Prediction model for respiratory-related mortality in microscopic polyangiitis with interstitial lung disease: multicenter REVEAL cohort study.

This study aimed to establish prediction models for respiratory-related mortality in microscopic polyangiitis (MPA) complicated by interstitial lung disease (ILD) using clinical characteristics. We enrolled patients with MPA with ILD between May 2005 and June 2021 in a multicentre cohort of Japanese patients with MPA (REVEAL cohort). We evaluated the demographic, clinical, laboratory, radiological findings, treatments, and the presence of honeycombing 1 cm above the diaphragm using chest high-resolution computed tomography (HRCT) on admission. We explored the risk factors predictive of respiratory-related mortality. Of 115 patients, 26 cases died of respiratory-related diseases during a median follow-up of 3.8 years. Eighteen patients (69%) died due to respiratory infection, three (12%) had diffuse alveolar hemorrhage (DAH), and five (19%) had exacerbation of ILD. In univariate analysis, older age, lower percent forced vital capacity (%FVC), lower percent diffusing capacity of carbon monoxide (%DLco), and the presence of honeycombing in the right lower lobe were identified as risk factors. Additionally, in multivariate analysis adjusted for age and treatment, %FVC, %DLco, and the presence of honeycombing in the right lower lobe were independently associated with respiratory-related mortality. We created prediction models based on the values of %FVC, %DLco, and presence of honeycombing on chest HRCT (MPF model). The 5-year respiratory-related death-free rate was significantly different between patients with MPA with ILD stratified by the number of risk factors based on the MPF model. Our study indicates that the MPF model may help predict respiratory-related death in patients with MPA with ILD.

Informations de copyright

© The Author(s) 2023. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For permissions, please email: [e-mail].