Solving The Interpretation Bottleneck for Cancer Precision Medicine

High-throughput molecular profiling technologies now allow the systematic identification of molecular drivers of cancer for most major tumour types. Clinical and functional studies have correlated these drivers with patient outcomes and helped develop targeted therapies. However, maintaining current and comprehensive interpretations of the clinical significance of variants represents a major bottleneck. To address this challenge, the Clinical Interpretations of Variants in Cancer knowledgebase (CIViC; civicdb.org) was created to provide a knowledge repository and sophisticated curation interface for expert-crowdsourcing the curation of actionable cancer variants. Importantly, all data are made freely available with a public domain license, daily and monthly data freezes, and public programmatic access. This has allowed adoption of CIViC variant interpretations into many research tools for variant annotation as well as commercial and non-commercial report generation workflows. To date, efforts have focused predominantly on small mutations, detected through targeted sequencing panels, and assumed a single-target-to-single-therapy paradigm. As sequencing costs decrease, whole genome, transcriptome, and epigenome approaches will replace these targeted methods. This will allow increasingly unbiased assay of molecular alterations of most types (large and small) and will also dramatically increase the number of variants of unknown clinical significance. Strategies to address these challenges will be discussed.