Residual disease in DCIS: an excision too far?
M. Morrow
DOI: https://doi.org/10.1097/00130404-200301000-00006
Abstract:In this issue of The Cancer Journal, Harris and co-aunegative margins. Hiramatsu et al6 also observed no significant increases in the risk of local failure in patients thors attempt to identify factors associated with residual ductal carcinoma in situ (DCIS) after an initial excision with close margins (< 1 mm). Fourquet et al7 reported an 8-year actuarial recurrence rate of 12% for patients shows positive or close margins.1 The patient population studied included only women who actually had breastwith margins of 1–2 mm, compared to with 11% of those with more widely negative margins. In contrast, conserving surgery. The failure of Harris et al1 to identify predictors of residual DCIS, particularly extensive DCIS, Chan et al8 reported a 37% local failure rate in patients with close margins (< 1 mm), compared with 4.5% for is not extremely surprising, because patients with residual disease extensive enough to require mastectomy those with a larger-width margin width. However, few of the women in the study of Chan et al8 received were not included in the study. The authors conclude that routine re-excision for patients with close or positive radiotherapy. margins remains appropriate clinical practice. The goal of breast-conservingsurgery for ductal carciThis probably does not come as a surprise to most noma in situ (DCIS) is to reduce the tumor burden to clinicians. Consensus exists that positive margins rethe level where it is likely to be controlled by radiotherquire re-excision,2 and this is supported by data from apy, or to completely remove all of the DCIS, allowing prospective randomized trials. In the National Surgical the patient to be treated by excision alone. Using serial Adjuvant Breast and Bowel Project (NSABP B24) study,3 subgross sectioning techniques in breasts with apparthe rate of ipsilateral breast recurrence after excision ently localized tumors, microscopic foci of malignancy and radiotherapy increased from 16.05 per 1000 women can be identified in 41% of cases at a distance of more per year when margins were free to 30.89 when margins than 2 cm from the reference tumor,9 yet this microwere involved. In the European Organization for Rescopic disease is controlled with radiotherapy in more search and Treatment of Cancer Trial (EORTC 10853),4 than 85% of women.4,10 If this is the case, why should a involved margins were associated with a hazard ratio of ‘‘close’’ margin be clinically significant? One explanation 2.07 for recurrence (P 4 0.0008). may be found in the growth pattern of DCIS. A study The need for re-excision in patients with close marof 60 DCIS lesions in which three-dimensional imaging gins is less clear. Unfortunately, the NSABP3 and EORTC4 of the ductal tree was used to map the extent of disease randomized trials do not address the issue of close mardemonstrated that 50% of the DCIS lesions grew in a gins, which are variably described as a 2-mm clearance discontinuous fashion, with gaps of 5 mm or more (the definition used in the study of Harris et al1), or a between tumor foci.11 This information suggests that a 1-mm clearance. In a study of 422 mammographically significant amount of DCIS could remain in the breast detected cancers with a median follow up of 9.4 years, after excision to a negative, but close, margin. Solin et al5 reported a 7% local failure rate for patients How should the practicing clinician use these apparwith close margins (tumor # 2 mm from the margin), ently conflicting pieces of information? Other studies compared to with a 9% local failure rate for those with examining the incidence of residual DCIS after an initial excision have identified tumor in 19% to 85% of cases, and large tumor size and the margins status of the initial From the Lynn Sage Breast Program, Northwestern University, Feinexcision were significant predictors of the presence of berg School of Medicine, Northwestern Memorial Hospital, Chicago, Illinois. residual tumor.12-15 The extent of margin involvement Received on November 14, 2002; accepted for publication December is a useful guide to the likelihood of identifying residual 4, 2002. tumor. Neuschatz et al15 found that the incidence of No benefits in any form have been or will be received from a commercial par ty related directly or indirectly to the subject of this article. residual tumor decreased from 85% to 31% as the extent Reprint requests: Monica Morrow, MD, Professor of Surgery, Northof margin involvement decreased from extensively posiwestern University, Feinberg School of Medicine, Director, Lynn Sage Breast Program, Northwestern Memorial Hospital, 251 East Huron tive to tumor within 1–2 mm of the margin. The precise Street, Galter Pavilion 13-174, Chicago, Illinois 60611. E-mail: margin that is encroached upon by tumor should also Copyright q 2003 Jones and Bartlett Publishers, Inc. be considered. The use of compression devices for speci-