NEW YORK – The evidence supporting variants of uncertain significance (VUS) should be considered when determining which VUS to report in diagnostic multi-gene panels, argue scientists from across academic and commercial labs in newly published research.
Current laboratory policies typically call for reporting all VUS in diagnostic testing, or when performed on a symptomatic individual.
This practice, said Heidi Rehm, a professor of pathology at Harvard Medical School and the first author of a study published in late July in Genetics in Medicine, is a holdover from when multi-gene panels were much smaller and focused on a narrower range of indications.
"The idea was [that] we should report variants of uncertain significance in diagnostic genetic testing because there's the ability to follow up on those results and clarify a [VUS] with additional clinical investigations," she said.
As our understanding of the variability in genetic diseases has grown, Rehm explained, gene testing panels have also grown, complicating a physician's ability to effectively follow up.
As an example, Rehm noted that some neurodevelopmental panels contain over 500 genes.
"You just get a lot of background noise because you're only testing genes that are slightly related to the phenotype," she said, adding that this has already become a problem in clinical practices.
"In addition to being a burden during the initial return of results, and raising the likelihood that mismanagement occurs with a VUS, it also has a high burden after the initial return," she stated. "For example, in our [Massachusetts General Brigham] Cancer Clinic, 17 percent of reports were subsequently updated and new versions sent to the clinic when VUS were reclassified, mostly moving to 'likely benign' [or] 'benign.'"
Whole-genome and exome sequencing, typically used in screening assays, go even further than multi-gene panels, interrogating thousands of genes at a time. Counterintuitively, this results in fewer reported VUS than with panel testing.
In their study — conducted on behalf of the Medical Genome Initiative, a coalition of healthcare and research organizations formed in 2019 — Rehm and her collaborators explain that this is largely because of different reporting standards. VUS identified through genome and exome sequencing will generally be reported only if they meet certain evidence criteria related to pathogenicity, clinical correlation, and segregation analysis, as well as family genetic information when available.
Interrogating data from 1,463,812 multi-gene panels, 42,165 exome sequencing tests, and 6,329 genome sequencing tests performed across 19 clinical laboratories, Rehm's group found that the rate of test results with at least one VUS without causal etiology was significantly lower for ES/GS tests (22.5 percent) than the rate from MGPs (32.6 percent).
In contrast to panel testing, Rehm explained that labs conducting genome and exome sequencing tests will more often report only variants with a high likelihood of being relevant to a given indication.
This is not always the case, however, as some studies have found that even VUS reported in genome and exome tests can sow confusion and consternation among both patients and physicians, particularly with respect to direct-to-consumer (DTC) genetic tests.
Earlier this year, for example, an American Heart Association (AHA) study found that incidental variants — rare variants unrelated to the phenotype or disease for which a test was ordered — were frequently reported on DTC genetic tests. Unsure how to interpret these findings themselves, patients often bring them to their physicians seeking guidance, but without sufficient genetics experience or access to genetic counselors, these doctors may also struggle to provide clarity.
As a result, the AHA study suggested not reporting VUS to patients at all.
In contrast to DTC tests, clinical labs often do employ genetic counselors to help guide physicians through test results.
Paul Kruszka, chief medical officer of GeneDx, which provided some of its de-identified testing data for the study, said in an email that the company has board-certified genetic counselors available to help answer the questions of both patients and providers.
"Exome and genome tests provide the best options in reducing uncertainty and making patient reports clearer and more useful, especially in cases where trio testing is available," Kruszka said in an email.
"Patients want access to testing that gives the highest chance of providing a clear genetic answer for their health concerns, and providers want to reduce the amount of uncertainty they have to navigate, as well," he added.
Rehm sees greater nuance in the results, acknowledging that the rationale for reporting all VUS in panel testing is not always without merit, as some VUS will become pathogenic or likely pathogenic over time and steps taken by physicians over the course of practice can help determine this.
She noted, however, that many VUS are borderline benign or likely benign and certain testing scenarios, such as in cancer testing, often involve little to no follow-up and/or low probabilities that a VUS becomes pathogenic or likely pathogenic.
In their study, Rehm and her colleagues call for changes to reporting practices both at the level of individual laboratory policies and of professional guideline recommendations, both of which she says are just beginning to happen, through two American College of Medical Genetics and Genomics (ACMG) working groups, for instance.
Rehm said one ACMG working group is developing the next sequence variant classification guidelines, which will give guidance for subdividing VUS into three tiers depending on evidence level.
Another working group, she said, will publish guidance on how to use those tiers for reporting and making other recommendations.
"For example," she said, "physicians rarely give phenotype data for panel testing, unlike exome and genome sequencing, where its required. But if they did, labs could use that data to decide which VUS are relevant to put on a report."
Doing so could also reduce patient anxiety over uncertain results and lower their chances of having unnecessary treatments done.
One study from the University of Pennsylvania Health System, for example, found that between 10 and 15 percent of women with pathogenic variants or VUS in genes not known to substantially increase ovarian cancer risk reported having opted to surgically remove their ovaries.
Similarly, a study from Stanford University found that women often undergo genetic testing without input from a genetic counselor and that many of those with average disease risk and VUS undergo preventive bilateral mastectomies, suggesting that in these cases, neither the patients nor their physicians had fully understood the test results.
Some individual labs have already begun changing their reporting practices. Mass General Brigham Laboratory for Molecular Medicine, Quest, and Baylor, for example, all assign VUS to tiers of evidence strength, which they use to determine VUS reporting. Additionally, Rehm noted that Ambry has made VUS reporting optional for their large neurodevelopmental multi-gene panel.
Current reporting norms, Rehm said, are something of an evolutionary result that now needs more intentional direction.
"There hasn't been time [for us] to catch up and say, maybe the strategy we're using today to just report all VUS in a panel shouldn't be the strategy," she said.