While ethnicity-based screening remains the consensus approach for identifying disease-causing mutations, the American population is becoming increasingly admixed, meaning that genetic laboratories often have to run multiple panels on the same individual.
"Many different ethnic panels may be relevant for the same individual, and testing for each ethnicity-based panel individually increases cost substantially," said Alice Tanner, assistant professor of human genetics and director of the DNA Diagnostic Laboratory at Emory Genetic Laboratory.
Additionally, Tanner said that studies have shown that people may not always be aware of all aspects of their actual ancestry, and that carrier rates for many severe Mendelian diseases are being "found to be higher than previously suspected, and many are being identified outside of their historically known ethnic groups."
To address the limitations of the ethnicity-based panels available, Tanner and colleagues at EGL worked recently with TessArae, a Potomac Falls, Va.-based company that develops targeted sequencing assays based on the Affymetrix GeneChip platform.
The result of that collaboration is a pan-ethnic carrier screening (PECS) assay that can determine a person's carrier status for more than 90 single-gene disorders in one test. Tanner and colleagues at EGL and TessArae profiled PECS in a recent Journal of Molecular Diagnostics paper. Tanner is the lead author on the publication.
According to Tanner, EGL had been offering carrier screening for cystic fibrosis and for mutations prevalent in individuals of Ashkenazi Jewish ancestry for some time before deciding to develop an expanded carrier screen that would include more mutations and disorders found in the general population. She said that EGL decided to work with TessArae because of its targeted sequencing platform's "rapid time to result" of less than 24 hours, "low running cost as compared to all other sequencing-based methodologies," and EGL's desire to detect only clinically-relevant mutations and not variants of unknown significance.
She also credited TessArae's GeneCipher analysis algorithms, which "return DNA sequence as the result with the speed, efficiency, and high multiplicity of microarrays." GeneCipher translates the array signal intensities to genomic DNA sequence and calls the sequence generated at each allele as wild-type, heterozygous, or homozygous for every disorder, "eliminating the need for bioinformatics resources at EGL to analyze the results," Tanner said.
Pan-ethnic paradigm
According to Tanner, much of the carrier screening performed today takes place in obstetricians' offices, and is carried out using ethnicity-based panels recommended by the American College of Medical Genetics and the American Congress of Obstetricians and Gynecologists.
While some pan-ethnic panels have been introduced, Tanner said "it is important for end users to realize that not all pan-ethnic panels are created equally" and that the "variation in both disorders and mutations included on the various commercial panels is quite significant, which is why an understanding of how and why a panel's specific content was chosen is so important."
Tanner noted that pan-ethnic panels are already commonly used in newborn screening, but that the "pan-ethnic paradigm" has not yet moved into adult testing, though such a screen "could be considered more equitable in providing informed decision-making about reproductive options." She added that the "decreasing cost of technology is making pan-ethnic carrier screening possible by allowing for screening of many disorders at once."
EGL and TessArae's new array includes all of the most commonly tested CF mutations and many of the less common ones, while excluding rare mutations that do not significantly increase the detection rate, Tanner said. PECS also includes other severe childhood onset disorders such as those included in newborn screening programs and other metabolic and lysosomal storage disorders, she added.
According to Tanner, the disease-causing variants were chosen based on their prevalence in diverse ethnic populations and from EGL's database of mutations amassed from the global patient populations referred to EGL for clinical testing. PECS also excludes some disorders that are included in other commercially available carrier screening panels and target milder or later onset diseases or rare mutations that are restricted to small populations. She did not name the panels.
Matthew Lorence, senior vice president of sales, marketing, and business development at TessArae, told BioArray News that PECS has a number of attributes that distinguish it from other panels.
According to Lorence, PECS "emphasizes the newborn screen," including more mutations in more genes than are included in ACMG's currently recommended newborn screening panel. It also includes the most commonly tested cystic fibrosis mutations and many of the less common ones, while excluding rare mutations that do not significantly increase the detection rate. Lorence noted that PECS tests for the "most clinically relevant disorders, avoiding controversial ones like adult onset diseases or extremely rare mutations found only in highly specific populations."
Another factor in the array design was EmBase, EGL's database of genetic mutations "amassed from patient populations around the world" that make the content of PECS "highly unique," Lorence added.
Once the content was defined, TessArae designed the PECS array, and Affymetrix subsequently manufactured it. TessArae and EGL then validated the assay. An internal study of genomic DNA isolated from 26 consenting unaffected volunteers, representing the general population, is in progress, Lorence said, and a total of 76 PECS assays representing 38 different extractions in duplicate have been performed. Of the 316 positive calls in the 76 assays, 304 were concordant by Sanger sequencing with 12 being false positive calls, Lorence said.
"This is not unusual since the parameters embedded in the GeneCipher algorithms were intentionally set to call every locus either as mutant … or uncalled unless the data clearly support a wild type call," he said. "This minimizes the risk of a false negative result."
According to Lorence, TessArae customers continue to employ ethnic-specific panels in community screening events targeted to specific ethnic groups. He said that Jewish Genetic Diseases UK recently held such an event, for which TessArae's customer, Great Ormond Street Hospital, performed all of the assay testing.
Still, "most of our customers see the majority of growth in the pan-ethnic panel offerings because they provide a much more comprehensive approach that is applicable to the general population at the same cost as the ethnic-specific panels," said Lorence. "The beauty of the TessArae development platform is that customers can develop multiple custom panels to cater to the needs of various markets within their community," he said. "The customization attributes of the PECS assay allow TessArae customers to be responsive to different situations while keeping the screening revenue in their laboratory," he added.
Several TessArae customers are currently offering PECS or variations of it as a commercial carrier screen, Lorence said. These include Columbia University Medical Center in addition to Great Ormond Street Hospital. He described other customers as a specialized center in molecular medicine and a premier institute in the Gulf Cooperation Council, which are using the assay "in alignment with their newborn screening initiatives in the Middle East;" as well as a "world-renowned academic and leading health care system in New England," which is "currently validating their PECS assay and plans to offer the carrier screen to their community by this summer."
RT-PCR and NGS
TessArae's technology distinguishes it from other approaches to carrier screening detection, and Lorence highlighted some of the differences between TessArae's array-based method and RT-PCR- and next-generation sequencing-based tests.
In contrast to RT-PCR, Lorence said that PECS "provides the genomic sequence of the detected loci plus several bases of flanking sequence, not just increased fluorescent signal." He noted that "most genetic disease testing has been based on Sanger sequencing, and that the data from PECS is therefore comparable to that method.
"PECS also generates sequence from the wild-type loci, as opposed to simply reporting a negative result, which might be due to amplification failure in an RT-PCR-based assay," Lorence said. He added that PECS is "not affected by the presence of unknown polymorphisms in the vicinity of the targeted locus, which may lead to negative or weakly-positive results in an RT-PCR-based assay."
For NGS, Lorence discussed a number of weaknesses, including sequencing of insertion and deletion (indels) mutations and the false positive variant rate in these types of mutations. "All NGS-based assays experience some amount of dropout, requiring some regions to be routinely completed with Sanger sequencing, adding to the cost of the assay," said Lorence. He also reiterated that PECS provided a faster sample-to-result time than sequencing, and that it does not detect variants of unclear clinical significance.
While PECS maintains these advantages over other methods, Madhuri Hegde, executive director of EGL, told BioArray News that the lab is in the process of transferring the assay to Illumina's MiSeq platform, as most of its genetic testing has already been moved to the MiSeq, and the pan-ethnic carrier screen will fit better into EGL's workflow if it is also available on that platform.
"NGS algorithms for indels are improving but the indels will need Sanger confirmation for some time for reporting," said Hegde. "EGL has developed [an] in-house informatics pipeline for panels and exomes so the transition to NGS for this panel is not difficult for us," she said.
"Most of EGL's molecular assays are already on NGS, so this is more of a lab workflow issue," said Lorence. "PECS is highly targeted and validated with respect to the indels that it was designed to detect," he added. "So, we are not speaking of unknown or untargeted indels, but targeted indels."In PECS, he continued, the "targeted indels, whether one or more bases, or hundreds of bases as is the case for nine of the deletions, are accurately and reproducibly detected, so the argument against NGS is not applicable."