BreastPRS Effectively Separates OncotypeDX Intermediate Risk Group to Low and High Risk Groups
Timothy M D'Alfonso, Ryan K van Laar, Rachel Flinchum, Nathan Brown, Linda Saint John, Sandra J Shin. Weill Cornell Medical College, New York, NY; Signal Genetics, LLC, New York, NY
Background: Multi-gene prognostic tools resulting in patients classified as 'intermediate risk' are of limited clinical utility, as the benefit of chemotherapy in patients in this risk group is unclear. BreastPRS (Signal Genetics) is a new molecular characterization assay developed from a meta-analysis of publically available genomic datasets. The assay utilizes three unique yet complementary gene signatures to determine molecular grade (103 genes), molecular subtype (280 genes) and risk of recurrence (200 genes). BreastPRS is a binary assay that could potentially re-classify intermediate risk patients into more meaningful prognostic groups. We sought to re-classify Oncotype Dx (ODx)(Genomic Health) intermediate risk cases using BreastPRS.
Design: 307 whole genome expression profiles were generated from formalin-fixed paraffin embedded tissue of patients' tumors with known ODx recurrence scores. A set of 120 genes was found to hierarchically cluster patients according to ODx risk group. This gene set was used to interrogate data from 522 untreated ER+ breast cancer patients obtained from public genomic data repositories, with associated outcome data. Those predicted to be intermediate risk according to the ODx-estimation (ODx-est) algorithm were further analyzed using BreastPRS. The recurrence free survival differences between resulting high and low risk predictions were compared using Kaplan Meier analysis and multivariate cox proportional hazards regression.
Results: The risk groups by ODx-est and ODx were concordant in 72% of cases (Chi Square P<0.001). Using ODx-est and BreastPRS, 522 ER+ untreated patients were classified into high/intermediate/low and high/low risk groups, respectively, with a high degree of statistical significance in outcome (15-year recurrence free survival, P<0.0001). The 124 patients predicted to be intermediate risk by ODx-est were then re-classified as high or low risk using BreastPRS. The reclassified risk groups exhibited differences in recurrence free survival that were statistically significant (P=0.0021, Hazard ratio: 4.38, 95% CI: 2.24 to 8.57). The significance of this stratification was independent when adjusted for age at diagnosis and tumor grade (P=0.025).
Conclusions: The data strongly suggest that BreastPRS can reclassify intermediate risk patients by ODx into two groups with significant differences in recurrence-free survival.
Wednesday, March 6, 2013 9:30 AM
Poster Session V # 37, Wednesday Morning