Bundibugyo, Bunia, and the trial that had to happen anyway

TL;DR

• On Thursday 2 July 2026, the first patient was enrolled in the PARTNERS trial at the Evangelical Medical Center in Bunia, Ituri province, eastern Democratic Republic of the Congo — the first randomised trial ever run inside an active outbreak of Bundibugyo virus (BDBV), a species of Ebola for which no approved vaccine or treatment exists.
• The trial tests Gilead's remdesivir (a broad-spectrum antiviral) and Mapp Biopharmaceutical's MBP134 (an experimental monoclonal-antibody cocktail), alone and in combination. A third WHO priority candidate — maftivimab, the middle component of Regeneron's approved Zaire-Ebola cocktail Inmazeb — will be added under an adaptive protocol once monotherapy stocks are validated.
• Primary endpoint: 28-day survival. Trial may require up to 1,000 participants over several months to yield a definitive answer — fewer if either drug proves highly effective.
• Outbreak scale as of 30 June (WHO): 1,406 confirmed cases, 301 suspected, 438 deaths in DRC alone, with spillover into Uganda. Case-fatality rate tracking at roughly 31% — consistent with historical BDBV lethality (25–40%).
• The scientific bet is not really about which drug wins. It is whether outbreak-embedded research can now be done fast enough, safely enough, and with enough community trust to matter in the outbreak that funded it.

The framework, before the news

In outbreak medicine there is a chronic, well-known failure mode: by the time a rigorous trial is designed, ethics-approved, resourced, and deployed, the outbreak it was meant to answer questions about is over. The data that arrives is data for the next outbreak — if there is one, and if the pathogen behaves the same way, and if the world remembers.

The 2014–16 West Africa Ebola epidemic was the industry's public confession of this problem. Trials began late. Some finished after transmission had already collapsed. The definitive answer on the Ebola vaccine rVSV-ZEBOV came after the epidemic that made it urgent. The lesson, learned publicly and often, was: research has to run inside the outbreak, not behind it.

Every outbreak since has been, in part, a test of whether the world can now do that. The 2018–20 outbreak in eastern DRC produced the PALM trial, which established Inmazeb (REGN-EB3) and later Ebanga (mAb114) as the first approved treatments for Zaire ebolavirus. It worked. It was also a specific species, in a specific window, with specific drugs.

Bundibugyo is a harder case. It has caused only a handful of documented outbreaks since it was first identified in 2007 near the Uganda–DRC border. Nobody has ever tested a drug against it during transmission. The existing Ebola vaccine and Inmazeb, both licensed for Zaire species, are not confirmed to work against BDBV. So when this outbreak was declared in May 2026, the cupboard was, technically, bare.

PARTNERS is the response to that bare cupboard. Whether it produces a licensed drug for Bundibugyo is one question. Whether it succeeds as an operational template is the more interesting one.

What happened

Speaking from WHO headquarters in Geneva on Thursday, Director-General Tedros Adhanom Ghebreyesus announced the enrolment of the first PARTNERS participant at the Ebola treatment centre inside Bunia's Evangelical Medical Center. AP correspondents in Bunia described the moment as marked by "urgency rather than ceremony" — health workers moving between wards, no press event, the first patient consented and dosed the same day.

The trial's operations lead, Professor Amanda Rojek of the University of Oxford's Pandemic Sciences Institute, framed it in a single sentence that is now doing the rounds in global health WhatsApp groups: "One of the key lessons from recent outbreaks is that research needs to happen alongside the response, not after it."

The outbreak continues to grow. WHO data as of 30 June: 1,406 confirmed cases, 301 suspected, 438 deaths in DRC, with additional confirmed transmission across the border in Uganda. The public health emergency declaration remains in force. The United States began an emergency vaccination effort using experimental BDBV-targeted vaccine candidates in June.

The trial, decoded

For readers who need the mechanics — this is where the briefing register earns its keep.

Design
Adaptive, randomised, controlled. Four arms initially: (1) standard care, (2) remdesivir alone, (3) MBP134 alone, (4) remdesivir + MBP134 combination. Maftivimab monotherapy will be added once GMP stocks are validated — the adaptive protocol allows arm insertion without a full re-approval cycle.

Primary endpoint
28-day all-cause mortality. WHO research adviser Dr Vasee Moorthy has been explicit that this is a survival trial, not a viral-load trial. In a pathogen this lethal, the outcome that matters is whether more people walk out of the treatment centre.

Sample size
"Up to 1,000 participants" cited by Moorthy, with the caveat that if either monotherapy proves highly effective the adaptive design can stop earlier. Realistically, if either drug produces an effect on the scale of what mAb114 showed against Zaire species (halving mortality from roughly 66% to under 35% in the sickest patients), a signal could emerge in the low hundreds of enrolments.

The drugs
Remdesivir is the outsider. It performed weakly against Zaire ebolavirus in PALM — one of the arms that was dropped. Its inclusion here rests on in-vitro data suggesting BDBV may be more susceptible than Zaire, plus the practical fact that Gilead already has manufacturing scale, cold-chain logistics, and a licensed-for-COVID product on hand. It is the pragmatic bet.

MBP134 is the scientific bet. A pan-ebolavirus antibody cocktail engineered by Mapp Biopharmaceutical from convalescent-donor sequences, designed specifically to hit conserved epitopes across Ebola species — including Bundibugyo, Sudan, and Zaire. Non-human primate data has been strong. Human data at outbreak scale, until this trial, was zero.

What isn't in the trial
No head-to-head with a licensed Zaire-Ebola drug (Inmazeb, Ebanga) — because using either off-label as a comparator would introduce ethical and regulatory questions the trial cannot resolve inside an emergency. This is a genuine gap in what PARTNERS will tell the world.

The quieter story: community trust as the load-bearing wall

The most quoted line from local reporting is not from a scientist. Nelson Dhebi, a shopkeeper in Bunia, told AP that while he supports the research and hopes for good outcomes, he worries the treatments could cause deaths, and that others should be part of the trial.

This is not an outlier voice. It is the median voice. Every outbreak-embedded trial since 2014 has failed or succeeded on the same axis: whether the community carrying the outbreak trusts the researchers enough to consent, comply, and follow up. In 2019, mistrust in Beni contributed to attacks on Ebola treatment centres and burial teams. In 2026, PARTNERS is being run at a site — the Evangelical Medical Center — chosen partly because it has existing standing in Bunia and Ituri that a WHO field tent would not.

Read against that history, Dhebi's statement is the correct level of scepticism. The right question the trial has to answer, in his terms, is not "does the drug work" but "does the treatment centre make my neighbour more likely to come home." Those are the same question. The trial's job is to make sure the second phrasing is the one people believe.

Stakeholder landscape

• Patients and communities in Ituri and North Kivu, DRC, and cross-border districts in Uganda. First-order stakeholders. Their consent and follow-up determine whether the trial produces usable data.
• DRC Ministry of Health and Ugandan MoH. Political owners of the response. Trial results will feed directly into national treatment protocols.
• WHO (Geneva). Sponsor, coordinator, and — de facto — the entity being publicly tested on whether it has learned the lesson it has been saying it learned since 2016.
• University of Oxford Pandemic Sciences Institute. Operations lead. The institutional descendant of the RECOVERY-trial playbook that reshaped COVID therapeutics through pragmatic, at-scale randomisation.
• Gilead Sciences and Mapp Biopharmaceutical. Drug sponsors. Both have quiet reputational and commercial upside if their asset works; both risk very little if PARTNERS' primary finding is "neither drug materially changes survival."
• Regeneron. Not a direct participant, but supplies the maftivimab component of Inmazeb and will be a stakeholder if monotherapy stocks enter the adaptive arm.
• US BARDA and CEPI. Funders of BDBV vaccine development in parallel. Their investment thesis benefits from any BDBV-specific therapeutic winning.
• The broader global-health research architecture. Watching whether the "research embedded in response" doctrine works when the outbreak is small, remote, and in a species without a commercial market.

Cross-layer connections

Three that are not obvious:

Manufacturing geography. MBP134 is a monoclonal antibody. Antibody manufacturing at scale still lives predominantly in the US, EU, and increasingly China. If PARTNERS produces a positive result for MBP134, the next question — quietly urgent — is whether any of that capacity can be deployed toward a pathogen with no commercial market and unpredictable outbreak timing. This is the same architecture problem that produced the mpox-vaccine bottleneck in 2022 and the yellow-fever supply crunch that periodically returns.

Vaccine–therapeutic sequencing. The US is running an experimental BDBV vaccination effort in parallel. If both work, the more consequential public-health question becomes the sequencing one: ring-vaccinate contacts, or treat cases and their close contacts prophylactically? PARTNERS' data on antibody kinetics, if published in the granular form the field has been pushing for since 2016, will shape that answer.

The AI-in-drug-discovery frame. MBP134 was engineered using structural biology and epitope-mapping approaches that are now being aggressively AI-augmented across the industry. A positive PARTNERS result would be one of the first outbreak-scale validations of the engineered-antibody-cocktail approach for a rare pathogen — a template with obvious relevance to Marburg, Lassa, Nipah, and other filoviruses and haemorrhagic fevers.

Uncertainty ledger

• Trial power. If enrolment slows because the outbreak decelerates, PARTNERS may again face the 2015 problem — an incomplete answer for the next outbreak, not this one. The adaptive design mitigates but does not eliminate this risk.
• Species specificity. In-vitro susceptibility for remdesivir against BDBV is suggestive, not decisive. If it fails, the field is back to MBP134 as the single hope inside this outbreak.
• Regulatory pathway. No agency has an approved BDBV therapeutic pathway. FDA and EMA precedent from the Inmazeb / Ebanga approvals for Zaire species is the closest template, and it is not a perfect fit.
• Cross-border containment. Uganda case count is climbing more slowly, but the porous Ituri–West Nile corridor is the same geography that produced the 2000 Gulu outbreak. A larger Ugandan cluster changes trial logistics.
• Community trust events. Any adverse event in an early trial arm, correctly or incorrectly attributed to the drug, could stall enrolment sharply.

What this means for you

Recommendations here are addressed to the general public and, where relevant, to the natural audiences of the story.

For the general public (anywhere). This is a scientific process story worth following, not a panic story. BDBV does not currently pose a material risk outside affected districts of DRC and Uganda. The right thing to do is follow the trial's progress through WHO situation reports and reputable science journalism, and to be sceptical of both hype ("miracle Ebola cure") and dismissal ("nothing ever changes") as the first data lands.

For clinicians and public-health professionals in unaffected countries. Watch for adaptive-protocol amendments and interim safety announcements. If MBP134 shows early efficacy, national stockpiling conversations will move quickly for at-risk states — and, in the medium term, for any country running significant travel or troop rotation through central Africa.

For research funders and biosecurity policy readers. The signal to track is not whether either drug wins. It is trial velocity: time from outbreak declaration to first patient enrolled (approximately eight weeks in this case), retention rates, and time-to-interim-analysis. Those metrics are the actual test of the "research inside the outbreak" doctrine.

For anyone tempted to make this a geopolitical or vaccine-policy story. Don't, yet. The evidence isn't there. It may become one — see the uncertainty ledger — but writing that story before the data lands is exactly the failure mode PARTNERS is designed to correct.

Honest disclosure: for the average reader with no direct connection to central African public health, there is no immediate action here. Attention is not action. Following the story carefully is the useful thing.

Bottom Line

The PARTNERS trial is the most important experiment being run in global health this month, and almost none of the importance is about the two drugs on the label. The real experiment is whether the world's outbreak-research architecture — WHO, Oxford, DRC MoH, sponsor pharma, and the community whose neighbours are dying — can now move at outbreak speed without breaking anything that has to hold. Bundibugyo is a small, cruel virus in a difficult province. If PARTNERS produces a clean answer inside this outbreak, the template it establishes will outlive the answer.

Sources

Tier 1

• BBC, "Ebola treatments trial begins in the Democratic Republic of Congo" — 2 July 2026
• Reuters, "Trial for Ebola treatment starts in DR Congo" — 3 July 2026
• Associated Press, "Researchers launch study on Ebola treatments as Congo outbreak worsens" — 2 July 2026
• Associated Press / The Washington Post, "Residents in eastern Congo cling to hope as a new Ebola treatment trial begins" — 5 July 2026
• WHO Director-General remarks, Geneva — 2 July 2026
• Reuters, "Congo says number of confirmed Ebola cases at 1,274, including 360 deaths" — 29 June 2026

Tier 2

• Politico, "Residents in Congo cling to hope as Ebola treatment trial begins" — 5 July 2026
• pharmaphorum, "Trial of Bundibugyo Ebola drugs starts in DRC" (Prof Amanda Rojek quote; PARTNERS trial design detail) — 3 July 2026