Utility of Epithelial Cell Biomarkers to Assess Severity and Discriminate Between Different Patterns of Interstitial Lung Disease in Patients with Rheumatoid Arthritis

Abstract

Background Several biomarkers have been studied in rheumatoid arthritis (RA) interstitial lung disease (ILD), but their clinical application has not been well established yet [1]. Given that usual interstitial pneumonia (UIP) is the most common pattern of ILD in patients with RA, the utilization of biomarkers that give information about the epithelial cell activation could be an interesting approach to assess the severity of lung disease in these patients [2].

Objectives To know the ability of CA19-9, CA125 and CEA to discriminate different patterns on HRCT and assess the severity of ILD in patients with RA.

Methods We conducted a cross sectional study. Patients with a diagnosis of RA-ILD who were evaluated in three different rheumatic diseases clinics between December 2021 and January 2023 were consecutively included. The diagnosis of RA was defined according to ACR/EULAR 2010 classification criteria. The diagnosis of ILD was confirmed by an experienced radiologist. We performed HRCT, lung function tests, joint disease evaluation, and blood extraction to determine the values of CA 19-9, CA 125, CEA, rheumatoid factor (RF) and ACPAs. The pattern found on HRCT was classified in UIP pattern, probable UIP, indeterminate for UIP, and suggestive of another diagnosis, as proposed the inter society consensus of 2022 [4]. The score proposed by Goh was calculated to determine the extension of lung opacities (inflammatory, fibrotic, and total extension scores) [5]. The joint disease activity was assessed using the DAS28 score. The values of the biomarkers in serum were measured by electrochemiluminescence.

T-test, Mann Whitney test, and Chi square were used for comparisons. The Spearman test was used for correlation analysis.

Results We included 38 patients, 29 were women (76.3%). The mean (SD) age was of 62.1 (11.5) years.

The median (IQR) of DAS28 was 3.3 (2.6-4). Regarding serologic tests, 35/36 (97.2%) and 36/37 (97.3) were positive for ACPAs and RF respectively.

In relation to lung disease, the mean (SD) of FVC% was 83.8 (23.4) and the median (IQR) DLCO was 59 (52-81). Eighteen patients (47.3%) showed UIP or probable UIP pattern in the HRCT, while 20 (52.7%) had another pattern. The median (IQR) of Goh score for total lung disease extension was 20 (8-25).

Patients with UIP pattern showed significantly higher values of CA19-9 [11.5 (7-35.7) vs 6.5 (2-12.5), p=0.017] and a trend to higher values of CA125 [22 (15-52) vs 14 (11.4-20.7), p=0.06] than those with other patterns. The values of CA19-9 showed a good correlation with total Goh score (r=0.52, p=0.02) and fibrosis Goh score (r=0.65, p=0.03) in patients with UIP pattern. Also, the values of CEA showed an acceptable correlation with total Goh score (r=0.46, p=0.05) and fibrosis Goh score (r=0.43, p=0.07) in patients with UIP pattern.

Conclusion Patients with RA ILD and UIP pattern showed higher plasmatic values of biomarkers that reflect epithelial cell activation. Also, the values of some of these biomarkers showed good correlation with the extension of the disease on HRCT. Therefore, these biomarkers could be a useful tool to identify a more aggressive clinical behavior in patients with RA ILD.