Genomic Study Uncovers Resistance Mutations in ER+ Breast Cancer

The genomic landscape of recurrent metastatic estrogen receptor–positive breast cancer differed significantly from the mutational profile of primary disease in a study that sheds light on acquired resistance mechanisms to anticancer therapies.

Nikhil Wagle, MD

The genomic landscape of recurrent metastatic estrogen receptor (ER)—positive breast cancer differed significantly from the mutational profile of primary disease in a study that sheds light on acquired resistance mechanisms to anticancer therapies, according to research presented at the 2016 San Antonio Breast Cancer Symposium (SABCS).1

The findings illustrate the importance of conducting molecular testing on metastatic tissue samples, said Ofir Cohen, PhD, in discussing the study at a press conference at SABCS. “The take-home message here is that tumors do evolve and the metastatic setting is different from the primary setting,” said Cohen, a postdoctoral researcher and computational biologist at the Broad Institute of MIT and Harvard and at the Dana-Farber Cancer Institute.

The research has implications for determining subsequent therapies for women with recurrent metastatic disease as well as for designing clinical trials and exploring novel drugs, researchers said.

“These resistant tumors remain the most common cause of breast cancer death, yet mechanisms by which this resistance develops are poorly understood,” senior investigator Nikhil Wagle, MD, deputy director of the Center for Cancer Precision Medicine at Dana-Farber, said in a statement. He also is an associate member at the Broad Institute.

In the study, investigators from the 2 centers conducted sequencing tests on 149 biopsies from patients with metastatic disease, most of whom had received at least 1 ER-targeting therapy consisting of tamoxifen, an aromatase inhibitor, or fulvestrant prior to the biopsy. They also had pretreatment samples from 44 matched primary tumors to compare.

Two types of next-generation testing were utilized: whole-exome sequencing in the tumor and the germline, which consists of the genes that encode all proteins in the cancer cell; and transcriptome sequencing, which analyzes tumor RNA.

Noteworthy results included new information on the prevalence of mutations in ESR1, ERBB2, and RB1. Cohen indicated that the overall prevalence of these mutations in the metastatic samples was 24% for ESR1, 8% for ERBB2, and approximately 6% for RB1. The frequency was higher in metastatic samples for which there were matched primary specimens: 92% for ESR1 (12 of 13 samples), 83% for ERBB2 (5 of 6 samples), and 60% for RB1 (3 of 5 samples).

Cohen indicated that mutations in these genes have been implicated in resistance to therapies—ESR1 to aromatase inhibitors, ERBB2 to drugs that target the estrogen receptor, and RB1 to endocrine therapies and the new class of CDK 4/6 inhibitors. Research into these theories is continuing.

Additional genes found to be significantly mutated in the metastatic samples included TP53, GATA3, PIK3CA, AKT1, and KRAS, according to Cohen’s presentation. Many of these genes have been associated with primary ER-positive breast cancer, although the frequency of the alterations in some instances is higher in the metastatic than the primary setting.

Cohen said that although it is challenging to assess the importance of mutations identified through sequencing, the analysis of the transcriptome in addition to the whole genome is valuable. He noted that tumors may evolve resistance through nongenomic mechanisms such as epigenetics.

“As long as those mechanisms leave a footprint on the transcriptome, the part of the exome and the genome that are expressed, sequencing the transcriptome together with the exome may yield relevant insight into the resistance state that do not derive specifically from exome mutations,” he said.

“Pre-existing events may highlight events that predispose to metastasis, supporting the idea that comprehensive characterization of primary tumors might help predict metastatic potential, while acquired events may suggest novel therapeutic approaches to overcome or prevent resistance and highlight the idea of periodic monitoring with technologies such as cell-free DNA from blood,” Cohen said,

The study presented at SABCS is part of a far-reaching plan to develop a “resistance atlas” for ER-positive metastatic breast cancer. More than 250 biopsies have been conducted thus far at the Dana-Farber Center for Cancer Precision Medicine, Cohen indicated. Patients are then followed with serial blood biopsies every 2 months.

Results of the testing are relayed to physicians for clinical decision making and to researchers to form the basis of further experimental and drug discovery studies.

The work is being funded through federal programs including the National Cancer Institute and by several nonprofit foundations.

Cohen O, Kim D, Oh C, Buendia J, et al. Whole exome and transcriptome sequencing of resistant ER+ metastatic breast cancer. Presented at: 2016 San Antonio Breast Cancer Symposium; December 6-10; San Antonio, TX. Abstract S1-01.

<<<

View more from the 2016 San Antonio Breast Cancer Symposium

With a growing number of tissue samples to study, researchers have been able to distinguish between pre-existing molecular events found in both primary and metastatic tumors and evolutionary, acquired events found only in metastatic specimens, Cohen said in a statement.