When tech entrepreneur Onno Faber learned he had an extremely rare genetic disease with no available treatments, he leaned on his tech skills and organized a Silicon Valley “hackathon” in 2017 to dig into his own DNA and perhaps help others with his condition.
Faber can now lean on a few million dollars. The company that emerged from the hackathon, RDMD, has raised $3 million in seed money to build a health-data repository for rare diseases. Faber (pictured above, right) is chairman and head of product; Nancy Yu (above, left), who spent two years on the business side at 23andMe, is cofounder and CEO. They are currently camped out at a popular shared workspace in downtown San Francisco.
Faber, who moved from the Netherlands to San Francisco in 2014 to build a video messaging company, has neurofibromatosis type 2. It causes noncancerous tumors in the nervous system that can affect hearing and balance and lead to other grave medical problems. Thanks to the hackathon, he learned that a drug approved for advanced breast cancer, lapatinib (Tykerb), might help him. He started taking it a few weeks ago. “We don’t know if it’s going to work, but we have a good theory,” he says. His insurance covers it, but he acknowledges that others in his situation might not be so lucky.
RDMD wants to gather, under one roof, the far-flung medical data of people with rare diseases—sometimes only a few dozen people are afflicted per disease. The company is using tech tricks to convert a range of health records in different formats, such as doctor’s notes, medical images, and hospital charts, into a single repository that researchers can use to push drug development forward. “The data out there in the world isn’t necessarily accessible for research,” says Faber.
Even though the hackathon last year was billed as “AI Genomics,” RDMD isn’t using artificial intelligence yet. Before you analyze with AI algorithms, Faber says, you have to build and structure the data. With their own software, RDMD is building what might be described as a reference library; every piece of data under its roof can be traced back to its original document. The audit trail is important if the data are to be used one day in clinical studies.
One goal of RDMD is to gather enough medical information to build what’s known as “natural history” studies, which help researchers understand the characteristics, timelines, and other aspects of rare diseases by following a group of patients over time who don’t receive treatment. Such studies can help in the design of diagnostics and clinical trials for diseases that have little or no clinical track record.
Ethan Perlstein, CEO of rare-disease drug developer Perlara in South San Francisco, CA, is an angel investor in RDMD’s seed round. He admits he can’t judge the efficacy of the startup’s software—it’s too early—but says, “I get the need for a dedicated platform” for the rare-disease community.
With little support from industry in rare disease research, many patients, caregivers, doctors, and scientists have learned to “self-assemble” and help each other, Perlstein says. “Ultimately the best patient registries and engagement will come from this kind of bottom-up approach.” His investment also allows him to follow closely RDMD’s progress in neurofibromatosis, for which Perlara might want to pursue drugs one day. (RDMD has recruited 150 patients to its service so far.)
Federal regulators are more open-minded than ever about faster drug development these days, but new medicines still require years of work. In the short-term, other than the hope for eventual treatments, what’s in it for the patients who consent to the use of their medical histories? RDMD’s answer: At no charge, the company will do the legwork to extract their records from hospitals and other healthcare providers, then give each patient a private online portal to keep their records up to date. Patients can also share their records, as needed, with doctors and others. Here’s an example of Faber’s RDMD record:
The company will make money by sharing patient information, stripped of personal details, with drug developers—a strategy that follows the lead of an increasing number of health service companies.
There are different models out there. For example, Patients Like Me of Cambridge, MA, hosts private communication forums for its members, then aggregates and sells the information that they post.
Aggregating and selling so-called “de-identified” patient data is also part of the playbook of 23andMe, which charges $100 and $200 for ancestry and disease-risk DNA tests but
