Cancer-associated gene-variant testing reports classify polymorphisms as “benign”, “likely benign”, “likely pathogenic”, “pathogenic”, or as variants of uncertain significance (VUS). As more patients are tested, test results accumulate and scientific understanding of the role of gene variants in cancer improves, gene variants may need to be reclassified — upgraded from likely benign, benign, or VUS to pathogenic or likely pathogenic; or downgraded from pathogenic, likely pathogenic, or VUS to benign or likely benign.
A recent analysis of test results at the Myriad Genetics lab, published in JAMA, showed that periodic reviews of new data pertinent to patient genetic variations led to the reclassification of a quarter (24.9%) of VUS findings during the 10-year period between 2006 and 2016.1 In the vast majority of cases (91.2%), VUSs were downgraded to benign or likely benign. Reclassifications were rare for pathogenic and likely pathogenic variants (0.7%) and benign or likely benign results (0.2%). Overall, 6.4% of initially identified unique gene variants were reclassified.1
The findings spotlight the need for all gene-testing labs to review records and alert clinicians when gene variants are reclassified, emphasized senior study coauthor Theodora Ross, MD, PhD, a professor at the UT Southwestern Medical Center in Dallas, Texas, and director of the UT Southwestern Cancer Genetics Program.
Continue Reading
Dr Ross’ team investigates the normal function of BRCA1, as well as the basic biology of cancer cells and how those cells resist cancer drugs. She published her first book on cancer genetics, A Cancer in the Family: Take Control of Your Genetic Inheritance, in 2016.
Cancer Therapy Advisor asked Dr Ross about her team’s study of gene-variant reclassification rates that harnessed a decade of results drawn from the genetic testing performed by Myriad Genetic Laboratories Inc.
Cancer Therapy Advisor (CTA): Should all gene tests be reviewed regularly for possible reclassification?
Dr Ross: Yes. With the way we examine data these days, gene-variant analysis can be automated, where variants are put into a database and when publications come out in peer-reviewed journals or patient data come in regarding that variant, the variant is flagged. Since this part can be automated, there may ultimately be minimal human capital that needs to go into it. Once there is new information, then the scientists — humans — need to pay attention and assess if [a variant] should be declassified and then reclassified. To my knowledge, artificial intelligence or AI is not able to do the latter. Human beings are still of use!
CTA: How should variant reclassification be explained to patients?
Dr Ross: We should tell our patients that no matter what their result is, to please check back in with us each year to determine if there is new information from the labs, if they need updated testing, and to keep us informed if any personal or family history facts change, such as new cancer diagnoses in the family. Clinicians have an obligation to contact patients when they receive an amended reclassification report, and genetic ‘check-ups’ can be a backup when that obligation is not met.
