Discovery of novel protein biomarkers for the detection of polyendocrine metabolic ovarian syndrome phenotype A

Abstract

Polyendocrine metabolic ovarian syndrome (PMOS) is a prevalent endocrine condition affecting women of reproductive age. PMOS is underrecognized/underdiagnosed and associated with secondary comorbidities that may result from delayed diagnosis/management. Novel biomarkers are needed to improve timely diagnosis. This retrospective, case-control study used biobanked serum samples from women with PMOS phenotype A ( n  = 51) and healthy controls ( n  = 37). We performed protein biomarker discovery using Olink Proximity Extension Assay and receiver operating characteristic-area under the curve (ROC-AUC) analysis. We conducted an overrepresentation pathway analysis against Reactome. We investigated 1,196 protein biomarkers and conducted a pathway analysis using the 145 top-performing biomarkers (AUC > 0.750). Top 10 performing biomarkers (AUC ≥ 0.961) were linked to relevant biological pathways, including immune system regulation (serum amyloid A4 and leukotriene A4 hydrolase), lipid, carbohydrate, and protein metabolism (fibroblast growth factor binding protein 1, carboxylesterase 2, inositol polyphosphate-1-phosphatase, inositol polyphosphate-1-phosphatase like, pro-glucagon, and leptin), nervous system development (semaphorin 4C and plexin B1), and cellular response (peroxiredoxin 1). These pathways/sub-pathways may contribute to PMOS pathophysiology and related comorbidities (insulin resistance, obesity, and fatty liver disease). We discovered novel circulating protein biomarkers that distinguish women with PMOS phenotype A from healthy controls, supporting diagnosis and enabling timely intervention/appropriate management.