187P - Summary of head-to-head comparisons of patient (pt) risk classifications by the 21-gene recurrence score (RS) assay and other genomic assays for early breast cancer (EBC)

Abstract

Background: Many genomic assays that assess recurrence risk in EBC are prognostic, but they differ in risk group stratification, which can affect clinical utility. Prospective outcomes of > 50K pts treated based on 21-gene RS results have shown that pts with low RS EBC can safely forgo chemotherapy. Because of its extensive validation and wide clinical use, the RS assay is a common comparator in head-to-head studies with other assays. Methods: Published/presented studies of the RS assay performed on same tumor samples with Breast Cancer Index (BCI), EndoPredict (EP) or EP+clinical features (EPclin), MammaPrint (MMP), and/or Prosigna (ROR) assays were reviewed. Study findings were summarized descriptively. Results: 14 studies were found that compared the RS assay with BCI (1), BCI, EPclin, and ROR (1), EP/EPclin (2), MMP (6), and ROR (4). Overall discordance in risk stratification ranged from 43% to 66% between assays (Table). The RS assay classifies 12% of pts as high risk, compared with EP (63%), EPclin (48%), and MMP (46%), assays with low/high risk groups, and compared with BCI (16%) and ROR (33%), assays that, like the RS assay, use three risk groups.Table187PDiscordancea Between the RS Assay and Other AssaysBCIROREP/EPclinMMPStudyb1-level2-levelOverall1-level2-levelOverall1-level2-levelOverall1-level2-levelOverallSestak 201637%5%42%Bartlett 2016c40%10%50%Alvarado 201537%10%46%Dowsett 201341%3%43%Sinn 201745%20%66%Varga 201329%/29%18%/21%47%/50%Clough 201338%19%57%Denduluri 201134%25%58%Maroun 201531%22%53%Shivers 201326%19%44%a. Overall=any discordance in risk classification between the RS assay and other; 1-level=discordance of one risk category (low ↔ intermediate or intermediate ↔ high); 2-level=discordance of two risk categories (low ↔ high). b. Four studies lacked risk classification information appropriate for inclusion in this table. c. Study used nonstandard RS cutoffs for the RS vs. MMP comparison. Open table in a new tab a. Overall=any discordance in risk classification between the RS assay and other; 1-level=discordance of one risk category (low ↔ intermediate or intermediate ↔ high); 2-level=discordance of two risk categories (low ↔ high). b. Four studies lacked risk classification information appropriate for inclusion in this table. c. Study used nonstandard RS cutoffs for the RS vs. MMP comparison. Conclusions: The five most common genomic assays in clinical use for EBC risk-stratify pts differently and thus are not interchangeable. Of these, the RS assay classifies the smallest proportion of pts as high risk. Legal entity responsible for the study: Zsuzsanna Varga Funding: Genomic Health Disclosure: Z. Varga: Consultant/advisor: Genomic Health, Roche. P. Sinn: Advisor: Genomic Health. D. McCullough, A. Lau, M.C. Stöppler, F.L. Baehner, C. Chao: Employment and stock ownership: Genomic Health. A. Seidman: Consultant/speaker: Genomic Health.