Utility of breast MRI for patients with breast-cancer scheduled for breast-conserving surgery.

1086 Background: Purposes of this study is to compare the operation results between MRI group and none MRI group in breast cancer patients underwent BCS and to evaluate the accuracy of breast MRI for assessing tumor size. Methods: Between January 2003 and October 2005, 1610 consecutive patients who underwent breast cancer surgery were studied retrospectively. 750 patients received BCS and 486 patients (64.8%) of MRI group and 264 patients (35.2%) of non-MRI group were studied. To evaluate the accuracy of breast MRI, we compared pathologic tumor size with tumor size with breast MRI and sonography. Results: There was no significant difference of pT and N stage between MRI group and non-MRI group in patients received BCS (p=0.065). In MRI group vs. non-MRI group, margin was positive at frozen section in 36 patients (13.6%) vs. 97 (20.0%) with statistical significance (p=0.006). In non-MRI group, 12 patients (4.5%) underwent additional breast resection and none was converted to mastectomy. In MRI group, 70 patients (14.4%) had to resect breast additionally for margin positivity and 2 patients (0.4%) were converted to total mastectomy. Additional resection was performed in MRI group more than non-MRI group (p<0.0001). Local recurrence rate after BCS was not different between MRI group (5.8%) and non-MRI group (5.3%)(p=0.766). 5-year LRFSR was not different between MRI group (94.4%) and non-MRI group (94.6%) (p=0.837). However, 5-year OSR of MRI group was 94.1%, rate of non-MRI group was 90.1% (p=0.044). Tumor size of invasive component and size on MRI was significantly different (p<0.0001), but there was no significant difference between tumor size of invasive component and size on breast sonography (p=0.717). However, tumor size including DCIS component and size on MRI was not different (p=0.548), but tumor size including DCIS component and size on sonography was significantly different (p<0.0001). Conclusions:Margin positivity and additional resection rate was higher in MRI group. OSR was slightly higher in MRI group than in non-MRI group. Breast sonography was more accurate for measuring invasive component than breast MRI, but MRI exhibited more accurate for measuring tumor size including DCIS component.