The meeting with Wilson ended on an uncertain note. Bennett could see that Wilson was having serious doubts. If he didn't want to do it, it was unlikely that the executive committee would approve the project.

For the next several days, Wilson mulled over what to do. It was clear to him that if the only goal was to find an effective treatment for PDE, he would oppose the project because the payoff in new knowledge would be too limited.

But Bennett's PDE animal work proved that gene replacement can slow down RP. If it worked with PDE, Wilson asked, why wouldn't it work with choroideremia?

There was another argument for proceeding: Regardless of what form of RP was ultimately chosen for treatment, the surgical procedure and vector would be the same; only the gene carried by the vector would be different. With the exception of animal testing, all the time-consuming preparatory work would be the same.

And the need to know how to deliver these genes was growing. Six RP genes had already been identified, and probably many more would be found in coming months. It was important, Wilson thought, to know what was the best vector to use in eye diseases, what was the best dose, and how to deliver it.

It was not clear at all that the project should be abandoned.

Maguire and Jacobson were sitting at the conference table with members of the executive committee when Bennett rushed in a few minutes late. She immediately went to the head of the table and started her presentation.

Recapping much of the material presented at the earlier meeting, she said that, as they moved ahead with their preparations, they and other laboratories would be screening blood to find more patients with PDE.

Alan Davis, director of the institute's vector laboratory, said five vectors would be studied to see which was best suited for delivering genes to eyes.

Eric Wheeldon, who ran the institute's animal facility, said 50 mice, 160 rats and four monkeys would be used to test the different vectors before one was put into humans.

Jennifer Dennin, the institute's administrative coordinator, had priced out all the things that would go into the study and estimated that it would cost $330,000 to prepare for the human trial.

The executive committee - composed of Wilson and five other researchers from Penn, the Wistar Institute and Children's Seashore House - asked several questions and thanked the scientists for coming.

Bennett, Maguire and Jacobson spent the next two weeks anxiously awaiting word from the institute. Every afternoon, Maguire and Jacobson would call Bennett to see if she had heard anything.

Finally, Bennett received the letter from Wilson. The committee had agreed to support planning for human trials of a drug against retinitis pigmentosa.

The emphasis would be on finding the best vector to deliver genes to the retina and determining the maximum safe drug dose.

Though Bennett intended to push ahead with the PDE work, Wilson wasn't so sure that this would be the RP gene defect finally chosen.

Whatever it was, he wanted everything to be in place so they could go to human trials as soon as it was chosen. A lot of momentum had been built up, and Wilson had no intention of losing it.

Through the fall, Bennett and her colleagues moved quickly to prepare for the animal studies.

They arranged for the institute's vector lab to make viruses for pilot studies. They asked the institute's animal facility to get and house the animals.

Bennett started assembling the human PDE gene for insertion into the vectors that would be used in the human trials.