Scholarlink

Purpose <p>This study aims to compare, select, and investigate MRI-based multiregional radiomics model to predict pathologic complete response (pCR) in Locally advanced rectal cancer (LARC) patients after neoadjuvant chemoradiotherapy (nCRT).</p> Methods <p>This retrospective study included 245 patients who underwent rectal MRI examination before nCRT were recruited and split into training (hospital 1, <i>n</i> = 177) and external validation cohort (hospital 2, <i>n</i> = 68). Pretreatment T2WI and ADC images were used to manually delineate volumetric region of interest. Intratumoral, peritumoral-2&#xa0;mm, peritumoral-3&#xa0;mm, peritumoral-5&#xa0;mm, and peritumoral-mesorectal fat (MRF) radiomics features were extracted. Clinical model was built based on clinical and MRI features. Diagnosis performance was compared among models. 3-year recurrence-free survival (RFS) was evaluated by Kaplan-Meier curve.</p> Results <p>cN stage (odds ratio = 2.62, 95%CI,1.13–6.11) was a risk factor to construct clinical model. Peritumoral-3&#xa0;mm radiomics model showed slightly higher performance than peritumoral-2&#xa0;mm, peritumoral-5&#xa0;mm, and peritumoral-MRF radiomics model, but there were no significant differences (all <i>P</i> &gt; 0.05). Combining T2WI and ADC based intratumoral and peritumoral-3&#xa0;mm radiomics model showed good performance with AUC of 0.793 and 0.759 in training and external validation cohort, respectively. The clinical-radiomics combined model showed better performance than clinical model (AUC, 0.829 vs. 0.599; <i>P</i> &lt; 0.001) and better performance to the intratumoral and peritumoral-3&#xa0;mm radiomics model (AUC, 0.829 vs. 0.793; <i>P</i> = 0.04) in training cohort. And the clinical-radiomics combined model showed better performance than clinical model (AUC, 0.793 vs. 0.599; <i>P</i> &lt; 0.001) and better performance to the intratumoral and peritumoral-3&#xa0;mm radiomics model (AUC, 0.795 vs. 0.759; <i>P</i> = 0.21) in external validation cohort. Kaplan-Meier survival curves showed high-risk patients defined by nomogram based non-pCR had a worse 3-year RFS than that of low-risk patients.</p> Conclusion <p>The nomogram integrating T2WI and ADC based intratumoral and peritumoral radiomics signatures and cN stage could preoperative predict pCR. The nomogram based pCR was associated with 3-year RFS.</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

MRI-based multiregional radiomics for pretreatment prediction of pathologic complete response to neoadjuvant chemoradiation therapy in locally advanced rectal cancer: a bicenter study

  • 期刊论文

  • 发表日期：

    2025年11月4日

查看全文
  • Fulin Lu,
  • Xin Li,
  • Xiaoli Chen,
  • Yisha Liu,
  • Hao Liu,
  • Bin Luo,
  • Xin Zhang,
  • Hang Li

摘要

Purpose

This study aims to compare, select, and investigate MRI-based multiregional radiomics model to predict pathologic complete response (pCR) in Locally advanced rectal cancer (LARC) patients after neoadjuvant chemoradiotherapy (nCRT).

Methods

This retrospective study included 245 patients who underwent rectal MRI examination before nCRT were recruited and split into training (hospital 1, n = 177) and external validation cohort (hospital 2, n = 68). Pretreatment T2WI and ADC images were used to manually delineate volumetric region of interest. Intratumoral, peritumoral-2 mm, peritumoral-3 mm, peritumoral-5 mm, and peritumoral-mesorectal fat (MRF) radiomics features were extracted. Clinical model was built based on clinical and MRI features. Diagnosis performance was compared among models. 3-year recurrence-free survival (RFS) was evaluated by Kaplan-Meier curve.

Results

cN stage (odds ratio = 2.62, 95%CI,1.13–6.11) was a risk factor to construct clinical model. Peritumoral-3 mm radiomics model showed slightly higher performance than peritumoral-2 mm, peritumoral-5 mm, and peritumoral-MRF radiomics model, but there were no significant differences (all P > 0.05). Combining T2WI and ADC based intratumoral and peritumoral-3 mm radiomics model showed good performance with AUC of 0.793 and 0.759 in training and external validation cohort, respectively. The clinical-radiomics combined model showed better performance than clinical model (AUC, 0.829 vs. 0.599; P < 0.001) and better performance to the intratumoral and peritumoral-3 mm radiomics model (AUC, 0.829 vs. 0.793; P = 0.04) in training cohort. And the clinical-radiomics combined model showed better performance than clinical model (AUC, 0.793 vs. 0.599; P < 0.001) and better performance to the intratumoral and peritumoral-3 mm radiomics model (AUC, 0.795 vs. 0.759; P = 0.21) in external validation cohort. Kaplan-Meier survival curves showed high-risk patients defined by nomogram based non-pCR had a worse 3-year RFS than that of low-risk patients.

Conclusion

The nomogram integrating T2WI and ADC based intratumoral and peritumoral radiomics signatures and cN stage could preoperative predict pCR. The nomogram based pCR was associated with 3-year RFS.