To assess the association between the enhancement pattern of the pancreatic parenchyma on preoperative multiphasic contrast-enhanced computed tomography (CECT) and the occurrence of postpancreatectomy acute pancreatitis (PPAP) after pancreaticoduodenectomy (PD). A total of 513 patients who underwent PD were retrospective enrolled. The CT attenuation values of the nonenhanced (N), arterial (A), portal venous (P), and late (L) phases in the pancreatic parenchyma were measured on preoperative multiphasic CECT. The enhancement pattern was quantized by the CT attenuation value ratios in each phase. Receiver operating characteristic (ROC) curve analyses were computed to evaluate predictive performance. Regression analyses were used to identify independent risk factors for PPAP. PPAP developed in 102 patients (19.9%) and was associated with increased morbidity and a worse postoperative course. The A/P ratio, P/L ratio, and A/L ratio were significantly higher in the PPAP group. On the ROC analysis, the A/L ratio and A/P ratio both performed well in predicting PPAP (A/L: AUC = 0.7579; A/P: AUC = 0.7497). On multivariate analyses, the A/L ratio > 1.29 (OR 4.30 95% CI: 2.62-7.06, p < 0.001) and A/P ratio > 1.13 (OR 5.02 95% CI: 2.98-8.45, p < 0.001) were both independent risk factors of PPAP in each model. The enhancement pattern of the pancreatic parenchyma on multiphasic preoperative CECT is a good predictor of the occurrence of PPAP after PD, which could help clinicians identify high-risk patients or enable selective enhance recovery protocols. Preoperative identification of patients at high risk for postpancreatectomy acute pancreatitis by enhancement patterns of the pancreatic parenchyma allows surgeons to tailor their perioperative management and take precautions. PPAP is associated with increased risk of postoperative complications and a worse postoperative course. A rapid-decrease enhancement pattern of the pancreatic parenchyma is related to the occurrence of PPAP. The A/L and A/P ratios were both independent risk factors of PPAP in each multivariate model.
Read full abstract