Stem Cell Transplantation for Ocular Surface Diseases
SCHEFFER C.G. TSENG, EDGAR M. ESPANA and MARIO A. DI PASCUALE
Table Of Contents
TWO TYPES OF OCULAR SURFACE FAILURE
LIMBAL STEM CELL DEFICIENCY
OCULAR SURFACE HEALTH IS MAINTAINED BY A STABLE TEAR FILM
Anatomically, the ocular surface encompasses the entire mucosal epithelial lining bordered by the skin at the superior and inferior eyelid margins. Histologically, this epithelial surface covers two major territories (i.e., the cornea and the conjunctiva). The primary function of the ocular surface is to provide clear vision during an open-eye state. To achieve this goal while maintaining comfort and preventing microbial invasion, the ocular surface has to be covered by a stable tear film. Therefore, the mechanism by which the ocular surface health is ensured is inherently built into the relationship between ocular surface epithelia and the preocular tear film. As recently summarized,1 compositional factors including the ocular surface epithelia, meibomian glands, and lacrimal glands and hydrodynamic factors such as eyelid blinking and closure are essential to constitute ocular surface defense that governs how preocular tear film stability is maintained (Fig. 1). These two major factors are integrated via two neuronal reflex arcs, triggered by the sensory drive of the first branch of trigeminal nerve (V1), and mediated by the parasympathetic branch and the motor branch of the facial nerve (VII) as the efferent output, respectively (Fig. 2). Such a neuroanatomic integration of ocular surface defense explains how external adnexal glands and eyelids can be integrated with ocular surface epithelia as a unit to maintain a stable tear film.
Based on this concept, one predicts that deficiency or alteration of any of these elements involved in such neuroanatomic integration can lead to an unstable tear film, a hallmark of various forms of dry eye.2 Therefore, investigation into dysfunction of this neuroanatomic integration is the first step toward better understanding of the pathogenesis of various ocular surface disorders, and is a prerequisite before any ocular surface reconstruction.
|TWO TYPES OF OCULAR SURFACE FAILURE|
According to the resultant epithelial phenotype defined by impression cytology 3, all severe ocular surface diseases can be classified to express one of the two major types of ocular surface failure. The first type is squamous metaplasia (Fig. 3) in which a normal nonkeratinized ocular surface epithelium transforms into a skin-like keratinized epithelium.3 Because of skin-like epithelial differentiation (skin-like), the ocular surface with squamous metaplasia becomes nonwettable, a hallmark of various dry eye disorders.2 If the underlying insults are primarily derived from poor ocular surface defense, such squamous metaplasia can be reversed to a normal state when ocular surface defense is restored, indicating that epithelial progenitor cells (i.e., stem cells) of the ocular surface are not intrinsically altered. Squamous metaplasia can be caused by insults not originating from ocular surface defense. In xerophthalmia caused by systemic vitamin A deficiency,4,5 squamous metaplasia can also be reversed by replacement of vitamin A. Nevertheless, squamous metaplasia can also be caused by various insults damaging epithelial progenitor cells, the basement membrane, and the underlying stroma leading to chronic inflammation and scarring as seen in various forms of cicatricial keratoconjunctivitis such as chemical burns, Stevens-Johnson syndrome, ocular pemphigoid, etc.6,7 Except for early active ocular cicatricial pemphigoid, in which systemic immunosuppression can reverse squamous metaplasia, the rest cannot be reversed by conventional medical treatments. The irreversibility of such abnormal epithelial differentiation strongly implies that epithelial progenitor cells (i.e., stem cells [SC]), may have been intrinsically altered.
|LIMBAL STEM CELL DEFICIENCY|
|The second type of ocular surface failure is limbal stem cell deficiency (LSCD) (Fig. 3), which is characterized by the replacement of the normal corneal
epithelium by an invading conjunctival epithelium.8 This pathologic state of LSCD can experimentally be created in rabbits
by damaging the limbal epithelium, which contains corneal epithelial
SC.9 Corneas with LSCD manifest poor epithelialization (persistent defects
or recurrent erosions), chronic stromal inflammation (keratitis
mixed with scarring), corneal vascularization, and conjunctival
epithelial ingrowth. Consequently, patients with LSCD experience
severe irritation, photophobia, and decreased vision, and are poor
candidates for conventional corneal transplantation. Because most of these
features can also be found in other corneal diseases, the sine qua non criterion for diagnosing LSCD is the existence of conjunctival epithelial ingrowth
onto the corneal surface (conjunctivalization). Clinically, the
presence of conjunctivalization may be suggested by observing the
loss of limbal Vogt's palisades under slit-lamp examination10 and by occurrence of late fluorescein staining,11 reflecting poor epithelial barrier function.12 However, the definitive diagnosis of conjunctivalization relies on the
detection of conjunctival goblet cells on the corneal surface by impression
cytology.8 Accurate diagnosis of LSCD is crucial to guide the choice of appropriate
procedures for transplanting limbal epithelial SC.|
Based on the underlying etiology, corneal diseases manifesting LSCD can be subdivided into two major categories (Table 1).8 In the first category, limbal epithelial SC are destroyed by known or recognizable offenders such as a chemical or thermal burn, Stevens-Johnson syndrome/toxic epidermal necrolysis, multiple surgeries or cryotherapies or medications (iatrogenic), contact lens, severe microbial infection, radiation, and antimetabolites including 5-fluorouracil and mitomycin C.8,13–15 A second category is characterized by a gradual loss of the SC population without known or identifiable precipitating factors. In this situation, the limbal stromal niche is presumably affected and progressively deteriorates by a variety of etiologies that include aniridia and coloboma, neoplasia, multiple hormonal deficiencies, peripheral ulcerative corneal diseases, neurotrophic keratopathy and idiopathic limbal deficiency.8,10,16–18 These diseases can also be categorized by the underlying cause being hereditary or not as summarized in Table 1. The aforementioned information explains why transplantation of epithelial SC and restoration of SC stromal environment is necessary in ocular surface reconstruction.
|Considering all of the basic concepts together, we propose that the general
strategy of ocular surface reconstruction is analogous to growing
a flowering plant in the garden (Fig. 4). Several key strategies are undertaken: (1) to restore
ocular surface defense so that a stable tear film can be maintained, analogous
to providing rain/water; (2) to restore stromal
substrate that is important to support existing or transplanted epithelial
SC, analogous to providing a top soil; and (3) to restore
epithelial SC, analogous to providing a good seed. These strategies
are similarly applied to corneal and conjunctival surface reconstruction. Two
major threats to successful ocular surface reconstruction are
severe deficiencies in ocular surface defense leading to the lack of
tear film protection and chronic inflammation or ischemia in the stroma
that continue to damage the stromal substrate.|
FIRST STRATEGY: RESTORE OCULAR SURFACE DEFENSE
Before reconstruction, one should realize that keratinization in squamous metaplasia and vascularization in LSCD are two important pathologic mechanisms to prevent such corneal complications as epithelial defects, ulcers, or melt in eyes without an unstable tear film. Therefore, one should avoid performing conventional corneal transplantation to correct these two types of corneal failures if the ocular surface defense cannot first be restored. Because ocular surface reconstruction aims to restore a clear avascular cornea, which will require a stable tear film to maintain its health (see “Basic Concepts” section), it is of utmost important to restore a sound ocular surface defense. If not restored or unable to be restored, poor ocular surface defense in these eyes represent a contraindication for corneal surface reconstruction with epithelial SC transplantation.
Deficiencies in ocular surface defense can be recognized by history taking, and external and biomicroscopic examinations, and by the use of special tests such as dynamic tear function test, dye staining and impression cytology. Measures taken to correct severe aqueous tear deficiency dry eye include punctal occlusion and frequent application of autologous serum drops for mechanical microtrauma caused by lid margin and lash abnormalities such as trichiasis, entropion, or meibomian gland orifice metaplasia, which can be corrected by high oxygen permeability (Dk) contact lens, scleral contact lens, or appropriate plastic surgeries including mucous membrane graft. Symblepharon that causes obliteration of tear meniscus, misdirected lashes, inability of adequate blinking and closure, and motility restriction should also be corrected by fornix reconstruction before corneal surface reconstruction. Exposure problems may be corrected by Botox-induced ptosis, high DK contact lens, scleral lens, or tarsorrhaphy.
SECOND STRATEGY: RESTORE STROMAL ENVIRONMENT
The state and severity of LSCD can be graded as partial or total. In partial LSCD, a portion of the limbal SC remains intact. In partial or total LSCD, if the central cornea still preserves a normal corneal epithelial phenotype and satisfactory vision, treatments should be directed to maintaining the remaining corneal epithelium and to activating the remaining limbal SC. Besides bandage contact lens or scleral lens to shield against continuous attrition, one new way of expanding remaining limbal or conjunctival epithelial SC in vivo is to transplant preserved amniotic membrane to restore stromal environment that has been destroyed in diseases mentioned in Table 1.
Amniotic membrane (i.e., the innermost layer of the placental membrane) consists of a simple epithelium, a thick basement membrane, and an avascular stroma. When appropriately procured and processed, cryopreserved amniotic membrane can be used as a surgical graft for substrate replacement and does not elicit immunologic reactions.19 A number of studies have shown that amniotic basement membrane facilitates epithelialization and maintains epithelial phenotype, while amniotic avascular stroma exerts anti-inflammatory, antiangiogenic and antiscarring effects [for reviews see Dua and Azuara-Blanco,20 Kruse et al.,21 Sippel et al.,22 and Tseng and Tsubota23), properties important for reconstruction of epithelium-lining tissues. For partial LSCD, debridement of conjunctivalized epithelium from the corneal surface24 with or without additional amniotic membrane transplantation25 is effective to restore the corneal surface. Because of its anti-inflammatory effect, amniotic membrane transplantation as a temporary patch (bandage) is also effective in restoring ocular surfaces during the acute attack of chemical burns26 and Stevens-Johnson syndrome27.
THIRD STRATEGY: RESTORE EPITHELIAL STEM CELLS
For corneal surface reconstruction, epithelial SC can be transplanted from the limbal epithelium. For conjunctival surface reconstruction, epithelial SC can be transplanted from the bulbar conjunctival epithelium, especially close to the fornix, where conjunctival epithelial SC are enriched.28 The following section is devoted to restoring corneal surfaces with LSCD.
When LSCD is total, that is, involving the entire limbal SC of an eye, amniotic transplantation alone is ineffective, and corneal surface reconstruction requires transplantation of limbal epithelial SC. According to the type and source of tissue removed to transplant limbal epithelial SC, several surgical procedures have been devised, of which each terminology used herein adopts what has been recommended by Holland and Schwartz.29
Conjunctival Limbal Autograft
When total LSCD is unilateral, conjunctival limbal autograft (CLAU) from the uninvolved fellow eye is advised.30 The surgical procedure for CLAU is shown in Figure 5. (To view surgical video, please contact Dr. Tseng at firstname.lastname@example.org.) In brief, the conjunctivalized pannus is removed from the corneal surface by peritomy followed by superficial keratectomy with blunt dissection in the recipient eye (Fig. 5A). The cicatrix was removed from the subconjunctival space (Fig. 5B). This invariably results in the recession of the conjunctival edge to 3 to 5 mm from the limbus (Fig. 5C). Two strips of limbal conjunctival-free grafts, each spanning 6 to 7 mm limbal arc length, are removed by superficial lamellar keratectomy at 1 mm within the limbus from the superior and inferior limbal regions (Fig. 5D) and by including 5 mm of adjacent conjunctiva. These two free grafts are transferred and secured to the recipient eye at the corresponding anatomic sites by interrupted 10-0 nylon sutures to the limbus and 8-0 vicryl sutures to the sclera. The size of the limbal removal in CLAU can be adjusted according to the visual potential of the donor eye and the extent of LSCD in the recipient eye, and the amount of the conjunctiva can be increased if the recipient eye also requires symblepharon lysis. Figure 6 illustrates how such a case with unilateral acid burn improved visual acuity, and maintained a smooth and stable corneal surface without vascularization and a clearer corneal stroma following CLAU (depicted in Fig. 5) without corneal transplantation.
In chemical burns, severe inflammation and ischemia in the acute stage is a threat against the success of transplanted CLAU, as noted in a number of studies,30,31 For this reason, CLAU is not recommended. Nevertheless, as aforementioned, transplantation of amniotic membrane as a temporary patch is an alternative because it can suppress the intense inflammation, facilitate epithelial wound healing, and prevent scrring in acute burns26 and Stevens-Johnson syndrome/toxic epidermal necrolysis.27 Although it is generally believed that the donor eyes with such limbal removal recover well without complication (Fig. 6H), scattered reports show that some donor eyes may become decompensated with pseudopterygium or partial LSCD, especially in eyes with subclinical LSCD. Figure 7 shows one such case. To preclude such a potential complication, one alternative is to transplant amniotic membrane as a graft to cover the defect after removal of CLAU in the donor eye and over the corneal surface before CLAU in the recipient eye so that the remaining and transplanted limbal epithelial SC can be expanded in the donor and recipient eye, respectively.32 Figure 8 illustrates how amniotic membrane transplantation can resurrect the corneal surface even after the removal over 11 clock hours of the limbus in the donor eye, and how both donor and recipient eyes can be helped after CLAU.
Living-Related Conjunctival Limbal Allograft Transplant
When total LSCD is bilateral, corneal surface reconstruction relies on transplantation of allogeneic limbal epithelial SC. To do so, one option is to living-related conjuctival limbal allograft transplant (lr-CLAL). The surgical procedure of lr-CLAL is identical to CLAU. Amniotic membrane can be used similarly to eliminate the concern of removing limbal SC from the healthy donor eye and to augment the effect of CLAU in the recipient eye. Nevertheless, unless the donor and the recipient are perfectly matched, the success of lr-CLAL depends on systemic immunosuppression and allograft rejection is still the main threat.33
Keratolimbal Allograft Transplant
The other option is to perform keratolimbal allograft (KLAL) from cadaveric donors. This surgical technique may restore SC in patients with bilateral LSCD or in patients with unilateral LSCD who do not wish to jeopardize the healthy eye with any surgical procedure including CLAU. Because of allogeneic transplantation, it is mandatory to administer systemic immunosuppression in the same manner as in lr-CLAL. Despite continuous oral administration of cyclosporin A (CSA), Tsubota et al.34 and Solomon et al.35reported that the long-term success of KLAL is approximately 40% to 50% in 3 to 5 years, while Ikari and Daya36 reported a 21.2% success rate in 5 years of follow-up. The lower success rate may be contributed by serveral limiting factors. The first limiting factor is that the ocular surface defense is not fully restored in some patients (see “First Strategy” section). This concept is suggested in a study showing that severe aqueous tear deficiency dry eye is a major cause leading to the failure,37 and in another study showing that uncorrected lid abnormalities are associated with failure.35 The second limiting factor is that oral CSA alone may not be sufficient to achieve adequate immunosuppression. As recently reported,34–36,38,39 among all diseases with total LSCD, Stevens-Johnson syndrome/toxic epidermal necrolysis has the worst prognosis when treated with KLAL or lr-CLAL. One reason for such a poor prognosis is the presence of chronic inflammation, which may enhance sensitization leading to allograft rejection. Indeed, ocular surface inflammation is an important clinical finding used to grade the severity of ocular surface diseases.40 Amniotic membrane transplantation has been used to augment the success of KLAL based on the fact that it may help suppress inflammation, and restore the damaged limbal stromal environment.38,41,42 Allograft rejection remains to be the most important cause limiting the success of KLAL. Signs of allograft rejection include telangiectatic and engorged limbal blood vessels, epithelial rejection lines and epithelial breakdown in severe limbal inflammation (Fig. 9). We thus believe that a combination of several immunosuppressive agents will have to be administered for lr-CLAL or KLAL in a manner similar to what is used in other solid organ transplantations.43 Table 2 describes one such protocol of immunosuppressive regimen and monitoring. It is advised that patients who are to receive such immunosuppressive regimen be screened and monitored by qualified internists, hematologists, rheumatologists, or oncologists, and be informed about all potential side effects. Furthermore, serologic screening of prior DNA viruses, and prophylactic use of oral Acyclovir are also important. As the field of immunosuppression evolves rapidly, we believe that there are other protocols that may improve the outcome of KLAL in the future. We also believe new techniques to suppress stromal inflammation are needed in the future. These new strategies of KLAL have been summarized in a recent review.44
*Acyclovir 200 mg 5/d PO is used as a prophylactic measure against viral activation starting three days before surgery.
CBC, complete blood cell count; PO, orally.
The surgical procedure of KLAL is schematically depicted in Figure 10 (To view surgical video, please contact Dr. Tseng at email@example.com). It should be noted that after the pannus is removed, the remaining corneal stroma is frequently clear especially in those eyes without ulceration and endothelial dysfunction (Fig. 11), resulting in significant improvement of vision without corneal transplantation. Nevertheless, if the remaining corneal stroma still has a deep scar, visual rehabilitation will require additional corneal transplantation (Fig. 12). If corneal transplantation will have to be performed, it is best to do so as a separate procedure 3 or 4 months later when the eye is not as inflamed (Fig. 12) because there is a high rejection rate for corneal transplants when combined with KLAL at the same setting.35,36,42,45 To avoid this complication, another solution is to perform deep lamellar keratoplasty rather than penetrating keratoplasty especially if there is no corneal endothelial dysfunction.
Ex Vivo Expansion of Limbal Stem Cells
Another new procedure of transplanting limbal epithelial SC is via ex vivo expansion of limbal SC. This technique, first demonstrated by Pellegrini et al.46,47 using 3T3 fibroblast feeder layer, can also be achieved by amniotic membrane with or without 3T3 fibroblast feeder layers for autologous48–51 or allogeneic52,53 limbal SC transplantation for treating total LSCD. The surgical procedure is schematically shown in Figure 13. A representative case is shown in Figure 14, which has been shown to result in a limbal epithelial phenotype on the corneal surface.51 The theoretical advantage over the autologous limbal SC transplantation (i.e., CLAU) or living-related allogeneic limbal SC transplantation (i.e., lr-CLAL) is that only a small limbal biopsy is needed, thus minimizing the risk to the donor eye. The theoretical advantage over the allogeneic limbal SC transplantation (i.e., KLAL or lr-CLAL) is that the allograft rejection might be reduced because only epi-thelial cells are transplanted, and antigen-presenting Langerhan's cells are eliminated during ex vivo expansion.
Unlike transplanting amniotic membrane, which is classified by the Food and Drug Administration (FDA) as a tissue, and hence does not require premarket approval for its use in ocular surface reconstruction, ex vivo expansion needs FDA approval for clinical uses in the United States. Dr. Tseng has obtained an Investigation New Drug (IND) No. 10313 application from the FDA to conduct a phase 1 clinical trial. Once the safety and efficacy is demonstrated in clinical trials, one may envision this new procedure may one day improve the outcome of ocular surface reconstruction.
|The basic science portion of this study was supported by Public Health Service Research Grant EY06819 (to S.C.G.T. via TissueTech, Inc.) from the Department of Health and Human Services, National Eye Institute, National Institute of Health, Bethesda, Maryland; preparation of this manuscript is supported by a grant from Ocular Surface Research & Education Foundation.|
9. Schermer A, Galvin S, Sun T-T: Differentiation-related expression of a major 64K corneal keratin in vivo and in culture suggests limbal location of corneal epithelial stem cells. J Cell Biol 103:49, 1986
15. Pires RTF, Chokshi A, Tseng SCG: Amniotic membrane transplantation or limbal conjunctival autograft for limbal stem cell deficiency induced by 5-fluorouracil in glaucoma surgeries. Cornea 19:284, 1999
23. Tseng SCG, Tsubota K: Amniotic membrane transplantation for ocular surface reconstruction. In Holland EJ, Mannis MJ, eds. Ocular Surface Diseases: Medical and Surgical Management. New York: Springer, 2001:226–231
28. Wei Z-G, Wu R-L, Lavker RM, et al: In vitro growth and differentiation of rabbit bulbar, fornix, and palpebral conjunctival epithelia. Implication on conjunctival epithelial transdifferentiation and stem cells. Invest Ophthalmol Vis Sci 34:1814, 1993
32. Meallet MA, Espana EM, Grueterich M, et al: Amniotic membrane transplantation for recipient and donor eyes undergoing conjunctival limbal autograft for total limbal stem cell deficiency. Ophthalmology 110:1585, 2003
42. Tseng SCG, Prabhasawat P, Barton K, et al: Amniotic membrane transplantation with or without limbal allografts for corneal surface reconstruction in patients with limbal stem cell deficiency. Arch Ophthalmol 116:431, 1998
47. Rama P, Bonini S, Lambiase A, et al: Autologous fibrin-cultured limbal stem cells permanently restore the corneal surface of patients with total limbal stem cell deficiency. Transplantation 72:1478, 2001
51. Grueterich M, Espana EM, Touhami A, et al: Phenotypic study of a case with successful transplantation of ex vivo expanded human limbal epithelium for unilateral total limbal stem cell deficiency. Ophthalmology 109:1547, 2002