Nomogram based on preoperative contrast-enhanced CT predicts early liver metastasis in pancreatic ductal adenocarcinoma

Background: Early liver metastasis (LM) of pancreatic ductal adenocarcinoma (PDAC) is a critical factor affecting curative resection and poor prognosis. However, contrast-enhanced computed tomography (CECT) demonstrates limited sensitivity in detecting early-stage LM from PDAC. We aimed to construct a nomogram from conventional CECT features and preoperative tumor markers for early prediction of LM from PDAC. Methods: The preoperative tumor markers and CECT images of 184 patients with pathologically confirmed PDAC (from January 2015 to November 2021) were retrospectively analyzed. We compared patients with early and non-early LM from PDAC in terms of CECT features and tumor markers. Finally, multivariate analysis was performed to identify independent risk factors for early LM from PDAC. Results: Preoperative carcinoembryonic antigen, carbohydrate antigen 125 (CA125), minimum distance between the tumor and superior mesenteric artery or splenic artery (TA distance), minimum distance between the tumor and superior mesenteric vein, portal vein, or splenic vein (TV distance), maximum hug angle between the tumor and superior mesenteric artery or splenic artery (TA angle), maximum hug angle between the tumor and superior mesenteric vein, portal vein, or splenic vein (TV angle), tumor longitudinal diameter, and peritumoral fat fuzzy (PFS) differed significantly between PDAC patients with early and non-early LM. Multivariate analysis revealed that TV distance [odds ratio (OR): 0.714, 95% confidence interval (CI): 0.514-0.948, P=0.028], TV angle >180° (OR: 6.718, 95% CI: 1.481-35.822, P=0.017), and PFS (OR: 5.893, 95% CI: 1.322-31.349, P=0.025) were independent risk factors for early LM from PDAC. The area under the curve (AUC) value, sensitivity, and specificity of the nomogram model constructed based on univariate analysis of the training and validation cohorts were 0.944 (95% CI: 0.892-0.995), 95.0%, and 88.6%, and 0.921 (95% CI: 0.8577-0.985), 94.1%, and 80.0%, respectively. Conclusions: A nomogram constructed based on preoperative conventional CECT features could predict early LM from PDAC, thereby assisting clinical decision-making regarding treatment.