The study took place in an Ebola treatment center in Nzérékoré, Guinea.

The study took place in an Ebola treatment center in Nzérékoré, Guinea.

Julien Rey/MSF

Many caveats on promising Ebola drug trial

SEATTLE, WASHINGTON—There’s a “slight hope” that it may help some people. That is the researchers’ sober assessment of an Ebola drug that made page one news earlier this month when it was touted as the first promising treatment for this deadly disease.

Those media reports relied on preliminary data on a trial of a new drug in Guinea that were leaked by a nongovernmental organization that was helping conduct the study, the researchers say. The New York Times broke the story on 4 February, and the same day, French President François Hollande issued a statement heralding the results as encouraging. “Healing is accelerated,” the statement read. Guinean officials announced they wanted to make the drug more widely available.

Yesterday was the first time the French team discussed their data publicly, speaking to the media at the Conference on Retroviruses and Opportunistic Infections (CROI) here.

In the opening press conference, the trial’s principal investigator, Denis Malvy of the University of Bordeaux, described the trial, called JIKI. Sixty-nine adults and adolescents took part in the study, from mid-December to mid-January. All participants received the drug, favipiravir, for up to 10 days. Their outcomes were compared with patients receiving standard care at the same clinics for the 3 months prior to the trial’s start. In all, 48% of the participants died, which is similar to the mortality rate in the historical control group. But when the researchers did a subset analysis of patients who entered the clinic with lower levels of Ebola virus in their blood, they found an increase in survival.

Malvy, who plans to describe the data in detail at a Wednesday morning presentation, explained that the trial did not initially intend to analyze the outcome of patients by different viral loads. But the standard test used to diagnose the disease gave this information, and the investigators saw an interesting pattern.

In 29 patients who arrived with high levels of virus, 27 died. “It is highly unlikely that favipiravir monotherapy will reduce mortality in this population,” Malvy said.

In the 40 patients who entered the trial with lower viral loads, only 15% died compared with a 30% mortality in the historical control group. “We have the initial impression that monotherapy with favipiravir could improve survival in the group with less advanced disease,” Malvy said, noting that the trial is ongoing. But Malvy also stressed that even when the trial is complete, it will not provide a clear answer about whether favipiravir worked, because, for ethical reasons, the researchers did not enroll an untreated control group at the same time. He called this  a “proof of concept” study that the researchers hope will inform future trials.

The improved mortality in the group may have not been related to favipiravir, Malvy said. For one thing, the standard of care in the clinics may have improved during the time of the study, making the comparison to the historical control misleading. The researchers also relied on a relatively crude measure of the two different viral loads. Other studies that have more carefully analyzed Ebola viral loads have found that at least four different levels correlate with different mortality rates, which suggests the supposed benefit of favipiravir seen might disappear if more precise measures were compared.

Xavier Anglaret, a colleague of Malvy’s at the University of Bordeaux who coordinates the JIKI study, said their data “are not a high level of evidence” and that the group’s aim was modest from the outset. “It’s only trying to find a signal,” Anglaret said.

Anglaret said the researchers had submitted an abstract to the CROI meeting prior to the New York Times story appearing. “We were all very angry” about the leak, he said. “We wanted to share the overall story ourselves.”

*The Ebola Files: Given the current Ebola outbreak, unprecedented in terms of number of people killed and rapid geographic spread, Science and Science Translational Medicinehave made a collection of research and news articles on the viral disease freely available to researchers and the general public.