Clinical and Scientific Considerations for Whole Eye Transplantation: An Ophthalmologist's Perspective

Over 40 million patients worldwide are legally blind from either retinal or optic nerve disease.¹ Retinal neurons, from photoreceptors to retinal ganglion cells (RGCs), have an intrinsically poor capacity for self-repair after damage and no capacity to self-regenerate. Thus, vision restoration in patients with late-stage macular degeneration or glaucoma will likely require strategies to replace damaged cells. Currently, two major approaches are being pursued: (1) cell type-specific replacement and (2) whole eye transplantation (WET). Both strategies are associated with unique advantages and challenges. For the TVST readership, in this opinion piece, we summarize many of the salient barriers clinician-scientists must overcome to enable approaches like WET to transition from a cosmetic procedure to a surgery that provides functional recovery. 

Targeted Cell Type-Specific Transplantation

Simply put, targeted, cell type-specific transplantation aims to replace non-functioning degenerated cells with healthy cells. The success of targeted cellular replacement strategies depends on the ability to generate a large volume of healthy, immune-compatible, high-grade cells that would need to be integrated into the neurosensory retina. Thankfully, advances in stem cell biology have enabled the production of patient-derived RGC-like and photoreceptor-like cells.^2,3 Athough these cells display morphologic and electrophysiologic characteristics similar to native cells, critical functional differences still limit clinical translation. Additionally, standardizing good laboratory practices that allow for affordable and reproducible production of cells for transplantation is needed. 

Even if the barriers mentioned above are addressed, cellular integration is the major challenge limiting the translation of cell-type-specific transplantation. In the case of photoreceptor disease, scientists have successfully transplanted human embryonic stem cell-derived retinal pigment epithelium (RPE) cells by loading them onto ultrathin, bioinert parylene sheets, for transplanting cells subretinally.⁴ The parylene sheet allows RPE cells to maintain their apical-basal orientation and for cells to be inserted into the correct tissue layer. One year after implantation, many patients reported >5 letter gain in best-corrected visual recovery in the treated eye and >5 letter loss in the untreated eye.⁴ Such efforts, however, are even more complicated if photoreceptors or RGCs need replacing.³ Restoring vision with photoreceptors, for example, requires cells to not only physically reside in the correct retinal layer, with the correct apical to basal orientation, but also for their dendritic processes and axons to synapse with native cells in the retina. RGCs have the added barriers of needing approaches that can direct axon regeneration out of the eye, into the optic nerve, to the diencephalon, and drive synaptic formation with intracranial neurons. Complicating matters further is that there is no guarantee that intravitreal delivery of a suspension of RGCs would result in the correct RGC subtype to integrate into the areas where they are “missing.” Few, if any, approaches have been described that can deliver this. 

Whole Eye Transplantation—An Alternative to Cell Type-Specific Transplantation

WET involves replacing the entire globe as an alternative to targeted cell-type specific replacement. Landmark work by surgeons at New York University (NYU), led by Dr. Eduardo D. Rodriguez, allowed for the successful transplantation of a human eye and face in a patient with severe facial injuries from electrocution.² Although this patient has no light perception vision, the allografted eye has survived 1 year without signs of phthisis, sympathetic ophthalmia, or infection. Successes like these are paving the way for the possibility of whole-eye transplantation to one day restore vision to patients who are debilitated by retinal and optic nerve disease. 

Optic Nerve Regeneration

A major advantage offered by WET over single-cell replacement approaches is that the neural network within the retina need not be re-established as it would be preserved in the donor eye. However, WET suffers from the major challenge that donor RGC axons need to “jump” past the optic nerve coaptation site. In contrast, transplanted RGC axons would at least have the intact, albeit non-functional, optic nerve to act as a scaffold. Whereas advances in gene therapy, neurotrophic factor administration, and electric field stimulation have shown promise in preclinical optic nerve crush injury models, researchers have yet to demonstrate efficacy in optic nerve transection models—a model that more closely mimics WET. Moreover, directing long-distance axon growth past the optic chiasm is still challenging and has only been shown by a handful of approaches.^5–8 

Another consideration is that even if RGC axons could jump across the severed nerve ending, it would be impossible to guarantee maintenance of the retinal-cortical map. For example, if the left eye were shifted clockwise during nerve coaptation, RGCs in the superior-nasal quadrant of donor retinas would end up synapsing with superior-temporal neurons in the host's geniculate nucleus. This limitation also plagues RGC-specific transplantation approaches; its effect on vision restoration is unknown. 

Donor Identification and Organ Procurement

At the outset, successful WET will depend on our ability to identify good donor candidates, who are likely to be limited in supply, similar to other organs (e.g. liver and kidneys). Success will depend on our efficiency at isolating and maintaining allograft viability during transport. Toward this goal, researchers have been developing bedside tools to assess ocular function in potential organ donors. Assessing ocular function is complicated by the fact that visual evoked potentials (VEPs) cannot be measured in patients who are brain-dead and medical records are unlikely to be complete. Handheld electroretinograms (ERGs) can be helpful but suffer from poor signal-to-noise ratio and tell us little about RGC function. Current handheld ERG devices can perform ERG, electrooculography (EOG), and VEP recordings, but tests of RGC-specific function (e.g. pattern ERG or photopic negative response) are currently unavailable. 

Although the optimal surgical approach for organ procurement is still under investigation, successful allografting will require immediate cannulation of the ophthalmic artery and vein to maintain perfusion and prevent thrombosis. Researchers continually optimize ECMO-like devices where globes can be placed and preserved during transport.⁹ Thankfully, not everything has to be created from scratch for WET. Standing on the shoulders of previous transplant science, many approaches propose to modify existing media, which has a complex cocktail of preservatives, antimicrobials, and immunosuppressants.¹⁰ Other proposed modifications include intravitreal delivery of nutrients to the neurosensory retina. Given that it takes several days to organize organ donation, there may be an opportunity to initiate such neuroprotective approaches before organ harvesting. 

In addition to technical challenges, vision scientists must fight dogma and push for paradigm shifts regarding their position in the organ harvesting priority line. Currently, cardiothoracic surgeons are first in line when it comes to donor harvesting. The central nervous system, however, unlike other commonly harvested organs (e.g. liver and kidneys), is susceptible to hypoxic and perfusion injury. Although there is no consensus regarding how long the retina can survive without perfusion, it is safe to say that the critical period is on the order of several minutes to an hour rather than several hours.^9,11–14 Given this, eye transplant surgeons must occupy a simultaneous priority status with cardiothoracic surgeons and be allowed to procure globes before cardiovascular structures are cross-clamped. Thankfully, pioneers in the field began breaking down these barriers years ago. In addition, NYU surgeons who pioneered the first combined face and eye transplant have studied surgical techniques to limit retinal ischemia time during transplantation. Nevertheless, we must continue to support our colleagues and lobby for eye harvesting to sit at the front of the organ transplant priority list. 

Once transported, the allograft must undergo quality control assessment before surgery in the recipient's orbit is commenced. Few reliable technologies exist that allow for functional assessment of an enucleated eye. Moreover, thresholds for what constitutes a functional organ with good potential for engraftment need to be established and validated. 

Surgical Considerations

Another significant challenge facing WET is balancing the need to perfuse the anterior chamber while preserving ocular movement. The orbit is perfused by the ophthalmic artery, indicating that globe viability can be maintained if stable anastomosis of the donor ophthalmic artery is made with a host vessel of the appropriate caliber and perfusion pressure (e.g. host ophthalmic artery or superficial temporal artery). What complicates the situation is that the anterior segment of the eye is solely perfused by vascular branches of the extraocular recti muscles. However, strabismus surgeons would warn that removal of more than two rectus muscles at a time puts the eye at risk for anterior segment ischemia. What about if all six are removed? Moreover, the time it takes for the arteries of the recti muscles to reperfuse the anterior segment after reattachment is on the order of weeks. In contrast, anterior segment ischemia sets in within days. Although this limitation could conceivably be overcome with transplantation of the entire muscle cone, this would prevent the host's cranial nerves from controlling eye movement, leading to total ophthalmoplegia. Indeed, the patient who underwent WET at NYU has complete ptosis and ophthalmoplegia 1 year after engraftment but no signs of phthisis.¹⁵ Successful WET will require the development of surgical approaches that both prevent anterior segment ischemia and preserve extraocular movements. 

Of course, the challenge of preserving eye movement is minor compared with that of anastomosing the optic nerve endings themselves. At a minimum, physical alignment of the donor and host nerve endings is needed to give donor RGCs axons a chance to grow into the host optic nerve. Optic nerve coaptation with suturing and other techniques can be translated from well-established peripheral nerve repair techniques.¹⁶ For example, nerve wraps and scaffolds can augment nerve coaptation.¹⁷ Nerve wraps can help maintain tissue alignment, provide drugs to minimize inflammation, and provide neurotrophic support. Electrospun scaffolds can benefit from being loaded with cells that could function as interneurons, connecting donor axonal endings to recipient axonal stumps,¹⁸ or proteins that lure donor RGC axons into the scaffold.¹⁹ Scaffolds, however, could potentially suffer from the “candy store” effect in which axons are lured into the scaffold but have no impetus to exit. Finally, innovations in robotic assistive approaches may be critical for increasing the success of these delicate surgical procedures. 

Immunosuppression

Significant insight into the role that the immune system plays in blocking and facilitating RGC axon growth has been gained over the past decade. Specifically, microglial activation is associated with RGC degeneration, whereas inhibition of microglial activity with micro-RNA results in robust axon regeneration after optic nerve crush injury.^20–22 Glial scars elaborated by astrocytes are also likely to be a significant barrier to RGC axon regeneration after WET, especially given the sizeable incisional injury. However, recruitment of neutrophils to the site of crush injury is thought to be pro-regenerative.^23,24 As the role of different immune system arms in inhibiting and enabling RGC axon growth continues to be elucidated, approaches that can mediate lineage type-specific modulation will be needed to enable WET to restore vision. 

Targeted immunosuppression will also be needed to prevent organ rejection after WET. The eye is traditionally thought to be “immune privileged”; however, even corneal transplants can undergo rejection, and it is unknown how rejection will manifest in the different components of the transplanted eye.^25,26 Crucial aspects, such as early indicators for immune rejection, forms of rejection responses, clinical signs and laboratory findings, and modalities for monitoring and diagnosis are not fully understood. A staging system for eye transplant rejection will need to be established, similar to classification criteria (e.g. BANFF) for solid organ and other vascularized composite allotransplants.²⁷ Unlike with solid organ transplants (e.g. liver or kidneys), the eye presumably will not be able to tolerate trauma from multiple biopsies to monitor for signs of rejection. The superficial location of the eye is favorable for noninvasive modalities like optical coherence tomography to monitor rejection in WET. 

The superficial location of the eye also makes WET amenable to local therapies that can reduce the amount of systemic immunosuppression needed to prevent rejection and, thus, lower the risk profile of transplantation. The ultimate goal is success while “doing no harm,” and the need for systemic immunosuppression can lead to adverse and sometimes life-threatening side effects. Given that eye transplants may include surrounding tissue such as skin, which is highly immunogenic, immunomodulatory therapies that manipulate the inherent properties of the recipient's immune system to “tolerate” the transplant with minimal or no immunosuppression are the ultimate goal.²⁸ Immune suppression will also likely be needed with cell-type specific transplantation approaches. 

Recipient Selection

Consideration should also be given to what the host receiving environment looks like. This question hints at the challenging topic of understanding who is a good candidate for WET. Is the atrophic host optic nerve an organized tube with formed collagen tunnels that can welcome growing axons, or are there significant debris and residual proteins that will repel regenerating axons? Will the axons know where to grow and which way to turn at the chiasm? The complex symphony of molecular and electrical signals that coordinated RGC axon growth during development are likely to be jumbled in the adult and, thus, unhelpful. Without directional cues, will regenerating axons get tangled? In addition, if regenerated RGC axons reach the host diencephalon, what will target structures look like? High-resolution magnetic resonance imaging (MRI) in patients with severe glaucoma demonstrates atrophic lateral geniculate nuclei,²⁹ suggesting that our visual system undergoes anterograde degeneration. This could imply that although patients with chronic, severe vision loss have the lowest risk profile for WET, they may also be the least likely to benefit. Unfortunately, these barriers would also plague RGC-specific transplantation approaches. 

Neural Plasticity and Brain Adaptation

Plasticity of the adult visual system is poorly understood but decreases with age. Preclinical studies suggest that with certain interventions, the critical window of the visual system can be extended or even reactivated. Will target neurons be receptive to forming new connections with regenerated axons? Is driving donor axon regeneration to the dendrites of target neurons in the brain sufficient to promote new synapse formation or are other signals needed? At a minimum, patients who undergo WET will require extensive vision therapy to help activate donor RGCs and hopefully drive synapse formation. More research will be needed to understand how much plasticity exists in the lateral geniculate nucleus.³⁰ Again, this is a barrier that would also plague RGC-specific transplantation approaches. 

Myelination

Even if we can overcome the abovementioned limitations, functional vision is unlikely to be restored if axons are not remyelinated. Rodent studies suggest that regenerated RGC axons are unlikely to spontaneously remyelinate.³¹ Insight into how to drive remyelination of central nervous system (CNS) axons can be found in the neuroimmunology literature, where demyelination is the pathophysiology that underlies diseases such as multiple sclerosis and neuromyelitis optica spectrum disease. Unfortunately, most of the effective therapies available to patients with these disorders are directed at immune modulation, as phase II/III clinical trials on remyelinating drugs (e.g. opicinumab) have failed to reach functional end points.³² Not only should myelination be promoted, but the timing of myelination is also likely to affect the success of RGC axon regeneration. Myelin debris released after trauma has been shown to inhibit RGC axon growth, and failure to remove myelin debris is thought to be one of the primary reasons underlying poor axon regeneration in the CNS.³³ Thus, not only promoting myelination but also coordinating when it occurs will be a critical component to the success of WET and RGC-specific transplantation approaches. 

Ethical and Psychosocial Considerations

The dictum that drives clinical decision making is that the benefit of a procedure should outweigh the risk to the patient. But how much vision can we expect to restore? WET raises ethical concerns, including informed consent, risks versus benefits, and equity in access to the procedure, including access to healthy eyes for donation. The lowest-hanging fruit is the restoration of circadian rhythm. Patients who are bilaterally blind not only suffer from a lack of image formation but also from failure to synchronize their circadian rhythm with light/dark cycles. They fail to know when to wake up and when to sleep. This debilitating symptom stems from a lack of input into the suprachiasmatic nucleus (SCN) that, from its name, sits above the optic chiasm. Although most animal studies fail to demonstrate RGC axon regeneration past the optic chiasm, almost all demonstrate axons that project to the SCN. This suggests that restoring circadian rhythm could be a low-hanging fruit for WET. Placing outcome measures such as improvement on a Snellen Chart or mean and pattern deviation on visual field testing as a bar for success after WET is likely too optimistic and may lead to the patient and the public's perception of failure. Which functional gains should be demonstrated in preclinical studies before initiating clinical trials? Ultimately, work to develop functional outcome measures beyond the Snellen Chart, such as the ability to navigate a sidewalk, needs to be established and validated by clinician scientists to help vision scientists discern which approaches for vision restoration after WET should be pursued. 

Although the surgical approach and immune suppression regimen for WET are still under investigation, these procedures are not risk-free. The most dreaded complications would be graft failure and sympathetic ophthalmia. Although graft failure could be readily remedied with an ocular prosthesis implant, damage to the other eye from sympathetic ophthalmia would be devastating. HLA matching can minimize but cannot guarantee against this possibility. For this reason, bilaterally blind patients are likely to be the first target population to receive WET. However, as mentioned earlier, these patients may be the most difficult to treat. 

Cost Considerations

Although this opinion piece focuses on sharing technical limitations associated with WET, the potential economic burden to society should also be considered. WET costs start with coordination between different surgical disciplines (e.g. ophthalmology, plastic surgery, and neurosurgery) and organ procurement services. Still, it will also need to involve postoperative care teams for immunosuppression and rehabilitation. On the other hand, the cost of blindness to a patient extends beyond the inability to participate in the workforce and mental wellness. Does society have the resources to perform WET for everyone who needs it? Who will have access to WET? 

Conclusions

Curing blindness via WET is an ambitious goal with significant barriers. An important limitation to the field's ability to make meaningful advancements despite incredible work being performed all around the world is, in part, because researchers tend to work in silos. Thankfully, government institutions, including the National Eye Institute (Audacious Goals initiative) and ARPA-H (The Human Eye Allotransplant initiative), provide funding opportunities to fast-track innovation and collaboration on these issues. Think tanks like RReSTORe bring disparate researchers to the same table for unified discussion. What is incredible about these efforts is that they are accelerating new, multi-disciplinary, convergent collaborations. Ultimately, optic regeneration will require a complex, combinatorial approach in which axon growth is supported, guidance cues are given, the immune system is modulated, axons are remyelinated, and brain connectivity is supported. Different clinical expertise, from ophthalmology to transplant surgery to neurosurgery to immunology, must be combined before realizing the potential of moonshot approaches like WET. 

Acknowledgments

KG and KW are supported by funds from ARPA-H (Agreement No. 1AY2AX000042-01).