Use of next-generation sequencing (NGS) to detect high frequency of targetable alterations in primary and metastatic breast cancer (MBC).

Abstract

10559 Background: The aim of this study was to assess the frequency of genomic alterations in breast cancer potentially treatable with approved targeted agents or investigational drugs in clinical trials. NGS was performed in a CLIA setting (Foundation Medicine). Methods: DNA was extracted from needle biopsies of 33 pre-therapy primary and 17 MBCs (mean age 52 yrs; 58% ER+, 20% HER2+, 30% triple negative) obtained prospectively for biomarker discovery and preserved in RNAlater. Patients with MBC received an average of 7 drugs (range 5-17) including adjuvant therapy before biopsy for this research; 13 biopsies were from soft tissues, 3 from liver and 1 from bone. Sequencing was targeted to 3230 exons in 182 cancer-related genes and 37 introns in 14 genes often rearranged in cancer. Average median depth was >1200x. Results: All biopsies yielded sufficient DNA. NGS revealed a total of 117 known driver mutations across 36 genes (per-tumor average=2.5, range 1-6), including 37 base substitutions (32%), 28 indels (24%), 42 amplifications (36%) and 10 homozygous deletions (9%). NGS identified HER2 gene amplification in 6/7 cases scored HER2+ by FISH. The average number of functionally important alterations was surprisingly similar, 2.3 in primaries vs 2.8 in heavily treated MBCs (p=0.32). Remarkably, 25/33 (76%) of primary and 14/17 (82%) of MBCs had at least 1 genomic alteration targetable with an FDA approved drug or novel agent in clinical trials. These included: ERBB2 alterations (n=9), PIK3CA mutations (n=8), NF1 mutations (n=4, candidate for PI3K/MEK inhibitors), AKT1-3 mutations (n=5, PI3K inhibitors), BRCA1/2, (n=6, PARP inhibitors), and CCND2 (n=3)/CDKN2A (n=3) mutations (CDK inhibitors). Numerous other alterations with less apparent therapeutic implications were also observed. Conclusions: Comprehensive NGS profiling in breast cancer needle biopsies showed high frequency of genomic alterations linked to a clinical treatment option or clinical trials of targeted therapies. These results demonstrate it is feasible to use NGS to guide targeted therapy. Prospective testing of the diagnostic/predictive value of this patient selection approach is currently under way.