Related Experiment Video
Updated: Jul 4, 2026
Optimization of a Multiplex RNA-based Expression Assay Using Breast Cancer Archival Material
Published on: August 1, 2018
Multiparametric MRI-based nomogram integrating clinicopathological factors for predicting HER2 expression status in
Yi Chen1,2, Xiaofeng Chen1,2, Bowen Yue1,2
1Department of Radiology, Meizhou People's Hospital, Meizhou, China.
Background:
Human epidermal growth factor receptor 2 (HER2) expression in breast cancer (BC) determines the options for targeted therapy. Multiparametric MRI (mpMRI) has the potential for noninvasive HER2 status prediction but remains underexplored.
Objectives:
To develop and validate an mpMRI-based nomogram incorporating clinicopathological factors for predicting HER2 status in BC patients.
Methods:
In this retrospective analysis, 313 BC patients were classified as HER2-overexpression, HER2-low, or HER2-zero on the basis of immunohistochemistry and fluorescence in situ hybridization. The patients were divided into training (n=232) and validation (n=81) datasets. Clinicopathological factors and mpMRI parameters were analyzed. Logistic regression identified independent predictors that were used to construct and validate the nomogram. Discrimination was evaluated by the area under the receiver operating characteristic curve (AUC).
Results:
CA125, Ki-67, the minimum apparent diffusion coefficient (ADC-min), and early-phase maximum enhancement (ME) differed significantly among the HER2 subgroups. The nomogram integrating these factors achieved AUCs of 0.762 (95% CI: 0.686-0.838) and 0.738 (95% CI: 0.594-0.882) in differentiating HER2-over/HER2-low from HER2-zero in the training and validation datasets, respectively. Differentiation between the HER2-over and HER2-low subtypes exhibited AUCs of 0.719 and 0.772, respectively.
Conclusions:
Our nomogram, which combines mpMRI and clinicopathological variables, effectively predicts HER2 expression in BC patients, providing a promising noninvasive clinical tool to guide targeted therapy selection.
