Predicting the likelihood of residual disease in women treated for ductal carcinoma in situ.

ABSTRACT

BACKGROUND: To identify women at risk for residual disease after excision of ductal carcinoma in situ (DCIS), we assessed the relationship between characteristics of the initial biopsy and the presence of residual DCIS at a subsequent operation. STUDY DESIGN: We identified 134 consecutive "paired" operations from 112 women who had undergone 2 or more operations for DCIS between February 1995 and December 1996. Cancer status of the margins, patient age and leading presentation, tumor subtype and grade, and the presence of multifocal-extensive disease were assessed as potential predictors. RESULTS: Residual DCIS was found in 60 patients (45%) in 2 of 12 patients (17%) with negative margins, in 11 of 36 (31%) with close margins (< 2 mm), in 30 of 52 (58%) with positive margins, and in 17 of 34 patients (50%) with margins of unknown status. Patients with positive or unknown margins were 7.7 and 8.3 times, respectively, more likely to have residual disease than patients with negative margins (95% CI 1.1-59.1; 1.1-66.4). Patients with clinical presentations were 8.0 times more likely to have residual disease than patients who presented with abnormal mammograms (95% CI 2.3-27.6). Multifocal-extensive DCIS was associated with residual disease (adjusted odds ratio [OR] = 7.7, 95% CI 2.9-20.5), as was comedo subtype (OR = 2.7, 95% CI 1.1-6.7). CONCLUSIONS: Positive or unknown biopsy margins, a clinical presentation, multifocal-extensive cancer, and the comedo subtype are associated with higher risk of residual DCIS.