Slender hopes unravel after great effort and expense, emphasizing the importance of rigorous analyses and interpreting borderline results with caution

New York City, February 3, 2020 – Treatment Action Group (TAG) acknowledges with profound disappointment today’s announcement that the HIV Vaccine Trials Network HVTN 702 study did not demonstrate efficacy in preventing HIV acquisition (https://www.nih.gov/news-events/news-releases/experimental-hiv-vaccine-regimen-ineffective-preventing-hiv).[1] Importantly, the HIV vaccines were safe and there was no evidence of any harm associated with the immunizations.

The trial represented the culmination of a long, good faith effort to build on the marginal evidence of HIV vaccine efficacy that emerged in 2009 from RV144, a trial conducted in Thailand by the Thai Ministry of Public Health and the US Military HIV Research Program.

Today’s news may represent a sobering reminder that results of borderline statistical significance need to be interpreted with great caution. Teetering slightly over the edge of the statistical threshold of significance does not represent definitive proof that a result is real and reproducible. Alternatively, it is possible that an approach that worked marginally in the Thai setting did not translate to South Africa, where HVTN 702 took place. This underlines the rationale for doing HVTN 702, which sought to reproduce the RV144 results in a different population using a vaccine combination more tailored to South African HIV strains and with an alternative adjuvant.

RV144 recruited 16,402 participants who were largely at a low risk of acquiring HIV via heterosexual sex. The results offered evidence of a 31.2% reduction in the incidence of HIV infection associated with receipt of a prime-boost HIV vaccine regimen consisting of ALVAC-HIV vCP1521 and AIDSVAX B/E.[2] The significance of the result was borderline, with a 95% confidence interval (CI) ranging from 1.1% to 52.1% (if the lower bound of the CI had been less than 1 the result would not have been considered statistically significant). In fact, two other approaches to analyzing the findings—the intention-to-treat analysis and the per-protocol analysis—did not produce evidence of significant efficacy in reducing HIV acquisition risk.

Much was made of an unplanned (“post hoc”) analysis of RV144 indicating that efficacy was higher (~60%) at an early timepoint of the trial, after one year.[3] Such post hoc analyses can be important, but must be treated very cautiously as they can look for and assess responses at time-points not specified in the protocol.

When the RV144 results were announced with great fanfare and suggested to represent the first evidence of “modest” HIV vaccine efficacy, TAG cautioned that “marginal” was a more appropriate term.[4] But the HIV vaccine field had no option other than to attempt to verify if the findings could be confirmed or perhaps even improved—the need for an HIV vaccine is great, and at the time there was a lack of other strategies considered likely to show efficacy.

The people who performed the massive amount of work required to initiate HVTN 702 and the many individuals who participated are to be saluted. The results, while not what anyone hoped for, are still important, and there are likely many lessons to be drawn from the data that will be generated. There were also critically important steps taken to bolster the ethical conduct of biomedical prevention trials, including dialogues with community and arrangements to provide pre-exposure prophylaxis (PrEP) to participants as needed—this work will have lasting benefits.

While potentially uncomfortable to confront, it is also important to look back at how the trial came about and ensure that any lessons can also be drawn from how RV144 was viewed and interpreted. A considerable amount of research has been conducted to evaluate possible immune “correlates of protection” in RV144, and there may now need to be reflection as to whether protection occurred at all.

We look forward to ongoing and planned analyses of the results of HVTN 702 and comparisons of the immunologic, virologic, and clinical outcomes of the two trials.

# # #

About TAG: Treatment Action Group (TAG) is an independent, activist and community-based research and policy think tank fighting for better treatment, prevention, a vaccine, and a cure for HIV, tuberculosis, and hepatitis C virus. TAG works to ensure that all people with HIV, TB, and HCV receive lifesaving treatment, care, and information. We are science-based treatment activists working to expand and accelerate vital research and effective community engagement with research and policy institutions.

[1] National Institutes of Health. News Release. Experimental HIV vaccine regimen ineffective in preventing HIV. No safety concerns found; NIH and partners discontinue vaccinations. February 3, 2020. https://www.nih.gov/news-events/news-releases/experimental-hiv-vaccine-regimen-ineffective-preventing-hiv.

[2] Rerks-Ngarm S, Pitisuttithum P, Nitayaphan S, et al. Vaccination with ALVAC and AIDSVAX to prevent HIV-1 infection in Thailand. N Engl J Med. 2009 Dec 3;361(23):2209-20. doi: 10.1056/NEJMoa0908492. Epub 2009 Oct 20. https://www.nejm.org/doi/10.1056/NEJMoa0908492

[3] Robb ML, Rerks-Ngarm S, Nitayaphan S, et al. Risk behaviour and time as covariates for efficacy of the HIV vaccine regimen ALVAC-HIV (vCP1521) and AIDSVAX B/E: a post-hoc analysis of the Thai phase 3 efficacy trial RV 144. Lancet Infect Dis 2012; 12:531–7. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3530398/

[4] Treatment Action Group. Cause for Caution on the Results of the ALVAC/AIDSVAX HIV Vaccine Efficacy Trial: More Marginal Than “Modest.” 2009 September 25. https://www.thebodypro.com/article/cause-caution-results-alvac-aidsvax-hiv-vaccine-efficacy-trial-ma