Gene therapy has been rhapsodized and vilified in its nearly two decades of human testing, helping some and making others sicker. But a new 12-month clinical trial has shown that, at least in one ocular disease, it appears safe and—perhaps even more impressive—effective.

The research, part of a phase I clinical trial to test the safety of the treatment, was published as a letter to the editor in The New England Journal of Medicine earlier this week and will be in the September issue of Human Gene Therapy. (The paper was co-authored by about a dozen researchers, two of whom own equity in a company that could profit from a commercialized version of this procedure.)

The researchers report that three young adults with severe vision impairment from a hereditary disease maintained improved eyesight a year after gene therapy was administered—and didn't suffer any health side effects in the meantime. Gene therapy, which often employs viruses to deliver the good genes to a body's target cells, has been known to trigger severe immune responses and was blamed for the death of an 18-year-old in 1999, who was receiving gene therapy for a hereditary metabolic disorder.

The test subjects suffer from Leber congenital amaurosis (LCA), a form of hereditary retinal degeneration that occurs in infants and young children and is relatively rare. Most people who have lost vision due to hereditary retinal degeneration have either no photoreceptors with which to perceive light or photoreceptors that don't work. "This disease has a little bit of both," explains lead study author Artur Cideciyan, an associate research professor at the Scheie Eye Institute at the University of Pennsylvania. "It's a complex disease."

The participants, ages 22 to 25, all had a mutation on the RPE65 gene, the signature of LCA. Doctors administered normal RPE65 (delivered via an rAAV2 vector) into a site on one of the retinas that had the most photoreceptors left. And, Cideciyan says, "Within weeks of the procedure there was a substantial, very significant improvement in sensitivity to see dim lights."

Those in the study weren't entirely blind before the therapy. In fact, Cideciyan notes, "some of them can read," but by that he means they can see the 'E' at the top of a standard vision chart. "That does not represent the visual inability that they have because they also have a large loss of sensitivity of light… So they may be able to read on a high-contrast chart, but they are further disabled in less light environments."

What has been more surprising to the researchers is not that the improvements have held for the year—that much had already been documented in animal studies—but that, at least in one of the patients, the therapy helped train the eye to see better. "It can completely bypass the dysfunction so it can give as much vision as there are retained receptors." One patient reported that she was newly able, about a year after the treatment, to see the digital clock in the family car. Cideciyan attributes that not to continued improvement of the photoreceptors themselves but to the retraining of the visual system to take advantage of the treated areas of the retina.

Despite having a checkered past, gene therapy seemed like a promising option for many patients with this disease. There were more volunteers than could be accommodated in this trial, Cideciyan noted.

But the treatment isn't out of the woods yet. The research team will continue the study for a full three years, as is required by the U.S. Food and Drug Administration (FDA) for a phase I clinical trial. And the participants will be monitored for 15 years for health impacts, as is required by the FDA for any gene-therapy trials. Following the end of the first trial, there will be two more—to test the treatment's efficacy—before it can be approved for commercial use on the market.

Cideciyan admits that the encouraging results are only for a rare condition, but he hopes that with continued research, it will "have huge potential for much more common diseases."