By Julia Karow
Baylor College of Medicine has opened a Whole Genome Laboratory that is initially offering a clinical whole-exome sequencing test for the diagnosis of genetic disorders and plans to add whole-genome sequencing in the "near future."
The laboratory, a joint effort of Baylor's Human Genome Sequencing Center and the Medical Genetics Laboratories of the department of molecular and human genetics, is among the first in the country to bring genome-wide next-generation sequencing to the clinic.
The Medical College of Wisconsin, for example, has been using whole-genome sequencing to diagnose children with rare genetic disorders (CSN 8/24/2011), and the Laboratory for Molecular Medicine at the Partners HealthCare Center for Personalized Genetic Medicine plans to introduce clinical whole-genome sequencing next year (CSN 6/15/2011). Also, Ambry Genetics recently launched a clinical diagnostic exome test for patients with severe unexplained conditions (CSN 10/5/2011), and GeneDx said it plans to introduce a similar test early next year (CSN 10/19/2011).
Baylor's CLIA-certified WGL, currently housed in the department of molecular and human genetics, has already received its first patient samples and is ready to take orders from physicians within and outside the US. It currently charges $9,000 for the test, a price that is likely to come down with technical improvements. The test has a turnaround time of about 90 days.
The WGL has already received prior authorization from health insurers for several patients.
According to Arthur Beaudet, department chair and professor of molecular and human genetics, Baylor has been setting up the lab over the past six months. While Beaudet and Richard Gibbs, director of the HGSC, have been spearheading the effort, the lab is run by Christine Eng, who is also the director of the DNA Diagnostic Laboratory within the Medical Genetics Laboratories, and a professor of molecular and human genetics.
The combined expertise of a genome center and an existing diagnostic lab helped get the WGL off the ground "relatively expeditiously," Beaudet said, although understanding each other's culture has sometimes been a challenge. "Those of use who are more clinical and doing routine molecular diagnostics have not really appreciated all the challenges of having a good pipeline to analyze the data, and those who have been running a genome pipeline now see that there are a lot of things that come up when you are going to validate everything, have standard operating procedures, and sign out [the results] by board-certified clinicians."
While physicians can order the exome sequencing test on any patient, including healthy individuals, Beaudet said the lab expects the majority of samples to come from children with unexplained disorders that are presumed to be genetic. These include intellectual disability, autism, epilepsy, and birth defects.
Gary Clark, chief of the neurology service at Texas Children's Hospital, estimated, for example, that a quarter of the patients in the hospital's clinic for pediatric neurology suffer from a genetic disorder that has not been previously described or for which the genetic cause is unknown.
Most of the patients will already have had other diagnostic tests performed to identify the cause of their problem — for example copy number arrays, or test for specific genes — with no success.
And while it is currently unknown for how many of these patients the exome sequencing test will provide a diagnosis, Beaudet said that some publications suggest a "double-digit number" percentage.
In some cases, the mutation found by the test will not be novel, but so rare that it was overlooked. "Maybe 10 to 20 percent will be cases where if you really looked carefully at the phenotype, and if you had taken advantage of every research sequencing test that's available in the whole world, you probably could have made that diagnosis," Beaudet explained. "Those will be the ones that will be the easiest."
The main benefit of a molecular diagnosis will be effective genetic counseling of the family, for example about the recurrence risk of the disorder in other children. It would also put an end to further testing, reducing healthcare costs and anxiety for the family. Only in a small fraction of patients — maybe 5 percent — will the knowledge lead to a change in treatment, Beaudet estimated.
Like existing tests, such as chromosomal analysis and copy number arrays, the exome sequencing test will not uncover all disease-causing mutations, either because they are located outside the exome, or because the test fails to cover portions of the exome. But even so, "we think we will make many diagnoses that would not be made otherwise," he said.
Currently, the WGL is equipped with two Illumina Genome Analyzer sequencers, on which it has validated the test. It plans to switch over to the Illumina HiSeq soon, which will drive down the cost of the test, and is in the process of validating it on that platform.
To capture the exome, the WGL uses a custom-built NimbleGen capture chip, VCRome 2.1, that targets 43 megabases of sequence and captures 99.5 percent of its target in a typical experiment, according to Donna Muzny, director of operations of the HGSC.
After analyzing the patient's exome data, the lab validates potentially disease-causing variants by Sanger sequencing or an equivalent method. The lab also analyzes these mutations in the patient's parents in order to determine which parent passed on the defect, or whether it occurred de novo.
If that analysis provides no answer, the lab may also sequence the exomes of the parents, allowing it to focus on de novo missense mutations in the child.
All results are reviewed by experts who are certified by the American Board of Medical Genetics, who will provide an interpretation of the results "that we want to be easily understood by the clinician and the family," Beaudet said. "I think it takes a large team of clinicians to really do this well."
In fact, about half of the test's cost is currently associated with the interpretation, and a board-certified geneticist could "easily spend a few hours working with the data, trying to develop an appropriate clinical interpretation, and sign it out," he said.
While most exome sequencing services merely generate lists of variants, he said, Baylor will be among the first to "say which variants we think are having a clinical impact, and which are not."
In its report, the lab will focus on variants that appear to explain the patient's phenotype, but it will also include pharmacogenomic and disease carrier results. "We don't want to overwhelm clinicians and patients with huge amounts of data, but we will have other data that could be useful and would potentially become part of the medical record," Beaudet said.
The lab has not discussed whether it would not report on any particular gene because of intellectual property issues, such as the BRCA genes, he added.
The test has a turnaround time of about 90 days, which he said is a "reasonable target" since the results are not urgent in most situations.
For some patients where all exome results come back negative, the WGL may move on to whole-genome sequencing "before too long," Beaudet said, but it is not clear yet when that will start.
It is also too early to say whether the lab would outsource whole-genome sequencing to a company like Complete Genomics or Illumina or provide it in house.
"At the moment, whole-exome is a reasonable compromise in terms of the cost and the available answers in the data, but a lot of people believe that whole-genome will supplant whole-exome relatively quickly," he said.
WGL is offering clinical exome sequencing as a laboratory-developed test. In validating the test for use in its CLIA lab, the lab entered "pretty novel territory," Beaudet said, because it is so different from existing tests.
"If you taken an old-fashioned test, you might want to have certain positive controls with many or all of the expected mutations you might see," he explained. "Here, there is no way you can have positive controls for all the mutations that exist in the genome."
While it is easy to confirm mutations found by other methods, and thus eliminate false-positive results, it is "much less clear" how to deal with false negative results.
Beaudet declined to discuss details of the validation at this time because "somebody might think that what we decided was adequate validation they might not see as adequate."
Demand for the test is hard to predict, and will depend on how often it provides a molecular diagnosis, how actionable that information is, and how often insurance pays for it, but Beaudet said that analyzing 100 samples per week is "within the realm of possibility."
Before running the test, Baylor requests prior authorization from health insurance companies, and has already been successful in several cases. Given that some diagnostic gene panels currently cost more than $10,000, Beaudet said he believes insurance firms "will be relatively accepting" of the exome sequencing test, which would cost less and provide more data.
Since presenting the test at the International Congress of Human Genetics and American Society of Human Genetics meeting last month, the lab has seen "a lot of interest," he said, and it is prepared to increase its capacity with additional sequencing instruments and new personnel for the interpretation as needed.
Have topics you'd like to see covered in Clinical Sequencing News? Contact the editor at jkarow [at] genomeweb [.] com.