Abstract B09: Molecular phenotyping of colorectal tumors in clinical practice: Assignment of extended prognostic subtypes by direct testing of endoscopic specimens

Introduction: Since the pioneering work of Fearon and Vogelstein in 1990, detailed mechanisms of colorectal neoplasia have intensively been studied (1). Only recently, however, importance of molecular subtypes for prognosis has been confirmed on large cohorts (2,3). The so-called Jass classification is set to become an part of diagnostics complementary to the companion predictive genotyping (4). The diagnosis is mostly based on examination of tumor tissue from endoscopy. It was an intention of this work to evaluate feasibility of molecular phenotyping in addition to the standard testing performed from the same source material. Patients and Methods: A total of 145 carcinomas (105 fresh biopsies or 40 FFPE sections, all stages) have been acquired over a 3-year period. The molecular subtypes were assigned based upon results of CIMP/MSI/BRAF/RAS (5). To partition the most common type arising from traditional adenoma-carcinoma pathway, we have additionally examined MLH1 methylation and APC and TP53 mutations (6). The methodology was based on previously validated protocols including MS-MLPA for the evaluation of CIMP, PCR fragment analysis MSI and high-sensitive denaturing CE assay (DCE) for somatic mutations in RAS, BRAF, TP53 and APC genes. By extending Jass classification we recognize 6 molecular subtypes with distinct features: CIMP status > negative > TP53 > positive > Traditional CIMP (median prognosis) > positive > MLH1 ------------> negative > KRAS -----------------------> positive > Traditional CIN antiEGFR resistant (median prognosis) -----------------------> negative > APC> positive > Traditional CIN antiEGFR sensitive (median prognosis) ------------> positive > BRAF --------------------> negative> MSI > positive > Familial MSI (good prognosis) --------------------> positive > MSI --------------------------------> negative > Serrated CIMP (poor prognosis) --------------------------------> positive> Serrated CIMP+MSI (best prognosis) Results: A total of 105/105 (100%) of fresh tissue samples and 33/45 (73.3%) of FFPE samples provided high-quality DNA for subsequent completion of the complete molecular testing panel. The distribution of the six types was (i) 3x Serrated CIMP (ii) 9x Serrated CIMP+MSI; (iii) 2x Familial MSI, (iv) 26x Traditional CIN antiEGFR resistant (v) 12x Traditional CIN antiEGFR sensitive and (vi) 18x Traditional CIMP. The remaining 68 carcinomas were resulting from other combinations. Conclusions: Methodology as well as logistics of sample processing has been optimized for use at endoscopy unit. The distribution of Jass molecular subtypes corresponded to the larger foreign cohorts with the dominating contribution from the Traditional CIN/CIMP subtypes. Until CIMP/CIN/MSI panels covering all markers have been developed a multi-tier testing according to the above algorithm is the most cost efficient. Assignment of molecular subtypes by an optimized protocol is useable in clinical management of patients and feasible in routine practice. Supported by IGA Ministry of Health project No. NT 14383 Literature 1. Fearon ER, Vogelstein B. Cell. 1990 Jun 1;61(5):759-67. 2. Phipps AI et al. Gastroenterology 2015; 148: 77-87.e2 3. Sinicrope FA et al. Gastroenterology 2015; 148: 88-99. 4. Modest DP et al. Ann Oncol. 2016 Jun 29. pii: mdw261. [Epub ahead of print] 5. Jass JR. Int J Colorectal Dis 1999; 14: 194-200 6. Kim JH, Bae JM, Cho NY, Kang GH. Oncotarget 2016; Epub ahead of print Citation Format: Marek Minarik, Tereza Halkova, Petra Minarikova, Barbora Belsanova, Stepan Suchanek, Jiri Cyrany, Jan Bures, Miroslav Zavoral, Lucie Benesova. Molecular phenotyping of colorectal tumors in clinical practice: Assignment of extended prognostic subtypes by direct testing of endoscopic specimens. [abstract]. In: Proceedings of the AACR Special Conference on Colorectal Cancer: From Initiation to Outcomes; 2016 Sep 17-20; Tampa, FL. Philadelphia (PA): AACR; Cancer Res 2017;77(3 Suppl):Abstract nr B09.