Diagnostic Accuracy of Ductal Carcinoma In Situ: Results of Eastern Cooperative Oncology Trial 5194.
Jean F Simpson, Melinda E Sanders, David L Page. Vanderbilt University Medical Center, Nashville, TN
Background: The lay press has recently highlighted concerns regarding the over-diagnosis of ductal carcinoma in situ (DCIS). Given the difference in natural history and therapeutic implications, reliably separating DCIS from atypical ductal hyperplasia (ADH) is critical. Misdiagnosis rates of 20% have been quoted, but are anecdotal; no systematic study of this issue has been undertaken. The Eastern Cooperative Oncology (ECOG) Trial 5194 was conducted to investigate whether women with small examples of DICS could forego radiation therapy. Particpating institutions were required to submit all H&E slides of DCIS for central review before a woman was deemed eligible for the trial. The current study was undertaken to examine the frquency of over-diagnosis of DCIS.
Design: Women were eligible for ECOG 5194 if (1) non-palpable DCIS lesions were sequentially embedded so that the extent of involvement could be accurately determined, (2) low or intermediate grade DCIS was 2.5 cm or smaller, (3) high grade DCIS was 1.0 cm or smaller, and (4) margins of excision were negative by at least 3 mm. Slides from the originating pathologist (OP) were reviewed centrally by us, with immediate feedback to the OP. Patients were eligible if the margin was <3 mm if a subsequent re-excision had negative (>3mm) margins. Data collected included patient age, laterality of DCIS, size, type, DCIS grade and margin distance. For the current study, diagnoses from the CP and OP were compared. All submitted cases had been diagnosed as DCIS by the OP.
Results: Slides from 711 women were received for central review. Six hundred sixty-one women were eligble and enrolled in this prospective trial. Median age was 60 years (range 28-88); median DCIS size was 6 mm. For intermediate or high grade DCIS, the CP diagnosis agreed with the OP diagnosis of DCIS in 99% of cases. The CP review disagreed with the OP diagnosis of low grade DCIS in 19% of cases; most were classified as atypical ductal hyperplasia on CP review. Other reclassified cases included lobular carcinoma in situ, usual hyperplasia without atypia, mucocele-like lesion, and papillary apocrine change.
Conclusions: The distinction between ADH and low grade DCIS is based on cytology, architecture, and extent of involvement. The cases that were downgraded to ADH on central review did not fulfill the extent criterion. Given the differences in natural history and therapeutic implication, careful evaluation of extent of involvement will assure an accurate diagnosis of low grade DCIS.
Monday, February 28, 2011 11:15 AM
Platform Session: Section C, Monday Morning