The Quest for the First FDA-Approved Video Game

A book excerpt from Power Play: How Video Games Can Save the World by Asi Burak and Laura Parker

Dr. Adam Gazzaley is a neurology professor at the University of California in San Francisco, and director of the Gazzaley Lab. The lab is best known for using technology, especially video games, to thwart cognitive decline. For years, Gazzaley’s work has focused on memory and attention—in particular, how these processes change with aging and dementia, and what we can do to keep aging minds active. His lab couples behavioral assessments with techniques like functional magnetic resonance imaging (fMRI), electroencephalography (EEG) and transcranial magnetic stimulation (TMS).

Dr. Adam Gazzaley

In person, Gazzaley cuts a striking figure. He’s tall and broad-shouldered, with a healthy crop of silver hair and dark brown eyes, and a placid, unwavering demeanor. One might call him intimidating, but there’s a keen sense of humor lurking beneath the surface, for those willing to tease it out. “Adam works really hard and plays really hard,” Brett Morrison, a neurology resident at Johns Hopkins Hospital and a friend of Gazzaley’s, once admitted.

Born in Brooklyn, Gazzaley developed an early interest in science as a devotee of Carl Sagan’s “Cosmos” series on PBS. After competing in a statewide science fair with a project on solar energy, Gazzaley petitioned his parents to let him attend the Bronx High School of Science, despite the fact the school was a near four-hour roundtrip by public transport from his house in Howard Beach. He passed the admission test and spent the next four years getting up at 5.30am to catch a bus.

Gazzaley felt like an underdog—Howard Beach was a working class neighborhood, and none of his friends understood his determination to go to college. After various Ivy League schools turned him down, Gazzaley was accepted into the State University of New York Binghamton—where, in his sophomore year as a biochemistry major, he was nearly expelled for blowing up a toilet. (His plan to create a chemical reaction in the lake behind campus with a large piece of sodium metal backfired—literally—when he tested it first in a toilet block with 30 other students watching on. “All that was left was two pipes sticking out of the wall,” he recalled. No one was hurt, but Gazzaley was placed on something called “social probation”—in other words, no parties.)

While trying to fulfill his humanities requirements, Gazzaley took a series of classes focused on different visions of the future; in one class, the lecturer showed students research on nanobots, and how they might one day be used in neurosurgery to complement—if not entirely replace—a brain surgeon’s hands. “It was almost like science fiction, and it just sparked this little bug in me,” Gazzaley later recalled. He took out every textbook he could find on neuroscience, and later applied, and was accepted, to Mount Sinai School of Medicine’s M.D.-PhD program.

Under the guidance of neurobiologist John Morrison, Gazzaley began looking into how aging alters cognitive function. While studying memory in a group of monkeys, Gazzaley developed a method for measuring changes in proteins called NMDA receptors, which play a role in memory formation. He published three papers between 1996 and 1997, which established that memory-related neural circuitry is not damaged by the process of aging, like it is in Alzheimer’s disease. Instead, Gazzaley reasoned, the cause of memory loss due to simple aging seems to come from subtle molecular changes in the brain, which prevent neurons from communicating with one another the way they’re supposed to.

It was a significant achievement for a young scientist. According to Morrison, Gazzaley was able to master what few scientists his age could: not just successfully answering a research query with tested data, but turning the results into something publishable. Another MSSM neuroscientist, Deanna Benson, said once that working with Gazzaley back then was more like working with a peer than a student, despite her initial skepticism of his fashion sense: “He dressed in tank tops and shorts. He used to lift weights a lot, too, so he looked kind of all buffed out—like the kind of guy who spent more time in the gym than he ever would spend in the lab.”

From 1999 to 2002, Gazzaley completed his residency in neurology at the University of Pennsylvania, where he met neurobiologist Mark D’Esposito, who at the time was using fMRI to study cognition. Gazzaley liked the idea of being able to visualize brain activity and, when D’Esposito later invited Gazzaley to come to California to do his postdoctoral training at UC Berkley, he accepted.

**

Soon after, Gazzaley became restless. His work to date had focused on things like distraction, multi-tasking, focus, decision-making and memory. What he had learned was that, while most people experience ups and downs within each of these areas, things get steadily worse as one ages, and Gazzaley’s research used fMRI to reveal what’s actually happening inside the brain during cognitive decline. But, while the work was satisfying, something was missing. Namely, he hadn’t found a way to fix the problem. “I felt that we weren’t doing anything to help the people that are troubled by it and frustrated by it and maybe even impaired by it,” he said recently. “It was intellectually fascinating, but my goal has always been to help people.”

This restlessness was also driven in part by Gazzaley’s growing profile. He was doing as many as 70 talks a year, and becoming increasingly embarrassed that none ended with any sense of hope. “I’d finish a talk and the [audience] would be like, ‘Wait, that’s it?’. It was like a movie where everyone dies in the end and the credits roll. And it’s like, the worst movie of all time. No matter how good the story is, or how interesting it is, when you in the audience, listening to this stuff, it’s just not satisfying in any way to end with bad news.”

But where could he begin? There was already a growing movement around “brain training” games, which included games and apps like Dr. Kawashima's How Old Is Your Brain?, a popular puzzle game for the Nintendo DS that singlehandedly raised the median age of the console’s audience by at least twenty years. Scientists, on the other hand, were skeptical. In 2014, a group of leading cognitive psychologists and neuroscientists released a statement warning that the scientific literature did not support the claims made by many brain training apps and games, namely that they help prevent cognitive decline.

In early 2016, the company behind one of the most popular brain training apps, Luminosity, agreed to pay $2 million to settle false marketing claims that the app alleviated symptoms of Alzheimer’s. The US Federal Trade Commission asked the company to contact its over 70 million users and offer them the chance to cancel their subscription to some 40 online games. (Despite this, the market shows no signs of slowing down: according to recent projections, the global cognitive assessment and training market is expected to grow from $2.4 billion in 2015 to $7.5 billion by 2020 for digital and paper-based combined. By comparison, the digital apps market is forecasted to reach $6 billion by 2020, according to a recent report that also predicts that biometrics-aided meditation will be the next big thing in consumer wellness, with more than one million adults in North America expected to take a self-administered annual brain health check-up via an iPad or Android tablet.)

Gazzaley was one the scientists who signed the 2014 statement, but later admitted to having doubts, chiefly because he feared that such a bold move could scare off potential investors interested in funding further research into brain training projects. He wasn’t interested in making a brain training game per se, but he liked the idea of interactivity. He thought the solution needed to be more immersive—something that could deliver long-term benefits, not just immediate satisfaction.

One method would be to simply devise a series of cognitive tasks that players could perform over and over again. But this wouldn’t take advantage of the brain’s plasticity. When we are young, our brains are more plastic—we can learn multiple languages with speed and efficiency, for example. As we age, our brains become increasingly rigid—but it’s still possible to get better at something over time. “The big question was what type of interactivity is even tolerable to do for any amount of time,” Gazzaley said. The second, equally important, question, was, how do you repeatedly engage someone in an activity enough times to determine a long-term benefit, not just an immediate one?

To get answers, Gazzaley called Daphne Bavalier, a cognitive neuroscientist at the University of Geneva studying the real-world effects of video games. Born in Paris, Bavalier got her PhD at MIT, followed by a postdoctoral fellowship at the Salk Institute in San Diego, where she became interested in brain plasticity and learning. While conducting a study into the effect of age on peripheral vision, Bavalier found that all the subjects participating in the study, most of them students, achieved a nearly perfect score on a certain digital field-of-view tasks[i]. “When we looked for commonalities between the pilot subjects, we found that all of them belonged to an action video game club,” Bavalier said later. Working together with an undergraduate lab technician named Shawn Green, Bavalier switched the project to study the impact of video games on cognition; the pair eventually published the results of the study in Nature, concluding that action video games helped to improve attention.

Bavalier was hooked. She wanted to find out the limits of video game-induced changes. She turned her attention to vision and contrast sensitivity (differentiating between subtle shades of the color grey) by conducting another study in which she found that forcing people to play action games actually improved their ability to detect contrast better and make better sense of “visual clutter”. For example, a driver with better contrast sensitivity has an easier time identifying the car directly ahead when there’s fog on the road. To determine whether it’s feasible to train a non-gamer to have the same skills, Bavalier asked a group of non-gamers to play action games for a few weeks, and then sent them home, forbidding them to touch another game. Every few months, she’d ask them back to the lab to check their vision; she found the positive effects of that short burst of game playing were not wearing off. “We looked at the effect of playing action games on this visual skill of contrast sensitivity, and we've seen effects that last up to two years,” Bavelier told NPR in 2010.

Taking things a step further, Bavalier partnered with UC Berkeley, McMaster University and the school of interactive games and media at the Rochester Institute of Technology to see whether action games could be modified to help patients with amblyopia, or lazy eye. “There is evidence that certain computer-based activities—like reading—are more tiring for vision than others,” Bavalier explained during a 2012 interview. Reading forces the eyes to work over a very narrow range of spatial frequencies at high contrast, which can impact vision differently than playing video games, which tend to be much richer. “Clearly, it is not the monitor resolution itself, but what is actually on the screen and how you interact with it that is important for vision,” Bavalier said.

Bavalier’s work seemed to show that, though the brain’s plasticity sags dramatically as we age, it is possible to curb the effects of this decline by stimulating the brain in the right way—in this case, through video games. Gazzaley was convinced. “I thought, yes, I can build a customized video game that specifically targets these deficits in the older adult brain, and see if the effects are passed on,” Gazzaley said.

So he reached out to a few close friends who happened to work at LucasArts, the video game division of the now Disney-owned Lucasfilm. (LucasArts ceased operations in 2013.) The problem was Gazzaley didn’t know what kind of video game to make. A first-person shooter where players navigate different environments killing enemies? A third-person adventure game where it’s more about puzzle solving and spatial awareness? Or a fast-paced action game, like the kind Bavalier had used? Gazzaley knew it needed to be something simple—because, after all, it was intended for an aging population—but he also wanted it to be engaging.

One night, he dreamt he was playing a driving game, but instead of racing against other drivers or avoiding barrels on the track, he was responding to differently colored signs. “I woke up and immediately sketched out some ideas in PhotoShop,” he said. (Apparently, it’s common for Gazzaley to dream solutions to the problem he’s working on.) He then sent the sketches to his LucasArts friends. The then-executive art director, Matt Omernick, introduced Gazzaley to one of the company’s engineers, Eric Johnson, and a game designer named Noah Falstein, now the chief game designer at Google, who agreed to help Gazzaley build his driving game.

The team began meeting in Gazzaley’s office and brainstorm design principles and goals over sushi and beer. Gazzaley began reading up on game design. The team spent months building a prototype of a game they called NeuroRacer, a 3D driving game similar to something one might play at the DMV to qualify for a driver’s license. In the game, participants steer a car along a winding road using their left thumb. Now and again, a sign pops up. If participants recognize the sign to be of a particular shape and color—something they are instructed to watch out for—they must shoot it down using a finger on their right hand. These two tasks test cognitive skills like focus, multitasking and working memory, which is the ability to temporarily retain multiple pieces of information. “[Both tasks] demand a lot of attention,” Gazzaley said. “They are both getting harder as you get better. If you trade off one, the other one will suffer invariably. One of the rules of the game is that you only level up if you improve on both tasks, not just one.”

That, of course, was the point. Gazzaley wanted participants’ brains to figure out how to multitask without either task suffering. He regards cognitive control disabilities as kind of a triad of working memory, attention abilities and goal management. NeuroRacer is an example of a goal management challenge. Participants have two goals, and they have to accomplish them both at the same time. There’s no strategy or shortcut involved. It’s not about inherent skill. It’s about practice. And the hypothesis was that, if someone could get better at goal management, their attention abilities and working memory would also improve because they similar parts of the brain.

The team began testing NeuroRacer on the thousands of volunteers who walked through Gazzaley’s lab for various studies. Finally, they recruited 180 adults between the ages of 20 and 70 for an in-lab study. The team found that older participants had a much harder time responding to the multitasking tasks than younger participants. (An expected outcome.) Next, 46 new volunteers aged 60 to 85 were asked to go home and play NeuroRacer for one hour a day, three days a week. After four weeks, the volunteers were tested again. On average, Gazzaley discovered that after playing the game as prescribed, the older participants fared much better in the multitasking requirement of the game (i.e. driving and reacting to signs at the same time) than both younger and older participants who had played NeuroRacer just once, as well as a control group who didn’t play the game at all.

The study also found that those who had played NeuroRacer for a month at home fared better in tests of two other cognitive skills that involve similar areas of the brain as multitasking, but which weren’t directly exercised in NeuroRacer. The volunteers got better at responding to infrequent important stimuli and also improved their working memory—just as Gazzaley had predicted. More importantly, these skills did not deteriorate over time after six months, even without practice. (The game was only accessible via the laptops provided by the lab, which were taken away after the study concluded.) “It was shocking for us,” Gazzaley recalled. “If we had known that was the case we would have tested much more than six months, but we just didn’t expect it.”

In September 2013, the results of the NeuroRacer study appeared on the cover of Nature. “These provide the evidence of how a custom designed video game can be used to assess cognitive abilities across the lifespan, evaluate underlying neural mechanisms, and serve as a powerful tool for cognitive enhancement,” Gazzaley concluded.

**

The Nature cover brought Gazzaley and NeuroRacer a lot of attention. But Gazzaley didn’t want people to assume this was the holy grail of neurogames. He thought of it more as a blueprint for how one should marry the two fields. He also realized he had to get NeuroRacer out of the lab and into people’s hands. The game would need a serious update first; it was created as an experimental prototype and not a commercial product. Gazzaley filed a patent for the methodology behind the game, and set about trying to find someone who could help him turn it into a real game.

While attending a health and technology conference in Boston in 2011, Gazzaley had hit it off with a health industry entrepreneur named Eddie Martucci, who was part of an investment startup named PureTech at the time. Martucci had come to the conference looking for inspiration. He wanted to form a company that could successfully combine neuroscience with entertainment software to create a new kind of medicine, one that could be both stimulating and effective. “Adam was giving his talk in the smallest conference room; it contrasted with magnitude of what he was talking about,” Martucci recently recalled.

Martucci approached Gazzaley after his talk, and the two went for lunch. They spent the next four hours talking. Martucci didn’t attend any other talks at the conference. A few months later, he invited Gazzaley back to Boston. “NeuroRacer immediately went on our shortlist of technology that we were looking for,” Martucci said. Martucci had set a high bar for his future company: he wanted a solid platform with the right scientific pillars that could be built into software. He imagined something that could be both a video game and a medical device, with FDA approval, which doctors could prescribe to patients as readily as a drug. He knew it would take a lot of time and money to achieve this, not to mention exhaustive clinical trials.

The result, Akili Interactive, was incorporated in late 2011. Gazzaley later came onboard as co-founder and chief science advisor. Martucci hired game designers and engineers, whose first task was rebuilding NeuroRacer into a consumer-grade action video game. The result, which took a year and a half, was a game called Project Evo. In this game, players have to guide an alien spacecraft through a canyon (instead of driving a car down a winding road). To move the spacecraft, players click on red fish that appear on the screen while at the same time ignoring distractions—like other animals or fish that also appear. The game is self-sufficient: it doesn’t need to be programmed by a doctor of clinician. Instead, it adapts naturally to each patient, increasing the speed of the ship and the speed and number of objects that the player must both click on, and avoid.

The difficulty of the game changes to keep players at a certain success rate, i.e. that sweet spot where things are just starting to go wrong. The game also adapts second-by-second to give players incentive to keep pushing themselves to succeed, but also to challenge them neurologically. The goal of the game is to get players to become better at assessing sensory processing and interference processing at the same time. When the game senses a plateau—i.e. that a player has stopped making large gains—it reconfigures itself to be closer to the player’s difficulty level. Akili calls Project Evo a “proprietary multitasking cognitive trainer”. It’s playable on mobile phones and tablets. “It’s better [than NeuroRacer] in pretty much every way: art, music, story, better interactivity, more challenges, more user friendly, higher rewards, more accessible, and it’s cloud-based,” Gazzaley said. “It’s just all the things we couldn’t do in the lab.”

Next, Martucci wanted to conduct a clinical trial for the game in exactly the same way a pharmaceutical company might conduct a new drug trial. The problem was that NeuroRacer wasn’t built to target a particular ‘population’: i.e. ADHD sufferers or Alzheimer’s sufferers. But sensory processing seemed to be a problem in a lot of neurological conditions. Still, if Martucci wanted FDA approval, he’d need to make Project Evo’s aims more specific. He spent nearly a year speaking to experts and mapping out the best populations where Project Evo could have the best effect. One of these was ADHD. Word of what Akili reached Shire, the pharmaceutical company behind Adderall, a leading ADHD prescription drug.

At the time, Shire was making equity investments in companies it wanted to work with in the future. Shire liked what Akili was proposing and invested in the company, partnering on a pilot study in 2013 that involved an ADHD-targeted version of Project Evo. (The results were released in October 2015 at the 62nd annual conference of the American Academy of Child and Adolescent Psychiatry.[ii]) But how do you do a drug trial when the drug is a video game? Akili went through the same process as a clinical drug trial, recruiting participants at four sites throughout the US, though online patient boards and doctor’s officers. “Patients who want to participate come into the site, get a full workup of their symptoms, and instead of going home with a drug, they go home with an iPad.” (In the recruitment process, Akili used the term “new medical device” rather than “video game”, so as not to scare off parents.)

The subsequent trial involved 40 children between the ages of 8-12 years in two groups: ADHD and “neurotypical”— a population that does not have whatever condition/disease is being studied. The kids were prescribed the game to play at home, on an iPad, once a day over a period of four weeks. They were assessed clinically both before and after the study, plus given a computerized assessment of attention, memory and impulsivity. Parents were also asked to weigh in on their children’s behavior before and after the study. After the trial, eleven of the 19 “outcome variables” in the computerized working memory test in the ADHD cohort showed significant improvement; parents also reported improvement in their children’s working memory and inhibition, leading to the conclusion that the game “may be effective” in improving attention, working memory and inhibition in pediatric ADHD population.

More importantly, the kids had a blast. “Although medication is the front line treatment for children with ADHD, there remains a need for the development of accessible, non-drug interventions,” the subsequent report concluded. “Delivery of an effective intervention as an action video game on a mobile device may in the future provide an engaging way to treat ADHD symptoms in an at-home setting.”

Pfizer then came knocking, wanting a similar drug trial for early Alzheimer’s. The company was specifically interested in Project Evo’s data capture feature, which collects about 60 data points per second. Akili developed an Alzheimer’s-specific version of Project Evo and, in partnership with Pfizer, plans to embark on a similar trial with people aged between 60 and 70 who have the neural markers for early Alzheimer’s. The goal is to test whether the game can identify those who are at risk for developing Alzheimer’s down the track, something Martucci says everyone in the pharmaceutical industry is currently trying to do.

Secondary platforms, games and programs are already in the works, so that when Project Evo gets FDA approval—it no longer seems to be a question of if—there will be something in the pipeline. As with any new drug, FDA approval process is lengthy and involved. Akili is submitting Project Evo as a medical device, not a drug—similar to a hip implant or a new electrode. The company has sought FDA’s feedback in past Project Evo clinical trials, and the response, according to Martucci, has been positive.

While the FDA cannot yet say a hard yes or no to whether the game will be approved, Martucci says the administration is “comfortable” with what Akili is doing. He expects the final approval to arrive before the end of 2017. Before that can happen, Akili has to do more to prove Project Evo’s effectiveness among other populations. A larger, randomized control trial on ADHD has been planned, involving a few hundred patients across ten sites throughout the US—a significantly larger group than the first trial. Another trial targeting autism is also in the works. “We want a label that reads ‘for the treatment for ADHD’, or ‘for the treatment of X disorder’,” Martucci said. “For these claims, we need to get FDA approval for each patient population. And we will.”

Later, in his lab, Gazzaley expressed a similar sentiment. “If five years from now doctors are pulling out prescription pads and writing, ‘Four weeks of iPad training’, as opposed to a drug—that will be incredibly exciting future.”

An excerpt from the book Power Play: How Video Games Can Save the World by Asi Burak and Laura Parker. Copyright (c) 2017 by the authors and reprinted by permission of St. Martin's Press.

[i] C.S. Green, D. Bavelier, “Action-Video-Game Experience Alters the Spatial Resolution of Vision,” Psychological Science (January 2007): 88-94, http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2896830/

[ii] Katherine Ellison, “Video Game Is Built to Be Prescribed to Children With A.D.H.D.,” New York Times, November 23, 2015, http://well.blogs.nytimes.com/2015/11/23/video-game-is-built-to-be-prescribed-to-children-with-a-d-h-d/