Understanding the mechanisms creating false positive lumpectomy margins.

OBJECTIVE: All breast surgeons deal with the frustration of initial pathologic close or positive margins that have no residual cancer upon re-excision. To understand the mechanisms that create false positive margins, specimen handling was standardized in a single surgeon's practice and margin issues were tracked.
METHODS: Prospectively over a 3.5-year period, needle-localized lumpectomies for the management of early-stage breast cancer were standardized in all aspects of specimen handling, including surgeon inking and specimen compression for specimen radiography for quality assurance. The current study reviews 220 such cases where the original lumpectomy included a small piece of overlying skin from over the target lesion. All specimen radiography was performed with compression at the skin to deep level to bias the "pancaking" effect of pushing tumor to specimen surface to the deep margin.
RESULTS: Of the 220 therapeutic lumpectomies performed for clinical stage 0-2 breast cancer in this fashion, 175 (79.5%) had negative margins by a distance of 10 mm or more. Margins less than 10 mm were classified as close and were present in 20 (9.1%) of cases. These were heavily biased toward margins closer than 2 mm. Positive initial margins accounted for 25 (11.4%) of cases. Of the 45 close or positive margin cases, 12 involved the deep margin only, and on re-excision none was found to have residual tumor. When other single margins were involved, re-excision found tumor in 5 of 14 cases (35.7%). When multiple margins were close or positive, 9 of 19 cases (47.3%) were found to have residual tumor at re-excision.
CONCLUSIONS: Specimen compression increases the incidence of false margin positivity. The best predictors of true margin positivity are multiple close or positive margins or margin positivity in a direction not associated with specimen ex vivo compression.