中国实用外科杂志 ›› 2022, Vol. 42 ›› Issue (02): 206-213.DOI: 10.19538/j.cjps.issn1005-2208.2022.02.15

• 论蓍 • 上一篇    下一篇

初始可手术切除Ⅳ期结直肠癌病人生存预测模型建立与验证

张    磊,时飞宇,胡晨浩,张    喆,阎    俊,王光辉,佘军军   

  1. 西安交通大学第一附属医院 普通外科,陕西西安 710061
  • 出版日期:2022-02-01 发布日期:2022-01-27

  • Online:2022-02-01 Published:2022-01-27

摘要: 目的    探究影响初始可手术切除Ⅳ期结直肠癌病人预后的独立危险因素,构建并验证个体化生存预测模型。方法    回顾性分析2014-01-01—2019-12-31西安交通大学第一附属医院行手术治疗的445例Ⅳ期结直肠癌病人的临床病例资料。通过R软件采用随机抽样的方法,选择356例(4/5)病人作为建模组,剩余89例(1/5)病人作为验证组。在建模组中采用Cox比例风险模型分析初始可手术切除Ⅳ期结直肠癌病人预后的独立危险因素。通过R软件构建初始可手术切除Ⅳ期结直肠癌病人总生存率(OS)、无进展生存率(PFS)列线图。在建模组和验证组中绘制受试者特征(ROC)曲线和校准图评价列线图预测效能。结果    年龄、切除效果、CRS评分以及术后辅助治疗为初始可手术切除Ⅳ期结直肠癌病人OS和PFS的独立危险因素。基于筛选出的危险因素成功构建预测初始可手术切除Ⅳ期结直肠癌病人的OS、PFS列线图模型。建模组与验证组5年OS和PFS的ROC曲线下面积分别为0.736、0.784、0.794、0.756。在建模组与验证组的校准图评价中校正曲线均展现了较好的拟合度。结论 高龄、姑息性切除、CRS评分3~5分、术后未行辅助治疗是初始可手术切除Ⅳ期结直肠癌病人预后的独立危险因素。据此建立的预测模型能较为准确预测初始可手术切除Ⅳ期结直肠癌病人的预后,为临床决策提供参考和帮助。

关键词: 结直肠肿瘤, 远处转移, 预后, 列线图

Abstract: Establishment and validation of a survival prediction model for patients with initially resectable stage Ⅳ colorectal cancer        ZHANG Lei,SHI Fei-yu,HU Chen-hao,et al. Department of General Surgery, the First Affiliated Hospital of Xi'an Jiaotong University, Xi'an 710061,China
Corresponding author: SHE Jun-jun, E-mail: junjunshe1975@sina.com
Abstract    Objective    To explore the independent risk factors affecting the prognosis of patients with initially surgically resectable stage Ⅳ colorectal cancer, and to construct and validate an individualized survival prediction model. Methods    The clinical data of 445 patients with stage Ⅳ colorectal cancer who underwent surgery in the First Affiliated Hospital of Xi’an Jiaotong University from January 1, 2014 to December 31, 2019 were retrospectively analyzed. Using R software to implement random sampling, 356 patients (4/5) were selected as the training group, and the remaining 89 patients (1/5) were selected as the verification group. In the training group, Cox proportional risk model was used to analyze the independent risk factors for prognosis of patients with initially surgically resectable IV colorectal cancer. Nomograms for predicting the overall survival (OS) and progression-free survival (PFS) of patients with initially surgically resectable stage Ⅳ colorectal cancer were established by the R language software. The accuracy of these models were evaluated by calculating the area under the receiver operating characteristic curve (ROC) of the training group and the verification group. Calibration curve was also used to evaluate nomograms' prediction efficiency. Results  Age, resection effect , CRS score and postoperative adjuvant therapy were independent risk factors for prognosis in patients with initially surgically resectable stage Ⅳ colorectal cancer. Based on the screened risk factors, nomogram models were successfully constructed to predict the OS and PFS of patients with initial operable stage Ⅳ colorectal cancer. The areas under the ROC curve of the 5-year OS and 5-year PFS in the training group and the verification group were 0.736, 0.784, 0.794, and 0.756, respectively. The calibration curves of both the training group and the verification group showed high consistence. Conclusion    Advanced age, palliative resection, CRS score of 3-5, and no postoperative adjuvant therapy were independent risk factors for the prognosis of patients with initially surgically resectable stage Ⅳ colorectal cancer. These survival prediction models may accurately predict the prognosis of patients with initially operable stage Ⅳ colorectal cancer, and provide reference and help for clinical decision-making.

Key words: colorectal neoplasms, distant metastasis, prognosis, nomogram