During the darkest months of the West African Ebola epidemic in 2014, scientists were almost empty-handed. Hundreds of patients were dying every week; drugs and vaccines weren't ready for testing. There was another strategy they could try, however: taking blood from people who had already survived Ebola and giving it to those who were sick. Blood from survivors is laden with antibodies against the virus that might just help new victims overcome Ebola as well. But a new study suggests the approach doesn’t work, and some scientists say it's time to bury the idea and move on.
Passive immunotherapy, as the strategy is called, has proven its merit with other diseases, including influenza and diphtheria, but whether it worked with Ebola was unknown. A study during the 1995 Ebola outbreak in Kikwit, in what is now the Democratic Republic of the Congo, showed that seven out of eight patients who received blood from survivors survived, but a later analysis showed that most of them received the transfusions late in their disease, when they were unlikely to die anyway. A 2007 study in which monkeys received convalescent blood showed no benefit. "Given these discouraging results and the risks of transmitting infection, whole-blood transfusions, even under desperate epidemic conditions, seem unwarranted," the authors of that paper warned.
When Ebola exploded in West Africa in 2014, and patients were dying in large numbers, the World Health Organization said it was worth trying anyway—especially because candidate drugs were in short supply.
The new study, led by researchers at the Institute of Tropical Medicine in Antwerp, Belgium, and Guinea's National Blood Transfusion Center, didn't use blood but plasma—essentially blood minus the red blood cells. (The cells are given back to the donors.) The logistics were daunting: A huge blue bus chockful of equipment to obtain and process blood was shipped in from abroad. Survivors were mobilized and convinced to donate plasma—not an easy task in West Africa, where blood is a symbol of strength.
The researchers hoped to recruit 130 patients; because the epidemic was already on the retreat when their study started, they managed to administer plasma to only 99 people, 15 of whom were excluded from the analysis for various reasons. (Guinea had its last Ebola case in October.) Because withholding a potential treatment was unacceptable to regulators and the local population, there was no placebo group, says lead author Johan van Griensven; instead, the control group was formed by 418 patients treated at the same center in the five previous months.
Of the 84 patients eventually included in the plasma group, 31% died, compared with 38% in the control group—a difference of 7%. When the researchers adjusted the data to correct for patient age and virus levels in their blood, the difference shrunk to 3%, and it was no longer statistically significant, the team reports today in The New England Journal of Medicine (NEJM).
The researchers haven't entirely given up on the idea that plasma might work, however. They didn't determine the antibody levels in the plasma that they administered; those measurements are now being done at a high-level biosafety lab in France. It's still possible that plasma from some donors with high levels of antibody—or particularly powerful antibodies—was effective, Van Griensven says. "This is not the end of the convalescent plasma story," says Calum Semple, a clinical virologist at the University of Liverpool in the United Kingdom and a co-author on the study. (The fact that administering plasma was possible in the midst of an outbreak, and that the treatment was safe and acceptable to donors and patients, is an important feat in itself, he says.)
The trial's disappointing outcome is "not very surprising," says virologist Thomas Geisbert of the University of Texas Medical Center in Galveston, one of the authors of the 2007 monkey study. He says doing the study didn't make much sense in the first place, and now it's time to let go of the idea altogether.
Geisbert’s hopes are on treatments that did do well in animal studies, including a lab-made cocktail of antibodies called ZMapp that protected monkeys from Ebola. A field trial of ZMapp that has enrolled around 70 patients in Liberia, Guinea, Sierra Leone, and the United States is ongoing; whether that cohort is large enough to produce definitive results is still unclear.
A second study published in NEJM today provides hints, but no solid evidence, about another strategy that might work against Ebola. The paper is based on a natural experiment: In August 2014, an Ebola treatment center in the Foya, in northern Liberia, ran out of the first-line malaria therapy, a combination called arthemeter-lumefantrin, that was given to all Ebola patients admitted to the center. For 12 days, until new supplies arrived, the center relied on another drug combo named artesunate-amodiaquine.
As it happens, a screening published in 2013 showed that amodiaquine has anti-Ebola activity in the test tube. So researchers set out to see whether patients on the new malaria drugs had a higher chance of survival. They did: In the NEJM paper, the team shows that 51% of the patients died during the 12-day interval, compared to 64% of those in the periods before and after the stock-out.
It's not clear that amodiaquine saved patients from Ebola, however, the researchers write. One possibility is that the drug doesn't do anything, but that arthemeter-lumefrantrine increases the risk of death instead. It's also possible that the patients in the 12-day interval somehow differed from those who came earlier or later. Still, if malaria treatment is a standard component of Ebola care, artesenate-amodiaquine may be the better choice, the team writes. "It's definitely an interesting finding," says Robert Garry of Tulane University in New Orleans, Louisiana—but more work is needed on the anti-Ebola effects of amodiaquine, he says.