Scholarlink

<p>The optimal treatment for esophageal cancer (EC) invading the muscularis mucosa (T1a-M3) or submucosa (T1b) remains debated. This study analyzed patients with stage T1a-M3/T1b EC from the SEER database (cycle 2004–2017). Cancer-specific survival (CSS) and overall survival (OS) were compared across five groups. Primary analysis used stable inverse probability treatment weighting (sIPTW), while sensitivity analysis applied propensity score matching (PSM) and external datasets. After sIPTW adjustment, endoscopic therapy showed significantly better CSS and OS than chemoradiotherapy (CRT) (both <i>P</i> &lt; 0.0001). Compared to radical resection, endoscopic treatment had superior CSS (<i>P</i> = 0.038) without observed significant difference in OS (<i>P</i> = 0.50). Four variable selection methods built the optimal predictive model, validated internally and externally. Internal AUC values at 1, 3, and 5 years were 0.730, 0.743, and 0.769, respectively; external AUC values were 0.765, 0.731, and 0.785.</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

Therapeutic outcomes of different treatment approaches for T1a-M3/T1b esophageal cancer and development of a prognostic prediction model

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年11月17日

查看全文
  • Aijing Zhu,
  • Shu Huang,
  • Shuaijing Huang,
  • Yang Zhou,
  • Yan Liang,
  • Chunxiao Yue,
  • Xiaoyu Bai,
  • Jiawei Li,
  • Yadong Feng

摘要

The optimal treatment for esophageal cancer (EC) invading the muscularis mucosa (T1a-M3) or submucosa (T1b) remains debated. This study analyzed patients with stage T1a-M3/T1b EC from the SEER database (cycle 2004–2017). Cancer-specific survival (CSS) and overall survival (OS) were compared across five groups. Primary analysis used stable inverse probability treatment weighting (sIPTW), while sensitivity analysis applied propensity score matching (PSM) and external datasets. After sIPTW adjustment, endoscopic therapy showed significantly better CSS and OS than chemoradiotherapy (CRT) (both P < 0.0001). Compared to radical resection, endoscopic treatment had superior CSS (P = 0.038) without observed significant difference in OS (P = 0.50). Four variable selection methods built the optimal predictive model, validated internally and externally. Internal AUC values at 1, 3, and 5 years were 0.730, 0.743, and 0.769, respectively; external AUC values were 0.765, 0.731, and 0.785.