Real-world outcomes of neoadjuvant treatment for HER2 positive early-stage breast cancer

e18791 Background: Patients with HER2 positive early-stage breast cancer (ESBC) with residual disease (RD) following neoadjuvant therapy derive benefit from treatment escalation. However, treatment modifications for patients who achieve a complete pathologic response (pCR) remains controversial. The Neo-Bioscore is a validated prognostic tool used for patients with ESBC treated neoadjuvantly. To understand the impact of pCR on outcomes in a real-world setting, and better refine patient selection for treatment escalation or de-escalation, we investigated the use of the Neo-Bioscore and/or other prognostic variables. Methods: Patients diagnosed from 2005 to 2014 with stage I-III HER2 positive breast cancer and treated with neoadjuvant systemic therapy were identified from the BC Cancer Breast Cancer Outcomes Unit database. Outcomes of interest included breast cancer specific survival (BCSS), overall survival (OS), distant disease- free survival (DRFS), and locoregional recurrence free survival (LRFS), and were assessed using the Kaplan-Meier method, comparing the pCR and RD groups. Multivariate analysis was used to assess clinical and pathologic prognostic factors beyond pCR. Results: Of the 417 patients meeting inclusion criteria, median age was 50, including 38.1% with clinical stage 2 and 58.3% with clinical stage 3 disease and 59.5% with hormone receptor (HR) positive tumors. 98.1% received chemotherapy and 92.6% received trastuzumab. Median follow-up was 9.6 years (9.2, 10.0). Overall, 193 patients (46.3%) achieved a pCR. Those with RD had higher clinical T stage (p < 0.01) and more HR positive disease (p < 0.01). Patients who achieved a pCR had significantly better 5-year BCSS (91.2% vs. 74.8%, p< 0.0001), OS (90.7% vs. 69.9%, p< 0.0001), DRFS (87.9% vs. 66.3%, p< 0.0001) and LRFS (98.9% vs. 92.5%, p= 0.007) in univariate analyses. In multivariate analyses, the neo-bioscore (1 vs 5: HR 0.05, 95% CI 0.003,0.666, p = 0.02), higher pathological T stage (ypT3 vs ypT0: HR 7.2, 95% CI 1.3, 40.9, p = 0.03) and pathological N stage (ypN1 vs ypN0: HR 2.2, 95% CI 1.1, 4.5, p = 0.03) were associated with BCSS, whereas patients’ age, ER status and baseline clinical T and N stage were not. Conclusions: In a real-world population, ESBC patients with HER2 positive tumors treated with neoadjuvant systemic therapy who achieved pCR derived significant improvements across all survival parameters compared to those with RD. However, distant relapses still occurred in 12% of patients who achieved pCR. Further analyses to identify patients with pCR who relapsed are underway to determine who may benefit from treatment escalation. The final pathological stage and neo-bioscore for patients with RD were prognostic in a real-world cohort and should be considered for adjuvant treatment decisions. Validation of these results in an independent cohort are also ongoing to better understand predictive and prognostic variables.