Assessment of Salivary Mirna, Clinical, and Demographic Characterization in Colorectal Cancer Diagnosis Publisher



Koopaie M¹ ; Manifar S² ; Talebi MM¹ ; Kolahdooz S³ ; Razavi AE⁴ ; Davoudi M⁵ ; Pourshahidi S⁶ Show Affiliations
Source: Translational Oncology Published:2024

Abstract

Aim: Colorectal cancer (CRC), as the third most frequent malignancy in the world, is the fourth major cause of cancer-related mortality. Its early detection contributes significantly to a reduction in mortality. The objective of this case-control research was to analyze the salivary expression of microRNA-29a (miR-29a) and microRNA-92a (miR-92a), and also to consider demographic, clinical, and nutritional habits for differentiation between CRC patients and healthy controls, especially in the early stages. Method: A standard checklist was used to obtain the demographic information, clinical features, and dietary habits of the case and control groups. Samplings of whole unstimulated saliva samples were obtained from 33 healthy persons and 42 CRC patients. Through real-time PCR, statistical analyses, and machine learning analyses, miR-29a and miR-92a salivary expression levels were evaluated. Results: The mean salivary expression of miR-92a and miR-29a in CRC patients was significantly higher than in healthy controls (p < 0.001). The area under the receiver operating characteristic curve for miR-92a and miR-29a salivary biomarkers was 0.947 and 0.978, respectively. The sensitivity and specificity values for miR-92a were 95.24 % and 84.85 %, respectively, whereas sensitivity and specificity for miR-29a were equal to 95.20 % and 87.88 %, respectively. Multiple logistic regressions considering demographics, clinical features, and nutritional habits led to values of 95.35 % and 96.88 % as sensitivity and specificity, respectively, and machine learning analysis led to values of 88.89 % and 86.67 % as sensitivity and specificity, respectively. Conclusion: CRC could be accurately diagnosed based on miR-92a and miR-29a levels in saliva. Statistical analysis and machine learning might develop cost-effective models for the distinction of CRC using a noninvasive technique. © 2024