ECP 2023 Abstracts

S244 Virchows Archiv (2023) 483 (Suppl 1):S1–S391 13 Results: We processed 83 ER, 82 PR, 78 Her2, and 83 Ki67 slides for validation. One round of algorithm calibration was performed for ER, PR, and Ki-67, while four rounds of calibration were performed for HER2. After the final calibration, the concordance between the Aperio and Visiopharm systems was 100% for ER, 100% for PR, 91% for HER2, and 100% for Ki67 cases. Review of 522 ER, 519 PR, 421 HER2, and 401 Ki67 cases post go-live shows a continuing need for fine tuning of the system to accommodate individual pathologist prac- tice and changing management guidelines. Conclusion: While the vendor-neutral digital image analysis allows for access to greater options for digital image analysis, there are continuing challenges for adapting digital image analysis in routine anatomic pathol- ogy workflow including quality control and pathology practice variations. E-PS-08-050 A selection of automatic or manual cache data creation procedures of whole slide image in cloud-based digital pathology system M. Yook*, C.K. Jung, S.H. Lee, S.K. Yoon, I.Y. Choi *Department of Medical Informatics, College of Medicine, The Catho- lic University of Korea, Republic of Korea Background & objectives: In cloud systems, store large sized data in object storage in general. This study examines the cache creation procedures of whole slide image (WSI) on cloud system from object storage to server’s memory through network attached storage (NAS). Methods: Digital pathology services delivered through online must not be designed to causes an unpleasant user experience. Therefore, this study set the limit of page loading time is below 3 seconds. The time is vary depending on a WSI file size hence measuring upper limit size for make below 3 seconds as a result. Results: As a result of testing with 22 different sizes of WSIs, ranging from 0.83MB to 26,414.57MB, the transferred size per second ranged from a minimum of 10.38MB to a maximum of 119.57MB. Among them, exclude less than 100MB WSIs because they were always tak- ing below 3 seconds unless it takes technical problem. The average of measured result for selected 11 WSIs were 102.17MB per second in consequences. Therefore, it needs to be less than about 300MB to be able to process in 3 seconds or less is calculated. Conclusion: According to the test results, automatic cache creation pro- cedure is appropriate only when the size of the WSI file to be loaded is less than about 300MB considering the user experience. However, 300 MB is a small number difficult to be founded in a typical original WSI. Therefore, it is necessary to consider reducing the size by degrading the quality or resolution of the WSI depending on the purpose of use. Funding: This research was supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HI21C0940). E-PS-09 | E-Posters Endocrine Pathology E-PS-09-001 Interobserver reproducibility of medullary thyroid carcinoma grading system C. Ariño-Palao*, J. Bernal-Jorquera, D. Roldán, M.G. Rosas-Hernán- dez, I. Ruz-Caracuel *Pathology, Ramón y Cajal University Hospital, IRYCIS, Madrid, Spain Background & objectives: Medullary thyroid carcinoma (MTC) grad- ing system has proven to be extremely useful to stratify patients accord- ing to their prognosis. Although simple, there is little information about pathologists intercomparison results. Methods: Forty-one MTCs from two hospitals were collected. Two teams composed of one pathologist and one trainee blindly reviewed one representative haematoxylin-eosin slide and its associated Ki67 immu- nohistochemistry. We annotated necrosis, number of mitosis/2mm2, Ki67 proliferative index and grade. Cases with discordant Ki67 pro- liferative index were manually counted using a digital image software (QuPath). Kappa correlation index was calculated for grade. Results: Both teams used the same criteria described in the new WHO Classification of Endocrine Tumours. Ki67 proliferative index was originally counted on a microscope. Both teams agreed on grade in 37 out of 41 cases (90%). Kappa agreement score was 0,69 for grade. In one case, discordances were due to not considering necrosis. The other three cases were very near the cut-off point of Ki67 index, with one team scoring them as high-grade and the other as low. These cases were finally counted on hot spot areas using a digital image software. After review and agreement, the case with necrosis and two cases with discrepant Ki67 index were classified high-grade. Conclusion: MTC grading system is notably reproducible between pathologists working in different locations. Main limitations are the presence of small foci of necrosis and stablishing Ki67 proliferative index in cases very near the cut-off point. Reproducibility can be increased if Ki67 proliferative index is assessed using a digital image software. E-PS-09-002 Paraganglioma with FUS::CREM gene fusion: a case report of an aggressive entity M.A. Bani*, V. Suybeng, D. Vasseur, M. Faron, E. Baudin, J. Scoazec *Department of Medical Biology and Pathology, Morphological pathology laboratory, Gustave Roussy Cancer Campus, Villejuif, France, Paris-Saclay university, Gustave Roussy Cancer Centre, Inserm US23, CNRS UMS3655, AMMICa, Villejuif, France Background & objectives: Gene fusions are rare events in paragan- gliomas and usually involves UBTF::MAML3. CREB family gene fusions have been described in diverse mesenchymal neoplasms. How- ever, paragangliomas harbouring FUS::CREM fusions have never been described. We herein describe a first case report. Methods: We report the case of a 52-year-old man with a history of a testicular seminoma who presented with anaemia. CT-scan revealed multiple pancreatic masses with 2 lung nodules. A left pancreatectomy was performed. Results: On microscopy, tumours appeared to be unencapsulated and infiltrating the pancreatic tissue. This tumoral proliferation had a dif- fuse growth with large confluent nests. Some foci of necrosis were found. Tumour cells had a basophilic to amphophilic granular cyto- plasm and a vesicular nuclei and conspicuous nucleoli. Mitotic activity appeared low (1 mitosis/2mm2). Tumour cells expressed synaptophy- sin and chromogranin B without expression of keratins. Some small cells were highlighted by PS100. ALK(D5F4) was expressed. There was no expression of GATA3, PAX8, PHOX2B, TFE3 and SALL4. Expression of FH and SDHB was retained. RNA sequencing showed a FUS::CREM fusion and no mutation was found. The final diagnosis was paraganglioma with FUS::CREM gene fusion. Conclusion: To the best of our knowledge, the FUS::CREM gene fusion has not been previously described in paraganglioma. This patient showed multiple pancreatic tumours with two lung metastases sustain- ing the aggressive behaviour of the spectrum of CREM fusion-positive neoplasms. The unexpected site and pathological diagnosis represent major pitfalls and emphasize the value of molecular testing in rare neoplasms. E-PS-09-003 Osteoclast-like giant cell-rich anaplastic thyroid carcinoma: a diag- nostic challenge A. Borda*, N. Berger, C. Satala, Z. Reti, E. Szasz, M. Decaussin-Petrucci *G.E. Palade UMFST Targu Mures, Romania