Head‑to‑head comparison of contrast‑enhanced mammography and MRI in assessing the tumor response to neoadjuvant therapy in breast cancer: a prospective, multireader study.

Purpose: To compare contrast-enhanced mammography (CEM) and MRI in the evaluation of the tumor response to neoadjuvant therapy (NAT) in breast cancer.

Materials and methods: Patients who underwent CEM and breast MRI before and after NAT were prospectively enrolled in this study. Three readers with different levels of experience independently assessed the images and classified the tumor response following NAT. Radiological complete response (rCR) was defined as the absence of a suspiciously enhanced area. The sensitivity, specificity, and accuracy of CEM and MRI in determining pathological complete response (pCR) were calculated per reader and averaged with a generalized estimating equation. The agreement in the residual tumor extent between imaging and pathology was assessed by intraclass correlation coefficient (ICC) analyses.

Results: Fifty-seven patients (age: 47 ± 10 years) were enrolled, 21/57 of whom (36.8%) achieved pCR. In determining pCR, the estimated average sensitivity, specificity, and accuracy were 0.778, 0.714, and 0.754 for CEM, respectively, and 0.714, 0.944, and 0.860 for MRI, respectively (all P values > 0.050). Both imaging modalities slightly overestimated the residual lesion extent (mean overestimations: 1.7 mm and 2.7 mm). In ICC analysis, CEM and MRI showed poor consistency with pathology (average ICCs = 0.4467 and 0.449, respectively).

Conclusions: Compared with MRI, CEM showed comparable but slightly lower accuracy, higher sensitivity, and lower specificity in predicting pCR after NAT. In assessing residual tumor extent, CEM and MRI showed similar performance and tended to overestimate the pathological extent. CEM could be an acceptable alternative to MRI in certain patients undergoing NAT.