ECP 2023 Abstracts

S19 Virchows Archiv (2023) 483 (Suppl 1):S1–S391 13 Although BAP1 loss and homozygous P16 loss on FISH are considered to be useful in pleural fluid cytology, molecular analysis is still not commonly practiced within our centre. Based on this re-audit we make the following recommendations; double reporting in cases where there is clinical concern and discussion with clinicians regarding simultane- ous cytology and biopsy sampling to reduce delays in diagnosis. OFP-05-002 Reproducibility of c-MET immunohistochemistry (clone SP44) interpretation in non-small cell lung carcinoma C. Bontoux*, V. Hofman, E. Chamorey, R. Schiappa, S. Lassalle, E. Long-Mira, J. Fayada, C. Bonnetaud, S. Goffinet, M. Ilié, P. Hofman *Laboratory of Clinical and Experimental Pathology, Hospital-Inte- grated Biobank (BB- 0033-00025), Pasteur Hospital, IRCAN Team 4, Inserm U1081, CNRS 7284, FHU OncoAge, Université Côte d’Azur, Nice, France Background & objectives: Emerging therapies for non-small cell lung carcinoma (NSCLC) with MET overexpression have shown promise. However, MET expression evaluation is challenging. We aim to help standardizing MET expression evaluation among pathologists for a better understanding of its role as a biomarker. Methods: 110 cases diagnosed with NSCLC and routinely evaluated for cMET expression in the Laboratory of Clinical and experimen- tal Pathology (Nice, France) in 2022/2023 were selected. All cMET stained-slides were digitized. Six pathologists (junior and senior) per- formed a H-score and grading assessment of cMET expression for all cases, both under the microscope and on screen. We analysed inter and intra-rater reproducibility. Results: Global intraclass correlation coefficient (ICC) for H-score classification (cut-off of 150) was 0.668 (IC95%[0.552-0.773]) with an ICC of 0.812 (IC95%[0.691-0.907] in biopsy samples. Global ICC for cMET group grading (using a semi-quantitative evaluation according to the % of 3+ stained cells) was 0.671 (IC95%[0.555-0.776]) with an ICC of 0.888 (IC95%[0.805-0.947]) in biopsy samples and 0.959 (IC95%[0.914-0.986]) for squamous cell histology. ICC was similar for junior and senior pathologists. Regarding intra-rater reproducibility, Cohen’s kappa coefficient var- ies from 0.348 to 1.0 and from 0.538 to 1.0 for H-score and group grading assessment, respectively. Cohen’s kappa coefficient was lower for junior versus senior pathologists using both methods. Data from digital assessment are awaiting. Conclusion: We demonstrate that the reproducibility of assessing cMET expression ranges from moderate to good. Certain factors, such as squamous cell histology or biopsy samples, can significantly improve reproducibility. Additionally, we showed that intra-rater reproducibility seems to be lower for junior pathologists. Our findings support the notion of tumour heterogeneity in cMET expression both within and between tumours, highlighting the importance of estab- lishing a consensus definition and providing further training in cMET expression evaluation, particularly for inexperienced pathologists. OFP-05-003 Novel ALK Dako-CD246 assay proves to be as competent as the FDA-approved ALK Ventana-D5F3 assay in identifying lung adenocarcinomas with ALK alterations: real-life validation of 188 cases (2011-2023) O.C. Eren*, C. Aydin Mericoz, P. Bulutay, A. Baygul, I. Kulac *Koç University Hospital, Department of Pathology, Turkey Background & objectives: Affordable and reliable, immunohisto- chemistry-based biomarkers show promise. FDA-approved ALK Ven- tana-D5F3 assay is a valid alternative to molecular tests for non-small- cell lung cancer. Recently introduced ALK Dako-CD246 assay may offer similar benefit. We present the first validation study on the latter. Methods: Primary lung adenocarcinomas between 2011-2023 were retrieved from archives of Koç University Hospital. Representative tumour blocks for each case were stained with Ventana-D5F3 and Dako-CD246 and evaluated by 3 independent reviewers. For a subset, Next Generation Sequencing (NGS) / Fluorescence in situ Hybridiza- tion (FISH) was carried out to verify results of IHC-based assays. Results: 188 primary lung adenocarcinomas were evaluated. Inter- assay agreement was found to be K=0.885 (p<0.001, “near perfect” category). In Ventana-D5F3 assay, 14 of cases were interpreted as positive and 171 as negative. In Dako-CD246 assay, 16 and 169 cases were allocated in the respective categories. Three cases (%1.6) in both assays could not be reliably evaluated. Concerning molecularly confirmed cases (n=38), concordant results were obtained in 92.1% (35/38, Ventana®) and 94.7% (36/38, Dako®) of cases, achieving high sensitivity (84.6%, Ventana®; 92.2%, Dako®) and specificity (96% for both). With NGS/FISH results taken gold standard, Kappa scores were K=0.875 (p<0.001) for Ventana-D5F3 and K=0.940 (p<0.001) for Dako-CD246, both in “near perfect” category. Conclusion: IHC-based assays provide a valid and reasonably priced alternative, especially in settings where molecular confirmatory tests such as NGS/FISH are not offered or accessible. In this regard, Dako- CD246 proves to yield concordant results with Ventana-D5F3, which is already promptly used in routine pathology practice. The present study being the first in the field, given its high inter-assay and molecular concordance, we propose that novel Dako-CD246 assay can be reliably used in identifying cases for ALK-targeted therapy. OFP-05-004 Discrimination of multiple primary lung adenocarcinomas from the intrapulmonary metastasis by using pathologic, radiologic, and molecular correlations P. Bulutay, B.K. Aktas*, G.E. Cecikoglu, C. Aydin Mericoz, I. Kulac, K.B. Ozer, S. Erus, S. Tanju, S. Dilege, C. Atasoy, P. Fırat *Koç University Faculty of Medicine, Department of Pathology, Turkey Background & objectives: Accurately differentiating between “multiple primary lung cancers (MPLC)" and "intrapulmonary metastasis (IPM)" is essential for prognosis and treatment. Given that histology-based dis- crimination can be prone to misinterpretation, molecular evidence can provide a valuable adjunct in reducing errors in selected cases. Methods: The study included seventeen cases, each of which had at least two synchronous or metachronous lung adenocarcinomas and all underwent molecular testing. Cases were evaluated blindly for the morphologic features, radiologic interpretation, and molecular alterations, and diagnosed as MPLC or IPM for each parameter. The results were brought together and their compatibility with each other was investigated. Results: Twelve cases (70%) were evaluated with next-generation sequencing (NGS) and 5 with single gene analysis (30%). At least one driver mutation was detected in 77% (13/17) of the cases. Among them, 2 cases supported IPM due to the same mutation types, while the remaining 11 cases were classified as MPLC due to different mutation types. Morphologic classification of 85% (11/13) of those was compat- ible with their molecular findings. Molecular findings were consistent with radiologic classification in the remaining 2 cases. Four cases were molecularly inconclusive, among them 2 cases were evaluated using single gene analysis. Radiological and morphological diagnoses were consistent in 74% of the cases, and inconsistent in 26%. Conclusion: The histology-based approach can accurately discriminate MPAC and IPM in most cases. Radiologic correlation enhances the reliabil- ity of morphological diagnosis and can assist in categorizing morphologi- cally indeterminate cases into a specific direction. Revealing the molecular alterations of each tumour focus enables evidence-based differentiation of MPACs and IPMs. However, since performing molecular tests on each