Listen to this article
Apple and Alphabet, two of the world’s largest technology companies, are taking divergent approaches in their efforts to revolutionise healthcare, as doctors, drug companies and regulators look to Silicon Valley to bring down the cost of clinical trials.
After Apple presented mixed results from a study designed to show that its Apple Watch could be used to detect heart problems, Verily — like Google, a subsidiary of Alphabet — has been opening up about its own effort to bring scientific rigour to the use of wearable devices for diagnosing health problems and testing treatments.
Jessica Mega, Verily’s chief medical officer, told the Financial Times that it was at an “inflection point” in the ability to store and process massive amounts of information that could transform clinical trials — and ultimately, how much we understand about human biology.
The cardiologist and former clinical trial investigator said wearables such as Verily’s Study Watch can collect far more data than could ever be found inside the “four walls of a clinic”, discovering far more by measuring participants continuously.
“Currently we have a few gigabytes per person of health data — that’s projected to be terabytes in the future. We need to think about rigorous studies to figure out what signals matter,” she said.
Unlike the Apple Watch, Verily’s Study Watch is designed specifically for trials and is given to participants by trial organisers.
Verily recently expanded its Project Baseline initiative, a partnership with Duke University and Stanford Medicine, which is collecting health data on more than 10,000 people using the Study Watch, as well as traditional methods, to discover clues that can predict disease. Now, in partnership with the American Heart Association, it is opening up a registry of women, often under-represented in research, who want to participate in clinical trials and test new technologies.
Large tech companies and start-ups are competing to revolutionise a clinical trials market that is worth $65bn, according to CB Insights. Trials are almost always essential for regulatory approval of new treatments, yet the process can take an average of 7.5 years and cost up to $2bn per drug, according to the research firm.
Even regulators such as the US Food and Drug Administration, which have stringent requirements on how trials are conducted, are pushing for more collaboration and data sharing, and the adoption of new technologies.
Scott Gottlieb, the outgoing FDA commissioner, said last month that “important therapeutic opportunities” risk being delayed or discarded because companies cannot afford to run the trials to validate them.
“If you’re a physician working on a clinical trial your experience is probably no different than 20 or 30 years ago,” said Gary Hughes, chief executive of Teckro, a software platform that tries to make the process of participating in trials easier for doctors and patients. “It’s still a people and paper process.”
Teckro raised $25m from investors including Peter Thiel’s Founders Fund in February.
In an effort to reduce costs, pharmaceutical and biotech firms are partnering with technology companies to speed up the process.
US biotech Amgen is using artificial intelligence to improve trial planning, so it does not waste time and money setting them up in places where there will be few participants. Some 80 per cent of trials fail to meet enrolment timelines, and one-third of Phase III clinical study terminations are because of enrolment difficulties, according to a study by Cognizant, an IT company.
Amgen has used off-the-shelf AI from large cloud providers such as Google and Amazon to analyse its previous trial data. Over the past five years, it has managed to shave three years off the time it takes to get a drug to market from the research stage, which averages 10 to 15 years.
A bigger prize than improving the efficiency of clinical trials is opening up new ways of diagnosis and testing treatments using wearable devices.
The Apple Heart Study, which announced its preliminary results last month, suggested wearables could be used to reach hundreds of thousands of participants. Lloyd Minor, dean of the Stanford School of Medicine, said the entirely virtual study was a “big leap forward in digital health . . . Technology has influenced and impacted every sector of the economy except for healthcare. Now we are positioned to be able to catch up.”
However, it also highlighted potential pitfalls. The study, conducted by Stanford and funded by Apple, discovered the Apple Watch could be used to detect atrial fibrillation — a leading cause of stroke — that often remains hidden because people do not notice the symptoms. But just 34 per cent of people who received a notification of possible atrial fibrillation had it confirmed by an electrocardiogram patch. The researchers said this may have been because it is an intermittent condition. It could be a sign of false positives.
Whether large tech or start-ups will win the battle to transform clinical trials depends on their reach, data, and ability to recruit employees with the expertise to separate the signals from the noise.
Neil Kurtz, an industry veteran on the board of Medidata, which makes software for clinical trials, said only a handful of people really understand how to use machine learning in healthcare. “Getting access to those people is going to be one of the key determinants of who wins and who loses,” he said.
Alphabet’s Verily is making the pitch that it can be a winner by designing its device specifically for trials, rather than as a consumer product with less oversight of who it is being worn by and when.
As well as Project Baseline, the Study Watch is being used in a five-year study led by the University of North Carolina and Harvard of 5,000 people recovering from trauma, such as experiences in war zones or sexual assault.
And in the Netherlands it is being used in the Personalised Parkinson’s Project, a multiyear study to identify ways to track the disease’s progression. Worn by newly diagnosed patients, the watch will collect vital signs, activity levels and information about the participant’s environment. Its data will be compared with clinical and genetic information to try to find ways to predict differences in patients and develop new treatments.
Verily’s Dr Mega said she was a “big proponent” of consumer devices, but that if someone were using data from a wearable to make treatment decisions, there needed to be an extra level of rigour: “If you give someone a therapy based on that data, you want to be sure it was from you and not from your friend or your dog.”
Get alerts on Pharmaceuticals sector when a new story is published