Scholarlink

Background <p>The present study aimed to develop a nomogram model for predicting overall survival (OS) in esophageal squamous cell carcinoma (ESCC) patients.</p> Methods <p>A total of 205 patients with ESCC were enrolled and randomly divided into a training cohort (<i>n</i> = 153) and a test cohort (<i>n</i> = 52) at a ratio of 7:3. Multivariate Cox regression was used to construct the radiomics model based on CT data. The mutation signature was constructed based on whole genome sequencing data and found to be significantly associated with the prognosis of patients with ESCC. A nomogram model combining the Rad-score and mutation signature was constructed. An integrated nomogram model combining the Rad-score, mutation signature, and clinical factors was constructed.</p> Results <p>A total of 8 CT features were selected for multivariate Cox regression analysis to determine whether the Rad-score was significantly correlated with OS. The area under the curve (AUC) of the radiomics model was 0.834 (95% CI, 0.767–0.900) for the training cohort and 0.733 (95% CI, 0.574–0.892) for the test cohort. The Rad-score, S3, and S6 were used to construct an integrated RM nomogram. The predictive performance of the RM nomogram model was better than that of the radiomics model, with an AUC of 0. 830 (95% CI, 0.761–0.899) in the training cohort and 0.793 (95% CI, 0.653–0.934) in the test cohort. The Rad-score, TNM stage, lymph node metastasis status, S3, and S6 were used to construct an integrated RMC nomogram. The predictive performance of the RMC nomogram model was better than that of the radiomics model and RM nomogram model, with an AUC of 0. 862 (95% CI, 0.795–0.928) in the training cohort and 0. 837 (95% CI, 0.705–0.969) in the test cohort.</p> Conclusion <p>An integrated nomogram model combining the Rad-score, mutation signature, and clinical factors can better predict the prognosis of patients with ESCC.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Survival outcome prediction of esophageal squamous cell carcinoma patients based on radiomics and mutation signature

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年1月31日

查看全文
  • Ting Yan,
  • Zhenpeng Yan,
  • Guohui Chen,
  • Songrui Xu,
  • Chenxuan Wu,
  • Qichao Zhou,
  • Guolan Wang,
  • Ying Li,
  • Mengjiu Jia,
  • Xiaofei Zhuang,
  • Jie Yang,
  • Lili Liu,
  • Lu Wang,
  • Qinglu Wu,
  • Bin Wang,
  • Tianyi Yan

摘要

Background

The present study aimed to develop a nomogram model for predicting overall survival (OS) in esophageal squamous cell carcinoma (ESCC) patients.

Methods

A total of 205 patients with ESCC were enrolled and randomly divided into a training cohort (n = 153) and a test cohort (n = 52) at a ratio of 7:3. Multivariate Cox regression was used to construct the radiomics model based on CT data. The mutation signature was constructed based on whole genome sequencing data and found to be significantly associated with the prognosis of patients with ESCC. A nomogram model combining the Rad-score and mutation signature was constructed. An integrated nomogram model combining the Rad-score, mutation signature, and clinical factors was constructed.

Results

A total of 8 CT features were selected for multivariate Cox regression analysis to determine whether the Rad-score was significantly correlated with OS. The area under the curve (AUC) of the radiomics model was 0.834 (95% CI, 0.767–0.900) for the training cohort and 0.733 (95% CI, 0.574–0.892) for the test cohort. The Rad-score, S3, and S6 were used to construct an integrated RM nomogram. The predictive performance of the RM nomogram model was better than that of the radiomics model, with an AUC of 0. 830 (95% CI, 0.761–0.899) in the training cohort and 0.793 (95% CI, 0.653–0.934) in the test cohort. The Rad-score, TNM stage, lymph node metastasis status, S3, and S6 were used to construct an integrated RMC nomogram. The predictive performance of the RMC nomogram model was better than that of the radiomics model and RM nomogram model, with an AUC of 0. 862 (95% CI, 0.795–0.928) in the training cohort and 0. 837 (95% CI, 0.705–0.969) in the test cohort.

Conclusion

An integrated nomogram model combining the Rad-score, mutation signature, and clinical factors can better predict the prognosis of patients with ESCC.