HIV-1 antibody 3BNC117 suppresses viral rebound in
humans during treatment interruption
Download the PDF here
"Using antibodies against HIV could be part of a "kick and kill" strategy, said study co-author Dr. Michel Nussenzweig - to kick the virus out of its hiding places and kill it.
Nussenzweig, a physician and investigator at Rockefeller University's Howard Hughes Medical Institute, says the first step is to flush the virus out of its cellular hiding places - a natural response to discontinuing anti-retroviral therapy. An antibody, or a combination of antibodies, then could move in for the kill, tagging cells that have been commandeered to make HIV and targeting them for destruction by the immune system.
Immunotherapeutic approaches have been highly effective in targeting certain cancers and are being aggressively pursued as way to manage HIV infection, to protect uninfected people from infection and to one day, perhaps, cure HIV. In fact, the 3BNC117 monoclonal antibody that is the subject of Wednesday's Nature Letter is one of three HIV antibody therapies likely to show promising results in suppressing the virus' rebound this year alone.
Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious and Diseases, called the effort reported in Nature "an excellent study" in "a very hot field."
Fauci said the motivating idea of such efforts is to explore whether HIV-infected patients could be given a more convenient alternative to an intensive regimen of pills. After their infection has been suppressed with anti-retroviral drugs, patients might receive intermittent infusions of one or more antibodies that would suppress the resurgence and replication of the virus. A cocktail of such antibodies not only might be more convenient and effective than anti-retroviral drugs at tamping down HIV infection, it might offer a safer treatment for patients who develop dangerous side effects to anti-retrovirals.
The therapy tested in this safety trial uses a cloned antibody found to neutralize more than 80% of HIV strains. The antibody originally was harvested from an "elite controller" - an HIV-infected patient whose immune system suppressed the virus' replication to extremely low levels without anti-retroviral drugs. Nussenzweig said that the antibody's origin - naturally occurring in an HIV-infected human - bodes well for its safety and noted that in the current trial, participants getting as many as four infusions over eight weeks had no significant adverse reactions.
Could such a therapy be a cure for HIV?
"That's a very high bar," Nussenzweig said. Driving HIV out entirely would take not only cocktails that include many antibodies, he noted; it would require some further engineering of antibodies to prime and direct the immune system's destruction of HIV reservoirs.
A more attainable goal for the near-term, Nussenzweig said, is using antibody therapy like this to protect against HIV infection in people who are still uninfected. By testing primates, Nussenzweig's group already has found that injections of the antibody 3BNC117 conferred protection against HIV infection for 23 weeks.
If antibodies are to be used to confer long-lasting "passive immunity" from infection, or to suppress HIV-infection rebounds for long periods, Nussenzweig added, they will have to be engineered to last longer inside the human body. His team already has found several mutations that could extend the life of the antibodies, and hope to pursue further studies on their safety and effectiveness, he said.
"In my dreams, it's like a flu shot. You give it once a year," said Nussenzweig, who conducted this work with funding from the Bill and Melinda Gates Foundation. But if it is to protect young women in Africa - a population highly vulnerable to HIV infection - an antibody-based shot needs to last at least three months, he said. That way, it could be given alongside long-lasting birth control shots, which have gained wide acceptance among young women there." http://www.latimes.com/science/sciencenow/la-sci-sn-hiv-antibody-therapy-20160622-snap-story.html
"The goal is a once-a-year shot for prevention and a combination approach for cure," he says, adding that this immunotherapy approach to treating HIV would resemble treatments now used against cancer.
HIV-1 antibody 3BNC117 suppresses viral rebound in humans during treatment interruption
Nature 12 May 2016
"Combinations of drugs are needed to maintain HIV-1 suppression in effective ART regimens. Similarly, combinations of antibodies were required to suppress viremia in humanized mice 6,7. We speculate that combinations of bNAbs will also be needed to increase the frequency of individuals that remain suppressed by antibody during ATI. Whether 3BNC117 can also impact the size and composition of the latent reservoir during ATI will require additional studies."
Johannes F. Scheid, Joshua A. Horwitz, Yotam Bar-On, Edward F. Kreider, Ching-Lan Lu, Julio C. C. Lorenzi,
Anna Feldmann, Malte Braunschweig, Lilian Nogueira, Thiago Oliveira, Irina Shimeliovich, Roshni Patel, Leah Burke,
Yehuda Z. Cohen, Sonya Hadrigan, Allison Settler, Maggi Witmer-Pack, Anthony P. West, Jr., Boris Juelg, Tibor Keler,
Thomas Hawthorne, Barry Zingman, Roy M. Gulick, Nico Pfeifer, Gerald H. Learn, Michael S. Seaman, Pamela J. Bjorkman,
Florian Klein, Sarah J. Schlesinger, Bruce D. Walker, Beatrice H. Hahn, Michel C. Nussenzweig & Marina Caskey
1Laboratory of Molecular Immunology, and 14Howard Hughes Medical Institute, The Rockefeller University, New York, NY 10065. 2Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114. 3Departments of Medicine and Microbiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, 19104. 4Department of Computational Biology and Applied Algorithmics, Max Planck Institute for Informatics, Campus E1 4, 66123 Saarbrücken, Germany. 5Division of Infectious Diseases, Weill Medical College of Cornell University, New York, NY 10065, USA. 6Division of Biology, California Institute of Technology, Pasadena, California 91125. 7Ragon Institute of MGH, MIT and Harvard, Cambridge, MA 02139, USA. 8Celldex Therapeutics, Inc., Hampton, NJ 08827, USA. 9Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, NY 10467. 10Center for Virology and Vaccine Research, Beth Israel Deaconess Medical Center, Boston, Massachusetts 02215. 11Laboratory of Experimental Immunology, Center for Molecular Medicine Cologne (CMMC), University of Cologne, 50931 Cologne, Germany. 12Department I of Internal Medicine, Center of Integrated Oncology Cologne-Bonn, University Hospital Cologne, 50937 Cologne, Germany. 13Howard Hughes Medical Institute, Massachusetts General Hospital, Boston, MA 02114
Interruption of combination antiretroviral therapy (ART) in HIV-1-infected individuals leads to rapid viral rebound. Here we report the results of a phase IIa open label clinical trial evaluating 3BNC117, a broad and potent neutralizing antibody (bNAb) against the CD4 binding site of HIV-1 Env1, in the setting of analytical treatment interruption (ATI) in 13 HIV-1-infected individuals. Participants with 3BNC117-sensitive virus outgrowth cultures were enrolled. Two or four 30 mg/kg infusions of 3BNC117, separated by 3 or 2 weeks, respectively, were generally well tolerated. The infusions were associated with a delay in viral rebound for 5-9 weeks after 2 infusions, and up to 19 weeks after 4 infusions, or an average of 6.7 and 9.9 weeks respectively, compared with 2.6 weeks for historical controls (p=<1e-5). Rebound viruses arose predominantly from a single provirus. In most individuals, emerging viruses showed increased resistance indicating escape. However, 30% of participants remained suppressed until antibody concentrations waned below 20 μg/ml, and the viruses emerging in all but one of these individuals showed no apparent resistance to 3BCN117, suggesting failure to escape over a period of 9-19 weeks. We conclude that administration of 3BNC117 exerts strong selective pressure on HIV-1 emerging from latent reservoirs during ATI in humans.
A fraction of HIV-1 infected individuals develops broad and potent serologic activity against the virus. Single-cell antibody cloning methods2 have uncovered the source of this activity as broadly neutralizing antibodies (bNAbs), which target different sites on the HIV-1 envelope spike protein, gp1601-3.
In animal models, bNAbs show potent prophylactic activity, suppress established viremia, and delay viral rebound during ATI4-8. In humans, a phase I clinical trial showed that 3BNC117 is generally safe and effective in transiently reducing viremia in chronically HIV-1 infected individuals9. A single infusion of 3BNC117 was well tolerated, rapidly decreased viral loads in viremic individuals by an average of 1.48 log10 copies/ml, with durable activity for 4 weeks9. In addition, 3BNC117 increased autologous antibody responses in HIV-1 infected individuals, and enhanced clearance of infected cells in humans and in humanized mice10 11. VRC01, a less potent bNAb that also targets the CD4-binding site, suppressed viremia by 1.14 log10 (12,13 and Fig. 1a and b).
To investigate whether 3BNC117 can suppress viral rebound from the latent reservoir during ATI in chronically suppressed HIV-1 infected humans, we conducted a phase IIa open label clinical trial. To select participants with 3BNC117-sensitive viruses in their latent reservoirs, we performed bulk viral outgrowth cultures of peripheral blood mononuclear cells (PBMC) from individuals whose viremia was suppressed by combination antiretroviral therapy (ART). The resulting isolates were screened for sensitivity to 3BNC117 using the TZM-bl assay (Supplementary Table 1). Of 63 individuals screened, only 11% yielded viruses that were fully resistant to 3BNC117 (IC50 >20 μg/ml), and 65% were sensitive to 3BNC117 IC50 at concentrations below 2.0 μg/ml. In contrast only 29% were similarly sensitive to VRC01 (Fig. 1a and b, Extended Data Fig. 1 and Supplementary Table 1). We enrolled HIV-1 infected individuals who were on suppressive ART with plasma viral loads < 50 HIV-1 RNA copies per ml for at least 12 months, had CD4 counts > 500 cells/mm3, yielded 3BNC117-sensitive outgrowth viruses (IC50 ≤ 2.0 μg/ml), and whose viral load at screen was < 20 copies/ml (Extended Data Fig. 1, Supplementary Tables 2 and 4, and Methods). Participants were enrolled in two groups: 8 in group A to receive two 30 mg/kg infusions three weeks apart, while 7 in group B received up to four 30 mg/kg infusions at two-week intervals (Fig. 1c and d, Supplementary Table 2). Two group A participants had viral loads > 20 copies/ml at the time of infusion and were excluded from further analysis (Supplementary Tables 2 and 4).
Analytical treatment interruption (ATI) was started 2 days after the first 3BNC117 infusion. ART was reinitiated and infusions were stopped after two consecutive plasma viral load measurements exceeded 200 copies/ml. All individuals on non-nucleoside reverse transcriptase inhibitors (NNRTI) were switched to an integrate inhibitor-based regimen (dolutegravir+tenoforvir disoproxil fumarate/emtricitabine) four weeks before ATI due to the long half-life of NNRTIs (Supplementary Table 2).
Both dosing regimens were generally well tolerated. The majority of reported adverse events were transient and grade 1 in severity (Supplementary Table 5). The mean CD4 T-cell count at baseline (day 0) was 747 cells/mm3, and the average change in CD4 T-cell counts between start of ATI and rebound was -127 cells/mm3. Although CD4-T cells declined modestly during viral rebound in some participants, CD4 T-cells returned to baseline by week 12 in most participants (mean 828 cells/mm3) (Extended Data Fig. 2 and Supplementary Table 4). 5 of 12 individuals tested showed measurable increases in the magnitude and/or breadth of T cell responses to HIV-1 12 weeks after ATI, compared to baseline (Extended Data Fig. 3). None of the participants experienced acute retroviral syndrome during rebound, and viremia was re-suppressed below 20 copies/ml in all participants within 2-7 weeks after restarting ART (Supplementary Table 4). We conclude that up to four 30 mg/kg infusions of 3BNC117 during ATI are generally safe and well tolerated.
By anti-idiotype ELISA9 the half-life of 3BNC117 during ATI was 19.6 days among group A participants, and 14.1 days among those in group B (Fig. 1e, f and Supplementary Table 4). These measurements are similar to previously reported values for 3BNC117 in HIV-1 infected individuals on ART9 (Fig. 1e).
All 6 group A participants maintained viral loads below 200 copies/ml during the first 4 weeks, with rebound 5-9 weeks after ART interruption (Fig. 2a and Supplementary Table 4a). In group B, rebound occurred 3-19 weeks after ATI, with 4 out of 7 (57%) participants remaining suppressed for at least 10 weeks (Fig. 2b and Supplementary Table 4b). The average time to rebound was 6.7 weeks in group A, 9.9 weeks in group B, and 8.4 weeks for all participants together, compared with 2.6 weeks for matched historical non-infused control individuals (Fig. 2c, Extended Data Fig. 4a, Supplementary Tables 4, 6 and 7). Altogether, 6 of the 13 infused individuals (46%) remained suppressed until at least 9 weeks after ATI. Relative to matched historical control individuals, the delay to rebound among all 3BNC117-infused participants was highly significant (p=<1e-5, weighted log-rank test, Fig. 2c, Extended Data Fig. 4, Supplementary Tables 4, 6 and 7 and Methods Statistical Analyses). We conclude that repeated infusions of 3BNC117 are generally safe and significantly delay HIV-1 rebound from the latent reservoir during ATI.
Time to viral rebound did not correlate with pre-ATI viral culture sensitivity to 3BNC117, nor to baseline levels of cell-associated HIV-1 DNA (Fig. 2d and Extended Data Fig. 4e). Therefore the significance of viral outgrowth sensitivity as an inclusion criterion is not clear. 3BNC117 levels at rebound were also variable, ranging from 6-168 μg/ml, but were directly correlated to the IC80 of the emerging virus (Fig. 2a, b and e)
To determine whether rebound was associated with resistance to 3BNC117, we compared pre-infusion and rebound viral outgrowth cultures. A majority (8/13) of participants had rebound viruses that were more resistant to 3BNC117 (IC80 > 3-fold higher, Fig. 3a and c, Extended Data Fig. 5a, Supplementary Table 3). Among group A participants, all but 707 had more resistant rebound viruses; however, among group B participants, 4/7 (710, 711, 712 and 715) showed similar pre infusion and rebound sensitivity to 3BNC117 (Fig. 3a and c, Extended
Data Fig. 5a, Supplementary Table 3). Among these 5 individuals, 711 was the earliest to rebound at 3 weeks, despite having viruses that were surprisingly sensitive to 3BNC117 as measured by IC50 (Fig. 2b, Extended Data Fig. 5a, Supplementary Tables 3 and 4). However, 100% neutralization was not achieved against 711 rebound or pre-infusion viruses, even at high (50 μg/ml) antibody concentrations, suggesting that 3BNC117 was not fully therapeutic (Extended Data Fig. 5, Supplementary Table 3). Thus, the only participant in the study to rebound within 3 weeks of ATI may have done so because of preexisting resistance to 3BNC117 by the dominant virus in the reservoir.
The other 4 participants that showed no change between pre and post-infusion culture sensitivity to 3BNC117, 707, 710, 712, and 715, all rebounded relatively late at 9, 19, 16 and 11 weeks after ATI, respectively (Fig. 2a and b, Fig. 3a and c, Extended Data Fig. 5, Supplementary Tables 3 and 4). In all of these individuals rebound was associated with relatively low antibody concentrations ranging from 6-41 μg/ml (mean 19.7 μg/ml) . This antibody concentration represents 9.6 fold the mean IC80 for the rebounding viruses which is consistent with previous reports on the relationship between suppressive 3BNC117 concentration and neutralization titer in macaques14 (Fig. 2, Extended Data Fig. 5, Supplementary Tables 3 and 4).
To determine whether viral rebound during ATI was associated with resistance to other bNAbs undergoing clinical testing, we examined sensitivity to 10-107415, which targets a different and non-overlapping epitope on the HIV-1 trimer (Fig. 3b and d, Extended Data Fig. 5, and Supplementary Table 3). With the exception of 703 and 711, the participants' rebound cultures did not show increased resistance to 10-1074. We conclude that rebound during ATI in the presence of 3BNC117 is infrequently associated with increased resistance to 10-1074.
To further characterize viruses emerging from the latent reservoir, we performed single genome sequencing (SGS) of viral RNA from the plasma and viral outgrowth cultures from 8 individuals. Phylogenetic analysis of these sequences indicated that all trial participants were infected with epidemiologically unrelated viruses (Extended Data Fig. 6). Given the limited sampling of the pre-infusion reservoir, rebound viruses did not always fall within the radiation of pre infusion viral isolates (Fig. 3e and f, Extended Data Figs. 7 and 8, Supplementary Figs. 1 and 2).
Remarkably, in 5 of 8 participants, all rebounding virus sequences clustered within a low diversity lineage, consistent with the clonal expansion of a single recrudescent virus (Fig. 3e and f, Extended Data Figs. 7 and 8, Supplementary Table 8). These data contrast with individuals undergoing ATI in the absence of antibody infusion, where virus rebound is consistently polyclonal, indicating the activation of multiple latently infected cells16-19. Thus, in addition to delaying rebound, 3BNC117 appears to restrict the outgrowth of viral genotypes from the latent reservoir.
Six of eight participants sequenced had rebound viral outgrowth culture and/or plasma sequences that indicated 3BNC117 resistance. For example, in 704, all rebound viruses carried a serine at position 456 (Fig. 4a and Supplementary Figs. 1 and 2), which may disrupt a highly conserved salt bridge that maintains the V5 loop's position and conformation20-22. Similarly, in 708 and 709, nearly all rebound viruses carried atypical residues at position 282, where a lysine residue typically forms a salt bridge with 3BNC117 (Fig. 4a and Supplementary Figs. 1 and 2)23. However, documented 3BNC117 resistance mutations24 were not universally identified among rebound viral strains (Fig. 4a and Supplementary Figs. 1 and 2). Only a minor fraction (3 of 23) of sequences in the rebound population of participant 701 had potential resistance-conferring residues in Loop D (274F, 282R), while the remaining sequences did not (Fig. 4a and Supplementary Figs. 1 and 2). Similarly, in 702 and 703, only a subset of rebound viruses carried a putative resistance-conferring A281D change1,23. Nevertheless, the frequency of this change increased markedly with time in both participants, indicating continued selection for 3BNC117 resistance (Fig. 4a and Supplementary Figs. 1 and 2). For participants 707 and 711, no sequence features were identified that would indicate 3BNC117 resistance.
To determine the sensitivity of rebound viruses to 3BNC117, we performed TZM-bl neutralization assays using pseudo viruses typed with SGS Env genotypes (Fig. 4b, Extended Data Figs. 7 and 8, Supplementary Figs. 1 and 2 and Supplementary Table 9). With the exception of participant 707, who rebounded 9 weeks after ATI at very low 3BNC117 titers, Env genotypes at rebound were more resistant to 3BNC117 than pre-ATI (Fig. 4b, Extended Data Figs. 7 and 8 and Supplementary Table 9). We conclude that viral rebound during ATI in the presence of 3BNC117 selects for the emergence of resistant variants, indicating strong selection pressure by this antibody on viral populations arising from the reservoir.
Antibody potency and half-life are directly correlated with HIV-1 prophylaxis in pre-clinical models. For example, VRC01, a CD4bs antibody that is less potent than 3BNC1171, is less effective than 3BNC117 in preventing SHIVAD8 infection in macaques8,25. Consistent with these observations, clinical trials with combinations of 3 less potent first generation bNAbs showed limited effects on viral rebound in the setting of ATI in chronically-infected individuals26,27. In addition, selective pressure as evidenced by escape mutations was only observed for 1 of the 3 antibodies used in the combination, 2G1226,27. In contrast, 3BNC117 alone significantly delayed viral rebound with nearly half of all individuals remaining below 200 copies/ml until at least 9 weeks, including 4 individuals who failed to develop resistance and only rebounded at low antibody concentrations. We speculate that the difference in efficacy between 3BNC117 and less potent bNAbs in the setting of ATI is due to increased potency and/or a longer half-life 1,9,13.
Nevertheless, the majority of the individuals we studied rebounded at high 3BNC117 serum concentrations. A single viral genotype displaying increased resistance to 3BNC117 established rebound in most cases. These viruses represent pre-existing dormant variants that emerged from the latent reservoir. The time to rebound did not correlate with the amount of viral DNA in circulating PBMCs; however, this is a poor measure of the HIV-1 reservoir, since most integrated proviruses in patients on ART are defective28. Instead, the delay in viral rebound may represent a measure of the frequency of 3BNC117- resistant variants in the latent reservoir.
Combinations of drugs are needed to maintain HIV-1 suppression in effective ART regimens. Similarly, combinations of antibodies were required to suppress viremia in humanized mice 6,7. We speculate that combinations of bNAbs will also be needed to increase the frequency of individuals that remain suppressed by antibody during ATI. Whether 3BNC117 can also impact the size and composition of the latent reservoir during ATI will require additional studies.