PREOPERATIVE RADIOLOGIC PREDICTORS OF SURGICAL MARGINS IN BREAST CANCER RESECTION

Abstract

The research determines the predictive capability of preoperative radiologic variables in the determination of surgical margin in breast cancer resection.  An evaluation on a prospective group of patients was conducted involving mammography, ultrasound, and contrast-enhanced MRI to evaluate important radiologic features, such as size of the lesion, tumor-to-margin distance, multifocality, enhancement kinetics, and radiologic suspicion of ductal extension.  Statistical analysis revealed that volumetric measurements and pattern of contrast enhancement in MRI were the most significant variables to predict the presence or near-presence of positive or close margins (AUC 0.87). The next best was tumor-to-margin distance on mammography and ultrasound (AUC 0.79).  Multimodal imaging significantly improved overall sensitivity and specificity as compared to single-modality assessment (p < 0.05).  Radiologic appearance of multifocal disease, ductal extension, or lesions measuring more than 3 cm at risk increased the propensity of the involvement of the margins hence the rationale of a thorough preoperative imaging in planning surgery.  The findings indicate that radiologic prediction can be used to reduce re-excisions dramatically, make surgeries more precise, and assist individuals in making the choice whether to undergo breast-conserving surgery or a more massive excision.  This analysis supports the use of advanced imaging measurements and predictive modeling in clinical activity to enhance oncological and aesthetic outcomes of breast cancer patients.  Future studies are interested in studying standardized thresholds and radiomics and artificial intelligence integration to improve margin prediction in different clinical settings.