Evaluation of intra-operative brain tumour diagnostic services - a large tertiary uk centre experience

Abstract INTRODUCTION Brain tumour intraoperative diagnosis (smear cytology, frozen section) is a commonly performed, routine diagnostic service. Currently, samples must be transported from the operating room (OR) to pathology, impacting turnaround time (TAT), carbon emissions (if cross-site), and motivation for repeat sampling. We performed a broad evaluation of current practice in a large, tertiary, UK brain tumour centre, to identify potential gains in real-time tissue diagnosis. METHODS All brain tumour samples (n=228) sent for intraoperative diagnosis in 2021 were analysed retrospectively. TAT was assessed by capturing different timepoints along the pathway. Concordance between diagnoses at the following stages was determined: preoperatively based on radiology, intraoperatively (frozen section or smear), provisional paraffin and final integrated. Additionally, we anonymously surveyed neurosurgeons’ opinions (n=18) on the current service. RESULTS The mean (±SD) specimen transportation time was 10.6±2.0 minutes, with an estimated total TAT of 30-60 minutes. Intraoperative diagnosis provided a slightly higher rate of concordance with provisional paraffin diagnosis than preoperative radiological diagnosis (89.5% vs 86.3%). Non-concordance was most commonly due to non-representative sampling (e.g., predominantly necrotic), with no repeat sample being sent/available intraoperatively. Prevailing neurosurgical opinion of the intraoperative diagnostic service was dissatisfaction or neutrality (50% and 39% of respondents), with a minority being positive (11%). Reasons for this included: intraoperative delay due to TAT (47%), perceived inaccuracy of results (41%), and perceived reduced out-of-hours availability (56%). CONCLUSIONS Current brain tumour intraoperative diagnostic practice relies on physical sample transportation and manual processing; the resultant long TAT causes surgeon dissatisfaction and dissuades repeat analysis in the case of non-representative sampling. Real-time tissue diagnostic technologies such as OR-sited probe-based confocal endomicroscopy, scanners and Raman spectroscopy should be considered to facilitate faster and repeated examination. The latter may have additional benefits in real-time expert pathology feedback, tumour margin-zone analysis and increased extent of resection.