SURG-07. DETECTION OF GLIOMA INFILTRATION AT THE TUMOR MARGIN USING QUANTITATIVE STIMULATED RAMAN SCATTERING MICROSCOPY

BACKGROUND Differentiating neoplastic tumor tissues from normal brain at the infiltrative tumor margin remains challenging. Stimulated Raman scattering microscopy (SRM) is a rapid, label-free technique that provides microscopic imaging of unprocessed tissues. However, it has never been studied as a tool to enhance extent of resection. METHODS In this single center study, patients with WHO II-IV gliomas undergoing surgical resection were included. Margin samples were taken with corresponding stereotactic coordinates. Each tumor margin sample was analyzed using (1) H&E, (2) SRM, and (3) immunohistochemical (IHC) stains for IDH1-R132H or p53. Specimens were imaged fresh with SRM, inked for orientation and placed in formalin for routine processing. Stained slides were scored by 3-neuro-pathologists using a semiquantitative 2-tiered scoring system for the presence or absence of residual tumor. SRM microscopy images were segmented into cellularity. The intraobserver measure of agreement between modalities for each pathologist was calculated using Cohen’s kappa. RESULTS A total of 31 patients and 179 margin specimens were acquired. Postoperative MRI confirmed that 169 (94%) of glioma margin sample coordinates were indeed at the tumor margins (10 of 179 were not true margins). All 10 samples which were not true margins were identified as having residual tumor by SRM. IHC semiquantitative scoring of margin specimen confirmed 72 of 128 samples (56%) had residual tumor. Similarly, H&E-scoring confirmed 82 of 169 samples (49%) and 82 of 167 samples (49%) by SRM-scoring had residual tumor. Intraobserver agreements between all 3 modalities confirmed agreement: IHC-SRM was near perfect (K-0.84, CI-0.750-0.94); IHC-H&E substantial, (K-0.67 CI-0.54-0.80); SRM-H&E agreement substantial (K-0.72 CI-0.62-0.83). Margin samples with residual tumor demonstrated higher tissue cellularity when compared with samples without tumor (p< 0.0001). CONCLUSIONS SRM allows for identification of residual microscopic disease at the infiltrative tumor margin and therefore may be a promising tool to enhance extent of resection.