Radiomics-clinical nomogram for preoperative tumor-node-metastasis staging prediction in breast cancer patients using dynamic enhanced magnetic resonance imaging

Abstract

Background: Breast cancer is one of the most commonly diagnosed malignancies in women worldwide, and the disease burden continues to aggravate. The tumor-node-metastasis (TNM) staging information is crucial for oncology physicians to develop appropriate clinical strategies. This study aimed to investigate the value of a radiomics-clinical model for predicting TNM stage in breast cancer patients using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI). Methods: DCE-MRI images from 166 patients with pathologically confirmed breast cancer were retrospectively collected, including early stage (TNM0–TNM2) and locally advanced or advanced stage (TNM3–TNM4). Included patients were divided into a training cohort (n=116) and a test cohort (n=50). The radiomics, clinical and integrated models were constructed and a nomogram was established to distinguish the TNM0–TNM2 stage from the TNM3–TNM4 stage. Receiver operating characteristic (ROC) curves, calibration curves and decision curve analysis (DCA) were employed to assess the predictability of the models. Results: Eighty-five patients were at the early stages, while 81 patients were at the other stages. In the training and test cohorts, the area under the curve (AUC) values for distinguishing early and advanced breast cancer were 0.870 and 0.818 for the nomogram, respectively. The nomogram calibration curves showed good agreement between the predicted and observed TNM stages in the training and test cohorts. The Hosmer-Lemeshow test showed that the nomogram fit perfectly in the two cohorts. DCA indicated that the nomogram displayed clear superiority in forecasting TNM staging over clinical and radiomic signatures. Conclusions: Compared to traditional imaging methods, the clinical-radiomics nomogram acquired by DCE-MRI could potentially be utilized to preoperatively evaluate the TNM stage of breast cancer with relatively high accuracy. It can be an effective method to guide clinical decisions.