Abstract 61P
Background
Gene expression signatures (GES) have emerged to predict prognosis and guide the use of adjuvant chemotherapy (CT) in patients with hormone receptor-positive HER2-negative (HR+/HER2-) early breast cancer (eBC). However, the predictive value of the currently available GES to novel anti-cancer drugs used, such as CDK4/6 inhibitors and PARP inhibitors, remains unknown.
Methods
We applied four commercially available prognostic GES (Oncotype Dx Recurrence Score (RS), PAM50-based Prosigna (ROR), MammaPrint 70-gene (Prog70), and EndoPredict (EPclin)) to 5,391 HR+/HER2- eBC patients with pN0 or pN1 disease. We then assessed in each genomic risk category several multigene signatures predicting sensitivity/vulnerability to endocrine therapy (ET; E2F4), CDK4/6 inhibitors (E2Fregulon & RBsig), PARP inhibitors (HRD), and chemotherapy (CT; DLDA30) at individual patient’s level.
Results
Vulnerability signatures consistently (p<0.001, Fisher’s exact test) identified genomic high-risk patients as more vulnerable to CT and PARP inhibition, but less sensitive to ET and CDK4/6 inhibition (Table). Oppositely, genomic low-risk patients were predicted as strongly vulnerable to ET and CDK4/6 inhibition, but less sensitive to CT and PARP inhibition. For example, the ROR-P “high-risk” patients included more patients predicted as sensitive to CT and PARP inhibitors than the ROR-P “low-risk” patients (17% vs 4%, and 17% vs 3% respectively), and included less patients predicted as sensitive to ET and CDK4/6i (22% vs 88%, and 20% vs 90% according to the E2F4 signature, respectively).
Table: 61P
| Prediction | ROR-P | RS | EPclin | Prog70 | |||||
| Low/int | high | low | high | low | high | low | high | ||
| DLDA30 (CT) | |||||||||
| resistant | 96% | 83% | 99% | 78% | 99% | 84% | 100% | 86% | |
| sensitive | 4% | 17% | 1% | 22% | 1% | 16% | 0% | 14% | |
| E2F4 (ET) | |||||||||
| resistant | 12% | 78% | 26% | 62% | 15% | 58% | 5% | 57% | |
| sensitive | 88% | 22% | 74% | 38% | 85% | 42% | 95% | 43% | |
| Rbsig (CDK4/6i) | |||||||||
| resistant | 10% | 80% | 25% | 64% | 12% | 59% | 3% | 58% | |
| sensitive | 90% | 20% | 75% | 36% | 88% | 41% | 97% | 42% | |
| E2Fregulon (CDK4/6i) | |||||||||
| resistant | 12% | 80% | 27% | 62% | 17% | 58% | 6% | 57% | |
| sensitive | 88% | 20% | 73% | 38% | 83% | 42% | 94% | 43% | |
| HRD (PARPi) | |||||||||
| resistant | 97% | 83% | 96% | 84% | 98% | 87% | 98% | 87% | |
| sensitive | 3% | 17% | 4% | 16% | 2% | 13% | 2% | 13% |
Conclusions
Our analysis provides the first vulnerability prediction frame of genomic risk patients assessed by routinely used GES. This underlines the need of caution in interpreting the results of active clinical trials recruiting patients according to genomic risk for the assessment of novel anti-cancer drugs in HR+/HER2- eBC such as NATALEE trial.
Legal entity responsible for the study
Institut Paoli-Calmettes.
Funding
Has not received any funding.
Disclosure
A. De Nonneville: Non-Financial Interests, Advisory Board: Gilead, Daiichi Sankyo, Seagen, Lilly, Novartis. A. Gonçalves: Financial Interests, Institutional, Invited Speaker: Novartis, Roche, MSD, AstraZeneca, Daiichi Sankyo. All other authors have declared no conflicts of interest.