Accuracy of Breast Magnetic Resonance Imaging (MRI) and Breast Ultrasound Compared to Pathology in Assessing Residual Tumor in Breast Cancer Patients Receiving Neoadjuvant Systemic Treatment at Siriraj Hospital

Objective: To evaluate the accuracy of magnetic resonance imaging (MRI) and ultrasound (US) in assessing residual tumor size compared to pathological findings in breast cancer patients who received neoadjuvant systemic treatment (NAST), and to examine the influence of imaging on surgical planning across different molecular subtypes. Materials and Methods: This retrospective study included 24 breast cancer patients who underwent NAST followed by surgery at Siriraj Hospital between 2016 and 2024. Preoperative breast MRI and breast US, performed within 3–6 weeks prior to surgery, were compared with pathological tumor size. Analysis focused on mass lesions, with nonmass enhancement (NME) considered in selected cases where it presented. Imaging findings were independently reviewed by a second, blinded radiologist. Concordance between imaging and pathology was assessed. Results: A total 24 patients were analyzed. MRI showed superior agreement with pathological tumor size, with 91.7% of cases falling within a ±10 mm margin, compared to 70.0% for US. Mean tumor sizes were 5.4 mm for MRI, 8.3 mm for US, and 10.9 mm based on pathological examination. Bland-Altman analysis revealed better agreement between MRI and pathology (limits of agreement: –23.5 – 34.5 mm) compared to US. These results highlight the superior accuracy and reliability of MRI over US for preoperative tumor size assessment. Conclusion: MRI demonstrates greater accuracy than US in evaluating residual disease following NAST. In case of invasive lobular carcinoma (ILC) subtypes, incorporating NME into imaging assessment may improve concordance with pathological findings. MRI should be considered a standard modality in preoperative evaluation to optimize surgical planning.