By Clinical Informatics News Staff 

July 31, 2015 | The headline results of a clinical trial of a potential Ebola vaccine, published today in The Lancet, have stirred enormous optimism and brought renewed attention to the catastrophic Ebola epidemic that is still ongoing in Guinea and Sierra Leone. The authors of the trial, conducted in the Maritime region of Guinea that includes the capital city Conakry, have reported 100% efficacy of the vaccine in preventing new cases of Ebola disease in an interim analysis.

Those results bear much closer scrutiny, and by no means guarantee that the vaccine, called rVSV-ZEBOV, will foil all cases of Ebola. Whatever the end results, however, the creative design of the study is already being celebrated as a template for future efforts to test out new public health measures quickly, in emergency situations, and without compromising statistical power. In Science Magazine, Michael Osterholm, director of the University of Minnesota’s Center for Infectious Disease Research and Policy, is quoted as saying, “This will go down in history as one of those hallmark public health efforts… Had this been a standard, straightforward randomized controlled trial, we would never had [sic] this answer.”

Vaccines in the trial were administered much like they would be in a full vaccination campaign in a resource-limited area, following the “ring vaccination” model used during the smallpox eradication of the 1970s. In a ring vaccination, each new confirmed case of the disease is followed by a round of vaccinations, which target everyone in close contact with the infected individual (or their clothes or body fluids), plus second-order contacts. As a result, a limited number of vaccines can be given to the people most at risk, forming a “ring” that restricts further transmission of the disease. In a clinical trial, the approach also has the advantage of increasing the chance that new cases of the disease will appear in the study population ― raising the trial’s statistical power, even when, as in this case, the overall incidence of the disease was already on the decline.

Although ring vaccinations are a common strategy in public health campaigns, they had never before been adopted in a clinical trial. The novel study design and analysis is complex, but worth following. Once a new case of Ebola was confirmed, all the first- and second-order contacts of the infected individual were randomized as a single block to receive either immediate vaccination, or delayed vaccination after 21 days. The randomization process also sorted blocks so that the immediate and delayed vaccination groups would be similar, with a mix of large and small blocks in urban and rural settings. Reasoning that the speed with which the vaccine might take effect was unknown, and that Ebola has a 10-day incubation period before symptoms appear, the study authors looked to see whether individuals enrolled in the trial developed Ebola more than 10 days after being randomized.

Comparisons among groups were flexible. All told, 96 index cases were found, with 50 blocks receiving immediate vaccination (including over 2,000 individuals who consented), and 46 blocks receiving delayed vaccination (including almost 1,500 individuals who consented). People who received immediate vaccinations were compared against all eligible people in the delayed vaccination blocks, whether or not they consented to be vaccinated, boosting the control group to almost 2,400.

Because so little was known about the vaccine’s effectiveness, it was not clear how many people would have to be enrolled for the trial to be statistically powered to find an effect. If the vaccine prevented a very high percentage of infections, the effect would be discernable early on; if it were only 50% effective, a longer trial would be required. Therefore, the authors chose to do a single interim analysis at around 100 blocks, and see if the results justified modifying the trial.

That interim analysis is what was published today, and it seems to have exceeded expectations. With no cases of Ebola in the treatment group, and 16 in the control group, it is apparent that 96 blocks were enough to find a strong effect of vaccination. At first analysis, the authors estimate that these results are consistent with the vaccine being roughly 75% effective ― in theory, it could be as much as 100%.

The model of offering delayed vaccinations to the control blocks also made it possible to track overall Ebola cases over a large population in the long-term. Even once individuals who received delayed vaccinations are taken into account, the trial has so far seen no cases of Ebola developing more than six days after vaccination, in a total vaccinated population of more than 3,500. There is also some evidence that being in a vaccinated block is protective against the disease even for those who were ineligible or did not consent to be vaccinated, which would be consistent with vaccinations stopping transmissions of Ebola; so far, however, the trial is not large enough to confirm this effect.

The trial was designed to be adaptive, and now that the vaccine’s efficacy is apparent, the protocol will shift to giving immediate vaccinations to everyone who consents. Future analysis will focus on learning how often the Ebola virus is able to overcome the vaccine, as well as flagging safety concerns, which will help inform public health efforts.

The most profound takeaway, of course, is that it is highly likely that a safe and effective vaccine has been developed for Ebola, which could help prevent new crises like the epidemic that has been blazing in West Africa since December 2013. The trial design itself, however, is also a notable turn in the public response to the outbreak. As the authors write, “This trial design is logistically feasible, even in resource-poor settings and in a crisis situation. The approach is successful when the incidence of Ebola virus disease is low in the general population and new cases are concentrated in family and community contacts.”

The rapid organization and execution of this trial is a credit for the World Health Organization, the primary sponsor; funders including the Wellcome Trust, Doctors without Borders, and the Norwegian and Canadian governments; and most of all, the Guinean and international health workers on the ground who worked a complex vaccination and patient monitoring plan into their already overwhelming activities to contain the outbreak. While the trial itself was well-coordinated, however, public health experts remain frustrated at the lack of preparedness that delayed testing of vaccines. As the Wellcome Trust observes in a blog post today, rVSV-ZEBOV was developed by the Public Health Agency of Canada, and licensed to a small biotech company, years before the current outbreak began ― but had never moved past preclinical studies.

“If the safety testing work had been done earlier,” the post says, “before the urgency of the outbreak[,] the vaccine’s efficacy could have been assessed much sooner, perhaps changing the course of the epidemic and ultimately saving lives.”