Clinical Factors Associated with hrCT-Confirmed Interstitial Lung Disease in Rheumatoid Arthritis: A Retrospective Case-Control Study.

Background/

Objectives

Rheumatoid arthritis-associated interstitial lung disease (RA-ILD) is a major contributor to morbidity and mortality in RA, yet early recognition remains challenging in routine care. The study aimed to identify clinical factors associated with hrCT-confirmed RA-ILD using a CT-verified case-control design.

Methods

A single-center retrospective case-control study was designed to include RA patients who underwent chest hrCT in routine care. Cases were patients with ILD on index hrCT (n = 79) and controls were RA patients with hrCT negative for ILD (n = 59). Data were manually abstracted from clinical interview, laboratory testing, RA activity and structural assessment, respiratory examination, pulmonary function tests (PFT), chest radiography, and hrCT. Predictors were extracted from the 12 months preceding the index scan. Univariate comparisons used nonparametric tests or χ2, as appropriate. Prespecified multivariable logistic regression estimated adjusted odds ratios (aORs). Sensitivity analyses included restriction to patients with available pre-index PFT, addition of respiratory examination variables, and a matched conditional logistic regression analysis.

Results

In the primary multivariable model, male sex was independently associated with RA-ILD (aOR 5.31, 95% CI 1.91-14.75), and COPD/asthma was also associated (aOR 2.82, 1.05-7.56). Adding dyspnea and Velcro crackles improved discrimination (AUC 0.797 to 0.850); Velcro crackles were independently associated with RA-ILD (aOR 5.11, 1.32-19.73). Findings were directionally similar in sensitivity analyses, though precision decreased in matched models.

Conclusions

In this CT-imaged real-world RA cohort, male sex, COPD/asthma, and Velcro crackles were associated with hrCT-confirmed RA-ILD; these findings should be interpreted as preliminary, as they apply to patients selected for imaging and should not be extrapolated to unselected RA populations without validation in larger, multi-center and/or prospective cohorts with systematic ascertainment.