PO-451 Targeted proteomics to improve therapy stratification of cancer patients

Abstract

IntroductionThe heterogeneity of tumours calls for patient stratification to select the most effective, personalised therapies. The NCT MASTER (Molecularly Aided Stratification for Tumour Eradication Research) program aims at comprehensive characterisation of cancer patients seen at NCT Heidelberg and Heidelberg University Hospital. SNVs, small InDels, CNVs, and gene expression data obtained by whole-genome/exome and RNA sequencing are assessed by a molecular tumour board for their potential clinical impact. Variants are assigned to therapeutic baskets (PI3K-AKT-mTOR signalling, RAF-MEK-ERK signalling, tyrosine kinase signalling, DNA damage response, developmental pathways, immune evasion, cell cycle), connecting actionable aberrations with targeted drugs. Evidence-based treatment recommendations are provided to treating physicians and are currently acted upon clinically in ~30% of cases. We here investigated whether targeted proteomics adds valuable information in the therapy stratification process.Material and methodsWe analysed 134 tumour specimens, representing 17 entities, from patients enrolled in the NCT MASTER using reverse phase protein arrays (RPPA) technology. 123 antibodies for key proteins and posttranslational modifications (PTMs) mapping into the NCT MASTER therapeutic baskets were incubated, and signal intensities bioinformatically processed. Genomic data were reevaluated by the molecular tumour board to update therapy recommendations and were then integrated with proteomic information.Results and discussionsHeatmaps were generated to display expression/PTM patterns. The heterogeneity of the cohort was reflected in the proteomic data; however, samples did not cluster by entity, tumour cell content, or genomics-based treatment recommendation. Next, 48 proteins reflecting activation of cancer signalling pathways were investigated in detail to evaluate whether proteomic data would add relevant information towards shaping therapy recommendations. Results were compared with recommendations based on genomic data alone. Disparate results were found for many cases mirroring what the heatmaps had suggested.ConclusionWe have systematically evaluated the potential benefit of proteomic profiling in the therapy recommendation process of the NCT MASTER program. The data suggest that targeted proteomics indeed adds an additional, clinically meaningful layer to treatment stratification. Implementation of a proteomic workflow in the NCT MASTER program is in progress.