This story has been updated to clarify that the Parabase test's correct diagnosis rate when phenotype was considered was specifically 32 out of 34 samples and to add additional comments by the company.
NEW YORK (GenomeWeb) — Researchers from Parabase Genomics with collaborators from several pediatric medical centers have published a small retrospective study of the company's targeted next-gen sequencing test for the diagnosis of genetic disorders in newborns, which the firm expects to launch in 2015.
The study, which appeared online in Genetics in Medicine last week, showed the test was able to correctly identify a known disease mutation in 27 of 36, or 75 percent of blinded retrospective samples. When the sequencing results were analyzed in light of phenotype, the number of correct calls rose to 32.
Parabase, which entered the newborn sequencing market in 2013, has been offering a diagnostic exome test for rare genetic diseases, called RareDx, since the beginning of this year. Physicians can also currently order an in silico-filtered exome test targeting the same genes as will be covered in the company's rapid targeted test, when it is launched clinically next year.
In the company's study, the Parabase researchers looked at both the current filtered exome panel and the targeted test it plans to launch next year, using a set of samples from Amish and Mennonite children from the Clinic for Special Children in Strasburg, Pa.
In eight samples, the researchers evaluated their filtered exome sequencing approach, which was completed for the eight samples sequenced in parallel, in 105 hours from sample isolation through interpretation. It correctly identified the known disorder in all eight samples.
The team then tested a total of 36 clinical samples, including these initial eight, using their 126-gene targeted panel. The researchers compared their calls to the Sanger-confirmed disease status of the samples in two different scenarios, one in which there was no phenotypic or clinical information to aid in analysis, and one in which phenotype and clinical reports were considered.
Blind to clinical information, the panel yielded a correct diagnosis in 75 percent, or 27 of the 36 cases.
The company has reported 99.8 percent analytical sensitivity for its filtered exome panel.
According to the authors, complications behind the 75 percent clinical concordance rate included inability to distinguish causal variants from other mutations, variant calling errors, lack of gene coverage, and the presence of compound heretozygotes with an intronic second mutation. With the aid of phenotype to help resolve issues in distinguishing causal variants, the number of correct calls rose to 32 out of 34.
In the study, the authors also reported being able to sequence DNA extracted from dried blood spots with similar performance to DNA extracted from larger blood sample volumes, something crucial in the neonatal setting where patients have precious little blood available.
"Most next-gen sequencing services require 10 or maybe down to 2 ml of blood, and one of the things we learned from [neonatologists] is that it's impractical to collect large amounts of blood from a newborn," said Andy Bhattacharjee, Parabase's president and chief scientific officer.
"This is one of the first validation studies focused on small samples, which essentially is 50 micro liters of blood as dried spots, where we looked at that head to head with standard blood samples and showed that the answers are roughly the same," he added.
Stephen Kingsmore, director of Children's Mercy Hospital's Center for Pediatric Genomic Medicine, where he has pioneered a rapid whole-genome sequencing method for NICU infants called STAT-Seq, told CSN in an email that the ability to successfully sequence from dried blood spots for neonatal applications has become widespread, a development which will likely be a boon to the many groups advancing or investigating NGS as a tool to diagnose NCIU infants, or potentially screen healthy infants.
According to Bhattacharjee, "most [other groups that have published on using dried blood spots] have boiled the spots to retrieve the DNA and which is single stranded and this works for amplicon strategy of sequencing but not great for adapter ligation or enzymatic incorporation for which you need double stranded DNA."
Another important demonstration of the Parabase paper, Bhattacharjee said, was a clinically useful turnaround time of 105 hours, or about four days, from sample isolation through analysis.
Currently, the company's exome sequencing has a turnaround time of four to six weeks. The more targeted filtered exome — which includes 126 genes related to newborn screening disorders, 84 hearing loss genes, seven genes related to failure to thrive disorders, eight hypotonia genes, and two hepatosplenomegaly genes — takes between two and four weeks.
With the targeted test demonstrated in its recent study, Parabase is aiming to offer a service that can be useful in the clinical timeframe of the neonatal intensive care setting, where rapid diagnosis, in days, not weeks, can mean the avoidance of thousands of dollars a day in hospital care.
"To me, these results are the first step needed to validate and to show that this process works with a tiny amount of blood and a turnaround in the clinically relevant time window — in a matter of hours to days rather than weeks to months. At that pace, we can use results immediately and it can change the care and outcome of the patient," said Richard Parad, one of the study's co-authors, a physician at Brigham and Women's Hospital, and a member of Parabase's scientific advisory board.
According to Bhattacharjee, Parabase is working on reducing its turnaround time to about a day. "Right now it's around 105 hours, compared to weeks for most tests in the NGS space. But we think we can reduce it down to a one-day turnaround, though that's still in the development stage," he said.
In the clinical setting, however, Parabase also confirms all positive results with Sanger sequencing, which adds another several days till a final, clinically-confirmed result.
Children's Mercy's Kingsmore expressed some skepticism of the utility of a [several hundred]-gene panel in the NICU setting, where, he told CSN, there are more than 5,000 genes that can cause disease.
Kingsmore's whole-genome STAT-Seq effort at Children's Mercy, currently offered in a research setting in anticipation of future CLIA approval, showed in a proof-of-principle study that it could provide an answer to patients in around 50 hours for an estimated cost of $13,500.
In Bhattacharjee's view, "there is not enough justification to use a whole genome test [at such costs]" for all NICU patients.
Parabase previously told CSN that it intends eventually to expand its targeted test to include around 500 genes, which would capture around 80 percent of newborns in the NICU with a suspected genetic disorder. Those testing negative with unexplained clinical symptoms could then reflex to the company's exome test.
Parabase's clinical exome test currently costs $5,000 and its filtered clinical exome tests is priced at $4,500.
In addition to the NICU, Parabase has said it is interested in advancing its targeted test in the newborn screening setting — an application they highlight in the recent paper — as a follow up to abnormal screens, or potentially as a replacement for current screening tools if costs can be lowered enough.
Parad also highlighted the potential utility of more widespread use of sequencing in the NICU for phenotyes that are more common, and for which the influence of genetics is not as well known, for example, symptoms like hyperglycemia.
The overall utility of broad next-gen sequencing — either targeted panels, or whole-exome or whole-genome sequencing — and whether it improves outcomes, efficiency, or economics compared to current testing paradigms is still unknown.
As part of a separate project, one of four funded under the National Human Genome Research Institute's Newborn Sequencing in Genomic Medicine and Public Health program, Parad and other colleagues at Brigham and Women's are trying to answer some of these questions by randomizing a larger cohort of children, both in the NICU, and healthy infants, to receive either NGS or standard newborn screening and follow-up care.
But, Parad said, the bar to clear, at least in the NICU, doesn't look very high. "The NICU can cost maybe $3,000-$5,000 a day, so if you can cut even a day off of that hospitalization you've paid for the genetic testing very easily."
Parabase Chief Commercial Officer Gabor Bethlendy told CSN in an email that the company is currently in the process of converting customers of its filtered exome to the full targeted panel with the aim of launching fully in 2015 with this new, shorter-turnaround test.
He said Parabase is also in the process of completing the first close of a sale of an undisclosed amount of Series B preferred shares. Hong Kong-based SXE Ventures is leading a group of investors in the round, which also includes MIT Angels and Walnut Partners.
The company also plans to move on Oct. 1 into a new 3,300-sqare-foot clinical lab and office facility in South Boston.