Breast Cancer Index re-stratifies 21-gene assay risk groups for late distant recurrence and extended endocrine therapy benefit: Results from the BCI Registry Study.

Abstract

529 Background: Patients with hormone receptor-positive (HR+) breast cancer face a prolonged risk for late distant recurrence (DR), even after 5 years of adjuvant endocrine therapy. The BCI (Breast Cancer Index, Biotheranostics) is a genomic assay that provides the risk of late DR and predicts the likelihood of benefit from extended endocrine therapy (EET) based on the H/I (HOXB13/IL-17BR) ratio. The 21-gene RS (Recurrence Score; Oncotype DX, Exact Biosciences) test provides the 9-year risk of DR and the benefit from adjuvant chemotherapy (CT) in HR+/HER2− breast cancer. However, the RxPONDER trial showed that RS was not predictive of CT benefit in premenopausal women with N1 disease. Here, we investigated the concordance between BCI prognostic score and H/I versus RS in HR+/HER2- patients from the BCI Registry study. Methods: The BCI Registry study is a prospective, multi-institutional study to evaluate the long-term clinical outcome, decision impact and medication adherence in early-stage HR+ breast cancer patients. Based on TAILORx and RxPONDER results, RS stratified patients into four risk groups: Age ≤ 50, No CT benefit (RS ≤ 15); Age ≤ 50, CT benefit (RS>15); Age > 50, No CT benefit (RS ≤ 25); and Age > 50, CT benefit (RS > 25). Kendall’s correlation coefficient (R) was used to estimate the correlation between BCI prognostic score, BCI (H/I) and RS as continuous variables. Results: BCI and RS results were reported for 847 patients (76% T1; 59% grade II; 85% N0). H/I classified 559 patients (66%) as H/I-Low and 288 (34%) as H/I-High, respectively, while BCI prognostic risk scores classified 434 patients (51%) as low-risk and 413 (49%) as high-risk for late DR. RS classified 30 patients (4%) as Age ≤ 50, No CT benefit, 38 (5%) as Age ≤ 50, CT benefit, 665 (78%) as Age > 50, No CT benefit and 114 (13%) as Age > 50, CT benefit. When analyzed as continuous variables, BCI H/I and prognostic score showed a weak correlation with RS with R=0.18 and R=0.2, respectively. Based on categorical groups, H/I had low concordance with RS for younger women (Age ≤ 50) but showed some correlation with RS for older women (Age > 50). Among those older than Age 50, H/I identified 31% of those with no CT benefit as likely to benefit from EET while 39% of those with CT benefit as not likely to benefit from EET. Conclusions: BCI H/I and prognostic risk scores exhibited low concordance with RS risk categories. H/I consistently re-stratified RS risk categories into distinct H/I-Low and -High groups, further substantiating that prediction of CT benefit does not equate to prediction of EET benefit. [Table: see text]