Diagnosing and staging epithelial ovarian cancer by serum glycoproteomic profiling

Minimally invasive technologies for early diagnosis of epithelial ovarian cancer (EOC) remain an unmet clinical need. CA-125, a tumor marker secreted into the circulation, is utilized to monitor treatment response and disease relapse in EOC, but has limited utility in accurately triaging patients with pelvic masses of unknown histology. To address this unmet need, we applied a novel blood-based glycoproteomic platform that relies on mass spectrometry coupled to machine learning tools, and identified glycopeptide biomarkers that differentiate between patients with benign pelvic masses and malignant EOC. We then used a subset of these markers to generate a classifier that discriminated between benign pelvic tumors and EOC with sensitivity and specificity of 83.5% and 90.1% in the training set and 86.7 and 86.7% in the testing set, respectively. On subgroup analyses, we noticed that patients with malignant EOC had higher levels of fucosylated markers, primarily of hepatic origin. Furthermore, patients with late-stage EOC (FIGO stage III and IV) had markedly higher levels of tri- and tetra-antennary glycopeptide markers containing fucose. We used these markers to build an independent algorithm that can differentiate between early- and late-stage EOC. Lastly, we detected a similar upregulation of fucosylated glycans and gene expression signatures suggestive of multi-antennary glycans in late-stage EOC tissues. We posit that common mechanisms - possibly driven by cytokines - affect both the tumor glycocalyx and liver-derived glycoproteins. In summary, we generated blood glycoproteomic profiles resemblant of distinct tumor states and identified biomarkers that differentiate between benign and malignant pelvic masses, and/or between early- and late-stage EOC. We also provide mechanistic insights suggesting a direct link between the tumor site and the circulating glycoproteome. These data may inform the development of robust clinical tests to diagnose and stage patients with EOC.