A series of animal and human studies published this week are providing a sad footnote to a controversial 2016 experiment in monkeys that pointed to a possible cure for HIV but was later flagged for a glaring omission in its design. The animal studies failed to replicate the promising results of the original paper, while a clinical trial in humans using a similar strategy largely failed as well.
In 2016, researchers published a study in the well-respected journal Science. They experimented with rhesus macaque monkeys, infecting them with the simian immunodeficiency virus (SIV), a close relative of the HIV virus that causes AIDS in people. Then they gave the monkeys a short combination dose of the standard HIV treatment – antiretroviral therapy (ART) – and an antibody that targeted a protein, called a4b7, found in immune cells.
Current treatments for HIV can drive levels of the virus so low as to make a person completely not contagious and relatively healthy. But HIV seeds itself into reservoirs of the immune system, where some amount of it is able to survive these drugs. If someone stops taking their daily dose of ART or the treatment otherwise starts to fail, the virus can then emerge from its dormancy and wreak havoc again. Earlier research had suggested HIV uses a4b7 to help pull off its disappearing trick inside cells, so it was hoped that using both treatments at once could eradicate the virus completely, or at least provide a sustained remission that would let people no longer need daily ART.
And in the 2016 study, that’s what it seemed to do for the monkeys. Nine months after their treatment, their immune systems looked perfectly normal and no traces of SIV could be found. It was understandably seen as the first step to an HIV cure.
In science, though, no single study is the final word. Researchers at the National Institute of Allergy and Infectious Diseases (NIAID) and elsewhere quickly began to carry out their own larger studies that would replicate the original study’s design. Another U.S. government study was planned to test out the human version of the a4b7 antibody – an existing anti-inflammatory drug called vedolizumab – in a small clinical trial with people.
“We conducted these studies because we felt that the results reported in the 2016 Byrareddy study had the potential to transform the HIV cure field if the findings were reproducible and generalizable,” Dan Barouch, director of the Center for Virology and Vaccine Research at Beth Israel Deaconess Medical Center at Harvard Medical and senior author of one of the studies, told Gizmodo via email.
As all this was going on, however, it came to light that one of the study’s co-authors, Francois Villinger, now a pathologist at the University of Louisiana at Lafayette, had provided the research team a different type of SIV virus than was stated in the paper. To spare the technical details here, this version of the virus wasn’t programmed to produce a protein made by SIV in the wild, at least not for the first weeks of infection. According to the authors, only Villinger was aware of the omission. And he told the journal’s editors that he used the modified virus because it more closely mimicked how HIV behaves in humans (wild SIV tends to kill monkeys very quickly, while HIV infection is usually slower to kill).
The lack of disclosure was enough for the journal to issue an editorial expression of concern about the study this March. But importantly, they didn’t go as far to say that the science itself was tainted or that the study needed to be retracted.
“While there was no malfeasance, there was clearly a major negligence in our studies reported in the Science paper,” Ansari Aftab, senior author of the 2016 study, told Gizmodo via email.
That brings us to this week’s studies, published in Science (or in the case of the clinical trial, in Science Translational Medicine). Across all three animal studies, which replicated many but not all aspects of the 2016 paper and included a larger sample of animals, the combination treatment failed to stave off SIV once the drugs were stopped. In 19 HIV-positive people given six months of vedolizumab along with ART, the virus started coming back to life within weeks of stopping ART for most of the subjects.
Two people in the trial did seem to have low viral levels even without ART. But a slim minority (around 4 percent) of patients taking ART as usual can experience remission once off their medication, so it’s not necessarily evidence the treatment did anything special. One of these patients eventually chose to go back on ART even though his HIV levels stayed low, while the other was forced to restart ART 46 weeks later once his levels began climbing back up.
Aftab and his colleagues say they stand by their results. And they point out that there are important differences between their study and the replication trials. Only two studies used the same type of SIV, for instance. And the monkeys weren’t always given the same treatment in the same way, nor is it clear whether they all originated from India (as opposed to China), as the monkeys in the 2016 study did – a change that could affect the results, as genetics can influence how diseases present and how well treatments work. It’s also possible, Aftab admitted, that the type of virus used in their study could have affected their results. As for the human trial, most of the volunteers already had a chronic HIV infection by the time they started ART and then the antibody, while the monkeys were all treated when they had an acute infection.
These factors and more could explain why the original study succeeded where the other trials didn’t. And even if that’s not the case, it doesn’t necessarily mean that targeting a4b7 isn’t a worthwhile strategy. But it’s also very much possible, if not likely, that the original study was a fluke.
In its updated expression of concern, Science said as much, stating that the evidence right now suggests the original study is “not robust and therefore does not provide a good basis for guiding work on therapies for HIV.”
Of course, it wouldn’t be the first time that a promising HIV cure has floundered under closer scrutiny. But as disappointing as all this is, that’s also just how research works. Sometimes the failures are as spectacular as the successes, but it’s the same story nonetheless – one step forward, three steps back, repeat for a decade or more, until something sticks. Aftab and his colleagues, for their part, plan to continue studying ways to finally bring down HIV for good.