Thromboangiitis obliterans (TAO) is often diagnosed late and characterized by high amputation rates. TAO-specific early diagnostic and disease-staging biomarkers are urgently needed. A staged mass spectrometry (MS)-based discovery-verification-validation workflow was utilized in this study. Ten healthy controls (HCs) and 20 TAOs were included for data-independent acquisition (DIA)-MS quantitative proteomic analysis for the discovery cohort. The DIA-MS analysis acquired 842 identified proteins and 470 quantifiable proteins. Twenty-three candidate biomarkers were further quantified using targeted proteomics based on parallel monitoring reaction (PRM) analysis in the verification stage. A 9-protein and a 7-protein serum biomarker panels were built by machine learning to accurately distinguish TAOs from HCs and active TAOs (A-TAOs) from inactive TAOs. A combined prognostic panel consisting of serum proteins and clinical indicators was established, allowing for risk stratification of A-TAOs. During the validation stage, an independent prospective validation cohort was recruited to validate the proteomic panels based on the enzyme-linked immunosorbent assay analysis, demonstrating the stability and robustness of the predictive models. This study presented the serum proteomic landscape of a TAO cohort and provided novel insights into further biological research. Meanwhile, serum protein signatures have a great potential to improve the early diagnosis and risk stratification in TAOs.
Keywords: biomarkers; mass spectrometry; protein signature; serum proteome profiling; thromboangiitis obliterans.