Scholarlink

<p>Compared with existing clinical prognostic indicators, the prognostic efficacy of radiomics in predicting outcomes for early-stage extranodal nasal-type NK/T-cell lymphoma(ENKTCL) patients who received combined radiotherapy and non-anthracycline-based chemotherapy is still unclear. This study aims to construct and validate a hybrid radiomics nomogram based on multicenter data.&#xa0;A total of 196 patients from Xijing Hospital were enrolled and were randomly divided into the training set (<i>n</i> = 147) and internal validation set (<i>n</i> = 49) at a ratio of 3:1, and 57 patients from the other three hospitals were used as the independent external validation set. Important radiomic features selected from 1316 features extracted from contrast-enhanced CT-based tumors were used to calculate radscore as the radiomic signature. The optimal clinical index was selected from four common models: International Prognostic Index(IPI), Korean Prognostic Index(KPI), prognostic index for NK/T-cell lymphomas(PINK), and nomogram-revised risk index(NRI), which was combined with radscore to construct a hybrid nomogram for predicting overall survival(OS).&#xa0;Twelve radiomic features that were most useful to predict OS were ultimately selected to calculate radscore. Compared with PINK, IPI, and KPI, NRI demonstrated superior prognostic performance. Furthermore, the radscore model outperformed NRI. A hybrid nomogram, named RadNRI, was constructed by integrating NRI and radscore. The RadNRI model demonstrated the highest performance with C-index of 0.815, 0.823 and 0.765 in the training, internal and external validation sets, respectively.&#xa0;A hybrid radiomics nomogram was constructed and validated in multicenter data, which demonstrates significant potential in forecasting prognosis of early-stage ENKTCL patients.</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

Development and validation of a radiomics nomogram for predicting outcomes following combined radiotherapy and modern chemotherapy in early-stage extranodal nasal-type NK/T-cell lymphoma: a multicenter study

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月8日

查看全文
  • Jie Gong,
  • Yongchun Zhou,
  • Shan Gao,
  • Qiufang Liu,
  • JingYi Liu,
  • Yanchun Wei,
  • Yanqun Zhang,
  • Changhao Liu,
  • Bingxin Hou,
  • Lei Zhang,
  • Xiaolong Li,
  • Hongjuan Dong,
  • Juan Feng,
  • Jingjing Ren,
  • Ning Su,
  • Mai Zhang,
  • Lina Zhao,
  • Mei Shi,
  • Jian Zang

摘要

Compared with existing clinical prognostic indicators, the prognostic efficacy of radiomics in predicting outcomes for early-stage extranodal nasal-type NK/T-cell lymphoma(ENKTCL) patients who received combined radiotherapy and non-anthracycline-based chemotherapy is still unclear. This study aims to construct and validate a hybrid radiomics nomogram based on multicenter data. A total of 196 patients from Xijing Hospital were enrolled and were randomly divided into the training set (n = 147) and internal validation set (n = 49) at a ratio of 3:1, and 57 patients from the other three hospitals were used as the independent external validation set. Important radiomic features selected from 1316 features extracted from contrast-enhanced CT-based tumors were used to calculate radscore as the radiomic signature. The optimal clinical index was selected from four common models: International Prognostic Index(IPI), Korean Prognostic Index(KPI), prognostic index for NK/T-cell lymphomas(PINK), and nomogram-revised risk index(NRI), which was combined with radscore to construct a hybrid nomogram for predicting overall survival(OS). Twelve radiomic features that were most useful to predict OS were ultimately selected to calculate radscore. Compared with PINK, IPI, and KPI, NRI demonstrated superior prognostic performance. Furthermore, the radscore model outperformed NRI. A hybrid nomogram, named RadNRI, was constructed by integrating NRI and radscore. The RadNRI model demonstrated the highest performance with C-index of 0.815, 0.823 and 0.765 in the training, internal and external validation sets, respectively. A hybrid radiomics nomogram was constructed and validated in multicenter data, which demonstrates significant potential in forecasting prognosis of early-stage ENKTCL patients.