Abstract GS4-11: How low is low risk: MINDACT updated outcome and treatment benefit in patients considered clinical low risk and stratified by genomic signature, age and nodal status
Background With 8.7 years follow-up, the prospective phase III randomized MINDACT trial (EORTC 10041/BIG3-04) continues to meet its primary objective, i.e. 95.1% (95%CI 93.1-96.6), 5-year distant metastasis-free survival (DMFS) in clinical high (C-High)/genomic low (G-Low) risk patients who did not receive adjuvant chemotherapy (ACT) (Cardoso et al., ASCO 2020). In addition, about half of the MINDACT patients had a low clinical risk (C-Low) defined by pre-specified clinical-pathological characteristics. Here, we evaluated the outcome of this C-Low population stratified by the 70-gene signature (MammaPrint®) (G-Low or G-High) for outcome considering age, and present data on the total G-low population (C-Low and C-High combined). Methods Of 6693 patients enrolled in the MINDACT trial between 2007 and 2011, 3337 were C-Low, characterized as mainly T1, grade 1 or 2, and node negative. We evaluated the pre-specified DMFS, distant metastasis free interval (DMFI), and overall survival (OS) rates at 5 and 8 years in the C-Low population: i) in patients with genomic low risk (C-Low/G-Low, n=2744) who were recommended to receive endocrine therapy only (for 99% HR+), and ii) in C-Low/G-High who received ACT or not following randomization (ITT, n=690, 81% HR+). Exploratory analyses by age, ≤50 and >50, were conducted for ACT vs no ACT received in C-Low/G-High. In parallel we estimated survival rates for all G-low patients if all would have followed the genomic low risk assignment and received no ACT (C-Low/G-Low, and C-High/G-Low randomized to no ACT double weighted, n=4130). We used Kaplan-Meier estimates for time to event endpoints and hazard ratios with 95%CI from Cox-regression models adjusted for stratification factors used for the randomization. Results C-low/G-low patients who were recommended endocrine therapy only (compliance > 79%, based on local guidelines) have excellent 5 and 8 year survival rates for all endpoints (Table 1). The estimated survival rates for all G-Low patients, if all would have followed the genomic low risk assignment and received no ACT, is excellent as well (Table 1), albeit this population includes both C-Low and C-High patients. The survival estimates for C-Low/G-High patients are for all endpoints a few percentage points lower than for the C-Low/G-Low group (Table 1). At 8 years of follow-up, in the relatively small subset of 690 patients with C-Low/G-High tumors assigned to ACT or not by randomization (ITT), a 1.5% (SE ±2.3%) higher DMFS is seen in the ACT group, and a 2.9% (SE ±2.0%) higher DMFI. This suggested benefit is mostly seen in patients under 50 years of age (absolute Δ in DMFS for ACT vs no ACT at 8 years: 5.4% for age ≤50 vs -0.3% for age >50). Conclusion Patients with a 70-gene G-Low risk tumor have an excellent 8 year outcome in the context of C-Low characteristics when recommended for endocrine therapy only, very close to the outcome in the larger group of all G-Low patients regardless of clinical risk. Stratification of C-Low patients in to G-Low and G-high provides meaningful information. The benefit of ACT in C-Low patients with a 70-gene G-High risk tumor needs further confirmation, especially relevant in younger women. Citation Format: Laura J van 't Veer, Fatima Cardoso, Coralie Poncet, Josephine Lopes Cardozo, Suzette Delaloge, Jean-Yves Pierga, Peter Vuylsteke, Etienne Brain, Giuseppe Viale, Sherko Kümmel, Isabel T Rubio, Gabriele Zoppoli, Alistair Thompson, Erika Matos, Khalil Zaman, Susan Knox, Florentine Hilbers, Aleksandra Peric, Bart Meulemans, Martine Picccart, Emiel J Th Rutgers. How low is low risk: MINDACT updated outcome and treatment benefit in patients considered clinical low risk and stratified by genomic signature, age and nodal status [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr GS4-11.