Margin Index Does Not Predict Residual Disease in DCIS

Pam Harrison

May 06, 2011

May 6, 2011 (Washington, DC) — Margin index is not a reliable method of predicting residual disease after breast-conserving surgery with close margins in patients with ductal carcinoma in situ (DCIS), according to a study presented here at the American Society of Breast Surgeons 12th Annual Meeting.

Investigator Carla Fisher, MD, from Washington University School of Medicine, St. Louis, Missouri, acknowledged that she was unable to identify an optimal margin index that reliably predicted the presence or absence of residual disease after excision of DCIS.

"A margin index of greater than 5 had a sensitivity of 84%," she noted, "but a specificity of only 37%. This is in comparison with our previous data in stage I and II breast cancer patients, where a margin index of greater than 5 was shown to be predictive of the presence or absence of residual disease."

Sensitivity measures the proportion of actual positives that are correctly identified as such, whereas specificity measures the proportion of negatives that are correctly identified, Dr. Fisher pointed out.

Comparison Study Yields Clues

For the study, the investigators reviewed patients undergoing breast-conserving surgery for DCIS at the Washington University School of Medicine from June 2004 to January 2010, and at Northwestern University, in Evanston, Illinois, in 2009. A total of 380 patients were identified, of whom 271 underwent no further surgery and were excluded from the study. "This left a re-excision rate of 29%," Dr. Fisher said, "and 109 patients underwent additional surgery for close or positive margins."

On additional surgery, 46 patients had positive margins and, in another 15, the investigators were unable to assess the size of DCIS on retrospective pathology; all were excluded from the study. This left 48 patients with close margins for whom a margin index was calculated. The margin index is equal to the closest margin (in mm) divided by the tumor size (in mm) × 100.

"Most patients were Caucasian with estrogen-receptor- or progestin-receptor-positive breast cancer," Dr. Fisher reported. Nuclear grade was variable; 21% were grade 1 DCIS, 44% were grade 2 DCIS, and 31% were grade 3 DCIS. In general, patients underwent excision with margins of 2 mm or less, although most patients had margins of 1 mm or less. Most patients also had a tumor burden of 4 cm or less.

"For patients who underwent re-excision, 40% were found to have residual disease," Dr. Fisher reported. "But we found no reliable cutoff value at which to predict residual disease in DCIS." For example, using a cutoff value above 1, the margin index had a sensitivity of only 5% and a specificity of 95%.

A cutoff value above 10 resulted in much greater sensitivity (84%), but a low specificity (21%); a cutoff value above 20 had a high sensitivity (89%) but a low specificity (11%).

In contrast, earlier work has shown that in stage I to II breast cancer, a margin index above 5 had a sensitivity of 85% and a specificity of 73%.

"We feel that margin status and size of DCIS are important in evaluating margins in [breast-conserving surgery]," Dr. Fisher concluded, "but we also feel that a larger and also prospective study is needed to definitely determine the role of margin index in DCIS."

Commenting on the findings, Sheldon Feldman, MD, from Columbia University in New York City, and cochair of the session, noted that measuring the extent of disease in DCIS is very difficult using standard pathology processing.

"DCIS fills the milk duct and does not present as a mass, so there is no discrete tumor that can be measured," he told Medscape Medical News. "There are challenges when you try to get a robust measurement of the size of DCIS, and it's not surprising they found that the margin index was not a reliable method."

New Device for Measuring Tumor Margins on Horizon

Findings that a margin index does not reliability predict residual disease after breast-conserving surgery in DCIS supports the need for other innovations to help detect positive margins in the operating room. The newly developed MarginProbe device might prove to be the first of those innovations.

Lorraine Tafra, MD, from Anne Arundel Breast Center, Annapolis, Maryland, presented the data from 664 women with breast cancer undergoing lumpectomy who were enrolled in a study designed to evaluate the intraoperative margin probe. Of this cohort, 596 were randomized in the operating room to either the device group or the control group. In the device group, the intraoperative margin probe was used in addition to standard-of-care methods to measure each margin of the excised lumpectomy specimen.

"The device has been engineered to give you only 1 of 2 answers — yes or no," Dr. Tafra told Medscape Medical News. "So when you do a reading of the margin, which we do in 5 or 6 sites, and 1 comes back positive, you can make a decision to widen the margins then and there, and it is very easy to use."

The primary end point of the study in which the device was evaluated was complete surgical resection, which measured the surgeons' ability to identify and resect all positive margins when present. Approximately 50% of patients had 1 or more positive margins on the lumpectomy specimen, as determined by later pathology.

As Dr. Tafra reported, the successful complete surgical resection rate was 72% in the device group and 22% in the control group (P < .0001). Significantly fewer patients in the device group than in the control group underwent re-excision (33 vs 62 patients), for a 47% reduction in the need for re-excision in favor of the device (P = .001).

"Margins are becoming more and more important because tumors are getting smaller and more and more patients are getting lumpectomies," Dr. Tafra told Medscape Medical News. "The challenge is to tailor your surgery and try to take out enough tissue so margins are clear, but not to take too much out, otherwise it will deform the breast. This device gives us a very quick and accurate assessment of the margins in the [operating room]."

Dr. Feldman participated in the intraoperative margin probe study, and noted that "we are very enthusiastic about this, as we want to remove the right amount of tissue, but not too much." The next step, he added, would be to develop a device that can test for the presence of residual disease within the lumpectomy cavity itself.

"We care more about what has not been removed than what has been removed, so the next step would be to develop a device that would allow us to double-check and see what's still in there."

The MarginProbe has not yet been approved by the US Food and Drug Administration, so it is not currently available.

Dr. Fisher and Dr. Feldman have disclosed no relevant financial relationships. Dr. Tafra reports receiving funding from Dune Medical Devices.

American Society of Breast Surgeons (ASBS) 12th Annual Meeting: Abstract 1661, presented April 30, 2011, and abstract 1750, presented April 29, 2011.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.