Nimg-54. confocal fluorescence endomicroscopy: a prospective multi-centered, assessor blinded pilot study for the diagnosis of brain tumors

Abstract Frozen section analysis remains the gold standard for intraoperative diagnosis during brain tumor surgeries; however, it is associated with extensive time consumption and the potential for freezing artifacts. In recent years, confocal laser endomicroscopy (CLE) has been actively studied as a promising alternative to address these limitations of frozen section analysis. In this prospective, multi-centered, assessor-blinded pilot study, our aim was to evaluate the diagnostic effectiveness of cCeLL - Ex vivo, a second-generation CLE model, compared to frozen section analysis for brain tumor diagnosis. We enrolled a total of 55 patients with newly diagnosed brain tumors between October 2022 and December 2022. Biopsy samples were obtained during surgical resection, and each tissue sample was divided into three smaller samples for permanent section analysis, frozen section analysis, and imaging acquisition using cCeLL - Ex vivo. The efficacy of both frozen section analysis and cCeLL - Ex vivo was assessed by comparing their diagnoses to the results from permanent section analysis. The diagnostic accuracy of cCeLL - Ex vivo was slightly higher than that of frozen section analysis, with accuracies of 89.2% and 86.5%, respectively. Both methods demonstrated a sensitivity of 92.2%. Notably, cCeLL - Ex vivo exhibited a higher specificity of 70% compared to frozen section analysis, which had a specificity of 50%. Additionally, the mean time from sample preparation to diagnosis was significantly shorter in cCeLL - Ex vivo (13 min 17 sec) compared to frozen section analysis (28 min 28 sec). These results indicate the promising potential of cCeLL - Ex vivo as a complementary tool for intraoperative diagnosis of brain tumors. Future studies are expected to provide more evidence regarding the potential clinical utility of CLEs in clinical settings.