Longitudinal Detection of Tumor-Specific Peptides in Cerebrospinal Fluid for Pediatric Brain Tumor Surveillance

Pediatric brain tumor survivors remain at high risk of recurrence, yet current surveillance strategies relying on neuroimaging and cerebrospinal fluid (CSF) cytology have limited sensitivity for early or minimal disease. Tumor-specific peptides (TSPs) derived from individual tumors represent a promising class of highly specific biomarkers for longitudinal disease monitoring through CSF-based proteomic analysis. In this study, tumor tissue and serial CSF samples from six pediatric brain tumor patients (five medulloblastomas and one atypical teratoid/rhabdoid tumor (ATRT)) were analyzed using an integrated proteogenomic workflow combining discovery and targeted mass spectrometry. TSPs were identified from resected tumor tissue and matched against shotgun CSF proteomic datasets to nominate candidate biomarkers. High-confidence peptides were synthesized as isotopically labeled standards and quantified longitudinally using targeted multiple reaction monitoring. Two TSP biomarkers derived from individualized pediatric brain tumors (one medulloblastoma and one ATRT) demonstrated robust detection in serial CSF samples and exhibited temporal concordance with radiographic disease course, declining with treatment response and increasing during disease progression. These findings establish the feasibility of detecting and longitudinally quantifying TSPs in CSF and support further investigation of individualized proteomic biomarkers for treatment response monitoring and disease surveillance in pediatric brain tumors.