Amniotic membrane transplantation: structural and biological properties, tissue preparation, application and clinical indications

Abstract

The amniotic membrane is a single epithelial layer of the placenta. It has anti-inflammatory, anti-scarring, anti-angiogenic and possibly bactericidal properties. The basement membrane of the amniotic membrane acts as a substrate to encourage healing and re-epithelialisation. It has been used in many ocular surface diseases including persistent epithelial defects (corneal or conjunctival), chemical or thermal burns, limbal stem cell deficiency, cicatrising conjunctivitis, ocular graft versus host disease, microbial keratitis, corneal perforation, bullous keratopathy, dry eye disease, corneal haze following refractive surgery and cross-linking, band keratopathy, ocular surface neoplasia, pterygium surgery, and ligneous conjunctivitis. This review provides an up-to-date overview of amniotic membrane transplantation including the structural and biological properties, preparation and application, clinical indications, and commercially available products.

Introduction

Amniotic membrane transplantation (AMT) has undergone a renaissance within ophthalmology since techniques for preparation, preservation and screening tissue for blood-borne viruses have improved in recent decades. Its intrinsic properties suppress inflammation, reduce scarring, inhibit angiogenesis, subdue neoplastic activity and provide some element of antimicrobial action. It serves as an excellent substrate for use in the repair of ocular tissue.

The use of amniotic membrane (AM) was pioneered in plastic surgery during the early 20th century as a substance for skin transplantation [1]. It subsequently saw application in transplantation into almost all mucosal surfaces as well as the conjunctiva in 1940 by de Rötth and colleagues [2]. It surpassed the standard of rabbit peritoneum with dried amniotic tissue serving as the donor tissue. It subsequently fell into disuse and this was the case until the mid-1990s when reports of preserved AM for the treatment of several ocular surface pathologies began to gain traction within the literature [3, 4]. Since then the indications for AMT in ophthalmology have expanded greatly, not only with direct transplantation but also as a vector for expanding limbal epithelial stem cells ex vivo for subsequent implantation into the eye afflicted by limbal stem cell deficiency (LSCD) [5].

Reported uses today span from corneal reconstruction, conjunctival repair, and promotion of ocular surface healing to vitreoretinal applications in macular hole repair, use in strabismus surgery and as an adjunct in glaucoma surgical procedures. The current review aims to provide insight into the biology of the AM, the benefits its inherent properties convey as well as explore the current reported applications of AMT.

Structural & biological properties

Amniotic membrane is composed of a single epithelial layer supported by a basement membrane beneath which is a mesenchymal stroma that is further subdivided into compact, fibroblastic and spongy layers [6, 7]. The entire membrane is 20 to 500 μm in thickness, with the spongy layer (or zona spongiosa), acting as the interface with the outermost aspect of the chorion in vivo [7]. The AM’s tractional resistance is reliant on the meshwork of collagen type I & II within the stroma, this is pervasive throughout the acellular compact layer, the more sparsely packed fibroblastic layer, and to a lesser extent within the proteoglycan and glycoprotein-rich zona spongiosa, which is dominated by type III collagen [7]. AM is an avascular structure relying on amniotic fluid, the chorion and fetal surface blood vessels for nutrient supply, and as such it relies heavily on anaerobic glycolysis [8].

In utero, the AM appears to provide some element of secretory activity, contributing to the homeostasis of amniotic fluid [9]. Although the precise functions of the AM are not known within the context of amniotic fluid homeostasis, it does appear critical to the maintenance of a stable pH of 7.10 throughout gestation [7]. This, however, does not appear to be one of the critical properties of AM that make it an attractive tissue for transplantation in ophthalmology. These relate to its main function of protecting the developing fetus [10].

Epithelial repair

The composition of the AM not only provides a physical barrier but promotes the migration of epithelial cells, and their differentiation and leads to strengthened adhesion of basal epithelial cells [11]. This is supported by observations in rabbits that corneal epithelial defect healing is more rapid when AM is applied compared to contralateral control eyes after excimer laser induced injury [12]. This is further reinforced by the inhibition of apoptosis of epithelial cells conferred by the basement membrane and extracellular matrix of the AM [9, 13].

The promotion of orderly epithelialisation is putatively contributed to by the similar collagen composition of the AM basement membrane and the corneal or scleral surfaces. This is reinforced by the trophic factors that are produced by the AM itself acting in a paracrine fashion to promote wound healing. The AM produces platelet derived growth factors (PDGF), which promote the survival, proliferation and migration of epithelial cells [14]. Furthermore, preserved AM has been found to contain significant concentrations of hepatocyte growth factor (HGF) and keratocyte growth factor (KGF), mimicking the trophic environment supplied by corneal keratocytes [15].

Anti-inflammatory, anti-angiogenic, and anti-fibrotic properties

AM possesses numerous mechanisms to suppress inflammation although the exact mechanisms are not fully elucidated, even though observations of reduced inflammatory reaction in multiple clinical settings, from ocular burns to infectious keratitis to Stevens-Johnson syndrome [5, 16]. One important aspect is the down-regulation of pro-inflammatory cytokines such as IL-1α and IL-1β in corneal epithelial cells by AM, as demonstrated by the culture of limbal epithelial cells on an AM matrix suppressing the expression of these cytokines despite pro-inflammatory stimulation such as bacterial lipopolysaccharides [17]. This is further promoted by increased apoptosis of neutrophils following the application of human AM to rabbit corneas following phototherapeutic keratectomy [18]. This is thought to be, in part, to be promoted by the expression of pro-apoptotic ligands, such as Fas (CD95), on the surface of mesenchymal AM cells [9].

TGFβ signalling within fibroblasts is well recognised to be a potent inducer of fibrosis and fibroblastic activity. When corneal and limbal fibroblasts are cultured with AM, they demonstrate suppression of TGFβ as well as TGFβ receptors [19]. This process further suppresses the differentiation of fibroblasts into myofibroblasts, with the consequence of reduced scar tissue formation and the observation of less scarring in corneas treated with AMT following alkali induced damage [20].

Anti-angiogenic properties are vital in the benefit of using AM over other techniques because this maintains the avascular surface of the cornea. The production of anti-angiogenic compounds such as thrombospondin-1, endostatin and tissue inhibitors of metalloproteinases (TIMP-1, 2, 3 and 4) has been demonstrated in vitro from human AM [21], supporting such a property within transplanted AM.

Antimicrobial and antiviral actions

Amniotic membrane appears to possess intrinsic antimicrobial properties [22] that have been noted since its resurgence in use in ophthalmology in the early 1990s. This in vivo demonstration may partly explain the clinical findings of improved outcomes in bacterial keratitis compared to treatment with antibiotics alone [23]. Some of this activity is a product of its inherent anti-inflammatory properties discussed above. However, additional factors that act as a component of the innate immune system – secretory leucocyte proteinase inhibitor, and elafin – have been demonstrated to be expressed by AM in humans [24]. AM permits the penetration of topically applied antibiotics, and may also increase the retention time and bioavailability of antibiotics.

Furthermore, the AM appears to demonstrate intrinsic antiviral properties, which have most frequently been ascribed to its expression of cystatin E [5, 25]. This glycoprotein purportedly acts as a cysteine proteinase inhibitor, hence suppressing viral infections. Antiviral potential is purported to be increased by inducing pro-inflammatory T helper cell apoptosis in the context of herpes simplex keratitis (HSK) in the BALB/c murine model [26]. Additionally in BALB/c mice with ulcerative herpes simplex virus-1 keratitis, AM application reduced expression of matrix metalloproteinases −8 and −9 (MMP−8 and −9), and increased localisation of tissue inhibitors of metalloproteinase-1 (TIMP-1) [27]. This resulted in reduced MMP activity, allowing for rapid healing of the corneal surface with a reduction in the detrimental effects of MMP on the extracellular matrix [27]. However, inherent antiviral activity is contested in vitro cell culture, as AM soaked with aciclovir inhibited herpes viral growth, however, the sole application of AM (without soaking in aciclovir) did not significantly affect viral replication. [28]. The limitation of this model is the lack of immune-system-pathogen interaction, which results in a significant amount of collateral tissue damage in vivo, and may be the main target of the beneficial effects of AM transplantation.

These features are complemented by an immunomodulatory effect of AM in the graft or transplant recipient. Despite epithelial and mesenchymal cells demonstrating limited human leukocyte antigens (HLA) class Ia and Ib antigens in human AM, they do not appear to lead to immunological rejection and hence AMT does not require immunosuppression [29].

Preparation of amniotic tissue

The AM is harvested from the placenta, preferably following a planned caesarean section with strict asepsis due to the possibility of contamination during vaginal delivery [29]. The tissue then requires further decontamination with repeated washes of sterile saline or an established cocktail of antibiotics/antimycotics [29]. An alternative approach of sterilisation with a dose of 25 kGy of gamma radiation has been employed. However, this is associated with adverse changes in the properties of the AM including basement membrane deformation, reduction in growth factors and TIMPs that convey many of the features of AM that make it a useful substrate for ophthalmic transplantation [30].

At this stage, the AM can be separated from the underlying chorion and subdivided into sections appropriate for ophthalmic use. The fresh AM may then be used, however, this comes with significant time constraints and logistical practicalities. Since the turn of the 21st century, several different methods of preserving amniotic tissue for transplantation have emerged.

Cryopreservation

Since the resurgence of usage of AM in the 1990s, cryopreservation has been the mainstay of treatment and storage. It involves the AM being placed in a storage solution supported by a carrier substrate at −80 °C. This solution has most frequently been a one-to-one ratio of 86% glycerol in Dulbecco’s Modified Eagle Medium (DMEM) [31]. Stable storage at −80 °C is accepted to allow the AM to be used after a maximum of 1 to 2 years [9, 29]. This method, however, carries concern regarding the potential damage to the integrity of the AM, courtesy of crystallisation within the tissue, reducing its effectiveness and concentrations of preferable growth factors and anti-angiogenic intermediates [5, 32]. Furthermore, the use of such stored AM requires a well-managed and coordinated logistics system capable of ensuring stable storage at −80 °C throughout the various stages until the AM is required at the appropriate location.

Lyophilisation (freeze-dried)

During lyophilisation, AM is rapidly frozen to temperatures as low as −80 °C and a high-vacuum is then employed to dry the tissue – this gives rise to the more common name of ‘freeze-drying’. The water content is reduced further by sublimation to a maximum of 5–10% to inhibit all enzymatic activity and is sterilised using gamma-irradiation [29, 33]. This removes the need for cold-storage and the logistical issues associated with cryopreservation. However, there are conflicting reports as to whether standard lyophilisation is deleterious or advantageous when compared to standard cryopreservation [29, 34].

Dehydration (heat-dried)

AM can be dried under a biohazard hood by exposure to air and subsequent sterilisation using gamma irradiation (e.g. Cobalt 60 source) to eliminate the need for freezing and the potential harm this inflicts on the biological tissue [29, 35]. However a more recent advance has been the use of low-temperature vacuum evaporation to dehydrate AM, aiming to minimise biochemical and structural damage to the tissue [36]. With this method, the AM is placed in a vacuum and dried at 3.5 °C to 6 °C following removal of the spongy layer, pre-treatment with raffinose and incubation with broad-spectrum antimicrobials [36].

The application of lyoprotective saccharides in the form of trehalose or raffinose combined with the antioxidant epigallocatechin prior to drying is advantageous in retaining many facets of fresh AM that make it an attractive transplant material [32]. This process has been demonstrated to preserve the biochemical stability of multiple growth factors and enzymes crucial to the anti-fibrotic and anti-inflammatory properties of AM [32]. This process again provides AM that is stable at ambient temperatures and so removes the obstacle of cold-storage logistics during distribution.

Amniotic membrane application

Although legal mandates differ between countries, the principles surrounding ensuring the safety of biological tissue remain the same for AM regardless of the geopolitical environment. All potential donors should have a thorough history and examination to exclude risk factors for undiagnosed blood-borne virus (BBV) transmission, as well as potential sexually transmitted infections. This should be further reduced with routine blood testing for BBV serology as well as polymerase chain reaction (PCR) testing for potential current viral load [9, 29]. This is often combined with repeat testing after a period of tissue quarantine to ensure no occult infection at the time of harvesting has become manifest in the donor.

The AM and placenta can become contaminated during the process of collection and preservation. Adequate screening at this stage should include screening for both aerobic and anaerobic bacteria as well as fungi. Despite all these methods as standard, there are still reports of some contamination post-operatively, most frequently with gram-positive bacteria [37].

The AM tissue may be placed epithelial side down (onlay/patch) or epithelial side up (inlay/graft). With the epithelial side down (onlay/patch), the AM acts as a temporary barrier or dressing, to assist with epithelial regrowth beneath the membrane. It is believed that the matrix traps inflammatory cells inducing apoptosis and downregulating the inflammatory response [9]. The AM may be tethered in place with or without sutures [38]. For more indolent, persistent defects, the AM can be placed epithelial side up to permit corneal or conjunctival epithelialisation over the membrane, with the AM acting as a permanent basement membrane integrating into the ocular tissue and being colonised by host keratocytes [39]. For even deeper lesions, multiple layers of AM may be used, including one epithelial side up and the other epithelial side down. This method is purported to have a higher success rate and lower rate of recurrence than the sole use of a graft or patch [39].

Further roles for AM have been elucidated in recent decades. AM may be used as a basement membrane scaffold for ex vivo expansion of stem cells for implantation and as a carrier substrate for stem cells on implantation. This method has not been standardised as yet, and numerous variations exist at present [40]. Denuded AM, when compared to AM with the epithelial cells in situ, appears to demonstrate some preferable characteristics in expanding limbal epithelial stem cells – promoting a more orderly structure with a higher number of desmosomes and hemidesmosomes [41]. This methodology has continued to gain traction and has been expanded to include autologous as well as allogenic stem cell transplantation as a means of improving the limbal epithelial stem cell population [42]. This method is more commonly referred to as cultivated limbal epithelial transplantation (CLET).

In 2012, Sangwan and colleagues pioneered a new technique – simple limbal epithelial transplantation (SLET). [43]. In this technique, there is no ex vivo expansion of the limbal stem cells. The affected cornea is prepared with a 360° peritomy, excision/recession of the corneal conjunctivalisation. AM is then placed overlying the corneal stroma, fixed with fibrin adhesive and tucked under the surrounding conjunctiva [44]. Following this, a small section of the superior limbus (2 mm × 2 mm) is excised from the healthy limbus of the fellow eye, divided into multiple (8 to 10) smaller sections and transferred onto the AM for in vivo expansion [45]. SLET appears to demonstrate improved outcomes in terms of greater anatomical and functional success compared to CLET [45], with success rates of up to 83.8% reported in a multicentre case series [46]. Numerous modifications to SLET have been described including glueless-SLET with autologous transplants inserted directly into radial superficial peripheral corneal incisions [47].

Dua et al. added a further step of using AM to redirect conjunctival ingress following limbal stem cell grafting to prevent admixture of conjunctival epithelium and regenerating corneal epithelium [48]. After dissection of the conjunctival pannus, recession of the adjacent conjunctiva, and placement of the limbal grafts along with an underlying AM graft, a further layer of AM is sutured peripherally covering the entire cornea and tucked under the adjacent conjunctiva. This redirects the aberrant regrowth of conjunctiva over the ‘shielding’ AM patch and allows the reepithelialisation of the underlying cornea unencumbered [48].

Clinical indications

Chemical injury

AM patches have been an attractive proposition in the scenario of acute chemical or thermal injury. Older case series have reported improvement in outcomes in moderate chemical burns with low rates of corneal vascularisation and improved epithelial healing [49]. This was clear with a group of patients, 10 of whom had chemical burns and three had thermal [50]. Those with moderate burns, according to the Roper-Hall classification, demonstrated good re-epithelialisation and low levels of inflammation, however those with severe grade IV burns did not, and also demonstrated LSCD due to limbal ischaemia [50].

More recent randomised control trials compared AM to conventional medical therapy in 50 patients with moderate grade II-III burns and 50 with grade IV burns [51]. These authors reported improved rates of epithelialisation compared to control treatment in moderate burns, but no difference in the severe burn group [51]. Interestingly, they did grade the patients according to Dua’s updated classification, but did not report this in terms of limbal ischaemia and did not divide patients on this basis [51]. A subsequent Cochrane meta-analysis, which included this study solely, did not advocate the use of AM in the acute phase of injury based on the lack of beneficial long term outcomes in the form of visual acuity, corneal vascularisation or symblepharon formation [52]. The most recent randomised trial in 2019 only examined severe Roper-Hall grade IV chemical injuries in 60 eyes of 60 patients [53]. Over the average of 20.3 months of follow-up, there was no difference in visual acuity, neovascularisation or time-to-healing of epithelial defects between standard medical therapy and AM patch, along with medical therapy [53]. Crucially, however, none of these studies have used the more recent Dua classification that may aid in stratifying patients based on limbal ischaemia and hence viability of limbal stem cells [54]. This simple stratification of patients may aid, in those with less severe burns receiving benefit from AM as they have adequate limbal stem cell populations to effectively re-epithelialise their corneal surfaces within the protected-conducive environment of an AM patch. It would appear that more studies may be necessary for this category of patients, with improved classification of patients according to the extent of limbal ischaemia and probability of recovery without the need for stem cell grafting.

In the context of extensive chemical or thermal burns, AM may be applied as a large patch running from the lower lid, across the palpebral conjunctiva covering the lower conjunctival fornix, running back across the bulbar conjunctival surface, over the cornea, and superior conjunctiva in a similar manner to the equivalent lower conjunctiva [55]. This can be sutured in place or held in place with a symblepharon ring. The technique has also been employed for severe Stevens Johnson Syndrome [56].

Cicatrising conjunctivitis

Stevens Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) can cause cicatrising conjunctivitis that may benefit from AM patching in the acute phase of the disease [38]. A comparison of AM adjuvant patching combined with medical therapy to medical therapy alone was investigated in a randomised prospective trial of 50 eyes of 25 patients [57]. The AM group not only demonstrated greater best corrected visual acuity than the sole medical therapy group but also demonstrated no corneal haze, LSCD, symblepharon or entropion/ectropion. This was a significant benefit compared to the medical therapy group, which demonstrated corneal haze in 44%, evidence of LSCD (corneal vascularisation and conjunctivalisation) in 24%, symblepharon in 16% and ectropion/entropion in 8% [57]. This outcome appears to tally with longer-term follow up of AM patch use in the acute phase of SJS and TEN [58]. In this case series of 55 eyes receiving AM patch with a median follow up of 2.5 years, 87% had achieved a best corrected visual acuity better than 6/12, none had developed total LSCD, only 5% developed corneal conjunctivalisation, and 25% had developed some degree of corneal neovascularisation. The most common complication was meibomian gland dysfunction occurring in 78% of patients and dry eye disease in 58% [58]. These studies serve to demonstrate the beneficial role of AM in the acute phase of SJS and TEN in reducing chronic complications.

Ocular graft versus host disease

Ocular graft versus host disease (GvHD) can result in a significant deterioration of the ocular surface with both severe corneal and conjunctival disease. As such, AM can be an option in the management of GvHD at various stages of the disease process. A case report of a young female with GvHD secondary to bone marrow transplantation demonstrates the efficacy of using AMT in the acute phase of GvHD with corneal perforation to seal the defect and allow for further surgical intervention. In this case, the patient underwent lamellar keratoplasty at a later time [59]. Those with chronic recurrent GvHD may be very difficult to manage. Another case report of a patient with GvHD secondary to bone marrow transplantation for aplastic anaemia underwent serial penetrating keratoplasty and lamellar keratoplasty combined with AMT and on occasion, tarsorrhaphy [60]. Alas, the treatments were not successful in saving the eye with recurrent corneal perforation leading to evisceration of the eye [60].

Multi-layered AMT in combination with standard medical therapy may prove beneficial with reports of successful treatment of corneal perforation secondary to GvHD, with one reported to demonstrate a stable corneal epithelium and stroma up to 20 months after the intervention [61]. More recently, the application of AMT in the context of GvHD has been reported in the context of limbal ischaemia and herpes zoster ophthalmicus (HZO). In this instance, the patient had limbal ischaemia and HZO secondary to immunosuppressive therapy combined with GvHD, yet demonstrated re-epithelialisation of the conjunctiva and cornea after successive AMT combined with medical therapy [62].

Dry eye disease

Dry eye disease is a common presentation in the ophthalmic outpatient clinic. The recently reported DRy Eye Amniotic Membrane (DREAM) study investigated 97 eyes of 84 patients who had failed on topical therapy for severe dry eye [63]. After AM application for an average of 5.4 days, 88% of patients demonstrated a statistically significant reduction in the overall dry eye workshop (DEWS) score from 3.25 pre-operatively to 1.44 at 1 week, 1.45 at 1 month and 1.47 at 3 months [63]. The outcome of the DREAM study reflects the results of previous studies using an AM preparation for patients on maximum medical therapy for dry eye disease, in which patients had AM in situ for an average 4.9 days with symptomatic relief for 4.2 months on average [64]. This is in keeping with the guidance from the DEWS 2017 guidance, which recommends the use of AM patches as an alternative to surgical punctual occlusion in step 4 of its ladder of dry eye disease treatment [65].

Recurrent corneal erosion syndrome

Recurrent corneal erosion syndrome (RCES) has attracted some attention in possible intervention with AM patches. However, this is based on a single study by Huang et al., in which 11 eyes of 9 patients with RCES underwent AM patch for 4 to 7 days [66]. In all but one eye there was resolution during the follow up of an average of 13.7 months with smooth and stable corneal epithelia [66]. However, a 2018 Cochrane review for the treatment of RCES did not include the use of AM [67], and further controlled clinical trials are required to provide evidence for AM use in RCES.

Corneal ulceration, keratitis and perforation

The use of AM to repair corneas with persistent ulceration and perforation has been one of the leading uses of AM in ophthalmology. AM has been used in the context of ulceration from wide ranging aetiologies: neurotrophic ulcers, herpetic keratitis, microbial keratitis, autoimmune disease and bullous keratopathy [10]. A recent retrospective analysis of 149 patients demonstrated 70% of patients achieved epithelial closure with a single AM transplant [68]. The patients with the worst outcomes were those with persistent ulcers following previous corneal surgery with only 52% achieving re-epithelialisation. Patients with neurotrophic ulcers demonstrated the best success rates of 93%, closely followed by bacterial and herpetic ulcers with rates of 80% and 85% respectively [68]. However, a recent meta-analysis of eighteen eligible studies including 390 eyes of 385 patients, demonstrated no significant difference in the healing of the corneal epithelium between ulcers of infectious or non-infectious origin when treated with AM transplantation [69]. Notably, however, these data may be skewed as included in non-infectious causes were neurotrophic ulcers, which according to Schuerch and colleagues is one of the best performing aetiologies when treated with AM transplantation [68].

Focusing on infectious keratitis, in a 2017 randomised clinical trial, early application of an AM patch in 49 eyes of 49 patients alongside standard antimicrobial therapy was compared to antibiotics alone in 50 eyes of 50 patients [23]. The investigators reported improved visual acuity at 6 months in the AM group with reduced corneal vascularisation and a smaller scar size, although the acute manifestations did not appear significantly different [23]. Furthermore, a recent retrospective study of AM application in the context of treating severe infectious keratitis associated corneal ulcers demonstrated the superiority of AM over standard antimicrobials or antivirals alone [70]. In this report, the 11 eyes treated with AM demonstrated epithelialisation faster than in eyes treated without AM; 3.56 weeks and 5.87 weeks, respectively [70]. This is in spite of patients treated with AM having larger epithelial defects initially. Additionally, AM appears to result in improved best-corrected visual acuity at the time of complete epithelialisation [70]. There are also reports of successful use of AM to treat Acanthamoeba keratitis resistant to medical therapy [71].

One severe consequence of infectious keratitis may be corneal perforation. Multi-layered application of AM was demonstrated to be successful in perforations up to 1.5 mm in a case series of 15 patients [72]. An advancement in this technique involved the addition of fibrin glue. A case series involving 14 eyes with corneal perforation treated with fibrin glue and AM demonstrated a success rate of 92.9% and concluded that the addition of fibrin glue allowed for the treatment of perforations up to 3 mm in diameter with AM transplantation [73]. A 2016 study used an AM roll-in filling technique (AMR) combined with intracameral injection of C₃F₈ gas to reduce aqueous leakage [71, 74]. This was in a series of 46 cases of corneal perforation due to a mixture of bacterial, herpes simplex and fungal keratitis. They reported 100% success, defined as cessation of aqueous leak, epithelialisation of the AM over the perforation site and maintained anterior chamber depth [74]. In a large retrospective case series of 119 eyes that underwent AM transplant for corneal perforation or corneal melt, 102 (86%) demonstrated successful anatomical reconstruction of the anterior chamber [75]. However, of these eyes, 23 required penetrating keratoplasty at a later stage, suggesting that AM may not provide a permanent solution. The authors suggest this may be related to the underlying aetiology of the perforation, with the implication that successful treatment of corneal melt and perforation with AM require careful patient selection prior to the procedure [75]. The addition of lamellar keratoplasty to AM transplantation offers an intriguing alternative when there is a lack of donor corneas for penetrating keratoplasty in the context of corneal perforation [76]. Although this particular study is very limited in numbers and the aetiology of perforation, with only one case of infectious keratitis included.

The addition of botulinum toxin-induced ptosis further improves the success rate of primary AM transplantation, reducing the reoperation rate from 44.6% to 34.4% in a retrospective analysis when compared to AM transplant alone [39, 77]. Similarly, tarsorrhaphy has been indicated as an adjunct to improve surgical outcomes [39]. When compared to tarsorrhaphy or bandage contact lens alone, AM transplants performed equally well with no significant difference in re-epithelialisation and improvement in best correct visual acuity in those with persistent neurotrophic ulcers [78].

Bullous keratopathy

Bullous keratopathy is another painful condition afflicting the cornea. AM transplantation has been demonstrated to be beneficial in improving pain to a similar extent as anterior stromal puncture, but it is beneficial in providing a sustained regular anterior epithelial surface following AM transplantation [79]. Combined anterior stromal puncture with an AM transplant has also been reported to be effective in a retrospective case series of 12 eyes in 12 patients, with resolution of pain and conjunctival inflammation in 11 out of the 12 patients [80]. This improvement in pain and corneal epithelialisation 3 weeks post procedure is also observed in the subset of patients with poor visual potential treated with AM transplantation [81]. Phototherapeutic keratectomy (PTK) combined with AM grafts may provide additional benefits in those with bullous keratopathy. In 11 of 12 patients with painful symptomatic bullous keratopathy, authors demonstrated complete resolution of symptoms with a mean follow-up of 38 weeks when treated with AM graft and PTK [82].

Haze following photorefractive keratectomy and cross-linking

AM has also been used to address corneal haze following photorefractive keratectomy (PRK) and other surface based procedures including corneal cross-linking. Early use of AM in rabbits treated with laser photoablation, demonstrated reductions in the inflammatory response and keratocyte apoptosis in the early post-operative period [83]. These leporine studies also demonstrated a reduction in the corneal haze following laser photoablation in the later post-operative period, weeks 7 to 12 [83]. However, in a human study comparing bandage contact lenses and AM after PRK in 40 myopic patients, authors observed no differences in visual outcomes, corneal clarity or speed of re-epithelialisation [84]. Moreover, a more recent study from the same group demonstrated no benefit of AM on corneal densitometry readings in a retrospective study of 78 eyes that underwent PRK for myopia [85]. However, there are previous reports of the efficacy of AM in severe corneal haze following previous PTK with improvements in vision from 6/30 to 6/12 with resolution of the corneal haze [86]. AMT has also been employed in the management of epithelial ingrowth following laser in situ keratomileusis (LASIK)[87].

Band keratopathy

The combination of AM transplants with the use of ethylenediaminetetraacetic acid (EDTA) chelation for band keratopathy has also been reported as an effective to maintain a stable corneal epithelium and reduce ocular surface pain [88, 89]. All 11 eyes in one study acquired re-epithelialised corneas within a mean of 10.6 days [89] whilst 15 of 16 eyes in another study demonstrated re-epithelialisation. The one eye that did not, was later discovered to have complete LSCD [88].

Ocular surface neoplasia

The excision of ocular surface neoplasia leaves a conjunctival defect that is amenable to repair with AM. Early work demonstrated efficacy in a series of 16 eyes that had large ocular surface neoplasms excised [90]. They demonstrated complete healing of the ocular surface and recurrence in only 1 patient with conjunctival intraepithelial neoplasia in a follow-up period of nearly 2 years. A retrospective analysis of 53 patients treated with AM graft and cryotherapy also demonstrated that the use of AM resulted in low levels of scarring and symblepharon, although this analysis was lacking in control subjects as a comparison [91]. A single case report of the addition of mitomycin C along with the use of an AM graft following excision of conjunctival intra-epithelial neoplasia resulted in no ocular complications for the patient, suggesting this method may benefit patients [92]. The use of thicker sections of AM for reconstruction of the ocular surface has also been demonstrated to be beneficial in a range of neoplasias, including squamous cell carcinoma, melanoma, sebaceous carcinoma and atypical pterygium, with only 2 of 12 cases demonstrating disease recurrence [93]. This method, however, may remove one of the other perceived benefits of AM grafting in its relative transparency when compared to other mucosal tissues used, thus allowing for monitoring of disease recurrence [39].

Enucleation or evisceration may be the sad consequence of malignant neoplastic disease of the eye. However, AM may still be of use in such situations offering a treatment for anophthalmic contracted orbits offering similar outcomes to mucous membrane grafting [94], which is further discussed below.

Pterygium

Treating pterygia has been one of the areas of most interest and controversy in the application of AM grafts. A Cochrane review compared AM to conjunctival autograft in 1947 eyes of 1866 patients [95]. The authors demonstrated a reduced risk of recurrence at 6 months in those treated with conjunctival autograft when compared to AM transplant with a risk ratio of 0.43 (95% confidence interval 0.30 to 0.62). This suggests that AM may be the inferior methodology for the treatment of pterygia. An interesting development is the combined method of conjunctival autograft with subconjunctival application of AM [96]. In a retrospective study of 493 eyes in 355 patients, the authors demonstrated a recurrence rate of 1.22% in a follow-up range of 6 months to 6 years [96]. When this is compared to the findings of the Cochrane review, in which the recurrence rate varied from 3.3% to 16.7% at 6 months in those with a sole conjunctival autograft, it appears the combination of AM and conjunctival autograft may be a promising incremental improvement in the treatment of pterygia.

Ligneous conjunctivitis

Ligneous conjunctivitis (LC) is a rare disease caused by a deficiency of plasminogen and results in thick wood-like pseudomembranes forming over the ocular mucosal surface [97]. In 2004, Barabino and colleagues reported the first use of AM grafts to treat LC, with promising results and no recurrence of the membranes at 36 months [98]. Similarly, the treatment of a case of LC with AM graft and topical cyclosporine for 6 weeks demonstrated no recurrence at 39 months [97]. This combination of cyclosporine and AM graft has proven beneficial in two cases followed up to 40 and 28 months respectively, with no recurrence of membranes in either case [99]. However, further investigation is required into this rare condition to validate the use of AM routinely in its management.

Limbal stem cell deficiency (LSCD)

LSCD is an area of great interest with regard to the use of AM, relating to the various surgical innovations from CLET to SLET that have been pioneered using AM as a crucial component in the success of these procedures. The progenitor stem cells crucial to the regeneration of the corneal surface reside at the limbus – hence the clinical interest in promoting regeneration of this population in LSCD. As described above, SLET appears to show similar benefits to CLET, whilst removing the technical difficulties associated with ex vivo expansion and culturing of stem cells [100]. Complete epithelialisation is observed by an average of 22.5 days [101], with success rates across multiple case series averaging 74.2% for partial LSCD and 76.8% for total LCSD [100]. Post-operative best corrected visual acuity is also better in SLET with 68.6% of patients demonstrating a two-line improvement compared to 51.5% with CLET at their latest follow-up [45].

Interestingly SLET demonstrates success rates of 62.5% in cases of failed primary CLET, and exceeded that of repeat CLET, which was successful in only 53.5% of cases [102]. One group has modified SLET with a second overlying AM to provide a ‘sandwich’ SLET technique which they report provides more protection for the transplanted stem cells, however, more data is required to examine the validity of this hypothesis [103].

A further advancement in the field of SLET has been the introduction of allogenic SLET (alloSLET) in cases of severe bilateral disease where conventional SLET would not be possible [100]. This, however, comes with the added burden of requiring immunosuppressants. The outcomes of the largest study of 30 eyes appear to show good success rates with corneal re-epithelialisation in 78.6% of alloSLET using cadaveric donors and 87.5% in those from living-related donors [44, 100]. A subsequent retrospective analysis has been published, comparing AM graft and alloSLET in 38 and 39 eyes, respectively, for Dua grade 4 or worse ocular chemical injuries [104]. This study demonstrated the acute use of alloSLET was associated with an improved outcome in terms of a lower risk of developing subsequent LSCD (Odds Ratio (OR) 0.137) and need for subsequent keratoplasty (OR 0.093) in those that underwent alloSLET [104].

The use of oral mucosa to provide progenitor cells on an AM substrate carrier has also been explored. The most frequently studied has involved cultivated oral mucosal epithelial transplantation (COMET), which demonstrates a re-epithelialisation of the corneal epithelium in a mean of 5.2 days [105]. This technique may be of use in those with bilateral complete LSCD. Interestingly, this method has been developed similar to SLET in the rabbit model with subsequent promise demonstrated for simple oral mucosal epithelial transplant (SOMET) in a human subject with bilateral LSCD, resulting in improved vision, regression of corneal vascularisation and a stable epithelium [106].

The complexity in the management of LSCD has prompted the recent publication of a consensus statement on the management of LSCD from The Cornea Society, which may provide some framework in which these newly emerging techniques may be integrated [107].

Macular hole, strabismus and glaucoma surgery

Although the primary use of AM in ophthalmic clinical care has been for the treatment of ocular surface disease, other areas of application have been emerging in recent years and are briefly discussed here.

Vitreoretinal specialists have started to employ AM for the treatment of macular holes and it has demonstrated some promise [108]. Patients with macular holes appear to show improved best corrected visual acuity following implantation of AM combined with pars plana vitrectomy and either silicone oil or gas tamponade [108]. The same group has also had success at implanting AM to repair retinal detachment associated with a retinal tear with evidence of neuroretinal ingrowth into the region of the AM [109].

Performing strabismus surgery combined with the direct application of AM to the extraocular muscles undergoing surgical intervention is another area of on-going research. There appears to be improved post-operative metrics in patients treated with AM, with less restriction of ocular movements, which is thought to relate to reduced fibrosis within the extraocular muscles and hence fewer adhesions [110]. A number of different application methods have been investigated to reduce the development of adhesions, whether that be epithelial or stromal side facing the muscle, sclera or Tenon’s capsule, with or without suture fixation. One recent review concluded that, regardless of the orientation of the AM graft applied to the muscle during strabismus surgery, there was a good effect of cryopreserved AM in reducing adhesions [110]. An additional benefit may be found specifically in those undergoing reoperation for strabismus, with a study of 14 patients demonstrating improved duction and reduced deviation in this subgroup of patients [111], although this study lacks a control group and further investigation is required.

Surgical interventions for glaucoma have also shown some mixed results when using AM. AMT augmented trabeculectomy, in which AM is inserted under the scleral flap, have shown increased success in primary interventions with lower IOP at 3 and 12 months post-operatively [112]. Additionally, AM augmentation of trabeculectomies combined with the use of mitomycin C increased the success rate in patients with refractory glaucoma [113]. However, the use of AM in bleb revision following a bleb leak has demonstrated mixed results, with some results demonstrating its efficacy in the long term, meanwhile others suggest it is not a legitimate alternative to the use of conjunctival advancement due to the risks of early bleb leakage [5, 114].

Oculoplastics

AM has shown promise in multiple oculoplastic procedures including fornix reconstruction in the context of contracture and symblepharon in the anophthalmic socket as well as cicatricial eyelid abnormalities [5, 94]. Out of 18 patients treated in a non-comparative study of AMT for contracted sockets, 16 achieved sufficient fornix depth to retain an ocular prosthesis [115]. When AM is compared to conventional mucous membrane grafting, no difference in post-operative fornix depth or socket volume in 20 patients with contracted anophthalmic sockets has been reported [116]. A recent application of ‘ultra-thick’ human AM in fornix reconstruction resulted in a good prosthetic fit for 3 patients with a follow-up range of 10–14 months [117].

What is notable, however, is that patients undergoing socket reconstruction with AM requires a healthy stem cell pool to allow re-epithelialisation, a significant disadvantage when compared to conventional mucous membrane transplantation [94]. Interestingly, those with severe symblepharon had a successful restoration of fornix depth from a combination of sutureless AM transplant along with mucous membrane transplantation, cicatrix lysis with intraoperative mitomycin C [118]. This provides the potential additional benefits of AM use where minimal healthy stem cells remain from a variety of underlying aetiologies including SJS, chemical or thermal burns, mucous membrane pemphigoid, xeroderma pigmentosa and GvHD [118]. The additional benefit of AM is in the context of diseases where there may be limited availability of oral mucosa for grafting, such as a case of SJS that had had previous reconstruction and required a repeat procedure following recurrent contracture of the socket [119].

Examples of commercially available preparations of AM

Omnigen® and OmniLenz®

NuVision^TM (Nottingham, United Kingdom) produce both Omnigen® and OmniLenz® with their proprietary Tereo^TM processing technique. This technique is an adaptation of low-temperature vacuum dehydration of AM along with the application of lyoprotective saccharides combined with the antioxidant epigallocatechin. This allows for the storage and distribution of the processed AM at ambient temperatures, as well as preserving many of the beneficial properties of AM, as discussed above.

Omnigen® is available in multiple forms with the epithelial surface marked for orientation (Fig. 1). Discs are supplied in various diameters as outlined in Table 1. A rectangular and various square sheets are also available, which may be used in more extensive ocular surface disease.

Fig. 1.

Omnigen® and OmniLenz® (A) OmniLenz® in situ over a persistent epithelial defect following glaucoma surgery and (B) healed corneal epithelium following removal of OminLenz® (C) Omnigen® graft sutured with a continuous nylon suture for a persistent epithelial defect, and (D) fluorescein staining demonstrating the closure of the defect with the AM graft. (Images provided courtesy of NuVision^TM).

Table 1.

Summary of two commercially available amniotic membrane preparations.

Manufacturer	Product subcategory	Processing	Storage requirements	Size	Additional features
NuVisionTM	Omnigen®	TereoTM processing – adapted low-temperature vacuum dehydration with lyoprotective saccharide and antioxidant application	Ambient temperature storage and distribution	20 mm disc
				25 mm disc
				50 mm disc
				15 mm square
				20 mm square
				25 mm square
				50 mm*75 mm
	OmniLenz® C			12 mm AM disc	16 mm contact lens carrier designed for corneal coverage
	OmniLenz® L			17 mm AM disc	18 mm contact lens carrier for corneal and limbal coverage
BioTissue®	Prokera Classic®	CryoTek® - proprietary cryopreservation	Cold-storage −80 °C to 4 °C and cold-storage logistics required		Incorporated symblepharon ring
	Prokera PLUS®				Double layered AM
	Prokera Slim®				Thinned AM section for patient comfort
	Prokera Clear®				6 mm central clear aperature for vision
	Amniograft®			15 mm*10 mm
				20 mm*15 mm
				25 mm*20 mm
				35 mm*35 mm
				50 mm*50 mm
				100 mm*50 mm
	Amnioguard®			10 mm*7.5 mm	Increased thickness 500–900 µm
				25 mm*20 mm
				30 mm*40 mm

OmniLenz® is AM preserved as per Omnigen® but in combination with a soft contact lens to allow for sutureless application of AM in the clinic setting without the need for a surgical theatre. This comes in a smaller diameter to the OmniLenz® C variety targeted at corneal coverage and a larger OmniLenz® L variety aimed at including limbal coverage.

Prokera®, Amniograft® and Amnioguard®

BioTissue® (Miami, Florida, USA) produce an alternative in the form of the Prokera®, Amniograft®, and Amnioguard® family of AM derived products. The AM is processed using the company’s proprietary CryoTek® processing method which is broadly a form of cryopreservation. This necessitates cold-storage logistics for delivery and after receipt of the product prior to use.

Prokera® is available in four varieties, all of which incorporate a symblepharon ring carrier. Two of these are thicker with the stated aim of reducing the chances of adhesions: the single layered Prokera® Classic and the double-layered Prokera® PLUS. Additionally, there is the thinner Prokera® Slim aimed to improve comfort for the patient and the Prokera® Clear, which has a central 6 mm “Clear-view” aperture aimed to maintain some visual acuity during treatment for monocular patients. Unfortunately, Prokera® is no longer available in the UK.

Amniograft® and Amnioguard® are both produced in various rectangular sheets for a variety of applications. The main difference is that Amnioguard® is produced as a thicker section of 500 µm to 900 µm. The sizes produced are outlined in Table 1.

OcuFix2

Ocufix2 is a dehydrated, non-viable cellular human amnion with chorion membrane (MiMedx, Marietta, Georgia, USA) that is sterilised using gamma irradiation and processed with aminoglycoside antibiotics. It is stored at ambient conditions with a 5 year shelf life. OcuFix2 allografts are procured and processed in the USA and are also available in the UK. It is available as a 16 mm disc sheet or various square/rectangular sheets and can be fixated by suturing or with tissue adhesives.

Visio AMTRIX®

Visio AMTRIX® (TBF Génie tissulaire, Mions, France) is a lyophilised AM that also uses gamma irradiation and is chemically treated in the sterilisation process. It can be stored at room temperature for up to 5 years. It has recently been reported as an effective treatment for persistent epithelial defects in a sutureless fashion with a bandage contact lens, akin to the Omnilenz® [120].

Examples of AMT in use are shown in figures 2–7.

Fig. 3.

AMT in a persistent epithelial defect following penetrating keratoplasty.

Fig. 4.

A A pigmented conjunctival lesion which was excised and sent for histological examination. B Sutured AMT following excision of the conjunctival lesion.

Fig. 5.

A Peripheral ulcerative keratitis (PUK) with inferior corneal thinning. B Fluorescein pooling in the area of thinning. C Anterior segment optical coherence tomography (OCT) of the thinned area. D Sutured AMT used as part of the management of this condition.

Fig. 6.

A Fluorescein staining demonstrating a persistent epithelial defect following microbial keratitis within a penetrating keratoplasty (PK) corneal graft. B AMT in the management of this persistent epithelial defect. C Resolution of the persistent epithelial defect.

Fig. 2.

Multi-layered AMT with glue in the management of corneal perforation in a herpetic neurotrophic cornea.

Fig. 7.

A Emergency tectonic penetrating keratoplasty (PK) performed in an eye with severe microbial keratitis and corneal perforation. B Persistent epithelial defect post-operatively, which was managed with an Omnigen® and OmniLenz®. C Partial resolution of the persistent epithelial defect.

Conclusions

AM is a semi-permeable, durable yet malleable and is an invaluable resource in a range of ophthalmic indications, providing a biological membrane that affords an environment conducive to epithelial regeneration, suppresses aberrant tissue repair, and inhibits angiogenesis and inflammation. Over recent decades AM has seen application as a protective patch aiding the native corneal epithelium to generate, as a permanent AM graft to provide a stromal surface for corneal epithelium to repopulate, as a carrier for limbal stem cell grafts and, more recently, as a plug for retinal repair. The range of functions has seen the clinical usage of AM in a breadth of pathologies including chemical injury, SJS, LCSD, infectious keratitis and corneal perforation.

Author contributions

Manuscript drafting: FWBS, CM. Critical revision: FWBS, JH, JLADB, SH, CM. Final approval: FWBS, JH, JLADB, SH, CM.

Competing interests

No author has any financial interest in any product mentioned in this manuscript. Dr McAlinden has submitted an ICMJE form for financial interests outside the submitted paper.