Aberrant serum N-glycan profiles have been observed in multiple cancers including non-small-cell lung cancer (NSCLC), yet the potential of N-glycans in the early diagnosis of NSCLC remains to be determined. In this study, serum N-glycan profiles of 275 NSCLC patients and 309 healthy controls were characterized by MALDI-TOF-MS. The levels of serum N-glycans and N-glycosylation patterns were compared between NSCLC and control groups. In addition, a panel of N-glycan biomarkers for NSCLC diagnosis was established and validated using machine learning algorithms. As a result, a total of 54 N-glycan structures were identified in human serum. Compared with healthy controls, 29 serum N-glycans were increased or decreased in NSCLC patients. N-glycan abundance in different histological types or clinical stages of NSCLC presented differentiated changes. Furthermore, an optimal biomarker panel of eight N-glycans was constructed based on logistic regression, with an AUC of 0.86 in the validation set. Notably, this model also showed a desirable capacity in distinguishing early-stage patients from healthy controls (AUC = 0.88). In conclusion, our work highlights the abnormal N-glycan profiles in NSCLC and provides supports potential application of N-glycan biomarker panel in clinical NSCLC detection.
Keywords: biomarker; diagnosis; glycomics; machine learning; non-small-cell lung cancer.
© 2023 Wiley-VCH GmbH.