Abstract P4-02-11: Accuracy of breast magnetic resonance imaging has limited value to reduce the margin-positive rate: A study in relation to the molecular subtypes

Abstract

Abstract Background: Newly released guideline standardizing a negative margin after breast-conservative surgery (BCS) as “no ink on tumor” by SSO-ASTRO stressed the importance of estimation tumor extent with comprehensive breast imaging studies. To evaluate clinical value of breast magnetic resonance imaging (MRI) in patients with BCS, we compared the degree of correlation between MRI-pathology and ultrasonography (US)-pathology according to subtypes. In addition, we investigated the margin-positive rates and secondary operation rates for margin clearance. Methods: We identified patients with invasive breast cancer who had preoperative breast MRI and ultrasound between 2011 and 2016. We excluded patients having large tumor more than 5cm or multiple tumors or undergoing mastectomy. Patients were classified into 4 subtypes based on the immunohistochemistry; luminal A, luminal B/HER2, HER2, triple-negative breast cancer (TNBC). Lin's concordance correlation coefficient was used to measure the agreement between the MRI or US and tumor extent. Tumor extent was defined as pathologic tumor size including in situ carcinoma. Margin-positivity was assessed based on intraoperative frozen examination. Results: A total 516 patients with single tumor undergoing BCS were included. Means of tumor size were 1.99 ± 0.91 cm by pathologic examination, 1.91 ± 1.01 cm by MRI, and 1.76 ± 0.92 cm by US, respectively. The correlation coefficient of MRI-pathology was significantly higher than that of US-pathology (r=0.6975 vs. 0.6211, P=0.001). A superiority of MRI than US in measuring pathologic extent was only observed in TNBC (r=0.8089 vs. 0.6014, P<0.001), whereas the agreement between the MRI or US and tumor extents was low in the HER2 (MRI: 0.3509, US: 0.3165). Also, the margin-positive rate was higher in HER2 (luminal A, 11.6%; luminal B/HER2, 17.5%; HER2, 29.6%; TNBC, 17.8%; P=0.0382). In the post-hoc test, the HER2 was more likely to have positive margin compared to Luminal A (P=0.0039). There is no significant difference in secondary operation as margin clearance according to the subtypes (P>0.999). Margin positive and re-excision rates according to the subtypes Luminal A (n=302)Luminal B (n=80)HER2 (n=27)TNBC (n=107)P valuePositive margin35 (11.6)14 (17.5)8 (29.6)19 (17.8)0.0382Re-excision14 (4.6)4 (5.0)1 (3.7)5 (4.7)>0.9999 Conclusions: Given a superiority of MRI to US in preoperative assessment, MRI-guided BCS did not reduce the margin-positive rate in TNBC. In the HER2, size correlation of MRI-pathology was very low, and the margin-positive rate was high. Collectively, our findings suggest that accuracy of MRI has limited value to reduce the margin-positive rate. Citation Format: Bae SJ, Ahn SG, Yoon C, Cha YJ, Jeong J. Accuracy of breast magnetic resonance imaging has limited value to reduce the margin-positive rate: A study in relation to the molecular subtypes [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P4-02-11.