Abstract P5-15-03: Prediction of positive resection margins in patients with non-palpable breast cancer

Abstract

Background Obtaining tumor free resection margins is essential in patients undergoing breast conserving surgery. Several risk factors associated with positive margins are described in literature. We developed a prediction model to predict positive resection margins in patients undergoing breast conserving surgery of non-palpable lesions. Methods A total of 576 patients with non-palpable invasive breast cancer underwent breast conserving surgery at five different hospitals in the Netherlands. A prediction model for positive resection margins was built using multivariate regression analysis and internally validated by bootstrapping. Results Positive resection margins were present in 69/576 (12%) patients. Factors associated with positive margins included microcalcifications on mammography (OR 1.8, 1.0-3.2), tumor not visible on ultrasound (OR 2.6, 1.2-5.6), presence of DCIS (OR 2.3, 1.3-4.0), multifocality (OR 3.5, 1.0-12.1), caudal location in the breast (OR 1.9, 1.1-3.5), and invasive tumor size (OR 1.83, 1.6-2.7). Together, these factors were able to moderately discriminate between patients with positive versus negative margins (area under the ROC 0.71, 95% CI 0.648 – 0.780). After internal validation the discrimination was slightly lower with an AUC of 0.694. Prevalence of positive margins was 5.2% in the highest risk quintile versus 26.3% in the lowest quintile. Conclusion A model predicting positive resection margins after breast conserving surgery in non-palpable breast cancer was built. This model is moderately able to differentiate between women with high versus low risk of positive margins, and may be useful for surgical planning (eg. preoperative MRI) and informing of patients. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P5-15-03.