Multivariate analysis of predictors of positive or close margins in breast-conserving therapy.

Abstract

105 Background: Local recurrence rates after breast-conserving therapy (BCT) are increased with positive or close margin lumpectomy. This group of patients usually requires a re-excision lumpectomy or a completion mastectomy, causing delays in systemic therapy, and a less desirable cosmetic outcome. The objectives of this study were to identify any predictive risk factors that might be associated with positive or close lumpectomy margins (<1mm), and thus have implications for the planning of appropriate primary surgical excision. Methods: This was a retrospective cohort study of 195 consecutive patients from a prospectively collected, single-institution database. Patients who underwent primary lumpectomy for ductal carcinoma in situ (DCIS), invasive ductal carcinoma, and invasive lobular carcinoma from October 2007 to July 2010 were reviewed. Statistical analysis of the data was performed using Chi-squared analyses, multivariate logistic regression and Student’s t-test to identify significant predictors of a close or positive margin following lumpectomy. Results: In this study, 53.8% had positive or close margins. Ninety-two percent of this group underwent re-excision (86.5%) and total mastectomy (13.5%). We also observed that 53.1% of patients who underwent second operation had no residual disease. Factors significantly associated with positive or close margins include a family history of breast cancer (OR=2.3, p=0.01), a non-palpable mass (OR=2.4, p=0.01), or an excisional biopsy (OR=6.02, p=0.022). Age, race, use of hormone replacement therapy, menopausal status, use of preoperative MRI or US, BIRAD, use of neoadjuvant chemotherapy, tumor size, staging, waiting time, histology, receptor status, and axillary status were not significantly correlated with positive or close margins. Conclusions: For patients who were recommended to have a lumpectomy, an increased risk of a positive or close margin was significantly associated with a family history of breast cancer, a nonpalpable mass, or an excisional biopsy. These predictors should be weighed in the decision for attempting lumpectomy with the goal of a negative margin versus the ability to obtain a satisfactory cosmetic outcome.