Combining novel miRNA biomarkers with CA19-9 to enhance early detection of pancreatic ductal adenocarcinoma.

Abstract

e16443 Background: Pancreatic ductal adenocarcinoma (PDAC) is the third leading cause of cancer-related death in the United States, with a 5-year relative all-stage survival rate of only 13% between 2013 and 2019. This poor patient outcome is primarily attributed to late-stage diagnosis when treatment options are limited. The limitations of commonly used biomarkers such as CA19-9, particularly the low sensitivity for early stage disease and lack of expression in certain populations, underscore the urgent need for non-invasive blood tests to enable early detection of PDAC. Methods: We performed small RNA sequencing on plasma samples from 92 pancreatic cancer patients and 130 healthy donors. miRNA expression was profiled through smrnaseq analytical pipeline from nf-core . Following quantile normalization of the expression profile, we identified seven potential miRNA biomarkers. Subsequently, we trained the classification model incorporating these novel biomarkers with CA19-9 concentrations to predict the clinical diagnosis using data from 45 pancreatic cancer patients and 45 healthy donors. The remaining samples were reserved for validation. Additionally, we randomly selected 10 pancreatic cancer patients and 10 healthy donors to conduct two independent sequencing experiments to assess the reproducibility of the classification model. Results: Our classification model based on CA19-9 and novel miRNA biomarkers achieves 83% accuracy with 72% sensitivity and 89% specificity. This approach significantly enhances sensitivity by 16% compared with the classification model using CA19-9 alone. We also demonstrate the robustness of our classification model with a reproducibility rate of 95%. Conclusions: We demonstrated the value of multi-modal approaches in advancing early pancreatic ductal adenocarcinoma detection. Our model enhanced diagnostic capability by integrating the distinct molecular signatures of miRNA and CA19-9. Such multi-modal diagnostic approach could be reliably translated into clinical settings, potentially improving the early detection of pancreatic cancer in routine medical practice.