Clin Radiol. 2026 May 12;98:107356. doi: 10.1016/j.crad.2026.107356. Online ahead of print.
ABSTRACT
PURPOSE: To compare the performance of preoperative contrast-enhanced mammography (CEM) and magnetic resonance imaging (MRI) in assessing index tumour size and detecting additional malignant lesions and to evaluate total disease extent using postoperative histopathology as the reference standard.
METHODS: Retrospective analysis of 52 women with biopsy-proven breast cancer who underwent both CEM and MRI between July 2019 and December 2023. Two radiologists independently reviewed studies at separate times to reduce recall bias and were blinded to pathology. Index lesion size, additional suspicious lesions, and total disease extent were recorded. Statistical analysis included intraclass correlation coefficient (ICC) for lesion size, Kappa statistics for additional lesion detection, and Bland-Altman plots to assess agreement with histopathology.
RESULTS: CEM detected 51 of the 52 (98%) index lesions; MRI detected all (100%). Mean index lesion size was similar (CEM 24.9 ± 22.9 mm vs MRI 25.2 ± 22.8 mm; ICC = 0.975). Additional lesions were identified in 23 of the 52 patients, with very good agreement between modalities (Kappa = 0.881, P<.001). Among 42 patients with histopathological data on total disease extent (including multifocal or multicentric disease), CEM measurements closely matched histopathology (mean 32.6 mm vs 32.6 mm), while MRI slightly overestimated extent (mean 35.0 mm vs 32.6 mm). Discrepancies >20 mm occurred in five patients, mainly in cases with non-mass enhancement.
CONCLUSION: CEM shows high concordance with MRI for measuring index lesion size and detecting additional suspicious lesions, with closer agreement to histopathology for total disease extent. CEM is a viable, efficient alternative to MRI for preoperative locoregional staging of breast cancer.
PMID:42119266 | DOI:10.1016/j.crad.2026.107356
