# Performance of an Idiopathic Pulmonary Fibrosis–Derived Multibiomarker Panel for Rheumatoid Arthritis–Associated Interstitial Lung Disease

## TL;DR

A blood biomarker panel linked to idiopathic pulmonary fibrosis was found to also predict interstitial lung disease in rheumatoid arthritis patients, suggesting shared disease mechanisms.

## Contribution

A multibiomarker panel originally associated with idiopathic pulmonary fibrosis is shown to predict rheumatoid arthritis-associated interstitial lung disease.

## Key findings

- RA patients with ILD had significantly higher IPF multibiomarker scores than those without ILD.
- High multibiomarker scores were associated with a 2.14-fold increased odds of prevalent ILD and a 2.45-fold increased risk of developing ILD over 15 years.
- The cumulative hazard of incident ILD reached nearly 20% in the highest biomarker quartile by 15 years.

## Abstract

To assess whether a panel of peripheral blood biomarkers associated with idiopathic pulmonary fibrosis (IPF) is also associated with interstitial lung disease (ILD) in patients with rheumatoid arthritis (RA) using three independent cohorts.

We first assessed the association of a panel of IPF‐associated biomarkers with prevalent ILD among two separate RA cohorts (n = 93 and n = 71). Concentrations of eight IPF‐related biomarkers (eotaxin, Flt‐3L, interleukin‐8, macrophage‐derived chemokine, monocyte chemoattractant protein 1, and matrix metalloproteinase 2/7/9 [MMP‐2/7/9]) were measured, standardized, and summed to generate a multibiomarker score. We subsequently validated the association of this score (minus MMP‐2) with prevalent and incident ILD in an independent multicenter prospective cohort of US veterans with RA (n = 2,507). Multivariable regression models were adjusted for relevant covariates in the validation cohort.

In both development cohorts, participants with RA‐ILD had significantly higher IPF multibiomarker scores than those with RA alone. In the independent validation cohort, participants in the highest quartile of multibiomarker scores had a significantly higher likelihood of prevalent ILD (adjusted odds ratio 2.14 [95% confidence interval 1.18–3.87]) and incident ILD (adjusted hazard ratio 2.45 [95% confidence interval 1.55–3.88]) than those in the lowest quartile. The cumulative hazard of incident ILD approached 20% by 15 years for those in the highest quartile, compared to <10% for all other quartiles.

A multibiomarker panel derived from IPF‐associated biomarkers was associated with RA‐ILD in separate development and validation cohorts. This overlap supports the concept of shared etiopathogenesis of IPF and RA‐ILD and illustrates the potential for peripheral blood biomarker panels to stratify ILD risk among patients with RA.