Chen Yi, Xie Tiansong, Chen Lei, Zhang Zehua, Wang Yu, Zhou Zhengrong, Liu Wei
Department of Radiology, Fudan University Shanghai Cancer Center & Department of Oncology, Shanghai Medical College, Fudan University, Shanghai, 200032, China.
Shanghai Institute of Medical Imaging, Shanghai, 200032, China.
Eur Radiol. 2025 May;35(5):2692-2701. doi: 10.1007/s00330-024-11143-2. Epub 2024 Oct 24.
To investigate the value of dual-layer spectral computed tomography (DLCT) parameters derived from primary tumors in predicting lymph node metastasis (LNM) of resectable pancreatic ductal adenocarcinoma (PDAC).
In this retrospective study, patients with resectable PDAC who underwent DLCT within 2-week intervals before surgery were enrolled and randomly divided into training and validation sets at a 7:3 ratio. The patients' clinical data, CT morphological features, and DLCT parameters were analyzed. Univariate and multivariate logistic analyses were used to identify the predictors and construct a predictive model, and receiver operator characteristic (ROC) curves were programmed to evaluate the predictive efficacy.
We enrolled 107 patients (44 patients with LNM and 63 patients without LNM). Among all variables, iodine concentration in the venous phase, extracellular volume, and tumor size were identified as independent predictors of LNM. The nomogram model, incorporating the two DLCT parameters and the morphological feature, achieved an area under the curve (AUC) of 0.877 (95% confidence interval [CI]: 0.803-0.952) and 0.842 (95% CI: 0.707-0.977) for predicting LNM in the training and validation sets, respectively. Furthermore, the AUC of the nomogram model was greater than that of morphological features of lymph nodes in the training (AUC = 0.877 vs. 0.570) and validation (AUC = 0.842 vs. 0.583) sets.
DLCT has the potential to predict LNM in patients with resectable PDAC and show a better predictive value than morphological features of lymph nodes.
Question Morphological features of lymph nodes are of limited value in detecting metastatic lymph nodes in pancreatic ductal adenocarcinoma (PDAC). Findings Dual-layer spectral computed tomography (DLCT) parameters and morphological features derived from PDAC lesions show good preoperatively predictive efficacy for lymph node metastasis. Clinical relevance The proposed DLCT-based nomogram model may serve as an effective and convenient tool for preoperatively predicting lymph node metastasis of resectable PDAC.
探讨原发性肿瘤的双层光谱计算机断层扫描（DLCT）参数在预测可切除胰腺导管腺癌（PDAC）淋巴结转移（LNM）中的价值。
在这项回顾性研究中，纳入术前2周内接受DLCT检查的可切除PDAC患者，并按7:3的比例随机分为训练集和验证集。分析患者的临床资料、CT形态学特征和DLCT参数。采用单因素和多因素逻辑回归分析确定预测因素并构建预测模型，绘制受试者工作特征（ROC）曲线评估预测效能。
共纳入107例患者（44例有LNM，63例无LNM）。在所有变量中，静脉期碘浓度、细胞外容积和肿瘤大小被确定为LNM的独立预测因素。包含两个DLCT参数和形态学特征的列线图模型在训练集和验证集中预测LNM的曲线下面积（AUC）分别为0.877（95%置信区间[CI]：0.803 - 0.952）和0.842（95%CI：0.707 - 0.977）。此外，列线图模型在训练集（AUC = 0.877对0.570）和验证集（AUC = 0.842对0.583）中的AUC均大于淋巴结形态学特征的AUC。
DLCT有潜力预测可切除PDAC患者的LNM，且比淋巴结形态学特征具有更好的预测价值。
问题 淋巴结形态学特征在检测胰腺导管腺癌（PDAC）转移淋巴结方面价值有限。发现 源自PDAC病变的双层光谱计算机断层扫描（DLCT）参数和形态学特征对淋巴结转移显示出良好的术前预测效能。临床意义 所提出的基于DLCT的列线图模型可作为术前预测可切除PDAC淋巴结转移的有效且便捷工具。
