ECP 2022 Abstract Book

Virchows Archiv (2022) 481 (Suppl 1):S1–S364 13 DNA, TMB and MSI genomic signatures in DNA, as well as RNA fusions and splice variants. Limit of Detection was as low as 1.6% variant allele frequency for small DNA variants, 2-fold change for gene amps, 10 support- ing reads for fusions, and 19 supporting reads for splice variants. Specificity was 99.9999% for small DNA variants, 99.9% for fusion variants and 100% for gene amps and MSI. Positive Percent Agreement (PPA) with whole exome sequencing (WES) for small DNA variants was 84.7% (382/451) for WES somatic and 99.8% (33,163/33,224) for WES germline variants. Conclusion: DNA small variant Negative Percent Agreement was 99.999% (70,000,481/70,000,907). PPA for gene amps was 92.3% (337/365), for MSI status 93% (40/43). PPA for RNA fusions and for RNA splice variants was 80.5% (70/87). Qualitative precision analysis across multiple operators, instruments, reagent lots, and days showed high concordance (>90% positive percent call). This CGP assay helps maximize the ability to find actionable bio- markers and help inform therapy decisions according to clinical guidelines that have the potential to improve patient management. MD-01-003 Evaluation of MET amplification in lung cancer via Idylla™ GeneFusion cartridges J. Siemanowski*, V. Welter, A. Ehteschami, A. Quaas, S. Merkelbach-Bruse *University hospital Cologne, Germany Background & objectives: MET amplification in lung cancer is known as resistance mechanism to epidermal growth factor receptor tyrosine–kinase inhibitors. This retrospective study used Idylla™ GeneFusion assay (Biocartis) cartridges to develop a delta Cq cut-off for discrimination of non-amplified versus MET ampli- fied samples. Methods: A cohort of 70 samples including 31 non-amplified and 39 MET amplified samples was analysed. MET amplification sta- tus was previously detected by fluorescence in situ hybridization (FISH). One to five 10 μm slices with a tumour cell content (TC) between 10% and 90% of the same FFPE tumour tissues were taken for analysis with the Idylla TM GeneFusion Assay. Results: For initial data analysis a threshold of -3.0 (Delta housekeeping gene-MET) was used for discrimination between non-amplified and MET amplified samples. Hereby 27 out of 31 non-amplified samples (specificity= 87%) and 26 out of 39 MET amplified samples (sensitivity=67%) were detected. After a threshold adjustment from -3.0 to -2.0 the specificity was lowered to 74% (23/31 non-amplified samples) but sensitivity increased to 84% (33/39 MET amplified samples). No further optimization was reached by implementing a % TC cut-off since false positives and false negatives were distributed at different TC content. All top-level amplifications (copy number gain (CNG) > 10) were true positive. Conclusion: This study showed that the Idylla TM GeneFusion Assay might be a promising screening tool for top level MET amplification assessment in lung cancer samples. Nevertheless, even after threshold adjustment both specificity and sensitivity within different levels of MET amplification remained lower than 90%. Therefore, this test with our proposed cut-off is not suited to identify MET amplification at lower thresholds. This is in line with other published PCR-based approaches using for example next generation sequencing. MD-01-004 Shallow whole genome sequencing accurately detects homolo- gous recombination deficiency in ovarian cancer J. Dagher*, N. Scamuffa, V. Vocat-Mottier, K. Lefort, L. de Leval, B. Bisig, E. Missiaglia *Institute of Pathology, CHUV, Switzerland Background & objectives: Homologous Recombination Deficiency (HRD) predicts benefit from PARP inhibitors. MyChoiceCDx test (Myriad) is approved to assess HRD, based on BRCA1/2 mutations and genomic instability (GIS) score. We assessed HRD by shallow whole-genome sequencing (sWGS) using Large-scale Genomic Altera- tion (LGA) score. Methods: Fifteen high-grade serous ovarian carcinomas underwent sWGS on a NextSeq550 (Illumina), using FFPE-extracted DNA (tumour content ≥25%). All samples had known tumour BRCA1/2 status (9 mutated, 6 wild-type), while 9 also had Myriad GIS score available. Raw reads were aligned to hg19 human genome assembly and analysed using shallowHRD software to compute LGA score, which was considered positive if ≥20. Results: We successfully performed sWGS on all samples, generat- ing an average of 22 million (M) reads per sample (11-42M) and achieving a mean coverage of 0.8X (0.3X-1.6X). LGA computation was still robust by artificially down-sampling to 5M reads (0.2X coverage). LGA score was positive in 8/9 BRCA -mutated cases, and negative in all wild-type cases (median 26 (11-40) vs. 6.5 (0-14), Wilcoxon rank test P=0.004). Correlation between LGA and GIS scores was statistically significant (Spearman rank rho 0.78; P=0.014), with only 1/9 cases showing discordant results. This was a BRCA wild-type sample with positive GIS score but negative LGA score, presumably due to low tumour cell content (25%) and low DNA quality. Conclusion: In our series of 15 high-grade serous ovarian carci- nomas, determination of LGA score by sWGS – using a predefined positivity cut-off ≥20 – was highly concordant with both BRCA1/2 mutational status and Myriad GIS score. Despite a limited number of samples analysed so far, this preliminary cohort shows promis- ing results, supporting further work to refine the cut-off value(s) and validate this tool as a predictive molecular biomarker for PARP inhibitor therapy, for patients with ovarian cancer or potentially other malignancies. MD-01-005 Microsatellite instability in intestinal T-cell lymphomas V. Rattina, D. Vallois, J. Bouilly, E. Lingre, V. Vocat Mottier, L. Veloza Cabrera, K. Lefort, B. Bisig, L. de Leval, E. Missiaglia* *Institute of Pathology, University Hospital Lausanne (CHUV), Switzerland Background & objectives: Microsatellite instability (MSI) is a consequence of defective DNA mismatch repair (MMR), leading to a hypermutant phenotype, with a high tumour mutational burden (TMB). We explored the potential impact of MSI in the oncogen- esis of primary intestinal T-cell lymphomas (ITCLs). Methods: Whole Exome Sequencing (WES) was performed on 54 ITCLs and matched non-tumour DNA: 34 MEITLs (Monomor- phic Epitheliotropic Intestinal T-cell Lymphomas) and 20 EATLs (Enteropathy-Associated T-cell Lymphomas). Mutation signatures were extracted from somatic variants and compared with COSMIC reference signatures (MuSiCa package). MSI status was predicted from instability scores (MANTIS software), and assessed by PCR (5 mononucleotide markers) on 43 samples. Results: We observed an overall median TMB of 1.8 non-syn- onymous somatic mutations per Mb, with a higher median in EATLs relative to MEITLs (2.3 vs 1.7/Mb). Nonetheless, the highest TMBs were detected in three MEITLs (maximum: 15/ Mb). Accordingly, these three cases showed signatures associated with MMR deficiency, and high instability scores as measured S59