Development and Validation of a Multiple Tumor Marker Nomogram for Predicting Preoperative Metastasis in Patients with Colorectal Cancer

Jiasi Wang,Xiao Chen,Zhaoping Zeng,Xuan Huang,Jiaojiao Zhao,Li Song,Qianrong Xie,Jie Liu,Yin Zhang,Yanpeng Chu,Xiuqin Zhang,Fanxin Zeng

DOI: https://doi.org/10.2139/ssrn.3523840

2020-01-01

Abstract:Background: Preoperative prediction of colorectal cancer (CRC) metastasis is critical to the decision of surgery, prognosis, and therapy strategy. In this study, we aimed to develop and validate a multiple tumor marker nomogram for predicting preoperative metastasis in patients with CRC. Methods: From January 2014 to December 2018, 687 patients with histologically diagnosed CRC were collected, who were randomly divided into primary cohort and validation cohort according to the ratio of 7:3. Clinicopathologic variables, including gender, age, preoperative histologic grade, carcinoembryonic antigen (CEA), alpha-fetoprotein (AFP), carbohydrate antigen (CA) 125, CA153, and CA19-9, were derived from electronic medical records. The least absolute shrinkage and selection operator (LASSO) regression model was carried out for variable dimensionality reduction and feature selection. Multivariate logistic regression analysis was conducted to develop predictive models of preoperative metastasis. We combined multiple tumor markers with clinical risk factors and presented them as multi-tumor markers nomogram. The performance of the established model was evaluated by the receiver operating characteristic (ROC) curve, calibration belt, and clinical usefulness. Findings: Based on the primary cohort patients, 19 potential features were reduced to 6 predictors by LASSO regression respectively. The model demonstrated good discrimination and ROC curve (AUC, 0.709) in preoperative CRC patients. Good calibration was shown for the probability of CRC metastasis in the primary and validation cohorts. Decision curve analysis illustrated that multi-tumor markers nomogram was clinically useful. Interpretation: The study proposes a reliable prediction nomogram that can be efficiently and conveniently utilized to facilitate the preoperative prediction of individually-tailored metastasis in patients with CRC.Funding Statement: This study was supported by the National Natural Science Foundation of China (81902861), the Scientific Research Fund of Sichuan health and Health Committee (No. 18PJ361) and the Scientific Research Fund of Technology Bureau in Sichuan Province (No. 2018138, No. 2018JY0324). Declaration of Interests: The authors declare no competing financial interests.Ethics Approval Statement: The retrospective analysis was approved by the Medical Ethics Review Board of the Dazhou Central Hospital. And the Medical Ethics Review Board abandoned the need for informed consent from participants in the study.