Patient-Level Evaluation of Optimal Margin Width for Lumpectomy in Ductal Carcinoma In Situ: An Analysis of Patients in the National Cancer Database Special Study

PURPOSE Re-excision for close margins in patients undergoing breast-conserving surgery (BCS) for ductal carcinoma in situ (DCIS) aims to reduce recurrence rates. The current recommended margin width of ≥2 mm is based on study-level data. Using a multicenter cohort, we aimed to determine whether patient-level analysis of margin width support that negative margins <2 mm confer higher recurrence rates than margins ≥2 mm. METHODS We conducted a retrospective, observational cohort study of women diagnosed with DCIS between 2008 and 2015 treated with BCS. Patient, diagnosis, treatment, and outcome data were abstracted by local registrars and from a national database. Final margin width was categorized as positive (tumor on ink), 0.01-0.9 mm, 1.0-1.9 mm, or ≥2 mm. The primary outcome was ipsilateral recurrence. The Kaplan-Meier method and multivariable Cox proportional hazards models were used to estimate the effect of margin width on recurrence risk. RESULTS Among 11,156 women in the final cohort, 72.2% (n = 8,059) received BCS + radiotherapy and 27.8% (n = 3,097) received BCS alone. Median age at diagnosis was 61 years; median follow-up was 66.6 months. A total of 445 (4%) women experienced a recurrence during follow-up. In a multivariable adjusted analysis, there was no significant difference in recurrence risk between the negative margin groups both in patients who had either BCS with radiation or BCS alone. In subgroup analyses by age, race, hormone receptor status, grade, size, endocrine therapy, and radiation treatment, only the age <50 years subgroup showed a significant difference in recurrence risk between narrow (0.01-1.9 mm) and wide (≥2 mm) margins. CONCLUSION In this multivariable analysis of 11,156 patients, negative margins <2 mm were not associated with a higher local recurrence risk compared with margins ≥2 mm except in patients younger than 50 years at diagnosis.