Background: Genomic tests can identify ER-positive Her2-negative localized breast cancer (BC) patients (pts) who may not drive any benefit from adjuvant chemotherapy (CT). Several genomic tests have reached a high level of analytical and clinical validity, as well as clinical utility in such situation. Recent results suggest that the safe de–escalation of adjuvant chemotherapy may be most beneficial in pts with clinical high or indetermediate risk, as assessed by classical variables or online tools, through the use of a genomic test. The clinical risk though remains quite uncertain with variable definition and grey zones. The present study aimed at determining if EPclin clinico-genomic test had a significant impact on treatment decision making among pts with predefined intermediate /borderline clinical risk.

Patients and methods: Women were eligible for the present study if they had complete surgical removal of a localized ER+ Her2- pN0 or pN1mi BC, and were considered by the multidisciplinary team meeting (MTM) of the center as being in a "grey zone" of uncertain CT benefit based on classical clinic-pathological assessment. The MTM1 proposed a decision (chemo/no chemo). After informed consent, an EPclin signature classified the tumor as low risk (EPclin Score < 3.3, no chemo advised) or high risk (> 3.3, a theoretical indication for adjuvant chemo). Primary end point was the proportion of change between initial adjuvant CT decision at MTM1 and final administration of CT (yes/no). A 5 steps Fleming design was planned, considering that a change rate of 15% or less was not acceptable (low clinical utility). A one-step design was used (unilateral α = 2.5% and β = 1%).

Results: 203 pts were included, of whom 198 are evaluable for the main end point. 74% of the tumors were grade 2, 72% T1 , 25% T2, 3% T3; 16% were pN1mi and 84% pN0. Median age was 57. EpClin® was low risk in 67% and high risk in 33% of the cases.

The global rate of decision change between MTM1 and final administration of CT was 70/198 (35.4% IC-95% = [28.7-42.1]), with 55 (27.7%) decreases and 15 (7.6%) increases in CT indication. 27% instead of 47% of all pts finally received CT (43% decrease in CT prescription). In the multivariate analysis, the factors associated to decrease for less CT were EPclin score (OR 0.11 [0.03-0.35]), proliferation (OR 1.08 [1.03-1.13]) and higher grades (OR 5.22 [1.04-26.08]); while EPclin score was positively (OR 76.6 [7.11-824.8]), but higher grades inversely (OR 0.06 [0-0.86]) associated to an increase in CT administration.

Of note, the change in final decision occurred after the MTM2 and after report of the results and discussion with the patient in 9 of the 70 cases (12% of changes; 4.5% of the whole population)(8 decreases and 1 increase in CT administration).

Conclusion: Adendom met its primary objective, with 35% of intermediate clinico-pathological risk pts getting significant therapeutic changes upon the receipt of an EPclin gene expression profile. 43% of these "intermediate risk" pts planned for CT avoided it. 12% of the changes were discrepant with the test's results and occurred after discussion with the patient.

Citation Format: Penault-Llorca F, Kwiatkovski F, Grenier J, Levy C, Leheurteur M, Uwer L, Derbel O, Le Rol A, Jacquin J-P, Jouannaud C, Quenel-Tueux N, Girre V, Foa C, Guardiola E, Lortholary A, Catala S, Lemonnier J, Delaloge S. UCBG 2-14: A prospective multicenter non-randomized trial evaluating the effect of EndoPredict® (EPclin®) clinico-genomic test on treatment decision making among patients with intermediate clinical risk [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P2-05-10.