Back to Search Start Over

A novel prognosis prediction model after completion gastrectomy for remnant gastric cancer: Development and validation using international multicenter databases

Authors :
Zhi Fang Zheng
Bin Bin Xu
Chao-Hui Zheng
Jun-Feng Zhou
Qi Yue Chen
Jian Wei Xie
Mark J. Truty
Jian Xian Lin
Jun Lu
Chang Ming Huang
Ping Li
Jia Bin Wang
Qing Liang He
Source :
Surgery. 166:314-321
Publication Year :
2019
Publisher :
Elsevier BV, 2019.

Abstract

Background Examined lymph node counts of remnant gastric cancer patients are often insufficient, and the prognostic ability of tumor-node-metastasis staging is therefore limited. This study aimed to create a simple and universally applicable prediction model for RGC patients after completion of gastrectomy. Methods A 5-year overall survival prediction model for remnant gastric cancer patients was developed using a test dataset of 148 consecutive patients. Model coefficients were obtained based on the Cox analysis of clinicopathological factors. Prognostic performance was assessed with the concordance index (C-index) and decision curve analysis. For internal validation, the bootstrap method and calibration assessment were used. The model was validated using 2 external cohorts from China (First Affiliated Hospital of Fujian Medical University, n = 46) and the United States (Mayo Clinic, n = 20). Results Depth of tumor invasion, number of metastatic lymph nodes, distant metastasis, and operative time were independent prognostic factors. Our model’s C-index (0.761) showed better discriminatory power than that of the eighth tumor-node-metastasis staging system (0.714, P = .001). The model calibration was accurate at predicting 5-year survival. Decision curve analysis showed that the model had a greater benefit, and the results were also confirmed by bootstrap internal validation. In external validation, the C-index and decision curve analysis showed good prognostic performances in patient datasets from 2 participating institutions. Moreover, we verified the reliability of the model in an analysis of patients with different examined lymph node counts (>15 or ≤15). Conclusion Utilizing clinically practical information, we developed a universally applicable prediction model for accurately determining the 5-year overall survival of remnant gastric cancer patients after completion of gastrectomy. Our predictive model outperformed tumor-node-metastasis staging in diverse international datasets regardless of examined lymph node counts.

Details

ISSN :
00396060
Volume :
166
Database :
OpenAIRE
Journal :
Surgery
Accession number :
edsair.doi.dedup.....8f8742835f875477acd4cf9ac6d814ef
Full Text :
https://doi.org/10.1016/j.surg.2019.05.004