Molecular phenotypes of colorectal cancer.

e15506 Background: Despite the positive results of colorectal cancer treatment, there is a group of patients that progresses despite successful treatment for I–II stage of the disease. it may be that modern predictors in do not fully cover the entire spectrum of molecular genetic traits, especially the tumor microenvironment. Perhaps expression analysis is not only tumor tissue, but resection margin could also reach a consensus on the molecular gene phenotypes of colorectal cancer with the achievement of the goal of prognostic efficiency and therapeutic and prophylactic diagnostic tactics. Methods: The expression level of 64 genes in tumor tissue in samples obtained from 172 patients with stage I–III CRC was analyzed by real-time Polymerase chain reaction (PCR). The mean follow-up time was 28.15±14.7 months. Data analysis and predictive modeling arrays using the R 3.6.3 Computation Evaluation Framework (R Foundation for statistical computing, Vienna, Austria) using additional packages rms 5.1-4, survival 3.2-7 and pROC 1.16.2. Results: As a result of statistical analysis, data were obtained on the presence of predictors of the studied unfavorable source code among the selected sets of genes carcinogenesis in colorectal cancer. The next step revealed progression status with BIRC5, IL2, SCUBE2, MYC and P16INK4A gene expression as in tissues tumors, and in resection margin. 2 regression models were built to predict survival without progression in patients with the inclusion of the level of expression in the tumor tissue- nor, the inclusion of the level of gene expression in tumor and resection margin. Model with inclusion only tumor transcripts statistics did not differ from models without predictors. Conclusions: In this study, it was possible to identify a group of genes associated with certain favorable prognosis and responsible in one way or another for proliferation, angiogenesis, immune status of the tumor. It can be concluded that the study of the level of expression of numerical genes in tumor tissue can become a promising direction for the selection of patients in the group of poor prognosis in colorectal cancer.