x1010, 2x1010, or 3x1010 total cells. There was evidence of an up to 2.9-fold expansion of CCR5-modified cells, with a regression to baseline over about 1 year. Overall roughly 1.6% of the total CD4 cells were modified, although the patient with the best result had about 10% of his cells replaced. Infusions were safe and well tolerated with only mild adverse events. Wilen reported (abstr. 47) on pre-clinical studies in the laboratory that are aimed at mutating the CXCR4 receptor. The CXCR4 receptor is used by a minority of clinical isolates, but the ability to protect cells against these more pathogenic viruses could be important. Cells with CXCR4-disruption that were challenged in the laboratory with CXCR4-tropic HIV survived and expanded. When these CXCR4-ZFN treated T cells were engrafted into a humanized mouse model and challenged with an X4 tropic virus strain, the same survival advantage of modified cells was seen, but a dual-tropic variant emerged that enabled infection of CCR5+ cells in the absence of CXCR4. It remains to be seen if an HIV-resistant immune system will have to be engineered by the disrupting both CCR5 and CXCR4 genes.
For the time being, we do not need to discuss how gene therapy to induce a functional or complete cure of HIV would be medically practical, as this is proof-of-concept research. Remember, that until recently triple therapy was said to be medically impractical in Africa. So these proof-of-concept studies are truly landmark ones. However, several challenges arise that are yet to be answered by gene therapy approaches designed to make HIV-resistant CD4+ T cells. First, the engineered cells must survive, maintain immune function, and expand to a population size sufficient to at least drop HIV viremia substantially and modify disease. Of these, only survival has been demonstrated so far. Next is the issue of the need to alter both the R5 and X4 receptor. Then there is the potential need to readminister cells periodically.
One approach to overcome the need for long-term supply of modified CD4+ cells was illustrated by Paula Cannon's presentation (abstr. 164). Using the same Sangamo zinc finger technology, her group reported the permanent knockout of the CCR5 gene in human hematopoietic stem cells. In pre-clinical, laboratory experiments transplantation of ZFN-engineered human HSC into a humanized mouse model resulted in the generation of CCR5-negative T cells that preferentially survived following challenge with an R5-tropic strain of HIV-1. This eventually led to steep declines in HIV-1 levels, in both the peripheral blood and gut mucosa of the animals, and restoration of normal levels of human CD4+ T cell levels throughout their lymphoid tissues. In addition, the site-specific insertion of a GFP reporter gene at the CCR5 locus resulted in an increase in GFP-positive CD4+ T cells, but not CD8+ T cells, following R5-tropic HIV-1 challenge. This demonstrated the potential to insert a new therapeutic, anti-HIV gene in the future. However, these mice were reconstituted with modified stem cells, not transplanted with modified cells in the background of pre-existing, HIV-susceptible, stem cells. In the real-life example of the Berlin patient, most of the patient's own stem cells were wiped out by chemotherapy. It remains to be seen if transplantation with modified cells can be accomplished at a sufficiently high level to provide enough CD4 cells to have a clinical effect, without the need to somehow destroy a patient's own HIV-susceptible, stem cells. This is one advantage of modifying CD4 cells, which will have a selective advantage in the presence of HIV over the patient's native cells, in comparison to stem cell modification, wherein unmodified stem cells may still provide targets for HIV replication.
This brings up an old issue of the ability of HIV to infect human hematopoietic stem cells. Collins (abstr. 162) reviewed recently published data from her laboratory suggesting that, contrary to previous reports (Weichold Blood 1998; Shen J Virol 1999; Zhang J Clin Invest 2007) HIV could use the CXCR4 receptor to productively infect multipotent hematopoietic stem cells. Further studies may be required to clarify the inconsistencies between these different reports.
Is there a measurable effect of intensification of ART in ART-suppressed patients? Several studies have shown no effect of intensification of ART on residual viremia, the low levels of virus that can often be measured by research assays in patients who are "undetectable." However, Buzon (Nat Med 2010) found that treatment intensification with the integrase inhibitor raltegravir led to an increase in unintegrated HIV DNA in some patients 2 weeks after intensification, and a normalization of immune activation markers. Gandhi (abstr. 51) found no evidence of this increase in HIV DNA, although they measured this parameter at a later timepoint than did Buzon. They also found no significant changes in immune activation.
Markowitz, however, asked if there might be some benefit of intensive ART when initiated during acute HIV infection (abstr. 148LB). They treated 14 newly infected patients 3-drug therapy with Truvada and boosted atazanavir or darunavir, and 26 acutely infected patients with 5-drugs that also included twice-daily raltegravir and maroviroc (n = 26). There was no impact of intensified ART on as-treated response rates, residual plasma viremia by single-copy assay, degree of immune reconstitution in peripheral blood or levels of CD8+CD38+ T cells, a marker of immune activation. There were 3 failures in the intensive arm for unclear reasons, but the burden of additional medications may have contributed to non-adherence that was not reported. However, another study (abstr. 517) of the same intensified ART in 20 patients has presented by a Thai/US Army group reported that such intensified therapy given before sero-conversion may be associated with lower on-therapy T cell activation levels and smaller HIV reservoirs as measured by colonic mucosal HIV DNA, compared with later ART initiation. Other evidence that intense control of viral replication could result in low levels of persistent infection was presented by Graf (abstr. 292), who found low levels of HIV integrated DNA and high levels of dead-end unintegrated viral genomes in natural elite suppressors.
Other tools to eradicate HIV infection: Doyon (abstr. 48) screened 640 compounds from a chemical library of natural products to find compounds that induced expression of HIV a cell line model of latency. Of these, they found 2 compounds that induced expression without severe toxicity. Both compounds appeared to re-activate latent HIV-1 expression via activation of the important T cell factor NF-κB, but by activating kinase signaling molecules in the cell in a slightly different way.
Vandergeeten (Abstr. 298) presented a different approach to HIV eradication. This model of HIV latency and persistence posits that infection in resting memory cells is so long-lived not primarily because the virus is shut down and latent, but because homeostatic proliferation maintains the reservoir by copying it. The idea that memory cells are maintained by this kind of low-level replication or proliferation is a central one in immunology. Chomont published experiments in 2009 (Nature Med) that suggested that infected cells could copy themselves without expressing the latent virus and without dying. In this work, Vandergeeten et al stimulated CD4 T cells isolated from subjects receiving suppressive HAART with IL-7 and IL-15 (1 and 10 ng/mL) in the presence of drugs to prevent new round of infection. They again suggested that IL-7 maintains the reservoir by inducing homeostatic proliferation of cells harboring integrated DNA with minimal effect on HIV reactivation both in vitro and in vivo, and oppositely that IL-15 induces viral reactivation in vitro as a consequence of differentiation and proliferation of effector memory cells. However, the in vivo evidence for the effect of IL-7 was weak: data from 3 patients given a dose of IL-7 showed that integrated HIV DNA increased in two of three patients, and no variance was given for these PCR assays. As in the earlier Chomont paper, this may have been an increase in defective HIV genomes, not replication-competent viral reservoirs. Further studies will be needed. The authors also suggested that IL-15 therapy may be used as a strategy to deplete the latent HIV reservoir, as its effect was opposite that of IL-7.