Misdiagnosis of Still's Disease Save

Know-it-now

• Diagnosis of Still's disease by classification criteria is just a start, not a final diagnosis.
• Non-response to cytokine inhibitors and glucocorticoids should lead to reconsideration of your initial "Still's disease" diagnosis.
• Specific red flags suggesting it's not Stills: onset before age 2, positive family history, atypical rash (panniculitis-like, non-evanescent), persistent organomegaly or cytopenias, coronary artery abnormalities, interstitial lung disease, and recurrent abdominal pain. These were featured in nearly every misdiagnosed case in this report.
• Diagnostic reassessment must be iterative. sJIA/Still's disease requires longitudinal reassessment across the disease course, just like seronegative rheumatoid arthritis. 

High fever, strange rashes, scary labs often lead to a consideration of Still's disease (AOSD or systemic JIA). Yet the dramatic onset of a systemic disorder may overlap with another, and possibly as rare, systemic disorder that should not be misdiagnosed or mistreated.  A current case review by Zhu et al. presents six cases initially diagnosed as systemic JIA, teaching several important lessons.

Systemic juvenile idiopathic arthritis (sJIA) (now preferably called Still's disease - the child or adult), is a diagnosis of exclusion. No pathognomonic biomarker exists, and classification criteria (ILAR, PRINTO, EULAR/PReS) were never intended to substitute for ongoing clinical judgment. A new retrospective case series from China reports on six pediatric patients who fulfilled sJIA classification criteria but were ultimately reclassified with alternative diagnoses, after a median diagnostic delay of 9.5 months. All six had fever; five had rash; half had arthritis. Three were labeled "refractory sJIA" before revision.

Patient 1 presented at 14 months with fever, rash, hepatosplenomegaly, and striking bilateral interstitial/parenchymal lung disease. Persistent hyperferritinemia and thrombocytopenia despite HLH-2014 therapy (dexamethasone/etoposide), with no identified pHLH mutation, ultimately supported suspected familial HLH. The missed clue: very early onset plus a poor response to HLH-directed therapy in a phenotype that never behaved like classic sJIA.

Patient 2 had fever, rash, lymphadenopathy, and recurrent oral ulcers from 6 months of age with a positive maternal family history - features present from the start but initially subordinated to the systemic inflammatory picture. Recurrence after stopping tocilizumab, combined with the ulcer history and family history, led to revision to PFAPA (Periodic Fever, Aphthous stomatitis, Pharyngitis, and Adenitis or the "Marshall Syndrome") - there is no genetic test for this.

Patient 3 had fever and panniculitis-like (not evanescent) rash, plus mild coronary dilatation at presentation; an atypical rash morphology and a vascular finding not explained by sJIA. Progression of coronary dilatation during steroid tapering, with vasculitis on follow-up ultrasound, confirmed Takayasu arteritis.

Patient 4, the oldest at onset (135 months) and the only female, had knee pain, anemia, hepatomegaly, and abdominal symptoms. The patient was glucocorticoid and tocilizumab-refractory. Abdominal CT eventually identified a hypervascular omental mass; resection confirmed inflammatory myofibroblastic tumor. The lesson: unexplained abdominal pain and treatment-refractory anemia warranted imaging much earlier.

Patient 5 was the youngest at true disease onset (0.3 months, though diagnosed with sJIA at 24 months), with migratory arthritis, prior suppurative appendicitis, and recurrent abdominal pain/obstruction — a history suggesting a monogenic autoinflammatory process from infancy. Difficulty tapering glucocorticoids and recurrent hyperferritinemia prompted genetic testing, revealing an NLRP12 variant (NLRP12-AID).

Patient 6 presented at 145 months with fever, rash, arthritis, and lymphadenopathy, and developed MAS. Compound heterozygous MVK variants with functionally confirmed reduced enzymatic activity established mevalonate kinase deficiency (MKD) — a reminder that abdominal pain and early-pattern periodicity can hide within an apparently classic sJIA presentation.