Development and internal validation of a kinetic heterogeneity-based nomogram by dynamic contrast-enhanced magnetic resonance imaging to differentiate benign and malignant breast BI-RADS 4 lesions.

BACKGROUND: Category 4 breast cancer lesions with benign and malignant characteristics show substantial overlap in the Breast Imaging Reporting and Data System (BI-RADS). This study aimed to develop and validate a nomogram based on kinetic heterogeneity of the dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) to distinguish benign from malignant BI-RADS 4 breast lesions.

METHODS: A total of 271 patients diagnosed with BI-RADS 4 breast lesions by MRI and confirmed by histopathology at the Affiliated Tumor Hospital of Nantong University between January 2018 and June 2023 were retrospectively enrolled. Patients were divided into a benign group (n=84) and a malignant group (n=187) based on postoperative pathological results. The cohort was randomly split into a training set (n=192) and a validation set (n=79) in a 7:3 ratio. Clinical risk factors and MRI features were collected and re-evaluated. Kinetic heterogeneity parameters, including volume, washout component (%), plateau component (%), persistent component (%), predominant peak, and heterogeneity, were extracted using MATLAB and SPM12 software. Statistical analyses compared clinical, imaging, and kinetic parameters between groups. Univariate and multivariate logistic regressions identified independent predictors of malignancy. Three predictive models were constructed: one based on kinetic heterogeneity, one on clinicoradiological features, and a combined model integrating both. A nomogram was developed from the combined model. Model performance was evaluated using receiver operating characteristic (ROC) curves and validated in the independent set. Decision curve analysis (DCA) and clinical impact curve (CIC) were used to assess clinical utility.

RESULTS: Multivariate logistic regression identified peak, heterogeneity, apparent diffusion coefficient (ADC) values, time-intensity curve (TIC) type, and peritumoral edema as independent predictors of malignancy in BI-RADS 4 lesions. Among these, peak, heterogeneity, and ADC values demonstrated strong discriminatory power, with areas under the curve (AUC) of 0.793 [95% confidence interval (CI): 0.723-0.863], 0.816 (95% CI: 0.750-0.881), and 0.773 (95% CI: 0.704-0.842), respectively. The kinetic heterogeneity and clinicoradiological models achieved AUCs of 0.863 and 0.819, respectively. The combined nomogram demonstrated superior diagnostic performance (AUC 0.928 training, 0.906 validation), with high sensitivity and specificity. DCA and CIC confirmed its clinical utility.

CONCLUSIONS: The DCE-MRI kinetic heterogeneity-based nomogram is a promising tool to differentiate benign and malignant BI-RADS 4 breast lesions. Prospective external validation is warranted to confirm its potential for improving clinical decision-making and reducing unnecessary biopsies.