Recurrence and relapse in nonsegmental vitiligo: phenotypic and autoimmune predictors from a large retrospective cohort.

Abstract

Nonsegmental vitiligo frequently recurs after treatment discontinuation despite successful repigmentation. Identifying patients at higher risk of recurrence/relapse is essential for guiding long-term management and maintenance strategies. To identify phenotypic and autoimmune predictors of recurrence/relapse in nonsegmental vitiligo, and to establish a clinically relevant risk-stratification framework. We conducted a retrospective cohort study using the Chang Gung Research Database (2015-2024). Patients with ≥ 6 months of documented clinical stability and ≥ 1 year of follow-up were included. Time from stability to first recurrence/relapse was analysed using multivariable Cox proportional hazards models adjusted a priori for age, sex, acral and facial involvement, baseline body surface area category (< 5%, 5-10%, > 10%), thyroid disease, other autoimmune diseases, anti-thyroid peroxidase, antinuclear antibody levels and family histories. A joint-exposure model assessed combined acral involvement and thyroid disease. Among 809 patients, 266 (32.9%) developed recurrence/relapse. Independent predictors included acral involvement [adjusted hazard ratio (aHR) 1.56, 95% confidence interval (CI) 1.20-2.03; P < 0.001], thyroid disease (aHR 1.42, 95% CI 1.07-1.89; P = 0.015) and other autoimmune comorbidities (aHR 1.73, 95% CI 1.23-2.41; P = 0.001). Joint-exposure analysis showed that patients with both acral involvement and thyroid disease had significantly elevated hazard (aHR 2.44, 95% CI 1.64-3.59; P < 0.001); whereas each factor alone conferred smaller or nonsignificant effects. Acral involvement, thyroid disease and other autoimmune comorbidities are key predictors of recurrence/relapse in nonsegmental vitiligo. Combined acral and thyroid involvement identifies a particularly high-risk subgroup with approximately 2.5-fold higher hazard. Incorporating these phenotypic and autoimmune risk factors into clinical assessment may support risk-stratified maintenance planning and more individualized long-term monitoring.