Even the researchers whose trial of a potential drug for Ebola made headlines last week worked hard to downplay the glimmer of efficacy it showed. “It is a weak signal in a nonrandomized trial,” Yves Levy, director of the French Institute of Health and Medical Research (INSERM) in Paris told Science about the data, which INSERM has not released. Weak or not, the report in The New York Times that favipiravir, a Japanese flu drug, had halved mortality in one group of Ebola patients in Guinea was one more piece of good news that is complicating prospects for trials of other Ebola drugs.
The Guinean government has already announced it wants to make favipiravir available to more people, and if the results hold up to greater scrutiny, they could force a change in the design of other clinical trials going forward. Meanwhile, the decline in new cases has investigators revamping or even canceling trials at a time when manufacturers finally have enough supplies to test some of the most promising experimental drugs. The toll of the outbreak ticked up last week, as Guinea, Liberia, and Sierra Leone—the three most affected countries—counted 124 confirmed cases, up from 99 cases the week before. As the World Health Organization’s (WHO’s) Bruce Aylward said at a press conference on 5 February: “The virus has told us this week, loud and clear, ‘I am not going to go away the way you’re expecting me to.’ ” Yet the numbers represent a sharp drop from the height of the epidemic in September when there were more than 700 cases reported in a single week in West Africa.
Just last week, the Wellcome Trust, a charity in the United Kingdom that is funding several Ebola trials, announced that the Liberian trial of brincidofovir, an antiviral developed by Chimerix of Durham, North Carolina, would be canceled because the company withdrew support. “It was rather a surprise to us and a bit of a mystery,” says Peter Horby, an investigator at the University of Oxford in the United Kingdom who headed the study. Chimerix said it made the decision after discussions with the U.S. Food and Drug Administration (FDA), noting that the trial was also having trouble recruiting patients. Horby says his group was planning to open a second trial site in Sierra Leone, where the numbers are far higher. FDA’s Luciana Borio says Chimerix also refused the agency's request to make public its correspondence with the company. Chimerix said it was concentrating on completing trials of the drug to treat other infections: cytomegalovirus and adenovirus.
Although one trial is canceled, others are about to go forward. Horby says his group hopes to start evaluating an RNA inhibitor called TKM-Ebola soon. The drug, made by Tekmira Pharmaceuticals of Burnaby, Canada, worked well in monkeys but has been in short supply.
Testing is also about to begin on the antibody cocktail ZMapp. Seen by many researchers as the best shot at treating Ebola because of promising monkey studies, ZMapp was used on nine patients last summer before the company behind it, Mapp Biopharmaceutical of San Diego, California, announced it had no more supplies. Now, the company says, it has enough doses to start a clinical trial in Liberia as early as this week. But there may be too few patients in that country for the experimental drug to prove its worth, says Clifford Lane, head of clinical research at the U.S. National Institute of Allergy and Infectious Diseases, which is launching the study in Monrovia with the Liberian Ministry of Health & Social Welfare.
So far, Guinea and Sierra Leone, where Ebola is still infecting dozens of people a week, have refused invitations to join the study. Their main stumbling block is trial design. ZMapp will be the first Ebola treatment that will be tested in a randomized, controlled study. “I think that’s the only way to tell whether these drugs are safe and effective,” Lane says.
The governments of Guinea and Sierra Leone, as well as Doctors Without Borders, which runs Ebola centers in those countries, have for ethical reasons been reluctant to participate in treatment trials that use a randomized, controlled design. Jeremy Farrar, head of the Wellcome Trust, also objects to the randomized, controlled trial design for Ebola drugs, given the high mortality rate of the disease. “Given the data we have from animals and individual patients, I would not feel comfortable being randomized,” he says.
Lane notes that the trial may not need many participants: If the drug is 100% effective and Ebola kills 50% of the people it infects, he says, as few as 30 people will need to receive ZMapp to determine whether it works. And even if there are not enough patients to provide a clear answer on efficacy, Lane says scientists might still get valuable data by looking at parameters like the blood levels of Ebola virus in those treated with the drug and those in the control arm.
The favipiravir study in Guinea illustrates the complexity of discerning clear answers without a robust control and the difficulties of communicating them. The trial data are reviewed every 20 patients by an independent monitoring board. On 26 January they evaluated the data from 80 patients. Because they detected a signal of efficacy, they asked the researchers to share the information with regulatory agencies in Guinea and France, Levy says. The INSERM researchers won’t make their data public until 25 February, at the Conference on Retroviruses and Opportunistic Infections in Seattle, Washington. “It is important to have a scientific debate about what these results really mean,” says Levy, noting that the meeting organizers insisted the data be embargoed. A researcher who had seen the data and asked not to be identified told Science that favipiravir did not help all of the patients treated with it at two trial sites in Guinea. In a subset of trial participants who had low levels of Ebola virus in the blood, however, the mortality was just 15%. In similar patients who entered the centers earlier and did not receive favipiravir, mortality was 30%. Marie-Paule Kieny, an assistant director-general at WHO, says it is difficult to make sense of the data at this point. “You can say it doesn’t mean anything or you can say it is promising. More research is needed to find out what really happened.”
Meanwhile, the study in Guinea is continuing and has now enrolled more than 100 patients. “The final result may still be different,” Levy says. But the preliminary data have already led Guinean authorities to expand the numbers of sites where favipiravir is to be used.
Other trials could prove harder to organize and interpret if favipiravir is distributed widely. A study testing the use of convalescent serum started in Guinea this week. “If there is a decision now to use favipiravir everywhere, what happens with that trial?” Kieny asks. The ZMapp trial may also be affected. That trial is designed to compare ZMapp with the standard of care. “If the standard of care changes, so does the control used in the trial,” Lane says. But he has not seen any results, he says. “The only data I have seen from that study are what was in The New York Times.”
*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.
*Correction, 13 February, 11:05 a.m.: This story incorrectly referred to a trial of ZMapp, an Ebola antibody cocktail, as being a placebo-controlled trial. As the story reports, ZMapp will be tested in a trial that, for the first time, uses a randomized, control group. But the control group will not receive a placebo. People in the control arm will receive the current standard of care, which includes providing intravenous fluids, balancing electrolytes, maintaining oxygen status and blood pressure, and treating other infections if they occur.