Merck’s Ebola vaccine offered substantial protection to people vaccinated during the 2018-2020 outbreak in the Democratic Republic of the Congo, with an effectiveness of 84% in those who had been vaccinated at least 10 days before being exposed to the virus, a new study reports. An earlier study had shown that people who were vaccinated but went on to develop the disease were more likely to survive the infection.
The analysis, published Tuesday in the journal Lancet Infectious Diseases, offers the first peer-reviewed assessment of how well the vaccine, Ervebo, works in a challenging real-world setting.
The study’s assessment of the vaccine’s effectiveness is lower than the 100% efficacy estimate that resulted from the Ebola Ça Suffit! (Ebola, That’s Enough) trial that was conducted in Guinea during the waning days of the 2014-2016 West African outbreak, or the 97.5% preliminary effectiveness estimate provided by the DRC’s National Institute for Biomedical Research (INRB) and the World Health Organization for the first year of the 2018-2020 outbreak in DRC.
But there were strong suspicions that both of those estimates were too high to be realistic. No vaccine protects everyone who receives it.
“Certainly when used in a real-world context, with all the challenges that that entails, we expect to see lower effectiveness than the [clinical trial] efficacy estimate,” Sophie Meakin, lead author of the new study, told STAT in an interview.
“The main takeaway is that the vaccine is highly protective against developing Ebola virus disease at 10 or more days after vaccination. And even if it’s not as high as the 100% that came from the Ça Suffit! trial, that’s to be expected.”
The study was conducted by researchers from the DRC Ministry of Health, the INRB, the University of Kinshasa, and Doctors Without Borders, which is known by the acronym for the French version of its name, MSF. Meakin is an epidemiologist with MSF’s Epicentre, a Paris-based unit focused on epidemiology and research.
A vaccine effectiveness of 84% means that in a group of vaccinated people who are exposed to the virus, 84% fewer people would go on to develop the disease when compared to what would happen in a similar group of unvaccinated people.
The vaccine, which protects against the Ebola Zaire virus, is one of only two licensed Ebola vaccines, and the only one recommended for use as an outbreak response tool because it is given in a single dose. The other Ebola vaccine, made by Johnson & Johnson, is given in two doses, 56 days apart, an unwieldy regimen in an emergency situation. The Merck product is licensed in both the European Union and the United States; the J&J vaccine is licensed in the European Union.
Meakin and her co-authors calculated the vaccine effectiveness estimate from a database that the DRC health ministry and MSF set up during the 2018-2020 outbreak, the second largest on record. Centered in the conflict-riven provinces of North Kivu and Ituri, on DRC’s borders with Uganda and Rwanda, there were nearly 3,500 confirmed infections and nearly 2,300 deaths in that outbreak.
Assessing interventions for Ebola such as vaccines or drugs is challenging. Many outbreaks have traditionally been in remote locations, where access to affected communities is limited and outbreaks have petered out after a few dozen cases.
The North Kivu-Ituri outbreak was horrific in terms of the damage it wreaked on an already embattled population. But the high number of cases produced a sizable cache of data that the health ministry and Epicentre have since been mining, to try to answer key questions about what works to prevent Ebola, or how best to treat Ebola patients when infection occurs. Meakin said Ebola treatment centers were all asked to collect the same set of data about each patient.
Earlier this year the two groups published a paper showing that people who had received the Merck vaccine but who went on to contract Ebola had a much lower risk of dying than unvaccinated people, even if they were vaccinated after being infected. Though full protection from the vaccine isn’t thought to develop until about 10 days after vaccination, the study reported that even people who had been vaccinated two or fewer days prior to developing symptoms — in other words, people who were probably incubating the disease when they were vaccinated — were about half as likely to die as unvaccinated people who became infected.
The HIV status of patients was not among the data collected, a fact that could potentially have had an impact on the vaccine effectiveness estimate. It’s conceivable that the vaccine would not be as protective for people whose immune systems are compromised, Meakin said; if some portion of the vaccinated people were HIV-positive or immunocompromised due to other health conditions, that could have pushed down the effectiveness estimate.
Heinz Feldmann, who led the work to develop Ervebo, agreed that is a possibility, though he noted that a study done in macaque monkeys infected with a simian equivalent of HIV found that the vaccine offered solid protection. The paper, published in 2008, showed that four of six of these monkeys survived after being exposed to Ebola.
Another possible explanation for the lower vaccine effectiveness estimate could be cold-chain breakdowns, noted Armand Sprecher, an MSF physician who has been involved in Ebola outbreak responses for nearly 25 years. Vaccines must be stored at specific temperatures before use, which can be hard to do in remote settings where electricity can be spotty.
Sprecher cautioned against drawing too much of a distinction between the differences in the effectiveness estimates. “What does it mean? It means [the vaccine’s] really good,” he said. “As vaccines go, something that’s sort of in the eighties or nineties percent efficacy is pretty good. With one dose.”
Feldmann, who is head of virology at the National Institute of Health’s Rocky Mountain Laboratories in Hamilton, Mont., expressed a bit of disappointment that the effectiveness estimate hadn’t topped 90%. But he noted it is important to now have an estimate available that reflects how the vaccine works in an outbreak setting, because it should temper expectations.
“I think this 85% or 84% comes most likely closer to the real situation,” Feldmann said. “It’s a very effective tool for outbreak management.”