Using a Microrna Panel of Circulating Exosomes for Diagnosis of Pancreatic Cancer: Multicentre Case-Control Study.

Gang Yang,Jiangdong Qiu,Jianwei Xu,Guangbing Xiong,Fangyu Zhao,Zhe Cao,Guangyu Chen,Yueze Liu,Jinxin Tao,Lianfang Zheng,Lei Wang,Renyi Qin,Taiping Zhang

DOI: https://doi.org/10.1093/bjs/znac375

2023-01-01

Abstract:Pancreatic cancer is one of the most lethal cancers worldwide and the 5-year survival rate is less than 11 per cent1. The accurate diagnosis of pancreatic cancer remains challenging as the sensitivity and specificity of carbohydrate antigen (CA) 19-9 for its diagnosis are only 79 and 82 per cent respectively2. Liquid biopsy including exosomes has been highlighted as a novel strategy for pancreatic cancer diagnosis, but large-sample, multicentre studies are urgently need3,4. The present study screened the microRNA (miRNA) profiles in circulating exosomes, identified the ideal exosomal miRNAs for pancreatic cancer diagnosis by a three-stage process in a three-centre cohort, and constructed a panel whose diagnostic value was superior to that of CA19-9. This multicentre study was conducted at three high-volume pancreatic centres in China. A total of 262 peripheral blood samples were obtained from patients with pancreatic cancer, chronic pancreatitis (CP), other pancreatic tumours (OPT), and negative controls (NC; healthy volunteers without chronic or malignant diseases). These were divided into three groups: discovery group (16), preliminary group (56), and large-sample group (190). Exosomal miRNAs were extracted from plasma, and a miRNA microarray was applied to screen the exosomal miRNA expression profiles in the discovery group. MicroRNAs that were differentially expressed between pancreatic cancer and other groups were selected for further validation in the preliminary group and large-sample group by real-time quantitative PCR (RT-qPCR). The exosomal miRNA panel was constructed based on miRNA expression values by logistic regression. The test performance of the panel, CA19-9, and their combination in discriminating pancreatic cancer from CP, NC, and OPT was evaluated using receiver operating characteristic (ROC) curves and areas under the curve (AUCs). Further details of methods are provided in the Supplementary material.