Scholarlink

Purpose <p>To evaluate the diagnostic performance of integrated ¹⁸F-FDG PET/MRI multiparametric imaging for predicting the ternary classification of human epidermal growth factor receptor-2 (HER2) expression in breast cancer, with particular emphasis on identifying the emerging HER2-low subtype.</p> Methods <p>This retrospective study included 111 pathologically confirmed breast cancer patients who underwent integrated ¹⁸F-FDG PET/MRI between August 2022 and February 2025. Patients were classified as HER2-zero (<i>n</i> = 27), HER2-low (<i>n</i> = 52), or HER2-overexpressing (HER2-oe, <i>n</i> = 32). Morphological, functional, synthetic MRI (SyMRI), and metabolic parameters were extracted. Univariate and multivariate logistic regression analyses were performed to identify independent predictors of HER2 expression status, and diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis. Internal validation was conducted using bootstrap resampling (1000 iterations), and calibration curves were generated to assess model performance.</p> Results <p>Proton density (PD) derived from SyMRI showed significant diagnostic value, with higher PD values observed in the HER2-low group compared with the HER2-zero group (<i>P</i> = 0.002; AUC = 0.728). Multiparametric models outperformed single parameters in identifying HER2-oe tumors. The combined “signal enhancement rate-proton density value-maximum standardized uptake value” (SER-PD-SUVmax) model achieved an AUC of 0.934 for differentiating HER2-zero from HER2-oe tumors, while the “largest tumor diameter-signal enhancement rate-maximum standardized uptake value” (LD-SER-SUVmax) model yielded an AUC of 0.849 for distinguishing HER2-low from HER2-oe tumors. Bootstrap validation demonstrated stable model performance with only slight reductions in AUC.</p> Conclusion <p>Integrated ¹⁸F-FDG PET/MRI multiparametric imaging may provide complementary information for assessing HER2 expression status in breast cancer and may be particularly helpful for the noninvasive identification of HER2-overexpressing tumors, while offering preliminary value for HER2-low characterization.</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

18F-FDG PET/MRI multiparameter imaging for predicting HER2-zero,-low and -overexpressing in breast cancer

  • 期刊论文

  • 发表日期：

    2026年6月1日

查看全文
  • Yumeng Bian,
  • Jue Wang,
  • Jinquan Chen,
  • Lijun Tang,
  • Tiannv Li,
  • Ziteng Ding

摘要

Purpose

To evaluate the diagnostic performance of integrated ¹⁸F-FDG PET/MRI multiparametric imaging for predicting the ternary classification of human epidermal growth factor receptor-2 (HER2) expression in breast cancer, with particular emphasis on identifying the emerging HER2-low subtype.

Methods

This retrospective study included 111 pathologically confirmed breast cancer patients who underwent integrated ¹⁸F-FDG PET/MRI between August 2022 and February 2025. Patients were classified as HER2-zero (n = 27), HER2-low (n = 52), or HER2-overexpressing (HER2-oe, n = 32). Morphological, functional, synthetic MRI (SyMRI), and metabolic parameters were extracted. Univariate and multivariate logistic regression analyses were performed to identify independent predictors of HER2 expression status, and diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis. Internal validation was conducted using bootstrap resampling (1000 iterations), and calibration curves were generated to assess model performance.

Results

Proton density (PD) derived from SyMRI showed significant diagnostic value, with higher PD values observed in the HER2-low group compared with the HER2-zero group (P = 0.002; AUC = 0.728). Multiparametric models outperformed single parameters in identifying HER2-oe tumors. The combined “signal enhancement rate-proton density value-maximum standardized uptake value” (SER-PD-SUVmax) model achieved an AUC of 0.934 for differentiating HER2-zero from HER2-oe tumors, while the “largest tumor diameter-signal enhancement rate-maximum standardized uptake value” (LD-SER-SUVmax) model yielded an AUC of 0.849 for distinguishing HER2-low from HER2-oe tumors. Bootstrap validation demonstrated stable model performance with only slight reductions in AUC.

Conclusion

Integrated ¹⁸F-FDG PET/MRI multiparametric imaging may provide complementary information for assessing HER2 expression status in breast cancer and may be particularly helpful for the noninvasive identification of HER2-overexpressing tumors, while offering preliminary value for HER2-low characterization.