Despite the vast amount of information currently available through research and patient care about human genetics, there are still significant gaps in our knowledge. The term "variant of unknown significance" (VUS) is used to describe a DNA change found in an individual that has not yet been reliably characterized as benign or pathogenic.
While the analogy has been made between the DNA VUS and the finding of an incidental mass on radiographic imaging, there is an overwhelming difference in the prevalence of these two findings. If enough DNA is analyzed in any given individual, then not only is there sure to be a VUS, but also the sheer number of variants of unknown significance in an individual’s complete genome is a bit staggering. In the case of the individual with Charcot-Marie-Tooth disease reported on last year (N. Engl. J. Med. 2010;362:1181-91), there were more than half a million VUS that had never been seen before and many more that were previously seen in individuals, but remain VUS because they have not yet been characterized as benign or pathogenic.
Michael F. Murray
In the medical literature, there are many reported cases in which the gold standard of proving the biological relevance of a given DNA variant has been pursued by using clinical data and by supporting those data with experiments in animal models, in vitro experiments, or both. In such cases, a strong argument for the conclusion of benignity or pathogenicity can be made. It is obvious, however, that such cases will be the exception rather than the rule, given the sheer volume of variants that exist. Our current technology for animal and in vitro modeling will never suffice to confirm the anticipated effect of each identified VUS, and new high-throughput experimental systems are needed.
"In silico" approaches to predict the anticipated effect of VUS have been developed and are being refined; among these are PolyPhen and SIFT (scale-invariant feature transform). The focus of these programs to date has been the effect of VUS that are found in the protein coding sequences (CDS) of genes. These computer programs attempt to predict the effect of a VUS on the function of a gene based on the effect of the change on the protein structure, the effect of the change on the functional areas of the protein, and the accumulated evidence regarding the tolerance of changes in the given site of the VUS in other humans and in non-human species. The ability of programs to predict the effect of VUS outside the coding region of genes (that is, greater than 97% of the human genome) lags behind the current tools for prediction in CDS.
The classic tool of documenting whether or not a variant segregates with disease in a family with multiple affected individuals remains the most powerful tool for judging the significance of a VUS. However, small family size or lack of clear disease in some individuals within a family makes the application of this tool impossible in most VUS cases.
As we move forward into an era when more and more of your patients have DNA information to interpret be prepared to counsel patients on the need to live with a degree of uncertainty about the meaning of some of their DNA results. At the same time, expect that the tools for interpretation will continue to improve as the human DNA database grows and the interpretation programs become more refined. The approaches applied in future VUS interpretive programs will likely be: high-throughput experimental systems, improved computational systems, and hybrids of experimental-computational systems.