Intratumoral and peritumoral ultrasound radiomics analysis for predicting HER2-low expression in HER2-negative breast cancer patients: a retrospective analysis of dual-central study

Abstract

Objective: This study aims to explore whether intratumoral and peritumoral ultrasound radiomics of ultrasound images can predict the low expression status of human epidermal growth factor receptor 2 (HER2) in HER2-negative breast cancer patients.

Methods: HER2-negative breast cancer patients were recruited retrospectively and randomly divided into a training cohort (n = 303) and a test cohort (n = 130) at a ratio of 7:3. The region of interest within the breast ultrasound image was designated as the intratumoral region, and expansions of 3 mm, 5 mm, and 8 mm from this region were considered as the peritumoral regions for the extraction of ultrasound radiomic features. Feature extraction and selection were performed, and radiomics scores (Rad-score) were obtained in four ultrasound radiomics scenarios: intratumoral only, intratumoral + peritumoral 3 mm, intratumoral + peritumoral 5 mm, and intratumoral + peritumoral 8 mm. An optimal combined nomogram radiomic model incorporating clinical features was established and validated. Subsequently, the diagnostic performance of the radiomic models was evaluated.

Results: The results indicated that the intratumoral + peritumoral (5 mm) ultrasound radiomics exhibited the excellent diagnostic performance in evaluated the HER2 low expression. The nomogram combining intratumoral + peritumoral (5 mm) and clinical features showed superior diagnostic performance, achieving an area under the curve (AUC) of 0.911 and 0.869 in the training and test cohorts, respectively.

Conclusion: The combination of intratumoral + peritumoral (5 mm) ultrasound radiomics and clinical features possesses the capability to accurately predict the low-expression status of HER2 in HER2-negative breast cancer patients.

Keywords: Artificial intelligence; Breast cancer; HER2 status; Radiomics scores; Ultrasound.

Fig. 1

The flowchart of breast cancer patients with HER2 negative enrolled in this study. HER2 indicates human epidermal growth factor receptor 2

Fig. 2

Illustration of four ROIs in BC ultrasound imaging. A intratumoral. B intratumoral + 3 mm peritumoral. C intratumoral + 5 mm peritumoral. D intratumoral + 8 mm peritumoral

Fig. 3

The flowchart of combined model constructed in this study

Fig. 4

Distribution of selected features associated with HER-2 expression. A Intratumoral segmentation. B Intratumoral + 5 mm peritumoral segmentation using LASSO regression

Fig. 5

Radiomics feature selection by LASSO: A, E Intratumoral. B, F Intratumoral + 3 mm peritumoral. C, G Intratumoral + 5 mm peritumoral. D, H Intratumoral + 8 mm peritumoral

Fig. 6

A Nomogram constructed based on the combined model. B–D depict the performance characteristics (ROC, calibration, and DCA curves) in the training cohorts. E–G present the same classifiers’ efficacy as applied to test cohorts