Модель прогнозирования колоректального рака, основанная на определении уровня экспрессии микроРНК-21 в слюне и плазме.

Introduction. The development of new approaches to screening colorectal cancer (CRC), including those based on such molecular genetic methods as the assessment microRNA(miRNA) expression, is an urgent clinical task. Purpose study. study was aimed at a model for CRC prediction miRNA-21 expression level in saliva and plasma. Materials Methods. levels plasma (P-miRNA-21) saliva(S-miRNA-21) samples obtained from 65 patients with diagnosed colon (CRC group) 66 healthy volunteers were measured by reverse transcription polymerase chain reaction expressed conventional units (CU). Correlation analysis gender, age, localization stage process performed. Univariate used identify predictors. Based combination leading factors, risk classes formed subsequent construction decision tree. ROC estimate quality simulated tree CRR predicting model. statistical significance fixed 0.05. Results. P-miRNA-21 (3.71±7.38 CU) S-miRNA-21 (9.67±18.52 higher than control (p=0.0089 p<0.0001, respectively). sensitivity specificity P-miRNA-21>1.6 CU S-miRNA-21>2 markers 52% 89% 61% 83%, respectively. No correlation or disease found. In univariate predictors were: (Relative (CI), RR [CI): 2.51 (1.76; 3.58)], [RR (CI): 2.41 (1.75; 3.31)], age >61 years 2.65 (1.67; 4.21)], high embryonic antigen (CEA) 2.4 (1.93; 2.99)], СA-19.9 2.27 (1.85; 2.78)]. An algorithm diagnostics proposed form S-miRNA-21, which has (88.7%) (76.6%) model, AuROC 0.86. Conclusion. Evaluation both blood can be suitable non-invasive biomarkers cancer, especially part diagnostic