Prediction model of neoadjuvant chemotherapy effect on HER2 positive breast cancer based on MRI parameters and multimodal ultrasound
10.3760/cma.j.cn115807-20240429-00143
- VernacularTitle:基于MRI参数、多模态超声构建HER2阳性乳腺癌新辅助化疗效果的预测模型
- Author:
Wenqiang CUI
1
;
Meili YU
;
Tingting ZHANG
;
Yongguang BAN
;
Shumao ZHANG
Author Information
1. 临沂市人民医院超声医学科，临沂 276000
- Publication Type:Journal Article
- Keywords:
Magnetic resonance imaging;
Multimode ultrasound;
Human epidermal growth factor receptor 2;
Breast cancer;
Neoadjuvant chemotherapy
- From:
Chinese Journal of Endocrine Surgery
2025;19(5):688-692
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To establish a predictive model of neoadjuvant chemotherapy for human epidermal growth factor receptor 2 (HER2) positive breast cancer based on magnetic resonance imaging (MRI) parameters and multimodal ultrasound.Methods:The medical records of 279 patients with HER2 positive breast cancer admitted to Linyi People’s Hospital, Tai’an Central Hospital Affiliated to Qingdao University and Shandong First Medical University affiliated Provincial Hospital from Mar. 2021 to Nov. 2023 were retrospectively analyzed, and randomly divided into a training set ( n=223) and a validation set ( n=56) according to the 8∶2 law. All patients received neoadjuvant chemotherapy combined with targeted therapy and were divided into non-pathologic complete response (NpCR) and pathologic complete response (pCR) groups according to chemotherapy effect. By comparing MRI parameters and multimodal ultrasound parameters of the two groups, and the pCR risk model after neoadjuvant chemotherapy for HER2 positive breast cancer patients was constructed and verified. Results:After 3 cycles of chemotherapy, the incidence of pCR in 223 patients in the training set was 42.15%. Among 56 patients in the validation set, the incidence of pCR was 42.86%. The apparent diffusion coefficient (ADC) of pCR group was higher than NpCR group ( P<0.05). The peak systolic flow velocity (PSV), resistance index (RI), maximum radial change rate, area under the curve (AUC) and peak intensity (PI) in pCR group were higher than NpCR group ( P<0.05), and the ultrasound elastic score in pCR group was lower than NpCR group ( P<0.05). ΔADC ( OR=4.141, 95% CI: 1.820-9.421), maximum diameter change rate ( OR=5.212, 95% CI: 2.291-11.857), PI ( OR=4.802, 95% CI: 2.111-10.923) and ultrasonic elasticity score ( OR=6.629, 95% CI: 1.595-8.256) were the influencing factors of pCR after neoadjuvant chemotherapy for HER2 positive breast cancer ( P<0.05). The sensitivity and specificity of pCR predicted by the training set model after neoadjuvant chemotherapy for HER2 positive breast cancer were 90.43% (95% CI: 82.15%-95.26%), 91.47% (95% CI: 84.91%-95.45%), and the AUC was 0.904 (95% CI: 0.834-0.968). Validation set model predicted the pCR sensitivity after neoadjuvant chemotherapy for HER2 positive breast cancer was 87.50% (95% CI: 66.54%-96.71), specificity was 90.63% (95% CI: 73.83%-97.55%), and AUC was 0.897 (95% CI: 0.821-0.954) . Conclusion:The histogram model based on ΔADC, maximum radial rate of lesion change, PI and ultrasonic elasticity score can be used to evaluate the risk of pCR after neoadjuvant chemotherapy for HER2 positive breast cancer.
