SAN FRANCISCO (GenomeWeb) – Preemptive genomic screening of asymptomatic individuals to determine their risk for developing disease is increasingly being tested in the research setting to determine how and whether it can be implemented on a large scale and whether certain conditions are best suited for this type of testing. A handful of commercial providers are also offering genomic screening for some limited conditions, like hereditary cancers.
These efforts have uncovered some benefits of testing — for instance, in a case reported by the Geisinger Health System, being able to detect a tumor early after discovering a hereditary cancer risk variant. But they have also revealed challenges: Researchers are finding higher-than-expected frequencies for some variants, leading to questions about their true population prevalence and penetrance. The results also raise questions about how to interpret findings in the absence of other risk factors and what measures to take when a disease risk variant is found.
Daniel Siu, CEO of Rainbow Genomics, which partnered with Color earlier this year to offer a 30-gene hereditary cancer screening panel, said in an interview that from the first 559 individuals tested — asymptomatic individuals of Chinese ethnicity — around 6 percent were positive for a risk-conferring variant. That's higher than expected, he said, based on the 1 percent to 3 percent prevalence that's been reported in the literature.
Rainbow caters to self-paying customers in Hong Kong and Japan, and Siu said most of the firm's customers are Chinese patients who travel for medical care. As such, the higher frequency of pathogenic variants might not be too unusual, Siu noted, because he expected to initially draw individuals with a family history of disease or other risk factors. "In a self-pay market, these individuals are the first to be attracted to pay out of their own pockets to receive the test," he said.
However, while the first couple of hundred cases were most likely those with a family history of cancer, as indicated by an initial diagnostic yield closer to 9 percent, the rate of pathogenic variants fell and stabilized at around 6 percent as Rainbow tested more individuals. Siu said the firm plans to keep monitoring the rate to try and tease out the role of self-selection, adding that once it reaches 1,000 or 2,000 tests, it should have a better idea about the true prevalence of pathogenic variants.
Other researchers who have been conducting proactive genomic screening studies, such as the Geisinger Health System, have also reported finding a higher-than-expected frequency of some hereditary cancer mutations. At the American Society of Human Genetics meeting earlier this month, Vida Abedi, a research scientist at Geisinger, reported additional results. For one type of rare neurological tumor, hereditary paraganglioma-pheochromocytoma syndrome, her team found pathogenic variants at a 100-fold higher rate than expected, for example.
Siu noted that Rainbow's findings have implications not only for the individual, but also for that person's family members. He said that while most have been willing to share the information with their family, "whether they take action is another story."
Whenever possible, Rainbow tracks the outcomes of patients, he said. In many cases, however, the patients travel from China to receive testing, and although Rainbow does provide post-test genetic counseling, it is "difficult to quantify" whether patients follow recommendations, he said. Some patients are based in Hong Kong, though, and Rainbow has partnerships with local physicians to track their outcomes.
Rainbow's results also reflect other findings that have suggested that frequencies of pathogenic variants vary depending on the ethnic population, and that in non-Caucasian populations, individuals are more likely to carry a variant of unknown significance due to a lack of research in non-Caucasian populations.
Siu noted, for example, that mutations in the MUTYH gene, which has been linked to an autosomal recessive form of familial adenomatous polyposis, accounted for half of the variants Rainbow reported. "Because over 3 percent of patients carry this variant, it is more likely a carrier status issue for Asians, so you'd want to make sure couples who both carry this variant are screened to assess risks for their children ahead of time," he said. Again, he noted, the firm would need to validate these findings in larger numbers of individuals.
Rainbow's findings are similar to results reported by others at the recent ASHG meeting.
Medcan, a corporate wellness firm in Toronto, Canada, which similar to Rainbow offers proactive genomic testing under a self-pay model and in partnership with company wellness plans, has found that a higher-than-expected proportion of individuals who opt for genomic testing are found to carry risk variants.
The firm offers Invitae's 139-gene Genetic Health Panel, which was recently increased to 147 genes and includes genes related to hereditary cancer and cardiovascular disorders.
Jessica Gu, a genetic counselor at Medcan, said in a presentation at ASHG said that of around 1,000 individuals, about 14 percent had variants that increase their risk for disease and 41 percent were carriers of a risk allele.
Gu added that positive results have led to a change in management for 81 percent of individuals and that only about one-third of those who received a positive result had a relevant family history that would suggest they may be at risk.
She noted that there are a number of challenges with screening average-risk individuals. Because most of what's known about genetic risk variants, such as prevalence and penetrance, is from studies in high-risk individuals, it's unclear whether the same will prove true for average-risk individuals. Other studies have shown that for some hereditary cancer risk variants, the prevalence in the general population is higher than previously estimated, and it is unclear whether these variants will have the same penetrance as has been reported in high-risk individuals. Gu said that in the population Medcan has tested, they found a higher prevalence of variants for hypertrophic cardiomyopathy and of some breast and ovarian cancer risk variants than previously estimated.
Interpreting such results, particularly for individuals without a family history, is challenging, Gu said.
Similar to what Rainbow's Siu noted — that individuals with a family history may be more proactive in seeking out testing — Bruce Korf, co-director of the Center for Genomic Medicine at HudsonAlpha, said that he suspects a similar situation for a population genotyping project he is conducting. Korf is a principal investigator for the Alabama Genomic Health Initiative, a state-funded five-year project to perform whole-genome sequencing or array-based genotyping for more than 10,000 individuals with and without disease.
The Alabama Genomic Health Initiative seeks to study best practices for returning actionable results to research participants and to create a "genomics-ready" population in the state, Korf said.
The project includes both a population cohort of adults without a disease phenotype and individuals suspected to have a genetic disorder but without a diagnosis. Up to 10,000 individuals from the cohort will be genotyped using Illumina's Global Screening Array and up to 1,000 individuals with a suspected genetic disorder will receive whole-genome sequencing.
For the population cohort, selected variants are being returned, including pharmacogenomics variants, hereditary cancer syndrome-related variants, and variants associated with hereditary cardiovascular disorders. Similar results will be returned for the whole-genome sequencing arm, as well as any variant thought to explain the individual's disorder.
Thus far, Korf said in a presentation at ASHG, just over 3,700 participants have enrolled in the genotyping portion of the study and 104 individuals with a suspected genetic disorder have enrolled in the sequencing arm.
In total, 43 actionable results have been returned for the unselected individuals who were genotyped. At just over 1 percent of the cohort, that's a bit more than would be expected since the array does not pick up all pathogenic variants. "We think that people with concerns or interest about their family history are enriched" in this study arm, he said.
Korf added that there is some concern among the researchers that individuals who perhaps should be receiving clinical rather than research testing because they have a family history, for instance, are instead opting to enroll in the population research study. It's not yet clear why these individuals are choosing to go this route, as opposed to pursuing standard medical care, but he said it could be because the "barrier is lower." Enrolling in the research study can be easier than scheduling an appointment with a physician and inquiring about genetic testing. He clarified that there is no evidence that people are enrolling in the research study because they are unable to access medical care, but said the team would look further into that.
One of the first results the team returned to an individual in the population study was a pathogenic variant in the MYH7 gene, conferring risk for hypertrophic cardiomyopathy. "When the counselor called the participant and explained the result, that he was predisposed to cardiomyopathy, the participant said: 'Well, that was true for my first heart.'"
Participants have the option of sharing their results with their healthcare provider and to allow results to be included in a biobank for further research. Interestingly, Korf said, while over 90 percent consented to research, only around 47 percent agreed to have their results shared with their healthcare provider. He added, however, that "virtually everyone who gets actionable results chooses to share them with their healthcare provider." Participants are just not opting to have results automatically shared, he said, and are instead providing them to others on their own terms.
Korf said that for both the population and disease cohorts, the researchers plan to study the utility of genomic information — how the participants use the information, whether actionable findings in one individual prompt testing for additional family members, what the negative outcomes for returning results to individuals without evidence of disease are, and whether negative results give participants false reassurance. "We're getting the sense, anecdotally, that that might be a real concern," he said.
Another interesting facet of the study is who has chosen to participate. At the moment, Korf said, 73 percent of the individuals who have enrolled are female. He theorized that those enrolling are doing so out of family concerns. In addition, about three-fourths of participants are white, and going forward, one focus will be to increase the diversity of those who enroll, Korf said.