NEW YORK (GenomeWeb) – Over the last few years, Radboud University Medical Centre's Department of Human Genetics in Nijmegen, the Netherlands, has embraced exome sequencing as a routine diagnostic test for heterogeneous genetic disorders, but researchers there continue to believe the future of such testing will be in whole-genome sequencing.
The department, which was among the first in the world to implement exome sequencing for routine diagnostic use, has also introduced a next-gen sequencing-based hereditary cancer test for BRCA1 and BRCA2, is conducting somatic tumor testing in collaboration with the pathology department, and has plans to explore pharmacogenomic testing in its hospital. In partnership with the human genetics department at Maastricht University Medical Center, it is also preparing to expand noninvasive prenatal testing for fetal trisomies once NIPT becomes widely available to women in the Netherlands.
Last year, the department performed more than 6,000 diagnostic exomes, up from about 2,000 in 2013, making it its most frequently conducted single test. About 1,000 exomes were run for clients outside the Netherlands.
"We focused a lot on making the exome into a really normal diagnostic test," said Joris Veltman, a professor of translational genomics at Radboud UMC's genetics department. "From being a test for specifically selected patients, it is now a test that we commonly use in our diagnostics." Veltman spoke with GenomeWeb during a recent interview in Nijmegen that also involved Erik-Jan Kamsteeg, a clinical laboratory geneticist and one of the diagnostic lab's team leaders.
Exome sequencing diagnostics is now offered for more than 20 disease categories, all genetic disorders that can be caused by mutations in a large number of genes. The lab's expertise in several areas, including genetic skin disorders and cardiovascular diseases, increased through a strategic partnership between the human genetics departments at Radboud and Maastricht universities, which started in 2014. While early on only clinical geneticists were allowed to request the exome test, any clinician with experience in genetics can now order it.
As part of the testing process, variants in genes known to be involved in a patient's particular disease category are interpreted first. If that does not yield a diagnosis, the analysis can be expanded to the exome, which requires a referral to a clinical geneticist to provide counseling about potential incidental findings.
However, unlike many centers in the US, the lab currently does not actively scour the exome for secondary findings in genes unrelated to the patient's disorder, such as the genes recommended by the American College of Medical Genetics and Genomics. Doing so would turn the test into a genetic screen under Dutch law, for which the lab is not licensed, Veltman said, and it would increase the cost and time involved. "Getting the diagnosis for what the patient came to us is still our focus — that's challenge enough," he said.
The department is also looking to improve the characterization of variants of unknown significance, both through better bioinformatic predictions of their function and through experimental validation. For example, its researchers are now combining exome sequencing with metabolomics for certain disorders, and perform RNA sequencing and other assays. "There is a good collaboration between research and diagnostics; we definitely feed back things into research where we follow up variants," Veltman said.
The center currently does not submit clinical variants it identifies to international databases, like ClinVar or Decipher, which could aid other clinicians in their interpretation of genetic testing results, but Veltman, who also holds a part-time appointment at Maastricht University, said there are ongoing efforts to establish a Dutch database that will collect variant data from all medical genetic centers in the Netherlands. "It's definitely something that has been under fierce debate and discussion for a long time," he said.
For many genetic diseases, exome sequencing has now become a first-tier test, he said. The exception is disorders such as intellectual disability that can be caused by copy number variations, for which microarrays are still run first because of their short turnaround time. Turnaround time for the exome test has decreased from about six months in early 2015 to about four months, Veltman said.
To assess the clinical utility of the test, Radboud researchers have been comparing the diagnostic yield and cost of exome testing and other genetic tests. In one such study involving 150 pediatric neurology patients that is scheduled to be published in the near future, they performed exome testing and standard diagnostic tests — both genetic and non-genetic tests, such as MRIs — in parallel and found that the exome had a diagnostic yield of 29 percent, whereas the other tests provided diagnoses in 7.3 percent of cases.
In terms of costs, replacing all standard genetic tests by exome sequencing would reduce overall costs by about 10 percent from $8,500 to $7,700. "That is not dramatically much, but it shows you that [exome sequencing] is not increasing the cost, which is what most people were a bit afraid of," Veltman said.
The Dutch healthcare system continues to cover diagnostic exome sequencing, as it did three years ago, at the standard rate for genetic testing of about €700 ($780). Starting next year, tests will be placed into different categories that will be reimbursed at different levels. However, the overall budget for genetic testing has not increased, so healthcare providers have set a limit for how many tests they will cover each year, even though patient numbers at Radboud UMC have been on the rise. "Every year, we have quite a lot of patients that we don't get reimbursed for because we exceed the number," Veltman explained.
The exome test and other genetic diagnostic services are offered through Genome Diagnostics Nijmegen, the human genetics department's ISO 15189-certified clinical laboratory. Of the roughly 6,000 exomes the lab analyzed last year, about 1,000 came from abroad, including 800 from a single customer. The lab currently charges €1,800 ($2,000) for a single exome and €3,500 ($3,900) for a parent-child trio, according to its website.
While the exome data analysis and interpretation is performed in Nijmegen, the sequencing is not. Three years ago, the department, which is equipped with several Illumina NextSeq and Ion Torrent PGM sequencers, started outsourcing exome sequencing to BGI Europe in Copenhagen, Denmark, which offers services on Illumina platforms. At the time, this was more cost effective and allowed the Dutch team to focus on the data analysis, clinical interpretation, and reporting of the results. "It turned out to be very robust," Veltman said. "The outsourcing worked very well; we are very satisfied with that."
Last year, then, after the volume of exome diagnostics had picked up sufficiently to make in-house sequencing cost efficient again, the department decided to purchase the Revolocity system from BGI's Complete Genomics. This would not only have enabled it to perform rapid exome sequencing, for example, for patients in the intensive care unit, but also to explore whole-genome sequencing. However, BGI abruptly canceled the commercialization of the Revolocity last fall as part of a strategic shift, leaving customers who had ordered the system in the lurch.
Radboud UMC's lawyers are still in discussions with BGI to come to a solution, Veltman said, and an annex built to house the new in-house genome center currently sits empty.
The second most frequent test performed by the diagnostic lab is a hereditary cancer test for BRCA1 and BRCA2, which the lab switched from a Sanger sequencing to a next-gen sequencing assay last year that uses molecular inversion probe technology for target capture. Last year, it ran about 2,000 of these tests.
The lab also conducted about 600 noninvasive prenatal tests for fetal trisomies in 2015 as part of a Dutch study conducted in women with high-risk pregnancies that involved four laboratories. Next year, the Dutch government is expected to make NIPT available to all pregnant women in the Netherlands, Veltman said, and three genetic centers are preparing to deal with a dramatic increase in demand for the test, among them Maastricht UMC, which will take over testing from Radboud. In collaboration with a group in Leuven in Belgium, the Maastricht lab is also looking to develop a next-gen sequencing test for preimplantation genetic diagnosis, he noted.
Whole-genome sequencing is "where we see the future" of diagnostic sequencing, Veltman said, because it promises to increase the diagnostic yield by providing information on noncoding parts of the genome and structural variations that exome sequencing misses. "There is definitely benefit to going to genomes and going to long sequencing reads," he said.
Other areas of innovation the department is pursuing are deep exome sequencing, which can shed light on genetic mosaicism, and rapid exome sequencing for urgent cases. For the latter, the Nijmegen lab has already sequenced the exomes of a small number of newborns in the intensive care unit using the labs' in-house NextSeq 500 system with the aim of providing results within a week.
The Radboud team has already been exploring clinical whole-genome sequencing in a research setting, publishing a study in 2014 that showed genome data could boost the diagnostic yield for intellectual disability over exomes. Since then, the department has invested in bioinformatics infrastructure and would be able to handle data for a few thousand whole genomes per year. The plan, Veltman said, is to conduct additional clinical utility studies that explore the added value of genome sequencing, using funding from the Dutch government.
Offering diagnostic whole-genome sequencing, though, still seems a little further off in the future than he originally expected. "The problem is that the genome is still, for our diagnostic service, too expensive to offer as a routine test," Veltman said. The cost of genome sequencing through an outside provider and interpreting it in house still hovers around €2,000 to €3,000, he estimated, about twice the cost of exome sequencing, and has not dropped significantly in the past few years due to Illumina's virtual monopoly on whole-genome sequencing technology.
Because of that, and since the quality of exome sequencing has improved and it is still unclear how much genome sequencing will add diagnostically, "we still see, at this moment, that exome sequencing has a place in diagnostics for the coming few years," he said. Exome sequencing also remains diagnostically interesting because at very high coverage, it is able to detect mosaic mutations that might explain a patient's phenotype.
Over the next several years, though, there will likely be a transition period where genome sequencing, for example, will be offered to patients with negative diagnostic exome tests, he said. "Once genomes are cheap enough, and there is the capacity to run them fast and affordably, we'll move there."
The ideal test, he added, would provide a genome sequence with all structural variants and long reads as well as very deep coverage, but for that, technology still needs to improve.
However, the department already has plans to investigate long-read sequencing for clinical use and is looking to purchase a system that provides long sequence reads, for example, from Pacific Biosciences or from Oxford Nanopore Technologies. This would have potential applications in analyzing complex regions of the genome, such as HLA testing, as well as in repeat expansion disorders, Veltman said. The team has already conducted a number of small collaborative studies that involved PacBio data, which it would like to expand on.
In terms of other new genetic tests, Veltman said, the Radboud researchers are looking into conducting a proof-of-principle study for next-gen-sequencing-based pharmacogenomic testing at their hospital. The idea is to provide a targeted, rapid, and inexpensive test to patients who have been prescribed but not yet taken certain drugs.