Statistics from Altmetric.com
Persistent corneal epithelial defects with or without ulceration are a serious and urgent clinical problem which can be complicated by microbial infections and thus threaten patients' vision. Experimental studies have revealed that corneal epithelial injuries with or without involvement of the stroma result in rapid healing so long as the eye retains its normal ocular surface defences and possesses healthy limbal epithelial stem cells (for a review see Tseng1). Therefore, whenever there is a persistent defect or ulcer the diagnostic examination should first be directed to ruling out limbal stem cell deficiency, followed by analysis of the integrity of the ocular surface defence governed by neuroanatomical integration of both trigeminal and facial nerves (for a review see Solomonet al 2). Although the actual aetiology may arise from multiple influences and be exogenously or endogenously triggered, the most common denominator is neurotrophic keratopathy, in which trigeminal denervation simultaneously abolishes reflexes controlling tear secretion and lid blinking and closure. When medical treatments fail and the defect or ulcer persists (for example, for more than 3 weeks), conventional surgical treatments become indicated and should include punctal occlusion, application of bandage, scleral lens or tissue adhesive, lamellar or full thickness corneal transplantation, tarsorrhaphy, and conjunctival flap.
In this issue of the BJO (p 1455) Prabhasawat and coworkers reported their clinical experience in using amniotic membrane transplantation for treating 28 patients with persistent corneal epithelial defect or ulcers. As detailed in their report, the majority of eyes (24 of 28, 86%) were indeed affected by neurotrophic keratopathy. Although these authors were not the first reporting the use of amniotic membrane for this indication, theirs is the largest series of seven to date.3-9 Consequently, for the first time they could subdivide their patients into three groups according to the severity of stromal thinning, and compare their outcomes by single layer or multilayers of membrane application. They noted an overall success rate of 82.1%, which falls within the reported range of 67% to 91%. Variable success rates may be attributed to differences in the underlying aetiology and severity, and in the accompanying abnormalities and treatments. Intriguingly, when multilayers were used the defect covered by the amniotic membrane healed significantly faster than that with a single layer. Consistent with what has been reported in seven previous series and two case reports,10 11 they also noted that ocular surface inflammation is markedly reduced following transplantation (in nearly all cases).
Rapid healing and reduction of ocular surface inflammation following amniotic membrane transplantation can be explained by the following action mechanisms (for details also see references cited in the report by Prabhasawat et al in this issue). Firstly, the amniotic membrane provides a new basement membrane, which is an important substrate for supporting adhesion and growth of epithelial progenitor cells, including stem cells. Secondly, the amniotic membrane exerts an anti-inflammatory effect. We have recently reported that expression of IL-1α and IL-1β is markedly suppressed when human limbal epithelial cells are cultured on the amniotic membrane stromal matrix, even if they are challenged by lipopolysaccharide, a bacterial endotoxin.12 Such an anti-inflammatory effect has been demonstrated in experimental injuries caused by excimer laser ablation,13 14 acute chemical burns,15 or herpes simplex virus type 1 induced necrotising keratitis.16 Thirdly, the amniotic membrane stromal matrix has a direct anti-scarring effect as evidenced by its suppression of TGF-β signalling and myofibroblast differentiation. Fourthly, the combination of the above three actions may help re-establish a microenvironmental niche that is conducive for the growth of epithelial progenitor cells. Fifthly, the amniotic membrane may promote nerve regeneration by maintaining nerve growth factor (NGF) signalling. Lambiase et al 17first reported that topical application of NGF is effective in healing neurotrophic ulcers. We have recently gathered experimental evidence to confirm that the amniotic membrane contains a large amount of NGF, and preferentially maintains the NGF signalling system for human limbal epithelial cells in culture (Touhami et al, submitted, 2001).
Collectively, the aforementioned mechanisms also explain why the stromal thickness may be restored in deep ulcers using multilayers of membrane, and why some patients regained more vision after transplantation as shown in this and other published reports. However, amniotic membrane transplantation alone cannot heal the persistent defects or ulcers caused by total limbal stem cell deficiency, as noted by Prabhasawat and coworkers and other investigators. Furthermore, amniotic membrane transplantation has a limited role in treating eyes with severe ischaemia (for example, caused by chemical burns or rheumatoid arthritis) and severe exposure (that is, lack of blinking), to which other surgical therapies need to be added. Nevertheless, amniotic membrane transplantation offers the following advantages over conventional corneal transplantation—relative ease of surgery and avoidance of allograft rejection. As a result, amniotic membrane transplantation becomes an attractive alternative for countries where there is a shortage of corneal tissues. Even if corneal transplantation is needed, it can be performed after amniotic membrane transplantation when the ocular surface is not inflamed.
Proprietary interest: Dr Tseng has filed a patent application for the method and uses of amniotic membrane, and has a financial interest in Bio-Tissue, which is a tissue bank providing amniotic membrane for research and clinical uses.