Combining Three Peripheral Blood Biomarkers to Stratify Rheumatoid Arthritis-Associated Interstitial Lung Disease Risk.

Abstract

The purpose was to evaluate a biomarker score consisting of MUC5B rs35705950 promoter variant, plasma matrix metalloproteinase-7 (MMP-7), and serum anti-malondialdehyde-acetaldehyde (anti-MAA) antibody for rheumatoid arthritis (RA)-associated interstitial lung disease (ILD) risk stratification. Using a multicenter cohort of US veterans with RA, we performed a cross-sectional study of prevalent RA-ILD and a cohort study of incident RA-ILD. Medical records were used to confirm clinical ILD diagnoses by chest imaging or lung biopsy pathology reports. A combined three-biomarker score (range 0-3) was calculated based on the presence of the MUC5B variant and elevations (upper 25% vs lower 75%) in MMP-7 and anti-MAA antibody concentrations. Multivariate logistic and Cox regression models were adjusted for clinical risk factors. Among 2,043 participants with RA, prevalent ILD was identified in 88 (88.7% male; mean age 63.6 years). The odds of prevalent RA-ILD were higher with increased biomarker score (adjusted odds ratio 12.21 [95% confidence interval (CI) 3.82-38.97] for score of 3 vs 0). A score ≥1 had 76.1% sensitivity but 48.6% specificity. Incident RA-ILD developed in 148 participants, with those having a combined biomarker score of 3 having the highest risk (adjusted hazard ratio 4.36 [95% CI 1.55-12.27]). The area under the curve for prevalent RA-ILD and Harrell's C for incident RA-ILD were highest when clinical risk factors were combined with the biomarker score. This three-analyte biomarker score was associated with both prevalent and incident RA-ILD, improving risk stratification beyond clinical risk factors. Although this score is inadequate for clinical implementation, these findings demonstrate the potential for biomarker scores in RA-ILD risk stratification.