When Anakinra Isn't Enough: Reaching Past IL-1
A young man's macrophage activation syndrome is worsening on maximal anakinra and steroids — the choice ahead is between a drug approved for exactly his diagnosis but resting on a very small evidence base, and the older, more toxic protocol with by far the longest track record.
Idris O., a 26-year-old graduate student in mechanical engineering, was diagnosed with adult-onset Still's disease four months ago and had been doing reasonably well on anakinra and a prednisone taper until ten days ago, when he developed recurrent high fevers, worsening cytopenias, and a ferritin that has climbed past 40,000 ng/mL despite his anakinra dose already being pushed to its upper labeled range. His fibrinogen has fallen below 100 mg/dL, his platelet count has dropped to 62,000/µL, and his transaminases have roughly tripled over the past four days — he now meets formal criteria for macrophage activation syndrome developing on top of his underlying AOSD, refractory to first-line IL-1 blockade and steroids.
Before this admission, Idris had no other significant medical history beyond the AOSD diagnosis itself four months ago, no prior hospitalizations, and had been back at his research lab full time for nearly six weeks, describing himself to his rheumatologist at his last outpatient visit as “finally feeling like a person again.” That recent, hard-won stability is part of what makes this admission land as more alarming to his care team than a first presentation of MAS might otherwise — not a new disease finding its footing, but a previously controlled one turning sharply worse.
What makes his particular case unusually informative for the choice ahead is a specific laboratory pattern, not just the severity of his decline: his free IL-18 level, sent once MAS was confirmed, has returned markedly elevated, disproportionately high relative to his ferritin compared to typical MAS laboratory profiles, pointing toward a disease genuinely dominated by the IL-18/interferon-gamma axis rather than a more generic hyperinflammatory picture. Two escalation strategies exist once IL-1 blockade fails, and they are not equivalent for a patient whose biology looks like his. The HLH-94 protocol, chemotherapy-based and the most established option across hemophagocytic syndromes broadly, carries substantial myelosuppression risk in a man already down to 62,000 platelets. Emapalumab, an anti-interferon-gamma monoclonal antibody, targets the effector his own labs implicate most directly — and since June 2025 it has carried a specific FDA approval for MAS in known or suspected Still's disease, adult-onset included, in patients with an inadequate response to glucocorticoids. That approval rests on the pooled EMERALD and NI-0501-06 trials, which required a ferritin above 684 ng/mL plus two of four laboratory findings to enroll; Idris clears the ferritin threshold roughly sixty-fold and meets three of the remaining four. He is critically ill but not yet in multiorgan failure, so there is time to weigh this — and what weighing it reveals is that the drug being treated here as the adventurous option is the one actually written for him.
Combined rheumatology, hematology, and critical care consult
Given how fast he's declining, I'd move to HLH-94 — etoposide and dexamethasone. I know emapalumab now carries the label for this exact indication. I also know that label rests on thirty-nine patients across two open-label single-arm trials with no control group, against decades of HLH-94 experience in severe refractory hemophagocytic disease of every etiology. An approval and a deep evidence base are not the same quantity, and I don't want to discover the difference while he's losing ground.
I'd point to emapalumab instead — it targets interferon-gamma, the specific downstream effector of the pathway his own labs are implicating most directly, and the trials behind its label enrolled adults and children with MAS in Still's disease who had failed high-dose glucocorticoids. That is a description of Idris, not a population I'm extrapolating toward. Fifty-four percent reached complete response by week eight. He's already down to 62,000 platelets; adding etoposide's myelosuppression on top of that is a real, immediate cost, not a distant one.
I'll grant you thirty-nine patients is thin. But HLH-94's depth is largely in primary HLH and malignancy-associated disease — it was never studied in Still's-associated MAS at all, so on his actual diagnosis your evidence base isn't deeper than mine, it's absent. Small-and-on-target still beats large-and-elsewhere.
I think his own free IL-18 result is the thing that should actually decide this, not either of your general priors about the two protocols. It's disproportionately elevated relative to his ferritin in a way that's genuinely more specific to IL-18/interferon-gamma axis-driven disease than a typical MAS picture — that's evidence about him specifically, not just a mechanistic argument in the abstract.
Given that, I'd support emapalumab as the next step, and I'd note the trials permitted concomitant anakinra, so continuing his is consistent with how the drug was actually studied rather than an improvisation on top of it. But with an honest, named limit, not open-ended: if he isn't showing meaningful improvement in ferritin, fibrinogen, and clinical stability within 48 to 72 hours, we move to HLH-94 without further debate. That gives the more targeted approach a real chance given what his biology suggests, while keeping a concrete deadline so his overall decline doesn't get lost waiting on it.
Agreed: emapalumab started today, anakinra continued, with an explicit, pre-agreed 48-72 hour deadline for reassessing ferritin, fibrinogen, platelets, and clinical trajectory — the critical care physician's biology-specific framing gave the hematologist/oncologist's urgency and the rheumatologist's mechanistic preference a shared, checkable basis rather than requiring either to simply defer.
Emapalumab continued, HLH-94 not pursued — treated as evidence the IL-18/interferon-gamma axis read was correct for his specific disease.
HLH-94 (etoposide/dexamethasone) started without further debate — the deadline agreed today is treated as binding, not renegotiated in the moment by whichever specialist is present then.