文献速递胃肠外科

胃癌术后条件生存概率预测:基于东西方大样本数据的研究

Prediction of Conditional Probability of Survival After Surgery for Gastric Cancer: A Study Based on Eastern and Western Large Data Sets.
PubMed 文献 · Surgery · 2018-04-21

该研究基于SEER数据库11551例胃癌患者构建并验证了预测条件生存概率的列线图,并在中国福建医科大学附属协和医院4001例患者中外部验证。结果显示术后5年总生存条件概率从切除时的41.6%升至3年后的80.4%,列线图在验证集中1、3、5、10年总生存预测准确度分别为0.77、0.81、0.82、0.82。

要点

  • 胃癌根治术后5年总生存条件概率随时间延长而升高,3年时达80.4%。
  • 基于SEER数据构建的列线图在东西方人群中验证准确度良好。
  • 列线图可动态评估患者术后不同时间点的生存概率。

核心结论

  • 基于东西方大样本数据开发并验证了首个胃癌条件生存概率列线图,可考虑生存时间长度进行动态预后评估。

研究设计与样本

  • 回顾性分析SEER数据库11551例胃癌患者,随机2/3为开发集、1/3为验证集,构建预测总生存和疾病特异性生存条件概率的列线图。
  • 外部验证采用中国福建医科大学附属协和医院4001例胃癌患者数据。

主要发现

  • 术后5年总生存条件概率在切除时点为41.6%,术后1、2、3年分别升至52.8%、68.2%、80.4%。
  • 术后5年疾病特异性生存条件概率从胃切除时的48.9%升至存活1、2、3年后的59.8%、74.7%、85.5%。
  • 性别、种族、年龄、肿瘤浸润深度、淋巴结转移、肿瘤大小、部位和分级与总生存和疾病特异性生存相关(P <.05)。
  • SEER验证集中总生存列线图在术后1、3、5、10年的准确度分别为0.77、0.81、0.82、0.82。
  • 福建协和医院验证集(n=4001)中总生存列线图在术后1、3、5、10年的准确度分别为0.76、0.79、0.77、0.77。

临床怎么用

  • 可利用该列线图动态评估胃癌患者术后不同时间点的条件生存概率,辅助随访和预后沟通。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: The dynamic prognosis of patients who have undergone curative surgery for gastric cancer has yet to be reported. Our objective was to devise an accurate tool for predicting the conditional probability of survival for these patients. METHODS: We analyzed 11,551 gastric cancer patients from the Surveillance, Epidemiology, and End Results database. Two-thirds of the patients were selected randomly for the development set and one-third for the validation set. Two nomograms were constructed to predict the conditional probability of overall survival and the conditional probability of disease-specific survival, using conditional survival methods. We then applied these nomograms to the 4,001 patients in the database from Fujian Medical University Union Hospital, Fuzhou, China, one of the most active Chinese institutes. RESULTS: The 5-year conditional probability of overall survival of the patients was 41.6% immediately after resection and increased to 52.8%, 68.2%, and 80.4% at 1, 2, and 3 years after gastrectomy. The 5-year conditional probability of disease-specific survival "increased" from 48.9% at the time of gastrectomy to 59.8%, 74.7%, and 85.5% for patients surviving 1, 2, and 3 years, respectively. Sex; race; age; depth of tumor invasion; lymph node metastasis; and tumor size, site, and grade were associated with overall survival and disease-specific survival (P <.05). Within the Surveillance, Epidemiology, and End Results validation set, the accuracy of the conditional probability of overall survival nomogram was 0.77, 0.81, 0.82, and 0.82 at 1, 3, 5, and 10 years after gastrectomy, respectively. Within the other validation set from the Fujian Medical University Union Hospital (n = 4,001), the accuracy of the conditional probability of overall survival nomogram was 0.76, 0.79, 0.77, and 0.77 at 1, 3, 5, and 10 years, respectively. The accuracy of the conditional probability of disease-specific survival model was also favorable. The calibration curve demonstrated good agreement between the predicted and observed survival rates. CONCLUSION: Based on the large Eastern and Western data sets, we developed and validated the first conditional nomogram for prediction of conditional probability of survival for patients with gastric cancer to allow consideration of the duration of survivorship.

来源:Surgery,查看原文。版权归原作者所有。

引用

  1. Zhong Q, Chen QY, Li P, et al. Prediction of Conditional Probability of Survival After Surgery for Gastric Cancer: A Study Based on Eastern and Western Large Data Sets. Surgery. 2018. PMID: 29685636. doi:10.1016/j.surg.2018.02.011

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。