SURG-18. REAL-TIME INTRAOPERATIVE MOLECULAR DIAGNOSIS AND SURGICAL GUIDANCE USING LASER SPECTROSCOPY

Abstract

BACKGROUND: Raman spectroscopy offers an optical label-free and non destructive real-time solution for identifying tissue during surgery. We evaluate its use during surgery in 35 patients with a range of brain tumors. METHOD(S): Raman spectroscopy was deployed using a surgical protocol in this single centre study of 35 consecutive craniotomies for tumor. Analysis of tissue was done in-vivo during surgery at the brain surface, on route to the tumor and in superficial and deep tumor zones. Sampling was also performed along the resection margin and all spectra were compared to matched core biopsy samples verified by routine histopathology. Ex-vivo tissue analysis was also performed at different time points to look at change in tissue signal characteristics over time. RESULT(S): Over 3,000 Raman spectra from the brain were recorded. The system correctly identified the whole range of brain tumors being operated on. Multivariate analysis using PCA/LDA was performed. Most easily identified tumors include meningiomas and metastases. Low grade and high grade gliomas were also identified with high accuracy. CONCLUSION(S): Raman spectroscopy represents a new modality for in-vivo tissue characterisation with high accuracy. Our study demonstrates it can correctly characterise a range of different types of brain tumors rapidly during surgery. Current challenges to more widespread adoption include the modifications necessary for it to work in the OR and patient-patient and machine-machine variability.