Affymetrix has provided more information about how it achieved US Food and Drug Administration clearance for its CytoScan Dx Assay chromosomal microarray analysis offering.
The Santa Clara, Calif.-based array maker in January gained regulatory backing for the test as an in vitro diagnostic, the first and only vendor to do so for an array-based assay. CytoScan is already widely used in CLIA-compliant laboratories as a laboratory-developed test, but the firm believes that FDA clearance will help it to win over new customers, better negotiate with payors, and enter foreign clinical markets, according to Eric Fung, vice president of R&D clinical applications at Affymetrix.
"I think that people who have already adopted cytogenetic arrays in their practice will actually convert over to the CytoScan Dx Assay for a couple of reasons," said Fung. "The most obvious is that because the CytoScan Dx Assay has a level of validation behind it that has been vetted by the FDA, that can help a laboratory that is doing things in an LDT format [convert] to an IVD format, and that might be from a compliance perspective to their advantage."
Fung spoke with BioArray News at the American College of Medical Genetics and Genomics annual meeting, held late last month in Nashville, Tenn. Affymetrix held a workshop at the conference, where the path to clearance as well as clinical validation data was discussed.
The story of how CytoScan, which relies on an array containing 2.6 million copy number markers, passed FDA muster is an interesting one, according to Fung. Since 2009, when the FDA made vendors aware of its desire to increase its oversight of their chromosomal microarray platforms, there have been many questions about how best to evaluate the accuracy of an assay that provides a high-resolution look across the entirety of an affected individual's genome, especially in a context where previously cleared array-based tests had detected lower numbers of analytes associated with certain conditions, such as drug metabolism or cancer recurrence.
Ultimately, Affymetrix submitted the assay for evaluation as part of FDA's de novo classification process, a regulatory route for what it considers low- to moderate-risk medical devices.
"I think, first and foremost, the thing to know about the clearance is that it is an analytical clearance," said Fung. "The claims are about the analytical performance and the vast majority of the data and the analyses that went into the submission were around the analytical performance," he said.
According to Fung, CytoScan's analytical performance was predicated on two things: analytical reproducibility and analytical validity. To establish analytical reproducibility, Affymetrix had to demonstrate that CytoScan Dx could measure "certain characteristics, across operators, across sites, across regions, and across arrays," said Fung. "What you measure of course is the discussion you have with the FDA and because the microarrays are for whole-genome analysis of copy number variation what we are looking at is the reproducibility of measuring copy number variation across all of these parameters," he added.
The company sent the same set of samples to testing sites at LabCorp, South Carolina's Greenwood Genetic Center, and Radboud University in the Netherlands. Each site was asked to run the samples repeatedly "so that we could do a comparison of within site reproducibility and across site reproducibility, and we measured the reliability, or reproducibility, of detecting those CNVs and measuring their length and their presence or absence," said Fung.
To establish analytical accuracy, Affymetrix and its partners verified detected CNVs with other techniques, including next-generation sequencing and quantitative PCR. Though the samples had in many cases undergone routine testing, including analysis by conventional methods, such as karyotyping, only sequencing and qPCR could provide comparable resolution to CytoScan Dx, Fung noted.
"The way the study was designed, in agreement with the FDA, is that we could not use ourselves as a predicate," he said. "This was one of the challenges, because if we detected something, and it wasn't detected by a previous method, it's very likely that it wasn't detected because that previous method did not have the same level of resolution, which is why we went into doing qPCR and whole-genome sequencing."
In addition to demonstrating analytical performance, Affymetrix also had to provide data generated using samples from the intended use population —patients with developmental delay and disability, in the case of CytoScan Dx. The FDA cited Affymetrix's analytical study in its Jan. 17 statement on the clearance of CytoScan Dx, where the company and its partners compared affected individuals with those who have a normal phenotype. For that study, Affymetrix worked with GGC again, as well as CombiMatrix and the Mayo Clinic.
All of the studies generated a significant amount of data for Affymetrix to sift through.
"I think of it as an orange," Fung said of the review process. "We would slice the data one way, and then the FDA would say, 'Well, what happens if you slice the data another way?'"
While the company and its partners are writing up papers describing both studies, Fung said that Affymetrix has turned its attention to demonstrating clinical utility for the test in order to gain wider reimbursement for CMA.
"We believe that with the changing payor environment, having FDA clearance will be beneficial to us," said Fung. "Actually, there are a number of payors who are not yet reimbursing for microarrays, which is astounding because microarrays are recommended in all of the guidelines."
CMA is now recommended as a first-tier test by the American Academy of Neurology, the Child Neurology Society, the American College of Medical Genetics, and the American Academy of Pediatrics.
According to Fung, combining information about the clearance process, CytoScan Dx's IVD status, the numerous publications that discuss its clinical utility, as well as the various recommendations from professional societies, will all "expand the reimbursement landscape for us."
FDA clearance should also play a role in expanding the physical landscape for the test. Fung noted that much of the data it used to attain clearance could be used again to obtain a CE-IVD marking, allowing CytoScan Dx's clinical use in the EU.
The process for gaining local regulatory approval in various Asian countries is more nuanced.
"In Asia, different countries have different requirements," Fung noted. "For instance, the Chinese FDA requires home country clearance, which we have, but they may or may not require additional clinical studies in the Chinese population," he said. "So we are obviously planning our strategy in different Asian territories, because each Asian territory has its own requirements."
Though the FDA cleared CytoScan Dx in January, the test is still not commercially available. Fung confirmed that Affymetrix will begin shipping CytoScan Dx later this quarter.
Of all the centers participating in the process, Greenwood, SC-based GGC supported arguably the most elements of Affymetrix's CytoScan Dx FDA submission by participating in the multisite reproducibility demonstration, conducting analytical validation of variations identified by the CytoScan, and providing a majority of the 960 samples used for the clinical validation component of the study.
Alka Chaubey, director of GGC's cytogenetics laboratory, discussed these various studies during Affymetrix's ACMG workshop. She told BioArray News that there were multiple advantages to using the FDA-cleared CytoScan Dx platform over other array-based LDTs.
According to Chaubey, the FDA clearance provides a "clear indication" of the clinical use of the CMA platform, "uniformity in reporting of CMA results back to the clinicians to enable easier interpretation and patient follow-up care," as well as the provision of reagents and arrays manufactured in a certified GMP/ISO facility intended to simplify a laboratory’s [quality control] and validation procedures."
Another advantage of the FDA approval is that it "potentially lowers the litigation risks for genetic services," because the clearance specifies that interpretation of data should be carried out by a specialist, Chaubey noted.
"The fact that a board-certified cytogeneticist or molecular geneticist are the qualified individuals to perform the interpretation is also a win for the genetics community," she said, "by simply ensuring that this is a highly specialized test and can-not be offered as an off-the-shelf genetic test."
Chaubey acknowledged that labs have been and will continue to offer CMA as an LDT. However, that still means that the same labs will have to ensure validation of the CMA platform and perform and maintain appropriate quality assurance and control programs, she said.
Adopting CytoScan Dx therefore might allow labs to sidestep some of those processes while establishing greater inter-lab concordance in regards to CMA testing.
"Different labs have been using different thresholds for reporting which is a huge problem for the clinicians serving these patient populations to interpret the copy number variations," Chaubey added. "Obviously, this was largely due to the non-availability of an FDA-cleared CMA platform until now."