Background/Purpose: Rheumatoid arthritis-associated interstitial lung disease (RA-ILD) is one of the most serious extra-articular RA manifestations and is more common in older patients. Registry-based RA-ILD studies have been limited to investigate incident RA-ILD and cause-specific mortality due to small sample size. Previous nationwide studies were limited due to the previous lack of a validated RA-ILD algorithm in administrative data. Therefore, we aimed to investigate prevalence, incidence, and cause-specific mortality of RA-ILD using a recently validated claims-based algorithm.

Methods: We performed a retrospective cohort study using US nationwide claims data from Medicare (2008-2017). RA was identified using a previously validated algorithm (2+ ICD-9/10 codes for RA separated by 7+ days and DMARD prescription [PPV 86%]; RA date [index date] was the date these criteria were met). RA-ILD was identified using a recently validated algorithm (2+ ICD-9/10 codes for ILD by a rheumatologist or pulmonologist separated by 7+ days [PPV 72%]; ILD date was the 2nd ILD code). We identified prevalent RA-ILD and covariates as of 365 days prior to index date. Among RA without ILD at baseline, Cox regression estimated HRs for incident RA-ILD by baseline covariates. We compared the risk for total mortality between patients with RA-ILD to RA without ILD using multivariable Cox regression adjusting for baseline covariates. For cause-specific mortality, Fine and Gray subdistribution hazard ratios (sdHR) were estimated to handle competing risks of alternative causes of mortality.

Results: Among a total of 509,787 patients with RA in Medicare (mean age 72.6 years; 76.2% female), 10,306 (2.0%) had prevalent RA-ILD at initial RA observation. Among RA without ILD at baseline, 13,372 (2.6%) developed RA-ILD during 1,873,127 person-years of follow-up (median 3.0 years/person). The incidence rate of RA-ILD was 7.14 per 1,000 person-years. Several baseline factors were associated with incident RA-ILD: male sex (HR 1.31, 95%CI 1.26-1.36), smoking (HR 1.35, 95%CI 1.29-1.41), biologic/targeted DMARD use (HR 1.34, 95%CI 1.29-1.40), and glucocorticoid use (HR 1.45, 95%CI 1.39-1.50, Table 1). During follow-up, 38.7% of RA-ILD died compared to 20.7% of RA without ILD (unadjusted HR 2.36, 95%CI 2.28-2.45). After multivariable adjustment that included confounders and possible mediators (such as comorbidities after RA-ILD onset), the association of RA-ILD with total mortality remained significant (HR 1.66, 95%CI 1.60-1.72, Table 2). Accounting for competing risk of other causes of death, RA-ILD had a sdHR of 4.39 (95%CI 4.13-4.67) for respiratory mortality and a sdHR of 1.56 (95%CI 1.43-1.71) for cancer mortality compared to RA without ILD.

Conclusion: RA-ILD was present or developed in nearly 5% in this nationwide study of older patients with RA. RA-ILD was associated with excess total mortality that was not explained by measured factors. Male sex, smoking, biologic/targeted DMARD use, and glucocorticoid use were associated with incident RA-ILD. RA-ILD was strongly associated with increased respiratory mortality compared to RA without ILD. The novel association of RA-ILD with increased cancer mortality requires further investigation.

Table 1. Incidence rates and hazard ratios (HRs) for incident RA-ILD by selected characteristics among patients with RA and no prevalent ILD at initial observation in Medicare (n=499,481).

Table 2. Hazard ratios (HRs) for total mortality and subdistribution hazard ratios (sdHRs) for cause-specific mortality comparing RA-ILD to RA without ILD in Medicare (n=509,787).

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/prevalence-incidence-and-cause-specific-mortality-of-rheumatoid-arthritis-associated-interstitial-lung-disease-among-older-patients-with-rheumatoid-arthritis-a-nationwide-cohort-study/