NEW YORK (GenomeWeb) – The Laboratory for Molecular Medicine at Partners HealthCare Personalized Medicine has launched a clinical exome sequencing test and has added new and updated gene panels to its portfolio of next-gen sequencing diagnostic tests.
Over the past year, the Cambridge, Mass.-based molecular diagnostic lab has also switched its gene panels from the Illumina HiSeq to the MiSeq platform, reducing turnaround time.
In addition, the lab has started to bill health insurance providers directly for exome and genome sequencing tests for patients within the Partners HealthCare System and plans to pursue direct billing for outside patients in the future.
The exome sequencing test, which the LMM performs in house on the HiSeq platform, complements the genome sequencing test it launched in 2013. Sequencing for that test is outsourced to Illumina's CLIA laboratory.
The exome test is 30 to 50 percent less expensive than the genome test – a single exome costs $6,500, compared to $9,000 for a single genome, and exomes for a family trio cost $9,500, compared to $18,000 for trio genomes. All prices include the clinical interpretation.
Like the genome test, the exome has a turnaround time of about 16 weeks, which can vary depending on the complexity of the case and the lab's workload.
Secondary findings are reported the same way for the two tests – unless a patient opts out, the exome test assesses variants in more than 350 genes, which include those recommended by the American College of Medical Genetics and Genomics.
According to Heidi Rehm, the LMM's director and an associate professor of pathology at Harvard Medical School, the lab has run genome and exome tests both as primary diagnostic tests and as follow-ups to more targeted tests.
Some conditions, such as hearing loss, are often associated with copy number changes, which are more difficult to detect from exome data than from targeted sequence data, she explained. For other patients, often those with ill-defined clinical features that do not point to a specific disease and associated gene panel, it makes more sense to start with an exome. "Both are happening today," she said.
All variants are annotated in an automated way as part of the exome and genome pipelines, allowing the lab to filter them according to different criteria, and to zero in on specific variants more quickly. This is different from the targeted panels, where each variant is still assessed one by one, Rehm said.
Overall, the LMM has run fewer than 25 exome sequencing tests so far, and a total of about 130 genome sequencing tests. The genome test has been supporting the MedSeq project, a National Institutes of Health-funded research study to explore the use of genome sequencing in the clinic.
Rehm said reimbursement for the exome and genome tests "is still a work in progress" but the lab has had success in obtaining approval for coverage.
While preparing to launch the exome sequencing test, the LMM transitioned all its gene panel tests from the HiSeq to the MiSeq, thus freeing up capacity on the HiSeq for the exome test. In the process, it expanded and updated many of the panels and added a new panel for myopathy disorders, called MyoGene Panel, that covers 43 genes.
Running the panels on the MiSeq, which has a quicker workflow than the HiSeq, will significantly decrease their turnaround time, Rehm said.
To improve the accuracy of variant interpretation, the LMM is increasingly relying on data sharing with other diagnostic laboratories, mainly through the ClinVar database, which Rehm is involved in building through the NIH-funded ClinGen project.
For example, she said she now routinely participates in e-mail exchanges with directors from other laboratories to discuss controversial variants, and they often come to a consensus, sometimes with immediate consequence for the patient in question. In one such discussion earlier this month between the LMM, GeneDx, and Emory Genetics Laboratory, for example, they came to the conclusion that a variant that had been called pathogenic by one lab but a variant of unknown significance by another lab was probably not the underlying disease cause of a patient who was tested by the third lab, which meant that the patient should discontinue an expensive enzyme replacement therapy.
"Whether [the system] creates efficiency as of yet I would not say is clear, because in some ways, the more resources and information you have, the longer it can take to evaluate a variant," Rehm said. "But I do think that the improved knowledge, and therefore the improved accuracy of variant interpretation, by being able to come to consensus on variants and discuss all of the evidence, is in the best interest of the patient."
Over the two years of its existence, the ClinVar database has become "a household term" among genetic testing labs, Rehm said, and is developing into a database of reference for variant interpretation. But because it is variant-centric, it does not capture other clinical information about a patient that might be important for the test interpretation.
For that reason, Rehm and her colleagues are working on a password-protected system to share case-level data through ClinVar. "If I could look at the world in 10 years, we would have this centralized variant interpretation repository, which would be ClinVar, [and] then connect this to other resources that contain genotypic and phenotypic case-level information that helps inform the variant understanding, and that would be more likely a federated network," she said.
Another data-sharing system Rehm and other laboratories have been working on is the Matchmaker Exchange, which will allow users to exchange genotypic and phenotypic profiles from unsolved exome and genome sequencing cases. Participants in that project met last week and are planning to launch the exchange later this year after working out existing bugs and building a user interface, she said.
In general, a big challenge of launching new genomic tests for clinical use has been "a lack of consistency across sites, ranging from differences in the technical to the interpretive side," Rehm said. This is not unexpected as technologies, bioinformatics tools, interpretation methods, and data sharing continue to change. While it is difficult to settle on standards, "we must try to do what we can to establish best practices while still enabling innovation," she said.
Asked about her opinion on the FDA's proposed regulation of laboratory-developed tests, she said that the topic is complex, and that she is in favor of partial regulation that targets the highest risk categories of tests, such as direct-to-consumer tests. "However, much is in the details of implementation, and we are all still learning how increased regulation might happen," she said.