With its high-resolution 3D images, Cone Beam Breast CT (CBBCT) has shown promising results for its application in neoadjuvant chemotherapy (NAC) monitoring in the initial studies. Comparing to MRI images, added features of the calcifications are showed on CBBCT images. The purpose of this study is to demonstrate the accuracy of CE-CBBCT and dynamic contrast material-enhanced breast MRI (DCE-MRI) for determining residual tumor size after NAC in breast cancers accompanied with and without calcifications.

Between January 2019 and April 2020, 66 women (mean age, 44.0 years± 8.7; range, 25-69 years) underwent CE-CBBCT and DCE-MRI after NAC. All patients received surgery after NAC. Surgical pathology results showed that 32 women were invasive breast cancer accompanied with calcifications, and 34 women were invasive breast cancer without calcifications. Tumor sizes were measured in CE-CBBCT and DCE-MRI at first phase and second phase images (60-90 and 120-180 seconds after contrast material injection, respectively). At histopathologic examination, overall tumor size (including both invasive and in situ components) and the size of invasive component alone were separately recorded as gold standard. Absolute agreements of tumor sizes between CE-CBBCT, DCE-MRI, and histopathologic examination were assessed by using intraclass correlation coefficient (ICC) analysis. Factors affecting size discrepancy were assessed by using multiple linear regression analysis.

The overall tumor sizes measured at 2nd phase CE-CBBCT and DCE-MRI showed higher agreement with histopathologic results than the overall tumor sizes measured at 1st phase CE-CBBCT and DCE-MRI (CBBCT ICC 1st phase vs 2nd phase: 0.68 vs 0.79; P<0.05. MRI ICC 1st phase vs 2nd phase: 0.72 vs 0.87; P<0.05). Calcifications and tumor subtypes were independently associated with greater size discrepancy (P<0.05). Using histopathologic results as gold standard, in CE-CBBCT, residual tumors with calcifications were overestimated compared with those without calcifications (P<0.05), but no significant difference was observed in MRI (p>0.05). In both CE-CBBCT and MRI, HER2-positive cancers were overestimated compared with (HER2)-negative cancers (12mm ± 5.0 vs 6mm ± 3.0, P <0.05) and triple-negative cancers (12mm± 5.0 vs 3mm ± 2.0, P <0.05).

Second phase CE-CBBCT and DCE-MRI are more accurate than first phase CE-CBBCT and MRI for evaluating residual breast tumor size after neoadjuvant chemotherapy. Overestimation of tumor sizes was found both in CE-CBBCT and MRI.

CE-CBBCT and DCE-MRI were all accurate to evaluate the residual breast tumor size after neoadjuvant chemotherapy. Overestimation of tumor sizes after NAC was found both in CE-CBBCT and MRI.