Alternatives to endokeratoplasty: an attempt towards reducing global demand of human donor corneas

The human eye is a complex anatomical structure with multiple layers that are responsible for maintaining clear vision. The cornea, the anterior most tissue of the eye, is made of several layers, including the posterior monolayer of endothelial cells [1]. The corneal endothelium facilitates stromal deturgescence by a ‘pump-and-leak’ mechanism required to maintain the transparency of the tissue. Damage or dysfunction of the endothelium may lead to corneal blindness. It is known that human corneal endothelial cells (HCECs) do not proliferate in vivo and are arrested in the G1 phase of the cell cycle after birth; hence, it becomes essential to maintain the endothelial cells throughout a lifetime [2]. HCECs can be damaged or can malfunction due to various pathological conditions that prompt the remaining cells to migrate. The migration allows spreading and covering of the damaged area that helps to avoid corneal swelling by maintaining the barrier function. Corneal swelling due to increased water retention in the stroma leading to partial or total corneal blindness is observed in events with low endothelial cell density, which can be treated by a corneal transplant [3,4].

The current approach of treating corneal endothelial dysfunction is corneal endothelial transplantation, or endokeratoplasty (EK), which may be restricted in many parts of the globe due to the limited availability of human donor corneas [4]. Therefore, an alternative treatment option for endothelial dysfunction is urgently required. Cell therapy using cultured HCECs has shown clinical success for the treatment of EK. Various media, growth conditions, isolation techniques and culture methods for the cultivation of HCECs have been investigated and reviewed [5–14]. Cell therapy is based on direct injection or transplantation of cells on a carrier. Scaffolds for culturing and transplanting these cells have also been thoroughly reviewed [1]. However, several challenges have been identified and continuous attempts to overcome the concerns have been made to increase the safety and efficacy of cell/tissue-engineered products [15]. As these techniques rely on the donor tissues for cell culture, novel methods such as descemetorhexis without EK (DWEK) or Descemet's stripping only (DSO), whereby the central diseased endothelium is excised, and the treatment relies on the peripheral cells migrating towards the center to heal the excised area has been evaluated. The advantage is that DWEK/DSO eliminates the requirement of the donor tissue. These methods have shown promising results for patients suffering from Fuchs endothelial corneal dystrophy (FECD), a leading cause of EK. Several clinical trials are ongoing to treat endothelial dysfunction. This review therefore presents the studies on alternative treatment options for EK, suggesting a path to reduce the ongoing demand of human corneal tissues.

ROCK inhibitor

One of the most common approaches of HCEC regeneration involves the use of ROCK inhibitor for the development of allogeneic ex vivo expanded cells for transplantation [16]. ROCK inhibitor is associated with cytoskeletal reshaping, contraction of smooth muscle and gene expression [17,18]. Considering the active role of ROCK inhibitor in physiological functions, investigation on the clinical applications of ROCK inhibitors has been carried out [19–21]. ROCK inhibitors such as fasudil and ripasudil are currently used in clinical practice. It has been reported that ROCK inhibitor Y-27632 allows the adhesion of HCECs to a substrate, and the inhibition of ROCK signaling may manipulate cell adhesion properties [22–24]. Okumura et al. reported that Y-27632 also inhibits the apoptosis of primate corneal endothelial cells (CECs) in culture [25]. FECD shows an abnormal endothelium, therefore needing replacement of the diseased tissue, which limits the use of direct injection of ROCK inhibitor. In such cases, ex vivo expansion of cells is more suitable for cell-based therapeutic treatments. It has already been demonstrated that ROCK inhibitor improves the engraftment of transplanted CECs and that injection of CECs in the form of a cell suspension can regenerate the corneal endothelium [22,26,27]. Although it has been noted that HCECs have a limited regenerative capacity in vivo, they have the potential to proliferate in vitro when supplemented with suitable growth factors [2]. Cultured HCECs in vitro could thus be considered as a prospective alternative treatment for corneal endothelial disease. Media supplements, specific growth conditions and techniques to isolate and cultivate HCECs have been studied [5–14].

Isolation & expansion of HCECs

Isolating single cells from the donor tissue is crucial for a successful cell culture. This can be achieved by enzymatic digestion followed by collection of cells after centrifugation. Peh et al. optimized the protocol by peeling the Descemet's membrane (DM) from the donor tissue and treating the DM with collagenase followed by trypsin digestion to form single cells [12]. The collected cells are plated on fibronectin collagen (FNC)-coated dishes to enhance adherence. The media (supplemented with ROCK inhibitor) is refreshed, and the cells are characterized at the end of the culture period. This method has shown to expand cells for potential treatment of multiple patients over the typical one donor to one recipient as observed in conventional transplant procedures.

Tissue engineering of corneal endothelium

Tissue engineering, as a multidisciplinary area, can successfully combine different fields in the race to encourage tissue regeneration. However, this goal is harder to achieve in HCECs due to their limited ability to proliferate in vitro. Tissue engineering can be helpful in producing scaffolds, cells and biomolecules to produce healthy tissue substitutes [1]. Despite HCECs' culture hardships, the attempts to produce or combine effective scaffold or fully engineered endothelial tissue is getting more attention.

Tissue engineering based on a desired scaffold

Scaffolds are 3D structures created to resemble the extracellular matrix (ECM) of the desired tissue in terms of its physical structure, chemical function and biological behavior. The scaffold must be biodegradable, biocompatible with the recipient cells and tissues, and should be non- toxic or mutagenic or express immunogenic reactions. Scaffolds are built from different materials that can be further classified as biological/natural, semi-synthetic or synthetic/artificial [28]. Biological/natural scaffolds are obtained from naturally available material and used after washing or decellularizing the tissue of interest without changing the architecture and function of the ECM. They can be of human or animal origin, have an ability to engage the progenitor's infiltration into the scaffold and regulate immune responses or the secretion of various molecules [29]. Semi-synthetic scaffolds are built from polysaccharides or proteins, which are part of animals, microorganisms or plants. They are easy to absorb subject to enzyme degradation, and they have an ability to induce cell signalling and tissue-to-scaffold communication [28]. Synthetic scaffolds are built from biodegradable polymers, whose structures can be easily designed into the desired 3D matrix. In addition to their biodegradable nature, higher pore rate and biocompatibility, they often do not provoke immune responses compared with natural scaffolds [28].

Biological/natural, semi-synthetic & synthetic/artificial scaffolds

Although amniotic membranes have been investigated as a potential scaffold for cultured HCECs, its translucent property creates a conflict when it comes to transplantation [26,30]. Therefore, modified denuded amniotic membranes have been reported to overcome the issue of lower transparency [31]. Other naturally derived biological scaffolds include aloe vera gel and silk fibroin, which have shown functionality when transplanted in animal models [32]. Human anterior lens capsule has also been described as a potential carrier of HCECs [33]. CECs have also been maintained on decellularized bovine corneas [34]. Decalcified Tilapia (Oreochromis mossambicus) scales have been proposed as an ideal scaffold following structural modifications. Besides being transparent, the morphology of the fish-scale scaffold is similar to the cornea, which makes it biocompatible for culturing a monolayer of endothelial cells [35].

The scaffolds that are based partially or fully on artificial and synthetic materials have been categorized into semi-synthetic or artificial scaffolds. Transparent gelatine fabricated scaffolds modified with heparin have been reported[36], as have gelatine hydrogels that fit the curvature of the posterior corneal surface sheets [37], plastic compressed collagen [38] and bioengineered substrate recapitulating chemo-mechanical properties of DM applications [39]. Co-cultures of corneal endothelial and epithelial cells have been reported on top of the collagen fibers, where endothelial cells were seeded on the silkworm (Bombyx mori) fibroin coated with collagen IV, FNC coating mix and a chondroitin sulphate-laminin mixture [40,41]. In both cases, HCECs showed proliferation capabilities and were proposed for use as a scaffold [40,41]. Chen et al. combined the silk fibroin with poly(l-lactic acid-co-ε-caprolactone) (P[LLA-CL]) to obtain biocompatible materials with good mechanical properties [42]. In addition, several other bioengineered materials such as collagen vitrigels [43], atellocollagen and gelatine hydrogel sheets [44] and collagen or silk fibroin [45] have been studied. Gelatine hydrogel sheets [44], for example, are one of the proposed scaffolds for corneal endothelial tissue regeneration. Recently, Van Hoorick et al. introduced the biodegradable amorphous polyester (poly[D,L-lactic acid]) combined with cross-linkable gelatines, which support HCECs in proliferation and DM fabrication [46]. It was also noted that real architecture for 3D tissue (RAFT) biomaterial is the simple and rapid method of production, which yields multiple reproducible constructs with limited variability between batches [38]. A review by Teichmann et al. showed that the application of thermo-responsive carriers to the engineering of corneal endothelial tissue is highly promising, as it probably poses less risk of inducing complications than other carrier-based approaches [47].

Tissue engineering can help produce scaffolds, cells and biomolecules, resulting in healthy tissue substitutes. However, it is difficult to maintain the stability of these sheets during transport and surgery [48]. Several different cell carriers have been studied for the purpose of endothelial layer construction, but the possibilities are limited by the specific requirements such as cytocompatibility, reproducibility, ease of manufacturing, transparency and flexibility for surgeons. So far, scaffold development has been proposed for the transplantation of cultured CECs in the eyes of animal models [1].

Tissue-engineered human corneal endothelial grafts for restoring corneal endothelial functions in animal models

Tissue-engineered human corneal endothelial grafts could be an alternative and reduce the challenge of the limited availability of donor corneas worldwide. Peh et al. reported laser dissection to isolate ultra-thin lenticules (100 μm) from human donor corneas. All tissue pieces were built from the intact DM-endothelium layer and corneal stroma. Human corneal endothelial cells (first or second passage) were seeded on top of the scaffolds and maintained until the day of surgery [14]. Rabbit corneas were used in this study, which was divided in three groups: engineered scaffolds with settled HCECs, DM stripped and no graft received and just the DM/stroma tissue rings without HCECs [14]. Because rodent corneas have the ability to regenerate the corneal endothelium, bullous keratopathy was induced. The rabbit corneas with tissue-engineered grafts showed gradual thinning of the stroma by the second week follow-up. The tendency was observed until the 28th day. It suggested that the corneal endothelial cells' pump-and-leak mechanism functioned normally [14]. End-stage immunostaining proved that the only human positive cells were located on top of the tissue-engineered grafts and not on top of the rabbit stroma with maintenance of hexagonal cell configuration confirmed using alizarin red staining. There was also no integration between the human and rabbit endothelium, proving that HCECs could still perform without adapting to the environment [14]. One of the biggest obstacles in culturing HCECs is differentiation of these cells via endothelial-to-mesenchymal transition (EnDMT), which limits the ability to build up the functional tissue-engineered scaffolding system. To exclude this issue, Zhao et al. isolated a subpopulation of HCECs, non-transfected monoclonal HCEs (mcHCEs). mcHCEs were cultured on top of modified amniotic membranes (mdAM) [49] to build the tissue-engineered human corneal endothelial construct (TE-HCE). mcHCEs were collected and seeded on top of the modified TE-HCE and the culture was prolonged for an additional 4 days. TE-HCE morphology was examined by scanning and transmission electron microscopes. Intercellular junctions, stained by alizarin red and ZO-1, proved that the constructed scaffold was suitable for grafting. Rabbit corneas were divided into two groups, which received fully constructed TE-HCE graft and just mdAM. Descemetorhexis was performed in both the groups. Six-month post-op follow-up was evaluated. Corneas grafted with TE-HCE showed minimal edema and corneal thickness in comparison with corneas with just mdAM, further suggesting normal functioning of the TE-HCE endothelial layer. Ex vivo examinations showed the hexagonal structure of the cells that managed to establish a monolayer producing the extracellular matrix [49].

Although, the study by Zhao et al. provides a brief insight into the possibility of using tissue constructs for endothelial cell regeneration, clinical study is needed for such constructs to be used as a replacement for human corneas. Although promising results have been achieved, tissue engineering has multiple challenges that must be faced before entering the clinical phase. Researchers have also identified different modes of administering cultured HCECs into the eye.

Injection of cultured HCEC suspension

Cell injection in an animal model

A preclinical study of cell-based therapy in a primate model showed that regeneration of corneal endothelium is possible. Monkey CECs (MCECs) from cynomolgus monkeys or HCECs when combined with ROCK inhibitor, Y-27632, were injected into monkeys' anterior chamber of the eye under good manufacturing practice (GMP) guidelines. MCECs had three experimental models: 5.0 × 10⁵ MCECs mixed with ROCK inhibitor and injected; 5.0 × 10⁵ MCECs injected; and no cells injected. In the HCECs, there were four experimental models: 5.0 × 10⁵ HCECs mixed with ROCK inhibitor and injected; 5.0 × 10⁵ HCECs injected; Descemet's stripping automated EK (DSAEK) using the precut tissues: and no HCECs injected after the stripping. After all the manipulations, the animals were set in facedown positions for 3 h (except for the DSAEK experimental group). The follow-ups were performed up to 1 year after surgery. There were no abnormalities recorded. In case of MCECs, the eyes not injected with cells or injected with cells without ROCK inhibitor exhibited haziness, compared with corneas injected with MCECs mixed with ROCK inhibitor, which restored corneal clarity. Other microscopical examinations proved this observation. Corneas were significantly thinner in the group with ROCK inhibitor and the endothelial cell density (ECD) was 2000 cells/mm². The injected cells showed expression of Na⁺/K⁺ATPase and ZO-1 and hexagonal morphology. There were no improvement and regeneration in the corneas without the injection of HCEC or a ROCK inhibitor supplementation. On the contrary, in the eyes injected with HCEC with ROCK inhibitor, transparency was restored in 1 week (it can be compared to the improvement after DSAEK). Corneas were visibly thinner in this group and endothelium was regenerated to proper hexagonal morphology with ECD of 2890 cells/mm². Graft rejections were observed in some cases of DSAEK and HCEC without ROCK inhibitor, but not in the group where cells were supplemented with ROCK inhibitor. In all groups, there was no cell aggregation or rise in intraocular pressure, but in the group injected with HCECs without ROCK, singular fibroblast cells were observed. Approximately 40–50% adherence of the injected CECs to the cornea was reported, which was calculated based on the number of CECs injected and the cell density of the regenerated corneal endothelium. The residual cells were thought to have flowed out by aqueous flow to other organs via the veins. The remaining cells adhered due to the presence of ROCK inhibitor, which increased the efficiency of the engraftment of CECs [48].

Xia et al. used New Zealand White (NZW) rabbits with stripped DMs of different diameters (5.5 and 8 mm, and half cornea). Isolated HCECs (from corneas cultured in Optisol-GS) were transduced with lentivirus to express green fluorescent protein (GFP) and combined with the superparamagnetic nanoparticles on the day of surgeries. These prepared cells were injected into the anterior chamber of adult rabbits (2–6 × 10⁵ magnetic HCECs/eye) after stripping the DM and endothelial cells. The magnets were applied on close eyelids on top of the corneas and the animals were positioned with the eye facing downward in contact with the magnet up to 3 hours. Animals were checked up to 3 months after surgeries. Rabbits receiving HCECs showed decreased corneal thickness and recovered the corneal clarity and transparency faster. The magnetic cells attached to the stripped area expressed ZO-1, NCAM, N-Cadherin and CD166 and hexagonal morphology. Rabbit corneas recovered faster after injection with the HCECs, magnetic particles helped with cell attachment into the stripped spaces and injected cells integrated into the rabbit corneas, expressing the typical markers and helping with tissue regeneration [50]. Based on three established culture techniques – the use of ROCK inhibitors, the use of conditioned medium obtained from mesenchymal stem cells (MSCs) and the inhibition of TGF-β signaling, an in vitro expansion protocol for clinical application has been developed and is currently being used under GMP guidelines in Kyoto Prefectural University of Medicine, Japan.

Cadaveric donor tissues with high endothelial cell counts are usually utilized for keratoplasty; therefore, tissues with low endothelial cell counts are mostly available for research. Ong et al. investigated a two-step incubation and dissociation approach to obtain single cells to be applied for corneal endothelial cell injection (CE-CI) therapy, which does not require cell culture, termed the simple non-cultured endothelial cells (SNEC) technique. The study on rabbits with bullous keratopathy showed an increase in mean corneal thickness at day 1 but reduced thickness thereafter, remaining clear at week 3. The control group without any cells remained thick. The advantages following the SNEC technique include utilization of donor tissues that are unsuitable for conventional keratoplasty procedures: use of a simplified method with less manipulation, thus bypassing the regulatory needs and requirements of GMP facilities and dropping the overall relative costs: and fewer challenges than technically demanding keratoplasty procedures. Cell therapy could thus present a viable treatment option, further addressing the challenge observed due to a global shortage of suitable donor corneas with less reliance on donor availability [51].

Cultured HCECs transplanted by the cell injection method

Kinoshita et al. used HCECs that were isolated from healthy donors and cultured under GMP conditions. The cells were passaged before the surgeries took place [52]. Following descemetorhexis, the passaged cells (1 × 10⁶) were supplemented with ROCK inhibitor and the resulting suspension was injected into the anterior chamber. Shortly after the injection, the patients were asked to stay in a prone position for 3 h to ensure cellular adhesion on the bare stroma. This technique was applied on 11 patients with bullous keratopathy. First examinations after 6 months showed that the cells attached to the stroma in each operated eye. The mean corneal ECD was more than 1000 cells/mm² in 10 eyes with a decrease in corneal thickness, maintenance of clarity and improved best corrected visual acuity (BCVA) in nine eyes after 6 months [52]. The corneas maintained their thickness and transparency at 2-year follow-up with mean ECD above 1500 cells/mm². There was no increase in intraocular pressure in 10 patients. This study suggested that cell injection supplemented with a ROCK inhibitor could serve as an alternative to conventional corneal transplantation methods. The 2-year follow-up ECD results were similar to those observed after standard keratoplasties. It should also be noted that not all the injected cells adhered to the corneal surface and therefore needed to be tracked to eliminate any potential risk. The 4-year data of the same study showed a decrease in cell density from 1924 cells/mm² to 1366 cells/mm² reported at 6 months and 48 months after surgery respectively [53]. None of the 11 cases showed any local or systemic adverse effects. However, the authors have reported that the study was limited in terms of a smaller number of patients and a shorter follow-up time, and it lacked comparative analysis to the EK procedures and overall associated costs of the cell-based therapy. The cell injection-based therapy shows potential for the treatment of patients with bullous keratopathy. However, treating patients with FECD, which is one of the leading cause of corneal blindness, will be the next challenging goal for cultured cell injection-based treatments.

Interestingly, the study by Kinoshita et al. [52] involved transplanted cells that were obtained from relatively young donors (7–29 years of age) with cell densities >2500 cells/mm². The availability of the old-aged (>60 years of age) donor tissues with endothelial cell counts <2200 cells/mm² is significantly high (~35% of all the donated tissues) compared with the young donors (<10%). If the cells from old-aged donors can be cultured, then the overall availability of the tissues would rise, especially as it will transform the tissues from ‘discarded’ or ‘research’ to ‘transplantation’ grade [54]. Although this would partially resolve the problem of sourcing the primary endothelial cells for culture, the heterogeneity, unpredictable culture outcomes, preservation conditions, plating density and other factors could limit the outcomes and will have to be further optimized. If transplanting half a million cells show similar visual outcomes as for those receiving a million cells, then it can certainly increase the efficiency of the cell injection technique. However, it would be worth investigating if 0.5 million cells would be enough when cells from old-aged donor tissues that have typical characteristics were transplanted and which patients would make the best candidates for low cell count injection. This would further treat a greater number of patients and broaden the reach of the cell injection technique to countries where sourcing young donor tissues can be extremely difficult. Moreover, the cell injection technique requires 1 million cells, which, if cultured on a cell sheet, can result in the transplantation of approximately five grafts of 8.5 mm diameter with ECD of 3000 cells/mm². Thus, the efficiency of this technique may need further investigation.

A study by Parikumar et al. demonstrated a helping material for cell injection therapy, called a nano-composite gel (NCG) sheet [55]. Three patients with bullous keratopathy were included in this study. HCECs were derived from cadaveric donor corneas and cultured in vitro. At day 26, the cells were prepared for injection. The patients received NCG insertion followed by cell injection. NCG was inserted into the anterior chamber close to the corneal endothelium. HCECs were injected into the anterior chamber between the affected corneal endothelium and the NCG. The sheets were removed at day 3 post-operation. The visual acuity of three patients was improved between day 3 and day 11 post-operation with decreasing corneal bullae. Moreover, there was no adverse effect observed in this study. From this study, it was noted that NCG can be used as a supportive material for cell injection. NCG prevented CEC washoff from the aqueous humor and secured the cells to be placed in a proper position. These results indicate the potential therapeutic use of HCECs, including their safety profile [55].

Add-ons to cell injection therapy

From the currently known techniques, as described earlier, transplantation of cultured CECs may be performed using a cultured corneal endothelial sheet or an injection of cultured CECs as a cell suspension. Despite the successful transplantation of cultured corneal endothelial sheets in an animal model [56,57], it has been found that transplanting such a fragile layer of cells on a scaffold is highly challenging when it comes to humans. In addition, the development of an artificial carrier is also a current obstacle for cell-sheet transplantation. Although the potency of injected cells has been studied, the main disadvantage remains, as the cells are removed by the flow of aqueous humor, resulting in poor adhesion of the injected cells to the corneal tissue [16]. Magnetic cell guidance using iron powder incorporated into cultivated CECs has been attempted [58], although not in humans so far. It has been noticed that adhesion of CECs plays an important role not only during cell cultivation but also after transplantation. Therefore, it was determined that the injection of cultured CECs with ROCK inhibitor could regenerate healthy corneal endothelium and recover corneal transparency in the monkey model. In terms of safety, no severe local or systemic adverse effects have been observed [59].

Jia et al. proposed a new technique called mini-sheet injection [60]. Rabbit CECs (RCECs) were cultured in vitro until they reached confluence. RCECs were dissociated with accutase and gently titrated into mini-sheets. Every mini-sheet containing aggregates of 4–10 cells were released after the treatment with accutase. Simultaneously, RCECs dissociated with trypsin-EDTA were prepared as single cells. The suspension of mini-sheets and single cells were injected into the anterior chamber. The rabbits were maintained in facedown position for 3 h. After the operation, corneal clarity and thickness rapidly recovered at day 7 in the mini-sheet group. In contrast to the single cells group, recovery was at day 14. The results from this study highlighted the advantage of injecting small group of cells (mini-sheet) compared with the single cells. The reason may be the quick settlement on the DM, making it less likely to be washed off by aqueous humor flow.

A recent study from Peh et al. evaluated the functionality of CECs delivered via tissue-engineered keratoplasty (TE-EK) and CE-CI in the rabbit model [61]. In this study, the lenticule of the stromal layer was cut by laser and prepared as a cell carrier material. The rabbits received cataract extraction followed by endothelium replacements. Corneal thickness was improved in both groups. In the CE-CI group, corneal thickness was improved at 1 week after the operation. Meanwhile, the corneal thickness in the TE-EK group was improved at weeks 2–3. The reason might be a smaller wound size in the CE-CI group. However, corneal thickness did not show a significant difference between the two groups after 2 weeks post-operation. The results from this study showed that both TE-EK and CE-CI can be alternative treatments for patients with endothelial cell diseases.

Alternatives to scaffold & cell injection techniques

Administration of ROCK inhibitor as eye drops

Apart from the injection technique that has already been studied on animal and human models, ROCK inhibitor eye drop (ripasudil) administration has also been studied recently. The effect and safety of selective ROCK inhibitor, ripasudil 0.4%, eye drops on CECs were monitored on six healthy eyes for 1 week. Morphological changes were only observed in the eyes instilled with ripasudil, thus highlighting that the phenomenon was regulated by ripasudil. Transient cell border irregularity and darker cells representing pseudo guttae-like features were observed at 1.5 h following the administration of ripasudil. No changes in corneal edema or ECD were observed after repetitive administration following 1 week. In fact, the morphological changes were reversible before the next administration. The study concluded that ripasudil eye drops can induce a transient functional, but reversible alteration without any endothelial cell loss or edema and, thus, can be safely used for short-term purposes [62].

Another study investigated the feasibility of using ROCK-Y27632 inhibitor eye drops for treating severe corneal endothelial damage due to surgical invasion in a rabbit model. Transparency increased in five out of six tested corneas and Ki67 positive cells were observed in the treated eyes compared with the control eyes. ROCK inhibitor eye drops have found to be effective in preliminary human studies for the treatment of bullous keratopathy. In conclusion, this study reported that ROCK inhibitor eye drops could be useful in treating acute corneal endothelial damage to prevent the progression of bullous keratopathy [62–65].

All the studies are summarized in Table 1 (animal models) and Table 2 (humans). The drugs that are currently under investigation for the treatment of endothelial disease are listed in Table 3.

DMEK & unmet challenges

DMEK is a technique whereby the damaged DM (ECM of the endothelial cells), along with the endothelium, is replaced by healthy donor tissues. However, the DMEK procedure requires skilled maneuvers and often leads to cell loss. Relatively new techniques, such as pre-loading DMEK grafts, could be useful in reducing the cell loss that occurs during preparation of the grafts [66–70]. When the fragile endothelial sheets are cultured and ready for transplantation on a potential scaffold, the entire unit, including the cells and scaffolds, could be pre-loaded into a specific device as ready-to-use tissue to reduce the manipulation and time required in the surgery. Although there are many potential culturing alternatives, the transfer of this highly fragile monolayer cell sheet into the anterior chamber and its stable fixation to the posterior cornea remain a surgical challenge that has not yet been met. With a limited number of corneal endothelial specific cell markers [71], it also becomes more challenging to characterize these cells for transplantation purposes.

DSO & DWEK

As there is a limited supply of donor corneas, techniques such as DSO and DWEK are becoming popular, as they avoid the use of donor tissues completely, which may in turn reduce the demand of developing novel surgical devices altogether [72]. Iovieno et al. demonstrated a successful DWEK with 4 mm central descemetorhexis. All five patients with central guttae and undetectable ECD presenting with a healthy peripheral corneal endothelium were enrolled in this study. At 3 months post-surgery, complete endothelial repopulation was observed in all the patients. Corneal clarity improved in four of five patients [73]. However, although early post-operative results of the DWEK technique have been promising, its longevity still needs to be determined. In another retrospective study, 17 eyes of 13 patients with FECD were assessed following DWEK. Post-operative endothelial cell counts increased steadily over the study period. DWEK was found to be successful following descemetorhexis, resulting in corneal clarity regardless of baseline demographics [74]. Davies et al. demonstrated a combination of DWEK and netarsudil (ROCK inhibitor) in patients with FECD. 20 eyes from 10 patients were included in this study. DWEK with post-operative netarsudil instillation was applied in one eye, while DWEK with delayed netarsudil was prescribed in another eye. A reduction of corneal clearance time was found in immediate netarsudil post-operation with an average time of 4.6 weeks, while the corneal clearance time was found to be 8 weeks in the delayed netarsudil group (p < 0.01) [75]. Macsai et al. showed that patients who underwent DSO with topical ripasudil 0.4% recovered vision quicker and showed significantly higher average ECD than those in an only-DSO group [76]. From these studies, it was observed that DWEK/DSO in combination with netarsudil could be used as an alternative treatment for FECD [74–76].

What besides ROCK inhibitor?

All cell-based therapies would require the cells to be produced under GMP standards, which is possible, using the medium supplemented with ROCK inhibitor [14]. Unfortunately, HCECs tend to decrease cell numbers at higher passages, thereby limiting the possibility of cell transplantation from one donor being used for more than one patient and overall tissue-engineering capacities. Hongo et al. managed to maintain the primary HCECs using a medium supplemented with SB203580 inhibitor for 18 months [77]. SB203580 is an inhibitor of p38 MAPK, which has a huge role in cell aging. p38 MAPK, activated due to cellular stress, has an influence on cellular activity, proliferation and apoptosis. Inhibition of this pathway not only shows higher proliferation capabilities of HCECs in vitro but also lowers the percentage of EnDMT [77]. Although, there is no preclinical study reporting the use of animal models or human trials with respect to the effectiveness of SB203580 inhibitor on the proliferation of HCECs, this study shows promising results and could facilitate endothelial cell culture with an ability of transplantation possibilities from a single donor to multiple recipients.

Artificial endothelial layer – a game changer?

A recent study showed that an artificial endothelial layer – EndoArt (EyeYon Medical, Ness Ziona, Israel) – made of flexible, hydrophilic acrylic material (6 mm diameter and 50 um thickness), which is equivalent to the cornea's posterior layer, functions as an artificial fluid barrier similar to the corneal endothelium. The deturgescence function of this device showed significant corneal edema reduction within 1 day post-operatively and remained stable at 17 months follow-up. Being an artificial implant, it is robust, does not rely on donor tissues, allows easy intraocular manipulation due to its thickness without the risk of damaging the implant and does not require immunosuppression [78].

GMP-compliant cell engineering

It is important to propagate HCECs under GMP conditions to reduce the risk of potential xeno-contamination and possible transfers of infectious pathogens. Although the regulatory guidelines usually differ between regions, it is mainly to ensure the safety, efficacy and quality of the cell/tissue-engineered product [14].

Tissue-engineered grafts are classified as advanced therapy medicinal products (ATMPs) in Europe; hence, they must be produced in compliance with European laws (Regulation [EC] no. 1394/2007 of the European Parliament and of the council of 13 November 2007) [14], whereas cell-based therapeutics are governed by the US FDA and must comply with the Code of Federal Regulations, Title 2 1, Part 211 (Current Good Manufacturing Practice for Finished Pharmaceuticals) and Part 1271 (Human Cells, Tissues, and Cellular and Tissue-based Products). Many enzymes, buffers, materials, growth factors and reagents have successfully been fabricated under GMP-compliant standards. These developments have increased the viability of clinically acceptable tissue-engineered therapeutic alternatives. It has also been clear that the maintenance of a large-scale GMP-compliant tissue-engineering facility would significantly increase the working and overall costs. It has been projected that the cost of processing each pair of cadaveric human corneas to derive tissue-engineered constructs would amount to US$24,265 [14] compared with a routine keratoplasty (US$7320), which is significantly higher. However, this could be compensated by the large number of tissue-engineered grafts, further increasing the affordability of EK.

Challenges

Although the culture and transplant methods seem reasonable, it appears that an optimal model is yet to be found. Multiple factors limit the success of the cell/tissue-based approach, such as donor factors that include but are not limited to the age of the donor, the region where the endothelial cells are isolated (central vs peripheral), the preservation condition of the tissues and the cultured HCECs, plating densities, coating solutions, universal growth media, evaluation and characterization of the cultured HCECs, carriers for culturing and transplanting HCECs, long-term outcomes of injection/DWEK models, approval from the regulatory bodies for compliance and assurance and cost–effectiveness to be globally accepted [79]. However, these challenges are part of any successful medical product; therefore, overcoming these with standardized procedures will be the next big step in the field of EK replacement.

Conclusion

It has been observed that although there are methods described for the culture and possible sheet transplantation of CECs in vitro, there is no significant clinical outcome reported so far, especially considering cultured HCECs. HCECs are challenging to culture, especially considering that older donor corneal cells do not grow as optimally as those of young donors. The availability of older donor corneas is higher and therefore the culture capacity of these tissues should be increased to reduce the overall burden of corneal tissues worldwide. Cell suspension, sheet transplantation or eye drops may be used in the future with better efficacy. However, it requires a long-term clinical follow-up to assess their feasibility. Although there is a huge potential for HCEC culture and transplant, there is not enough evidence or literature that proves its efficiency in real terms and therefore the culture and its transplantation technique needs to be refined, especially for clinical applications. However, with the current trends and scientific knowledge that has been generated over the years, a possible cultured HCEC transplantation does seem to be a reality in the near future.

An understanding of cell biology and cell characteristics in different conditions of the cornea is essential for improving the cell culture technique. There are several challenges in every step of corneal endothelial cell culture for promoting cell proliferation. However, overcoming the unresolved issues such as an increased donor pool for culture, the effect of corneal preservation conditions, the achievement of homogenous culture, an imbalance between areas where culture is needed and where available, and regulatory and economic barriers would increase the potency of cell/tissue engineering-based products. Advances in cell culture and tissue engineering thus bring hope that effective and safe techniques for the treatment of corneal endothelial cell diseases may soon be developed.

Future perspective

As human tissues are precious, their optimal use is of high priority. Therefore, a single cornea is now used to treat multiple patients thanks to selective transplant methods [80]. A rising interest has been noted in potential hybrid techniques between DSO and DM endothelial transfer and conventional circular DMEK. Techniques such as quarter-DMEK and hemi-DMEK have been recently introduced, potentially allowing the harvesting of two endothelial grafts from a single donor corneoscleral rim [81,82]. This would further reduce the need of human donor corneas; however, these type of tissues would require special devices for consistent graft preparation and loading due to their specific sizes and shapes. Although advanced surgical techniques and cell therapies have the potential to treat multiple patients, they still rely heavily on donor tissues. Therefore, techniques such as DWEK/DSO that reduce the reliance on human tissues would be ideal if proven successful in the long term without contraindications in the future. A GMP-certified synthetic or artificial cornea that can be mass manufactured and cost-effective would be a game changer in the future and with products like EndoArt that reality seems on the horizon.