Herceptin® resistance in breast cancer : a pathwaybased approach

Abstract

Background and Aims: Trastuzumab (Herceptin®) has been hailed as one of the successes of personalized medicine for the treatment of HER2-positive breast cancer. Nonetheless, resistance to the drug and tumour recurrence are important caveats of the targeted therapy. This project aimed to study trastuzumab resistance by means of a pathway-based approach. Methods and Results: In a retrospective clinical review of all the patients diagnosed with HER2-positive breast cancer and receiving trastuzumab treatment between 2010 and 2015 (n = 119), treatment modality was the most significal predictor of event-free survival; specifically, Cox modelling showed neo-adjuvant therapy and treatment for advanced/metastatic disease to carry a hazard ratio (HR) of 4.0 and 21.0, respectively, over adjuvant therapy. A subset of patients with complete response to adjuvant trastuzumab sustained for two or more years who developed distant metastasis without local recurrence was identified; these patients would be ideal candidates for liquid biopsy (circulating tumour cell detection and characterisation), to monitor response to treatment and detect early metastasis. Immunohistochemical staining to observe total and phosphorylated (activated) HSP27 (pHSP27ser78) in archival tissues of the same patients (n = 93) showed a significant association between a higher pHSP27ser78/total HSP27 ratio and event-free survival (HR = 2.3), largely driven by lower total HSP27 protein expression. The role of activated HSP27 was corroborated, by means of high throughput phosphopretein profiling (DigiWest®), in two cell line models: UACC-893, which is intrinsically trastuzumab-resistant, and T15, generated by long-term culture of SKBR3 cells in 15μg/mL trastuzumab for induced trastuzumab resistance. In these two-cell lines, elevated pHSP27ser78/total HSP27 ratios and protein kinase C (PKC) activation, in conjunction with HER2 protein overexpression when compared to the trastuzumab-sensitive cell line, SKBR3, supported the hypothesis of HER2 receptor recycling-mediated resistance. In another intrinsically resistant cell line, JIMT-1, CIP2A overexpression (mean fold change in expression over SKBR3 = 14.3), observed by qPCR, was accompanied by a moderate response to the PP2A reactivating drug FTY720, indicating a role for deregulation of PP2A, a key member of the PI3K/Akt pathway, in trastuzumab resistance. DigiWest® phosphoprotein profiling also revealed other distinct and overlapping trastuzumab pathways, such as MEK/ERK pathway activation, not necessarily through upstream RAS/RAF, in all the three trastuzumab-resistant cell lines, and β-catenin stabilisation by inhibitory phosphorylation of GSK3β in JIMT-1 and UACC-893. Finally, a working model of circulating tumour cell isolation and characterisation by the detection of cell surface proteins (EpCAM, HER2) and RNA biomarkers (EPCAM, KRT19, ERBB2, FN1) was optimised. Conclusion: Collectively, this study shed more light into the different pathways implicated in intrinsic and acquired trastuzumab resistance, and paved the way for further studies in this area.