Months of Leg Nodules and No Biomarker to Say Whether They're the Whole Story
A single patient with several months of cutaneous polyarteritis nodosa and no systemic findings. The disagreement isn't about what to give her skin — it's about how hard to keep watching for a disease that hasn't announced itself yet, and may never.
Marisol V., 35, freelances as a graphic designer from a home office she's decorated with framed prints of her own work, and has spent the better part of this year also documenting, in a notebook she brought to this visit, every new tender nodule and ulcerating patch that's appeared on her lower legs — a habit that turned out to be genuinely useful once the rheumatologist started asking how long this had actually been going on. Several months, by her own careful count, recurring and slow to heal, sometimes overlapping before the last one finished. A skin biopsy found medium-vessel necrotizing vasculitis, the pattern that defines cutaneous polyarteritis nodosa — a real vasculitis, not merely inflamed skin, but one classically confined to the skin and nothing else. Her workup so far agrees with that limited picture: no fevers, no weight loss, no neuropathy, no abdominal pain, normal renal function, a normal cardiac exam, ANCA and hepatitis serologies both negative. By every marker checked, this looks like exactly what its name says: cutaneous, and only cutaneous. The one lab drawn with a specific drug already in mind is her G6PD, sent before anyone wrote for dapsone and returned normal — the screen that has to precede a sulfone, since dapsone's hydroxylamine metabolite drives oxidant hemolysis and a deficiency found afterward is found too late.
That's also the honest natural history of this disease in the large majority of cases. Daoud and colleagues (Br J Dermatol, 1997) followed 79 patients with cutaneous PAN and reported a course that was prolonged but benign, with systemic PAN developing in none of them — which cuts two ways for Marisol. It means aggressive systemic immunosuppression would be treating a risk that, for most patients, never materializes. It also means her several months of recurring, slow-to-heal lesions are not the outlier they feel like: a protracted relapsing course is the described norm here, not a warning sign, and reading her duration as evidence of impending progression gets the natural history backwards. Where the concern legitimately lives is on a much longer horizon — Morgan and Schwartz's review notes a series in which two of twenty patients did develop systemic PAN, at eighteen and nineteen years of follow-up. There is no validated biomarker — no antibody, no imaging finding, no laboratory threshold — that identifies at any single visit which patient that will be. So the question her notebook actually raises is not whether these months mean something, but how many years anyone intends to keep watching.
Treating what's there, watching for what isn't yet
Everything about her presentation matches the well-described, reassuring natural history of cutaneous PAN — it overwhelmingly stays cutaneous, and systemic immunosuppression carries real toxicity that isn't justified by a low-probability future risk. I'd manage this with dapsone and topical measures, the standard approach for limited disease, and not treat her as though systemic PAN were already present when nothing on her workup supports that.
I agree with the treatment plan for what's actually in front of us — I'd just push back on how much reassurance to take from the 'single-organ' label itself. That's a description of what's been observed so far, not a guarantee of what happens next. I'll grant that her several months don't themselves mean anything — Daoud's series describes exactly this kind of protracted relapsing course as the norm, so I'm not going to argue her duration is a red flag. I'd want periodic labs and a targeted review of systems at defined intervals, not because I think she has systemic disease today, but because missing an evolving case would have real stakes.
What I am proposing is that the watching outlast the treating. The progressions that have been reported turned up at eighteen and nineteen years, which is an argument for a surveillance horizon measured in decades, not for anything different today. That's a different ask than escalating her treatment.
I'd name directly why this doesn't resolve into a clean disagreement: there's no validated biomarker that distinguishes limited from evolving-systemic cutaneous PAN prospectively, so the actual decision in front of us isn't a treatment-escalation question at all right now — it's a surveillance-intensity question. Treat what's actually present with dapsone — her G6PD came back normal, which is the check that had to clear before a sulfone regardless of how limited the disease is, and I'd still recheck a count at two and four weeks, since dapsone-induced hemolysis is reported in patients with normal G6PD levels too. Match therapy to disease as it currently exists rather than to a feared future version of it, and pair that with a defined surveillance interval rather than either dismissing the duration concern or treating it as grounds for immunosuppression she doesn't yet need.
Agreed on treatment: start dapsone for the cutaneous disease, matching therapy to what's actually present, with the G6PD screen already back normal and blood counts rechecked at two and four weeks.
Explicitly not agreed, and left open rather than smoothed over: how long surveillance should continue, and what specifically would count as evidence of systemic progression significant enough to change the treatment plan. No biomarker exists to settle it, and the room did not pretend one did.