RNA Based Approaches to Profile Oncogenic Pathways From Low Quantity Samples to Drive Precision Oncology Strategies

Precision treatment of cancer requires knowledge on active tumor driving signal transduction pathways to select the optimal effective targeted treatment. Currently only a subset of patients derive clinical benefit from mutation based targeted treatment, due to intrinsic and acquired drug resistance mechanisms. Phenotypic assays to identify the tumor driving pathway based on protein analysis are difficult to multiplex on routine pathology samples. In contrast, the transcriptome contains information on signaling pathway activity and can complement genomic analyses. Here we present the validation and clinical application of a new knowledge-based mRNA-based diagnostic assay platform (OncoSignal) for measuring activity of relevant signaling pathways simultaneously and quantitatively with high resolution in tissue samples and circulating tumor cells, specifically with very small specimen quantities. The approach uses mRNA levels of a pathway's direct target genes, selected based on literature for multiple proof points, and used as evidence that a pathway is functionally activated. Using these validated target genes, a Bayesian network model has been built and calibrated on mRNA measurements of samples with known pathway status, which is used next to calculate a pathway activity score on individual test samples. Translation to RT-qPCR assays enables broad clinical diagnostic applications, including small analytes. A large number of cancer samples have been analyzed across a variety of cancer histologies and benchmarked across normal controls. Assays have been used to characterize cell types in the cancer cell microenvironment, including immune cells in which activated and immunotolerant states can be distinguished. Results support the expectation that the assays provide information on cancer driving signaling pathways which is difficult to derive from next generation DNA sequencing analysis. Current clinical oncology applications have been complementary to genomic mutation analysis to improve precision medicine: (1) prediction of response and resistance to various therapies, especially targeted therapy and immunotherapy; (2) assessment and monitoring of therapy efficacy; (3) prediction of invasive cancer cell behavior and prognosis; (4) measurement of circulating tumor cells. Preclinical oncology applications lie in a better understanding of cancer behavior across cancer types, and in development of a pathophysiology-based cancer classification for development of novel therapies and precision medicine.