By Julia Karow
Next-gen sequencing for clinical applications and medical research were big themes at this year's Advances in Genome Biology and Technology meeting in Marco Island, Fla., reflecting the technology's continued move into medicine over the last year.
Three conference sessions had a medical theme this year, focusing on the clinical translation of genomics, medical sequencing and genetic variation, and cancer and transcriptomes.
And, probably for the first time at this meeting, conference attendees discussed the term "CPT code" — the number assigned to medical and diagnostic services by the American Medical Association to define reimbursement by insurance — in connection with next-gen sequencing-based tests.
"Last year, there were just some very preliminary studies that were really at the interface between the basic sciences and the clinic," said Elaine Mardis, co-director of the Genome Institute at Washington University. "What we saw this year … is a focus on groups that are coming up with very defined content, looking very carefully at inherited diseases, and beginning to really transition many of the single-gene or small panel gene tests into now multiple-gene tests that really harness the capabilities of next-generation sequencing data."
NGS Gene Panels
Heidi Rehm, director of the Laboratory for Molecular Medicine at the Partners Healthcare Center for Personalized Genetic Medicine in Boston, presented one of the more applied examples of sequencing in a clinical context.
Rehm's lab, which focuses on genetic testing for inherited cardiovascular disease, cancer, and hearing loss, recently merged several existing gene panel tests — some of them resequencing array-based — into a pan-cardiomyopathy panel that contains 46 cardiomyopathy genes and uses a combination of next-gen sequencing and Sanger sequencing.
The lab also recently converted its resequencing array-based OtoChip hearing loss test into a next-gen sequencing-based OtoGenome test, and plans to expand it from 19 to 73 genes this spring.
One of the challenges with these tests, Rehm said, has been the need to confirm results from next-gen sequencing by Sanger sequencing, and to fill in missing areas that are not covered, also using Sanger.
The need to interpret every variant identified in the target region has led to a jump in novel variants, which take considerable time to interpret through searches of databases, literature, and the use of functional prediction tools. For example, even after testing thousands of patients for hypertrophic cardiomyopathy, she said, about 17 percent of the identified mutations are novel, she said.
Rehm also called for a new, publicly available and curated clinical variant database, populated with data submitted by clinical laboratories. If a grant under submission at the National Institutes of Health is funded, she will lead the development of such a database, initially focused on eight disease areas.
Developing software to deliver clinical reports to physicians has been another focus of her team's work. Eight years ago, they began building a system to track knowledge about genetic variants and support laboratory reporting. The system, called GeneInsight Suite, is already used by a number of clinical labs. A clinical interface for this tool was added two years ago that allows doctors to receive patient alerts on variants via e-mail and obtain updated reports within the GeneInsight Clinic interface. Rehm said that a manuscript detailing her lab's experience with this system has been accepted for publication in a journal and will appear shortly. The software has also been registered with the US Food and Drug Administration as a class 1 medical device.
Besides moving gene panel tests to next-gen sequencing, the LMM is collaborating with genetics clinics at Brigham and Women's Hospital and Massachusetts General Hospital to develop a whole-genome sequencing interpretation service, and is currently outsourcing the sequencing to Illumina's CLIA laboratory (CSN 6/15/2011).
Diagnosing Severe Childhood Disorders
Darrell Dinwiddie, from the Center for Pediatric Genomic Medicine at Children's Mercy Hospital in Kansas City, presented another promising example of the clinical use of next-gen sequencing. He and his colleagues, led by Stephen Kingsmore, the center's director, have developed a test to diagnose more than 600 severe childhood diseases — all Mendelian disorders — using targeted resequencing (CSN 8/9/2011).
The test, originally developed by Kingsmore's team at the National Center for Genome Resources, covers about 8,400 genomic regions from 526 genes, as well as the mitochondrial genome. It is currently available under a research protocol to physicians at Children's Mercy Hospital, where it is validated in a CLIA and CAP-approved lab. Its turnaround time is about four weeks, and it will cost on the order of $750.
It does not include carrier testing and testing for adult-onset diseases as there would be no immediate benefit for the children tested, Dinwiddie explained.
Generating about 4 gigabases of data for each sample on the Illumina HiSeq platform results in at least 16x coverage for 98 percent of the targets, which are selected by Illumina's TruSeq capture method. The test's sensitivity for substitutions, insertions and deletions, splicing, gross deletions, and SNPs is more than 97 percent, and its specificity is 100 percent.
Insertions, deletions, and copy number variants account for between 5 percent and 20 percent of all disease-causing variants, Dinwiddie pointed out, but are difficult to find, and the test uses a two-step method to detect them.
To interpret the results, the researchers map clinical symptoms to genes and diseases that match these symptoms.
They also classify the identified variants into categories defined by the American College of Medical Genetics, ranging from previously reported variants that are a recognized cause of a disorder to variants that are unlikely to cause a particular disease.
The researchers are in the midst of a clinical validation of the test, which is nearly completed. For the first phase of the validation, they analyzed 96 samples that included cell lines from Coriell and HapMap samples. Phase two, a blinded study, involves 384 samples with known disease-causing mutations from Children's Mercy Hospital, and phase three comprises 110 samples with a known and 110 samples with an unknown molecular diagnosis from collaborators in Germany.
As an example of a successful use of the test in patients who received it under a research protocol, Dinwiddie presented the case of two siblings, age 5 and 9, who had undergone five years of testing that cost more than $23,000 but did not provide a molecular diagnosis.
Using their targeted sequencing test, he and his colleagues were able to identify known mutations in the aprataxin gene, which causes ataxia, and confirmed these results by Sanger sequencing. Both parents turned out to be carriers of the disease.
The molecular diagnosis even resulted in a new treatment for the children: since mutations in the aprataxin gene result in a deficiency of coenzyme Q10, the children now receive this as a medication. While their doctors are still working on the correct dosage, Dinwiddie said that the children are making "progressive improvement."
On a research basis, Kingsmore's team has also been exploring the use of "emergency" whole-genome sequencing in newborns.
During an Illumina user meeting held at the conference, Kingsmore reported analyzing the genomes of two newborns with life-threatening genetic diseases in a neonatal intensive care unit.
Molecular diagnoses are almost never made for these patients, he explained, and doctors need to make quick decisions on their management, mostly based on their symptoms.
The genomes of the two patients, including their mitochondrial genomes, were sequenced by Illumina at high coverage on its new HiSeq 2500 sequencing system, which generated results within 50 hours, starting from sample preparation to interpreted results.
For the analysis, the researchers focused on a number of candidate genes that were suggested by the patients' phenotype and symptoms, with the aim to rule out a subset of treatable conditions where knowing the presence of a mutation could save the children's lives.
They assembled a variety of software tools, enabling them to come up with a "rapid diagnostic answer." As in the targeted resequencing test, they did not look at carrier status or adult-onset diseases.
The researchers were unable to identify mutations that were either known to cause disease or novel and likely to cause disease, and both children eventually died. But the genome analysis was at least able to rule out certain disease causes, thus reducing unnecessary testing and the use of empirical therapies.
Positive results could have suggested other diagnostic tests for confirmatory testing, and would have been useful for prognostic assessments, genetic counseling of the parents, and targeted therapies in a subset of patients.
Thus, the study provided proof of concept for real-time genome sequencing to provide relevant medical results in life-threatening situations where a monogenic disease is suspected. A side benefit, he added, is a complete pharmacogenetic profile that may alter treatment choice.
Have topics you'd like to see covered in Clinical Sequencing News? Contact the editor at jkarow [at] genomeweb [.] com.