ECP 2022 Abstract Book

Virchows Archiv (2022) 481 (Suppl 1):S1–S364 13 Background & objectives: There is inconsistent data on cytokeratin 7 (CK7) immunostaining patterns in clear cell renal cell carcinomas (CC-RCC). The aim of this study was to assess the CK7 expression pattern of CC-RCC. Methods: We analysed 30 cases of CC-RCC from our institution, diagnosed between 2014 and 2022. CK7 immunostaining was performed. The tumours were divided into low-grade CC-RCC (LGCC-RCC) (WHO/ISUP grade I and II) and high grade CC- RCC (HGCC-RCC) (grade WHO/ISUP III and IV). Architectural pattern was also noted. The patterns of staining were divided into three groups: negative, focal and diffuse. Results: The mean age at diagnosis was 62,5 years (33 - 83 years-old). 22 cases (73,3%) were classified as LGCC-RCCs and 8 cases (26,7%) as HGCC-RCCs. 14 cases (46.7%) showed CK7 immunoreactivity, with 3 cases displaying a diffuse expression pattern and 11 cases with focal staining. 13 cases of the LGCC- RCCs group (59,1%) and 1 case (12,5%) of the HGCC-RCC group showed expression of CK7. 10 out of 16 cases with macrocystic or microcystic architectures showed CK7 expression. Conclusion: Our results showed that CK7 immunoreactivity in CC- RCCs can have a variable pattern and extent of expression. LGCC- RCCs revealed a higher rate of CK7 positivity regarding HGCC-RCCs, leading to a possible prognostic role of CK7 regarding these entities. Thus, the possibility of expression of CK7 by CC-RCC is an important characteristic that pathologists should keep in mind, particularly in the interpretation of small tissue samples. PS-26-034 Morphological characterization of chromophobe renal cell car- cinoma: review of 30 cases G. Miranda*, J. dos Santos, J. Correia-Pinto, J. Cunha, T. Amaro *Unidade Local de Saúde de Matosinhos - HPH, Portugal Background & objectives: Chromophobe renal cell carcinoma (ChRCC) is a renal neoplasm traditionally described by having sheets of large polygonal cells with pale to eosinophilic cytoplasm. The aim of this study was to document the histopathological features of ChRCC cases in our institution Methods: We reviewed retrospectively the slides of cases diag- nosed between 2014 to 2022. Immunohistochemistry for CK7, CD117 and CD15 were performed in all cases. 30 cases of ChRCC were identified from all 229 cases of renal neoplasms from our institution in this period. Results: ChRCC represented 13% of all renal neoplasms. The mean age at diagnosis was 66.3 years old (47 – 87 years). 73.3 % of the patients were men and 26.7% women. The histologi- cal types were divided between solid sheet-like, tubulocystic and papillary patterns and the cytoplasm features were divided between pale cytoplasm versus eosinophilic cytoplasm. The architectural patterns were 18 cases (60%) of a predominant solid pattern, 11 cases (36.7%) with predominant tubulocystic areas, and 1 case (3.3%) with papillary architecture. 18 cases (60%) had abundant pale cytoplasm and 12 (40%) predominant eosinophilic cytoplasm. 56.7% had diffuse positivity for CK7, 96.7% showed CD117 expression and only 13.3% had CD15 focal positivity. Conclusion: Our study demonstrated a similar distribution in age at diagnosis and gender in comparison to the existing literature, described the architectural diversity of this neoplasm and its immunohistochemical findings. To avoid misdiagnosis, this pathological heterogeneity of ChRCC is an important feature that pathologists should have in mind. PS-26-035 Quantitative, rules-based grading for noninvasive papillary urothelial carcinoma D. Berman*, A. Slotman, K. Lindale, D. Winkowski, R. Baird, A. Simpson, R. Gooding *Queen’s University, Canada Background & objectives: Noninvasive papillary urothelial car- cinoma (NPUC) places great burdens on patients and healthcare systems. Histopathologic grading for NPUC guides management, yet subjective criteria and poor reliability limit its value. Here we develop an accurate, quantitative, and explainable NPUC grading algorithm. Methods: Clinical data including recurrence and progression were collected for 201 NPUC patients presenting from 2008-2016. Image analysis was performed on 641 H&E stained tissue microarray (TMA) cores harvested from their transurethral resections. Histologic features extracted using Visiopharm software were analysed individually and in combination using regression and decision tree models. Survival analysis for Kaplan-Meier curves were calculated with log-rank statistics. Results: Whole slide grade and stage were significantly associated with recurrence-free survival. TMA-based nuclear morphometry, cellular organization and automated mitotic figure counts differentiated low- from high-grade NPUC. As a single variable, variation in (standard deviation of) nuclear area distinguished between high- and low-grade with 82% accuracy. Multivariate models combining nuclear size and shape-related variables with mitotic activity increased balanced accuracy to 88%. To adapt these TMA-based findings to whole slide images, we are using deep learning to select and analyse hotspots that are enriched for high-grade features. Conclusion: This work represents the first demonstration of explainable quantitative criteria for NPUC grading, and points to mitotic index and variation in nuclear size as key quantitative vari- ables. Ongoing work will build a clinically-oriented user-friendly platform that works on whole slide images and optimizes the prog- nostic value of NPUC grading. Funding: Ontario Molecular Pathology Network and Ontario Institute for Cancer Research PS-26-036 Percutaneous renal biopsy of kidney tumours: what comes next? D. João*, A. Sanches, A. Luis *Centro Hospitalar Vila Nova de Gaia, Portugal Background & objectives: Kidney tumours are frequently both diag- nosed and treated through partial or radical nephrectomy after imagio- logical detection. Nevertheless, many are first approached by percuta- neous needle biopsy. We aim to study and compare biopsy and surgical specimen diagnosis, grade and clinical outcomes. Methods: We reviewed all cases of kidney tumour biopsy received in our department during the last 5 years (n=56). For each, we determined the biopsy diagnosis; the diagnosis in the nephrectomy specimens (if surgical treatment), patient management (surveil- lance, cryotherapy, surgical treatment, systemic treatment or pal- liative care), and clinical follow-up. Results: In biopsy, clear cell renal cell carcinoma was the most frequent diagnosis (n=13;23%), followed by a differential diagnosis between oncocytoma and the eosinophilic variant of chromophobe renal cell carcinoma (n=7;13%). Regarding patient management, 16 (29%) were submitted to cryotherapy, in 14 (25%) was opted surveillance, 6 (11%) received palliative care and 3 (5%) systemic S177