Value of quantitative imaging parameters combined with serum tumor markers in prognostic evaluation after pancreatic ductal adenocarcinoma surgery
- VernacularTitle:影像学定量参数联合血清肿瘤标志物对胰腺导管腺癌术后预后的评估价值
- Author:
Ningning FENG
1
;
Yueshan ZHANG
1
;
Han GAO
1
;
Baoming YANG
1
Author Information
- Publication Type:Journal Article
- Keywords: Pancreatic Neoplasms; Tomography， X-Ray Computed; Multiparametric Magnetic Resonance Imaging; Biomarkers， Tumor; Prognosis
- From: Journal of Clinical Hepatology 2026;42(6):1375-1382
- CountryChina
- Language:Chinese
- Abstract: ObjectiveTo investigate the application value of a predictive model constructed based on quantitative imaging parameters and serum tumor markers in predicting the postoperative prognosis of patients with pancreatic ductal adenocarcinoma （PDAC）. MethodsA retrospective study was conducted among 146 patients with PDAC who underwent radical resection in Hebei Medical University Fourth Hospital from April 2020 to March 2023， and the patients were divided into a training set with 102 patients and a validation set of 44 patients at a ratio of 7∶3 using the completely randomized method. All patients underwent enhanced computed tomography （CT） and multi-parametric magnetic resonance imaging scans， and CT values in the arterial phase， the venous phase， and the delayed phase were recorded， as well as apparent diffusion coefficient （ADC） at b = 800 s/mm2 and signal intensity （SI） of T2 weighted imaging. The serum levels of carbohydrate antigen 19-9 （CA19-9） and carcinoembryonic antigen （CEA） were also measured. The independent-samples t test was used for comparison of continuous data between groups， and the chi-square test was used for comparison of categorical data between groups. Kaplan-Meier survival curves were plotted for survival analysis； the univariate and multivariate Logistic regression analysis was used to investigate the association of clinical and imaging indicators with prognosis； the least absolute shrinkage and selection operator （LASSO）-Cox regression model was used to identify the important influencing factors for prognosis， and a prognostic prediction model was constructed. The receiver operating characteristic （ROC） curve was used to analyze the predictive value of the model in predicting prognosis in both the training set and the validation set. ResultsThe Kaplan-Meier survival curve analysis showed a median survival time of 33.00 months in the training set and 32.00 months in the validation set， with no significant difference between the training set and the validation set （P>0.05）. The univariate analysis showed that patient age， degree of tumor differentiation， lymph node metastasis， tumor stage， vascular invasion， CA19-9， CEA， ADC value， and SI value were significantly associated with the survival prognosis of patients （χ²=5.906， 13.116， 12.807， 17.277， 14.611， 7.275， 14.339， 9.506， and 13.137， all P<0.05）. The LASSO-Cox multivariate regression analysis showed that six factors were incorporated into the regression model， i.e.， tumor stage （HR=8.934， 95%CI： 3.215 — 21.562， P<0.001）， lymph node metastasis （HR=2.971， 95%CI： 1.298 — 5.647， P=0.002）， CA19-9 （HR=3.948， 95%CI： 1.758 — 8.994， P<0.001）， CEA （HR=1.965， 95%CI： 1.083 — 3.664， P=0.039）， ADC value （HR=2.873， 95%CI： 1.307 — 6.037， P=0.003）， and SI value （HR=3.107， 95%CI： 1.264 — 7.339， P=0.001）. The nomogram model constructed based on these factors had an area under the ROC curve of 0.845 （95%CI： 0.774 — 0.915） in the training set and 0.919 （95%CI： 0.870 — 0.967） in the validation set. ConclusionThe combined model of quantitative imaging parameters and serum tumor markers constructed based on LASSO-Cox regression can effectively predict the postoperative prognosis of PDAC patients， thereby helping to identify high-risk patients and guide decision-making for adjuvant therapy.
