Justin Cooper , 2025-04-30 14:58:00
April 30, 2025
3 min read
Key takeaways:
- Eight clusters of related biomarkers were significantly associated with RA-ILD.
- The biomarkers led to a greater jump in accuracy vs. “the strongest genetic risk factor identified to date.”
Biomarkers in blood and plasma can be used to predict rheumatoid arthritis-associated interstitial lung disease more accurately than clinical risk factors alone, suggesting potential new opportunities in screening and treatment, data show.
“One of our main focuses is how to predict which patients with rheumatoid arthritis are at highest risk for developing interstitial lung disease (ILD), which is a major cause of death in this population,” Austin M. Wheeler, MD, a rheumatologist at the University of Nebraska Medical Center, told Healio. “We’ve also been looking at clinical and genetic risk factors, but circulating proteins in the blood provide us with another method to risk-stratify these patients, and provide some new insights on the different biologic pathways involved in RA-ILD.”
To assess RA biomarker signatures in the peripheral blood that could potentially improve RA-ILD risk stratification, Wheeler and colleagues analyzed data on 2,001 patients (mean age, 63.7 years; 88.7% men) from the Veterans Affairs Rheumatoid Arthritis registry. Patients’ serum or plasma samples were assessed for various peripheral biomarkers, including autoantibodies, pro-inflammatory cytokines/chemokines, adipokines, alarmins and matrix metalloproteinases.
The researchers also assessed the presence of “the strongest genetic risk factor identified to date” for RA-ILD: the MUC5B rs35705950 promoter variant, the researchers wrote. The biomarkers were organized and investigated collectively using principal component analysis.
“Principal components analysis takes the large number of biomarkers and groups them together into components of biomarkers that tend to go together,” Wheeler told Healio. “This allowed us to group the 65 biomarkers into 15 component groups. We were then able to use the results to measure the association of the component groups with ILD.
“We also looked at which biomarkers were represented within the component groups associated with ILD to identify important biologic pathways in this disease,” he added.
According to the researchers, who published their findings in Arthritis & Rheumatology, the process yielded a total of 15 principal components, or clusters of related biomarkers. After adjusting for clinical factors, eight were significantly associated with RA-ILD.
These were:
- Cytokines involved in innate immune response (IL-3, IFN-2a, IL-15 and IL-5) (adjusted OR = 1.16; 95% CI, 1.05-1.29);
- anti-malondialdehyde-acetaldehyde adduct antibodies to albumin, fibrinogen, collagen and/or vimentin PC4 (adjusted ORs included 1.28 [95% CI, 1.11-1.47], 1.18 [95% CI, 1.01-1.37] and 1.26 [95% CI, 1.07-1.49]);
- anti-CCP antibodies and rheumatoid factor (adjusted OR = 1.17; 95% CI, 1.02-1.34);
- adipokines and cytokines (FL, leptin, fractalkine and FGF-21) (adjusted OR = 1.45; 95% ci, 1.23-1.66);
- alarmins (IL-17e-IL-25 and TSLP) (adjusted OR = 1.21; 95% CI, 1.03-1.41);
- tissue remodeling and neutrophil chemotaxis-related biomarkers (MMP9 and IL-8) (adjusted OR = 1.2; 95% CI, 1.03-1.4).
Wheeler said he was surprised by “just how many different biologic pathways seem to be involved in RA patients with ILD.”
“With this study design, we don’t know whether that is a cause or consequence of ILD, but there are many different immunologic and fibrotic mechanisms active in these cases,” he said. “This points toward potential new opportunities for development of screening tests and therapeutic pathways to target.”
A predictive model for RA-ILD using logistic regression yielded an area under the curve of 0.63 (95% CI, 0.582-0.679) when based only on clinical risk factors, such as older age and smoking.
The model’s accuracy increased when all 15 principal components were added (AUC 0.739; 95% CI, 0.693-0.786) and again when the MUC5B promoter region was added (AUC 0.751; 95% CI, 0.707-0.802), according to the researchers.
Wheeler found it notable that ILD was identified accurately enough in the model based on principal component biomarkers that the MUC5B promoter variant added “relatively little to the predictive performance.”
“This could suggest that we are underutilizing the predictive ability of circulating biomarkers in ILD identification/prediction right now,” he said.
Wheeler said his team plans to build on these findings.
“The goal is to transition from identifying cases to predicting who is at risk of developing ILD to guide screening approaches, as well as identifying who is at risk of progression to severe ILD,” he said. “That really gets at the key underlying issue, which is how do we use our resources effectively to screen/monitor the right patients, while not subjecting low-risk patients to unnecessary testing?”
For more information:
Austin M. Wheeler, MD, can be reached at austin.wheeler@unmc.edu.
Perspective
Rheumatoid arthritis-associated interstitial lung disease (RA-ILD) is a major contributor to illness and death among individuals with RA. Despite its clinical importance as an extra-articular manifestation of RA, the number of clearly defined risk factors for RA-ILD remains limited.
Consistently reported clinical risk factors include advanced age, male sex, cigarette smoking and severe joint involvement. Seropositivity also plays a notable role, with associations observed between RA-ILD and the presence of rheumatoid factor, anti-cyclic citrullinated peptide antibodies, and anti-malondialdehyde-acetaldehyde adduct (anti-MAA) antibodies. In particular, high levels of IgA and IgM anti-MAA antibodies have been linked to a significantly increased risk for developing RA-ILD. These antibody isotypes appear especially relevant, as MAA is abundantly expressed in the lung tissue of RA-ILD patients.
Beyond clinical features and autoantibody profiles, genetic predisposition is increasingly recognized as a key contributor to RA-ILD, especially in cases exhibiting a usual interstitial pneumonia (UIP) pattern. The most strongly associated genetic variant identified thus far is the MUC5B rs35705950 promoter polymorphism. This single nucleotide polymorphism has been validated in multiple studies across independent populations. MUC5B encodes a mucin protein involved in airway mucus production. The rs35705950 variant is a gain-of-function mutation in the gene’s promoter region, leading to elevated MUC5B expression.
Although this overexpression contributes to fibrosis in idiopathic pulmonary fibrosis (IPF), it also plays a role in UIP patterns seen in RA-ILD and other autoimmune lung diseases, such as antisynthetase syndrome.
Importantly, the rs35705950 T allele is not only a powerful genetic risk factor for IPF, but also has implications for clinical management. In IPF, its presence has been associated with better survival outcomes, particularly among patients receiving antifibrotic therapy. According to the NIH, this genetic marker is increasingly being used to inform prognosis and guide treatment decisions.
David A. McLain, MD, MACR, FACP, FRCP
Executive director, Alabama Society for the Rheumatic Diseases
Symposium director, Congress of Clinical Rheumatology
Disclosures: McLain reports no relevant financial disclosures.
Sources/Disclosures
Collapse
Disclosures:
Wheeler reports being supported by the Rheumatology Research Foundation. Please see the study for all other authors’ relevant financial disclosures.
