Prognostic model for oversurvival and tumor-specific survival prediction in patients with advanced extrahepatic cholangiocarcinoma: a population-based analysis

Abstract

Background: The prognosis of patients with extrahepatic cholangiocarcinoma (ECCA) must be determined with precision. However, the usual TNM staging system has the drawback of ignoring age, adjuvant therapy, and gender and lacks the ability to more correctly predict patient prognosis. Therefore, we determine the risk factors of survival for patients with advanced ECCA patients and developed brand-new nomograms to forecast patients with advanced ECCA's overall survival (OS) and cancer-specific survival (CSS).

Method: From the Epidemiology and End Results (SEER) database, patients with advanced ECCA were chosen and randomly assigned in a ratio of 6:4 to the training and validation subgroups. The cumulative incidence function (CIF) difference between groups was confirmed by applying Gray's and Fine test and competing risk analyses. Next, the cancer-specific survival (CSS) and overall survival (OS) nomograms for advanced ECCA were developed and validated.

Results: In accordance with the selection criteria, 403 patients with advanced ECCA were acquired from the SEER database and then split at random into two groups: a training group (n = 241) and a validation group (n = 162). The 1-, 2-, and 3-year cancer-specific mortality rates were 58.7, 74.2, and 78.0%, respectively, while the matching mortality rates for the competition were 10.0, 13.8, and 15.0%. Nomograms were generated for estimating OS and CSS, and they were assessed using the ROC curve and the C-index. The calibration curves showed that there was a fair amount of agreement between the expected and actual probabilities of OS and CSS. Additionally, greater areas under the ROC curve were seen in the newly developed nomograms for OS and CSS when compared to the 7th AJCC staging system. The advanced ECCA patients were divided into groupings with an elevated risk and those with a low risk and the Kaplan-Meier method was used for the survival analysis, which showed that survival time was shorter in the high-risk group than in the low-risk group.

Conclusion: The proposed nomograms have good predictive ability. The nomograms may can help doctors determine the prognosis of patients with advanced ECCA as well as provide more precise treatment plans for them.

Keywords: Advanced ECCA; Cancer-specific survival; Nomogram; Oversurvival.

Fig. 1

Cumulative cancer-specific and competing mortality stratified by patient characteristics: A stage; B M stage; C N stage; D surgical treatment

Fig. 2

Overall survival rates stratified by patient characteristics: A age; B T stage; C N stage; D M stage; E TNM stage; F grade; G surgical treatment; H chemtherapy treatment; F radiation treatment

Fig. 3

Multivariate analyses of survival in patients with advanced ECCA: A Overall survival; B Cancer-specific survival

Fig. 4

Nomograms predicting 1-, 2-, and 3-year OS (A) and CSS (B) of patients with advanced ECCA

Fig. 5

ROC curves for the nomogram for 1-, 2-, and 3-year OS prediction. A the training cohort; B the validation cohort

Fig. 6

Decision curve analysis of 1-year, 2-year, and 3-year OS. A, B, C DCA curves of 1-year, 2-year, and 3-year OS in the training cohort; D, E, F DCA curves of 1-year, 2-year, and 3-year OS in the validation cohort; modle A,the prediction nomogram; modle B, the 7th AJCC stage

Fig. 7

Calibration plots of 1-year, 2-year, and 3-year OS for advanced ECCA patients. A, B, C Calibration plots of 1-year, 2-year, and 3-year OS in the training cohort; D, E, F Calibration plots of 1-year, 2-year, and 3-year OS in the validiation cohort

Fig. 8

Kaplan–Meier OS and CSS curves of patients with advanced ECCA based on the new risk stratification system. A, C Kaplan–Meier curves of the raining cohorts; B, D Kaplan–Meier curves of the validation cohorts; A, B Kaplan–Meier OS curves; C, D Kaplan–Meier CSS curves