Impact of Cavity Shave Margins on Margin Status in Patients with Pure Ductal Carcinoma In Situ

BACKGROUND: We examined the impact of cavity shave margins (CSMs) on margin status in patients with pure ductal carcinoma in situ (DCIS) undergoing partial mastectomy (PM). METHODS: One hundred and nine patients from 2 multicenter, randomized controlled trials were identified with pure DCIS (no invasive cancer). Surgeons performed their best PM, with specimen radiography and resection of selective margins per surgeon discretion. Patients were then randomized to have CSM resected or not. A positive margin was defined as <2 mm from ink. RESULTS: Median patient age was 63 years; median size of DCIS was 1.20 cm; 43.6% of patients had high-grade DCIS; and 58 (53.2%) patients were randomized to take CSM. The shave and no-shave groups were well-matched for age, race, ethnicity, palpability, grade, and size of DCIS. Although 33 (56.9%) of the patients in the shave group had a positive margin before randomization, only 12 (20.7%) had a positive margin after randomization to CSM (p < 0.001). In the no-shave group, 17 patients (33.3%) had a positive margin. Controlling for size and grade of DCIS, taking CSM resulted in a nearly 65% reduction in the positivemargin rate (odds ratio 0.366; 95% CI, 0.136 to 0.981; p = 0.046). Size of DCIS remained an independent predictor of positive margins in the model (odds ratio 1.646; 95% CI, 1.227 to 2.209; p = 0.001). CONCLUSIONS: CSM reduces positive-margin rates in patients with pure DCIS, and can be a practical solution for DCIS patients who tend to have a high rate of margin positivity. ((C) 2020 by the American College of Surgeons. Published by Elsevier Inc. All rights reserved.)

Automated summary

The study found that in patients with pure DCIS, cavity shave margins (CSM) can significantly reduce positive-margin rates, making it a practical solution for DCIS patients who tend to have a high rate of margin positivity. Size of DCIS remained an independent predictor of positive margins in the model.