Retinal Ganglion Cell Repopulation for Vision Restoration in Optic Neuropathy: A Roadmap From the RReSTORe Consortium

Major Points and Findings:

In optic atrophy the cells that have died are retinal ganglion cells (RGCs), the nerve cells whose long fibres make up the optic nerve. Restoring lost sight would mean replacing those cells and having the new ones grow a fibre all the way to the brain. This 46-page paper is the consensus statement of the RReSTORe Consortium, more than 200 researchers worldwide who work on that problem. The paper reports no trial and no patients. It is a review and research roadmap based almost entirely on laboratory and animal work.

The paper gives the assessment of the researchers most invested in optic nerve regeneration. They believe it “may be feasible”. They also state that the field has “yet to develop a definitive cure” for vision loss from optic neuropathy “or even a neuroprotective treatment that substantially slows disease progression”, that fewer than 1% of transplanted ganglion cells currently survive in animals, and that no one has yet shown transplanted cells reconnecting an eye to the brain. The roadmap deals with ganglion cells made from pluripotent stem cells or converted from the retina’s own support cells. It does not cover mesenchymal stem cells.

Aim:

To identify the most pressing unanswered questions standing between today’s laboratory results and a treatment that replaces retinal ganglion cells in glaucoma and other optic neuropathies, and to propose experiments to answer them.

Methods:

From January to April 2022 consortium members took part in a virtual consensus-building process, then met for a day-long workshop in Denver, Colorado, on 29 April 2022 in five parallel discussion groups: RGC development and differentiation; transplantation methods and models; RGC survival, maturation and host interactions; inner retinal wiring; and eye-to-brain connectivity. The paper summarises each, ending with a table of research goals. The paper represents expert consensus. It has no search protocol or grading of evidence and is therefore not a systematic review.

Results:

1. Making the cells: RGC-like cells can now be produced at scale from embryonic or induced pluripotent stem cells, which the authors count as a key achievement of the past decade. Several problems remain open. Lab-made cells are immature, RGCs inside retinal organoids die over time, and more than 86% of human RGCs are of the “midget” type needed for sharp vision, for which no production protocol exists. Nobody knows which of the roughly 18 human RGC subtypes must be replaced for useful sight. A second route, converting the retina’s own Müller glia into neurons, has produced RGC-like cells in injured adult mice, but no fibre growth into the optic nerve has been shown.

2. Getting them into the eye: The consortium reports the following figures and obstacles.

• Survival of transplanted primary RGCs is around 1%. For human cells placed in primate eyes it is typically below 1%.
• Injecting more cells does not improve survival. In one rat study more donor cells survived after injecting 40,000 than after 60,000. Suggested reasons are clumping, shear stress in the needle, competition for nutrients and a stronger immune reaction.
• Most studies follow animals for up to 3 months.
• The inner limiting membrane, the thin sheet lining the retina, physically blocks injected cells. Enzymes that dissolve it help in rodents but can inflame the retina, and surgical peeling is the more translatable option.
• Injecting loose cells into the vitreous has worked only in rodents, whose eyes have a large lens and a narrow vitreous cavity. In larger eyes the cells are likely to be trapped far from the retina and to clump, so a scaffold placed on the retinal surface after vitrectomy will probably be needed.
• Typical laboratory rodents are 2 to 6 months old, equivalent to a person of 20 to 30, while most patients are elderly with advanced disease, scarring and chronic inflammation.
• “It is not possible to put a needle into the eye without causing some inflammation”, so an anti-inflammatory regimen will probably be needed. Donor cells face rejection unless they are the patient’s own, and must have limited ability to multiply, to avoid tumours.

3. Keeping them alive: The vitreous is low in oxygen and RGCs depend heavily on mitochondrial energy. Ideas under study include nicotinamide (vitamin B3) and pyruvate, which have given short-term improvements in visual function in glaucoma patients in their own right. Others are hypoxic preconditioning, blocking cell-death pathways (a caspase-2 siRNA is already in clinical trials for optic neuropathies) and masking the “eat-me” signal that makes microglia engulf donor cells. One side observation was that a week after stem cell-derived RGCs were transplanted, more of the host’s own RGCs survived an optic nerve crush, possibly through transferred extracellular vesicles.

4. Wiring into the retina: New RGCs must connect with bipolar and amacrine cells. These partner cells are relatively spared in optic neuropathy, so the circuitry to plug into should still be there. In practice few donor cells reach the ganglion cell layer and fewer still form circuits.

5. Reaching the brain: Over the past 15 years, manipulating pathways such as PTEN/mTOR and SOCS3 has made injured RGCs in rodents regrow fibres over long distances. The authors hope the same can be done in transplanted cells. Several obstacles remain. The lamina cribrosa, a collagen sieve at the nerve head, is absent in rodents and its effect on regrowth is essentially unstudied. The others are steering at the optic chiasm, rebuilding an orderly map in the brain, regrown fibres that lack myelin, and the brain’s visual relay centres, which shrink and lose cells within 90 days of losing their input.

Conclusions:

The consortium writes that its discussions “yielded more questions than answers”. It calls for work in large animals (pigs, cats, tree shrews, primates), in aged and late-stage disease models, and for imaging that can track single cells in living eyes.

The authors expect that, to establish safety, the first patients in any trial will be those with advanced disease and severe to complete vision loss. They expect the best chance of benefit in people with recent, acute loss, and suggest optic pathway glioma in neurofibromatosis type 1, or acute ischaemic optic neuropathy, as early candidates. They propose modest first targets. Restoring even basic light perception through one RGC type (the intrinsically photosensitive cells) could reset the disturbed sleep-wake cycle common in people who are totally blind.

Background Information:

For people with optic atrophy today, no therapy available anywhere replaces lost retinal ganglion cells or regrows a human optic nerve. The roadmap, written in 2023, speaks of first-in-human RGC transplantation only as something still to come, and it does not mention mesenchymal stem cells. The realistic aim of any current cell-based treatment, including MSCs, is support of surviving, poorly functioning cells through secreted factors. That effect has not been confirmed in controlled trials for optic neuropathy.

Timing recurs throughout the paper. In the mouse crush model almost 90% of RGCs die within three weeks, and the brain targets begin to waste soon after. The consortium’s view is that any regenerative approach is most likely to work early, and that long-standing atrophy, the situation most patients seeking treatment are in, will be the hardest case.