Erythromelalgia: Identification of a corticosteroid-responsive subset.

Abstract

Corticosteroids (CS) may benefit certain patients with erythromelalgia. Our objective was to determine clinical predictors of corticosteroid-responsive erythromelalgia. Patients with erythromelalgia who received CS were identified and stratified into corticosteroid nonresponders (NRs), partial corticosteroid responders (PSRs), complete corticosteroid responders (CSRs), and steroid responders (SRs = PSRs + CSRs). In the study variable analysis, P < .05 was considered statistically significant. The median (interquartile range) age of the 31-patient cohort was 47 years (26-57 years), and 22 (71%) were female. Fourteen (45%) were NRs, 17 (55%) SRs, 8 (26%) PSRs, and 9 (29%) CSRs. A subacute temporal profile to disease zenith (<21 days) was described in 15 (48%) patients, of whom 13 (87%) were SRs (P = .003; odds ratio [OR] = 0.069 [95% confidence interval {CI}, 0.011-0.431]). Six (67%) CSRs reported a disease precipitant (eg, surgery, trauma, or infection; P = .007; OR = 12.667 [95% CI, 2-80.142]). SR patients received CS sooner than NR at 3 (3-12) versus 24 (17-45) months (P = .003). A high-dose CS trial (≥200 mg prednisone cumulatively) was administered to 17 (55%) patients, of whom 13 (76%) were SRs (P = .012; OR = 8.125 [95% CI, 1.612-40.752]). This was a retrospective case series. An infectious, traumatic, or surgical precipitant and subacute presentation may portend CR erythromelalgia. A transient "golden window" where CS intervention is useful may exist before irreversible nociceptive remodeling and central sensitization occurs.