544 Background: ASCO Tumor Marker Guidelines 2007 recommended clinical routine use of the invasion markers uPA (urokinase-type plasminogen activator) and its inhibitor PAI-1 for risk assessment in N0 breast cancer pts (Harris et al, J Clin Oncol. 2007;25:5287). Ten-year follow-up data of the confirmatory prospective clinical uPA/PAI-1 trial (CHEMO-N0) will be presented at the ASCO 2009 Meeting (Harbeck et al.). We conducted a further prospective trial (NNBC 3-Europe) addressing the direct comparison between risk-assessment by uPA/PAI-1 (UP) and clinico-pathological factors (C), and optimization of adjuvant chemotherapy for high-risk N0 pts: FEC*3-Doc*3 vs. standard FEC*6.
METHODS
From December 2002 to January 2009, 4,040 pts were recruited by 151 centres. For uPA and PAI-1 ELISA, tumors samples were snap frozen und processed. Risk assessment was based on grade and in G2 tumors, either by UP or by a St. Gallen adapted C algorithm. High-risk pts were randomized for adjuvant chemotherapy. All laboratories took part in central quality assurance (QA) for uPA/PAI-1 determination.
RESULTS
Risk assessment was performed for 1,590 pts based on C and for 2,445 pts based on UP. End of accrual was planned when 2,572 high risk pts were randomised for chemotherapy and thus ends Jan 31, 2009. Current status shows 1,291 high-risk pts in the FEC-Doc and 1,294 in the FEC arm. In the low-risk arm, 1,450 pts have endocrine therapy, only. By both risk stratifications, 42% of the G2 tumors were assigned as low risk, but in total, more pts were classified as low-risk by UP (39%) than by C (31%). Median uPA level were at 2.40 ng/mg protein (below cutoff), median PAI-1 at 15.9 (above cutoff). In the QA programm, mean coefficients of variations are 12% for uPA and PAI-1 determination.
CONCLUSIONS
The prospective biomarker trial using uPA/PAI-1 has just ended accrual with more than 4,000 patients. Routine uPA/PAI-1 determination is reliable and feasible even in core needle specimens. Using tumor grade and uPA/PAI-1, adjuvant chemotherapy may be avoided in about 39% of N0 breast cancer pts. Cooperation: EORTC Patho-Biology Group, AGO Breast Group, GBG, Unrestricted grants by Sanofi-Aventis, Pfizer, American Diagnostica, Martin-Luther-Universitaet Halle-Wittenberg. [Table: see text].