STAT, November 5, 2015
If you enter Health Nucleus, a new facility in San Diego cofounded by J. Craig Venter, one of the world’s best-known living scientists, you will get a telling glimpse into the state of medical science in 2015.
Your entire genome will be sequenced with extraordinary resolution and accuracy. Your body will be scanned in fine, three-dimensional detail. Thousands of compounds in your blood will be measured. Even the microbes that live inside you will be surveyed. You will get a custom-made iPad app to navigate data about yourself. Also, your wallet will be at least $25,000 lighter.
Venter, who came to the world’s attention in the 1990s when he led a campaign to produce the first draft of a human genome, launched Health Nucleus last month as part of his new company, Human Longevity. He has made clear that his aim is just as lofty as it was when he and his team sequenced the human genome or built a flu vaccine from a genetic sequence delivered to them over the Internet.
“We’re trying to show the value of actual scientific data that can change people’s lives,” Venter told STAT in some of his most extensive remarks yet about the project. “Our goal is to interpret everything in the genome that we can.”
Still, the initiative is drawing deep suspicion among some doctors who question whether Venter’s existing tests can tell patients anything meaningful at all. In interviews, they said they see Health Nucleus as the latest venture that could lead consumers to believe that more testing means improved health. That notion, they say, could drive customers to get procedures they don’t need, which might even be harmful.
“I think there is absolutely no evidence that any of those tests have any benefit for healthy people,” Dr. Rita Redberg, a cardiologist at the University of California at San Francisco and the editor-in-chief of JAMA Internal Medicine, said when asked about Venter’s new project.
Venter has a black belt in media savvy — he can make the details of molecular biology alluring for viewers of 60 Minutes and TED talks alike — but off screen he has earned a reputation even from his critics for serious scientific achievements. His non-profit J. Craig Venter Institute, which he founded in 1992, now has a staff of 300. Scientists at the institute have explored everything from the ocean’s biodiversity to the Ebola virus.
Last year, at age 67, Venter cofounded Human Longevity, a company based in San Diego with branches in Mountain View, Calif., and Singapore that is building the largest human genome-sequencing operation on Earth, equipped with massive computing resources to analyze the data being generated. The firm’s database now contains highly accurate genome sequences from 20,000 people; another 3,000 genomes are being added each month.
Franz Och, the former head of Google Translate and an expert on machine learning, is leading a team that’s teaching computers to recognize patterns in the company’s databases that scientists themselves may not be able to see. To demonstrate the power of this approach, Human Longevity researchers are using machine learning to discover how genetic variations shape the human face.
“We can determine a good resemblance of your photograph straight from your genetic code,” said Venter.
Venter and his colleagues will be publishing the results of that study soon — most likely generating another round of headlines. But headlines don’t pay the bills, and at a company that’s got $70 million in funding from private investors, bills matter. The company is now exploring a number of avenues for generating income from its database. It has partnered with Discovery, an insurance company in England and South Africa, to read the DNA of their clients. For $250 apiece, it will sequence the protein-coding regions of the genome, known as exomes, and offer an interpretation of the data.
Health Nucleus could become yet another source of income for Human Longevity. The San Diego facility can handle eight to 12 people a day. There are plans to open more sites both in the United States and abroad. “You can do the math,” Venter said.
Health Nucleus is already taking in customers who can afford the tests, drawing them with a high-end video that promises “individualized services including advanced genomes and other measures combined with advanced laboratory testing,” and declaring the program “creates a model for precision medicine, monitoring wellness in treating disease before it occurs.”
The battery of tests that Health Nucleus customers get for their steep payment include two kinds of genome sequencing. They get their own genome sequenced, along with the genomes of the microbes that live in their body. Blood tests measure a wide range of metabolic compounds produced by both the customers and their microbes. To measure the cardiovascular health of customers, Health Nucleus will use a method called 4D echocardiography, which captures the three-dimensional shape of the heart over time. With a technique called carotid intimal media thickness testing, Health Nucleus staff will measure the plaque on blood vessel walls.
As part of their exam, Health Nucleus customers also climb into a powerful magnetic resonance imaging scanner. The scanning reveals fine details of the brain, including the volume of different regions and the shape of brain blood vessels. “We can say something truly quantitative and predictive about a person’s risk for Alzheimer’s dementia,” said Dr. David Karow, Health Nucleus’s radiology consultant and an assistant professor at the University of California at San Diego.
Using a method called diffusion-weighted imaging, Health Nucleus also creates whole-body scans, which map the three-dimensional structure of fat deposits in customers and detect small tumors, which can be more successfully treated than ones discovered later.
Researchers then put together what the Health Nucleus website describes as “a personalized care plan delivered to you and your physician.”
The plan will be based in part on the customers’ genetic variants that raise their risk of diseases such as diabetes or cancer. To help customers understand the results, Health Nucleus doesn’t simply provide an abstract list of measurements. Instead, the company loads the information into a program that generates a three-dimensional likeness of a customer. Customers can then zoom inside to explore their imaged bodies and get visualizations of their other test results. As Human Longevity discovers more about the genome and other aspects of human biology, the program will be automatically updated.
“It’s the health care equivalent of Google Earth,” said Dr. Brad Perkins, the chief medical officer of Health Nucleus.
Many of the tests provided by Health Nucleus are experimental and do not yet have FDA approval for clinical use. Health Nucleus plans to avoid the headaches that testing companies like 23andMe and Theranos have recently experienced by operating for the time being under a clinical research protocol. Both Health Nucleus customers and their primary doctors will be required to give informed consent to be part of the research, which is designed to assess the value of making individual risk assessments.
While Health Nucleus may turn out to be a good source of revenue for Human Longevity, the opportunity to do research was the original motivation for the program.
“We wanted to do precise measurements on human biology that you can’t get from any other source,” said Venter.
After removing identifying details about Health Nucleus customers, Human Longevity researchers will add their data into the company’s database. Venter predicted that they will be able to probe the test results for connections that would go missing in ordinary research — such as, for example, the link between certain genes and the size of different structures in the brain.
Dr. Robert Green, a clinical geneticist at Harvard Medical School who is investigating how to make genomes useful in medical care, said that Health Nucleus could spur some remarkable new findings. “I think it’s a fantastically visionary plan,” he said.
But Green also has some serious reservations about Health Nucleus — especially about the information that customers will receive. “It really depends on what they give back,” he said. “What appears at first blush to be medically relevant information could be helpful, could be confusing, or could actually be harmful.”
Dr. Eric Topol, the director of the Scripps Translational Science Institute in La Jolla, Calif., and the author of “The Patient Will See You Now,” is another data evangelist who shares Green’s concerns. It’s possible that Health Longevity’s database will eventually get powerful enough to definitively reveal a lot of hidden risks for disease. But today the state of the art is a long way from that dream. “Right now we have tremendous amounts of noise with minimal signals,” Topol said.
To illustrate how Health Nucleus might help people stay healthy, Venter likes to bring up his own experience of having his genome sequenced. He discovered he has a genetic variant that puts him at risk of melanoma. By learning how to recognize the early stages of the cancer, he was able to spot an early growth and have it successfully treated. Otherwise, Venter said, he might not have caught it for months, at which point his future might have been far dicier. “Worst-case scenario, we wouldn’t being having this conversation,” he said.
But Timothy Caulfield, the research director at the Health Law Institute at the University of Alberta, doubts that most healthy people will learn much more from their genome than they would from standard medical procedures. “Stepping on your weight scale is more predictive of future health issues than almost all the information you’ll get from a genome scan,” said Caulfield.
Dr. H. Gilbert Welch, a professor of medicine at the Dartmouth Institute and the author of “Less Medicine, More Health,” wonders what effective treatments would come from a Health Nucleus genome scan. Consider what might happen if customers found out they have an elevated risk of developing Alzheimer’s disease. “What are we supposed to do? Play more Sudoku? How does this help people?” he asked.
To make sense of the microbiomes of Health Nucleus customers — the hundreds of species of microbes that live in and on our bodies — Health Nucleus will rely on data coming from a team of Human Longevity scientists led by Karen Nelson, the president of the J. Craig Venter Institute. Analyzing thousands of samples, they have already amassed a DNA database bigger than all the publically available microbiome sequences put together. But they expect to increase their database by a factor of ten in the next few years.
“This is the ultimate playground,” Nelson said.
With the help of machine learning, Nelson and her colleagues are finding changes in the activity of the microbiome that signal the early stages of a variety of diseases. Someday they may be able to deliver this kind of information to Health Nucleus customers, but not for years. For now, Nelson has far more modest goals, such as detecting the DNA of pathogens that go unnoticed undetected in ordinary exams. “We’re not trying to dictate health, but I think we’re going to pick up signatures,” she said.
As for the imaging done in a Health Nucleus exam, Redberg, the JAMA editor, sees little evidence that it will benefit healthy people. Three-dimensional echocardiography, she said, “makes for nice pictures, but it’s very hard to interpret clinically.” And research on whole-body imaging, she said, has indicated that it shouldn’t be used for routine screening.
“You weren’t finding anything that you needed to know about, and you were finding a lot of things that you were better off not knowing about,” she said.
Venter and Karow defend their use of whole-body imaging by pointing out that the technology has evolved tremendously in recent years. The most advanced scans let radiologists do a better job of distinguishing between malignant tumors and harmless growths without the need for a biopsy. The research they’re going to do at Health Nucleus, they argue, will shed light on how effective such scans are on promoting good health.
But if that’s what Health Nucleus is for, Venter’s critics say, then it should be set up as a randomized, controlled experiment, in which they can compare the long-term outcomes of people who get the tests and to that of people who don’t.
“I’m amazed that they are doing this,” Topol said. “I respect Venter as a leading scientist of our era, but there’s nothing to support it. The best thing would be to run a study to see if it actually helped people. Show us some data that this is ready for prime time.”
Copyright 2015 STAT News. Reprinted with permission.