Analysis of risk factors and construction of prognostic model for liver metastasis in T 1 stage colorectal neoplasms patients / 国际外科学杂志

ABSTRACT

Objective:To explore the risk factors associated with liver metastasis in T 1 stage colorectal neoplasms patients and establish a prognostic model. Methods:Clinicopathological data of T 1 stage colorectal neoplasms patients from the Surveillance, Epidemiology, and End Results (SEER) database between January 1, 2004, and December 31, 2019 were included. The differences in clinicopathological characteristics between patients with and without liver metastasis were compared using Chi-square test, Wilcoxon rank-sum test, and multivariate Logistic regression analysis. Survival curves were plotted using Kaplan-Meier method, and the Log-Rank test was used to compare survival differences between the two groups. Univariate and multivariate analysis of prognostic factors were performed using the proportional Cox regression hazards regression model. Patients were randomly divided into training set and validation set at a ratio of 6∶4 using simple random sampling method. A nomogram model was established based on independent prognostic factors based on the results of the multivariate Cox regression analysis. The predictive ability of the model was evaluated using time-dependent receiver operating characteristic (ROC) curves and calibration curves. Results:A total of 28 258 T 1 stage colorectal neoplasms patients were included in the study. The multivariate Logistic regression analysis for liver metastasis occurrence indicated that the neuroendocrine neoplasms, elevated carcinoembryonic antigen (CEA) levels, larger neoplasms size, positive lymph node metastasis, and presence of cancer nodules were statistically significant differences ( P<0.05). For T 1 stage colorectal neoplasms patients with liver metastasis, the results of the multivariate Cox regression analysis showed that age, primary site, ethnicity, chemotherapy, and surgical treatment were independent prognostic factors ( P<0.05). The nomogram constructed based on these five prognostic factors had time-dependent ROC areas under the curve of 0.758, 0.797 and 0.729 for 1-year, 3-year, and 5-year survival rates, respectively, 0.749, 0.857, 0.871 in the validation set. The calibration curves of the nomogram in the training and validation sets were close to the 45-degree diagonal line. Conclusions:Neuroendocrine neoplasms, elevated CEA levels, larger neoplasms size, positive lymph node metastasis, and presence of cancer nodules are independent risk factors for liver metastasis in T 1 stage colorectal neoplasms. Age, primary site, ethnicity, chemotherapy, and surgical treatment are independent prognostic factors. The nomogram constructed based on these clinical characteristics has good discrimination and calibration abilities.