Clonal evolution of breast cancer treated with palbociclib and fulvestrant: a circulating tumour DNA analysis of the PALOMA-3 trial


Year:

Session type:

Theme:

Ben O'Leary1,Ros Cutts1,Yuan Liu2,Sarah Hrebien1,Xin Huang2,Matthew Beaney1,Kerry Fenwick1,Fabrice Andre3,Sibylle Loibl4,Sherene Loi5,Isaac Garcia-Murillas1,Cynthia Huang Bartlett2,Massimo Cristofanilli6,Nicholas Turner7
1The Institute of Cancer Research,2Pfizer,3Institut Gustave Roussy, France,4German Breast Group,5University of Melbourne, Peter MacCallum Cancer Centre,6Robert H Lurie Comprehensive Cancer Centre, Feinberg School of Medicine,Chicago,7The Institute of Cancer Research and The Royal Marsden Hospital

Abstract

Background

Combination treatment using a CDK4/6 inhibitor, such as palbociclib, with endocrine therapy is now a standard of care for advanced hormone receptor positive (ER+) breast cancer. Little is known about mechanisms of resistance to these combinations in breast cancer patients, or how resistance evolves.

Method

Paired sequencing of the circulating tumour DNA in plasma collected before and after treatment was conducted for 193 patients in the PALOMA-3 trial, which randomised 521 patients between palbociclib and fulvestrant or fulvestrant plus placebo. Paired exome sequencing was performed for 14 patients who had received palbociclib and fulvestrant. Comparisons were made between the mutation profiles and clonal status of patients before and after treatment.

Results

Longer time on palbociclib and fulvestrant was observed for patients who had detectable new acquired mutation, as compared to patients who did not acquire any new mutation (p=0.0066) implicating acquired mutations in the evolution of resistance. Subclonal evolution was confirmed in 12/14 patients with exome sequencing data. The largest difference between baseline and end of treatment was seen in the acquisition of PIK3CA mutations (p=0.00018) irrespective of treatment arm. Different ESR1 mutations were commonly observed to be both lost and acquired during treatment within both treatment arms, though there was strong evidence for Y537S variant selection (p=0.006). Six patients acquired RB1 mutations, all of these being in the palbociclib-treated group (4.8%, 6/125, p=0.041).

Conclusion

The ESR1 mutation Y537S is selected on endocrine treatment with both palbociclib and fulvestrant and fulvestrant alone, implicating it as a resistance mechanism to fulvestrant. Together with the observation that acquisition of PIK3CA mutations is also common in both treatment arms, the data suggest that mechanisms of resistance to each element of the CDK4/6 inhibitor/endocrine combination may evolve separately. Furthermore, RB1 mutations may confer resistance to palbociclib, but were only seen in a small number of patients (~5%).