Biochemical and metabolic tissue discrimination using a dual-modality Raman-fluorescence endonasal spectroscopy probe for pituitary adenoma surgery.

PURPOSE: A major challenge in pituitary adenoma surgery is distinguishing adenoma from normal pituitary gland and adjacent structures. Accurate intraoperative identification is critical to optimize tumor resection, hormonal remission, and preservation of pituitary function. Building on our previous preclinical validation of a bayonet-style endonasal Raman probe, we evaluated a dual-modality Raman spectroscopy and tissue fluorescence system for ex vivo discrimination of pituitary adenoma, normal gland, and surrounding sellar tissues. METHODS: A prospective study was performed in 32 patients undergoing transsphenoidal resection of adenomas. During surgery, Raman spectroscopy and near-infrared fluorescence measurements were acquired ex vivo from excised tissues. A limited number of spectra were obtained from normal gland. Histopathology was the gold standard. Fluorescence intensity was normalized for integration time and compared statistically across tissue classes. Spectral angular mapping (SAM) was applied to standard-normal-variate (SNV)-normalized Raman spectra for tissue classification. RESULTS: 502 spectra were analyzed. Both pituitary gland and adenoma showed higher fluorescence than surrounding tissues. Fluorescence alone achieved an AUC of 0.90, with 86% accuracy, 88% sensitivity, and 86% specificity for discriminating adenomas from normal surrounding tissues. Raman spectroscopy analyzed using the SAM metric provided complementary biochemical contrast (87% accuracy, 76% sensitivity, 96% specificity). A two-stage approach, combining fluorescence preselection with Raman classification, improved discrimination between normal gland and adenoma: 94.5% sensitivity, 85.7% specificity. CONCLUSIONS: Raman spectroscopy and tissue fluorescence enable label-free discrimination of pituitary adenoma from normal gland and surrounding structures in ex vivo human tissue. These findings support the feasibility of optical guidance, to be further validated in ongoing in vivo clinical studies.