Prediction of death probability in adenocarcinoma of the transverse colon: competing-risk nomograms based on 21,469 patients

Hongbo Su,Shanshan Wang,Shuping Xie,Liying Huang,Yunlong Pan,Jun Lyu

DOI: https://doi.org/10.1007/s00432-023-04913-w

2023-01-01

Journal of Cancer Research and Clinical Oncology

Abstract:Purpose Tumors in parts of the colon other than the transverse colon have been well studied, but little is known about adenocarcinoma of the transverse colon (ATC).The aim of this study was to construct nomograms using competing-risk model for accurately predicting the probability of cancer-specific and non-cancer-specific death in patients with ATC. Methods Data on eligible patients recorded during 2000–2019 in the Surveillance, Epidemiology, and End Results database were extracted and screened. Factors that influencing prognosis were screened for death from ATC (DATC) and death from other causes (DOC) using competing-risk analysis, including univariate and multivariate analyses based on Gray’s test and the Fine–Gray model, respectively. Independent prognostic factors were identified and nomograms were constructed. For comparison, we also constructed a Cox model and an AJCC stage-only competing-risk model (AJCC model) for patients with DATC. Performance evaluations of the nomograms and comparison between the models were performed using calibration plots, Harrell’s concordance index (C-index), receiver operating characteristic (ROC) curves, and the areas under the ROC curve (AUCs). The nomograms and models were validated using a validation cohort. The net reclassification index, integrated discrimination improvement, decision curves, and risk stratification were not assessed because no accepted methods were suited for competing-risk model. Results This study included 21,469 patients with ATC, and 17 and 9 independent influencing factors were identified for the construction of DATC nomograms (DATCN) and DOC nomograms (DOCN), respectively. In both the training and validation cohorts, the calibration curves indicated good agreement between the nomogram-based predictions and the actual observations in the two nomograms, respectively. The C-index of the DATCN was higher than 80% (80.3–83.3%) at 1, 3 and 5 years in both the training and validation cohorts, significantly outperforming the AJCC (76.7–78%) and Cox (75.4–79.5%) model. The C-index of the DOCN was also higher than 69% (69.0–73.6%). In terms of ROC curves at each time point, those of the DATCN were very close to the upper-left corner of the coordinate axis in both the training and validation cohorts, and their AUCs were larger than 84% (84.2–85.4%).The AUCs of the AJCC (78.4–81.1%) and Cox (79.4–81.5%) models were significantly lower ( p < 0.05), and the curves were closer to the diagonal. The ROC curves of the DOCN was similar to those of the DATCN, and the AUCs were 68.5–74%. The DATCN and DOCN therefore had good consistency, accuracy, and stability, respectively. Conclusion This study was the first to construct competing-risk nomograms for ATC. These nomograms have proved useful for accurately assessing patient prognoses and allowing more-individualized follow-up strategies, thereby reducing the mortality.