Clinical and serological predictors of relapse in pemphigus: a study of 143 patients

Abstract

Background: Pemphigus is an autoimmune bullous disease mediated by autoantibodies targeting epithelial cell–cell adhesion molecules. Predictors of relapse have not yet been clearly identified. Aims: To identify factors at diagnosis and during follow-up that could be predictors of relapse. Methods: Clinical and immunopathological data at diagnosis, clinical remission and first relapse from patients with pemphigus vulgaris or foliaceus and at least a 36-month follow-up were collected retrospectively. Based on the autoantibody profile at diagnosis, three serological patient subsets were devised: (i) anti-desmoglein (Dsg)1-positive and anti-Dsg3-negative; (iii) anti-Dsg1-negative and anti-Dsg3-positive; and (iii) anti-Dsg1-positive and anti-Dsg3-positive. Results: Data from 143 patients were collected. No significant differences were found between relapsers (n = 90) and nonrelapsers (n = 53) for time to remission or for anti-Dsg1 and anti-Dsg3 titres at diagnosis and remission. In the analysis of all patients, a higher risk of relapse was found for a body surface area (BSA) score of 3 compared with BSA < 3 (OR = 3.30, 95% CI 1.17–9.28; P = 0.02) and for a positive titre of either anti-Dsg1 or anti-Dsg3 autoantibodies at remission compared with both being negative (OR = 2.42, 95% CI 1.21–4.85, P = 0.01). In patients who were anti-Dsg3-positive and anti-Dsg1-negative at diagnosis, failure to achieve anti-Dsg3 negativity at clinical remission was a significant predictor of relapse (OR = 7.89, 95% CI 2.06–30.21; P < 0.01). Similarly, failure to achieve anti-Dsg1 negativity at clinical remission was a significant predictor of relapse in patients with both anti-Dsg1 and anti-Dsg3 positivity at diagnosis (OR = 5.74, 95% CI 1.15–28.61; P = 0.03), but not in those who were anti-Dsg1-positive/anti-Dsg3-negative at diagnosis (OR = 1.08, 95% CI 0.27–4.30; P = 0.91). Conclusion: Regardless of pemphigus subtype, autoantibody titre negativity at clinical remission in patients classified based on their anti-Dsg1 and anti-Dsg3 profile at diagnosis and BSA were useful tools in predicting relapse.