Serum glycomic profile as a predictive biomarker of recurrence in patients with differentiated thyroid cancer

PURPOSE: Thyroid cancer recurrence following curative thyroidectomy is associated with increased morbidity and mortality, but current surveillance strategies are inadequate for early detection. Prior studies indicate that tissue glycosylation is altered in thyroid cancer, but the utility of serum glycosylation in thyroid cancer surveillance remains unexplored. We therefore assessed the potential utility of altered serum glycomic profile as a tumor‐specific target for disease surveillance in recurrent thyroid cancer. EXPERIMENTAL DESIGN: We employed banked serum samples from patients with recurrent thyroid cancer post thyroidectomy and healthy controls. N‐glycans were enzymatically released from serum glycoproteins, labeled via permethylation, and analyzed by MALDI‐TOF mass spectrometry. Global level and specific subtypes of glycan structures were compared between patients and controls. RESULTS: We evaluated 28 independent samples from 13 patients with cancer recurrence and 15 healthy controls. Global features of glycosylation, including N‐glycan class and terminal glycan modifications were similar between groups, but three of 35 individual glycans showed significant differences. The three glycans were biosynthetically related biantennary core fucosylated N‐glycans that only varied by the degree of galactosylation (G0F, G1F, and G2F; G: galactose, F: fucose). The ratio of G0F:G1F that captures reduced galactosylation was observed in patients samples but not in healthy controls (p = 0.004) and predicted thyroid cancer recurrence (AUC = 0.82, CI 95% = 0.64–0.99). CONCLUSIONS: Altered N‐glycomic profile was associated with thyroid cancer recurrence. This serum‐based biomarker would be useful as an effective surveillance tool to improve the care and prognosis of thyroid cancer after prospective validation.