Background: HER2 activating mutations occur in 2-5% of metastatic breast cancer (MBC) patients. These mutations cluster in the kinase domains and at amino acids 309-310 in the extracellular domain. The MutHER, SUMMIT, and PlasmaMATCH clinical trials have shown neratinib monotherapy or neratinib plus fulvestrant combination produce clinical benefit in 28% to 46% in HER2-mutated MBC patients, but median progression-free survival was only 3.6 to 5.4 months. In order to improve the knowledge and outcomes for patients with HER2-mutated MBC, we compared the mutational landscape and gene expression of HER2-mutated MBC patients to HER2-amplified and HER2-wild type MBC patients.

Methods: De-identified data from a cohort of stage 4 breast cancer patients (n=1,834) sequenced with the Tempus xT (DNA-seq of 595-648 genes, whole exome-capture RNA-seq) solid tumor assay was retrospectively analyzed. The most recent sample of the patient was used for analyses. Patients were stratified by HER2 mutational status: HER2-wild type (HER2-wt), HER2-amplifications (HER2-amps), or HER2-mutants (HER2-muts). Additionally, a sub-analysis was conducted among HER2-mutants to compare kinase domain mutations to other HER2 mutations. Patient demographic characteristics were compared between groups along with the prevalence of individual gene alterations (pathogenic/likely pathogenic short variants and copy number alterations), adjusted for false discovery.

Results: Within the cohort, 62 (3.4%) patients harbored HER2-muts tumors and 125 (6.8%) patients had HER2-amps tumors. Three patients, whose tumors had both HER2 mutation and amplification, were classified among the HER2-muts. Relative to the HER2-wt cohort (median = 53 yo), HER2-muts patients were older (median = 55 yo) while HER2-amps patients were younger (median = 49 yo) (P<0.001). Significant differences were observed in genomic alterations co-occurring with HER2-muts compared to HER2-wt and HER2-amps, including CDK12 (4.8% vs 0.2% vs 74%), CDH1 (44% vs 11% vs 4.8%), ESR1 (4.8% vs 21% vs 8.8%), TOP2A (3.2% vs 0.5% vs 14%), and ERBB3 (11% vs 0.9% vs 0.8%). Of note, the majority of CDK12 mutations are amplifications (96%). Median HER2 mRNA log10 gene expression differed among the three cohorts (HER2-muts (3.79), HER2-wt (3.56), HER2-amps (4.54); P<0.001). Among HER2-muts patients, 46 (74.2%) patients had HER2 kinase domain mutations (HER2-kinase). Relative to other HER2-muts patients, HER2-kinase patients displayed a lower prevalence of CDH1 mutations (35% vs 69%; P = 0.018). No significant differences were noted
in ERBB3 co-mutations, however all 7 were missense variants (5 of which are p.E928G variants) and occurred among HER2-kinase patients.

Conclusions: Real-world data show increased HER2 mRNA expression in both HER2-mut and HER2-amps MBC patients. Co-occurring genomic alterations were different among all three groups. Notably, ERBB3 and CDH1 alterations co-occurred commonly in HER2-mut MBC patients, while ESR1 alterations co-occurred in only 4.8%. All ERBB3 co-mutations occurred with HER2 kinase domain mutations, while CDH1 co-mutations were less prevalent in the HER2-kinase group. Understanding the frequency and spectrum of HER2 mutations in MBC, as well as co-mutations, will help to guide combination treatment strategies and improve the clinical benefit of targeted therapy in HER2-mutated MBC.

VIEW THE PUBLICATION

VIEW THE POSTER