SAN FRANCISCO (GenomeWeb) – Close to 20 percent of healthy individuals may harbor potentially disease-causing variants, according to two recent studies.
Researchers from the MedSeq Project, reporting in the Annals of Internal Medicine, and scientists from Stanford University, whose study appears on the BioRxiv preprint server each said that sequencing asymptomatic adult cohorts found that a surprising number of them carried actionable variants in monogenic disease genes.
With the declining costs of next-generation sequencing, many envision a future in which whole-genome or exome sequencing is performed not just to diagnose existing disease, but as a screen to understand people's predisposition for disease and ultimately to prevent it. The two research groups sought to address whether such a scenario is currently feasible.
Although both studies evaluated relatively small numbers of individuals — 51 sequenced in the MedSeq study and 70 in the Stanford one — the researchers said the work adds to growing evidence that genomic sequencing for disease predisposition will eventually be a reality.
"I do envision a world where everyone gets their genome sequenced at or before birth and the information is used to manage people's health," Mike Snyder, who led the Stanford study, said. "Right now, family history is used, which I would argue is quite poor. And I think genomics can improve on that quite a bit, especially for rare, Mendelian disorders."
Both groups are now looking for funding to expand their studies to evaluate larger numbers of individuals.
"This was an informative experience, but expanding to greater numbers of patients will expand our estimates about variant prevalence and what are the true health and economic outcomes of introducing sequencing to a healthcare system," Jason Vassy, a principal investigator in the MedSeq Project and lead author of the Annals of Internal Medicine study, said.
Manuel Corpas, scientific lead of Repositive who was not involved with either study, said that both studies found a "substantially higher rate of actionable findings," at 17 percent to 21 percent, than what has been reported in previous studies, which is only about 1 percent to 5 percent. He said that while such results were interesting, the sample sizes in both studies were small and the cohorts were not very diverse, making it difficult to draw strong conclusions until results had been found in larger groups of more diverse individuals.
Vassy agreed that larger studies are needed, as is the need to study asymptomatic people, since both studies found much higher than previously estimated proportions of individuals harboring potentially disease-causing variants.
Such studies would help to shine a light on the population prevalence and penetrance of disease-causing variants.
"A lot of these disease-causing variants and genes were first described in individuals with strong family history or a severe phenotype," Vassy said. "Those striking families allowed the genes to be discovered, but we don't know the population prevalence of those variants and whether there are milder phenotypes, reduced penetrance or expression, or later onset."
The MedSeq Project was funded with a four-year, $9.6 million grant from the National Human Genome Research Institute as part of its Clinical Sequencing Exploratory Research Project. One part of the project included sequencing the whole genomes of 51 healthy patients and conducting a family history evaluation of the participants. The findings were compared to 50 patients who only had a family history evaluation. Sequencing results were delivered in the context of patients' primary care physicians, rather than genetic counselors. The goal was to see whether such an approach was feasible in the primary care setting, whether sequencing uncovered clinically relevant information not discovered by a family history evaluation alone, and how sequencing impacted healthcare costs.
At a number of prior conferences investigators reported on some of the preliminary top-line findings from MedSeq, including the finding that 21 percent of individuals harbored a potentially pathogenic variant and that sequencing added only relatively small healthcare costs compared to costs incurred without genomic testing.
The publication of the results, though, sheds light on the types of variants uncovered by genomic testing and also included an evaluation of the action that physicians took after receiving results.
"There's a lot of concern in the field that primary care physicians, or other non-geneticists, are not prepared to handle the onslaught of genetic information," Vassy said. And, by and large, he added, the physicians handled the patients appropriately, often referring patients to a genetic counselor.
In total, of 11 patients who had a pathogenic variant, the physicians were judged by an external panel of geneticists to have handled eight of the cases appropriately. For one case, the panel found that the physician under-evaluated the importance of a pathogenic variant, and in another case the panel found that the physician did not report back correct information on the inheritance of a variant. In the third case, the panel was neutral on its evaluation, finding that the physician under-evaluated a patient's subtle clinical symptoms, but the corresponding variant was classified as being of uncertain significance.
Another interesting finding was that of the 11 individuals who were deemed to have an actionable variant, only two had any evidence of disease. Vassy said that the group considered variants actionable and returned them back to physicians if they were classified as pathogenic or likely pathogenic, according to American College of Medical Genetics and Genomics guidelines. In addition, he said the team returned back variants of uncertain significant that "highly favored" being pathogenic. The two individuals showed symptoms of disease also had variants classified as pathogenic, he said. And, there was only one individual with a pathogenic variant with no symptoms of disease. The other eight individuals had either likely pathogenic or a VUS that favored pathogenic.
For the nine individuals with actionable variants and no signs of the disease, all but one also did not have a family history of disease, Vassy said. That one individual had a variant that was associated with a disease that can cause severe blistering and rashes when exposed to sun or certain medications. After the physician discussed that result with the patient, the patient did acknowledge experiencing some minor symptoms and also identified a family member who experienced much more severe rashes.
"That's not typically what you think to report when a primary care physician asks you about family history," Vassy said. "It was only after the fact of finding the variant."
The Stanford study found a similar proportion of individuals with actionable mutations in disease-related genes. Out of 70 individuals who received exome sequencing, 12 had actionable results. The Stanford cohort differed from the MedSeq cohort in that the individuals were involved in a diabetes study, so it included four participants who had previously been diagnosed as diabetic and 20 participants who were pre-diabetic.
Snyder said that the exome sequencing helped shed light on some of the participants' known health issues — including resolving one patient's misdiagnosis — and also identified important disease predisposition variants, in particular for inherited cancer. Five participants had variants in cancer predisposition genes.
Similar to the MedSeq study, not all participants with actionable variants had evidence of the disease or a family history. Two patients, however, did. In one patient, exome sequencing uncovered that the patient was heterozygous for a likely pathogenic variant in a gene that causes maturity onset diabetes of the young. The patient had previously been diagnosed as type 2 diabetic. MODY is often misdiagnosed, Snyder said, as was the case with this patient, and has "a very different treatment strategy," he said. The sequencing results enabled the patient to discuss potential changes to treatment with a physician and she also had her three children tested for the genetic variants.
In another patient, researchers found a likely pathogenic variant in a gene associated with dilated cardiomyopathy. The specific variant had not previously been reported as causative, but the patient had a family history of the disease. After identifying the variant, physicians performed a follow up echocardiogram and ultimately decided to place the participant on blood pressure-lowering drugs.
Snyder said that the exome sequencing was part of a larger study at Stanford, called the Integrated Personal Omics Profiling study, in which researchers perform not only exome sequencing, but also transcriptome sequencing and metabolomic, proetomic, and microbiomic analysis, and combine that with integrated health analyses of the patients. The ultimate goal, he said, "is to define people's healthy state and then catch disease at the earliest possible time so it's best managed."
He said the researchers are in the process of evaluating the data for about 100 participants. Thus far, he said, aside from the genomic information, the metabolomics data has been the most useful, and he said the researchers are looking at how to link the metabolome and genome data.
Like the MedSeq group, Snyder said his group is looking for funding to expand the study to a larger group of people.
Going forward, Vassy said that at the very least, these recent studies make the argument that even larger studies are needed, such as the National Institute's of Health's All of Us study.
"We hope that our results give a glimpse, a microcosm of what this will look like in a primary care setting," he said.