A Nomogram to Predict Malignancy in Small and Non-parallel BI-RADS 4A Breast Lesions for Risk Stratification and Clinical Decision Support

Background Due to high false-positive outcomes, breast Imaging Reporting and Data System (BI-RADS) 4A breast lesions are often confusing for surgeons. This investigation aimed to identify specific factors of small and non-parallel BI-RADS 4A breast lesions and develop a predictive model to stratify the malignancy risk. Methods For this retrospective study, 282 patients with ultrasound-detected BI-RADS 4A lesions and planed to undergo surgery were recruited in the First Affiliated Hospital of Nanjing Medical University from January 2020 to December 2023. Logistic regression analysis was used to pinpoint independent risk factors and develop a predictive tool to distinguish between benign from malignant cases. The performance of these models was assessed by the receiver operating characteristic (ROC) curve and the decision curve analysis (DCA). Results Malignancy was confirmed in 58 out of 282 cases (20.6%). The final predictive nomogram incorporated age, menopausal status, and margin. The area under the ROC curve was 0.747 in the training cohort and 0.741 in the testing cohort. DCA indicated significant clinical net benefit across a range of threshold probabilities. Moreover, patients were categorized into low-, medium- and high-risk groups according to the malignancy risk calculated by the model. The malignancy rate in the low-risk category was minimal, recorded at 10% (5/50) in the training set and 4.8% (1/21) in the testing set. Conclusion We successfully established a robust nomogram for assessing the malignancy risk of small and non-parallel BI-RADS 4A breast lesions. This model can distinguish different risk levels, identifying a lower probability of malignant breast lesions in the low-risk group.