May 03, 2017
The Oxford-based firm is also in discussions with multiple pharmaceutical and biotechnology companies that wish to use its test, which is run on a custom Affymetrix microarray, to select patients for use in their drug development programs.
“Our plan is that by the end of this year, we will be offering this test from a CLIA-approved partner in the US,” said CEO Richard Pither.
According to Pither, Cytox has tapped Norcross, Georgia-based Akesogen to be the company’s US partner, based in part on its experience with the Affymetrix platform. Cytox is also interested in seeing its assay validated as a laboratory-developed test in a CLIA environment, Pither noted, which will demonstrate that it adheres to a high regulatory standard, enabling the company’s pharma clients to incorporate the assay into therapeutic development planning.
The company could also use validation data should it US Food and Drug Administration 510(k) clearance for the kit.
“Our plan aligns with our partners’ in pharma, but also ensures there is a clinical diagnostic beyond that,” Pither said.
When asked about a timetable for pursuing FDA clearance, Pither said that the firm is focused on serving the needs of its pharma and academic customers, but will “develop a strategy for full regulatory development and commercialization in both the US, EU, and other major territories.”
Cytox was founded in 2004 to commercialize technology developed at the University of Oxford based on spotting cell cycle kinetic irregularities in lymphocytes that were shown to have an association with increased Alzheimer’s disease risk. However, it required access to fresh blood and live lymphocytes, an approach that the company ultimately found untenable.
“That technology was interesting but we felt it wasn’t robust enough to turn into a routine clinical test and the complexities of doing that within the context of clinical trials was such that we didn’t believe the assay could be developed as a robust clinical tool,” said Pither.
Four years ago, when Pither joined Cytox from GE Healthcare, the company decided on a different approach: using microarrays to measure markers associated with the underlying phenotype the company was able to identify using the lymphocyte test. Cytox ultimately developed a custom chip called the variaTECT SNP Research Array and a polygenic risk score algorithm called SNPfitR that now form the core components of its Alzheimer’s test.
According to Pither, the variaTECT SNP Research Array includes 130,000 SNP markers associated with Alzheimer’s disease, and is run on the Affymetrix GeneTitan Multi-Channel Instrument, which enables customers to process plates of arrays in a scalable format. The system is sold by Thermo Fisher Scientific, which acquired Santa Clara, California-based Affy last year.
The company’s SNPfitR algorithm works directly with the output of the GeneTitan, Pither said.
“It allows you to analyze that data using one of three polygenic risk score algorithms we have embedded in that software,” he noted.
Pither said that Cytox decided to use Affymetrix technology, because Affymetrix “has the best genotyping platform for clinical use.” Cytox also makes use of the Affymetrix’s logistics chain. Affymetrix has shipped the assay to partners in the UK, the US, and Australia, he noted.
The array and the SNPfitR algorithm were developed in partnership with the laboratory of John Hardy, a neuroscientist and molecular geneticist at University College London. Using these tools, Cytox can determine if an individual is positive or negative for beta-amyloid, the protein linked to the deadly neurodegenerative disease.
“Essentially you are asking 130,000 times whether a specific variant is present, yes or no,” said Pither of the approach. “The algorithm takes the yes or no answers and adds them up to give you an overall risk score, 1 or 0, 1 being very high risk, 0 being very low risk,” he said. “That is new and potentially transformational in the way we think about identifying Alzheimer’s disease risk.”
In autumn, Cytox introduced an early-access genetic testing service, targeting the assay to clinical researchers and pharma clients for use in stratifying samples obtained from clinical trial subjects who are either pre-symptomatic or show mild cognitive impairment and are at risk of developing Alzheimer’s.
The company discussed the validation of the test on a cohort of 1,000 samples at the Alzheimer’s and Parkinson’s Diseases and Related Neurological Disorders conference, held last month in Vienna. In the talk, they described two algorithms within the SNPfitR architecture that allow the company to identify true amyloid positivity and true amyloid negativity. According to Pither, the samples were obtained from the Australian Imaging, Biomarker & Lifestyle Study of Ageing, as well as partners at the University of Pennsylvania.
With the early-access offer underway, and in the wake of the presentation in Vienna, Cytox is gearing up for its US launch with Akesogen later this year. Kevin Banks, Cytox’s global head of market development, said that the firm selected Akesogen as a partner because of its “deep experience with the Affymetrix platform.” Akesogen received a $7.5 million grant in 2014 to genotype samples on the Affymetrix Axiom platform as part of the US Department of Veteran’s Affairs’ Million Veterans Program.
“We wanted this technology to be robust, to be clinically validated,” Banks said. By partnering with Akesogen, he said Cytox “took the technology risk out of the equation.” He noted that the company also has experience running clinical trials, which may help Cytox down the road should it seek 510(k) clearance for its assay.
Mark Bouzyk, founder and CSO of Akesogen, said that the firm is well positioned to assist Cytox as it enters the US market. He noted that Akesogen runs a CLIA-certified and CAP-accredited lab, and that it has run about 250,000 samples on the Axiom platform as part of the MVP. In addition to offering Cytox’s assay as a clinical offering, he said that Akesogen will also run the test for pharma clients.
According to Bouzyk, Akesogen and Cytox will pilot the technology on samples from Emory University’s new Healthy Aging Study, a large clinical research study targeting 100,000 residents of the Atlanta area, where Akesogen is based.
“We are … working with Cytox to bring this technology into the broad market in coming years with new promising therapeutic opportunities in the pipeline make this a highly attractive proposition for early detection of AD risk,” said Bouzyk.
Courting pharma
Cytox is currently in discussions with a “large number” of pharma and biotech companies that are interested in using its test as part of various Alzheimer’s disease drug development programs, according to Pither. The company believes it can save its clients money by enabling them to identify early-stage, pre-symptomatic individuals at risk of developing the disease who can then be enrolled in programs to develop drugs for preventing disease progression. Other companies see an opportunity for using Cytox’s test as a companion diagnostic to identify those individuals most likely to benefit from a treatment before they develop any symptoms.
“The challenge in the field is how to identify those people before the changes are apparent,” said Pither. “We know some new drugs are moving through with quite encouraging results at the moment, and we are talking to many of the pharma engaged in development with a view to incorporating our test for screening and stratification in clinical trials,” he said.
Drug developers currently rely on PET scans to identify patients for such trials, which cost about $10,000 per person, Pither estimated. However, the success rate of identifying amyloid-positive patients using PET scans is only 10 percent or 20 percent in asymptomatic cases.
“They are spending about $100,000 to find one or two subjects,” said Pither. Cytox’s test could be used to determine which patients are amyloid positive, who would then be followed up by PET scan, he said. Pither noted that the predictive rate of Cytox’s test for identifying amyloid-positive subjects is greater than 90 percent.
“We can very significantly reduce the number of failures by incorporating our test upstream of PET imaging,” Pither said.
The test’s significant potential cost savings has allowed the firm to negotiate pricing with its pharma clients, too, Pither noted.
“Our starting point is that we are going to reduce your screening failure rate by between 80 percent and 90 percent,” said Pither. “It’s a value-driven price proposition.” He said that pricing for the CLIA assay it aims to launch later this year will be “assessed on a case-by-case basis, contingent upon the specifics of the study and the nature of the partnership.”
In terms of competition, Pither acknowledged protein arrays designed to detect amyloid-positive individuals. Some companies such as Meso Scale Discovery and GenScript offer protein or peptide arrays for Alzheimer’s disease research. However, Pither said these arrays have “proven difficult to validate,” noting that protein markers “have a huge dynamic range” and are senstive to handling. “Handling SNPs in DNA is a far more stable analyte,” he said.
23andMe did obtain FDA clearance to predict late-onset Alzheimer’s disease risk as part of its consumer genomics offering, “but they are not making claims about accuracy there,” said Pither. “There is no associated risk or algorithm software.”
And last month a team of US and Norwegian investigators said it had developed a polygenic risk score that could be used to predict late-onset Alzheimer’s disease in a manner that’s tailored to a given individual’s age.
The investigators published the results of their study in the journal PLoS Medicine.
“I don’t see direct competitors yet, but we’d be naive to think that they aren’t coming,” Pither said.