膵臓癌の早期発見に関する新しい希望

オリジナルソース

タイトル: Multiethnic radio genomics reveals low-abundancy microRNA signature in plasma-derived extracellular vesicles for early diagnosis and subtyping of pancreatic cancer

概要: PurposeCurrently there is a lack of effective methods to accurately detect pancreatic cacer. In our study, we develop a liquid biopsy signature of EV miRNAs based on associated radiomics features of patients tumors in order to provide new insights for the early diagnosis of pancreatic cancer. Experimental DesignA total of eight datasets enrolled in this study, featuring clinical and imaging data from different benign pancreatic lesions and malignant pancreatic cancers as well as small RNAseq data from cargo of plasma extracellular vesicles of tumor patients. Radiomics Feature Extraction and different features analysis performed by limma packages. Feature selection was performed by Boruta algorithms and radiomics related signature model was build and validated by lasso regression algorithms. Radiomic signature related to low abundance EV miRNAs was analyzed by weighted gene co-expression network analysis. The diagnosis ability of above miRNA are validated by ten machine-learning algorithms. The shared target of candidate miRNAs were predicted and clustered followed by subsequently probing for predicting survival benefit of the patient, drug sensitivity of tumor cells and functional differences. ResultsA total of 88 significant radiologic features demonstrate differences between benign lesion and pancreatic cancer. Three radiomics factor related signature related a plasma EV-miRNAs triplet possessing high accuracy in diagnosis cancer from benign lesions. Moreover, clustering miRNA and there predicted molecular signaling partners in tumor tissue identified tow molecular subtypes of pancreatic cancer. Cluster stratification separates low risk tumors in terms of severely prolonged overall survival time of patients, higher sensitivity to immune therapies. We also propose the potential of purposing selected targeted drugs to specifically targeting the molecular activation markers in high-risk tumor cluster. ConclusionOur three radiogenomics related blood plasma extracellular vesicle microRNA signature is a useful liquid biopsy tool for early diagnosis and molecular subtyping of pancreatic cancer, which might treatment decision making. Statement of translational relevanceThe identification of a low-abundance microRNA signature in plasma-derived extracellular vesicles offers significant translational potential for the early diagnosis and subtyping of pancreatic cancer, particularly across diverse ethnic populations. This discovery could lead to the development of non-invasive liquid biopsies that improve early detection rates, a critical need for a cancer with notoriously poor prognosis due to late diagnosis. By incorporating this microRNA signature into clinical practice, oncologists may be able to detect pancreatic cancer at earlier, more treatable stages, enhancing patient survival rates. Additionally, the subtyping capability of this signature could guide personalized treatment strategies, allowing for more targeted therapies based on specific cancer subtypes. This could ultimately reduce the need for invasive diagnostic procedures and optimize treatment efficacy, reducing adverse effects and improving outcomes. The integration of radiogenomics and liquid biopsy technologies promises to be a powerful tool in the future of cancer medicine, particularly in underserved populations.