An Immunohistochemistry-based Decision Flowchart to De-escalate Axillary Surgery for Biopsy-Diagnosed Pure Ductal Carcinoma In Situ (DCIS) of the Breast.

The treatment of ductal carcinoma in situ (DCIS) of the breast and invasive breast cancer (IBC) is considerably similar, despite their different biology. Axillary surgery is increasingly de-escalated by omitting sentinel lymph node biopsies (SLNBs) for hormone receptor-positive, human epidermal growth factor receptor 2-negative (HER2-) early-stage IBCs, supported by recent clinical trial data. The American Society of Clinical Oncology/National Comprehensive Cancer Network guidelines recommend SLNB for patients with DCIS treated with mastectomy, although some patients treated with lumpectomy still undergo SLNB. We aimed to identify markers to predict the upstaging risk after a biopsy diagnosis of pure DCIS and provide data to justify de-escalation of axillary surgery for patients with DCIS. The following features were retrospectively reviewed for 258 DCIS: nuclear grade, necrosis, histological calcifications, stromal architecture, DCIS architecture, tumor-infiltrating lymphocytes, estrogen receptor (ER), and progesterone receptor. Immunohistochemistry for HER2 was performed. Statistical analyses were performed with upstaging to (micro-)invasive carcinoma as the outcome variable. Clinical and radiological data were included in multivariable logistic regression. Solid architecture (P = .001) and radiological size (>20 mm; P = .025) were independently associated with upstaging risk. Two algorithms were created, based on surgery type, ER/HER2 status, and upstaging risk. High upstaging risk was defined as solid architecture and/or radiological size >20 mm (model 1) or solid architecture and/or absent histological calcifications (model 2). Each model identified a single patient with upstaging to ER+/HER2- IBC, without a positive SLNB. Both models showed that SLNBs may be safely omitted in low-risk ER+/HER2- DCIS, even when treated with mastectomy. However, model 1 identified fewer patients as low risk than model 2: 10 patients (17%) vs 32 patients (55%) within the group of 58 ER+/HER2- DCIS treated with mastectomy. Model 2 seems, therefore, more promising for further validation. Routine implementation of this decision flowchart requires retrospective validation before moving to prospective clinical trials but could substantially de-escalate axillary surgery for future patients with DCIS.