No abstract

Despite the great improvement of patient outcomes by trastuzumab, a monoclonal antibody targeted on HER2-positive breast cancer, approximately 23% of patients with early-stage disease treated with adjuvant trastuzumab either fail to respond or experience recurrence within 10 years, highlighting the importance of identifying which HER2-positive patients would benefit from trastuzumab upfront. Efforts to identify biomarkers predictive of response to trastuzumab in initial breast tumor core biopsies have been complicated by the clinical and biological heterogeneity of HER2-positive tumors. Therefore, we identified a trastuzumab-resistant (TrR) signature that accurately predicts response to trastuzumab quantitively and qualitatively in vitro and in vivo, via repurposing transcriptome profiles in an engineered cell line model. We additionally demonstrated that our TrR signature was associated with tumor progression and capable of stratifying patient prognosis. Our study further illustrated the possible mechanism of this resistance as being less inherited cytotoxic T cell infiltration and failure to secrete Interferon-γ upon trastuzumab treatment in TrR tumors. These findings highlight the potential clinical application of TrR signature in treatment management and identifying possible immunotherapy interventions.

Background/Objective: Breast carcinomas with HER2 receptor amplification and overexpression account for approximately 20% of all breast cancers, which also confer more aggressive phenotypes and are associated with poor prognosis. Up to 23% of patients with early-stage HER2-positivie (HER2+) breast cancer treated with adjuvant chemotherapy and trastuzumab experience disease recurrence within 10 years, highlighting the importance of identifying which HER2+ patients respond to this treatment and those that do not. Efforts to identify biomarkers predictive of response in the initial breast tumor biopsy to trastuzumab have been complicated by the clinical and biological heterogeneity of HER2+ tumors. Therefore, we aim to develop a trastuzumab-resistant (TRR) signature which could faithfully predict patients' response to this treatment using transcriptome profiles in engineered cell line models. We further investigated the possible mechanisms mediating TRR by bioinformatics and biochemical approaches in order to reveal promising new therapeutic targets TRR HER2+ patients. Methods: Publically available transcriptome profile (GSE15043) of genetically engineered isogenic models of TRR cell lineages in mammary HER2+ BT474 cells (initially trastuzumab sensitive, TRS) are used to generate the TRR signature using Bioconductor in R. Performance of this gene signature was tested on independent in vitro and in vivo transcriptome datasets using Receiving Operation Curve (ROC). Signature scores were determined by calculating the correlations between medium centered gene expression levels within the signature and gene expression levels for that gene within a given sample, following quantile normalization. Three pairs of TRR vs. TRS cell lines were used for accessing their profiles of proliferation, migration, genomic stability and signature scores. Genes of interest in the signature were further perturbed in TRR cells for drug sensitivity assay. Results and Conclusion: TRR cell lines have significant more malignant phenotypes in comparison to TRS cell lines, indicated by their shorter doubling time, faster migration and more innate double strain breaks. 43 differentially regulated genes between BT474R (engineered TRR cell line) and BT474 are identified and the area under the curve (AUC) of ROCs are 0.75, 0.88, and 0.84 for microarray data of CCLE, RNA sequencing data of CCLE and patient dataset GSE62327, respectively. In addition, TRR signature is significantly enriched in non-responding cell lines, TRR patients and patients relapsed within 3 years after treatment. TRR cell lines – BT474-C5, T47D and HCC1954 exhibit significant higher TRR signature scores in comparison to TRS cell lines by the 43-gene PCR array. Interestingly, overexpression of one of gene candidates, S100A8, which is also a key cytokine in regulating antibody-dependent cytotoxicity (ADCC), conferred trastuzumab resistance in TRR cell lines. This finding could assist in clinicians' diagnosis and selection of treatments for HER2+ breast cancer with a more accurate risk evaluation and offering promising alternatives such as immunotherapy. Citation Format: Yin J, Sand A, Sun C, Piacsek M, Beres A, Tjoe JA. Signature-guided biomarker discovery and therapy for trastuzumab-resistant HER2-positive breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P5-03-12.