Developing biomarkers for predicting clinical relapse in pemphigus patients treated with rituximab

Abstract

Rituximab is an effective therapy for pemphigus, although relapses are common. To identify biomarkers to predict relapse of pemphigus following rituximab treatment. In this retrospective cohort study, 62 patients with pemphigus treated with 99 rituximab cycles provided longitudinal clinical scoring and biomarker data, including levels of CD19+ B cells, CD4+ T cells, and desmoglein 1 (Dsg1) and desmoglein 3 (Dsg3) autoantibodies. An extended time-variant Kaplan-Meier estimator and extended Cox model were applied. Relapse was rare before B-cell repopulation. Univariate analysis revealed low CD4 count (<400cells/μL) to predict relapse (P<.001). A positive result of testing for Dsg1 (>20 IU) was predictive of relapse among patients with mucocutaneous disease (hazard ratio, 6.40; P=.019); a positive result of testing for Dsg3 (>20 IU) was predictive in patients with mucocutaneous and mucosal disease (hazard ratio, 32.92; P<.001). Multivariable analysis revealed that every CD4 value increase of 200 decreases the hazard ratio for relapse by 35% (P=.029). A positive result of testing for Dsg1 increases the risk for relapse by a factor of 12.32 in patients with mucocutaneous disease (P=.001); positive result of testing for Dsg3 increases risk for relapse by 28.38 in patients with mucosal and mucocutaneous disease (P=.006). Limitations include the retrospective design and inconsistent follow-up. Relapse is associated with B-cell repopulation, low CD4+ T -cell count, and positive result of testing for Dsg1 and Dsg3.