Even with brilliant idea and deep pockets, risks high for start-ups

The monkeys could not only use a joystick to play a pinball-like game, but also control the action on the screen just by thinking. An array of electrodes affixed to their brains eavesdropped on a cluster of about 30 neurons and instantly interpreted what the monkeys intended to do. It was perhaps the most sophisticated interface between a living brain and a computer yet developed.

"I thought that it was an incredible technology that really needed to be moved forward," recalls Surgenor, who had been an executive at Haemonetics Corp. and Genzyme Corp.

In 2003, Surgenor became chief executive of Cyberkinetics Inc., helping the company raise more than $40 million, go public, and shepherd the technology he'd seen into human testing. Just last month, he finished shutting down the company, selling off the last of its assets to another local neurotechnology company for $350,000.

What happened in the intervening six years offers a glimpse of how challenging it can be even for a well-funded start-up to bring a breakthrough technology to market - especially in the regulated world of medical devices.

Cyberkinetics had the benefit of several deep-pocketed backers who believed the company was developing something that might change the lives of patients who'd lost the ability to move or speak as the result of a spinal cord injury or a neurodegenerative disease such as amyotrophic lateral sclerosis. With an implant, they might be able to control a computer or direct the movement of a prosthetic limb simply by thinking.

Early on, Boston-based Oxford Bioscience Partners gave the company $4 million, and Thermo Electron Corp. founder George Hatsopoulos ponied up $1 million.

"We believed this was cutting-edge technology for the future," says Jeff Barnes, a partner at Oxford.

In June 2004, Cyberkinetics implanted the device in its first human patient, a 25-year-old named Matthew Nagle who was paralyzed from the neck down after being stabbed. By thought alone, he was able to turn lights on and off, control the movement of a cursor on a computer screen, and open and close a prosthetic hand. That first implant resulted in a shower of press for the Foxborough company.

Even though the BrainGate was implanted in four other patients as part of the initial clinical trial, the company found it didn't work consistently enough - although in some patients, on some days, it seemed to respond perfectly to the neurons firing in their motor cortex, the part of the brain responsible for movement. "It's probably the most complicated product I've ever seen," Surgenor says. "There was a device component, a software component, and a human component."

It was clear the BrainGate would require more development. And then there was the question of how big the eventual market was. The total number of people suffering from spinal cord injury or ALS were "in the single-digit thousands," Surgenor says. How many of them would opt for the implant procedure, which entailed not just drilling a hole in the skull, but leaving an exit point for wires that carried the brain's signals?

In October 2004, Cyberkinetics went public on the thinly traded OTC Bulletin Board, through a reverse-merger with a Nevada shell company that had purportedly been established to mine copper, nickel, and platinum. The idea, explains Barnes, was that the public listing would enable a broader range of investors - not just venture capital firms. "I wasn't the number one cheerleader for that," Surgenor says, though the company raised more money as a public entity than it did as a privately held company. (At its peak, in late 2004, the stock traded at $6.50.)

In 2007, the company, which employed about 40 people at its peak, used its stock to acquire a Purdue University spin-out, Andara Life Science Inc. Andara was developing an implantable electrical stimulator intended to restore sensation and some movement in patients who'd experienced a severe spinal cord injury. The company filed for a special "humanitarian device exemption" with the Food and Drug Administration, which was created to lower the hurdles of FDA approval for devices that are targeted at small patient populations and show at least probable benefit. (Cyberkinetics estimated that about 4,000 people might be candidates for the Andara implant.) Clinical data on the Andara showed it was more likely to restore feeling than movement in patients.

BrainGate was essentially sidelined so the company could focus on the Andara stimulator, which they told investors might be commercialized as early as late 2007 - depending on the FDA.

But by November 2008, the company only had enough cash to last a month, and it received an FDA letter asking more questions about its Andara application. The company's shares were trading at a penny. Surgenor was spending much of his time negotiating with lenders, and he realized that the company simply didn't have any more time to wait out the FDA.

As Surgenor set about winding down the company, he managed to ensure its intellectual property had a chance to find a way to the market. He sold the still-not-approved Andara stimulator to Waltham-based Neurometrix Inc. and a Utah-based division of Cyberkinetics, which makes products for researchers studying the brain-computer interface, to employees. With funding from the National Institutes of Health, work continues at Brown University on a second-generation of the BrainGate technology - this one with a wireless link from the implanted electrode to the outside world.

Oxford Bioscience Partners lost its entire investment in Cyberkinetics. Barnes dubs it one of the most challenging investments the firm has done. "I still believe that the technology can be developed into a great product," he says.

"I'm not afraid of doing things that are pioneering," says Surgenor, who is setting up a consulting firm called Red Sky Partners. But, "the ability to finance things that are exciting but have an unclear path to the market is just really tough. And BrainGate was the poster child for that."

Scott Kirsner can be reached at kirsner@pobox.com.