1. Ductal Carcinoma in Situ with Microinvasion: Prognostic Implications, Long-term Outcomes, and Role of Axillary Evaluation Rahul R. Parikh, MD 1, Bruce G. Haffty, MD 1, & Meena S. Moran, MD 2 1 Department of Radiation Oncology, The Cancer Institute of New Jersey, UMDNJ-Robert Wood Johnson Medical School, New Brunswick, NJ 2 Department of Therapeutic Radiology, Yale University School of Medicine, New Haven, CT Presentation No: 1083 Yale School of Medicine

2. Purpose/Methods DCIS has a relatively good prognosis with breast conserving surgery and radiation (CS+RT). Axillary evaluation is not routinely performed for pure DCIS lesions because of the low prevalence of nodal metastases (0-7%). –The role for surgical exploration of the axilla remains unclear for patients with DCIS with microinvasion (DCISM). We assessed the association of age, race, family history, margin status, histology, and receptor status, with DCIS or DCISM in a large cohort of patients treated with CS+RT (n=393). Long-term outcome was evaluated as a function pathology (DCIS vs. DCISM) and axillary evaluation for LRFS, DMFS, and OS. Median patient age at diagnosis = 55.8 years Median follow-up = 12.91 years

3. Extent of Axillary Evaluation DCIS (%)DCISM (%) Total no. of patients32172 No. of patients with SLN biopsies12 (3.78%)4 (5.55%) Positive SLN00 Axillary dissection (median # nodes, range)58 (10.5, 1-23)42 (13, 1-35) Positive axillary dissection0 (0%)1 (1.38%) * Note: All patients that had SLNBx went on to have Axillary dissection (standard clinical practice during this time interval)

4. Patient/Tumor Characteristics Histology Papillary59 (15.0%) Cribiform62 (15.8%) Solid56 (14.2%) Comedo102 (26.0%) Unknown114 (29.0%) Estrogen Receptor Positive10 (2.5%) Negative12 (3.1%) Unknown371 (94.4%) Progesterone Receptor Positive7 (1.8%) Negative11 (2.7%) Unknown375 (95.4%) CharacteristicsNo. (%) Age (years) < 50258 (65.6%)  50 135 (34.4%) Family history Positive136 (34.6%) Negative237 (60.3%) Unknown20 (5.1%) Margin Status Positive98 (24.9%) Negative (within 2mm) 295 (75.1%) CharacteristicsNo. (%) DCISM did NOT correlate with, young age, race, family history, margin status, histology, receptor status, use of hormonal therapy or chemotherapy. (all p > 0.05)

5. Prognostic Factor 10-Year LRFS10-Year DMFS10-Year OS RR (95% CI)p-valueRR (95% CI)p-valueRR (95% CI)p-value Univariate survival analysis Margin Status Positive1.38 (0.55-3.46 0.49 3.49 (0.49-24.85) 0.21 1.38 (0.37-5.10) 0.62 Negative1.00 (Ref.) Histology Comedo0.95 (0.65-1.40) 0.81 0.85 (0.27-2.67) 0.78 1.22 (0.58-2.57) 0.58 Non-comedo1.00 (Ref.) Pathology DCIS1.00 (Ref.) 0.36 1.00 (Ref.) 0.77 1.00 (Ref.) 0.95 DCISM1.58 (0.58-4.30) 0.72 (0.07-6.95) 1.03 (0.28-3.82) Axillary Evaluation Yes1.62 (0.73-3.60) 0.23 3.27 (0.46-23.28) 0.23 0.71 (0.19-2.64) 0.61 No1.00 (Ref.) In UVA, margin status, histology, pathology, and axillary evaluation were NOT independent predictors of LRFS, DMFS, or OS (p > 0.05).

6. Survival Figure 1. 10 year LRFS for DCIS = 90.7% & DCISM = 89.0% (p = 0.36, log-rank test). Figure 2. 10-year DMFS for DCIS = 98.5% & DCISM = 97.9% (p = 0.77, log-rank test). Figure 3. 10-year OS for DCIS = 93.2% & DCISM = 95.7% (p = 0.93, log-rank test).

7. Conclusions –DCISM did NOT correlate with local relapse, young age, race, family history, margin status, histology, or adjuvant therapy (all p> 0.05). –Similar to DCIS, the prevalence of nodal metastases in the DCISM cohort was low (1.38%) in this large dataset. –In UVA & survival analysis, pathology (DCIS vs. DCISM) was NOT an independent predictor of LRFS, DMFS, or OS (p > 0.05). –Surgical evaluation of the axilla was NOT an independent predictor of LRFS, DMFS, or OS in Cox proportional hazards analysis (p > 0.05). –Given the low incidence of loco-regional and distant failures in this large, retrospective dataset, the role for surgical evaluation of the axilla remains unclear. –TAKE HOME POINT: Patients with DCISM may be treated similarly to patients with DCIS. Yale School of Medicine