Establishment and validation of a diagnostic model for evaluating synchronous liver metastasis in pancreatic cancer using comprehensive blood biochemical indicators

利用综合血液生化指标建立并验证胰腺癌同步肝转移诊断模型

阅读:1

Abstract

The liver is the most common metastatic target organ of pancreatic cancer (PC). Currently, imaging examination is effective for detecting liver metastases (LM) of PC, but some small metastases are difficult to detect, making it necessary to establish a comprehensive diagnostic model with which to predict LM. A total of 59 patients with PC were enrolled as the training cohort and 16 patients with PC were included as the external validation cohort. The 59 patients in the training cohort were divided into LM and No-LM groups. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors for synchronous liver metastasis (SLM) in PC. Based on these findings, a diagnostic model was constructed and a nomogram was developed to facilitate practical application. The accuracy and reliability of this diagnostic model were then evaluated using the area under the receiver operating characteristic curve (AUC), Hosmer-Lemeshow (HL) curves and decision curve analysis (DCA). Multivariate analysis identified CEA [odds ratio (OR)=1.05, 95% CI: 1.01-1.08], CA153 (OR=1.18, 95% CI: 1.06-1.31), white blood cells (WBC; OR=1.71, 95% CI: 1.08-2.72) and platelets (PLT; OR=1.01, 95% CI: 1.00-1.03) as independent risk factors. In the training and external validation cohorts, the diagnostic efficacy of the model's AUC was 0.92 and 0.90, respectively, with sensitivities of 0.96 and 0.83, and specificities of 0.86 and 0.75, respectively. The HL and DCA curves indicate the excellent calibration and clinical net benefit of the model. In conclusion, the diagnostic model integrating CEA, CA153, WBC and PLTs shows high predictive performance for identifying SLM in patients with PC.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。