ECP 2022 Abstract Book

Virchows Archiv (2022) 481 (Suppl 1):S1–S364 13 E-PS-15-003 Prediction of biochemical recurrence in TMPRSS2-ERG- positive prostate cancer by neural network based on pathway enrichment V. Pavlov*, A. Kobelyatskaya, E. Pudova, A. Snezhkina, K. Nyushko, D. Kalinin, G. Krasnov, A. Kudryavtseva *Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Moscow, Russia Background & objectives: Prostate cancer (PCa) is one of the most common diseases in men. Biochemical recurrence occurs in 20-50% of patients after radical prostatectomy. Nearly half of PCa cases are of the TMPRSS2-ERG molecular subtype. Methods: The study used RNA-Seq data for 154 locally advanced PCa samples from TCGA. Data analysis was performed in the R (v.3.6.3) using the edgeR (v.3.24.3). GSVA was performed using the GSVA package (v.1.34.0) with the GO, KEGG, and Reactome databases. To build a fully connected neural network (FCNN) as a predictive model, the keras and tensorflow libraries were used. Results: We identified 7 differential enriched biological pathways between groups of biochemical recurrence (BCR) and biochemical recurrence free (BRF) cases within TMPRSS2-ERG-positive PCa: phytol metabolic process (GO:0033306, rs = -0.31), fatty alcohol metabolic process (GO:1903173, rs = -0.31), positive regulation of amyloid fibril formation (GO:1905908, rs = -0.47), heparanase activity (GO :0030305, rs = -0.38), positive regulation of ryanodine- sensitive calcium-release channel activity (GO:0060316, rs = -0.44), centriolar satellite (GO:0034451, rs = 0.34) and Regulation of gene expression in endocrine-committed (NEUROG3+) progenitor cells (R-HSA-210746, rs = 0.49). The model based on combination GO:1905908 + GO:0034451 + GO:0030305 + R-HSA-210746 + GO:0060316 was most promising with AUC = 0.94. Conclusion: Thus, we have identified changes in regulation of biological pathways that are involved in phytanic acid metabolism, amyloid bodies formation, neuroendocrine transformation, and cell differentiation, evidently inhibiting apoptosis and inducing tumour cell proliferation for BCR group within TMPRSS2-ERG-positive PCa. Based on changes in the above biological pathways, we built the neural network as a predictive model. This work was funded by the Russian Science Foundation grant no.18- 75-10127. This work was performed using the equipment of EIMB RAS “Genome” centre ( http://www.eimb.ru/ru1/ckp/ccu_genome_c.php) . E-PS-15-004 Is next generation sequencing the new golden standard for ALK testing? The diagnostic journey of an atypical case R. Pirlog*, N. Piton, F. Marguet, A. Lamy, J. Sabourin *Iuliu Hatieganu University, Romania Background & objectives: Identification of ALK rearrangements in lung adenocarcinomas (LUAD) is done by immunohistochemistry (IHC) followed by fluorescent-in-situ hybridization (FISH). However, multiple discordances between IHC and FISH analyses have been reported. Therefore, a more specific method to identify these altera- tions is needed. Methods: We present a case of a patient diagnosed in 2013 with a primary LUAD and in 2018 with a metastatic recurrence. ALK IHC was performed using clone D5F3. FISH analysis was carried out using Vysis ALK Break-Apart FFPE FISH Probe Kit. In 2021 the molecular analysis was carried out by Next Generation Sequencing (NGS) using Archer® FusionPlex Lung panel kit. Results: In 2013 the right inferior lobe resection showed an infil- trative LUAD, pT2aN0MxL0N0V0. The tumour was ALK-positive on IHC but negative on FISH. Molecular biology analysis was negative for EGFR, KRAS, and BRAF genes. In 2018 was identified a 9 mm LUAD was histologically compat- ible with the primary lesion from 2013 pT2aN0MxL0N0V0. The tumour was intensively positive for ALK on IHC but considered negative on FISH. Interestingly, the FISH analysis revealed a loss of the red signal in 75% of the nuclei analysed. Molecular biology analysis remained negative. In 2021 NGS analysis using the Archer® FusionPlex Lung panel kit (Archer®) revealed an unusual fusion between ALK exon 2–EML4 exon 13. Conclusion: The unusual loss of the red signal suggested the presence of an atypical rear rangement in the 5’ part of the ALK gene which required additional investigation. Using a NGS panel we identified an atypical EML4-ALK fusion which was not detected by break-apart conventional FISH. We highlight the need for a new gold standard for ALK rearrangements testing represented by RNA NGS, a more sensitive approach that could identify patients with rare molecular phenotypes and increase the addressability of targeted therapies. E-PS-15-005 Extractions made easy: fully automated extraction and quanti- tation of nucleic acid using Genexus™ purification system and ready to use consumables T. Jayaweera*, N. Siepert, B. Hradecky, E. Ostrowska, T. Delacour, K. Bramlett *ThermoFisher Scientific, USA Background & objectives: Reliable nucleic acid (NA) extraction and quantitation can be challenging in most labs due to lack of expertise and processes in place. Here we report fully automated Genexus™ purification system to extract and quantify NA with ready to use consumables. Methods: Four purification kits with automated workflows were developed to extract NA from FFPE, liquid biopsy, whole blood, PBL, BMA, cells, and tissue samples. An interactive user interface with touch screen allows intuitive run setup with customizable sample information, elution volumes and ability to select onboard quantitation. If opted, extracted nucleic acid is then quantified onboard using a relevant Qubit™ assay. Results: To evaluate Genexus™ purification system per- formance, various samples and input levels were tested and compared with existing extraction methods. gDNA (n=140) and total RNA (n=144) were extracted from sample inputs as low as 50 μL. DNA (n=72) and RNA (n=58) was also extracted from FFPE tumour resections as well as CNBs and FNAs. cfTNA (n=24) extractions were performed using up to 8mL of plasma. Study showed comparable NA extraction efficiency and quantitation accuracy to existing MagMax™ kits and manual Qubit™ assays respectively. All extracted NA was successfully used for downstream applications such as quantitative PCR or NGS confirming the high quality of extracted NA. Conclusion: This report demonstrates the use of Genexus™ puri- fication system to reliably extract and quantify NA from a variety of challenging clinical research samples and across multiple input levels to be successfully used in downstream applications. Ready to use consumables with REACH compliant reagents, fully auto- mated extraction and quantitation allows the Genexus™ purifica- tion system to be used with minimum user interaction, training and expertise. S305