Boetes et al.23 | Retrospective | Patients with invasive lobular carcinoma treated surgically | 34 (36 cases of breast cancer) | False negative rate for MRI was 0% compared with 3% and 14% for US and XRM, respectively; when 2 radiologists retrospectively reviewed the exam results, the percentage of correctly identified size of cancer was 47% and 75% for MRI (r = 0.81, P < .01); this was the most accurate and most highly correlated method |
Schelfout et al.17 | Prospective | Women with lesions on XRM, US, and/or CBE | 212 | 96%, 37%, and 41% of multifocal disease was detected by MRI, XRM, and US, respectively; 95%, 18%, and 9% of multicentric disease was detected by MRI, XRM, and US, respectively; 100% of bilateral breast cancers were seen on MRI; 56% of bilateral breast cancers were seen on XRM and US |
Van Goethem et al.24 | Prospective | Patients with dense breasts planning to undergo surgery | 67 | 65/67 patients had breast cancer confirmed pathologically; MRI was 98% sensitive for initial lesion compared with 83% and 70.8% for XRM and US, respectively; extent of cancer was underestimated by 12.5% of MRI results as opposed to 37% and 40% for XRM and US results, respectively; multifocal/multicentric disease was picked up 100% of the time by MRI as opposed to 35% and 30% of XRM and US, respectively |
Bedrosian et al.18 | Retrospective | Patients diagnosed with invasive breast cancer who had MRI preoperatively | 267 | MRI was 95% sensitive for detecting primary breast cancer; planned management of 26% of patients (N = 69) were changed due to MRI results—in 49/69 (71%) of the patients, postsurgical pathology confirmed that the change in management was appropriate |
Rieber et al.28 | Prospective | Patients suspected of having breast malignancy based on results of XRM, US, or CBE | 43 | Sensitivity of MR mammography for diagnosis of primary cancer, contralateral cancer, and multifocal disease was 100%, 100%, and 95.2%; similar values were 93%, 100%, and 92.5% for PET, respectively |
Fischer et al.20 | Prospective | Patients with breast abnormalities after XRM, CBE, PE, and US | 463 | 66 patients (14.3%) had their planned therapy accurately changed as a result of MRI of the breast with 16 patients (3.5%) undergoing unneeded open biopsy |
Esserman et al.21 | Prospective | Patients with breast cancer with planned surgical correction | 57 (58 cases of breast cancer) | Preoperative MRI identified degree of disease accurately in 54/58 cases of breast cancer; anatomic detail identified on MRI was accurate 98% of the time whereas anatomic detail identified on XRM was accurate in 55% of cases |
Rodenko et al.22 | Retrospective | Patients with infiltrating lobular carcinoma who had MRI and XRM performed preoperatively | 20 | Postoperative pathology correlated with preoperative MRI findings in 85% of patients; The correlation with preoperative XRM was 32% (P < .0001) |