Comparison of the prognostic performance between OncoMasTR and OncotypeDX multigene signatures in hormone receptor-positive, HER2-negative, lymph node-negative breast cancer.

12074 Background: Multigene prognostic signatures (MGPS) enable identification of early stage breast cancer (BC) patients requiring less aggressive treatment. OncoMasTR is a new MGPS found via a novel transcriptional network analysis method that identified genes – Master Transcription Regulators (MTRs) – that regulate previously identified prognostic biomarkers. The optimised OncoMasTR signature consisting of just 3 MTRs (OM) and incorporating clinicopathological information (OMclinical) has been clinically validated. We examined OncoMasTR’s prognostic performance alone and in comparison to OncotypeDX. Methods: We measured MTR expression levels by RT-qPCR in tissue from Irish patients (n = 367) enrolled in the Trial Assigning Individualized Options for Treatment (TAILORx) study through Cancer Trials Ireland. OM and OMclinical numeric risk scores and risk category (high or low) were blindly calculated using the OncoMasTR algorithm. OncoMasTR scores and OncotypeDX recurrence scores (RS) were independently compared on measures of risk classification, BC recurrence (all, local and distant), correlation and concordance over 9 years of follow-up. Results: OM (LRχ2= 12.92, p = 0.0003) and OMclinical (LRχ2= 17.21, p < 0.0001) provided more prognostic information than RS (LRχ2= 4.74, p = 0.0294). OM classified 39% of samples as low risk (no local or distant recurrence, 0.0% recurrence) and 61% as high risk (9.0% recurrence; 4.5% local, 4.5% distant). OncotypeDX classified 17.3% of samples as low risk (1.9% recurrence; 1.9% local, 0% distant), 63.3% as intermediate risk (7.0% recurrence; 3.5% local, 3.5% distant), and 19.4% as high risk (8.0% recurrence; 3.2% local, 4.8% distant). There was moderate correlation between OM and RS numeric risk scores (r = 0.46, p < 0.0001) and low concordance between the OM and RS risk categories. Higher correlation and concordance were observed among high risk samples. Conclusions: OM, OMclinical and RS were significantly prognostic for recurrence in TAILORx samples. OM and OMclinical demonstrated exceptional sensitivity by classifying all patients who had any recurrence to the high risk category.