Rheumatoid arthritis-related interstitial lung disease: associations, prognostic factors and physiological and radiological characteristics--a large multicentre UK study.

OBJECTIVES: The prevalence of interstitial lung disease (ILD) in RA is ∼5%. Previous work identified increasing age, active articular disease and articular damage as risk factors for RA-associated ILD (RA-ILD). The roles of high-resolution CT (HRCT) and lung function testing in defining the nature and extent of pulmonary involvement have recently been explored. This study is the first to examine predictive and prognostic factors for the development of RA-ILD and to report on the physiological and radiological characteristics of the condition from a large multicentre UK network. METHODS: We collected data from centres across the UK on patients with both RA and ILD (proved on HRCT) diagnosed over a 25-year period from 1987 to 2012 using a standard pro forma. Potential predictors of RA-ILD were analysed. Baseline lung function data were recorded and related to HRCT findings. We analysed HRCT for subtype and extent of lung involved and examined the relationship between these and both all-cause and pulmonary mortality. We compared our results with case controls matched for age and gender using computer-generated selection from the RA population from one contributing centre. RESULTS: A total of 230 patients were identified from across the UK with proven RA-ILD diagnosed over 25 years. Median age at diagnosis was 64 years and the male:female ratio was 1:1.09. Univariate analysis showed anti-CCP antibody titres to be the single most strongly associated predictor of RA-ILD. Male gender, age at onset, smoking and RF were all independently associated with RA-ILD on multivariate analysis. Vital capacity (VC) was preserved in limited disease but reduced in extensive disease, while gas transfer was reduced in both. Usual interstitial pneumonia (UIP) was the most common subtype on HRCT and both this and extensive disease were associated with increased all-cause mortality. CONCLUSION: This is the largest study of RA-ILD in the UK. Anti-CCP antibodies were strongly associated with RA-ILD in both sexes. Smoking was strongly associated with ILD in males, which may explain the higher frequency of RA-ILD in men. The predominant HRCT pattern was UIP and most patients had limited disease at presentation. The presence of UIP and extensive disease are associated with increased mortality. Baseline gas transfer is a useful screening tool for ILD, while the preservation of VC at baseline might predict limited disease on HRCT.