Automated Multiplex Plasma Proteomics for Quality Assessment and Clinical Triage: A Proof-of-Concept in Indeterminate Thyroid Nodules

Abstract Fine-needle aspiration biopsy of thyroid nodules frequently yields indeterminate results, leading to unnecessary diagnostic surgeries. We evaluated a targeted plasma proteomics approach to differentiate benign follicular thyroid adenoma (FTA) from malignant papillary thyroid carcinoma (PTC). Using LC-MRM mass spectrometry with stable isotope-labeled standards (SIS), we quantified a multiplex protein panel in plasma samples from 267 patients. The analytical workflow incorporated an MS-based assessment of preanalytical hemolysis to prevent bias associated with thyroid-related blood composition changes and multiple rounds of randomization allowed assessment of batch effects. While previously proposed low-abundance thyroid cancer markers were rarely detected in the cohort, a signature of high-abundance plasma proteins successfully stratified the patients. The malignant PTC phenotype was characterized by a significant increase in apolipoprotein A1 and myosin-9, along with elevated CD14 and kininogen-1. Benign FTA was associated with higher levels of acute-phase proteins, including ceruloplasmin and complement C1q subunit C. Simple PCRE model with 20 parameters based on 4 major plasma proteins allowed to build a classifier with specificity that can be adjusted with regard to epidemiology and medical resource. This study establishes a foundational mass spectrometry-based framework that utilizes systemic proteomic profiles to differentiate underlying pathologies and improve clinical decision-making regarding patient management. By enabling additional stratification of patients with Bethesda IV cytology, who are otherwise frequently subjected to unwarranted diagnostic operations, this multiplex MS approach provides a potential non-invasive triage tool to reduce unnecessary surgical interventions for indeterminate thyroid nodules.