Impact of systemic inflammation and disease activity on the incidence of interstitial lung disease in patients with rheumatoid arthritis - a nested case-control study within the German biologics register RABBIT.

backgroundTo investigate the association between the development of incident interstitial lung disease (ILD) in patients with rheumatoid arthritis (RA) and the disease activity of RA with its various components, especially C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR).

methodsWe analysed data from RA patients, observed in the German biologics register RABBIT between 2001 and 2021. In a nested case-control study, patients with a reported incident ILD diagnosis during follow-up were matched 1:5 to patients without ILD. Matching criteria were sex, age, RA duration, date of enrolment and observation time. Defined by a directed acyclic graph (DAG), we adjusted the conditional logistic regression models for rheumatoid factor, smoking, chronic obstructive pulmonary disease and tuberculosis/chronic viral infections to investigate the impact of disease activity/systemic inflammation. Mean and categorized values were analysed within 12 months prior to ILD and during the entire observation time. Additionally, two sensitivity analyses were performed, using validated ILD cases only and considering ILD cases with an observation time of more than 12 months.

resultsWe identified 139 RA patients with incident ILD and matched them to 686 controls. In 94 cases the diagnosis could be validated, and 98 cases had a follow-up of > 12 months. The averaged DAS28 composite score (including ESR or CRP) was not associated with developing RA-ILD (odds ratios 1.16 [95% confidence interval: 0.97-1.40] and 1.06 [0.86-1.29], respectively). However, single measures of inflammation, log ESR (1.86 [1.35-2.57]) and log CRP (1.55 [1.21-1.97]), were significantly associated with an increased RA-ILD risk. A higher risk for ILD was also revealed for persistently high inflammation. Other DAS28 components showed no significant associations with RA-ILD. These results were consistent for values over the entire observation time of a patient and within 12 months prior to the ILD. Sensitivity analyses confirmed these findings.

conclusionHigher levels of systemic inflammation, as indicated by ESR and CRP, but not joint counts or patient's global assessment, were significantly associated with the occurrence of incident ILD in patients with RA. As possible predictor for the development of RA-ILD, systemic inflammation should be monitored closely and independently of joint count results.