Background: The presence of tumor at the inked margins (IMs) of breast specimens is associated with an increased risk of local recurrence after breast-conserving therapy for invasive breast carcinoma and ductal carcinoma in situ (DCIS). Given the importance of margin status, some have advocated the use of shaved margins (SMs) as a means of conducting a more complete examination of the specimen margins than could be done with sections taken perpendicular to the IMs. However, it is not known whether these two methods of margin assessment provide comparable information.
Methods: To address this issue, the authors studied 22 consecutive breast reexcision specimens (10 DCIS, 6 infiltrating ductal carcinomas, and 6 infiltrating lobular carcinomas) in which the specimen surfaces were inked, the margins were shaved, and tumor was present in at least one of the SM sections. A total of 199 SMs were examined. The SMs were originally embedded in a way that permitted histologic sections to be cut opposite the inked surface. Sections of SM stained with hematoxylin and eosin (H & E) were reviewed and scored for the presence and extent (number of low-power fields) of cancer. The remaining tissue from the SM was then removed from the blocks, cut perpendicular to the IM, and reembedded to permit visualization of tumor in relation to the IM. Sections were then cut from two different levels of each reembedded block and stained with H & E. An SM was considered positive if tumor was present anywhere on the section. An IM was considered positive when tumor extended to the inked surface.
Results: Although all 22 excisions had at least 1 positive SM, tumor was present at an IM in only 12 specimens (55%). Among 69 positive SMs, the corresponding IM was positive in only 42 (61%). The likelihood of a positive IM increased with the number of low-power fields of involvement by invasive carcinoma or DCIS on the SM, as follows: 19% with 1 low power-field, 67% with 2 low-power fields, and 97% with > or = 3 low-power fields (all P < 0.02). When the SM was negative, the corresponding IM was negative in 98% of cases.
Conclusions: Many patients with positive SMs do not have positive IMs. A positive SM more reliably predicts a positive IM when tumor involves > or = 3 low-power fields of the SM. The authors conclude that the clinical implications of a positive SM may not be the same as those of a positive IM. Clinical outcome studies are needed to define further the implications of positive SMs. [See editorial counterpoint on pages 1453-8 and reply to counterpoint on pages 1459-60, this issue.]