Nomogram for predicting tumor-stroma ratio in pancreatic ductal adenocarcinoma using dual-energy computed tomography.

Abstract

Background: This study aimed to develop and validate a nomogram to predict both the tumor-stroma ratio (TSR) and the overall survival (OS) of patients with pancreatic ductal adenocarcinoma (PDAC) using preoperative dual-energy computed tomography (DECT) parameters.

Methods: 153 patients with histopathologically confirmed PDAC who underwent preoperative DECT scans were retrospectively reviewed and divided into high- and low-TSR groups based on histological analyses of surgical specimens. Several DECT parameters of the primary tumor were measured, including the normalized iodine concentration (NIC), effective atomic number, slope of the energy spectrum attenuation curve (K), CT values (40-100 keV), and extracellular volume fraction (ECVf), and analyzed alongside clinical and radiological data. Univariate and multivariate logistic regression models were used to identify independent predictors, which were then incorporated into radiology, DECT, and nomogram models. The association of the nomograms with OS was assessed using Kaplan-Meier curves and Cox regression analysis.

Results: CT-reported lymph node status, NIC_venous, K_venous, and ECVf were identified as independent predictors of the TSR and included in the nomogram model. The nomogram demonstrated high predictive accuracy with an area under the receiver operating characteristic curve of 0.934 in the training set and 0.891 in the validation set, outperforming the radiology model (0.715 and 0.692, respectively). Patients with a high predicted TSR exhibited worse OS than those with a low predicted TSR.

Conclusion: The DECT-based nomogram model provides a noninvasive and accurate preoperative prediction of the TSR and prognosis of patients with PDAC and may assist in individualized risk stratification and treatment planning.