You are here
Protein expression by genetic mutations identified in gene panels (hotspots) and efficacy of targeted treatments.
Background: Treatment decision support by next generation sequencing of gene panels is currently limited to the analysis of somatic (tumor) data from DNA sequencing without taking into consideration expression of mutated genes. We present here a supercomputer-driven, cloud-based integrated analysis of genomic (DNA) and transcriptomic (RNA) sequencing data to 1) directly identify driver variants between somatic and germline (normal) DNA and 2) to determine expression of identified mutations in a cohort of 3,784 patients, thereby establishing therapeutic relevance of the mutated genes overcoming the limitations of panels. Methods: This large scale 3,784 patient genomic (DNA and RNA sequencing) data set from 19 anatomical tumor types was processed to detect DNA variants (germline vs somatic) and RNA expression, and to establish not only the existence but also the expression level of hotspot mutations in the following oncogenes: PIK3CA, KRAS, NRAS, AKT1, BRAF, IDH1, CTNNB1, and IDH2. Results: Of the 3,784 patients in this analysis, 720 were found to have mutations in the oncogenes listed above. Remarkably, only 38 (5.3%) of these patients had better than 90% expression by RNAseq, and 36 patients (5.0%) with identified hotspot mutations had no or low ( < 10%) expression. For example, mutations at position E545 in the PI3K protein encoded by the PIK3CA gene, which has been targeted by both pan-PI3K and mutant-targeted drugs in clinical trials, showed low or no expression in 12% (5/41) of breast cancer patients and not a single patient showed relatively maximal ( > 90%) expression. Similarly of the 204 thyroid cancer patients with a BRAF V600 hotspot mutation, 7.5% (15/204) had low or no expression and none had relatively maximal expression. Conclusions: These findings illustrate that genetic mutations in gene panels (hotspots) do not always result in protein expression. Given that many gene mutations were not expressed, we conclude that an informed molecularly-driven clinical treatment decision requires insight into downstream protein expression and not just DNA alterations alone.
Abstracts by Stephen Charles Benz:
Identifying patient-specific neoepitopes for cell-based and vaccine immunotherapy targets in breast cancer patients by HLA typing and predicting MHC presentation from whole genome and RNA sequencing data.Meeting: 2016 ASCO Annual Meeting | Abstract No: 11606Category: Tumor Biology - Tumor-Based Biomarkers
Genomics, transcriptomics, and proteomics in the clinical setting: Integrating whole genome and RNA sequencing with quantitative proteomics to better inform clinical treatment selection.Meeting: 2015 ASCO Annual Meeting | Abstract No: 11093