Proteomic evaluation of HER-2 expression and activation in tumors that are RNA overexpressed, but not DNA amplified.

Abstract

e15093 Background: HER2 evaluation with IHC and FISH has limitations in terms of prediction for response to anti-HER2 therapies. With a growing number of options for Her-2 directed therapy, prediction of response has taken on increasing importance. Next-Generation Sequencing (NGS) tests report on ERBB2 RNA overexpression, but there is a dearth of data to support RNA as being predictive of responses to HER-2 therapy. Proteomic based signatures such as the HER2 Activation Response score (HARPS) have been demonstrated to be a strong predictor of response to a variety of HER2 targeted agents in breast cancer independent of HER-2 IHC/FISH. Methods: Using an institutional molecular tumor board database, we identified tumors from 2018 to 2024 for which the Next-Generation Sequencing (NGS) identified “RNA overexpression”. Using a LMD-RPPA based quantitative proteomics assay, we assessed IHC activity, HER2 activation (phosphorylation at residue Y1248) and HER-2 Activation Response Signature (HARPS) for all samples. Results: 53 tumors were available for Proteomic analysis. 35/53 were DNA-RNA+, 14 were DNA+RNA+, and 4 were DNA+RNA-. 14/14 tumors with DNA+RNA+ had IHC activity3+ or 2+ by quantitative RPPA based HER-2 expression, while among the 17 tumors with the top tertile highest quantitative expression, 14/17 had IHC+ (3+ or 2+) or CN gain >8 copies. We found no correlation with HER2 activation or HARPS with protein HER2 expression. We found 5 patients, all among the 9 samples with Glioblastoma Multiforme (GBM), that were HARPS+ and highly HER2 activated, only one of which having a DNA copy number gain ( 4/5 DNA-RNA+, and 1/5 DNA+RNA+). HER2 RNA expression did not correlate with HER2 FISH, IHC, RPPA HER2, HAPRS, or HER2 phosphorylation. Distribution across the HER2 protein and HER2 protein activation continuum distributed randomly with RNA over-expression not demonstrating a clear trend towards activation compared to the RNA negative samples. Conclusions: Tumor samples with RNA over-expression did not correlate with HER2 activation or an activation signal. The finding of a highly HER2 activated tumors in GBM is worthy of further exploration. Median Her2 phosphorylation score at residue Y1248 for three different categories of tissue: those that were reported on the NGS profile to be both DNA amplified and RNA overexpressed, DNA non-amplified and RNA overexpressed, and neither DNA amplified or RNA expressed. HER-2 DNA/RNA profile (DNA amplification/RNA overexpression) Median Her2 Y1248 phosporylation score DNA+RNA+ (n=14) 152.86 DNA-RNA+ (n=35) 0.00 DNA-RNA- (n=4) 55.42