Abstract
BACKGROUND: This study aimed to develop an early diagnostic method integrating proteomic biomarkers and clinical parameters for screening interstitial lung disease (ILD) in patients with newly diagnosed rheumatoid arthritis (RA) through a multi-phase research strategy. METHODS: A three-phase study was conducted: (1) Discovery: Tandem mass tag (TMT)-labeled quantitative proteomics with liquid chromatography-tandem mass spectrometry (LC-MS/MS) analyzed serum protein profiles in 5 RA-ILD and 5 RA-non-ILD patients, identifying candidates via bioinformatics. (2) Verification: Enzyme-linked immunosorbent assay (ELISA) validated candidates in an independent cohort (13 RA-ILD vs 14 RA-non-ILD). (3) Application: Biomarkers combined with clinical indicators (Krebs von den Lungen-6 [KL-6], age, sex) were evaluated in 110 patients (51 RA-ILD vs 59 RA-non-ILD) to build a predictive model. RESULTS: Proteomic analysis identified matrix metalloproteinase-3 (MMP3), von Willebrand factor (VWF), and other significantly differentially expressed proteins. ELISA validation confirmed that serum MMP3 and VWF levels were significantly higher in the RA-ILD group than in the RA-non-ILD group (p = 0.025 and 0.027, respectively). Expanded validation demonstrated superior diagnostic performance when combining MMP3 and VWF with KL-6 (area under the curve [AUC] = 0.90). The nomogram prediction model based on univariate analysis exhibited excellent discrimination (AUC = 0.89) and calibration. CONCLUSION: This systematic study from discovery to validation identified MMP3 and VWF as potential biomarkers for RA-ILD. The integrated predictive model combining these biomarkers with clinical parameters (KL-6, age, sex) provides a potential tool for early ILD screening in RA patients, offering novel strategies for early diagnosis and intervention of RA-ILD.