SURG-24. Intraoperative Assessment of Glioma Using Miniaturized Epi-fluorescence Widefield Microscope: A Prospective Comparative Study of Ex-vivo and In-vivo Applications

Abstract BACKGROUND Intraoperative histopathological diagnosis of gliomas relies on frozen section (FZ) analysis, which is time-consuming and has variable diagnostic accuracy. This study aimed to evaluate the performance of a Miniaturized Epi-fluorescence Widefield Microscope (MEW-M), implemented using the EndoSCell™ device, for both ex-vivo and in-vivo intraoperative assessment of gliomas. DESIGN This study was conducted between May 2023 and August 2024. Ex-vivo evaluation was performed on core specimens from tumor mass, while in-vivo assessment was conducted for margin specimens at tumor-brain interface. SETTING Prospective, single-center observational conducted at the Department of Neurosurgery, The First Affiliated Hospital of Nanjing Medical University. PARTICIPANTS Eligible participants were aged between 18 and 85 years with glioma confirmed by MRI. Patients were excluded if they had severe cardiovascular or circulatory disease, pregnancy, participation in other trials within 3 months, or inability to provide informed consent. A total of 70 patients were enrolled. RESULTS The study cohort comprised 70 patients (mean age, 55.9±13.9 years; 40 [57%] males), with glioblastoma as the predominant tumor type (67%, 47/70). For tumor core specimens, both pathologists and neurosurgeons achieved 100% sensitivity and accuracy. For margin specimens, pathologists exhibited 100% sensitivity with 88.9% specificity (AUC = 0.944, accuracy 96.0%), while neurosurgeons maintained 100% sensitivity with 74.1% specificity (AUC=0.870, accuracy 90.7%). Both groups showed identical performance for low-grade gliomas (sensitivity 100%, specificity 75.0%, accuracy 91.3%, AUC=0.875), while for high-grade gliomas, pathologists exhibited superior specificity (90.0% versus 70.0%) and accuracy (98.4% versus 95.1%) compared to neurosurgeons. Temporal analysis revealed significantly reduced procedural duration for MEW-M compared with FZ analysis (3.42±1.04 vs 32.93±11.37 minutes). In tumor cavity assessment, MEW-M detected residual tumors in 74% (49/66) of cases. CONCLUSIONS The MEW-M system may serve as a valuable adjunct for intraoperative decision-making during glioma resection procedures, potentially optimizing surgical workflow while maintaining diagnostic reliability.