Patricia F. Fitzpatrick Dimond Ph.D. Technical Editor of Clinical OMICs President of BioInsight Communications
Company’s Back on Track after Meeting FDA Requirements
On October 21, privately held 23andMe launched an approved Personal Genome Service (PGS). The company, which said it had worked closely with the FDA to validate its DNA analysis offerings, had received a stern warning letter from the agency in 2013 about its direct-to-consumer (DTC) spit kits for DNA collection and reports generated from gene analysis data.
“I am hopeful for 2015, it has been quite a transformation for the company, to really change and go through this entire process,” noted CEO Anne Wojcicki, who added that 23andMe revamped its approach to communicating and dealing with the FDA. The result: in February it won its first approval for a test for Bloom syndrome. Along with this authorization, the FDA also classified carrier screening tests as class II devices, which the agency intends to exempt from premarket review.
“We had to make a decision as a company whether we were going to stay in the business of personal genomic analysis,” said Angela Calman-Wonson, 23andMe’s vp of communications. “The classification gave us a road map for what we still had to go through for the validation process. The company went with genomic detection of autosomal recessive carrier status because it was the best understood area of genetics and emphasized that it was not testing for the diseases [themselves].”
In its warning letter, the FDA expressed concern that reports to its customers might encourage them to interpret and act on the results themselves without the benefit of interpretation. The company replied that the reports provided a potential “first step in prevention” encouraging PGS users to “take steps toward mitigating serious diseases” such as diabetes, coronary heart disease, and breast cancer.
The letter pointed out that on the basis of information in the reports, patients might begin to self-manage their treatments by changing their medication doses, or abandon certain therapies, citing for example false genotype results for 23andMe’s Warfarin drug response test. In that case, the agency said, “unreasonable risk of illness, injury, or death to the patient due to thrombosis or bleeding events “could occur from treatment with a drug at uncalibrated doses.”
The agency also described its efforts to encourage 23andMe to comply with its validation requirements, citing 14 face-to-face meetings and telephone calls, hundreds of emails, and dozens of written communications during which it said it provided specific feedback on study protocols that would prove 23andMe’s tests were useful as diagnostics and ensuring they yielded consistent results.
The company complied with the agency's demands and decided to maintain access to health-related reports for customers who received such data or purchased a test kit before the FDA warning letter dated November 22, 2013. Customers who purchased kits from the firm on or after that date have only received ancestry-related information and their raw genetic data.
As part of its validation process for each autosomal recessive test, the company performed accuracy studies at two lab sites using samples with known variant status. Results of each 23andMe test were compared with Sanger sequencing results. Only samples that passed quality control and produced a genotype for both sequencing and the 23andMe test were included in the calculation. 23andMe test results demonstrated at least 99% agreement with sequencing.
In the case of Bloom Syndrome, 23andMe performed two separate studies to demonstrate the test’s accuracy in detecting Bloom syndrome carrier status. One study conducted at two laboratories tested a total of 70 unique samples, including those from known carriers of the disease. An additional study evaluated 105 samples at the same two laboratories. Both studies showed equivalent results in detecting carrier status of Bloom syndrome when the same samples were tested.
The company’s carrier status test now includes approved tests for multiple genetic disorders, including beta thalassemia and related hemoglobinopathies, Bloom syndrome, congenital disorder of glycosylation type 1a (PMM2-CDG), and cystic fibrosis, among others.
“The long game here is not to make money selling kits, although the kits are essential to get base level data,” Patrick Chung, a 23andMe board member, told FastCompany in a 2013 interview. “Once you have the data, [the company] does actually become the Google of personalized healthcare.”
In January, 23andMe announced an agreement with Genentech to generate whole genome sequencing data for approximately 3,000 people in 23andMe’s Parkinson’s disease community. The goal of the collaboration is to identify new therapeutic targets for treating Parkinson’s.
Under terms of the deal, 23andMe can conduct additional research on the generated data once the project is completed and share the data, scrubbed of patients' identifying information, with other researchers. Only data from 23andMe customers who have consented to participate in the research will be shared. The deal provides 23andMe with an upfront payment from Genentech of $10 million, with further milestones of as much as $50 million. The agreement is the first of ten 23andMe says it has signed with large pharmaceutical and biotech companies, including Pfizer.
And apart from leveraging its data for drug discovery efforts at partner companies, 23andMe also intends to conduct its own research to try to identify new therapies for both common and rare diseases. To further this effort, 23andMe has created a therapeutics group that focuses on the discovery of novel treatments through human genetics.
A recent Series E financing will go a long way to funding an internal lab and the scientists to staff it. Fidelity Management & Research Company led the round, joined by new investors Casdin Capital, WuXi Healthcare Ventures, and Xfund, as well as existing investors including Illumina, New Enterprise Associates, MPM Capital, and Google Ventures.
To heads its own research enterprise, 23andMe recruited Richard Scheller, Ph.D., as its CSO and head of therapeutics. Dr. Scheller retired in December 2014 after a 14 year stint at Genentech where he was the executive vp of research and early development.
“With Dr. Scheller joining the team, we are putting significant resources into translating genetic information into the discovery and development of new therapies for our customers and the world,” said Wojcicki. “This is a major step forward to realizing our mission to help people benefit from the human genome.”
But moving from genomic analysis, collection, and storage to drug discovery entails quite a leap, and significant economic and technical resources, as the company is well aware. The company will continue to seek collaborations with drug development partners as it builds its own internal research resources.
While it is unusual for a health data company to move into drug discovery and development, “obviously they [23andMe] have credibility in this area as a result of the partnerships they have with a number of pharma companies,” said Dan Bradbury, founder of BioBrit, a life sciences consulting and investment firm in La Jolla, CA. “I think they will be taken seriously but it will take a number of years to establish the capability” to develop drugs, he said.