The Struggle to Build a Massive ‘Biobank’ of Patient Data
Posted March 19, 2018 6:16 p.m. EDT
Updated March 19, 2018 6:17 p.m. EDT
This spring, the National Institutes of Health will start recruiting participants for one of the most ambitious medical projects ever envisioned.
The goal is to find 1 million people in the United States, from all walks of life and all racial and ethnic groups, who are willing to have their genomes sequenced, and to provide their medical records and regular blood samples.
They may choose to wear devices that continuously monitor physical activity, perhaps even devices not yet developed that will track heart rate and blood pressure. They will fill out surveys about what they eat and how much.
If all goes well, experts say, the result will be a trove of health information like nothing the world has seen. The project, called the All of Us Research Program, should provide new insights into who gets sick and why, and how to prevent and treat chronic diseases.
The All of Us program joins a wave of similar efforts to construct gigantic “biobanks” by, among others, the Department of Veterans Affairs, a British collaboration and private companies like Geisinger Health Systems and Kaiser Permanente.
But All of Us is the only one that attempts to capture a huge sample that is representative of the U.S. population. “It will be transformative,” said Dr. Francis Collins, director of the National Institutes of Health.
It will also be expensive.
In 2017 alone, the budget for All of Us was $230 million, of which $40 million came from the 21st Century Cures Act. Congress has authorized an astounding $1.455 billion over 10 years for the project.
While supporters say the results will be well worth the money and effort, others have begun to question whether All of Us is just too ambitious, too loaded with cumbersome bureaucracy — and too duplicative of smaller programs that are moving much more quickly.
In the three years since the All of Us program was announced, not a single person’s DNA has been sequenced. Instead, project leaders have signed up more than 17,000 volunteers as “beta testers” in a pilot phase of the program. They supplied blood and urine samples, had measurements taken, and filled out surveys.
Dr. George D. Yancopoulos, the president and chief scientific officer of the biotech company Regeneron, said the NIH did not have much to show for three years of planning. Regeneron has been deeply involved in similar public and private efforts, sequencing the DNA of more than 300,000 participants. The beta testers constitute just 1.7 percent of the program’s target, Yancopoulos noted, and the investigators have collected only the simplest data, not genetic sequences.
“At this rate, when will they complete their 1 million-person target?” he wondered. “And at what taxpayer cost?”
“I think someone needs to ask tough questions about whether this is the best use of precious NIH resources,” he added. “Should the funding instead go to individual researchers who are doing truly basic and innovative science?”
Two large health providers — Geisinger and Kaiser Permanente — both backed away from grants to participate in All of Us.
David Ledbetter, executive vice president and chief scientific officer of Geisinger, said that the program’s complexity made it too time-consuming: conference calls upon conference calls, meetings upon meetings, without much progress.
“We decided it was not the right expenditure of our time,” he said. Geisinger gave back its award that was potentially worth $50 million over five years.
Geisinger has enrolled more than 180,000 participants in a biobank of its own, and the health system already has years of their medical records. Regeneron is sequencing the participants’ DNA and has completed more than 100,000.
Ledbetter said the NIH program would be “very valuable someday.” But Geisinger, he said, did not want to wait.
“Someday is today,” he said.
Kaiser Permanente, too, is now far ahead in developing its own biobank. Originally, the company expected that the federal project would profit from Kaiser’s experience with recruiting and data analysis, said Elizabeth McGlynn, vice president of Kaiser Permanente Research.
“We were not able to engage as a scientific partner,” McGlynn said. “We felt increasingly that we were just being asked to give access to our members.” DeCode Genetics, a subsidiary of Amgen, a biotech company, is working with a biobank of 160,000 people from Iceland. Dr. Sean E. Harper, Amgen’s executive vice president for research and development, says it is hard to imagine the complexity of analyzing the data.
“It took about 20 years and over $1 billion of investment to get to the point where we are able to routinely extract from the data the necessary information to validate or invalidate drug targets,” he said. Sequencing the DNA is the easy part, he said. “The hard part is to get all these medical records and lab tests curated in a computer system where they are query-able and to perfect the analytics.”
Despite these concerns, All of Us has contracted with scientists at just about every leading university, as well as with companies like Verily, a subsidiary of Alphabet, parent company of Google.
“We will have an unprecedented amount of data at a scale never done before,” said Eric Dishman, director of the program.
Investigators have grand plans for all that data once it becomes available.
Dr. Atul Butte, director of the Institute for Computational Health Sciences at the University of California, San Francisco, hopes to find the earliest signs of disease, especially of Type 2 diabetes.
“Do you go back and forth from diabetes for a while?” he asks. “Is it preventable?”
The Veterans Affairs Department began building a similar biobank, called the Million Veteran Program, in 2011 with a very lean budget: just $250 million over the past seven years. The agency has recruited 650,000 vets so far and has years of their medical records, including prescription data. Investigators expect to sequence the DNA of 100,000 participants in the next two years, at a cost of $70 for each person’s entire genome. The data will be public.
The British program, called the U.K. Biobank, has half a million participants with complete medical records and additional data for some, including body and brain scans. Regeneron has committed to sequencing the DNA of all of the participants by the end of 2019. After a six- to 12-month period of exclusivity, the company will make that data public.
But what All of Us is attempting to do is much more complex, Dishman said. The U.K. project and the program at Geisinger lack a representative range of racial and ethnic groups; the VA’s biobank has relatively few women.
All of Us will be built to reflect the United States population. The San Francisco General Hospital Foundation, for instance, was given a grant to recruit lesbian, bisexual, gay and transgender participants.
“I think what the U.S. project adds is that it reflects the diversity of the U.S.,” said Dr. Sekar Kathiresan, a geneticist at the Broad Institute and an investigator with the NIH program.
That is partly why planning for the project has dragged on. And the diversity of participants makes the daunting task of retrieving medical records even more difficult.
Americans tend to have medical records stored slapdash all over the place, and they change insurers and medical plans frequently. There is little uniformity in the country’s electronic health systems.
Even the most straightforward part of the project — DNA sequencing — is formidable.
“There are not enough sequencing machines in the U.S. to just focus on our project,” Dishman acknowledged. All of Us will depend on prices falling for sequencing — and more sequencers being built. For participants, there will be a reward, Collins noted. The program will give them genetic information and health data, and tell them how they compare to others in the population. To do that, All of Us plans to enlist genetic counselors. Yet right now, there are not nearly enough counselors to handle the task.
The counseling issue also bothered Kaiser, McGlynn said, and it was a factor in the decision to return the grant money.
“Genetic counselors are in terribly short supply,” she said. “We wanted to be sure we were well organized to deliver results in a way that was ethical and not scary to members.”
The accumulation of tremendous caches of medical data is raising troubling questions about what participants should be told. The U.K. Biobank, for instance, considered returning results to participants, but decided against it after an experiment.
It involved participants getting whole body scans for the program. For some subjects, a radiologist systematically assessed the scans to see if anything seemed abnormal. If it did, the patient and the patient’s doctor were informed.
As it turned out, 20 percent of the participants had abnormalities. They often went on to have other tests, some of which were invasive and involved major surgery, like removal of a lobe of the lung.
Yet in the end, only 1 in 8 with abnormal scans actually had a medical problem, and even then there was nothing they could do about it most of the time, said Dr. Rory Collins, chief executive of the U.K. project. “What we are trying to do is not provide care to individuals, but to generate a resource that can provide health information,” he said. “Feedback can cause more harm than good.” Other experts disagree with the British approach. At Geisinger, participants are told if they have a genetic variant that might affect their disease risk.
They are offered genetic counseling if they want it — and so far, about two-thirds do. The medical system has sufficient counselors to handle the demand, said Adam H. Buchanan, co-director of the counseling program.
Given the substantial obstacles, will the NIH project, which has not even really begun, be worth the immense expense and effort?
Collins, an adviser to All of Us, thinks it will. Huge amounts of data will be needed to really understand interactions between genetics, environment and lifestyle.
“Half a million people isn’t enough. Even a million isn’t enough,” he said.
Ledbetter was more circumspect. “I think the idea is great,” he said. “It is ambitious. It is expensive. It will take a while.”
“I wish the NIH well.”