Preoperative MRI for non-invasive assessment of pancreatic acinar cell composition

This study explores the predictive power of preoperative MRI in evaluating pancreatic acinar cell composition. We aimed to determine whether MRI-derived signal intensity ratios could serve as reliable biomarkers for preoperative assessment.

In this retrospective study, 47 patients (30 males, 17 females; mean age: 64.6 years) who underwent pancreatoduodenectomy between January 2022 and May 2024 were included. Preoperative MRI signal intensities from T1-weighted sequences, dynamic contrast-enhanced phases, and diffusion-weighted imaging were quantified. The acinar cell ratio was determined via histopathological analysis of the pancreatic resection margin. Pearson and Spearman correlation tests assessed associations between MRI parameters and acinar cell content, while a logistic regression model evaluated their predictive accuracy.

A significant correlation was observed between various MRI parameters and pancreatic acinar cell ratio. The strongest association was found for the pre-contrast pancreas T1-weighted to venous phase signal intensity ratio (Ratio-3), which yielded a correlation coefficient of 0.696 (p < 0.0001). ROC analysis demonstrated that Ratio-3 had an area under the curve (AUC) of 0.91, with 94 % sensitivity and 80 % specificity in predicting an acinar cell ratio exceeding 40 %.

Ratio-3 (Pre-contrast pancreas T1W/Venous phase signal) shows strong potential as a non-invasive biomarker for assessing pancreatic acinar cell composition. These findings may improve preoperative risk stratification, aiding in pancreatic fistula prediction and surgical planning.