Article
Intraoperative in-situ raman spectroscopy for brain tumour delineation
Intraoperative in-situ Raman Spektroskopie zur Abgrenzung von Hirntumoren
Search Medline for
Authors
Published: | May 25, 2022 |
---|
Outline
Text
Objective: Maximal safe tumor resection (MTR) is one of the biggest challenges in Neurosurgery. Raman spectroscopy could be utilized to achieve MTR, since it is a label-free technique to acquire a biochemical fingerprint of tissue within seconds. However, clinical translation to intraoperative settings remains demanding. Hence, we performed in-situ Raman spectroscopy during brain tumor surgery with an easy-to-use handheld Raman probe to prove feasibility during neurosurgical procedures and investigate Raman-based biomarkers for intraoperative tumor delineation.
Methods: Intraoperative Raman spectroscopy was performed in 30 patients with primary and secondary brain tumors (10 glioblastoma WHO IV, 6 glioma WHO II-III, 3 meningioma WHO I-II, 11 metastases). A commercially available fiber-optic Raman probe was used with an excitation wavelength of 785 nm and acquisition time of 2 s per spectrum. Between 5-15 spectra of tumor-, and adjacent tissue were acquired per patient at different loci. Reference histopathology was performed at measurement positions. Both, Near-Infrared Fluorescence- and Raman spectra were analyzed.
Results: Measurements prolonged the duration of surgery about 5 min at most. Consistently to previous ex vivo studies, Raman bands assigned to lipids (e.g. 1437 cm-1) were significantly reduced across all tumor entities (p<0.0001), whereas protein bands (e.g. 1240-1280 cm-1) were significantly increased (p<0.0001). Fluorescence intensity also showed significant reduction for all types of tumors compared to the adjacent tissue (p<0.0001).
Interestingly, we found a strong positive correlation of Fluorescence- and lipid band intensity and in contrast a negative correlation for Fluorescence- and protein band intensity.
Conclusion: We showed successful clinical application of intraoperative Raman Spectroscopy through minimal delay of surgical workflow. Typical tumor features of Raman spectra, previously described in ex-vivo studies were confirmed in an in-vivo setting. Thus, utilizing intensities of Raman bands related to lipids and proteins might be a suited approach to contribute achieving MTR long-term. The differences in fluorescence intensity of tumor and adjacent tissue are yet to be understood. However, our results strongly suggest the feasibility of a fluorescence based tissue classification. Further development could lead to a standalone Near-infrared Fluorescence application, which would simplify the intraoperative setup.