Risk prediction model for mortality in microscopic polyangiitis: multicentre REVEAL cohort study

Abstract

Background To establish refined risk prediction models for mortality in patients with microscopic polyangiitis (MPA) by using comprehensive clinical characteristics. Methods Data from the multicentre Japanese registry of patients with vasculitis (REVEAL cohort) were used in our analysis. In total, 194 patients with newly diagnosed MPA were included, and baseline demographic, clinical, laboratory, and treatment details were collected. Univariate and multivariate analyses were conducted to identify the significant risk factors predictive of mortality. Results Over a median follow-up of 202.5 (84–352) weeks, 60 (30.9%) of 194 patients died. The causes of death included MPA-related vasculitis (18.3%), infection (50.0%), and others (31.7%). Deceased patients were older (median age 76.2 years) than survivors (72.3 years) (P < 0.0001). The death group had shorter observation periods (median 128.5 [35.3–248] weeks) than the survivor group (229 [112–392] weeks). Compared to survivors, the death group exhibited a higher smoking index, lower serum albumin levels, higher serum C-reactive protein levels, higher Birmingham Vasculitis Activity Score (BVAS), higher Five-Factor Score, and a more severe European Vasculitis Study Group (EUVAS) categorization system. Multivariate analysis revealed that higher BVAS and severe EUVAS independently predicted mortality. Kaplan–Meier survival curves demonstrated lower survival rates for BVAS ≥20 and severe EUVAS, and a risk prediction model (RPM) based on these stratified patients into low, moderate, and high-risk mortality groups. Conclusions The developed RPM is promising to predict mortality in patients with MPA and provides clinicians with a valuable tool for risk assessment and informed clinical decision-making.