Genome-wide CNV testing will uncover clinically actionable incidental findings

see Incidental copy-number variants identified by routine genome testing in a clinical population

As genome-wide studies enter the clinic, evidence continues to mount that potentially clinically actionable incidental findings can be expected with some regularity. In this issue, Boone et al. describe an array of incidentally discovered copy-number-variant (CNV) mutations that may increase the risk of adult-onset disease and may be clinically actionable. The group performed array comparative genomic hybridization (aCGH), which can detect deletions, duplications, and rearrangements as small as one exon, on just over 9,000 individuals who had been referred because of suspicion of a genetic disorder. In addition to detecting CNVs relevant to the referring condition, the research team identified 83 CNVs affecting late-onset-disease genes. Frustratingly, half of the variants discovered were of unknown clinical importance. The authors suggest that, given the nearly 1 in 100 chance that potentially clinically relevant incidental genetic alterations will be revealed, patients should be made aware of this possibility before being asked to consent to testing. —Karyn Hede, News Editor

Parsing the incidentalome

see An informatics approach to analyzing the incidentalome

Given the large number and heterogeneous nature of genomic variants generated any time genome-scale analysis is performed, informatic approaches will need to be developed that assist in clinical decision making regarding their interpretation and to guide the return of such results. One such attempt is described by Berg et al. in this issue. Using an automated filtering system that categorizes genomic variations by potential clinical significance, the authors conducted a test run of 80 whole-genome sequences gleaned from publicly available sources. After they had applied a series of filters designed to reduce the probability of identifying variations unlikely to cause disease, the algorithm effectively reduced the number of variants requiring human review and identified incidental variants with likely clinical relevance, including about nine variants per person indicating carrier status for recessive disorders. The proposed filtering approach excludes typically benign missense mutations as well as variants within intergenic regions and introns. Although it decreases overall sensitivity, the authors argue that a stringent cutoff ensures high specificity, reduces the need for manual review of results to a manageable level, and is both necessary and advisable when dealing with incidental findings that represent, by definition, a low a priori risk of disease. The proof-of-concept approach is flexible and amenable to changes as more information about disease risk associated with various variants becomes available. The results also demonstrate the inadequacy of current mutation databases, an important issue because the ultimate utility of any analytical scheme in a genomic context will require a well-curated and regularly maintained universal, publicly available database of known gene variants. —Karyn Hede, News Editor