Seattle's Sage Bionetworks seeks a drug-discovery revolution
When Stephen Friend formed the nonprofit Sage Bionetworks, his aim was nothing short of revolutionizing the way drug discovery is done so patients don't have to wait for help.
photographed by John Lok
SPRING 2010. A young man steps to a microphone in a hotel ballroom. A sea of brainiac biologists and fancy-pants pharma execs sit before him.
"I'm a 22-year-old college dropout," he begins.
Then he proceeds to show the roomful of egos how medicine's got it all wrong.
It's wrong in the way patients are viewed. Wrong in the way drugs are developed. Wrong in the way research is conducted. Wrong.
Well, he said it much better than that.
A few years earlier, Josh Sommer explained, he had a tumor removed from his skull, a rare bone cancer called chordoma. In short order, he learned that these tumors often return again and again; that the only known treatment is surgery; and the average survival is seven years.
Sommer learned, as well, that researchers had studied this disease. Yet a lot of their work was "sitting locked inside labs." Neither he nor his doctors could get at it.
"I'm four years into a diagnosis with an average survival of seven years," he noted. In other words, step on it, people.
The crowd, recalled conference organizer Stephen Friend, fell silent. All those brainiacs and fancy pants knew the problem wasn't chordoma research, per se. It was them. It was how they kept their own cancer research to themselves. Their diabetes research. Parkinson's, Alzheimer's, heart disease, you name it.
That's the way it goes, says Friend, a biologist himself. Science plods. Patients wait.
In the world of medical research, patients are at the bottom of the food chain. This earnest young dropout had just forced these guys to look the problem straight in the eye.
That's why Friend invited him in the first place. Friend had been one of those fancy pants; a brainiac, too. Now, he had another calling — kick-starting a revolution. It is a revolution that involves technological advancement, for sure. But the biggest shift will have to occur in the hearts and minds of scientists themselves. If Friend has his way, his colleagues will start doing something we were supposed to learn in kindergarten: sharing.
What he's talking about is an "open-source movement" for biology, sort of a Wikipedia for the pharmacopeia.
In Friend's milieu, this is a tectonic cultural shift. Universities will need a shake-up, from the way research is done to how promotions are made. Funders like the National Institutes of Health will have to change the way grants are awarded. Pharma will transform, too.
If you say to Friend, "That seems kind of . . . hard," he will pause. He will fix his gaze. And he will smile.
"It's almost undoable."
BEFORE WE get to the revolution, let's establish something up front: Medicine is failing us. We're talking mainly here about drug discovery.
It takes too long to develop new drugs; it costs too much; and the drugs that make it out of the pipeline work for only about half the people who take them.
This is not how it was supposed to be. A decade ago, when scientists announced they had sequenced the human genome — cracked the code of DNA — we were promised a whole new world of personalized medicine. Doctors would take a blood sample, run it through a DNA sequencer and out would pop what essentially was a personalized medical manual. That's the basic idea, anyway. You have a defective copy of gene ABC? We'll prescribe this drug instead of that one because it works best on people with that defect.
This is already happening, albeit to a very limited extent. For instance, doctors treating heart disease know that a certain gene indicates how patients will respond to an anticoagulant.
But personalized medicine? Not so much.
The head of the National Institutes of Health and many others are still convinced we'll get there. So is Friend.
He's got cred. In the early 1980s, as an oncology researcher, he led a team that cloned the first cancer-susceptibility gene. In 1997 he cofounded an upstart South Lake Union biotech company, Rosetta Inpharmatics. When that was gulped up by Merck, he headed up the pharma giant's cancer research. Forbes magazine called him "one of the last great dreamers" in the drug industry.
Friend remembers the moment he realized that his whole industry was terribly flawed. It was 2008, and he was talking with another scientist in a hallway at the University of Washington. Friend asked after his research. The scientist wouldn't say. It was Friend's ninth such conversation in two weeks. Turns out scientists, like journalists, are afraid of getting scooped.
Friend knew the drill: Wait until the findings get published. He knew that could be years. Maybe never.
Most of us have come to accept that this is just the way it works. Pharma is all about patents and profits, right? They have to keep secrets. In academia, it's just a different Darwinian fight. Biologists compete for grants, for tenure, for publication. And yes, for glory.
"If patients knew the secrecy in academia, they would throw up," Friend says. After all, we fund a lot of this research with our tax dollars and donations. Some of us participate in it, too — giving our blood, our medical records, our bodies, even. Yet we can't learn what they're doing.
After eight years in academia, eight in biotech and eight in Big Pharma, Friend opted for a change. "Each of these places thinks they're going to cure diseases by themselves," he says. Ha!
"You could say, in fact, that there's a medical-industrial complex that needs to be taken down."
With a rap like that, you might think Friend has no, well, friends in the field. You'd be wrong.
JERALD RADICH, an oncology doc and researcher at the Fred Hutchinson Cancer Research Center, is among them. He's seen many advances in his 25 years in practice. Yet every time a patient walks in the door with leukemia, he's got to tell them how much he doesn't know.
The first-line treatment is chemotherapy, and in most people, the drugs quickly start knocking down the leukemia cells.
But in 10 to 20 percent of people, the typical chemo drugs don't do anything. It's not luck, Radich says. It's not the patient's will to survive. It must have something to do with their biology.
But infuriatingly, the only way to know if a particular patient is going to be a "nonresponder" is to put him through chemo. By the time he gets other therapies, he's wasted precious weeks and gotten a whole lot sicker.
"If somebody has no hope and is undergoing a toxic therapy that's going to take time . . ." Radich trails off. It's ugly.
Yet science hasn't been able to move the percentages for years. Doctors just keep giving the chemo, and the treatment keeps failing, and there's no way to know who's who.
Dr. Michael Cunningham wants to know who's who, too. He's medical director of Seattle Children's Craniofacial Center and treats patients with a disease called craniosynostosis.
Kids like Olivia Bush. Normally, at birth, a baby's skull bones aren't fully fused. This allows the skull to expand as the brain grows, until, at around age 2, the skull begins to fuse over the soft spots and growth slows significantly.
But in craniosynostosis, the fusion happens in the womb. The brain grows but the skull doesn't have enough give. The brain presses the skull outward where it can, but gets squished where it can't. It can be "extremely deforming," Cunningham explained. If untreated, these kids are also more likely to have brain damage than their peers.
Cunningham diagnosed Olivia's craniosynostosis at around 4 months. By then, her brain was really pushing her forehead out.
"It was starting to look like the keel of a ship," says her dad, Todd Bush.
The surgery involves cutting slits in the side of her skull and removing pieces of bone.
"It took the breath away from us," he recalls. It was an ordeal, but the surgery was a success. Now in third grade, Olivia plays tennis and soccer and is a voracious reader.
Still, her parents were left with a lot of questions. Could it be something during the pregnancy? Or is it an inherited trait?
This is not a trivial matter.
If the Bushes wanted to have more kids, would those children have the disease, too? Would they be brain damaged? What if Olivia wants to have children?
Cunningham believes some families have almost no chance of reoccurrence, while others have a big chance. He's been studying the disorder for years and believes the key is in the child's DNA. But he still hasn't found a way to tell which families are most likely to have recurrence, and why.
Until doctors know, they can't begin to prevent it.
The truth is, Friend says, science has no idea how a lot of disease works.
This is where the revolution comes in.
IN 2009, FRIEND left his pharma job and formed a nonprofit, Sage Bionetworks. It doesn't do lab experiments, it doesn't have patients and it doesn't make drugs. But Friend believes Sage can help accelerate research.
Typically, scientists latch onto a single gene or protein, convinced that it's the key to a particular disease. Next they identify a new compound that tamps it down (or maybe builds it up). Years of lab work are involved in that alone. Then, if it makes it past the laboratory stage — which it usually doesn't — they've got to test the compound for toxicity. Only then can clinical trials begin to see what happens in the real world. "The vast majority of these things just fail," says Jonathan Derry, Sage's research director.
The thing is, humans have tens of thousands of genes, and even more proteins. Searching one-by-one could take forever.
A second, and larger, problem is that genes and proteins don't play solo. They're part of a system. One gene may be turned off, but there's another one that makes up for it, and a third one that might counteract that, and on and on. It's further complicated because each gene can make many different RNAs, the working copies of DNA.
How does this all fit together? Science doesn't know the half of it.
"If you diagramed the relationships," explains Jonathan Izant, vice president at Sage, "it looks like a spiderweb chart." Only more complex. "It's referred to by the highly technical term of 'hairballs.' "
Hairballs. Tens of thousands of genes and proteins and RNA, interconnected and matted so there's no clear way to tease them apart and figure out what it all means.
Friend, along with many others, thinks the way toward medical progress is to do exactly that: Find the patterns amid the chaos. Make a map and you've made a model of how a disease works.
Only then can docs run your DNA through a sequencing machine and get an answer that means anything.
So where does Sage come in? They have people like Derry, a biologist who leads a team of high-level mathematicians and computer nerds.
You might say he conducts very brainy fishing expeditions. Derry and his ilk write complex algorithms that can dig into genomic data. He sorts through the DNA and RNA and proteins, as well as the health information gathered on regular doctors' visits, to look for patterns. Based on what pops out of his computer, he tries to make predictions about the real causes and effects of disease, whether a particular new medication might help or hurt. Then someone else takes those predictions into the lab.
Here's the thing: You can't make the map based on data from a few patients. You need samples by the hundreds, the thousands, to find the patterns.
The necessary data, however, isn't widely available. Which brings us back to the sharing thing.
We're talking about sharing on a scale so large you'd need to build a new kind of computer platform to handle it.
Even five years ago, this kind of research, called systems biology, wasn't even possible because the large-scale data hadn't been generated. Enter, Sage.
FOR THE PAST two years, Friend has been going around the globe, talking to pharma execs, universities, the FDA, the White House — anyone who will listen. Among other things, he's encouraging scientists to release information for the Sage database. Sage, in turn, will make it available to other scientists.
Friend is quick to point out he's not alone in pushing these ideas. The government is beginning to put sharing requirements — loose ones, anyway — into grant contracts. They've created another kind of central repository to dump data. Other organizations have been on similar tracks.
Some researchers are skeptical that systems biology will find the answers. Finding a single protein is hard enough, they say. More data will lead only to more confusion.
But some scientists are buying in, taking Friend up on the offer of what he calls "shock-and-awe technologies."
Cunningham has signed on. By letting Sage take a look at data from hundreds of craniosynostosis cases, he hopes Derry and the team can help find links to sort out the cause.
Radich is sharing data with Sage, too. He wants to figure out why some patients with acute myeloid leukemia are nonresponders. Cancer consortiums all across the country have joined in, as well.
Friend says they're working with researchers on projects involving Alzheimer's, schizophrenia, Huntington's disease, colon cancer and more. Sage has secured $20 million in funding, from government, industry and foundations.
Surprisingly, perhaps, the patent-chasing private sector is willing to give sharing a try. Friend says a number of pharma companies have signed on for specific projects. The way pharma sees it, nine out of 10 drug trials fail anyway, Friend says. "If it was eight out of 10, they'd double their profitability." Turns out academics are a harder sell. There are turf issues. Egos. And questions of funding, too, since everyone's competing for a limited pool. You don't win the big grants for being big-hearted.
Izant recalls a scientist's email string that was inadvertently forwarded. It said something like, "who in the hell are these people and why should I listen to them telling me what I should do with the data I have spent years putting together?"
Friend says these researchers will just be left behind. He believes that if scientists had earlier access to others' data, they could connect the dots with their own research. And it's just as important to share things that don't work. No sense in repeating that experiment.
Patients might not know it yet, but they actually have the real power, Friend believes. Sage is developing tools to allow patients to be stewards of their own data, whether it's at their doctors' office or in pharma's hands. The Josh Sommerses of the world will one day be able to say, you want me to join your research? Then share your data.
If patients refuse to play with the nonsharers . . . it's a whole new world.
Maureen O'Hagan is a Seattle Times staff reporter. John Lok is a Times staff photographer.