Recurrent Erythema Nodosum With No Trigger Found: How Far to Treat an Idiopathic Diagnosis
A single patient on her fourth flare of erythema nodosum with three clean workups behind her. The disagreement isn’t about what’s causing it — the workup has already answered that as honestly as it can — it’s about how aggressively a recurring, functionally disruptive, genuinely idiopathic condition deserves to be treated.
Dara O., a 29-year-old woman, is the second-oldest of five siblings and the one who drives their widowed father to most of his own appointments now that he’s stopped trusting himself behind the wheel after dark. Over the past eight months she has had three separate episodes of tender, red, raised nodules across both shins, each one arriving over a few days, staying painful enough that she avoids bumping her legs against furniture for three to five weeks, and then fading into the yellow-green bruise-like discoloration erythema nodosum is known for before finally clearing.
Each episode has triggered a full workup, and each has come back clean: a chest X-ray and, after the second episode, a chest CT showed no hilar or mediastinal lymphadenopathy, ruling against sarcoidosis; an ASO titer and throat culture were both negative for streptococcal infection; stool studies for Yersinia and Campylobacter were negative; an interferon-gamma release assay for tuberculosis was negative; she takes no new medications and isn’t on oral contraceptives, both common drug triggers; and she has none of the gastrointestinal symptoms that would raise inflammatory bowel disease. The episodes have alternated legs, left shin, then right, then left again, which her dermatologist notes argues against any single localized cause. Three episodes, three negative workups — by any standard definition this is now idiopathic recurrent erythema nodosum, not erythema nodosum still awaiting a diagnosis. Each flare has cost her real function: she teaches an evening fitness class twice a week and sat out entirely during two of the three episodes, unable to tolerate the standing the class requires. A fourth began on that same left shin this morning. What is worth being clear about before anyone reaches for a prescription is what the evidence behind every option actually is: no comparator-controlled trial has ever been run in idiopathic recurrent erythema nodosum. The choice ahead is not between a proven treatment and an unproven one, but between two unproven ones, separated by their toxicity rather than by their evidence.
Deciding how hard to treat a diagnosis of exclusion
I’d start potassium iodide with this flare. Horio et al., in Archives of Dermatology in 1981, reported substantial improvement in 11 of 15 patients treated for erythema nodosum, typically within days to a couple of weeks — well ahead of the natural several-week course. Four episodes in eight months, two of which cost her the fitness class she teaches entirely, is no longer a situation where documenting another flare and waiting is a fully satisfying answer.
You’re right that four episodes is a different situation than one — but case-series evidence without a comparator arm can’t distinguish a real drug effect from the same spontaneous resolution erythema nodosum already has most of the time.
Erythema nodosum, idiopathic or not, is characteristically self-limited within three to six weeks regardless of intervention. Potassium iodide carries real, monitorable thyroid-function risk she’d have no clear stopping point for. Starting an agent with that risk profile on case-series evidence, for a condition that clears in a comparable timeframe untreated, trades a known good long-term prognosis for a new and real one.
There’s a middle path that addresses the actual problem — recurrence, not any single flare. Colchicine reduces neutrophil chemotaxis and inflammasome activation, a plausible mechanism in a neutrophilic panniculitis like this. Yurdakul et al.’s 2001 double-blind trial in Behçet’s syndrome is the only randomized evidence anyone has for colchicine against erythema nodosum lesions, and it found a real effect there; closer to her, Hrin et al. published a retrospective case series in 2023 on colchicine in erythema nodosum itself. I’ll say directly that one randomized trial in Behçet’s plus one uncontrolled series is a real evidence gap, not a settled equivalence — but it carries a lower, more familiar monitoring burden than potassium iodide, and it’s aimed at the four-episode pattern itself rather than treating each flare as its own isolated event.
I’d start it with a defined one-year reassessment, treat this flare’s pain with an NSAID regardless of that decision, and hold potassium iodide in reserve rather than adopting it now.
Agreed: naproxen for this acute flare, colchicine started today for recurrence prevention with a defined one-year reassessment point, and no further workup escalation given three prior negative evaluations.
Not agreed, and left as a real open branch rather than smoothed over:
Move to potassium iodide at that point, given colchicine’s evidence base is extrapolated rather than direct.
First confirm colchicine was actually taken as prescribed and consider extending the trial before escalating to an agent with real endocrine risk.