Cell-based subretinal implant may partially restore vision in geographic atrophy
Key takeaways:
- Retinal implants have a history of challenges, and questions remain about their utility.
- A bioengineered RPE stem cell-derived retinal implant has shown ability to partially restore vision in geographic atrophy.
SEVILLE, Spain — A cell-based subretinal implant shows promise in the treatment of geographic atrophy, achieving the unprecedented milestone of partially restoring vision.
During his keynote lecture at the Congress on Controversies in Ophthalmology, Mark S. Humayun, MD, PhD, discussed cell-based approaches for GA and early studies examining autologous same-eye retinal pigment epithelium (RPE) transplantation, which led to vision improvement.
“This was important because everyone looked into these eyes with geographic atrophy and said, ‘The retina is already dead. There’s no way to improve vision,’” he said.
However, the high rate of complications, up to a 45% rate of retinal detachment with this approach, led the researchers “to try to find the RPE from somewhere else.”
He then spoke about the controversy surrounding the source of stem cells.
Pluripotent stem cells can give rise to most cell types in the body. Either embryonic stem cells are used, which are allogenic and may require immunosuppression, or induced pluripotent stem cells that are autologous but have a high cost because they are custom developed for each patient. Because they need a lot of genetic manipulation, there is also the risk for malignant transformation, Humayun said.
The other option is adult stem cells, which are limited to giving rise to only a few cell types.
He also compared the pros and cons of RPE cells injected with a suspension vs. substrate-based delivery of a polarized RPE monolayer. Suspensions are easier to manufacture and administer. However, substrates are necessary if one wants polarized (ie, oriented correctly) RPE cells and are easier to visualize and remove if needed. Polarization, he said, is a key factor for the production of pigment epithelium-derived factor, a growth factor that is important for protection of the retina and survivability of the transplanted RPE cells.
In case of a scaffold, the next question is whether it should be erodible or non-erodible. Erodible scaffolds have the advantage of leaving just the cells in place as they dissolve, but the dissolution process may lead to inflammation. Non-erodible scaffolds do not have to deal with the unpredictability of durability and overcome the problem of inflammation associated with dissolution.
Based on these observations, Humayun and colleagues developed a bioengineered RPE stem cell-derived retinal implant on a substrate that mimics the natural Bruch’s membrane.
“After 7 years of preclinical development, we presented our data to the FDA, and FDA allowed us to start phase 1/2a trials on geographic atrophy,” he said.
All patients received a low dose of tacrolimus for 60 days in the perioperative period. At the time of surgery, following vitrectomy, the macula was hydrodissected and the device was implanted in the subretinal space.
“We had to spend a lot of time engineering the scaffold, which included showing that folding and unfolding the scaffold did not dislodge the cells,” Humayun said.
This allowed implantation through a small retinotomy of only 1.5 mm of a large subretinal scaffold with the stem cell-derived RPE to cover about 87% of the area of GA. This could be seen clearly because the implant was engineered to have a different refractive index on OCT.
“The fixation of these patients after the patch surgery returned to the center, and that was very exciting,” Humayun said.
Vision improved in 27% of the subjects, with a mean gain of 8.2 letters.
“In geographic atrophy, the drugs we are using only slow down lesion growth. Achieving some degree of visual improvement with the implant was very exciting,” he said. “We just published that the same subjects that had improvement in vision at 1 year also maintained it 3 years later.”
Side effects were minimal. Some retinal bleeding was observed around the retinotomy in three subjects out of the first cohort of seven patients, but after adding more diathermy to the retinotomy, no case was reported in the second cohort of nine patients.
Why did the vision improve in these subjects? There are several hypotheses that Humayun suggested in his lecture.
“It could be that there were dormant photoreceptors where there were just nuclei, and then outer segments grew when we put healthy RPE. It could be that there were partially damaged outer segments that improved in the presence of healthy RP photoreceptors. Thus far, our results do not support eccentric fixation. Lastly, there might be some other mechanism that we do not understand,” he said. “Whatever the explanation, visual improvement in geographic atrophy is indeed very rare, and it’s very encouraging to see it after this implant.”
Humayun said that the implanted cells were highly mismatched and yet were well accepted without sign of rejection.
“More than 50% of the [human leukocyte antigen] alleles were mismatched between the donor and the host. You would never put a kidney or a liver into anybody with this kind of high mismatch. We provide only immunosuppression for 60 days in the perioperative period,” he said. “The reasons for graft survival are also controversial. Is the subretinal space relatively immune privileged, even in GA and after subretinal surgery? Does the biocompatible scaffold lead to minimal inflammation? Does the monolayer of polarized RPE lead to minimal inflammation and therefore integrate better with the host?”
A multicenter phase 2b study will begin later this year.
Collapse
Humayun M. Keynote lecture: Retinal implants to restore vision. Presented at: Congress on Controversies in Ophthalmology; April 4-5, 2025; Seville, Spain.
Read more about