Severe Mpox After the Negative Trial: Does Tecovirimat Still Have a Role
A single patient, living with poorly-controlled HIV, in severe pain from mpox proctitis that supportive care alone isn't touching. The disagreement is what is left of a drug after two negative randomized trials, neither of which put a patient like him in a comparative arm.
D.L., a 36-year-old graphic designer, was diagnosed with HIV eight years ago and has cycled on and off antiretroviral therapy through several difficult stretches — a job loss, an eviction, a move across the country to be near family — landing him back in care only two months ago with a CD4 count of 140. Five days ago he developed a mpox lesion at the anal margin that has since progressed to severe, worsening proctitis: pain severe enough that he can't sit without wincing, rectal bleeding, and a fever that hasn't broken despite around-the-clock analgesia. Supportive care — the mainstay of mpox management for most patients — is not controlling either his pain or his systemic symptoms, and his low CD4 count independently predicts a harder, more prolonged course than the disease typically takes in someone with intact cell-mediated immunity.
The evidence he is being weighed against is no longer a single trial in a distant population. PALM007 randomized 597 patients with clade I mpox in the Democratic Republic of the Congo — mean age 16, most of them children — and found no reduction in time to lesion resolution. STOMP then randomized adults with clade II mpox, his own clade, across seven countries, roughly a third of them living with HIV, and found no benefit on lesion resolution, on pain, or on viral clearance; its authors read the two trials together as evidence that tecovirimat lacks clinical efficacy against both clades. Any argument for giving it to him has to survive that, not just PALM007. What survives is narrow and specific: STOMP's randomized arms excluded severe disease and severe immunocompromise by design, routing those patients into a separate open-label arm that was never built to estimate efficacy and closed early, and the CDC's own current position is that tecovirimat's role in severe immunocompromise, advanced HIV included, remains undetermined and is why the expanded-access protocol still exists. His CD4 of 140, drawn on re-engagement two months ago and not yet recovered on restarted therapy, is what puts him inside that carve-out rather than inside either randomized population. The same immune picture cuts against him from the other direction, though, and the team has to hold both: STOMP's virologic analysis found treatment-emergent tecovirimat resistance concentrated in exactly the participants with advanced HIV and profoundly impaired cellular immunity, where it tracked with persistently higher viral loads and slower clinical recovery. The gap the trials left open for him is the same gap in which the drug has most often failed.
ID consult, hospital day two
I'd hold off on tecovirimat and focus everything on aggressive pain control, supportive management, and getting his ART right. And I want to name what has changed since the last time this argument was had on this ward: it is no longer PALM007 on its own. STOMP randomized adults with his clade, a third of them living with HIV, and found nothing — not on lesion resolution, not on pain, not on viral clearance. Its authors put the two trials side by side and concluded the drug lacks clinical efficacy against clade I and clade II both.
One negative trial in a distant population invites a carve-out, and I'd have argued for one. Two negative trials on two continents is a different thing, and the honest reading is that we should stop looking for the population where it works and start looking for a drug that does.
I'd offer it under the expanded-access protocol, and the second trial is exactly why I'd still say that. Look at how STOMP was built. Severe disease and severe immunocompromise weren't underrepresented in the randomized arms by bad luck — they were routed out of them by protocol, into an open-label arm never designed to estimate efficacy, which then closed early. He is not a patient STOMP enrolled and failed to help. He is a patient STOMP declined to randomize. CDC has kept the expanded-access pathway open for people with advanced HIV for precisely that reason, and says so in as many words: the role of the drug in severe immunocompromise is undetermined.
You're right that this is the same shape of argument people use to wriggle out of any trial they dislike, and I'd hold myself to the harder version of it: the exclusion has to be structural, written into the protocol, not something I inferred from a baseline table afterward. Here it is written in. That is the whole of my claim — not that he'll benefit, but that neither trial measured whether he would.
I'd treat him, and I'd rather ground it in what I'm watching at the bedside than in how we settle the generalizability argument: five days of worsening, treatment-refractory pain and unbroken fever in a man with a CD4 of 140, not responding to supportive care. But I don't want to walk past the finding neither of you has picked up. STOMP's virologic analysis found treatment-emergent resistance clustering in the participants with advanced HIV and profoundly impaired cellular immunity — higher viral loads that stayed higher, slower recovery. The subgroup we're claiming as his carve-out is the same subgroup in which the drug most often stopped working.
So I'd treat, but not open-endedly. A defined course, a stated endpoint, and a decision point where we either see movement or we stop and say so — because if he is the patient in whom resistance emerges, a long uninterrupted course is the way we'd find out too late. This isn't an argument for routine tecovirimat in every mpox case. It's an argument for a bounded trial of therapy in a patient failing everything else, with the resistance data telling us how to run it rather than whether to.
Agreed: tecovirimat started, analgesia escalated with an acute pain service consult, and antiretroviral therapy continued alongside close monitoring for immune reconstitution concerns given his recent CD4 nadir.
Not agreed: the infectious disease physician's underlying caution about explaining away negative trial results case by case remains a real, standing concern for the group, not resolved by this one decision. The team was explicit that this was a judgment specific to his severity and immune status, not a template for offering tecovirimat to every mpox patient whose picture differs even slightly from PALM007's enrolled population.