S. M. Nazarian1, M. K. Pomponio1, M. M. Goldbach1, C. Huang1, A. D. Williams2, J. Tchou1 1University Of Pennsylvania,Surgery,Philadelphia, PA, USA 2Lankenau Medical Center,Surgery,Wynnewood, PA, USA
Introduction:
Pathologic complete response after neoadjuvant chemotherapy (NAC) has been associated with improved outcomes in clinical trial settings. Whether NAC is associated with improved outcomes in large population studies is unclear. We sought to compare the clinical outcomes of patients with Her2/neu+ (Her2+) breast cancer who did or did not receive NAC at a single institution.
Methods:
A single institution database was queried for all cases of non-metastatic Her2+ breast cancer diagnosed between January 1, 2009 and December 31, 2015. Those with no nodes examined were excluded. Stata/MP 14.2 (College Station, TX) was used for all analyses.
Results:
Her2+ breast cancer was identified in 587 women during the study period, 157 of whom received NAC and 430 who did not. Those who received NAC had distinct clinical characteristics. NAC patients were younger (median 46 years, interquartile range (IQR) 39, 55 versus median 55 years, IQR 46, 63, for no NAC, P = 0.0001). As expected, NAC patients presented with a higher clinical stage (median stage 2 IQR 2,3 compared with median stage 1 IQR 1, 2, P = 0.0001) and their tumors were more often estrogen receptor negative (36.5% versus 27.9% P = 0.044). In terms of treatment, NAC patients were more likely to undergo a mastectomy (65.4% versus 47.2%, P < 0.0001).
With respect to outcomes, NAC patients recurred at higher rates (23.7% versus 9.4%, P < 0.0001) including local recurrence (2.5% versus 0%, P = 0.002) and distant recurrence (21.1% versus 8.1%, P < 0.0001). Nonetheless, patients who responded well to NAC had lower rates of overall recurrence (P = 0.009) and distant recurrence (P = 0.010, see table).
Conclusion:
In this single-institution cohort, Her2+ patients receiving NAC had multiple clinical characteristics suggestive of more aggressive disease and suffered higher rates of recurrence. Our results raise the possibility of confounding by indication, e.g. NAC may be selected partly based on unrecognized provider bias. This highlights the need for better risk stratification in these high-risk patients and warrants further study.
