Biomarker analyses indicate which advanced breast cancer patients are most likely to benefit from CDK4/6 inhibitors
- Biomarker analyses were performed on 568 tissue specimens from patients with advanced breast cancer (ABC)
- Oestrogen receptor positive (ER+) patients were more susceptible to treatment with cyclin-dependent kinase (CDK) 4 and 6 inhibitors
- There was not a subgroup of ER+ patients identified who did not benefit from palbociclib with letrozole (PAL + L)
Recent analytical data were presented at this year’s European Society for Medical Oncology Congress, providing an insight into which ABC patients benefit the most from CDK 4/6 inhibitors. The results showed that ER+ patients were more responsive to the treatment, explained Dr Richard Finn, Associate Professor of Haematology and Oncology at the Geffen School of Medicine, UCLA and the Jonsson Comprehensive Cancer Center, US.
In early 2015, the PALOMA-1 phase I/II study led to the accelerated approval of palbociclib by the Food and Drugs Administration (FDA), contingent on the confirmation of benefit in further studies. For this reason, the phase III trial PALOMA-2 was devised. The study included 666 postmenopausal, ER+ or HER2 negative patients, randomised 2:1 to PAL + L or placebo + letrozole (P + L). Patients were endocrine-therapy naive.
“As it stands today, the best predictive marker for women who will get a benefit from a CDK 4/6 inhibitor is ER+.”
Dr Richard Finn, Associate Professor of Haematology and Oncology, Geffen School of Medicine at UCLA and the Jonsson Comprehensive Cancer Center, Los Angeles, US
Results showed a significant increase in median progression-free survival (PFS); 24.8 months with PAL + L versus 14.5 months with P + L (hazard ratio (HR) 0.58, P<0.001). The work also suggested that ER+ patients were more susceptible to treatment with CDK 4/6 inhibitors. For this reason, biomarker analyses were performed to confirm the observations.
Tissue samples from the original diagnostic or metastatic tumours were collected. “In this study we went back and looked at centralised review of ER staining as well as several proteins involved in the cyclin D, CDK pathway,” Dr Finn explained. Immunohistochemistry tests were used to identify the presence of ER, retinoblastoma protein (Rb), p16 and cyclin D1. Kinase 67 (Ki-67) was also measured as a proliferative marker.
Qualitative results were produced (ie positive or negative for markers) and a quantitative result was developed by using an H-score. Positivity was defined by an H-score of ≥1.
85% (n=568) of patients from the PALOMA-2 study had tumour tissue available, 566 of whom were evaluable for detecting ER. A central review confirmed that 89% of patients were ER+. ER median H-scores were 120 for those receiving PAL + L (interquartile range (IQR) 45-170) and 110 for those receiving P + L (IQR 38-158). PFS improvement with PAL + L was observed in all ER quartiles; all ER+ patients responded to PAL + L regardless of the level of ER protein found in the tumour.
Effect of biomarkers on efficacy
| Biomarker | Median PFS (months), 95% confidence interval (CI) | HR
(95% CI) |
P value | |
| PAL + L | P + L | |||
| ER+ | 24.9 (n=338) | 16.3 (n=166) | 0.57 | <0.0001 |
| ER- | 15.6 (n=40) | 5.4 (n=22) | 0.41 | 0.0030 |
| Rb+ | 24.2 (n=345) | 13.7 (n=167) | 0.53 | <0.0001 |
| Rb- | NE (n=29) | 18.5 (n=22) | 0.68 | 0.1619 |
| Cyclin D1+ | 24.8 (n=370) | 13.8 (n=179) | 0.56 | <0.0001 |
| Cyclin D1- | 11.1 (n=5) | 8.1 (n=10) | 1.0 | 0.9964 |
| p16+ | 24.8 (n=305) | 13.8 (n=161) | 0.52 | <0.0001 |
| p16- | 16.8 (n=59) | 13.8 (n=25) | 0.73 | 0.3221 |
| Ki-67 ≤15% | NE (n=179) | 19.2 (n=75) | 0.54 | 0.0015 |
| Ki-67 >15% | 19.2 (n=189) | 11.0 (n=110) | 0.60 | 0.0006 |
A trend towards a PAL + L benefit for p16- patients was noted, however, there were too few specimens to draw conclusions. For patients that were cyclin D1+ (97%), benefit did not vary with H-score and Ki-67 values did not show a patient group to have better or worse PFS with PAL + L.
“I think the take home-message I would like to leave people with after the biomarker work is, as it stands today, the best predictive marker for women who will get a benefit from a CDK 4/6 inhibitor is ER+,” Dr Finn summarises.
Studies are still ongoing in order to further analyse these (and other) biomarkers using various methods. They aim to look at cyclin D1 amplification and p16, as well as analysing expression of CDK 2/4/6, cyclin E1/2, Rb and p16. Studies are also looking into the analysis of HER2 using fluorescent in situ hybridisation (FISH) and into progesterone receptor expression by immunohistochemistry.
The research builds on a wealth of existing evidence suggesting that biomarkers are useful for providing care that is more targeted and, most likely, more effective.
Based on Finn R, Jiang I et al. Biomarker analyses from the phase 3 PALOMA-2 trial of palbociclib (P) with letrozole (L) compared with placebo (PLB) plus L in postmenopausal women with ER+/HER2– advanced breast cancer (ABC) (LBA15). Presented on Saturday 8 October 2016.
[button link=”https://medonline.at/infocenter-esmo/” color=”blue” target=”blank” size=”large”]<< Back to Infocenter ESMO 2016[/button]