Impact of the 21-gene assay on treatment decisions in high-risk patients with ER-positive HER2-negative early breast cancer: Results of the KARMA D...

Date 03 May 2019
Event ESMO Breast Cancer 2019
Session Poster lunch
Topics Personalised/Precision Medicine
Breast Cancer
Presenter Antonio Llombart
Citation Annals of Oncology (2019) 30 (suppl_3): iii1-iii26. 10.1093/annonc/mdz095
Authors A. Llombart1, R. Andrés2, A. Anton Torres3, N. Martinez4, C. Rodriguez5, P. Sánchez-Rovira6, B. Rojas7, M. Ruiz Borrego8, B. Bermejo9, E. Martinez de Dueñas10
  • 1Medical Oncology, Hospital Arnau de Vilanova, 46015 - Valencia/ES
  • 2Medical Oncology, Hospital Lozano Blesa, 50009 - Zaragoza/ES
  • 3Oncology Department, Hospital Miguel Servet, 50009 - Zaragoza/ES
  • 4Medical Oncology, Hospital Universitario Ramon y Cajal, 28031 - Madrid/ES
  • 5Medical Oncology, Hospital Universitario de Salamanca, 37007 - Salamanca/ES
  • 6Medical Oncology, Complejo Hospitalario de Jaen, 23007 - Jaen/ES
  • 7Medical Oncology, Hospitall Vall d´Hebron, 08035 - Barcelona/ES
  • 8Medical Oncology, Hospital Universitario Virgen del Rocio, 41013 - Sevilla/ES
  • 9Medical Oncology, Hospital Clinico Universitario de Valencia, 46010 - Valencia/ES
  • 10Medical Oncology, Hospital Provincial de Castellón, 12002 - Castellón/ES

Abstract

Background

Patients (pts) with ER+, HER2– early breast cancer (EBC) do not universally benefit from adding adjuvant chemotherapy (CT) to hormonal treatment (HT). The 21-gene Oncotype DX Breast Recurrence Score® (RS) assay is uniquely designed and validated to guide CT treatment decisions in this population.

Methods

The primary objective of the KARMA Dx study was to assess the impact of the RS result on treatment recommendations in high-risk patients for whom CT benefit was uncertain. Secondary objective was the evaluation of physicians’ confidence in their recommendation. This study was performed before the recent publication of the TAILORx trial that definitively determined RS cut-points for CT benefit. Eligible EBC pts were ER+, HER2– candidates for CT based on high-risk clinicopathologic features: · Cohort A: grade 3, N0/N1mic · Cohort B: grade 1-2, N1-3, or · Cohort C: cT > 2 cm N0 by imaging (not T4) and planned for neoadjuvant CT Recommendations were captured before and after availability of individual RS results. Primary outcome was the percent change in CT recommendations post- vs. pre-RS results.

Table: 28P

CohortChanges in recommendationRecurrence Score (RS) groupsTotal n
0-1718-3031-100
All patients (n = 209)CT + HT to HT alone99 (86%)39 (53%)2 (10%)140 (67%)
No change16 (14%)35 (47%)18 (90%)69 (33%)
Total n1157420209
A: N0, grade 3 (n = 30)CT + HT to HT alone5 (56%)3 (27%)1 (10%)9 (30%)
No change4 (44%)8 (73%)9 (90%)21 (70%)
Total n9111030
B: N1 (1-3), grade 1-2 (n = 150)CT + HT to HT alone78 (91%)31 (54%)0109 (73%)
No change8 (9%)26 (46%)7 (100%)41 (27%)
Total n86577150
C: Neoadjuvant (n = 29)CT + HT to HT alone16 (80%)5 (83%)1 (33%)22 (76%)
No change4 (20%)1 (17%)2 (67%)7 (24%)
Total n206329

Results

219 pts with median age of 55 years were recruited in 7 Spanish centers and 209 pts were included in the final analysis. Recommendations changed from CT + HT to HT alone for 67% (140) of patients. The table shows changes in recommendations by cohort and RS group. Considering that 85% of patients had RS 0-25, using these TAILORx cut-points might further affect CT reduction. Physicians’ confidence in their final recommendation increased in 34% of cases.

Conclusions

RS testing resulted in an overall 67% reduction in CT recommendation in pts considered candidates for CT based on clinicopathologic criteria. These results reached 76% for patients in cohort C. The findings of the present study indicate the significant potential of the 21-gene assay to spare CT in high-risk EBC pts.

Editorial acknowledgement

Clinical trial identification

Legal entity responsible for the study

Hospital Arnau de Vilanova.

Funding

Genomic Health Inc.

Disclosure

A. Llombart: Grants/research support: Genomic Health Inc.; Stock shareholder: MedSIR. All other authors have declared no conflicts of interest.