Ultrasound-Based IHC4-Associated Radiomic Model for Predicting Late Recurrence in ER-Positive Breast Cancer: A...

Area of Science:

• Radiology and Oncology
• Biomarker Discovery
• Breast Cancer Research

Background:

• Estrogen receptor-positive (ER-positive) breast cancer recurrence is a significant clinical concern.
• Accurate prediction of late recurrence is crucial for optimizing treatment strategies, including extended endocrine therapy.
• Current prognostic markers may require improvement for precise risk stratification.

Purpose of the Study:

• To develop and validate an ultrasound-based radiomic model associated with the immunohistochemical four-marker (IHC4) score.
• To assess the model's ability to predict late distant recurrence (DR) in ER-positive breast cancer.
• To evaluate the model's potential for risk stratification and guiding decisions on extended endocrine therapy.

Main Methods:

• A retrospective multicenter study involving 523 patients with ER-positive breast cancer.
• Development of a support vector machine (SVM)-based radiomic model using IHC4-associated features.
• Validation of the model's performance using area under the receiver operating characteristic curve (AUC) in training and external cohorts.

Main Results:

• A seven-feature IHC4-associated radiomic model was constructed.
• Consistent model performance was observed across all cohorts with AUCs ranging from 0.79 to 0.84.
• The radiomic score effectively stratified late DR risk and was associated with progression-free survival (PFS) in multivariable analysis (HR=4.246, P=.001).

Conclusions:

• The developed IHC4-associated radiomic model demonstrates potential as a noninvasive biomarker.
• This model can aid in predicting prognosis for patients with ER-positive breast cancer.
• The findings support the model's utility in risk stratification and potentially informing treatment decisions.