Quenching the Fever Without Blinding the Warning Signs
Steroid-refractory adult-onset Still's disease needs a biologic, but the two reasonable choices don't just differ in convenience — one of them can also mask the exact lab pattern that would warn of macrophage activation syndrome developing underneath it.
Priya D., a 29-year-old elementary school teacher, was admitted five days ago with daily spiking fevers to 39.5°C, an evanescent salmon-pink rash that appears with each fever and fades between spikes, and polyarthralgia she initially attributed to “overdoing it” at a weekend hiking trip. A rheumatology workup — negative ANA and rheumatoid factor, ferritin markedly elevated at 8,400 ng/mL, and a compatible clinical picture meeting Yamaguchi criteria — confirmed adult-onset Still's disease after infectious and malignant causes were reasonably excluded. High-dose prednisone, started on hospital day three once infection and malignancy had been reasonably excluded, has not controlled her fevers through day four, and the team is now discussing a biologic.
She has no other significant past medical history — no autoimmune disease in her family that she knows of, no prior hospitalizations, and until six days ago had never spent a night in a hospital as an adult. That clean baseline matters here mainly by contrast: nothing about her history would have predicted this admission, and nothing in it currently explains why her fevers have stayed uncontrolled through four days of high-dose steroids that, in most AOSD presentations, bring at least partial relief by now. What makes today's decision harder than a routine steroid-refractory AOSD case is her most recent trend, not her diagnosis: ferritin has continued climbing over the past forty-eight hours, and her fibrinogen — previously elevated, as expected in active AOSD — has begun to fall toward the normal range faster than her clinical picture would predict. That specific combination, a still-rising ferritin paired with a falling fibrinogen, is one of the earlier laboratory signatures of macrophage activation syndrome developing on top of AOSD, before the full HScore criteria are met, and before the more obvious signs — falling platelets, rising transaminases — have necessarily appeared. It matters directly to which biologic gets chosen next. Shimizu et al. showed that tocilizumab masks the clinical and laboratory signature of Still's-associated MAS, and a single-center series of twenty AOSD patients on tocilizumab found six who went on to develop MAS, several of them afebrile and with ferritin lower than the diagnostic criteria expect. An IL-6 receptor blocker suppresses the acute-phase ferritin response independent of whether MAS is actually progressing underneath it — meaning that starting it now could make her labs look reassuring precisely while the process it would obscure keeps advancing, a distinction that would not matter nearly as much if her trend line looked stable rather than actively moving.
Rheumatology consult, day 4 of high-dose steroids without response
Given where her labs are trending, I'd start anakinra, not tocilizumab. AOSD and MAS run along largely the same IL-1-driven pathway, and anakinra is itself a real treatment for MAS if it fully develops — so we'd be covering both the disease in front of us and the complication we're worried about with a single choice, with a drug short-acting enough to pull back quickly if something changes.
I want to name the case for tocilizumab honestly before we set it aside: Kaneko's randomised, double-blind, placebo-controlled trial showed it controls adult-onset Still's disease that has failed glucocorticoids, which is a better grade of evidence than anakinra has in adults, and dosing every four weeks instead of daily injections matters for a patient who is exhausted and has already told us the daily-injection burden of anakinra worries her.
That's a real advantage — but it doesn't answer the concern about her labs specifically, and I don't think convenience should be the tie-breaker if there's a genuine safety reason to prefer the other drug here.
This is exactly the pattern that worries me from the MAS side of things, not the AOSD side: a still-climbing ferritin against a fibrinogen that's already starting to fall is an early signature, before she meets full criteria for the syndrome. And the Shimizu work is specific — tocilizumab suppresses the ferritin response independent of whether MAS is actually progressing — which means starting it now wouldn't just fail to help if MAS is beginning, it would actively hide the one signal we'd be watching for it.
I'm not arguing tocilizumab is wrong for AOSD generally — it isn't. I'm arguing it's wrong for her specifically, this week, given where her own numbers already sit. Anakinra keeps the warning system intact while still treating the disease in front of us.
Agreed: anakinra started today, prednisone continued at current dose pending response, and ferritin/fibrinogen/platelets/triglycerides trended every twelve hours specifically to catch any further movement toward MAS early — the critical care physician's lab-trajectory argument settled the choice once it was named directly, with the hospitalist's convenience point accepted as real but not decisive here.
Explicitly agreed as a contingency, not left implicit: if her ferritin/fibrinogen pattern continues worsening despite anakinra, the team moves toward formal MAS-directed management rather than simply escalating the AOSD regimen further — named now so it isn't relitigated at 2 a.m. if the labs turn.