Abstract
OBJECTIVE: To compare the accuracy of contrast-enhanced digital mammography (CEDM) and MRI, including maximal intensity projection (MIP) images, for measuring the tumour size of breast cancer. METHODS: We included 52 females (mean age, 50.9 years) with surgery due to breast cancer, and measured maximum diameter of main mass on mediolateral oblique (MLO) and craniocaudal (CC) views of mammography and CEDM; sagittal, axial MIP images, and early dynamic contrast-enhanced MRI (CEMRI) before surgery. Bland-Altman plot, intraclass correlation coefficient, and univariate linear regression analysis were used to evaluate the maximum size between imaging and pathology including only invasive component (OP(invasive)) or with carcinoma in situ (OP(max)). RESULTS: Mean OP(invasive) was 15.5 mm, and overestimation rate was similar or higher than underestimation rate on all images except CC view of mammography and axial MIP image of CEDM. Mean OP(max) was 21.7 mm, and underestimation rate was higher than the overestimation rate. All parameters of CEDM and CEMRI showed good agreement ( k > 0.75) with OP(invasive), with the most favourable result being the CC view of CEDM and axial MIP image of CEMRI. CONCLUSION: All views of CEDM and MRI provided accurate measurements of tumour size. Axial plane CEDM and MRI would be the first choice for image review and treatment planning, with the highest accuracy obtained by using CC view of CEDM. ADVANCES IN KNOWLEDGE: Previous studies have not compared the measurement of the tumour size using detailed sequences; in our study, we discovered that CEDM can be an alternative modality to CEMRI.