High-Throughput Profiling of Colorectal Cancer Liver Metastases Reveals Intra- and Inter-Patient Heterogeneity in the EGFR and WNT Pathways Associated with Clinical Outcome

Introduction: In cancer progression the development of distant metastases is the crucial adverse event. In colorectal cancer 70% of patients develop metastases in the liver. Within the BMBF-founded consortium MetastaSys, the role of the WNT and EGFR pathway in these metastases was investigated. In order to address the phenomenon of intermetastatic heterogeneity metastasis profiling was applied to several metastases of two patients. Methods: Four and two sequential liver metastases of two pilot cases, as well as corresponding normal liver tissue were comprehensively characterized in terms of clinical annotations, standardized histopathological and molecular subtyping. In all metastases the mutational status of KRAS, NRAS, BRAF and PIK3CA was assessed, as well as alterations in DNA mismatch repair enzymes and microsatellite instability. Fresh frozen samples were characterized by an experienced pathologist (tumor cell content, amount of stroma, inflammatory infiltrate and necrosis) to assure high quality of tissue. Samples were subsequently profiled via transcriptome sequencing (RNA-Seq) and proteomic profiling with Reverse-Phase-Protein-Arrays (RPPA). Results: A standardized data analysis pipeline for an integrated analysis of RNA-Seq and protein array data with clinical and genetic data was established. Dimensionality reduction of the transcriptome data revealed a distinct separation between normal tissue and metastases samples. Interestingly, the correlation between healthy samples of both patients was higher compared to the correlation between different metastatic samples of an individual patient. In different metastases of the same patient heterogeneous expression patterns were identified in the EGFR and the WNT pathway with relevant implication for clinical treatment approaches. Conclusion: High-throughput profiling of gene expression allows a clear discrimination between normal and metastatic liver tissue. Identified master regulators could identify drivers of intrapatient metastasis heterogeneity. Based on our findings and the positive validation of the established methodology including innovative high-throughput profiling and comprehensive bioinformatics approaches we started analyzing a large number of metastatic patient samples within the consortium. Overall design: Two conditions of rectal cancer patient (I+II) data separates by the event of distant metastasis: 1. good prognosis, no event (3 samples), 2. bad prognosis, event (8 samples); remaining patients are categorized as bad prognosis, event (43 samples).