Three decades of advances in HIV treatment and prevention have curbed mortality from the HIV epidemic, but every year since 2010 2 million more people around the world have been infected with HIV. This is a sad reminder that we still have not put the brakes on this clever viral menace.

Millions of lives and the hopes of an entire generation thus hang on the development of an effective HIV vaccine. Those hopes were rekindled in November 2016 with the opening of a major new HIV vaccine trial in South Africa. Called HVTN 702, the trial will enroll 5,400 adults to receive five injections of an investigational HIV vaccine or a matching placebo.

This latest step in a global effort to develop an effective HIV vaccine builds on a legacy of failure, hubris, and reinvention. Each stage of this history has much to teach us about the development of preventive vaccines in general.

Failure

The legacy of HIV vaccine failure kicked off in earnest in 2003 when disappointed researchers announced the failure of the phase III trial of AIDSVAX, that day’s leading HIV vaccine candidate. In over 5,000 high-risk subjects in the United States and other western countries, the risk of HIV infection was nearly identical among vaccine and placebo recipients.

Undaunted, HIV vaccine scientists confronted this news with admirable perseverance, and scientific creativity. They hypothesized that an effective HIV vaccine required not the antibody responses elicited by AIDSVAX (and other existing vaccines) but instead the T cell immune responses observed in patients with milder cases of HIV disease. Promising preclinical and epidemiological findings ensued, and soon a new fleet of vaccine candidates that elicited bigger T cell immune responses advanced into human clinical trials.

This period of fecund investigation, and the renewed hopes stoked by it, should inspire present-day investigators developing vaccines against other infectious scourges not to give up either.

Hubris

In the buildup to new phase III clinical trials of this second wave of HIV vaccines, many HIV vaccine scientists transitioned from reasoned confidence in the importance of T cell immune responses to messianic conviction that they knew how an effective HIV vaccine should work. This unfounded conviction led to an embarrassing moment from which vaccine researchers in many fields can learn today.

Editorializing in the journal Science, some of the world’s most prominent HIV vaccine researchers argued in 2004 that an investigational vaccine study called Rv144 was not only likely to fail but that funding it could lead to “a crucial erosion of confidence by the public and politicians in our capability of developing an effective AIDS vaccine collectively.” Why? Because AIDSVAX had already failed, plus the Rv144 regimen (a combination of AIDSVAX plus another untested vaccine) elicited T cell immune responses that the editorialists said were too weak to be effective. Other vaccine candidates, they wrote, were more likely to succeed because they generated greater magnitude T cell immune responses.

This editorial, which encapsulated the sentiments of many HIV immunologists, drew controversy. Fortunately, as we shall see, it did not derail the Rv144 study. But the preeminence of the belief that T cell immune responses were indispensable to vaccine efficacy went unchecked for at least a decade.

Hubris invites repudiation, and three years later it came when the results of the STEP study were announced in 2007.

STEP was a multinational phase III study in 3,000 subjects of an adenovirus-vectored HIV vaccine hyped as eliciting massive T cell immune responses. Expectations were intense, so researchers were stony-faced when they announced that the STEP trial had been halted early due to vaccine failure. While safe and highly immunogenic, the STEP vaccine was utterly ineffective in preventing HIV infection. Even worse, the vaccine may have increased the risk of HIV infection among selected participants.

This was a seminal moment. HIV clinicians knew they would have to tell HIV-infected and at-risk people around the world that an HIV vaccine was still decades away. Scientists watched as their cherished hypotheses were questioned and HIV vaccine funding priorities were reconsidered.

The STEP defeat teaches an important lesson about the dangers of dogma. HIV vaccinologists are not the only researchers who worship at the altar of T cell immune responses to vaccines. Tuberculosis vaccines, for instance, now are vetted in pre-clinical and early-phase vaccine studies for their ability to elicit T cell immune responses. That is how we have done it for years, and pre-clinical studies suggest such immune responses are important to vaccine efficacy. Yet, in humans vaccinated with the current tuberculosis vaccine, bacille Calmette Gurine, T cell immune responses do not correlate with vaccine efficacy. (And, in an eerie echo of the STEP trial, a tuberculosis vaccine candidate that elicited strong T cell responses recently failed in a major clinical trial.)

The pitfalls of HIV vaccine development remind us of the dangers of scientific dogmatism. Hopefully we will learn these lessons in global vaccine development efforts to come.

Reinvention

The failure of the STEP trial hurt, as did the 2013 failure of another T cell vaccine trial called HVTN 505. Those failures, however, did prompt a reconsideration of the scientific dogma in HIV vaccinology. Scientists rekindled their interest in other kinds of immune responses. Perhaps antibody responses still merited investigation after all.

Enter Rv144, the vaccine regimen disparaged in that 2004 Science editorial. Much to detractors’ surprise, investigators announced in 2009 that recipients of the Rv144 vaccine exhibited 31.5 percent protection from HIV infection. This wasn’t a raging success but it did suggest a path forward. Three years later, immunologists announced that certain antibody responses, while short lived, clearly predicted vaccine protection from HIV infection in the Rv144 trial. We dismissed antibody responses too soon, it turned out, but we were back on track.

The future

The HVTN 702 trial that opened in November is the next step in the reinvention of HIV vaccine development. The new regimen being investigated in HVTN 702 capitalizes on the Rv144 results by enhancing and prolonging its protective antibody responses.

This is a second chance for HIV vaccine development. We should approach this historic opportunity wisely, and with the proper scientific modesty. We are not sure to this day which immune responses define a protective HIV vaccine. Antibody responses are in vogue but T cell immune responses and other types of immunity too may be part of the secret sauce of an effective HIV vaccine. Instead of using unproven immunological hypotheses to prioritize vaccines, we should admit uncertainty, and we should let clinical trials pick the winner.

We do have some certainties in HIV vaccine science, though. The creativity of HIV vaccine scientists is inspiring. The trust put in their hands by the 5,400 South African research subjects now receiving the HVTN 702 vaccine regimen is also a welcome reminder that human solidarity will beat the HIV epidemic in time. Most importantly, the 2 million people who contract HIV infections every year are begging for progress. Today, as the first doses of the most promising HIV vaccine in history are given, that progress has come. In 2017, we should commit to extend that same spirit of creativity, solidarity, and scientific modesty to the entirety of our global vaccine development efforts.