Radiomic Signatures Derived from Hybrid Contrast-Enhanced Ultrasound Images (CEUS) for the Assessment of Histological Characteristics of Breast Cancer: A Pilot Study

SIMPLE SUMMARY: The hybrid images obtained at the CEUS present an overlap of the tumor in grayscale over the enhancement pattern, which allows an appropriate evaluation of the entire lesion without including in the analysis the peritumoral area. We hypothesized that the spatial heterogeneity of tissue enhancement differs depending on histological characteristics. Therefore, the aim of this study was to evaluate the diagnostic performance of radiomic features extracted from standardized hybrid CEUS data for the assessment of hormone receptor status, HER2 status, tumor grade and Ki-67 in patients with primary breast cancer. The clinical benefit of this study would be to use CEUS and bioinformatics tools for improved biopsy planning or even guiding treatment decisions. ABSTRACT: The purpose of this study was to evaluate the diagnostic performance of radiomic features extracted from standardized hybrid contrast-enhanced ultrasound (CEUS) data for the assessment of hormone receptor status, human epidermal growth factor receptor 2 (HER2) status, tumor grade and Ki-67 in patients with primary breast cancer. Methods: This prospective study included 72 patients with biopsy-proven breast cancer who underwent CEUS examinations between October 2020 and September 2021. Results: A radiomic analysis found the WavEnHH_s_4 parameter as an independent predictor associated with the HER2+ status with 76.92% sensitivity, and 64.41% specificity and a prediction model that could differentiate between the HER2 entities with 76.92% sensitivity and 84.75% specificity. The RWavEnLH_s-4 parameter was an independent predictor for estrogen receptor (ER) status with 55.93% sensitivity and 84.62% specificity, while a prediction model (RPerc01, RPerc10 and RWavEnLH_s_4) could differentiate between the progesterone receptor (PR) status with 44.74% sensitivity and 88.24% specificity. No texture parameter showed statistically significant results at the univariate analysis when comparing the Nottingham grade and the Ki-67 status. Conclusion: Our preliminary data indicate a potential that hybrid CEUS radiomic features allow the discrimination between breast cancers of different receptor and HER2 statuses with high specificity. Hybrid CEUS radiomic features might have the potential to provide a noninvasive, easily accessible and contrast-agent-safe method to assess tumor biology before and during treatment.