Tumor glycosylation in cancer therapy: biomarkers, resistance states, and translational opportunities

Tumor glycosylation has emerged as a potentially informative layer in translational oncology because it connects oncogenic signaling, metabolic adaptation, immune recognition, and therapeutic response through measurable molecular changes. Across cancers, recurrent glycan alterations, including branched, hypersialylated, or hyperfucosylated surface N-glycans, truncated mucin-type O-glycans, checkpoint-associated glycoforms, glycan-lectin suppressive interfaces, and HBP-OGT-O-GlcNAc-centered stress programs, may influence target accessibility, immune exclusion, receptor persistence, and treatment tolerance. These alterations are relevant across immunotherapy, targeted therapy, chemotherapy, radiotherapy, and adaptive resistance states. Rather than treating glycosylation as a single hallmark, this review focuses on a restricted set of recurrent, assayable, and potentially clinically interpretable glyco-readouts that may refine clinical interpretation in defined treatment settings. Checkpoint-associated glycoforms can refine assessment of target blockability, surface glycan architectures can indicate altered therapeutic accessibility or steric shielding, lectin-linked signatures can reflect suppressive immune-excluded microenvironments, and HBP-OGT-O-GlcNAc-related readouts can help identify repair-tolerant resistance states. These readout classes are better considered complementary to established markers rather than universal biomarkers. A use-case-first route provides the most credible framework for clinical translation. Clinical translation will depend less on the breadth of glyco-signals described across cancers than on whether individual readouts can be linked to defined treatment questions, measured through fit-for-purpose tissue or liquid assays, and shown to improve clinical interpretation in specific treatment settings. On that basis, the near-term translational focus remains on a limited set of clinically interpretable readouts rather than on glycosylation as a broad biomarker layer. Progress will depend on scenario-specific analytical qualification, prospective validation in well-defined cohorts, and biomarker-guided combination strategies. Overall, tumor glycosylation is most clinically useful as a selective source of decision-relevant readouts for patient stratification, resistance interpretation, and rational therapeutic design.