Anti-Ro52/TRIM21 seropositivity in interstitial lung disease: effect modification by disease subtype and myositis-specific antibody context

Anti-Ro52/TRIM21 shows contradictory prognostic associations in interstitial lung disease (ILD), protective in some cohorts and adverse in others. We evaluated whether it acts as an independent predictor or a phenotypic marker and whether its effect is modified by disease subtype and myositis-specific antibodies.

We retrospectively identified 538 adults with radiologically confirmed interstitial lung disease (228 anti-Ro52-positive) from a 2010-2020 single-center cohort. Cox proportional-hazards models were sequentially adjusted for demographic, smoking, malignancy, and pulmonary-function covariates. Pre-specified effect modification by connective tissue disease (CTD) status and myositis-specific antibodies was assessed by likelihood-ratio interaction tests with false-discovery-rate correction. Robustness to missing data was supported by multiple imputation (m = 20) and a missing-not-at-random (MNAR) tipping-point analysis.

Ro52 positivity was associated with younger age, CTD predominance, and lower complement levels. Pre-specified effect modification was suggested: anti-Ro52 trended toward lower mortality in CTD-ILD (HR 0.73, 95% CI 0.49-1.07, P = 0.104) but did not independently increase mortality within the anti-MDA5-positive subgroup (HR 1.73, 95% CI 0.66-4.54). Instead, the adverse prognostic effect of anti-MDA5 was numerically greater among anti-Ro52-positive than Ro52-negative patients (MDA5 HR 10.57, 95% CI 3.60-31.03 vs. 2.56, 95% CI 0.97-6.74), although the formal interaction term did not reach statistical significance (P = 0.055). The fully adjusted (Model 6) HR was 0.72 (95% CI 0.52-0.99; P = 0.045); further adjustment for CTD status (Model 7) attenuated the association to non-significance (HR 0.75, 95% CI 0.54-1.04, P = 0.091), and IPCW weighting and time-varying-age analyses were likewise non-significant; malignancy independently predicted mortality (HR 1.81, 95% CI 1.15-2.88; P = 0.012).

Anti-Ro52 is a context-dependent phenotypic marker of immune-mediated ILD, not a universal prognostic biomarker. Its prognostic association is heterogeneous across ILD phenotypes: it is not independently protective in CTD-ILD (adjusted HR 0.73, P = 0.104), whereas the adverse prognostic effect of anti-MDA5 was numerically greater among anti-Ro52-positive than Ro52-negative patients (MDA5 HR 10.57 vs. 2.56), although this interaction did not reach statistical significance (P = 0.055); CTD-status adjustment similarly attenuated the Ro52 association (HR 0.75, P = 0.091). These findings support stratified prognostic models integrating Ro52 with autoantibody and disease-subtype context.