The clinical impact of substituting multigene assays: Comparative clinical analysis of PAM50-ROR versus 21-gene assay

Date 03 May 2019
Event ESMO Breast Cancer 2019
Session Poster lunch
Topics Breast Cancer
Translational Research
Presenter Christos Markopoulos
Citation Annals of Oncology (2019) 30 (suppl_3): iii1-iii26. 10.1093/annonc/mdz095
Authors C.J. Markopoulos1, L. Hochheiser2, J. Hornberger3, E. Hytopoulos4, M.C. Stoppler5, M. Gitlin6
  • 1Surgery, National & Kapodistrian University of Athens, 11521 - Athens/GR
  • 2Medicine, University of Vermont, Burlington/US
  • 3Oncology, Genomic Health Inc., 94063 - Redwood City/US
  • 4Biostatistics, Genomic Health, Redwood City/US
  • 5Medicine, Genomic Health Inc, 94063 - Redwood City/US
  • 6Outcomes Research, BluePath Solutions, Los Angeles/US



The National Cooperative Cancer Network Guidelines (v3.2018, Oct 2018) state that 21-gene assay (Oncotype DX Breast Recurrence Score®, Genomic Health, Inc.) is the preferred multigene assay (MGA) to predict chemotherapy (CT) benefit for early stage breast cancer (ESBC). We assessed the potential effects of substituting another MGA, PAM50-ROR (Prosigna®, NanoString), for the 21-gene assay on clinical outcomes.


We conducted a decision analysis using Markov Chain Monte Carlo simulation (rjags package in R) to estimate posterior distributions in a cohort of 100,000 women with ESBC. The discordance of PAM50-ROR with the 21-gene Recurrence Score® result threshold for beneficial effects of CT (Sparano NEJM 2018) was extracted from the only three published discordance studies. Effects of adjuvant CT on distant recurrence, overall survival, and toxicities were based on Chandler et al. (JCO 2018) outcomes over 25-year time horizon. Primary endpoints are incidence of distant recurrence, incidence of chemotoxicities, and overall survival.


About 1 in 4 women (25%; IQR 16 - 28) tested with PAM50-ROR have results discordant with 21-gene assay, resulting in an average increase of 2,466 of 100,000 women tested receiving CT (IQR -1,647 – 9,702). PAM50-ROR classifies 7% of women as high risk who have RS ≤ 25 (with chemotoxicities and no chemobenefit), 11% as low/intermediate risk who have RS > 25 (with increased distant recurrence and decreased survival among those not treated with CT), and 25% as intermediate who have RS ≤ 25, of whom >7% are likely to receive CT. The consequent effect of treatment choices based on discordant test is an increased incidence of distant recurrences (3,052; IQR 1,212 – 3,319), increased incidence of toxicities (696; IQR -324 – 2,228), and decreased years of overall survival (34,255 y; IQR 16,586 – 38,118).


Use of PAM50 ROR compared with the 21-gene assay may lead to an increase in CT recommendations and potentially less desirable clinical consequences. Implementation of new incentives to create value-based oncology care needs safeguards to focus use towards optimal interventions.

Editorial acknowledgement

Clinical trial identification

Legal entity responsible for the study

Genomic Health, Inc.


Genomic Health, Inc.


C.J. Markopoulos: Consultant: Genomic Health Inc.; Educational grants: AstraZeneca, Pfizer, Novartis. L. Hochheiser: Consultant: Abbott (providing guidance and direction as part of a Payor Advisory Board); Manuscript consultant: Genomic Health, Inc. J. Hornberger, E. Hytopoulos, M.C. Stoppler: Employee, owns stock: Genomic Health, Inc. M. Gitlin: Consultant: Genomic Health.