Differential impacts of Oncotype DX and Mammostrat prognostic indices upon the management of ER+ N0 breast cancer.

Abstract

599 Background: Oncotype DX (Genomic Health, Redwood City CA) and Mammostrat (Clarient, Aliso Viejo CA) are two distinct prognostic measures currently marketed to facilitate adjuvant chemotherapy decision making for ER+ breast cancer patients and their physicians. Both assays define three prognostic strata—favorable, intermediate, and unfavorable. Both assays were also validated in the same retrospective cohorts, but there is significant discordance between these assays, suggesting that assay selection may affect clinical decisions. Methods: We have previously reported that Oncotype DX significantly reduces adjuvant chemotherapy use in 89 consecutive ER+, N0 patients for whom this assay was ordered at our institution. Mammostrat assays were performed on 46 of these cases for which tumor blocks were available. Decision analysis was applied to determine changes in management that would have been most likely if Mammostrat had been substituted for Oncotype DX. Results: Oncotype DX and Mammostrat were concordant for prognostic strata in just 11 (24%) cases. Oncotype DX predicted a more favorable prognosis than Mammostrat in 27 (59%) cases, while Mammostrat predicted a more favorable prognosis in the remaining 8 (17%) cases. As shown in the Table, Oncotype DX reduced chemotherapy utilization whereas Mammostrat would have increased it. Conclusions: Despite being validated in identical patient cohorts, Oncotype DX and Mammostrat are frequently discordant and may often affect adjuvant treatment decisions in opposite directions. [Table: see text]