Ctni-01. Raman-Assisted Rapid Intraoperative Detection of Idh Mutation (Rapid-Idh) in Gliomas for Pathology-Level Precision Resection: A Multicenter, Prospective Clinical Trial

Gliomas represent the most prevalent primary intracranial malignancies in adults. Surgical resection is the primary treatment for gliomas. IDH mutation serves as a critical molecular marker for glioma classification and surgical planning. It informs both the extent of planned resection and, when assessed intraoperatively, the delineation of tumor margins. However, intraoperative detection of IDH status remains unavailable, limiting precision-guided resection. Here, we present RAPID-IDH, a label-free, AI-powered Raman spectroscopy platform for real-time, intraoperative molecular diagnosis of IDH mutation status in gliomas. We established the Neuro-Oncology Raman Atlas (NORA), comprising 5,536 Raman spectra collected from brain tumor specimens between 2020 and 2024. Using 2,815 glioma spectra, we trained and internally validated an XGBoost-based classification model, which was subsequently evaluated in a prospective, blinded multicenter study. RAPID-IDH achieved an AUC of 0.946 (95% CI, 0.924–0.964) in the internal cohort (n=563) and 0.937 (95% CI, 0.892–0.975) in the external cohort (n=156), demonstrating high accuracy and generalizability (NCT 06363162). Single-cell sequencing further revealed altered tryptophan metabolism as the biological basis of IDH-associated Raman spectral signatures, bridging molecular alterations and optical phenotypes. RAPID-IDH enables IDH mutation detection in under 3 seconds per sample, allowing repeatable assessments and enabling dynamic tumor boundary mapping during surgery. This technology integrates molecular diagnostics with intraoperative decision-making, representing a transformative advance in precision neurosurgery that enhances resection accuracy and may improve patient outcomes. We anticipate our research to be the beginning of a new paradigm in intraoperative rapid detection. Future studies will focus on developing optical detection methods applicable to other molecular subtypes of glioma and a broad range of tumor types, further advancing precision medicine in surgery.