ECP 2023 Abstracts

S43 Virchows Archiv (2023) 483 (Suppl 1):S1–S391 13 of six months (0-2408 months), the median overall survival was of 6±0.5 months. No difference in overall survival was seen associated with the expres- sion of CD44 (p=0,504), HER2 (p=0,599), BerEp4 (p=0,899), EGFR (p=0,081), mismatch repair proteins (p=0.197) and HIF1 (p=0,65). PD-L1 and NTRK was negative in all cases. Conclusion: Colorectal brain metastases are associated with a poor prognosis. In our study, no association was found between CD44, HER, mismatch repair proteins, BerEp4, EGFR, PD-L1, HIF1 and NTRK. Although not showing statistical significance there seems to be a pos- sibility for CD44 and EGFR to predict prognosis and guide therapy. These findings should be confirmed in further multicentric studies. OFP-11-004 Diagnostic accuracy and feasibility of an ultra-fast digital confo- cal microscopy scanner for real-time intra-operative brain tumour diagnosis W. Bolton, P. Ramakrishnan, D. Reveendran, V. Crispi, O. Adebola, R. Sinha, R. Digby*, A. Chakrabarty, R. Mathew *Leeds Teaching Hositals NHS Trust, United Kingdom Background & objectives: Real-time intra-operative brain tumour tissue analysis shortens turnaround times and facilitates repeated sampling, improving diagnostic accuracy and guidance to maximise resection. The aim of this present study is to evaluate the diagnostic accuracy of Histolog® within 50 brain tumour patients. Methods: Varying brain tumour types was conducted in adult patients undergoing brain biopsies or tumour debulking surgery. Multiple, freshly excised tissue samples were stained, and scanned via Histolog® confocal microscopy (60 seconds) in parallel to standard diagnostic pathways including frozen section and smear cytology. Blinded con- cordance analysis between Histolog® images and gold standard histo- pathology was performed by a Consultant Neuropathologist. Results: A total of 50 cases are included in this analysis. Cases studied included glioma, metastasis, meningioma, schwannoma, and pituitary adenoma. Blinded concordance analysis revealed that in every case (n=50; 100%) the Histolog® image was able to confirm the presence of abnormal vs normal tissue, and in n=34 (68%) of instances the diag- nosis was precisely confirmed with Histolog® alone. Histolog® images demonstrated specific diagnostic features such as clusters of cohesive epithelioid cells (metastatic carcinoma), sheet-like variably cellular and pleomorphic cells (gliomas), diffuse sheet-like monomorphic round nuclei (pituitary adenoma), elongated spindle cells (schwannoma), and nodular architecture and oval nuclei (meningioma). Conclusion: Ultra-fast confocal microscopy scanning with His- tolog® demonstrates non-inferior diagnostic accuracy when com- pared to current gold standard but significantly reduces the intra- operative time to achieve diagnosis. We are the first team in the world to demonstrate the feasibility of this technology for real-time intra-operative brain tumour tissue diagnosis. Future studies will investigate the effect of real-time margin zone analysis and explore the use of machine learning applied to the Histolog® images for automatic computer aided diagnosis. OFP-11-005 In the molecular era, is histopathological examination still effective in demonstrating the relationship with recurrence in meningiomas? Y. Adali*, S. Ekmekci, Ü. Küçük, Y. Pekçevik, E.E. Pala *Izmir University of Economics Faculty of Medicine, Turkey Background & objectives: Meningiomas are usually slow growing tumours which are divided into three grades according to their histo- pathological and some molecular findings. In this study, it was aimed to document the meningiomas and to investigate the relationship between histopathological findings and recurrence. Methods: 266 cases diagnosed with meningioma between 2017-2022 were included in the study. Sections of the cases were re-evaluated by 2 pathologists in terms of histological type, grade, number of mitosis, Ki-67 proliferation index (Ki-67); and presence of necrosis, pattern- free pattern, small cell change, hypercellularity, macronucleolus, brain invasion. Tumours diagnosed in 2022 were not included in the recur- rence focused analysis. Results: 111(41.7%) cases were meningothelial, 67(25.2%) atypical, 45(16.9%) transitional, 18(6.8%) fibrous, 8(3.0%) anaplastic, 6(2.3%) angiomatous, 4(1.5%) psammomatous, 3(1.1%) microcystic, 2(0.8%) metaplastic, and 2(0.8%) clear cell histological type of meningiomas. It was noted that the grade, mitosis, and Ki-67 were statistically sig- nificant with the presence of necrosis, pattern-free pattern, small cell changes, hypercellularity, macronucleolus, and brain invasion (all p values 0.000). The presence parameters other than grade, mitosis and Ki-67 shows a statistically significant correlation with the absence of necrosis, patternless pattern, small cell changes, hypercellularity, macronucleolus, and brain invasion (all p values less than 0.001). Simi- larly, a statistically significant correlation was found with recurrence with increased grade, mitosis and Ki-67. Conclusion: The most common subtypes are defined as menin- gothelial, fibrous (replaced by atypical in our study) and transitional. Meningiomas are graded according to the mitosis, hypercellularity, macronucleolus, small cell changes, pattern and necrosis. In the present study, a significant relationship between grade and mitosis and Ki-67 was shown, while a negative correlation with other histopathological findings was noted. In this context, it should be underlined that not only histopathological findings, but also molecular features should be included in the grading of meningioma in daily practice. OFP-11-006 Histologic definition of enhancing nodule and FLAIR hyperinten- sity region of glioblastoma, IDH-wild type: a clinico-pathologic study on a single-institution series G. Broggi*, R. Altieri, V. Barresi, F. Certo, G. Barbagallo, M. Zanelli, A. Palicelli, G. Magro, R. Caltabiano *Department of Medical and Surgical Sciences and Advanced Tech- nologies "G. F. Ingrassia", Anatomic Pathology, University of Catania, 95123 Catania, Italy Background & objectives: Since the extent of resection beyond the enhancing nodule (EN) in glioblastoma IDH-wild type (GBM, IDHwt) is a highly-debated topic in neuro-oncology, we aimed to provide a comprehensive description of the histopathology of EN and FLAIR hyperintensity regions. Methods: 33 patients (20 males and 13 females with a mean age at diagnosis of 56 years) affected by GBM, IDHwt were enrolled in this study; for each of them EC and FLAIR hyperintensity regions were sampled intraoperatively using neuronavigation and 5-aminolevulinic acid (5-ALA) fluorescence. A total of 109 histological samples were collected and evaluated. Results: In 29/33 (88%) cases, the samples from ECs exhibited classic GBMmorphology, consisting of hypercellularity, increased mitotic activity, necrosis and/or microvascular proliferation (MVP). In 4/33 cases (12%), the specimens from the ECs showed hypercellular and mitotically-active high- grade diffuse astrocytic tumours, IDHwt, lacking both necrosis and MVP. We also found that FLAIR hyperintensity regions consisted of: (i) fragments of white matter focally to diffusely infiltrated by tumour cells in 76% of cases; (ii) a mixture of white matter with reactive astrogliosis and grey matter with perineuronal satellitosis in 15% and (iii) tumour tissue in 9%. Conclusion: Since it has been demonstrated that local disease recur- rence often arises in peritumoral areas of GBM, IDhwt and that radi- ologically-defined FLAIR hyperintensity areas of GBM IDHwt are often visible beyond the conventional EN, a deeper knowledge of the histology of FLAIR hyperintensity areas in GBM, IDHwt may serve