Factors Associated with Lymph Node Assessment in Ductal Carcinoma in situ: Analysis of 1988–2002 Seer Data

Abstract

Ductal carcinoma in situ (DCIS) represents 20-30% of mammographically detected breast cancers, but the role of lymph node assessment (LNA) in women with DCIS remains unclear. Using the 1988-2002 Surveillance, Epidemiology, and End Results (SEER) Program data, we conducted a case-control study to identify variables associated with (1) LNA in DCIS patients and (2) use of axillary lymph node dissection (ALND) compared with sentinel lymph node biopsy (SLNB). Using separate multivariable logistic regression models, we identified patient and tumor-related factors associated with LNA (1988-2002) and with the method used (recorded only in 1998-2002). We report adjusted odds ratios (aORs) and 95% confidence intervals (CIs). Of 23,502 women with DCIS, 37% underwent mastectomy and 63% underwent breast-conservation therapy (BCT); 6,650 cases (28%) underwent LNA. Women younger than 80 years (aOR 1.47; 95% CI 1.24-1.75) or who had mastectomy (aOR 11.06; 95% CI 10.30-11.90), tumor size greater than 9 mm (aORs ranged from 1.27-1.97 for 10-mm increments from 10 to 50 mm or more) or poorly differentiated grade (aOR 1.33; 95% CI 1.11-1.55) were more likely to have had a LNA. From 1998 to 2002, 10,637 women underwent resection for DCIS (21% mastectomy; 79% BCT); of these, 2,219 (21%) had LNA (73% mastectomy; 27% BCT). Mastectomy patients were 3.52 times more likely to receive ALND (95% CI 2.71-4.57) than SLNB, after controlling for other factors. Optimal guidelines for use of LNA in DCIS have not been defined. However, there appeared to be a persistent and excessive utilization of ALND for LNA in women with DCIS (1998-2002).