Cancer drugs have cured countless mice, only to fall short in people. It’s a well-known problem with drug development: Preclinical tests done on animals can be inaccurate, and that drives up the cost and time it takes to make a successful therapeutic.
Several startups are developing technologies to improve preclinical testing—microchips that act like human organs, for instance, are a popular approach. A startup called Vium has a different idea, what it calls a “living informatics” approach that combines big data techniques with sensor technologies to try to boost the speed, efficiency, and success rate of animal tests themselves.
Vium has been operating stealthily under the name “Mousera” for about three years, but this morning made its first announcement including details about the technology it’s developed. The San Mateo, CA-based startup has already attracted $33 million in venture financing so far from Lux Capital, Data Collective, Dolby Family Ventures, AME Cloud Ventures, and Peter Thiel’s Founders Fund. It’s also got some high-profile advisors and board members, including Harvard Medical School geneticist David Sinclair and former Pfizer CEO Jeff Kindler (a Lux partner). Vium CEO Timothy Robertson is a former executive from Redwood City, CA, digital medicine company Proteus Digital Health; co-founder Joe Betts-Lacroix (Vium’s chief technology officer) helped form handheld computer developer OQO.
The company gave just a few tidbits about its approach when operating as Mousera. In this report from MedCity News, for instance, Robertson said the company’s goal was to make preclinical testing more efficient by industrializing certain processes that are manual.
“This is an enormously lucrative space that is completely underserved by technology,” Robertson told me today. “That’s why we’ve been in stealth, we wanted to maintain our first mover advantage.”
In its announcement today, Vium says it’s “applying life sciences, digital technology, and large-scale interpretive methods to living systems” to have researchers “make better informed decisions” more quickly than with existing methods. The approach consists of two key pieces. One is what the company calls a “Digital Vivarium,” a facility in which rodent cages are equipped with sensors, high-definition cameras, and other technology to accumulate all sorts of digital data on the animals. That information is then transmitted to a cloud service where it is stored and analyzed. Vium says scientists can monitor the animals and access the information from anywhere via an online portal, and that its system has controls to minimize human interaction with the animals, which should reduce possible study errors.
According to Robertson, there are a few key advantages to this approach. One is speed. Typically, if a researcher wants to run an animal study, it’ll call up a contract research organization, which will send out a representative, and then “spend typically weeks” going back and forth with the CRO on how to run the study and getting a price quote. Vium replaces this with an online research suite; a researcher types in how the experiment should be run, and gets a price.
Another difference is the amount of data being collected. With traditional methods, Robertson says, a technician walks around with clipboard, looks in a cage, logs an observation, and that’s “the only data point you get for a whole day.” All those observations are put into a report that can take weeks to put together, Robertson says. Vium replaces that process with a sort of digital health tracking system, in which an array of sensors track the mice all day, capturing data about their behavior and physiology, and computers log the information. Researchers can adjust their experiments or make decisions on experimental drugs as the data come in, rather than waiting for a report. And because they’re getting more data, researchers can use fewer animals, Robertson says.
Additionally, this should lead to fewer human interactions with the mice, which view humans as predators and go into fight or flight mode when they’re picked up. Less interaction, Robertson says, can lead to less variable results.
Vium has done “a number of scientific studies” over the last three years to prove its claims, Robertson says. The startup has been applying its approach to testing drugs for a variety of diseases like rheumatoid arthritis, multiple sclerosis, lupus, leukemia, and liver fibrosis, and has been working with select customers in an invitation-only beta mode. The company will be putting its technology on display at the BIO International Convention next week in San Francisco, and is now officially making the technology broadly available to large and small pharmaceutical companies and academic labs.
Vium charges a fee for each experiment. Robertson wouldn’t disclose specific pricing figures, but says the costs per experiment are “very comparable” to quotes from a traditional CRO.
But Vium also aims to strike deals with companies that want to bring the technology in-house. That’s a much more sizeable investment, something in the “seven to eight figure range,” according to Robertson, but Vium has to prove that its technology is worth it to customers first. He says discussions are underway, but it’ll take many such deals before Vium starts to make a big dent in the preclinical drug testing world.
“We’ve seen the playbook of having a field where there’s no technology,” Robertson says. “You bring in technology and you win.”