Abstract P2-09-05: Risk Group Selection in Primary Breast Cancer According to ASCO Recommended Biomarkers Onkotype DX and uPA/PAI-1: First Experience from Prospective Multicenter WSG Plan B Trial

Abstract

Abstract Background: Both the Oncotype DX multi-gene assay and invasion factors uPA/PAI-1 are guideline-recommended (ASCO, AGO) biomarkers for decision support regarding adjuvant chemotherapy in primary breast cancer (BC). Material and Methods: The West German Study Group (WSG) Plan B trial (planned n=2,448) is evaluating anthracyline-free adjuvant chemotherapy (6x TC) vs. standard 4xEC-4xDOC in HER2- negative BC with 0-3 positive lymph nodes, Oncotype DX is used as selection criterion where pts with RS> 11 are randomized to one of the two chemotherapy arms and pts with RS ≥11 treated with hormonal therapy alone. uPA/PAI-1 (both low vs. either/both high), measured by ELISA, is obtained as an optional risk factor. Results: By June 2010, 1064 patients had been randomized in Plan B (96 recruiting centers). In 153 patients (27 centers), uPA and PAI-1 had been measured; for 131 (84 N0, 47 N+) of these, Oncotype DX Recurrence Score® (RS) results were also available. When considered as continuous variables, RS was weakly positively correlated (Spearman's coefficient) with uPA (rs=0.18, p=0.04) and with PAI-1 (rs=0.23, p=0.01). When considered as risk categories/(Table 1), there was a weak concordance between RS and uPA/PAI-1, using either the standard RS cutoff points (18 and 30) or the TAILORx trial cutoff points (11 and 25). Table T: Association between RS and uPA/PAI-1 categories Assignment of high risk was most strongly concordant between uPA/PAI-1 and RS (RS > 25: 22/25 patients had high uPA/PAI-1; RS >30: 14/15 patients had high uPA/PAI-1). This high-risk concordance extends to N0 patients (RS>25: 17/19 N0 pts had high uPA/PAI-1; RS>30: 11/12 pts had high uPA/PAI-1). Discussion: For the first time, risk groups in primary breast cancer according to both Oncotype DX and uPA/PAI-1 have been compared. These preliminary data show that the high RS group seems highly concordant with the prospectively assessed invasion markers uPA/PAI-1. However, within low and intermediate RS, uPA/PAI-1 could still identify a substantial collective at risk; low uPA/PAI-1 could define a clinically relevant low-risk collective within risk groups that would be classified as “intermediate” according to the multigene assay Oncotype DX. Additional recruitment and outcome assessment of the ongoing multicenter WSG Plan B trial will address the clinical significance of these findings. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P2-09-05.