A Contact Lens-Based Technique for Expansion and Transplantation of Autologous Epithelial Progenitors for Ocular Surface Reconstruction

Abstract

Background. 

A healthy cornea is reliant on a distinct population of stem cells (SC) that replace damaged or aging epithelium throughout life. Depletion of the SC pool or damage to the niche can result in a blinding and painful condition known as limbal-SC deficiency (LSCD). Although current treatment strategies for reconstituting the ocular surface for patients suffering LSCD are promising, they are complicated by transferring autologous or allogeneic progenitors in the presence of animal, human, and synthetic products. We report on the safe and efficacy of a unique autologous SC transfer technique that utilizes an Food and Drug Administration-approved contact lens (CL) as the SC substrate and carrier for patients with LSCD.

Methods. 

Three patients with LSCD due to aniridia (n=1) and posttreatment for recurrent ocular surface melanoma (n=2) were included. Limbal (n=2) or conjunctival biopsies (n=1) were harvested and progenitors expanded ex vivo on therapeutic CLs in the presence of autologous serum. Cell-laden CLs were transferred to the patient's corneal surface and clinical outcome measures were recorded (follow-up range, 8–13 months).

Results. 

A stable transparent corneal epithelium was restored in each patient. There was no recurrence of conjunctivalization or corneal vascularization, and a significant improvement in symptom score occurred in all patients. Best-corrected visual acuity was increased in all eyes after the procedure.

Conclusion. 

Ex vivo expansion of ocular surface epithelium in the presence of autologous serum and transplantation with the aid of a soft CLs is a promising new technique capable of achieving ocular surface rehabilitation.

Limbal stem-cell deficiency (LSCD) resulting in ocular surface failure is the final common endpoint after damage to the limbus from a variety of insults including chemical or thermal burns, multiple surgeries, radiation therapy, cryotherapy, and chemotherapy, or severe microbial infection. LSCD can be manifested through disorders characterized by mucous membrane dysfunction such as Stevens-Johnson syndrome and ocular cicatricial pemphigoid. Aniridia is a congenital form of LSCD that develop from mutations in Pax6. Complications of LSCD include corneal conjunctivalization, neovascularization, ulceration, inflammation, and scarring which result in pain, loss of vision, and vulnerability to infection (¹).

Treatment for LSCD includes application of a bandage soft-contact lens (CL), intensive nonpreserved lubrication, and autologous serum eye drops (²). However, these therapies are rarely successful when total limbal epithelial stem-cell (LESC) failure occurs. In cases of total LESC failure, a population of stem cell (SC) must be introduced to regenerate and restore the ocular surface. Strategies to correct LESC failure have relied on harvesting large segments of limbus from the healthy contralateral eye (unilateral cases) (³); however, SC depletion in the donor eye is a risk. An alternative method utilizes allogeneic grafts (bilateral cases) from living relatives or cadaveric donors. However, persistent corneal defects have been noted, and to prevent rejection, long-term systemic immunosuppressive therapy is required (⁴). For unilateral limbal dysfunction, cultured autologous limbal progenitors have been transplanted (⁵). A decade ago, Pellegrini et al. (⁶) pioneered a method of culturing LESC from small (1 mm²) grafts harvested from a healthy eye (in unilateral cases), expanded in culture, and maintained on a layer of lethally irradiated mouse 3T3 feeder cells. Epithelial progenitors were removed from the feeder layer as intact sheets and transferred directly to the patient's cornea using a soft CL as the carrier. Other carriers used have included paraffin gauze, porcine collagen shields, synthetic products, and fibrin gel (¹); however, all these methods are cumbersome as the cell sheet is free floating. Amniotic membrane alone (for partial LSCD) or as a substrate for ex vivo expansion of LESC (in total LSCD) minimizes these problems but still requires a surgical procedure (¹). However, donor-to-donor variability, limited availability, cost of screening, preparing and storing, and its inherent semiopaqueness, which can impede visual acuity, are disadvantages.

Despite the promising clinical outcome reported for current autologous transplantation procedures to treat LSCD (¹), to our knowledge most methods use animal products, foreign human tissue (^4–6) or non-Food and Drug Administration (FDA) approved biomaterials (⁷) thereby increasing the risk of xenobiotic infection, and side effects from the biomaterials. Investigators have trialed synthetic and biodegradable polymers as cell carriers in animal (⁸) and culture (¹) models of LSCD, but their efficacy in humans is not yet established (⁸). We developed a novel autologous technique that circumvents these problems by using an FDA-approved soft CL as the substrate, carrier, and bandage to protect the eye during transplantation and healing. We report herein on the remarkable ocular surface restoration in three patients with LSCD with no adverse side effects.

METHODS

Protocols relating the use of human cells and tissue were approved by the University of New South Wales Ethics Committee (HREC-06290) and carried out in accordance with the tenets of the World Medical Associations Declaration of Helsinki. The clinical trial was registered in Australia (ACTRN-012607000211460) and informed, written consent was obtained. The three patients included in this study were consecutive referrals to the Corneal Unit (Prince of Wales Hospital, Sydney) with longstanding total LSCD. Each had failed all prior maximal therapy including autologous serum drops, therapeutic CL wear, preservative free lubricants, and limbal SC allografts. Patient demographics were collected on a proforma and entered into a database. They included diagnosis, prior medical and surgical management, acceptability of current therapy, signs of dry eye, ocular symptoms score (severity graded 0–4) (²), facial analogue score (⁹), best-corrected visual acuity (BCVA), blepharitis score, and conjunctival injection score (¹⁰).

For autologous transplantation epithelial biopsies (1 mm²) consisting of superior forniceal conjunctiva (for bilateral cases) or superior limbal (for unilateral cases) epithelium were excised from the fellow eye under local anesthesia (Minims Tetracaine Hydrochloride 1%; Chauvin Pharmaceuticals, Bausch and Lomb, United Kingdom) and transported to the laboratory in basal growth media on ice. These sites are known SC rich regions and protection by the upper eye lid is known to aid healing after biopsy (¹). Serum was obtained at time of biopsy, prepared from 20 mL of whole blood (²) and incorporated (10% final) into growth media (Eagles minimum essential medium [EMEM]) containing antibiotic supplements as previously described (¹¹). Each biopsy was placed on a siloxane-hydrogel (Focus Night & Day) (lotrafilcon A, CIBA Vision) extended wear CL in 24-well culture plates (Corning Inc, NY) in an isolated, humidified incubator set to 5% CO₂. Cell growth was monitored daily, media was changed every other day and routinely assessed for microbial contamination. At confluence (∼10 days), the cell-laden CL (with biopsy attached) was rinsed and transported to the clinic in growth media.

Before CL insertion, 5% betadine was applied to the ocular surface. The cell-laden CL was rinsed saline and inserted into the patient's eyes under topical anesthetic (Minims Benoxinate Hydrochloride 0.4%; Chauvin Pharmaceuticals, Bausch and Lomb) using sterile forceps. A superficial keratectomy was performed to remove irregular epithelium and conjunctivalization (¹²) before CL insertion. All patients were continued on their prior topical and systemic therapy.

To assess matrix components and enzymes synthesized by proliferating progenitors, cadaveric tissue was obtained from the Lions Eye Bank (Sydney, NSW), cut into small segments, placed in culture plates, propagated, enzymatically released as single-cell suspensions, and seeded on CLs. Cadaveric limbal tissue was placed on the CL surface, cultured precisely as described earlier (¹¹), harvested, formalin fixed and processed for paraffin embedding, and sectioned for histologic and immunohistochemical evaluation as previously described (^13–16) using mouse monoclonal and rabbit polyclonal antibodies (Table 1).

TABLE 1: Primary antibodies used for immunohistochemistry

RESULTS

Three eyes of three patients underwent autologous cultured LSC transplantation. A transparent corneal epithelium was restored in each patient and the donor sites healed without sequelae. Patients' demographics and outcome measures are summarized in Table 2.

TABLE 2: Patient demographics and summary of results after ex vivo autologous stem-cell culture and transplantation

Patient 1

A 34-year-old White man presented with bilateral poor vision and recurrent persistent epithelial defects (PEDs) due to bilateral total LSCD. He had aniridia with congenital nystagmus, glaucoma, and was a bilateral psuedophake. In the right eye he had had a living-related right keratolimbal graft, which had failed. His topical therapy in both eyes was fluorometholone 0.1% (FML liquifilm, Allergan, Australia) daily and preservative free polyvinyl alcohol 1.4% (Refresh tears, Allergan) twice daily in both eyes. Because of a steroid responsive increase in intraocular pressure he was unable to use preservative free dexamethasone or prednisolone. In the right eye, he was using timolol 0.5% (timoptol-XE, Merck Sharp & Dohme, Australia) daily and retinoic acid 0.01% (Sydney Eye Hospital Pharmacy, Australia) once every 2 days. His systemic medications were mycophenolate mofetil (CellCept, Roche, Switzerland) 500 mg two times per day for immunosuppression after the keratolimbal graft and minocycline 50 mg daily (Minomycin, Sigma Pharmaceuticals, Australia) for posterior lid margin disease. His BCVA was 3/60 right and 1/60 left. The right corneal epithelium was irregular and opaque centrally with mild central subepithelial opacity, and the left cornea was conjunctivalized (Fig. 1A). Superior bulbar conjunctival biopsies were taken from the left eye and cultured on a CL. A superficial keratectomy (Fig. 1A) was performed to remove the abnormal epithelium from the cornea, and a cell-laden CL was inserted. At day 1, the CL was found rolled up under the upper lid. A second cell-laden CL was inserted (Fig. 1B) and removed 22 days later. Figure 1(C) shows the clinical appearance at 13 months. At 13 months, his BCVA had improved to 6/60, and the corneal epithelium was transparent and less irregular centrally, there was mild persistent subepithelial opacity.

FIGURE 1.: Clinical features of patients with limbal stem cell deficiency before and after stem-cell transplantation. Slit lamp photographs of patient 1 (A–C), patient 2 (D–F), and patient 3 (G–I). Patient 1 (a 34-year-old man) is shown after superficial keratectomy (A), with the contact lens (CL) in situ (B), and at 13 months postoperatively, showing an intact, stable epithelium with mild superficial subepithelial haze. Patient 2 (a 71-year-old woman) is shown preoperatively (D), with the CL in situ (E), and at 13 months postoperatively with an intact, transparent corneal epithelium and reduced vascularization (F). The limbal explant was incorporated within the epithelium (F, arrowhead). Patient 3 (a 69-year-old woman) is also shown preoperatively (G), with the CL in situ (H), and at 8 months postoperatively, a relatively smooth epithelium with reduced corneal vascularization is shown. Deep and superficial corneal stromal scarring remain.

Patient 2

A 71-year-old White woman presented with a right painful eye and poor vision from total LSCD (Fig. 1D). The right eye had multiple surgical excisions, cryotherapy, and topical mitomycin C to treat recurrent superficial spreading conjunctival melanoma arising from primary acquired melanosis. Her topical therapy was minims prednisolone 0.5% (Chauvin pharmaceuticals) two times per day to the right eye. Her BCVA was 6/36 right and 6/6 left. The right cornea was conjunctivalized in four quadrants and the bulbar conjunctiva markedly injected (Fig. 1D). Three superior limbal biopsies were excised from the normal left eye and cultured on CLs for 10 days. A superficial keratectomy was performed prior to CL placement (Fig. 1E). Biopsy remained within the corneal epithelium after CL removal (Fig. 1F, arrowhead) at day 19. At 13 months the right BCVA was 6/12 and the eye was comfortable. The central corneal epithelium was intact and transparent. Histologic analysis of the gelatinous pannus removed before CL placement revealed recurrent melanoma. Systemic investigations were negative for metastatic melanoma. Topical treatment with mitomycin C was commenced.

Patient 3

A 69-year-old White woman presented with right LSCD after multiple excisions, cryotherapy and topical mitomycin C for primary acquired melanosis and focal invasive malignant melanoma of the conjunctiva. Her right eye was painful, photophobic and had poor vision. At referral her topical therapy was Refresh tears plus (Allergan) and Polyvisc ointment at night (Paraffin, wool fat; Ioquin Company, Sydney, Australia). The preserved lubricants were ceased and Cellufresh (Allergan) commenced. On examination her right eye had a BCVA of count fingers and the left eye 6/9. The right cornea was conjunctivalized and the bulbar and palpebral conjunctiva injected and scarred (Fig. 1G). Three superior limbal biopsies were taken from the normal left eye, and a cell-laden CL was inserted after superficial keratectomy (Fig. 1H). Deep corneal vascularization and scarring was noted postkeratectomy. The CL was removed after 14 days. On last review, 8 months after the biopsy, her right eye was comfortable and her BCVA had increased to 6/60. She had a regular, intact corneal epithelial surface, and there was a reduction in the deep corneal vascularization (Fig. 1I).

Ex Vivo Expansion of Ocular Surface Progenitors on Contact Lenses

Epithelial cells sporned from limbal (Fig. 2A–F) and conjunctival (Fig. 3G–I) explants as early as 2 to 3 days after placement on the CL and reached the CL edge after 10 days (Fig. 2I, white arrow), which was the time chosen to transplant epithelial progenitors. CLs removed after the procedure demonstrated occasional remnant cell colonies (Fig. 2J and K).

FIGURE 2.: Limbal and conjunctival tissue explants cultured on therapeutic contact lenses. Limbal (A–F) and conjunctival (G–I) biopsies were harvested from patients (Patient 3, A–F; Patient 1, G–I; Patient 2, J and K) undergoing our autologous stem-cell transplantation procedure and placed as explants on the concaved surface of sterile siloxane-hydrogel contact lenses. Epithelial growth and migration (arrowheads) was noted as early as 2 to 3 days in culture (A, B, and G) and progressively expanded to reach confluence (F) and the contact lens edge (white arrow). Asterisk in (H) identifies the contact lens serial number. The area encompassed by the rectangle in (A) is magnified in (B). After the transfer procedure, contact lenses were removed from the ocular surface, fixed in formalin, and stained with hematoxylin as a whole mount (J and K) to identify remaining cells. Photomicrographs were taken at progressive intervals (days 2–10). Original (×40, A, C–I; ×200 B; ×1000 J and K).

FIGURE 3.: Matrix components in cells derived from limbal explants cultured on contact lenses (CLs). Human limbal explants were removed from the CL surface after 10 to 12 days in culture and processed for immunofluorescence (A–K) or immunohistochemical (H and L) assessment. Sections incubated with an appropriate isotype control antibody developed no immunoreactivity (A). Sections exposed to an anticollagen I (B), fibronectin (D), MMP-3 (F), and MMP-1 (G) were illuminated with green fluorescence. In contrast, collagen IV (C) and laminin (E) were not detected and only dihydrochloride 4′,6-diamidine-2-phenyl indole-stained nuclei (blue) were evident. Cells that lined the CL surface formed multiple layers (A, white arrow), and their nuclei were intensely (H, red arrows) or moderately (H, blue arrows) reactive to p63 (pan A4A antibody). Likewise, these cells were also reactive toward the α-specific isoforms of p63 (I and J), ABCG2 (K), and displayed cytoplasmic p75 reactivity (L). Formalin-fixed human cadaveric corneal tissue (M–O) was sectioned and stained using the same collagen type I antibody. No staining was identified in the central cornea (M), but was apparent in the peripheral cornea, particularly in the basement membrane that lined the basal epithelium (N, arrowheads) adjacent to Bowman's layer (N and O, arrows). This pattern of staining was more prominent at the limbus (O, arrowheads) with basement membrane and cell-associated reactivity observed. Bowman's layer is present in (N) as denoted by the arrows, but is absent to the left of the arrows indicating the beginning of the limbal zone. Immunoreactivity displayed in panels (M–O) was developed with 3-amino-9-ethylcarbazole (AEC) and nuclei counterstained with hematoxylin, except for panel (H). Original (×1000) under oil immersion.

Matrix Proteins and Enzymes Deposited on Contact Lenses

In a previous investigation, we showed that limbal-derived epithelium cultured on a CL substrate displayed a progenitor-like phenotype (¹¹). Here, we extended those studies to determine the expression of extracellular matrix constituents and matrix metalloproteinases (MMPs) in human cadaveric limbal tissue explants that were placed on therapeutic CLs and harvested when cells reached confluence. After 10 days in culture, two to three layers of epithelial cells were noted (Fig. 3A, arrow). Of the matrix proteins that were investigated only collagen type I (Fig. 3B) and fibronectin (Fig. 3D) were identified on CL- adherent epithelial cells. Collagen type IV and laminin was not detected (Fig. 3C and E). Antibodies directed against stromelysin (MMP-3; Fig. 3F) and collagenase-1 (MMP-1; Fig. 3G) illuminated the same cells. In whole human corneal specimens, collagen type I was absent from the central cornea (Fig. 3M), became apparent in the peripheral cornea (Fig. 3N) and was abundant at the limbus (Fig. 3O), where deposits were identified between basal limbal epithelial cells and the basement membrane (Fig. 3N and 3O, arrowheads).

Limbal Epithelia Cultured on CLs Display a Stem Cell-Like Phenotype

In addition to the SC-like properties identified in our previous investigations (¹¹), limbal epithelial cells nurtured on a CL were immunoreactive not only for p63 using an antibody that detects all six isoforms of this transcription factor (Fig. 3H, arrows), but also using isoform-specific reagents (¹⁷), p63α and ΔNp63α (Fig. 3I and J, respectively) positive cells were identified along the CL surface. Another well-characterized limbal SC marker, often referred to as the “universal” SC marker, is the ATP-binding cassette subfamily G, member 2 (^{18, 19}), this antigen (Fig. 3K), and the low affinity nerve growth factor receptor, p75 (¹⁶) (Fig. 3L) were both localized to a proportion of CL-adherent cells suggesting maintenance of a limbal SC phenotype under these conditions.

DISCUSSION

Our prospective pilot clinical trial for patients with LSCD has demonstrated the efficacy of using a siloxane-hydrogel CL as (1) a substrate for culturing ocular surface epithelial progenitors, (2) a device for carrying progenitors to the ocular surface, and (3) a bandage to protect the transplanted SC during transfer thereby promoting corneal repair and regeneration. Our novel technique (¹¹) bares many advantages over traditional methods for treating such patients, the most important of which is its autologous nature. Of particular note, early work in human skin identified persisting bovine serum proteins in keratinocyte cultures that had been switched to serum-free conditions (²⁰), and circulating antibodies to fetal bovine serum proteins were detected in burns patients many months after receiving cultured keratinocytes (²¹), potentially contributing to graft rejection by the activation of an immune response. In our system, epithelial cells were nurtured ex vivo in autologous serum, expanded on an FDA-approved soft CL and subsequently placed on the patient's ocular surface. By using our procedure, cells were transplanted within 10 days of obtaining a biopsy, offering flexibility in scheduling the procedure. Moreover, multiple biopsies were harvested to ensure the procedure was not reliant on a single tissue explant, and if complications arose (which was the case for patient 2), a second lens was readily available. Opting to leave the explant on the CL during transplantation was regarded advantageous as we reasoned that progenitor cells would continue to spawn from the tissue and provide signals for dispersal, integration, and regeneration of the cornea.

Reported mean success rate for cultured autologous LESC therapy is approximately 75% at between 6 and 29 months. However, this figure derives from multiple studies in which there was considerable heterogeneity in patients enrolled and techniques used, along with variable follow-up (¹). Similarly, our small case series had differing etiologies of LESC failure and used a novel technique, therefore making it difficult to compare our results with those previously published. It is possible that with larger patient numbers and longer follow-up, a subgroup of patients may emerge that experience significant benefits over others. Caveats to keep in mind when comparing clinical outcome between patients are (1) the type of cell culture protocol, (2) the subjectivity of clinical assessments made pretransplantation and posttransplantation, (3) the type of pretransplant and postprocedure management, (4) number and type of prior ocular surgical procedures and (5) coexistent ocular and systemic morbidities.

Identifying other SC sources for patients with LSCD is a major research focus as there is a shortage of donor corneas. Alternative sources have included embryonic SC (²²) and the use of transplanted autologous oral mucosal epithelium (²³). The former approach is yet to be trialed on humans; however, caution should be exercised as embryonic SC cultured in the presence of mouse 3T3 feeders begin to express animal glycoproteins (²⁴). The latter holds more promise (²³), despite midterm follow-up results that identified peripheral corneal neovascularization (²⁵). One should be cautious when interpreting results from this procedure, particular when treating patients with Stevens-Johnson syndrome and ocular cicatricial pemphigoid, as the autologous oral mucosal biopsies may be compromised by the patient's systemic disease. Arguably, the epithelium closest in lineage to the cornea is the conjunctiva, a tissue that harbors its own SC reservoir (²⁶) but is morphologically, biochemically, and functionally diverse from the cornea. As with oral mucosal transplants, autologous cultured conjunctival epithelium (²⁷) and conjunctival grafts (²⁸) have been used successfully to treat patients with LSCD. In our small cohort, patient 1 was transplanted with autologous conjunctival progenitors that rapidly revitalized the corneal surface. Although the precise mechanism by which conjunctival progenitors repopulate and stabilize the corneal surface is not fully understood, these cells have the capacity to transdifferentiate after encountering niche-specific signals (^{29, 30}).

We have propagated and compared the phenotype of limbal and conjunctival epithelium form tissue explants as well as from subcultured secondary cells seeded on CLs and found remarkable phenotypic (CK-15, p63α, and ATP-binding cassette subfamily G, member 2) and functional resemblance between these two epithelia using our cell expansion conditions (data not shown). These results could explain the successful reconstitution and maintenance of a healthy ocular surface in patient 1 with aniridia who received an autologous conjunctival transplant. This is likely as recent studies have demonstrated that murine corneal and conjunctival holoclones have an identical ocular gene expression signature (³¹). The same investigators also provided evidence of corneal epithelial oligopotency, suggesting that corneal cells have the reprogramming capacity to switch phenotype if provided with the appropriate conjunctival stromal signals. Our data, along with previously published reports on the reconstitution of the corneal surface with conjunctiva (^{32, 33}), is further evidence that these two distinct epithelia might be interchangeable. This close relationship has also been observed by others who identified corneal epithelial clusters within the conjunctival epithelium (³⁴).

Long-term corneal health after transplantation has always been a concern, irrespective of the treatment strategy used and survival of LESC after transplantation is currently a topic of intense debate as investigators cannot identify allogeneic LESC beyond 3 to 9 months (^{35, 36}), raising questions about the mechanism of LESC longevity and whether or not SC are actually transplanted (³⁷). It is tempting to speculate that donor SC provide signals that activate dormant or quiescent LESC (³⁸) alternatively bone marrow-derived SC might be a potential source of SC that regenerate the corneal surface (³⁹). Once SC are isolated from their native microenvironment and cultured on foreign substrates, their genetic program may switch from one that supports SC selfrenewing to one that promotes progenitor differentiation. Given that the lifespan of a transient amplifying cell is said to be 6 to 12 months (³⁵) this may be one reason why the ocular surface remains relatively stable over this period. Indeed in our study we have shown success in one patient up to 8 months and 2 patients for more than a year.

MMP-3 is an enzyme capable of cleaving proteogylcans and fibronectin, and MMP-1 has highest specific activity against interstitial fibrillar collagens type I, II, and III (⁴⁰). The coexpression of selective matrix proteins (collagen type I and fibronectin) and matrix enzymes (MMP-1 and MMP-3) by CL-adherent epithelial cells implies a healthy culture environment conducive to maintaining progenitors (^{41, 42}). The absence of collagen type I in the central cornea (Fig. 3M), but presence in the peripheral cornea (Fig. 3N), and abundant expression in the limbal zone (Fig. 3O) is further evidence to suggest this matrix component is an essential factor that may support stem/progenitor cells in their microenvironment (⁴¹). A similar differential pattern of collagen and laminin expression has been demonstrated between the central corneal and limbal region and postulated to be due to diverse functional and differentiation status of epithelial cells in this critical demarcation zone or the contribution of stromal signals that may differ between these two narrow regions (⁴³). Moreover, the expression of fibronectin by our CL-adherent progenitors is of practical and clinical importance as this ubiquitous extracellular matrix protein is known to inhibit terminal differentiation in cultured human keratinocytes (⁴⁴). The absence of collagen type IV and laminin expression on CL-grown epithelial cells is intriguing; however, this may be one reason why progenitor cells are able to migrate from a synthetic substrate to the native corneal surface as a comprehensive basement membrane has not yet been synthesized.

A limitation of our study is that we were unable to provide direct evidence to demonstrate the actual transfer of precursors from the device to the cornea. Studies using allogeneic (chromosome mapping) or autologous cells (prelabeled with tracer dyes) may be warranted to addressed this issue. We did, however, observe several healthy epithelial colonies that remained on the CL surface after transplantation (Fig. 2) suggesting that the majority of cells that covered the CL surface had most likely transferred as indicated by the rapid corneal regeneration in our patients.

The technique described herein has been successful in reconstituting a healthy ocular surface in three patients with LSCD. This novel procedure can be performed by a corneal surgeon who is required to remove the biopsies and an experienced laboratory researcher with standard cell culture equipment. However, the greatest advantage over current conventional treatment strategies is the autologous nature of the method.