The Predictive Value of Dual-layer Spectral Detector CT Parameters on Lymph Node Metastasis of Nonfunctional Pancreatic Neuroendocrine Tumor
10.11714/jsysu.med.YX20260084
- VernacularTitle:能谱CT参数对胰腺无功能性神经内分泌肿瘤淋巴结转移的预测价值
- Author:
Mingjie CHEN
1
;
Ke LI
1
;
Jiadan LUO
1
;
Jiayi YANG
2
;
Huasong CAI
1
;
Shiting FENG
1
;
Yanji LUO
1
;
Siya SHI
1
Author Information
1. Department of Radiology， The First Affiliated Hospital， Sun Yat-sen University， Guangzhou 510080， China
2. Guangzhou Xinhua University， Guangzhou 510080， China
- Publication Type:Journal Article
- Keywords:
nonfunctional pancreatic neuroendocrine tumor;
lymph node metastasis;
dual-layer spectral detector ct;
least absolute shrinkage and selection operator analysis;
diagnostic efficacy
- From:
Journal of Sun Yat-sen University(Medical Sciences)
2026;47(5):868-879
- CountryChina
- Language:Chinese
-
Abstract:
ObjectiveTo investigate the value of preoperative dual-layer spectral detector CT （DLCT） parameters in predicting lymph node metastasis （LNM） of non-functional pancreatic neuroendocrine neoplasms （NF-pNENs）. MethodsA retrospective study was conducted on patients with pathologically confirmed NF-pNENs who underwent preoperative contrast-enhanced DLCT examination between September 2019 and July 2024. Regions of interest （ROIs） were manually delineated on three representative slices of the solid component of the lesion， and the mean values were recorded. Conventional CT parameters， monochromatic energy images （40 keV， 70 keV， 100 keV）， iodine concentration maps， and effective atomic number （Zeff） maps in the pancreatic parenchymal phase and portal venous phase were acquired. A total of 34 parameters， including clinical characteristics， imaging features， conventional CT and DLCT quantitative parameters， were analyzed. Intergroup differences were compared between the LNM and non-LNM groups. Independent predictive factors were screened using logistic regression and least absolute shrinkage and selection operator （LASSO） regression. Receiver operating characteristic （ROC） curves were generated for each predictive model， and the area under the curve （AUC） was calculated to evaluate the predictive efficacy. ResultsA total of 82 patients with NF-pNENs were enrolled， including 37 males and 45 females， with a mean age of 49.7±14.7 years. Thirty patients were performed with LNM and 52 were with non-LNM. After univariate logistic regression， LASSO regression and multivariate stepwise logistic regression analyses， body mass index （BMI）， organ invasion， normalized CT value in the portal venous phase （nCTv） and normalized iodine concentration in the portal venous phase （nICv） were identified as independent predictive factors. The AUCs of the clinical-radiological qualitative model （BMI+organ invasion）， conventional CT model （nCTv）， DLCT model （nICv） and comprehensive model （organ invasion+nICv） were 0.815 （0.719， 0.910）， 0.750 （0.642， 0.858）， 0.829 （0.733， 0.925） and 0.865 （0.879， 0.953）， respectively. The comprehensive model combining clinical-radiological and DLCT characteristics exhibited superior predictive performance. ConclusionDLCT parameters are more effective than conventional CT parameters in predicting LNM in NF-pNENs， enabling non-invasive preoperative prediction and assisting clinical decision-making.
