ESR1 Mutations in Cell-Free DNA of Breast Cancer: Predictive “Tip of the Iceberg”

In this issue of JAMA Oncology, Chandarlapaty and colleagues¹ report a correlative analysis of the 2 most frequent hormone-binding domain estrogen receptor (ESR1) mutations (Y537S and D538G) in hormone receptor–positive, HER2-negative advanced breast cancer using archival tumor specimens from the BOLERO-2 trial² to identify correlations between these mutations and efficacy of the mammalian target of rapamycin complex 1 (mTORC1) inhibitor everolimus. The BOLERO-2 trial² was a double-blind, placebo-controlled, phase 3 trial of patients who recurred and/or progressed during therapy with nonsteroidal aromatase inhibitors (AIs), who were randomized to everolimus plus the steroidal AI exemestane or exemestane only, and the trial demonstrated that everolimus plus exemestane substantially improved progression-free survival (PFS). Using next-generation sequencing, approximately 5% of evaluable samples (most of those analyzed were primary tumors) from patients in BOLER0-2, compared with 2% of primary tumors from The Cancer Genome Atlas (TCGA) study, contained ESR1 mutations.³ The Chandarlapaty et al study¹ reports a higher ESR1 alteration frequency (28%) using digital drop polymerase chain reaction (ddPCR) mutation detection of cell-free (cf) DNA for specific ESR1 mutations in these patients with advanced breast cancer, and is consistent with the overall reported frequency of ESR1 mutations in about 20% of metastatic breast cancers.⁴ The use of sensitive detection methods, such as mutation-specific ddPCR, has transformed our understanding of the genomic basis of breast tumors, and has revealed the extreme genomic complexity and heterogeneity of metastatic breast cancer. In the BOLERO-2 study, overall survival (OS) was 32 months for patients with wild-type ESR1 vs 20.73 months with Y537S D538G mutations. The reported shorter OS for patients harboring ESR1 mutations suggests that they may represent potential poor prognostic factors in breast cancer; however, it must be stressed that without a no-treatment control arm, significant clinical conclusions about whether mutations were predictive of everolimus response cannot be determined.