Deriving Clinical Action from Whole-Genome Analysis

Abstract

Pathology is charged with the accurate assessment and characterization of disease symptoms. Much of this effort revolves around the proper diagnosis of cancer type. Previous to recent technological advances, these assessments and characterizations primarily consisted of ‘reading’ the stained tumor tissue to assess the status of cells, their configuration and type. Under today’s standards in pathology, molecular character ization of disease includes the assessment of the mutational status of individual genes, or small panels of genes, via proprietary commer cial platforms including targeted geneexpression arrays and quantitative RT-PCR. Table 1 shows a selection of clinically actionable variants used routinely in cancer care. These are individual or small panel tests that are largely delivered via a ‘send-out’ model for interpretation within a clinical time window. In the near future, one imagines that all of these and many more, for example, a large fraction of those listed in the solid tumor test directory [101] – variants that include SNP, copy number, structural variants or patterns of gene regulation – can and will be replaced by large-scale approaches, and in particular by whole-genome ana lysis (WGA). The benefits of moving from single testing to large-scale testing are straightforward. A large-scale approach will improve throughput, data management, general efficiency, speed of information transfer and result in better eco nomics for the hospital. In a few cases, the pressing need for tumor subtyping for the selection of gene-targeted therapies or determination of eligi bility for clinical research trials of new drugs has already driven a shift to panels of genes encompassed within laboratory developed tests (LDTs) [1,2].