Cultured Autologous Oral Mucosal Epithelial Cell Sheet (CAOMECS) Transplantation for the Treatment of Corneal Limbal Epithelial Stem Cell Deficiency

Abstract

Purpose.: Total bilateral corneal limbal epithelial stem cell deficiency (LSCD) cannot be treated with the surgical transplantation of autologous limbus or cultured autologous limbal epithelium. Transplantation of allogenic limbal epithelium is possible but requires immunosuppressive treatments. Cultured autologous oral mucosal epithelial cell sheet (CAOMECS) is a transparent, resistant, viable, and rapidly bioadhesive cell sheet, cultured with the UpCell-Insert technology (CellSeed, Inc., Tokyo, Japan), which allows for grafting onto the patient's corneal stroma without suturing. It has therefore been proposed as an alternative treatment for LSCD.

Methods.: The objectives were to assess the safety and efficacy of CAOMECS, using a prospective Gehan's design. Safety was measured in terms of ocular adverse events during the study period, and efficacy was measured using a composite criterion based on epithelial defect, punctate epithelial keratopathy, conjunctival epithelium on the cornea, number of vascular pediculi, and vessel activity.

Results.: CAOMECS was found to be safe and effective. In total, 26 eyes of 25 patients received a graft. Two patients experienced serious adverse events classified as not product related. Twenty-five patients were included in the efficacy analysis, as one patient was lost to follow-up. The treatment was found to be effective in 16 of 25 patients at 360 days after grafting. Of the 23 patients who completed follow-up at 360 days, 22 had no ulcers, and 19 showed a decrease in the severity of the punctate epithelial keratopathy.

Conclusions.: CAOMECS is a well-tolerated and safe tissue-engineered product. These results suggest its efficacy for reconstructing the ocular surface in patients with total bilateral corneal LSCD.

The corneal limbal epithelium is located at the transitional zone between the cornea and the bulbar conjunctiva. The limbal epithelial stem cells contained in the basal layer ensure the renewal of the corneal epithelium and serve as a barrier between the corneal and conjunctival epithelia. ^{1 –3} Injury to the corneal limbal epithelium, resulting from severe exogenous trauma such as thermal burns, chemical injuries, or endogenous eye disease, may cause partial or total limbal stem cell deficiency (LSCD). ^{4 –6} Total LSCD is a dysfunction or destruction of the entire limbal epithelial stem cell population, resulting in the progressive destruction of the corneal epithelium. ⁷ LSCD is characterized by conjunctival epithelial ingrowth, neovascularization, chronic inflammation, opacification, recurrent corneal erosion and persistent ulcers, the destruction of basement membrane components, and fibrous tissue ingrowth. It leads to functional impairment such as discomfort, chronic pain, photophobia, and dry eye, and may ultimately lead to visual impairment and even blindness. ^{2,8 –10}  

Autologous limbal graft (for application either directly or after ex vivo expansion) cannot be used in cases in which there has been bilateral damage to the patient's ocular surfaces by injury or disease. The alternative treatment is the transplantation of allogenic limbal epithelium in the form of a keratolimbal allograft from cadaveric tissue or a living HLA-matched donor. ^{11 –14} Allogenous limbal graft is not regarded as a standard treatment, ¹¹ since it requires systemic immunosuppression and has a success rate that tends to decrease gradually over time (graft survival rate of 40% at 1 year and 33% at 2 years). ¹⁵ Treatment with autologous oral mucosa grafted onto ocular surfaces ¹⁶ is considered to be safe, despite the risk of contamination. However, it is not effective in improving vision, probably because it contains opaque subepithelial fibrous tissue and it has a heterogeneous thickness. 

The CAOMECS is manufactured by using a novel temperature-responsive culture well, UpCell-Insert (CellSeed, Inc., Tokyo, Japan), ¹⁷ and is harvested without proteolytic processing, retaining cell-to-cell junctions as well as the deposited extracellular matrix of the basal membrane of the sheet. ^{18 –20} Moreover, it makes it possible to transplant the sheet with no support or substrates, which may cause opacity. Thus, the cell sheet is transparent, resistant, viable, and rapidly bioadhesive, which means it can be grafted onto the corneal stroma without suturing. CAOMECS transplantation is proposed for the treatment of patients with total bilateral LSCD with moderate or severe symptoms, for whom any other treatments are not applicable. The efficacy of CAOMECS transplantation has been suggested by the presence of epithelium replacement, both in nonclinical studies using rabbit LSCD models ^18,19 and in a clinical study including four patients with total bilateral LSCD with severe loss of vision (<1/10) in Japan. ²⁰ However, it was not possible to make a precise analysis of the risks and benefits from this preliminary study. 

We performed a clinical trial to confirm the safety and efficacy of CAOMECS with a prospective, noncomparative, two-stage Gehan design. 

Material and Methods

Design

This prospective, noncomparative and single institution study was designed according to the Gehan two-stage protocol. ²¹ The lowest limit of therapeutic activity considered to be of clinical importance was chosen as 30%, and the selected rejection error was 10%. In accordance with Gehan's two-stage design, seven patients were included in the first stage. 

The number of success 3 months after grafting during the preliminary phase was used to determine the number of patients needed for the follow-up phase (16–18, according to our initial hypothesis). 

Patients

The research adhered to the Declaration of Helsinki. Patients gave written informed consent before entering the study, which was approved by the local ethics committee. Patients 18 years of age or older were eligible for the study if they had total bilateral LSCD diagnosed by the same ophthalmologist with a slit lamp examination, as having a complete disappearance of the palisades of Vogt in the limbal area and a pathologic corneal epithelium, ²² with persistent defects, recurrent ulcers, chronic keratitis with scarring and punctate epithelial keratitis (PEK), and neovascularization with conjunctival invasion. Only one eye in each patient (the most severely affected) was grafted with a CAOMECS. 

Patients were not included if they presented with one of the following: acute systemic infection, history of an acute phase of ocular inflammation in the previous year, history of neoplastic disease, glaucoma (diagnosed by measuring intraocular pressure with air or applanation tonometer), total symblepharon (comprising eyelid aperture), a history of hypersensibility or allergy to antibiotics or serum, pregnancy, or infectious disease. 

Treatment Procedure

Harvesting and Culture of Oral Mucosal Epithelial Cells.

After sterilization of the oral cavity and local anesthesia, a 3 × 3-mm oral mucosa biopsy was excised from the cheek for epithelial cell culture as previously described. ²³ After enzymatic treatment, the cells were seeded on UpCell-Inserts (CellSeed, Inc.), placed in a six-well plate containing mitomycin C (MMC)–inactivated 3T3 fibroblasts as feeder layers, and cultured at least 4 days after confluence. Before grafting, the sheet was detached by reducing the temperature to <30°C and transferred to a PVDF ring. 

The risk of viral contamination was assessed by testing with a microbial detection system (BacT/Alert; bioMerieux, Marcy l'Etoile, France) on different media for HIV1 and -2, hepatitis B and C, herpes virus types 1 and 2, herpes virus and Mycoplasma, and sterility. 

The characterization of epithelial cells was performed as previously described ^22,23 using the K3/76 marker (anti-AE5 monoclonal antibody staining; Progen Biotechnik, Heidelberg, Germany). The presence of progenitor cells was tested for using anti-p63 (BD Pharmingen, Le Pont de Claix, France) and anti-β1-integrin (Santa Cruz, Heidelberg, Germany) stainings. All staining was performed using both immunocytology on parallel cell culture and immunohistology on the CAOMECS. The presence of a basement membrane was assessed with anti-laminin 5 staining (Millipore, Molsheim, France) on the CAOMECS. 

To assess colony forming efficiency (CFE), 20 or 40 cells/cm² were seeded on MMC-inactivated feeder layers. After cultivation for 12 days under the same conditions, CFE was determined as the ratio of the number of cells that produced large regular clones containing putative epithelial stem versus total seeded cells. ²⁴  

Transplantation of the Cell Sheets onto the Eye.

The conjunctival and subconjunctival scar tissue and all residual epithelium were removed from the patient's cornea up to 3 mm outside of the limbus, to re-expose the corneal stroma. After electrocoagulation was performed on all limbal vessels and the stroma was dried, the harvested CAOMECS was placed directly onto the exposed stromal bed (the total diameter of the graft was 20 mm). No sutures were required since the presence of basement membrane ensured immediate adhesion. The grafted corneal surface was immediately covered with a soft contact lens (Night and Day; Ciba Vision, Duluth, MN) to protect the CAOMECS during the first days. After surgery, topical antibiotics (0.3% ofloxacin) and corticosteroids (0.1% betamethasone) were initially applied four times a day and then three times a day. During the first week after surgery, betamethasone (4 mg per day) was administered orally to reduce postoperative inflammation. One month after surgery, the administration of topical corticosteroids was changed from 0.1% betamethasone to 0.1% fluorometholone. Since most patients had severe dry eye (due to defective tear production), preservative-free artificial tears were used at each patient's discretion throughout the study period. In cases of symblepharon, fornix reconstruction was always performed in patients before transplantation. 

Trial Outcomes

The primary endpoint for efficacy was measured at 360 days after grafting using the following criteria (the normal condition is italicized and underlined) by the same ophthalmologist (who was also the performing surgeon) through a direct observation using the slit lamp examination with and without fluorescein test:

•  
  Epithelial defect: no ulcer or mild (ulcer diameter <5 mm), moderate (diameter between 5 and 7 mm), or severe (diameter >7 mm) ulcer
•  
  PEK: mild, moderate, or severe
•  
  Conjunctival epithelium on cornea: absence or presence
•  
  Corneal vascularization: number of vascular pediculi (stem vessel branching out from the limbus)
•  
  Corneal vascularization, vessel activity: inactive (absence of blood), small-diameter vessels, different-diameter vessels, large-diameter vessels, and vasodilated vessels

The treatment for a patient was considered to be a success if all five criteria either improved or remained stable if normal at inclusion. The treatment was regarded as a failure if at least one criterion worsened or if one criterion that was subnormal at inclusion remained stable. 

The secondary endpoints were the patient's functional signs assessment (by interview with the ophthalmologist) covering photophobia (no complaint, fear of luminosity corrected by sunglasses, frequent winking to light, and inability to open the eye), watering (no complaint, dry, very dry, and difficulty opening the eye), and pain (no complaint, occasionally, repetitive pain during the day, and permanent); the assessment of clinical efficacy based on physical signs at the slit lamp examination (corneal opacification from clear to severe dense opacity); and evaluation of visual acuity according to the ability to see hand motion (at 20, 50, and 100 cm), ability to count fingers (at 20, 50, and 100 cm), and Snellen's scale (<1/20, 1/20, 1/10, and 10/10). 

Safety was assessed by documenting ocular adverse events that occurred during the patients' follow-up. 

Follow-up

Patients were followed up for 1 year after the CAOMECS graft, with seven medical visits scheduled: after surgery and 15, 30, 60, 90, 180 and 360 days after transplantation. 

Statistics

Two populations were defined. The safety population was defined as all patients receiving the graft. The efficacy population was defined as all patients with the evaluation criterion available at 360 days; patients with serious adverse events (SAEs) were recorded as having failed grafts. The analysis was descriptive. The success rate is presented with the corresponding 95% score confidence interval. ²⁵  

Results

A total of 25 patients were initially included, one patient was lost to follow-up after 30 days and was replaced, giving a total of 26 registered and treated patients (26 eyes were treated, but one patient was included twice, because he was treated in one eye and subsequently in the other eye). All patients received the CAOMECS graft between October 29, 2007, and June 29, 2009. The median age of the patients included in the study was 52 years. The other characteristics and medical histories of the patients are presented in Table 1. Two patients had simultaneous CAOMECS and corneal stromal grafts. 

Table 1.

 

Demographic Characteristics and Medical History

Patient	Sex	Age at CAOMECS Graft (y)	Grafted Eye	Initial Disease	Surgery before CAOMECS
001	M	48	RE	Corneal burn	CG
002	F	38	RE	Neuroparalytic keratitis	None
003	F	69	RE	Rosacea keratitis	None
004	M	40	LE	Corneal burn	AM
005	F	34	LE	Lyell syndrome	LK + AM
006	M	70	RE	Corneal burn	LK
007	F	54	LE	Corneal burn	AM + CG
008	M	61	RE	Severe trachoma	CG
009	M	41	LE	Lyell syndrome	LK
010	M	80	RE	Rosacea keratitis	CG
011	M	59	LE	Corneal burn	CG
012	M	38	LE	Corneal burn	LK + AM
013	M	38	RE	Corneal burn	AM + PS
014	M	53	LE	Corneal burn	AM + PS
015	F	51	RE	Contact lens hypoxia	CS
016	F	71	LE	Lyell syndrome	AM
017	M	50	LE	Lyell syndrome	AM
018	M	60	RE	Congenital aniridia	CS
				Contact lens hypoxia
019	F	25	RE	Cystinosis	None
020	M	69	LE	Severe Groenow dystrophy	CG
021	M	46	RE	Hepatitis c	AM
022	F	40	LE	Neuroparalytic keratitis	LK
023	F	33	RE	Congenital aniridia	None
024	F	54	RE	Congenital aniridia	CS
				Contact lens hypoxia
025	M	59	LE	Rosacea keratitis	LK + AM
026	M	55	RE	Corneal burn	AM + PS

 AM, amniotic membrane; CG, corneal graft; CS, cataract surgery; LK, lamellar keratoplasty; PS, palpebral surgery.

Characterization of Tissue-Engineered Epithelial Cell Sheets

Transplantable and transparent oral mucosal cell sheets (Fig. 1A) were produced after culturing the patient's oral mucosal cells for approximately 3 weeks. Expressions of K3/76 were observed in the cells (Fig. 1 C1) and cell sheets (Fig. 1 C2) of all patients. These data suggest that all transplanted cell sheets derived from oral mucosal epithelial cells showed similar morphologic characteristics to normal corneal epithelium. 

Figure 1.

 

CAOMECS quality controls. Evaluation of the transparency, which accommodated Arial letter size 10 (A). Appearance of the colonies in the colony-forming assay (B). Immunological controls on cells (left) and the final product (right) using a monoclonal antibody for keratin 3/76 as the identity marker (C1, C2); for the evaluation of putative stem cells, anti-p63 (D1, D2) and anti-β1-integrin staining (E) and, as basement membrane marker, anti-laminin 5 staining (F) were used. Scale bar, 5 μm.

The expression of p63 and β1-integrin as markers of putative epithelial stem cells (Figs. D1, D2, E) as well as the CFE of 3.4% ± 2.06% with the presence of holoclones (Fig. 1B) demonstrate the presence of putative stem cells. 

The extracellular expression of laminin 5 verified the preservation of the extracellular matrix component of the basal membrane (Fig. 1F). 

Safety

Overall, seven patients had ocular adverse events (Table 2). Two of these events were serious. The first SAE was related to an inflammatory event. One month after the CAOMECS graft, inflammation persisted and neovascularization appeared. This patient presented a corneal perforation at 7 months after grafting. However, this event was due to the removal of the conjunctival and subconjunctival scar tissue and was therefore classified as “not related to the experimental product. ” Moreover, the untreated eye also presented with perforation. The second SAE occurred in a patient who experienced a massive stroma graft rejection 4.5 months after the simultaneous transplantation of CAOMECS and corneal stromal grafts (the immunosuppressive treatment consisted of local corticosteroids). An ulcer over the whole donor corneal graft surface with strong vascularization required amniotic membrane surgery. 

Table 2.

 

Adverse Ocular Events

Patient	Description	Start Date*	End Date*
004	Pain and corneal recurrence	26/05/2008	12/11/2008
005	Meibomian cyst	07/01/2008	21/01/2008
005	Eye inflammation	17/12/2007	—
005	Corneal perforation†	20/06/2008	23/06/2008
011	Increased intraocular pressure	15/12/2008	16/12/2008
019	Keratitis	31/08/2009	07/12/2009
019	Amniotic membrane graft	02/11/2009	02/11/2009
020	Conjunctival operation	15/01/2010	18/01/2010
024	Corneal graft rejection†	06/10/2009	—
025	Inflammation	17/08/2009	01/09/2009
025	Pain and graft complication	14/01/2010	05/02/2010
025	Symblepharon	17/05/2010	—

*  Dates are dd/mm/yyyy.

†  Serious event leading to patient withdrawal from the study.

The adverse events in the other five patients were due to hypoxia from wearing contact lenses for too long or an inflammatory reaction. For two patients, it was also due to an associated pathology: hypertension caused by corticosteroid treatment resulting in glaucoma in one case and optical nerve atrophy in the other. 

Efficacy

Intermediate Analysis.

According to Gehan's design, an intermediate analysis was performed at 3 months after grafting for the first seven graft recipients (Table 3). 

Table 3.

 

Details of the Evaluation Items for the 25 Patients at Inclusion, 90 Days, and 360 Days

 

The results of the intermediate analysis rejected the inefficacy of the treatment with CAOMECS, with four successes observed. The patients were observed for a further 9 months, and, in line with Gehan's method, 18 more patients were included in the study and followed for 1 year. 

Final Analysis.

Twenty-five patients were evaluated for the main criteria (Table 3). The two patients who experienced SAEs were not assessed at 360 days after grafting, and their grafts were regarded as failures. The trial outcome at 360 days after grafting was regarded as a success in 16 (75%) of 25 patients (score 95%CI, 45–80). The grafts in the two patients with simultaneous grafting were failures. When considering the 23 patients assessed at 360 days after grafting, the number of improved patients was 19 for PEK (example in Figs. 2C, D), 22 for the number of vascular pediculi, and 20 for vessel activity. Concerning persistent epithelial defect, 11 patients of the 12 with ulcers at inclusion were improved (example in Figs. 2A, B). Eleven patients of the 15 with conjunctival epithelium on the cornea at inclusion were improved. The evolution of corneal vascularization is shown in Figure 3 (before grafting, A, C, E, and after, B, D, F). 

Figure 2.

 

Evaluation by the ophthalmologist using a slit lamp with fluorescein. Photographs of two patients' eyes, both taken with one drop of 0.5% fluorescein after four blinks. (A) Chronic corneal ulcer and (B) re-epithelialization of the chronic ulcer 360 days after CAOMECS graft; the surface of the cornea epithelium is irregular, which explains the stasis of fluorescein in some places. (C) Superficial punctate keratitis and (D) reduction in the keratitis 360 days after CAOMECS.

Figure 3.

 

Evaluation by the ophthalmologist using a slit lamp examination without fluorescein. Photographs of three patients' eyes were obtained. In all the patients, before grafting (A, C, E), the corneas were opaque with severe neovascularization. At 360 days after transplantation (B, D, F), the corneas became clearer with a regression of the neovessels. (B) The opacity was due to the stromal scar; (D, F) the corneas appeared more transparent.

Regarding the secondary evaluation criteria, the two patients who experienced SAEs were not assessed at 360 days after grafting; therefore, the following results include 23 patients. 

Concerning photophobia, watering, and pain, 22 patients showed improvements in at least one criterion, and 7 showed improvements on all three criteria (Table 4). Seventeen patients improved in visual acuity (Table 4). 

Table 4.

 

Evolution in Functional Signs and Visual Acuity between Inclusion and Day 360

	Photophobia			Watering			Pain			Visual Acuity
	Inclusion	Day 360	Evolution	Inclusion	Day 360	Evolution	Inclusion	Day 360	Evolution	Inclusion	Day 360	Evolution
001	No complaint	No complaint	Steady	Very dry	No complaint	Improve	Occasionally	No complaint	Improve	CF50	CF100	Improve
002	Fear of luminosity	No complaint	Improve	Very dry	No complaint	Improve	Occasionally	No complaint	Improve	CF100	3/10	Improve
003	Fear of luminosity	Fear of luminosity	Steady	Very dry	No complaint	Improve	Occasionally	No complaint	Improve	HM20	CF100	Improve
004	Frequent, winking to light	Frequent winking to light	Steady	Dry	No complaint	Improve	No complaint	No complaint	Steady	CF100	1/10	Improve
006	Fear of luminosity	Fear of luminosity	Steady	Dry	No complaint	Improve	Occasionally	No complaint	Improve	1/10	CF100	Worsen
007	Frequent winking to light	Fear of luminosity	Improve	Very dry	No complaint	Improve	Occasionally	No complaint	Improve	CF20	CF50	Improve
008	Frequent winking to light	Fear of luminosity	Improve	Very dry	No complaint	Improve	No complaint	No complaint	Steady	CF50	CF100	Improve
009	Fear of luminosity	No complaint	Improve	Very dry	No complaint	Improve	No complaint	No complaint	Steady	CF100	2/10	Improve
010	Fear of luminosity	No complaint	Improve	Dry	No complaint	Improve	No complaint	No complaint	Steady	HM50	CF100	Improve
011	Frequent winking to light	No complaint	Improve	Very dry	No complaint	Improve	No complaint	No complaint	Steady	CF100	1/20	Improve
012	Frequent winking to light	No complaint	Improve	No complaint	Dry	Worsen	Occasionally	No complaint	Improve	1/10	1/10	Steady
014	Fear of luminosity	No complaint	Improve	Dry	Dry	Steady	Occasionally	No complaint	Improve	HM100	HM50	Worsen
015	Frequent winking to light	No complaint	Improve	Very dry	No complaint	Improve	Occasionally	No complaint	Improve	1/10	2.5/10	Improve
016	Inability to open eye	Fear of luminosity	Improve	Dry	No complaint	Improve	Occasionally	No complaint	Improve	HM100	CF20	Improve
017	No complaint	No complaint	Steady	Difficulty opening the eye	No complaint	Improve	No complaint	No complaint	Steady	CF100	1/10	Improve
018	Frequent winking to light	Fear of luminosity	Improve	Dry	No complaint	Improve	Occasionally	No complaint	Improve	1.4/10	CF100	Worsen
019	Frequent winking to light	No complaint	Improve	Dry	No complaint	Improve	Repetitive pain	No complaint	Improve	HM50	HM20	Worsen
020	Fear of luminosity	No complaint	Improve	Dry	No complaint	Improve	No complaint	No complaint	Steady	HM100	CF20	Improve
021	Fear of luminosity	No complaint	Improve	No complaint	Dry	Worsen	Occasionally	No complaint	Improve	1/10	2/10	Improve
022	Frequent winking to light	No complaint	Improve	Very dry	No complaint	Improve	No complaint	No complaint	Steady	3/10	4/10	Improve
023	Frequent winking to light	No complaint	Improve	Dry	No complaint	Improve	Occasionally	No complaint	Improve	CF50	CF50	Steady
025	Frequent winking to light	Frequent winking to light	Steady	Dry	Dry	Steady	Occasionally	No complaint	Improve	HM20	HM100	Improve
026	Frequent winking to light	Frequent winking to light	Steady	No complaint	No complaint	Steady	Occasionally	Occasionally	Steady	HM100	CF50	Improve

 Photophobia scale: no complaint, fear of luminosity corrected by sunglasses, frequent winking to light, inability to open eye. Watering scale: no complaint, dry, very dry, difficulty opening the eye. Pain scale: no complaint, occasionally, repetitive pain during the day, permanent. CF: count fingers (20, 50, and 100 cm); HM: hand motion (20, 50, and 100 cm).

Concerning corneal opacification, 12 patients had moderately dense opacity partially obscuring details of the iris at inclusion: 9 remained stable and 3 moderated (decrease in the initial opacity); 9 had mild opacity (2 worsened and 7 remained stable), 1 had severe dense opacity that moderated and the last patient had a clear cornea or a trace of haze and remained stable. 

Discussion

In this study, the CAOMECS was regarded as a success based on a composite criterion in 16 (64%) of 25 patients assessed at 360 days after grafting. We also found that 22 of the 23 patients experienced improvements in photophobia, watering, or pain and that visual acuity improved in 17. The CAOMECS was also safe; two SAEs occurred but were not related to the product. 

In 5 of 23 patients, there was reconstruction of the corneal epithelium without ulcers or neovessels. Our results demonstrate that the CAOMECS can restore cornea epithelial function while repressing neovascularization and conjunctivalization and improving visual acuity in cases in which the stroma was sufficiently healthy. For the patients whose stromas were heavily damaged, the CAOMECS transplantation reduced the neovascularization process. 

Concerning vascularization, the expected invasion by active vessels initially eliminated by limbal electrocoagulation did not occur. After 1 year, there was a notable decrease in neovessels in the corneas of all patients, although in nine patients there was a partial recurrence of neovessels on a localized part of the cornea where there was no apparent loss of the CAOMECS. The reason has not been identified at this stage; however, it could be hypoxia (wearing extended-wear contact lenses for excessive lengths of time), high levels of dryness, and/or poor quality of the stromal bed. Previous studies on oral mucosal epithelial sheet transplantation also indicated the partial recurrence of neovessels. ^20,26 Thus, the biological activity of the oral epithelial sheet intended to inhibit neovascularization may be somewhat less effective than in real corneal epithelium. The lacrimal problem required most patients to use artificial tears. 

The new temperature-responsive UpCell-Insert technology (CellSeed, Inc.) ²⁷ makes it possible to harvest CAOMECS without enzymatic treatment, preserving cell–cell junctions and the basement membrane component, such as laminin 5, and giving rise to a cohesive sheet that can be prepared without a carrier, such as fibrin gel and amniotic membranes. Basal membrane protein preservation verified by the laminin 5 extracellular deposition allows for easy transplantation without suturing (in contrast with other culture techniques), and immediate adhesion between the CAOMECS and stroma. ²⁰ The absence of suturing is the most innovative part of this technique. Sutures are known to induce pain, discomfort, and inflammation and to have an impact on the patient's quality of life. In this study, we found that the criteria regarding patient comfort improved after the graft. 

Considering that all grafts had at least a 1-year follow-up and knowing that the lifespan of transient amplifying cells is believed to be less than 1 year, ²⁸ we suggest that CAOMECS contains stem cells, as demonstrated by the CFE at 3.4%. These putative stem cells ensure the constant renewal of corneal epithelium after CAOMECS transplantation, required to avoid conjunctivalization and neovascularization. 

The main limitation of this study is the absence of a control group. However, Gehan's two-stage design was adapted to this particular case in which the patients did not have any alternative therapeutic solution and in which no natural positive evolution was expected. We chose to select patients according to etiology and length of evolution (selection of pathologies known to destroy limbal epithelium) and confirmed by a clinical and slit lamp examination. We did not perform impression cytology to ascertain total LSCD, since it had not been approved by the time the study was initiated. Indeed, at the beginning of the clinical trial commonly used markers had no sufficient predictive value: for instance, the detection of conjunctival goblet cells using conventional cell cytology collected by means of impression cytology of the corneal epithelium ^4,8 did not seem to us a good diagnostic criterion, given the high variability in the density of goblet cells, which may even disappear from the ocular surface for certain LSCD etiologies. ^{29 –32} The fact that the main outcome was measured by a single ophthalmologist who was not masked could be regarded as a weakness of the study. However, the clinical signs were evaluated by one of the few French specialists in the cornea and LSCD disease, thus ensuring good grading quality. 

In conclusion, we demonstrated that CAOMECS is a well-tolerated and safe tissue-engineered product for the treatment of human LSCD. CAOMECS reconstructed the patients' ocular surface (i.e., new epithelium and less vascularization or less vessel activity), resulting in better visual acuity in cases of healthy stroma. CAOMECS should also be beneficial in patients with total bilateral LSCD, whose corneal stroma is severely damaged and cannot receive a secondary graft (donor cornea) due to the presence of neovessels, which would hinder its success. A decrease in the rejection risk for a secondary donor cornea graft is expected in recipients of a CAOMECS transplantation after 1 year. This was the case in 11 patients, who are currently in follow-up for assessment. 

Footnotes

 Supported by CellSeed, Inc., Tokyo, Japan. The funder was not involved at any stage of the clinical protocol or during the study period, data management, and statistical analysis phases.

Footnotes

 Disclosure: C. Burillon, CellSeed, Inc. (F); L. Huot, CellSeed, Inc. (F); V. Justin, CellSeed, Inc. (F); S. Nataf, CellSeed, Inc. (F); F. Chapuis, CellSeed, Inc. (F); E. Decullier, CellSeed, Inc. (F); O. Damour, CellSeed, Inc. (F)

The authors thank Hospices Civils de Lyon for sponsoring the study, CellSeed for providing the UpCell Insert, Morgan Dos Santos for laminin 5 labeling, and Kim Barrett for help with the English text.