The Role of Mastectomy in De Novo Stage IV Inflammatory Breast Cancer.

INTRODUCTION: The role of modified radical mastectomy (MRM) in patients with de novo stage IV inflammatory breast cancer (IBC) remains controversial. We evaluated the impact of MRM on outcomes in this population. METHODS: Ninety-seven women presenting with stage IV IBC were identified in an institutional database (2007-2016) and were stratified by receipt of MRM or no surgery (non-MRM). Demographic, clinicopathologic, and treatment factors were compared. Local-regional recurrence patterns were described and survival analyses were conducted. RESULTS: All patients initially received chemotherapy. Fifty-two patients (53.6%) underwent MRM; 47 received post-mastectomy radiation. Differences between the non-MRM and MRM groups included tumor receptor subtypes (hormone receptor-positive [HR+]/human epidermal growth factor receptor 2-positive [HER2+]: 4.4% vs. 19.2%; HR+/HER2-negative [HER2-]: 31.1% vs. 44.2%; HR-negative [HR-]/HER2+: 24.4% vs. 15.4%; and HR-/HER2-: 40.0% vs. 21.2%; p = 0.03), number of metastatic sites (3 vs. 2; p = 0.01), and clinical partial/complete response to chemotherapy (13.3% vs. 75.0%; p < 0.001). Of the 47 patients who completed trimodality therapy, 6 (12.8%) had a local-regional recurrence. Median overall survival (OS) was 19 months in the non-MRM group and 58 months in the MRM group (p < 0.001). On multivariable analysis, clinical N3 disease (hazard ratio 2.16, 95% confidence interval [CI] 1.07-4.37; p = 0.03) as well as tumor subtypes HR+/HER2- (hazard ratio 4.98, 95% CI 1.15-21.47; p = 0.03) and HR-/HER2- (hazard ratio 7.18, 95% CI 1.66-31.07; p = 0.008) were associated with decreased OS. Partial/complete response of distant disease to chemotherapy (hazard ratio 0.43, 95% CI 0.24-0.77; p = 0.005) and receipt of MRM (hazard ratio 0.52, 95% CI 0.29-0.93; p = 0.03) were independently associated with improved OS. CONCLUSIONS: In our retrospective study, MRM in de novo stage IV IBC patients is an independent factor associated with improved OS. Our findings strongly support the need for prospective randomized trials evaluating possible survival benefits of MRM in de novo stage IV IBC patients.