Congo Rolls Out a Mismatched Ebola Vaccine After 3,000 Deaths

Congo’s Bundibugyo Ebola outbreak has killed 3,007 people among 6,186 confirmed cases, the National Institute of Public Health said, and there is still no licensed shot for this virus. Figures through Aug. 31 also list 1,409 recoveries and 830 people in isolation or treatment, a fatality rate of 48.6%.

The virus has reached 60 health zones in six provinces. Ituri still accounts for 81.9% of confirmed infections. Health authorities began putting Ervebo, a vaccine licensed for a different Ebola species, into the arms of front-line workers in late August, even as WHO repeated that the product should stay inside a study.

Why Kinshasa Is Injecting a Zaire Vaccine

Bundibugyo virus, one of the orthoebolaviruses, had caused only two known outbreaks before this year. Uganda recorded 131 cases and 42 deaths in 2007. Congo recorded 38 laboratory-confirmed cases and 13 deaths in 2012. Case fatality then sat at 30% and 50%. Those waves were small enough that no company finished a Bundibugyo-specific product.

Ervebo, Merck’s single-dose rVSV vaccine, is licensed and WHO-prequalified for Zaire ebolavirus, the species behind 15 of Congo’s previous 16 Ebola outbreaks. About 345,000 people in Ituri already received it during the 2018 to 2020 Zaire epidemic. That familiarity is part of why officials reached for the same vial this time.

The supporting science is thin and mixed. A University of Texas primate study found that 3 of 4 animals given an rVSV Zaire vaccine survived Bundibugyo challenge, against 1 of 4 unvaccinated controls. Banked ferret sera after Ervebo showed ferret antibodies that cross-reacted with Bundibugyo, a limited humoral response. Human samples from West African Ervebo recipients, published in the New England Journal of Medicine, produced Bundibugyo-binding antibodies about eight times lower than the Zaire response at three months.

Outbreak surveillance presented to WHO’s vaccine advisers found nine confirmed Bundibugyo patients with a recorded Ervebo history, and all nine survived. In one probable health-worker cluster, 3 of 3 vaccinated workers lived and 5 of 5 unvaccinated workers died. Those counts are small. They are also the closest thing responders have to human outcome data.

THE OUTBREAK CLOCK

1. February 19, 2026: CDC modelers put the likely animal-to-human spillover around this date, weeks before anyone named the virus.
2. May 15, 2026: Congo declares an Ebola outbreak in Ituri after Bundibugyo is confirmed in laboratory samples.
3. May 17, 2026: WHO declares a Public Health Emergency of International Concern. Africa CDC follows a day later with a continental emergency.
4. July 2026: The PARTNERS treatment trial enrolls its first patients against a virus that still has no approved drug.
5. August 20, 2026: The International Coordinating Group assigns 70,000 Ervebo doses from the Gavi-funded stockpile.
6. August 31, 2026: Updated WHO emergency guidance again confines Ervebo to research protocols.
7. September 1, 2026: The new school year begins while most new cases still fall outside known contact lists.

Steve Ahuka-Mundeke, a Congolese virus specialist who helped run earlier campaigns, has argued that using Ervebo now would also rebuild protection against a future Zaire wave and test whether the shot does anything to Bundibugyo. That two-for-one logic is how a mismatched product reached Kisangani.

SAGE Still Limits Ervebo to Research Protocols

WHO’s first emergency note on May 28 said evidence of cross-protection was “very limited” and that Ervebo should not be used as a regular outbreak tool outside controlled research. After new animal, lab, and field observations, SAGE met again in an extraordinary session. The updated Ervebo only in research protocols guidance, dated Aug. 31, did not flip that core line.

Since evidence remains insufficient to support the programmatic use of the Ervebo vaccine for the prevention of BDBV disease, WHO recommends Ervebo should only be used for BDBV within the context of a research protocol.

World Health Organization, emergency guidance after SAGE, Sept. 2, 2026

WHO has issued new emergency guidance on the use of licensed Ebola vaccines during Bundibugyo virus disease (BDBV) outbreaks, following an Extraordinary Meeting of the Strategic Advisory Group of Experts on Immunization.

Since evidence remains insufficient to support the… pic.twitter.com/zWjvHpGEeN

— World Health Organization (WHO) (@WHO) September 2, 2026

The International Coordinating Group still 70,000 Ervebo doses released to Congo from a global stockpile of 500,000, splitting 20,000 for a Phase 3 trial and 50,000 for front-line and health workers. The first shipment has arrived. Africa CDC’s Emergency Consultative Group, reviewing the same file on Aug. 14, said there was still no reliable evidence of clinical efficacy and backed a ring-vaccination trial plus compassionate use for front-line staff under a study protocol.

MSD, which makes Ervebo, has said it holds no safety or efficacy data on Bundibugyo and that stockpile doses move only at UNICEF’s request. Yves Lévy, an INSERM immunologist whose team measured the cross-reactive antibodies, called those lab findings “very encouraging.” Encouraging is not the same as a licensed claim. The product going into arms remains off-label.

Bundibugyo-specific shots are further back. Oxford’s ChAdOx1 BDBV candidate dosed its first volunteer in July. Moderna’s mRNA-1469 candidate, backed by CEPI, dosed first participants on Aug. 4 at three sites in Canada. Those trials cannot change the next month of transmission in Ituri.

Five Control Targets Remain Below the Line

The U.S. Centers for Disease Control and Prevention, reading Congo’s daily situation reports, said in a Sept. 1 Notes from the Field that this is now the second-largest Ebola outbreak on record. As of Aug. 21, Congo had 5,458 confirmed cases and 2,606 confirmed deaths, about 5,000 cases in 100 days. The agency’s line on the geography was blunt: the pattern indicates uncontrolled expansion.

On the five missed outbreak control targets that usually break an Ebola chain, the three-week window from July 31 to Aug. 21 missed every one that had fresh data.

THE CDC SCORECARD AS OF AUGUST 21

National treatment-unit occupancy averaged 64%, under the 80% ceiling, but some zones reported 140% occupancy and could not isolate every infected patient. Africa CDC has said recorded cases may be about three times too low because surveillance and contact tracing remain weak. If that multiplier holds, the 6,186 confirmed infections are a floor, not a census.

A June CDC projection had already sketched this path. With only 20% of infected people isolated, 65% of simulations reached at least 20,000 cases and 69% reached at least 4,000 deaths within about 90 days of intervention. Jean Kaseya, director general of Africa CDC, said in June, “If we don’t stop this outbreak now, for sure it will be the largest Ebola outbreak ever.” He also said, “There is huge, huge community transmission.”

West Africa’s 2014 to 2016 epidemic still sits above this one, at more than 28,000 cases and more than 11,000 deaths. Congo’s own previous high-water mark, the 2018 to 2020 North Kivu Zaire outbreak, recorded 3,317 cases and nearly 2,300 deaths. This Bundibugyo wave has already passed that national record and is Congo’s 17th Ebola outbreak since 1976.

Unpaid Crews, Gold Camps, and Open Roads

The people asked to close those gaps have been walking off the job. In Ituri, some clinicians told colleagues they had not been paid wages or bonuses since the May 15 declaration. Strikes hit the Elikya treatment center in Bunia in July. Hygiene and decontamination crews later stopped work too. In mid-August, six Ebola checkpoints in Haut-Uélé stopped filing reports during a pay strike; five were still silent almost two weeks later. Three checkpoints in Ituri were on strike as of late August. In North Kivu, only 5 of 18 checkpoint staff had been paid by Aug. 19. Community-engagement workers in North Kivu and Tshopo went unpaid for three months, government reports show.

Edouige Makosi, a protester in Bunia, said, “With this way of managing things, Ebola will not end in this province.” Dr. Ben Bakule, a community investigator, described a late-May attack in Tutu, in Djugu territory, when a group of young men came at him and his colleagues during contact tracing.

WHO counted at least 155 infected health workers and 45 deaths as of Aug. 9, a 29% fatality rate among staff, plus 12 attacks on health care since the May 17 emergency declaration. Those infections cluster outside the dedicated treatment centers, where infection-control supplies and protocols are thinner.

WHAT KEEPS FEEDING NEW CHAINS

• Pay stoppages: Checkpoint clerks, burial teams, and nurses have repeatedly halted work over unpaid wages, punching holes in the daily trace.
• Armed groups: Ituri and North Kivu remain conflict zones, and attacks on ambulances and treatment sites cut access for days at a time.
• Gold camps: Artisanal miners move between crowded pits with little sanitation; authorities in Mongbwalu, a main gold town, believed more than 80 people had died before the outbreak was even named.
• Hidden deaths: 59% of confirmed deaths still occur outside treatment units, which means funerals and home care keep seeding new infections.
• The wrong kit: No licensed Bundibugyo vaccine or drug exists, so the pharmaceutical half of the response is a trial and an off-label Zaire shot.

Papy Baraka, a driver in Bunia, has cut the number of passengers he takes. “The disease puts everyone at risk, and daily life has slowed to a crawl. We are afraid,” he said. Jean-Claude Angwanzia left Mambassa, in the Ituri epicenter, because he did not feel safe.

Two Trial Drugs Against a Virus With No License

The licensed Zaire monoclonal antibodies Inmazeb and Ebanga were not built for this species. WHO’s therapeutics group in May prioritized MBP-134, a two-antibody cocktail from Mapp Biopharmaceutical with broad orthoebolavirus activity, plus Gilead’s remdesivir, for field trials. Obeldesivir was set aside for post-exposure study. Maftivimab, one antibody from a licensed Zaire cocktail, was listed as a later add-on.

The PARTNERS platform trial, run by Congo’s National Institute for Biomedical Research with Oxford and the Institute of Tropical Medicine in Antwerp, began enrolling in early July. Patients are assigned to MBP-134, remdesivir, both, or supportive care alone. Gilead and the U.S. government donated enough product for 1,200 patients. By Aug. 25, day 100 of the WHO emergency, the trial had enrolled its 200th participant, the fastest randomized Ebola treatment study on both launch time and pace. Two hundred treated patients against 3,007 recorded deaths is the gap the trial cannot close on its own.

A previously healthy 39-year-old health worker infected in Ituri was evacuated to Germany, received MBP-134 under an FDA emergency investigational authorization, plus remdesivir and supportive care, and was discharged on day 22 after symptoms began. Viral RNA was undetectable in blood, throat, urine, and stool by day 13 and in semen by day 25. That is one carefully managed case, not a field protocol.

Incubation runs from 2 to 21 days. Early fever, fatigue, muscle pain, headache, and sore throat look like malaria, which is why laboratory confirmation and isolation have to be fast. They have not been. Uganda had 20 confirmed cases and 2 deaths as of Aug. 12, then cleared its last imported patient. France reported one imported case on June 24 and no onward spread. The risk WHO assigned to Congo remains very high; the global risk remains low. WHO advises against travel or trade bans.

School Gates Opened on Unmapped Chains

The new school year began on Sept. 1 while CDC’s last published contact figure showed only 15% to 20% of new cases coming from people already on a watch list. Parents in the eastern provinces have been asked to send children into classrooms anyway. Authorities say prevention steps are in place. Families who watched neighbors die at home have reasons to doubt that a handwashing station is enough.

Safe burial coverage still reaches only about half of affected health zones. Treatment beds overflow in the hottest zones. Checkpoints go dark when the pay does not arrive. Miners keep moving. The vaccine now being offered to the people who staff those jobs is a Zaire product inside a research wrapper, and SAGE has not blessed it as a campaign tool.

Congo has run 16 Ebola responses before this one and knows the playbook: find the sick, list their contacts, isolate, test, bury the dead without infecting the family. That playbook is on the page. On the ground in Ituri it is running at 10.6 contacts per case, and the only licensed vial in the cooler was made for a different virus.

Disclaimer: This article is news reporting and analysis of an ongoing outbreak for general information. It is not medical advice, a treatment recommendation, or guidance on vaccination, travel, or clinical care for Ebola or any other illness. Readers who may have been exposed, who have symptoms, or who work in affected areas should consult a qualified physician or the local health authority before making health decisions. Case counts, death totals, trial results, and vaccine guidance are those published by the cited health agencies as of the dates given in the story and will change as the outbreak and the studies move.