By Monica Heger
This story was originally published Aug. 5.
Children's Mercy Hospital is in the midst of a clinical trial to validate a sequencing-based diagnostic assay for 592 childhood diseases that it expects to complete by September, offer to physicians within the hospital by October, and launch more broadly by 2012.
The test will be the first product launched through the recently established Center for Pediatric Genomic Medicine at the Kansas City, Mo.-based hospital. The new center aims to focus on five areas of pediatric genomic medicine: the diagnosis of recessive childhood disorders, newborn screening of rare diseases, preconception screening, prenatal diagnostics, and forming an orphan disease treatment consortium.
The center is using the Illumina HiSeq 2000 and is evaluating an Ion Torrent PGM. It will offer its tests through the hospital's CLIA-certified, CAP-accredited laboratory.
Currently, the team is validating the diagnostic test in a clinical trial of around 1,000 previously diagnosed clinical samples on the HiSeq, using Agilent's SureSelect in-solution enrichment kit.
The test was originally developed by the National Center of Genome Resources, but is moving to Children's Mercy Hospital for the clinical launch. Darrell Dinwiddie, formerly at the NCGR, but now the director of lab operations at the Center for Pediatric Genomic Medicine, gave an update on the test and the newly created center at last week's Next-Generation Sequencing and Genomic Medicine Summit in San Francisco.
Dinwiddie said that the test is being launched through Children's Mercy Hospital because NCGR is more of a research institute and does not have the clinical infrastructure of a hospital. "We need to be in a CLIA lab, have access to patients and physicians — things you couldn't do in a research setting," he told Clinical Sequencing News.
Aside from Dinwiddie, several other members of the NCGR team that developed the test have also moved to Children's Mercy Hospital, including Stephen Kingsmore, who will direct the Center for Pediatric Genomic Medicine; and Neil Miller, who will head the informatics team.
Twenty physicians representing every specialty at the hospital are serving as primary points of contact for the hospital's doctors within the center. Additionally, the center employs 12 genetic counselors and seven MD- or PhD-level clinical geneticists, as well as several bioethics specialists.
The hospital has contributed around $5 million in funding for the new center, including $3 million in lab equipment (including sequencers), $1 million in computing, and another $1 million in start-up funding.
The test for 592 recessive childhood diseases was originally developed as a carrier screen, and the NCGR team published a validation study testing different sequencing and enrichment technologies of the test last year in Science Translational Medicine (IS 1/18/2011). Now, however, the plan is to initially offer the test as a diagnostic for children and later as a preconception carrier screen.
One reason for the switch is that a diagnostic test will allow the team to incorporate phenotype information when interpreting the results to help determine which variants are relevant. That additional data will then be used to strengthen the test's prediction ability when launched as a preconception carrier screen.
Additionally, launching the test as a diagnostic for children could have an immediate impact on patient care. As many as 30,000 patients in the US could receive the test each year, and Dinwiddie estimates that about 90 percent of those patients could receive a disease diagnosis, at a cost of $600 per test. This would be a huge improvement over the current method of diagnosing rare childhood disorders, which relies on tests that focus on specific genes or syndromes. Not only is this approach costly, with a price tag of up to $10,000 per test, but only about half the patients receive a definitive diagnosis, he said.
"Timely diagnosis in affected children is important, even if treatment doesn't currently exist," said Dinwiddie. Aside from "eliminating a patient's diagnostic odyssey," a diagnosis can help prioritize investigational treatments or rule out other diseases and unnecessary treatments. It can also provide information about risk to other family members.
[ pagebreak ]
The test will sequence known disease genes from 592 recessive childhood disorders. Around 2.6 million nucleotides will be sequenced to an around 20-fold coverage. It has an accuracy of 98.8 percent, a sensitivity of 94.9 percent, and a specificity of 99.99 percent, Dinwiddie said.
The test identifies both known and unknown variants, although not all unknown variants will be reported to the physician. "We're spending a lot of time on deciding how to report unknown variants," said Dinwiddie.
The team is following guidelines set by the American College of Medical Genetics. In addition, it is using a prediction database to help determine if the mutation is deleterious, looking at information on the variant's frequency in the population, the conservation of the particular amino acid, and trying to determine whether the variant's predicted function relates to the patient's clinical presentation.
For instance, said Dinwiddie, if an unknown variant is found in a gene related to neurodevelopment, but the patient's clinical presentation is immunodeficiency, then the variant is likely not pathogenic.
Dinwiddie said he expects the clinical trial to be completed in September. The test will become available for the hospital's physicians by October, and more broadly available outside the hospital by 2012. The test costs around $600, which is currently covered by the hospital, although the team is working with insurance companies on a reimbursement strategy.
Turnaround time for getting results back to physicians will be four weeks in the first year, but is expected to decline to two weeks by the third year. The team will also increase the number of diseases it tests for to 1,200 by 2013, at which point the cost of the test is expected to fall to $300.
Aside from the targeted sequencing test, Dinwiddie said the center has institutional review board approval to do exome sequencing in certain patients — specifically, for patients with an unknown disorder for whom a "long list of previous genetic testing" has been done, and "particularly if we believe novel variants might be involved."
A committee — comprising genetic counselors, the patient's physician, bioethicists, and the 20 physicians that are the primary points of contact within the hospital — meets to discuss prospective candidates for exome sequencing. If the team decides to proceed, the appropriate consent and assent forms are obtained, and after the sequencing and analysis are done, the committee reconvenes to discuss the results, how confident they are in the results, what type of validation should be done, and what results to return to the physician.
Going forward, Dinwiddie said that the Center for Pediatric Genomic Medicine will also be looking to develop tests for newborn diagnostics, and eventually prenatal diagnostics.
It is also in the process of evaluating its Torrent PGM for clinical applications. The PGM would be particularly useful for smaller gene panels, or in cases where fast turnaround time is critical, he said.
The team will continue to run the 592 disease test on the HiSeq because of the large amount of sequencing for each test, and also because the team can multiplex up to 96 samples. But for applications like newborn testing, the PGM could be ideal, said Dinwiddie, especially in situations where "a diagnosis is needed in the first days of life in order to start treatment and save the patient's life."
Additionally, the PGM could be useful for screening for drug metabolism markers to figure out if a specific treatment will be effective and non-toxic in a patient.
Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at mheger [at] genomeweb [.] com.