Clinical characteristics and pulmonary-cause mortality in rheumatoid arthritis-associated interstitial lung disease by lung-onset status.

OBJECTIVES: Interstitial lung disease (ILD) is a major extra-articular manifestation of rheumatoid arthritis (RA), but the significance of pulmonary manifestations preceding or coinciding with articular disease remains unclear. This study aimed to compare the clinical and prognostic characteristics of RA-associated ILD (RA-ILD) stratified by lung-onset status. METHODS: In this single-center retrospective cohort study, 1,344 hospitalized patients with confirmed RA between 2008 and 2020 were screened and 192 RA-ILD patients with complete pulmonary function, HRCT, and follow-up data were included. Patients were classified into lung-onset and non-lung-onset groups based on the temporal relationship between ILD recognition and RA diagnosis. Clinical, pulmonary function, and HRCT characteristics were compared. Pulmonary-cause mortality was analyzed using Fine-Gray competing-risk regression, with cause-specific Cox regression as a complementary analysis. RESULTS: Among the 192 patients, 51 (26.6%) were classified as lung-onset RA-ILD. Respiratory manifestations and radiological disease burden were more prominent in the lung-onset group. TLC (72.8% vs. 86.3% predicted) and DLCO (50.9% vs. 57.5% predicted) were lower in the lung-onset group. On baseline HRCT, definite usual interstitial pneumonia (UIP) pattern (25.5% vs. 12.8%) and fibrotic nonspecific interstitial pneumonia (f-NSIP) pattern (17.6% vs. 7.8%) were more frequent. Total ILD extent (21.67% vs. 8.33%) was higher, and extensive involvement (total ILD extent > 30%) (27.5% vs. 7.8%) was more common in the lung-onset group. On follow-up CT scans, fibrotic abnormalities increased, whereas GGO and visually estimated total ILD extents decreased in both groups. Among 82 deaths, 49 were pulmonary-cause deaths. Lung-onset status was associated with pulmonary-cause mortality in the multivariable Fine-Gray model (subdistribution hazard ratio 3.704, 95% confidence interval 2.054-6.681), with consistent findings in the cause-specific Cox analysis, but was not significantly associated with all-cause mortality. CONCLUSIONS: Lung-onset RA-ILD was associated with more severe pulmonary functional impairment, greater radiological disease burden, and a higher cumulative incidence of pulmonary-cause mortality. Early recognition of RA in patients presenting with unexplained ILD may facilitate pulmonary risk stratification and longitudinal follow-up. Key Points • We designed this study to identify the clinical and prognostic features of patients with lung-onset RA-ILD. • Lung-onset RA-ILD was associated with more obvious respiratory symptoms, more severe pulmonary functional impairment and more extensive HRCT disease burden. • Patients with lung-onset RA-ILD had a higher cumulative incidence of pulmonary-cause mortality, and lung-onset status was associated with pulmonary-cause mortality in both Fine-Gray and cause-specific Cox analyses, but not with all-cause mortality. • Early recognition of RA in patients presenting with unexplained ILD may improve pulmonary risk stratification and management.