Gene Therapy Restores Sight

By Brandon Keim

Wired Science

September 22, 2008

With the help of gene therapy, two people who once were blind now can see.

The individuals -- their identities remain confidential -- are participants in an early-stage clinical trial of gene therapy for Leber's Congenital Amaurosis, a rare and untreatable form of congenital blindness.

Though the trial was designed to test the therapy's safety rather than its efficacy, its benefits were so impressive that the researchers decided to publicize their results.

"One of the patients said that the dim red light from his alarm clock had gotten so bright that it bothered him," said Artur Cideciyan, a University of Pennsylvania opthamologist and co-author of the study. "He had to turn away from it while he was sleeping."

Cideciyan's study, published today in the Proceedings of the National Academy of Sciences, is one of three simultaneous trials of gene therapy for Leber's Congenital Amaurosis, also known as LCA. Results from the first two were published in April in the New England Journal of Medicine: No safety problems were observed, and the results hinted at improvements of the sort described today.

Though the teams studied only one type of LCA, and LCA is only one type of blindness, researchers say the findings are broadly encouraging. They also say that, though the latest study involved just three people, of whom one experienced no significant improvement in vision, the results are powerful.

"The fact that you can see any kind of benefit is fabulous. It's a disease for which our therapeutic alternatives have been zero," said Oregon Health and Science University opthamologist Tim Stout, who was not involved in the study. "Does this mean that every gene therapy for every gene in every inherited eye is going to be a slam dunk? No, but I'm encouraged. It gives us some idea that the therapeutic manipulation of gene expression may be beneficial to eye diseases in general."

The patients possessed defective versions of the RPE65 gene, which normally produces an enzyme that maintains a protective layer of cells underlying light- and color-detecting photoreceptor cells. Cideciyan's team used a virus to introduce healthy versions of the gene into their patients' eyes, stimulating enzyme production allowing the remaining photoreceptors to function normally.

Improvement began in just over a week, and remained after 90 days -- the study's endpoint. Tests showed that two patients needed 63,000 times less light to see at a level comparable to their pre-therapeutic condition. Their eyes did take hours to adjust to low-light conditions, but the improvement was still radical.

Because the patients' eyes made use of just a few surviving photoreceptor cells, even greater improvements might be experienced if the therapy is applied earlier -- during childhood, or even infancy.

Before that can happen, however, the technique needs to be proven both effective and safe. The field of gene therapy has been troubled by a reputation for uncertainty and hazard -- a reputation that, given the difficulties of developing any experimental therapy, is not entirely deserved, but nonetheless persists.

According to National Eye Institute geneticist Brian Brooks, the virus used to deliver the genes is well-understood, having caused minimal reactions in this and other studies, and eyes appear to undergo little of the inflammatory response that gene therapy can trigger elsewhere. Stout added that animal tests suggest the changes induced by the virus may last for a lifetime.

Other blindness-causing diseases, such as age-related macular degeneration and retinitis pigmentosa, are more complicated and have different causes than the form of LCA studied by Cideciyan's team, but may respond to other gene therapies.

The results "are encouraging for any retinal degeneration where there is some structurally intact retina where you can intervene prior to significant degeneration," said Brooks.

"The virus was designed to bring in its package, and get the cells to produce a protein -- in this case, RPE65," said Cideciyan. "But in theory, it could be anything."

blog.wired.com/wiredscience