CYBERSPACE – Continuing improvements in HIV treatment and progress toward a cure notwithstanding, an effective vaccine will be necessary to gain the upper hand in the decades-long fight against the pandemic. The final day of the all-virtual 2020 Conference on Retroviruses and Opportunistic Infections (CROI) included both a look back on the failed HVTN702/Uhambo trial, and a plenary talk on how future vaccination efforts might do better.
The Uhambo trial, a phase IIb/III trial of a prime-boost vaccine that combined ALVAC-HIV from Sanofi Pasteur and an adjuvanted, two-component gp120 protein subunit vaccine from Glaxosmithkline plc, was built on the lessons of RV144 “Thai” trial and adapted to its African location.
Those adaptations, when they were tested, resulted in a vaccine that met all prespecified antibody and T-cell responses, showed stronger gp120 binding, and was more immunogenic than the vaccine and vaccination strategy used in the RV144 trial, HIV vaccine trials network (HVTN) principal investigator Larry Corey told the audience.
But despite all that, the vaccine showed no evidence of efficacy at all, and Uhambo was halted at its first interim analysis, only roughly six months after it had finished enrolling. The Kaplan-Meier curves for the vaccine and placebo groups “never separated,” said trial lead Glenda Gray, who is the president and CEO of the South African Medical Research Council.
Corey listed a number of possibilities for why the vaccine turned out to be a dud despite its seeming promise.
Though the vaccine looked immunogenic by the criteria used in the trials, “there might be unknown immune correlates of protection,” he said. Both viral and host genetics also differ from those of the RV144 trial.
Viral genetics in southern Africa are “vastly more diverse” than those in Thailand, especially in Thailand a decade ago, while the immune alleles FCGR2C and HLA*02, both of which were shown to be associated with protection during the RV144 trials, are less frequent in African populations.
Finally, Corey said, exposure to HIV is “more frequent and possibly stronger” in the conditions that predominate in southern Africa, where a much greater percentage of the population is infected. “The challenge exposure that women in South Africa experience may require a much higher level of immunogenicity.”
Session co-chair Galit Alter, group leader at the Ragon Institute, said that “the bottom line is that we must not be discouraged but remain firm in our commitment… We have much to learn from 702 and we remain optimistic.”
“The pipeline is rich,” she said. “We have many products in the queue.”
Those products include three others that are in advanced trials. The Imbokodo (HVTN705) and Mosaico (HVTN706) trials are testing vaccines, while the MAP (HVTN703/704) trial is looking at the ability of passive vaccination to provide lasting protection.
Meanwhile, back in the lab…
Focusing on the early clinical and preclinical parts of the HIV vaccine pipeline, Shane Crotty, a professor at the La Jolla Institute of Allergy and Immunology, gave an overview of strategies that could be useful to strengthen the immune response to antigens.
There are 28 licensed vaccines for humans, “and almost certainly, 26 of those work on the basis of neutralizing antibodies,” Crotty said. And while that “definitely does not mean it’s the only possible way to make a vaccine… it’s been the approach that’s worked most often in the past.”
Crotty took his audience through a mix of methods for development and vaccine design studies aimed at meeting the multiple challenges of HIV vaccine design with an equally multipronged approach – starting with an issue that is not limited to HIV vaccines.
For a long time, Crotty said, vaccine studies have been “quite frustrating,” because “to some extent immunologically they are a black box problem.”
The reason is that B-cell development happens in germinal centers in the lymph nodes, but the B cells used to study vaccines are extracted from blood samples, meaning that “you are guessing or inferring from blood what might be happening on the lymph nodes,” Crotty said.
In 2016, researchers developed a way to take samples directly from lymph nodes of nonhuman primates, which allowed a much better assessment of whether there are responses to vaccine candidates.
Using direct lymph node sampling, as well as humanized mouse models where they could precisely control the frequency of VRC-01 precursor antibodies, Crotty’s team and those of frequent collaborators William Schief and Darrel Irvine were able to show that antibodies that produce precursors of the bnAb VRC-01 were able to elicit immune responses, but those immune responses were outcompeted by those from antibodies to other targets, due both to the rarity of the VRC-01 precursor cells and the fact that the binding affinity is weaker than that of other antigens.
“It’s not that the antibody response you want is impossible, but that those B cells got outcompeted,” Crotty said.
The competitive advantage of “easier” but ultimately non-neutralizing antibodies “can go to the point of excluding neutralizing antibodies, if you don’t deal with this problem,” he said. “The good news is, you can deal with this problem… even rare B cells, if they had a high enough affinity, could compete quite well over time,” including to a candidate antigen aimed at eliciting VRC-01 that is now in a phase I trial evaluating its safety and immunogenicity, the IAVI G001 trial.
Crotty and his team have also been looking for methods to improve B-cell responses to antigens. In a separate study, published in Cell in 2019, they showed that compared to a bolus injection, slow delivery of antigen could lead to an up to 20-fold stronger neutralizing antibody response and altered the dynamics of immunodominance to favor responses by less frequent precursor cells.
In their experiments, the team used osmotic pumps to deliver the antigens over a period of time, a delivery method that would be feasible for small trials but not for large-scale trials, let alone rollout of a vaccine. But, the authors noted in their paper, “less cumbersome slow delivery immunization technologies are worthy of further development, including degradable encapsulating biomaterials and depot-forming adjuvants that make antigen available over time … Such technologies may be able to rescue protective immune responses to antigens that have previously failed by conventional bolus immunization.”