Medical Breakthroughs Set to be Fast-Tracked by Innovative New Health Agency
In 2007, Matthew Might's son, Bertrand, was born with a life-threatening disease that was so rare, doctors couldn't diagnose it. Might, a computer scientist and biologist, eventually realized, "Oh my gosh, he's the only patient in the world with this disease right now." To find effective treatments, new methodologies would need to be developed. But there was no process or playbook for doing that.
Might took it upon himself, along with a team of specialists, to try to find a cure. "What Bertrand really taught me was the visceral sense of urgency when there's suffering, and how to act on that," he said.
He calls it "the agency of urgency"—and patients with more common diseases, such as cancer and Alzheimer's, often feel that same need to take matters into their own hands, as they find their hopes for new treatments running up against bureaucratic systems designed to advance in small, steady steps, not leaps and bounds. "We all hope for a cure," said Florence "Pippy" Rogers, a 65-year-old volunteer with Georgia's chapter of the Alzheimer's Association. She lost her mother to the disease and, these days, worries about herself and her four siblings. "We need to keep accelerating research."
We have a fresh example of what can be achieved by fast-tracking discoveries in healthcare: Covid-19 vaccines.
President Biden has pushed for cancer moonshots since the disease took the life of his son, Beau, in 2015. His administration has now requested $6.5 billion to start a new agency in 2022, called the Advanced Research Projects Agency for Health, or ARPA-H, within the National Institutes of Health. It's based on DARPA, the Department of Defense agency known for hatching world-changing technologies such as drones, GPS and ARPANET, which became the internet.
We have a fresh example of what can be achieved by fast-tracking discoveries in healthcare: Covid-19 vaccines. "Operation Warp Speed was using ARPA-like principles," said Might. "It showed that in a moment of crisis, institutions like NIH can think in an ARPA-like way. So now the question is, why don't we do that all the time?"
But applying the DARPA model to health involves several challenging decisions. I asked experts what could be the hardest question facing advocates of ARPA-H: which health problems it should seek to address. "All the wonderful choices lead to the problem of which ones to choose and prioritize," said Sudip Parikh, CEO of the American Association for the Advancement of Science and executive publisher of the Science family of journals. "There is no objectively right answer."
The Agency of Urgency
ARPA-H will borrow at least three critical ingredients from DARPA: goal-oriented project managers, many from industry; aggressive public-private partnerships; and collaboration among fields that don't always interact. The DARPA concept has been applied to other purposes, including energy and homeland security, with promising results. "We're learning that 'ARPA-ism' is a franchisable model," said Might, a former principal investigator on DARPA projects.
The federal government already pours billions of dollars into advancing research on life-threatening diseases, with much of it channeled through the National Institutes of Health. But the purpose of ARPA-H "isn't just the usual suspects that NIH would fund," said David Walt, a Harvard biochemist, an innovator in gene sequencing and former chair of DARPA's Defense Science Research Council. Whereas some NIH-funded studies aim to gradually improve our understanding of diseases, ARPA-H projects will give full focus to real-world applications; they'll use essential findings from NIH research as starting points, drawing from them to rapidly engineer new technologies that could save lives.
And, ultimately, billions in healthcare costs, if ARPA-H lives up to its predecessor's track record; DARPA's breakthroughs have been economic game-changers, while its fail-fast approach—quickly pulling the plug on projects that aren't panning out—helps to avoid sunken costs. ARPA-H could fuel activities similar to the human genome project, which used existing research to map the base pairs that make up DNA, opening new doors for the biotech industry, sparking economic growth and creating hundreds of thousands of new jobs.
Despite a nearly $4 trillion health economy, "we aren't innovating when it comes to technological capabilities for health," said Liz Feld, president of the Suzanne Wright Foundation for pancreatic cancer.
Individual Diseases Ripe for Innovation
Although the need for innovation is clear, which diseases ARPA-H should tackle is less apparent. One important consideration when choosing health priorities could be "how many people suffer from a disease," said Nancy Kass, a professor of bioethics and public health at Johns Hopkins.
That perspective could justify cancer as a top objective. Cancer and heart disease have long been the two major killers in the U.S. Leonidas Platanias, professor of oncology at Northwestern and director of its cancer center, noted that we've already made significant progress on heart disease. "Anti-cholesterol drugs really have a wide impact," he said. "I don't want to compare one disease to another, but I think cancer may be the most challenging. We need even bigger breakthroughs." He wondered whether ARPA-H should be linked to the part of NIH dedicated to cancer, the National Cancer Institute, "to take maximum advantage of what happens" there.
Previous cancer moonshots have laid a foundation for success. And this sort of disease-by-disease approach makes sense in a way. "We know that concentrating on some diseases has led to treatments," said Parikh. "Think of spinal muscular atrophy or cystic fibrosis. Now, imagine if immune therapies were discovered ten years earlier."
But many advocates think ARPA-H should choose projects that don't revolve around any one disease. "It absolutely has to be disease agnostic," said Feld, president of the pancreatic cancer foundation. "We cannot reach ARPA-H's potential if it's subject to the advocacy of individual patient groups who think their disease is worse than the guy's disease next to them. That's not the way the DARPA model works." Platanias agreed that ARPA-H should "pick the highest concepts and developments that have the best chance" of success.
Finding Connections Between Diseases
Kass, the Hopkins bioethicist, believes that ARPA-H should walk a balance, with some projects focusing on specific diseases and others aspiring to solutions with broader applications, spanning multiple diseases. Being impartial, some have noted, might involve looking at the total "life years" saved by a health innovation; the more diseases addressed by a given breakthrough, the more years of healthy living it may confer. The social and economic value should increase as well.
For multiple payoffs, ARPA-H could concentrate on rare diseases, which can yield important insights for many other diseases, said Might. Every case of cancer and Alzheimer's is, in a way, its own rare disease. Cancer is a genetic disease, like his son Bertrand's rare disorder, and mutations vary widely across cancer patients. "It's safe to say that no two people have ever actually had the same cancer," said Might. In theory, solutions for rare diseases could help us understand how to individualize treatments for more common diseases.
Many experts I talked with support another priority for ARPA-H with implications for multiple diseases: therapies that slow down the aging process. "Aging is the greatest risk factor for every major disease that NIH is studying," said Matt Kaeberlein, a bio-gerontologist at the University of Washington. Yet, "half of one percent of the NIH budget goes to researching the biology of aging. An ARPA-H sized budget would push the field forward at a pace that's hard to imagine."
Might agreed. "It could take ARPA-H to get past the weird stigmas around aging-related research. It could have a tremendous impact on the field."
For example, ARPA-H could try to use mRNA technology to express proteins that affect biological aging, said Kaeberlein. It's an engineering project well-suited to the DARPA model. So is harnessing machine learning to identify biomarkers that assess how fast people are aging. Biological aging clocks, if validated, could quickly reveal whether proposed therapies for aging are working or not. "I think there's huge value in that," said Kaeberlein.
By delivering breakthroughs in computation, ARPA-H could improve diagnostics for many different diseases. That could include improving biowearables for continuously monitoring blood pressure—a hypothetical mentioned in the White House's concept paper on ARPA-H—and advanced imaging technologies. "The high cost of medical imaging is a leading reason why our healthcare costs are the highest in the world," said Feld. "There's no detection test for ALS. No brain detection for Alzheimer's. Innovations in detection technology would save on cost and human suffering."
Some biotech companies may be skeptical about the financial rewards of accelerating such technologies. But ARPA-H could fund public-private partnerships to "de-risk" biotech's involvement—an incentive that harkens back to the advance purchase contracts that companies got during Covid. (Some groups have suggested that ARPA-H could provide advance purchase agreements.)
Parikh is less bullish on creating diagnostics through ARPA-H. Like DARPA, Biden's health agency will enjoy some independence from federal oversight; it may even be located hundreds of miles from DC. That freedom affords some breathing room for innovation, but it could also make it tougher to ensure that algorithms fully consider diverse populations. "That part I really would like the government more involved in," Parikh said.
Might thinks ARPA-H should also explore innovations in clinical trials, which many patients and medical communities view as grindingly slow and requiring too many participants. "We can approve drugs for very tiny patient populations, even at the level of the individual," he said, while emphasizing the need for safety. But Platanias thinks the FDA has become much more flexible in recent years. In the cancer field, at least, "You now see faster approvals for more drugs. Having [more] shortcuts on clinical trial approvals is not necessarily a good idea."
With so many options on the table, ARPA-H needs to show the public a clear framework for measuring the value of potential projects. Kass warned that well-resourced advocates could skew the agency's priorities. They've affected health outcomes before, she noted; fundraising may partly explain larger increases in life expectancy for cystic fibrosis than sickle cell anemia. Engaging diverse communities is a must for ARPA-H. So are partnerships to get the agency's outputs to people who need them. "Research is half the equation," said Kass. "If we don't ensure implementation and access, who cares." The White House concept paper on ARPA-H made a similar point.
As Congress works on authorizing ARPA-H this year, Might is doing what he can to ensure better access to innovation on a patient-by-patient basis. Last year, his son, Bertrand, passed away suddenly from his disorder. He was 12. But Might's sense of urgency has persisted, as he directs the Precision Medicine Institute at the University of Alabama-Birmingham. That urgency "can be carried into an agency like ARPA-H," he said. "It guides what I do as I apply for funding, because I'm trying to build the infrastructure that other parents need. So they don't have to build it from scratch like I did."
Podcast: The Friday Five weekly roundup in health research
The Friday Five covers five stories in health research that you may have missed this week. There are plenty of controversies and troubling ethical issues in science – and we get into many of them in our online magazine – but this news roundup focuses on scientific creativity and progress to give you a therapeutic dose of inspiration headed into the weekend.
Covered in this week's Friday Five:
- Sex differences in cancer
- Promising research on a vaccine for Lyme disease
- Using a super material for brain-like devices
- Measuring your immunity to Covid
- Reducing dementia risk with leisure activities
One day in recent past, scientists at Columbia University’s Creative Machines Lab set up a robotic arm inside a circle of five streaming video cameras and let the robot watch itself move, turn and twist. For about three hours the robot did exactly that—it looked at itself this way and that, like toddlers exploring themselves in a room full of mirrors. By the time the robot stopped, its internal neural network finished learning the relationship between the robot’s motor actions and the volume it occupied in its environment. In other words, the robot built a spatial self-awareness, just like humans do. “We trained its deep neural network to understand how it moved in space,” says Boyuan Chen, one of the scientists who worked on it.
For decades robots have been doing helpful tasks that are too hard, too dangerous, or physically impossible for humans to carry out themselves. Robots are ultimately superior to humans in complex calculations, following rules to a tee and repeating the same steps perfectly. But even the biggest successes for human-robot collaborations—those in manufacturing and automotive industries—still require separating the two for safety reasons. Hardwired for a limited set of tasks, industrial robots don't have the intelligence to know where their robo-parts are in space, how fast they’re moving and when they can endanger a human.
Over the past decade or so, humans have begun to expect more from robots. Engineers have been building smarter versions that can avoid obstacles, follow voice commands, respond to human speech and make simple decisions. Some of them proved invaluable in many natural and man-made disasters like earthquakes, forest fires, nuclear accidents and chemical spills. These disaster recovery robots helped clean up dangerous chemicals, looked for survivors in crumbled buildings, and ventured into radioactive areas to assess damage.
Now roboticists are going a step further, training their creations to do even better: understand their own image in space and interact with humans like humans do. Today, there are already robot-teachers like KeeKo, robot-pets like Moffin, robot-babysitters like iPal, and robotic companions for the elderly like Pepper.
But even these reasonably intelligent creations still have huge limitations, some scientists think. “There are niche applications for the current generations of robots,” says professor Anthony Zador at Cold Spring Harbor Laboratory—but they are not “generalists” who can do varied tasks all on their own, as they mostly lack the abilities to improvise, make decisions based on a multitude of facts or emotions, and adjust to rapidly changing circumstances. “We don’t have general purpose robots that can interact with the world. We’re ages away from that.”
Robotic spatial self-awareness – the achievement by the team at Columbia – is an important step toward creating more intelligent machines. Hod Lipson, professor of mechanical engineering who runs the Columbia lab, says that future robots will need this ability to assist humans better. Knowing how you look and where in space your parts are, decreases the need for human oversight. It also helps the robot to detect and compensate for damage and keep up with its own wear-and-tear. And it allows robots to realize when something is wrong with them or their parts. “We want our robots to learn and continue to grow their minds and bodies on their own,” Chen says. That’s what Zador wants too—and on a much grander level. “I want a robot who can drive my car, take my dog for a walk and have a conversation with me.”
Columbia scientists have trained a robot to become aware of its own "body," so it can map the right path to touch a ball without running into an obstacle, in this case a square.
Jane Nisselson and Yinuo Qin/ Columbia Engineering
Today’s technological advances are making some of these leaps of progress possible. One of them is the so-called Deep Learning—a method that trains artificial intelligence systems to learn and use information similar to how humans do it. Described as a machine learning method based on neural network architectures with multiple layers of processing units, Deep Learning has been used to successfully teach machines to recognize images, understand speech and even write text.
Trained by Google, one of these language machine learning geniuses, BERT, can finish sentences. Another one called GPT3, designed by San Francisco-based company OpenAI, can write little stories. Yet, both of them still make funny mistakes in their linguistic exercises that even a child wouldn’t. According to a paper published by Stanford’s Center for Research on Foundational Models, BERT seems to not understand the word “not.” When asked to fill in the word after “A robin is a __” it correctly answers “bird.” But try inserting the word “not” into that sentence (“A robin is not a __”) and BERT still completes it the same way. Similarly, in one of its stories, GPT3 wrote that if you mix a spoonful of grape juice into your cranberry juice and drink the concoction, you die. It seems that robots, and artificial intelligence systems in general, are still missing some rudimentary facts of life that humans and animals grasp naturally and effortlessly.
How does one give robots a genome? Zador has an idea. We can’t really equip machines with real biological nucleotide-based genes, but we can mimic the neuronal blueprint those genes create.
It's not exactly the robots’ fault. Compared to humans, and all other organisms that have been around for thousands or millions of years, robots are very new. They are missing out on eons of evolutionary data-building. Animals and humans are born with the ability to do certain things because they are pre-wired in them. Flies know how to fly, fish knows how to swim, cats know how to meow, and babies know how to cry. Yet, flies don’t really learn to fly, fish doesn’t learn to swim, cats don’t learn to meow, and babies don’t learn to cry—they are born able to execute such behaviors because they’re preprogrammed to do so. All that happens thanks to the millions of years of evolutions wired into their respective genomes, which give rise to the brain’s neural networks responsible for these behaviors. Robots are the newbies, missing out on that trove of information, Zador argues.
A neuroscience professor who studies how brain circuitry generates various behaviors, Zador has a different approach to developing the robotic mind. Until their creators figure out a way to imbue the bots with that information, robots will remain quite limited in their abilities. Each model will only be able to do certain things it was programmed to do, but it will never go above and beyond its original code. So Zador argues that we have to start giving robots a genome.
How does one do that? Zador has an idea. We can’t really equip machines with real biological nucleotide-based genes, but we can mimic the neuronal blueprint those genes create. Genomes lay out rules for brain development. Specifically, the genome encodes blueprints for wiring up our nervous system—the details of which neurons are connected, the strength of those connections and other specs that will later hold the information learned throughout life. “Our genomes serve as blueprints for building our nervous system and these blueprints give rise to a human brain, which contains about 100 billion neurons,” Zador says.
If you think what a genome is, he explains, it is essentially a very compact and compressed form of information storage. Conceptually, genomes are similar to CliffsNotes and other study guides. When students read these short summaries, they know about what happened in a book, without actually reading that book. And that’s how we should be designing the next generation of robots if we ever want them to act like humans, Zador says. “We should give them a set of behavioral CliffsNotes, which they can then unwrap into brain-like structures.” Robots that have such brain-like structures will acquire a set of basic rules to generate basic behaviors and use them to learn more complex ones.
Currently Zador is in the process of developing algorithms that function like simple rules that generate such behaviors. “My algorithms would write these CliffsNotes, outlining how to solve a particular problem,” he explains. “And then, the neural networks will use these CliffsNotes to figure out which ones are useful and use them in their behaviors.” That’s how all living beings operate. They use the pre-programmed info from their genetics to adapt to their changing environments and learn what’s necessary to survive and thrive in these settings.
For example, a robot’s neural network could draw from CliffsNotes with “genetic” instructions for how to be aware of its own body or learn to adjust its movements. And other, different sets of CliffsNotes may imbue it with the basics of physical safety or the fundamentals of speech.
At the moment, Zador is working on algorithms that are trying to mimic neuronal blueprints for very simple organisms—such as earthworms, which have only 302 neurons and about 7000 synapses compared to the millions we have. That’s how evolution worked, too—expanding the brains from simple creatures to more complex to the Homo Sapiens. But if it took millions of years to arrive at modern humans, how long would it take scientists to forge a robot with human intelligence? That’s a billion-dollar question. Yet, Zador is optimistic. “My hypotheses is that if you can build simple organisms that can interact with the world, then the higher level functions will not be nearly as challenging as they currently are.”
Lina Zeldovich has written about science, medicine and technology for Popular Science, Smithsonian, National Geographic, Scientific American, Reader’s Digest, the New York Times and other major national and international publications. A Columbia J-School alumna, she has won several awards for her stories, including the ASJA Crisis Coverage Award for Covid reporting, and has been a contributing editor at Nautilus Magazine. In 2021, Zeldovich released her first book, The Other Dark Matter, published by the University of Chicago Press, about the science and business of turning waste into wealth and health. You can find her on http://linazeldovich.com/ and @linazeldovich.