70% of breast cancer (BC) patients have PI3K pathway gene mutations. PIK3CA mutations are used as biomarkers of response to PI3K inhibitors (PI3Ki), but mutant PIK3CA does not always correlate with response to PI3Ki. There is a clinical need to identify novel biomarkers of response to treatment for these BC patients. Non-coding mutations are capable of regulating gene expression and could play a role in maintaining pathway activation or activating compensatory mechanisms and predict therapy response in BC. Non-coding mutations surrounding kinome genes were obtained from BC patient whole genome sequencing data (n=915). The Activity by Contact bioinformatic model was used to predict interactions between mutated regions and surrounding transcription start sites. Two mutations with a predicted interaction were identified in the ERBB2/CDK12 gene loci of HER2+ BC patients (chr17:39959176/39959179). However, identification of a cell line with these mutations to be used as a model for their further study was not possible. Instead, a three-base insertion in the same predicted regulatory region (chr17:39959177) was identified through sequencing in multiple HER2+ BC cell lines and was selected for functional analysis. Mutation frequency for the chr17:39959177 mutation was validated in a cohort of metastatic HER2+ BC patients via MassArray, where it was observed in 75% of the patients analysed (n=73). Patients with the chr17:39959177 mutation had worse response to standard therapy (p=0.006) whilst it was associated with a significant decrease in both progression free and overall survival (p=0.019). These findings highlight the potential of this mutation as a biomarker for HER2+ BC treatment response and suggest a possible function in BC biology. Further analysis of the chr17:39959177 mutation revealed that it is located inside the HER2 amplicon in an area with high transcriptional activity in HER2+ BC cells. Its activity as an enhancer of gene expression was confirmed in vitro via luciferase assays. In silico gene expression analysis demonstrated how presence of the chr17:39959177 mutation is associated with increased gene expression up to 300kb away from the mutation site. Once the regulatory effect of the chr17:39959177 mutation had been confirmed, in silico analysis of protein expression was conducted to identify altered signalling pathways that could be used as potential drug targets. Mutant cell lines had a dysregulation of two proteins involved in DNA damage repair and three proteins involved in inflammation. In addition, analysis of in silico drug sensitivity demonstrated that mutant cell lines had increased sensitivity to drugs targeting DNA repair mechanisms and inflammation signalling. These findings highlight how presence of the chr17:39959177 mutation has the potential to identify which patients are less likely to respond to trastuzumab treatment and/or may have a worse outcome. This mutation is associated with altered signalling that could be targeted with novel drug combinations to circumvent treatment resistance. The data herein supports the need for further research to validate its clinical role.