MicroRNA Classifier Based on EUS-FNA for the Early Detection of Occult Liver Metastasis in Pancreatic Ductal Adenocarcinoma.
Objective: This study aimed to investigate whether microRNAs (miRNAs) quantified from primary pancreatic ductal adenocarcinoma (PDAC) samples could serve as potential biomarkers for identifying PDAC with occult liver metastasis.
Background: Undetectable occult liver metastasis impair the survival of PDAC. Novel biomarkers are immediately required to detect occult liver metastasis in PDAC.
Methods: Primary PDAC samples were collected using endoscopic ultrasound-guided fine needle aspiration (EUS-FNA). In the discovery stage, miRNA profiles were assessed in tissues from non-metastatic PDACs and PDACs with liver metastasis (n=10 each) using small RNA sequencing. Liver metastasis-associated miRNA signatures were assessed in a training cohort (n=114) and validated in two independent cohorts (n=116), including one nested case-control cohort.
Results: Between October 01, 2020, and July 01, 2023, 250 patients were recruited. Data from the discovery cohort consisted of 34 miRNAs that were significantly deregulated in PDACs with liver metastasis, compared with non-metastatic PDACs. One miRNA classifier (Cmi) comprising three miRNAs (miR-483-5p, miR-6734-5p, and miR-548q) was established in the training cohort and evaluated in two validation cohorts for the predictive efficacy of liver metastasis. In the nested case-control cohort, Cmi identified occult liver metastasis with high efficacy based on EUS-FNA. Additionally, considering the variation in cancer antigen 125 (CA125) between groups and its association with metastasis, Cmi was combined with CA125. The area under the receiver operating characteristic curve for Cmi ranged from 0.798 to 0.814, which was superior to the AUCs of CA125 alone (P<0.05). It demonstrated comparable efficacy to the combined Cmi+CA125 model (P>0.05 across all cohorts).
Conclusions: The proposed miRNA-based model holds promise in detecting occult liver metastasis and guiding clinical stratification for optimizing PDAC treatment.