Goggles and a gene injection give some retinitis pigmentosa patients limited object detection

Ten people were treated, and eight completed the object tests ^[13]. Four of those eight did consistently better at detecting the notebook with the goggles, across sessions spanning roughly a year to a year and a half ^[14]. That is 50% of the people tested ^[21] and 40% of the people treated ^[22]. The other four tested showed no consistent improvement ^[14].

The comparison is within each person. Everyone received the injection ^[3], and the object tests ran with and without the goggles while the untreated eye was covered ^[13][8]. The trial primarily assessed safety, with light sensitivity and object detection examined alongside ^[9]. Sensitivity was the dimmest light a participant could detect through the goggles, before and after treatment. Seven improved, from about twofold to 62-fold ^[10]. Six reached roughly a fourfold gain or more, which the team considered clinically meaningful ^[11]. Six of ten is 60% ^[23]. That threshold came from earlier research and was not in the trial protocol ^[12].

Sahel, of the University of Pittsburgh, and Roska, of the Institute of Molecular and Clinical Ophthalmology Basel, coauthored the NEJM paper ^[7]. They told ScienceAlert: "The gene therapy delivers a light sensor to the blind retina, making it light-sensitive again. The goggles provide light stimulation at the wavelength to which this sensor responds" ^[4]. A camera in the goggles captures changes in brightness, and a processor turns them into patterns of amber light that a projector directs onto the retina ^[5].

Why some people responded and others did not is open. "Differences between patients in the increase in sensitivity may reflect how well the gene therapy vector reaches the retinal cells, or how many retinal cells remain, or other variables such as training to wear the goggles," Sahel and Roska said ^[25]. More training sessions tracked with better performance, though the plausible range reached down to no link at all ^[26]. Recordings from the brains of five chosen participants were consistent with visual signals getting through ^[27]. Of those five, four had already responded in behavioral tests, and for some estimates the confidence interval included zero ^[27]. In our view, that makes the recordings a corroboration of known responders and little more.

The trial did not test function outside the clinic. The object tasks were detecting, locating and touching a notebook or a small box of staples, and telling the orientation of a bar ^[13]. Sensitivity is the dimmest light detected ^[10].

Variants in more than 100 genes can cause the disease, and that complicates efforts to develop a treatment for each cause ^[1]. With optogenetic therapy, the nerve cells that outlast the disease gain a fresh way to react to light ^[2], and here the virus delivers ChrimsonR to retinal ganglion cells ^[3]. In our view, it does not depend on which gene is mutated.

On safety, Sahel and Roska said: "Safety was very good and is not a limitation" ^[19]. They also noted that the results rested on 10 participants ^[20]. Across nine participants, the team logged 34 eye-related adverse events, most of them mild or moderate ^[16]. That is nine of ten people, or 90% ^[24]. Of 70 systemic events, six severe, the principal investigator, with oversight from a data and safety monitoring board, judged none related to the gene therapy or injection ^[18].