As a pediatric ophthalmologist, I do my best to assure that every young patient I examine will have a lifetime of perfect sight. The condition that I battle most commonly is amblyopia, or “lazy eye,” in which the eye is healthy but does not develop vision — simply because the brain doesn’t receive proper input when a child’s visual system is “learning” how to see.
When I can diagnose amblyopia early enough, I can treat it with an eye patch or eye drops to block the “good” eye, giving the eye with amblyopia time to catch up. But amblyopia does not fight fairly: about half of affected kids have no visible signs of the condition. As a result, amblyopia silently steals the sight of hundreds of thousands of children — many of whom will never get their vision back because treatment started too late.
It is this problem that inspired me to develop the Pediatric Vision Scanner (PVS). I am also trained as an engineer, and more than 20 years ago, while working with my mentor David Guyton, MD, at Johns Hopkins, we had an idea of how to help pediatricians and school nurses readily identify kids with amblyopia and its companion condition, strabismus (misaligned eyes).
When the eye focuses on a target, the image lines up with a particular area of the retina, the fovea. Knowing that the fovea has certain optical properties, Dr. Guyton wondered whether we could use polarized laser light to scan the retina and find the fovea. If so, this would tell us where the eye was looking.
Our prototype PVS did just that. We scanned both retinas at the same time, and if both eyes weren’t looking directly at a smiley-face target inside the machine at exactly the same time, it indicated a likely problem. In the early 2000s, I moved to Boston Children’s Hospital, where we did the first clinical testing of the PVS.
We were shocked at how well it worked. Kids with normal eyesight were able to pass the test, and the kids with a problem could not. The PVS was able to classify children with 97 percent accuracy — unheard of in the vision screening world.
At that point I thought, “Wow, I can just call up a company or a venture capitalist and they’ll see what a great product we have and start producing it and the world will be freed from the burden of amblyopia.”
That did not happen.
The companies and investors we talked to either didn’t believe the PVS could work that well, or they didn’t think a company selling the PVS would be profitable enough to be worth the risk. “Plus,” they all said, “you still need to get through FDA.”
I realized that it was probably going to take years to get the PVS into the world, and that I would need to start my own company to be sure that the leadership would have the necessary patience and persistence to see it through. I enlisted Justin Shaka, who had just left Boston Children’s to get his MBA at Carnegie Mellon.
The two of us spun out REBIScan. We licensed the inventions (which had been patented by Boston Children’s and Johns Hopkins) and raised money as best we could with a combination of grants, gifts and small investments from friends, family, the Massachusetts Lions Club District 33-Y and Boston Children’s Technology Development Fund. Over the next several years, we were able to build prototypes and loan them out to a few centers to conduct clinical trials.
The trials went incredibly well, again showing accuracy in the 95 percent range — way better than any other screening technology out there. Again I thought, “Okay, now we have proven the technology and we are going to be flooded with investments and start producing the device.”
But the potential large investors continued to sing the same tune: “Call us when you get FDA.”
FDA submission… and resubmission
We had heard horror stories about getting medical devices through the FDA, but we thought we would be different. After all, there were already several (less accurate) vision screeners on the market, and many other scanning laser devices had already cleared the FDA.
But we were not different. The hoops we had to jump through seemed endless. Our paper application to the FDA was thicker than a credit card standing on end. Months later we received a list of dozens of “deficiencies” in our application. These took many more months to work through as we conducted additional studies, gathered additional data and responded to questions. But that just led to more deficiencies, more questions, more studies.
Some of the testing we were asked to do made no sense at all. For example, we had shown high reproducibility when we re-tested 100 children a full year after their original test, and 100 percent consistent results in young adults over eight consecutive tests (two devices x two operators x two tests each). And we showed that the device’s output was 99.7 percent repeatable when tested hundreds of times on an optical bench. But the FDA wanted us to perform repeated testing (three devices, three operators, three tests) on a series of young children — 27 consecutive tests — and still get the same result every time! While most children can sit through one test, or maybe two or three, it takes some concentration to fixate exactly on the smiley face, and we were not aware of any child who could do so 27 consecutive times with consistent results. The FDA’s new requirement was going to be impossible to meet.
To appeal, we created a video of repeated testing in a single child. It clearly showed the difficulty: She was all smiles on the first test, but by the fourth test she was sighing in boredom and by the 10th test she simply refused to look at the machine.
Nonetheless, we were told that the repeated testing was required. The process of appealing and negotiating took months to work out.
When things seemed the most dire, I had to remind myself, “The FDA is here to protect us from snake oil salesmen.” And that proving safety and effectiveness would help pediatricians make an informed choice to use the PVS and, in the long term, benefit the kids receiving the test.
More than once during this trying 35-month period, we came close to running out of money. But just as the till was running empty, another supportive investor was willing to step in and write a check to keep us going.
Finally, on June 8, 2016, I got a text from Justin Shaka, now CEO of REBIScan. We could now market the PVS!
I like to cycle, and I remember a day when I was climbing a hill, heading for a landmark. The hill was enormous. I climbed and climbed, keeping the top in sight, and with sweat and effort and persistence I at last reached the top. And what was just over the crest of the hill? Not the landmark — another hill, this one even higher.
That is sort of where we are now. With FDA clearance behind us, we still need to raise enough investor funds to build a mass-produced version of the PVS. As we enter discussions with potential investors, even with FDA approval, we expect some will want more clinical results, or signed purchase orders, or evidence that we are making a profit before they will invest. In the medical device field, with so many hazards to be negotiated, only a very few visionary investors are willing to accept those risks to move the field forward.
So the long climb to market continues. But every time I see another child with vision loss whose amblyopia has been missed, I gain the strength to work and sweat and push — until every child is screened by the PVS at every well visit, and until every child with amblyopia is caught in time to receive sight-saving treatment. Or, as we say at Boston Children’s Hospital, until every child is well.