Scientists 'Resuscitate' Post-Mortem Eyes, Reviving Light Responses Up to 10 Hours After Death

A new device developed by researchers in Spain, Israel and the UK has restored blood flow and measurable light responses in whole eyes removed from the body, a development its authors say brings whole eye transplantation and new pre-clinical testing platforms a step closer to reality.

Published in bioRxiv, the study describes ECaBox (Eyes in a Care Box), a custom-built perfusion and monitoring system that cannulates the ophthalmic artery of an excised eye and pumps an oxygenated Ringer's solution through its vasculature. The team, led by researchers at the Centre for Genomic Regulation in Barcelona, Bar-Ilan University in Israel and King's College London, report that the approach preserved retinal structure and cell viability in pig eyes for up to 24 hours post-mortem, and restored electroretinogram (ERG) light responses for up to 10-12 hours after death.

Why it matters for the sector

Vision loss affects an estimated 2.2 billion people globally, according to World Health Organization figures cited in the paper, and durable treatments for degenerative retinal disease remain limited. While whole organ transplantation is now routine for hearts, lungs, kidneys and other organs, whole eye transplantation (WET) has long been considered unachievable because of the retina and optic nerve's extreme sensitivity to even brief interruptions in blood supply.

The researchers argue that a system capable of keeping an intact, whole eye functionally alive outside the body, rather than an isolated retina in a dish, is a meaningful step toward two practical goals for the field: extending the viability window for donor eyes destined for transplantation, and providing a more physiologically realistic platform for testing new retinal therapies before they reach clinical trials.

What the researchers did

The team first developed a cannulation and perfusion technique in pig eyes, chosen for their anatomical similarity to human eyes, including a comparable retinal vasculature and a "visual streak" analogous to the human macula. A 26-gauge cannula was inserted into the ophthalmic artery shortly after enucleation, secured with a custom 3D-printed eye holder, and connected to a closed-loop perfusion circuit that regulated flow, pressure and oxygenation in real time.

Using retinal fundus imaging, fluorescein angiography, optical coherence tomography and vascular casting analysed by Micro-CT, the group confirmed that perfusion restored flow throughout the retinal and choroidal vasculature, down to the capillary level, in roughly 90 per cent of well-cannulated eyes. A deep-learning model (Attention U-Net) was used to reconstruct and quantify the vascular networks, allowing the team to show that vessel density declined the longer an eye went unperfused before treatment.

Histological grading, cross-checked against an AI classifier trained to score degeneration severity, showed retinal structure held up well for up to 10 hours post-mortem without intervention, but deteriorated significantly by 24 hours. Cold storage at 4°C, the standard approach for preserving donor organs, was not sufficient to protect retinal cell viability over the same period. Perfusion with oxygenated solution, by contrast, significantly rescued cell viability, cell size and structural integrity at the 24-hour mark.

The electrophysiology finding

Perhaps the most striking result for clinicians is the functional data. Using corneal ERG electrodes, the same non-invasive approach used in human patients, the team recorded genuine light-evoked electrical responses in 15 of 36 perfused pig eyes (41.6 per cent), with signals persisting for up to 10 hours, and in one case 12 hours, after death. In three eyes, switching the perfusion pump off caused the ERG signal to fade, and restarting perfusion brought it back, direct evidence that the response was dependent on restored circulation rather than residual retinal activity.

The authors note this challenges the assumption that a retina's capacity to respond to light necessarily ends with death, and draw parallels with clinical phenomena such as amaurosis fugax, where vision returns once blood flow to the retina is restored.

The group also successfully applied the cannulation and perfusion technique to six pairs of human donor eyes, obtained under ethics approval from Vall d'Hebron Hospital in Barcelona from donors ineligible for corneal donation, confirming vascular perfusion by imaging and improved retinal cell viability compared with non-perfused contralateral eyes from the same six donors.

Caveats

This is an early-stage, peer-review-pending preprint, and the authors are candid about its limitations. Not every eye tested produced a positive ERG signal, and the reasons for that variability, ischaemia time, baseline intraocular pressure and baseline ERG amplitude were all ruled out as explanatory factors, remain unclear. Human data are limited by the small number of available donor eyes and the advanced age and comorbidities typical of donors whose corneas are unsuitable for transplant. No human eye in the study underwent ERG testing. The device has not been tested in a transplantation setting, and functional light response is a long way from restored vision, a point underscored by the world's only whole-eye transplant recipient to date, who regained retinal electrical activity following face-and-eye allograft surgery but did not regain sight.

What's next

The authors say they are developing a portable, surgery-room version of ECaBox designed to minimise ischaemia time in eyes from heart-beating donors, and see the detailed vascular reconstruction generated in this study as a foundation for "digital twin" modelling of ocular blood flow, potentially useful for planning perfusion strategies and testing pharmacokinetics in silico before physical trials.

For eyecare professionals, the work remains preclinical, but it adds to a small but growing body of evidence, including last year's first-in-human combined face-and-eye transplant, that the field's long-standing assumption of irreversible retinal ischaemic damage may not be as absolute as once thought.