The amniotic membrane was prepared and preserved using the method previously described by Tseng and colleagues. 8 The human placenta is obtained from elective cesarean sections in women who have been previously screened with serologic tests for human immunodeficiency virus, hepatitis B, hepatitis C, and syphilis. The placenta is washed under a laminar flow hood with a sterile solution containing 50 μL/mL penicillin, 50 μL/mL streptomycin, 100 μL/mL neomycin, and 2.5 μL/mL amphotericin B. The amnion is blunt dissected from the chorion. It is then mounted on fragments of nitrocellulose paper, with the epithelium uppermost. The fragments of amniotic membrane are preserved at −80°C in sterile recipients containing equal amounts of Dulbecco's modified Eagle medium (Life Technologies Inc., Paisley, Scotland) and glycerol.
The same surgeon carried out all the surgical interventions (O.G.) under peribulbar anesthesia. In the 10 cases of persistent epithelial defect, any poorly adhered epithelium was removed from the margins of the defect using microsponges and a 0.12 forceps. The amniotic membrane was implanted with the epithelial side uppermost. When the membrane was used to cover the entire cornea without extending beyond the limbus (C), it was secured using interrupted 10-0 nylon sutures parallel to and 0.5 mm from the limbus, with the knots buried (Fig. 2). When the amniotic membrane implant was applied to the entire cornea extending beyond the limbus in 360 degrees by 3 mm (C+L), it was doubly secured using both continuous 10-0 nylon sutures on the peripheral cornea and interrupted 9-0 Vicryl sutures at the conjunctiva with episcleral bites (Fig. 1). In all patients, chloramphenicol and dexamethasone ointments were applied at the end of the surgery, with ocular occlusion for the first 24 hours. A therapeutic contact lens was fitted once the occlusion had been removed in five of the 10 cases in which the amniotic membrane covered the cornea alone to maintain the membrane in position longer. In the remaining five cases and in the additional 10 cases in which the membrane exceeded the limbus, therapeutic contact lenses were not used. Postoperative treatment consisted of dexamethasone drops twice daily and chloramphenicol drops three times daily in the patients fitted with a therapeutic contact lens, and the same in ointment form in those without.
A second implant was used that did not extend beyond the limbus (C) combined with a therapeutic contact lens to maintain the membrane in place longer in cases in which improvement had been shown but not complete epithelialization, after early loss of the original amniotic membrane implant (cases 3 and 7 in Table 1). In case 6 (Table 1), in which early detachment of the implant with partial epithelialization had also occurred, we preferred to perform a lateral tarsorrhaphy. In this case, we opted for a lateral tarsorrhaphy instead of a second amniotic membrane implant because the patient presented with a mild floppy eyelid syndrome.
There were no intra-or postoperative complications. In the early postoperative period, all cases showed a decrease in ocular hyperemia, which included the uncovered conjunctiva surrounding the amniotic membrane. The amniotic membrane implant remained in place for a mean of 12.5 days (with a minimum of 3 days and a maximum of 34 days). In group 1, which consisted of 10 patients with persistent epithelial defect, complete epithelialization occurred in three cases and partial epithelialization in three. There was no response in four cases. The amniotic membrane implant detached within the first week in six of the 10 cases and in the remaining four cases between 2 and 3 weeks. Complete epithelialization was achieved in three of the four cases in which the amniotic membrane remained in place for 2 or more weeks. In contrast, complete epithelialization occurred in none of the six cases in which the amniotic membrane implant was lost during the first week. Of these patients, half showed no response, and half achieved only partial epithelialization. Of the cases in which only partial epithelialization occurred, one patient achieved complete epithelialization after a second implant (case 3), another patient showed no change despite another implant (case 7), and the third patient, with an associated floppy eyelid, proceeded to tarsorrhaphy (case 6). A therapeutic contact lens was applied postoperatively in two of the five cases with type C implants. In case 4, the implant remained in situ for 14 days with complete epithelialization achieved, whereas in case 6, the implant became detached on day 6, probably owing to an associated floppy eyelid.
In the second group, consisting of 10 patients with surgically induced epithelial defects, complete epithelialization was achieved irrespective of the length of time the amniotic membrane implant remained in place. Complete epithelialization was observed in eight of the 10 cases at the time of detachment or removal of the amniotic membrane. Partial epithelialization was observed in the two remaining cases when the implant became detached (at 3 and 8 days). However, both achieved complete spontaneous epithelialization within a few days. The application of a therapeutic contact lens postoperatively over the amniotic membrane in the five cases of this group in which the implant did not extend beyond the limbus (C) increased the time the amniotic membrane remained in situ. In cases 16 and 17, in which a contact lens was not used, the amniotic membrane became detached within 2 weeks. In cases 18, 19, and 20, the contact lens and amniotic membrane remained in place for more than 4 weeks, until both were removed. The mean follow-up period was 17 months (range, 12–23). No recurrence of the epithelial defect occurred in any of the cases that had responded to the amniotic membrane implant.
In recent years, various studies have shown the amniotic membrane graft to be effective in the treatment of a variety of pathologies: corneal ulcers, 5,10,15 bullous keratopathy, 11 and in the reconstruction of the ocular surface in cases of partial limbal stem cell deficiency. 8 The surgical technique used in all these situations involves covering the epithelial-stromal defect with a fragment of amniotic membrane, occasionally using more than one layer of amniotic membrane. This amniotic membrane fragment was sutured edge to edge to the surrounding healthy epithelium (Fig. 3), without extending beyond the margins of the epithelial defect. In this way, the amniotic membrane acts as a basal membrane, allowing reepithelialization over it, stemming from the surrounding healthy epithelium. Thus, we can treat corneal ulcers before they progress to perforation, 5,10 improve the symptoms in bullous keratopathy associated with poor visual potential, 11 or reconstruct in cases with severe ocular surface pathology. 8 However, with this technique, the amniotic membrane becomes trapped beneath the newly formed epithelium, which can considerably affect corneal transparency for several months at least. Therefore, this technique should not be used in cases in which an epithelial defect alone is involved or in those in which reepithelialization could achieve a full recovery of the visual acuity. In this report, we present our results using amniotic membrane implantation as a patch. The entire corneal surface is covered with the amniotic membrane, overlapping the edges of the epithelial defect, so that epithelialization occurs beneath the amniotic membrane (Fig. 4). In this way, corneal transparency is completely restored once the amniotic membrane either detaches or dissolves. The amniotic membrane used in this manner provides many of the advantages given by a therapeutic contact lens in that it helps to stabilize the corneal epithelium, protect against rubbing of the eyelids, decrease symptoms, and avoid drying of the cornea. 3 It also has a variety of biologic properties that have been previously demonstrated to favor epithelialization, which include high growth factor concentrations, 13 along with a reduction in inflammation and scarring. 11 This combination of mechanical and biologic effects suggests, at least theoretically, that the amniotic membrane implant could be a superior option to the therapeutic contact lens. However, the length of time that this type of implant remained in situ was the main problem found in our study. The amniotic membrane implant became detached within the first week in more than half of the cases, despite the use of more fixation sutures in some of these patients. This was probably owing to the mechanical effect of the eyelids because the epithelialization had taken place beneath the amniotic membrane and it was not tethered. We did not observe any differences in corneal epithelialization or length of time that the amniotic membrane remained in place between the two implantation techniques (C and C+L). Nevertheless, in those patients in whom the amniotic membrane did not extend beyond the limbus (C), the application of a therapeutic contact lens postoperatively succeeded in increasing the time that the implant remained in place. On examining the results of the 20 cases of the two groups, we found that four of five cases retained the membrane for 2 or more weeks with this implantation technique.
There was a difference in the degree of epithelialization between the two groups of patients. Group 1 comprised patients who showed difficulty in epithelialization. In these patients, we observed a relationship between the length of time that the amniotic membrane implant remained in place and a successful outcome. Complete epithelialization was achieved in three of the four patients in whom the amniotic membrane implant remained in place for 2 or more weeks (cases 1, 2, 4, and 5). The opposite was true when the membrane remained in place for less than a week, which was not long enough for complete epithelialization in all cases.
Because most cases of persistent epithelial defect respond to medical treatment with preservative-free lubricants or therapeutic contact lenses, these should be used as first-line treatments. Only when these options fail should surgical treatment be considered. In such cases, a corneal amniotic membrane implant (C) combined with a therapeutic contact lens could be a therapeutic option that some patients may prefer as an alternative to tarsorrhaphy.
Group 2 included patients with surgically induced epithelial defects. These patients presented fewer problems with regard to epithelialization than the patients in group 1. Epithelialization was achieved in all group 2 patients, regardless of the time that the implant remained in place. At the same time, two distinct groups of patients can be found in group 2. The first five patients (cases 11 to 15) presented with subepithelial corneal degeneration associated with congenital aniridia or Salzmann's degeneration. Epithelialization is also achieved in these patients without the use of an amniotic membrane implant on the epithelial defect left after superficial keratectomy.
However, these patients show a certain degree of hypofunction at the limbal stem cell level, and we frequently obtain a whirlwind-type pattern of epithelialization. With the use of the amniotic membrane implant to cover the cornea and limbus (C+L), we have seen more rapid and smoother epithelialization, especially in the cases of Salzmann's degeneration. This could be explained by the presence of growth factors 15 and the anti-inflammatory properties 11 of the amniotic membrane. Finally, in cases 16 to 20 of group 2, a corneal implant that did not overlap the limbus (C) was used to cover an amniotic membrane graft in the treatment of neurotrophic corneal ulcers. Various studies have demonstrated that a small graft at the site of the ulcer, consisting of one or more layers, can achieve corneal epithelialization. 5,10 In these cases, we always use the remaining fragment of amniotic membrane to cover the graft and corneal surface. In this way, the second implant acts as a therapeutic contact lens by protecting the original graft and providing larger quantities of growth factors and anti-inflammatory substances. Thus, epithelialization is promoted over the stromal graft and beneath the implant, which acts like a patch.
There is still dispute as to how the amniotic membrane should be implanted, with either the basement membrane or the stromal surface uppermost. When we want the epithelialization to occur over the amniotic membrane (as in corneal ulcers and conjunctival defects), the amniotic membrane should undoubtedly be placed with the basement membrane uppermost, where it acts as an excellent substrate for epithelialization. However, when used as a therapeutic contact lens, with the aim of inducing epithelial growth beneath it, it remains unclear which is the better implantation method. A recent study 14 used the amniotic membrane to cover the entire ocular surface including the conjunctival fornices in the treatment of acute chemical burns. The amniotic membrane was applied with the stromal surface uppermost in five eyes and the basement membrane uppermost in eight. Although the results in general were favorable, the criteria used in choosing the cases were not specified in the article nor were the results for each group given separately. We obtained favorable results in our series with the application of the amniotic membrane with the basement membrane uppermost in all cases. We have also applied the amniotic membrane with the stromal surface uppermost in subsequent cases (unpublished data). Although only a few patients were involved with a short follow-up time, little variation was observed with respect to epithelialization, decrease in inflammation, or the length of time that the amniotic membrane remained in place. In our view, there is probably little difference between the two forms of implantation in these cases because the basement membrane plays a less important role, and growth factors and anti-inflammatory substances can diffuse to the surrounding tissues. However, further studies using a larger number of cases will help to clarify this point.
Despite the small number of cases in the study, we believe that three significant conclusions can be made. First, the amniotic membrane implant used as a patch can be useful and safe to achieve healing in some cases of persistent epithelial defect that have not responded to medical treatment. Second, when the amniotic membrane implant remains in place longer, the number of successful cases appears to increase. Finally, the implantation technique C plus postoperative soft contact lens achieved longer duration of the amniotic membrane implant than the other possible implantation techniques (C and C+L). However, we should remember that this is a surgical approach. For this reason, we consider that the treatment of a persistent epithelial defect should be primarily medical with the use of preservative-free artificial tears, with or without therapeutic contact lens. This treatment is usually successful in most cases. In cases in which medical treatment has failed, an amniotic membrane implant can be a safe and useful alternative that some patients may prefer over tarsorrhaphy. Actually, our preferred technique of implantation is to use an amniotic membrane corneal implant without extending beyond the limbus with a postoperative soft contact lens to increase the time that the implant remains in place.
To evaluate whether amniotic membrane implantation is superior to other alternative therapies (soft contact lens alone or tarsorrhaphy), prospective comparative studies with larger case numbers should be performed.
In the same way, amniotic membrane can be used like a therapeutic contact lens to protect another underlying amniotic membrane graft or to help epithelialization after corneal surgery in cases of superficial keratectomy or in which extensive deepithelialization has occurred.
1. Macaluso DC, Feldman ST. Pathogenesis of sterile corneal erosions and ulcerations. In: Krachmer JH, Mannis MJ, Holland EJ. Cornea. Fundamentals of cornea and external disease, vol. 1. St. Louis: Mosby, 1997: 199–214.
2. Smiddy WE, Hamburg TR, Kracher GP, et al. Therapeutic contact lenses. Ophthalmology 1990; 97: 291–5.
3. McDermott ML, Chandler JW. Therapeutic uses of contact lenses. Surv Ophthalmol 1989; 33: 381–94.
4. Kim JC, Tseng SCG. Transplantation of preserved human amniotic membrane
for surface reconstruction in severely damaged rabbit corneas. Cornea 1995; 14: 473–84.
5. Lee S, Tseng SCG. Amniotic membrane
transplantation for persistent epithelial defects with ulceration. Am J Ophthalmol 1997; 123: 303–12.
6. Tseng SCG, Prabhasawat P, Lee S. Amniotic membrane
transplantation for conjunctival surface reconstruction. Am J Ophthalmol 1997; 124: 765–74.
7. Shimazaki J, Yang H, Tsubota K. Amniotic membrane
transplantation for ocular surface reconstruction in patients with chemical and thermal burns. Ophthalmology 1997; 104: 2068–76.
8. 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 1998; 116: 431–41.
9. Azuara-Blanco A, Pillai CT, Dua HS. Amniotic membrane
transplantation for ocular surface reconstruction. Br J Ophthalmol 1999; 83: 399–402.
10. Kruse FE, Rohrschneider K, Völcker HE. Multilayer amniotic membrane
transplantation for reconstruction of deep corneal ulcers. Ophthalmology 1999; 106: 1504–11.
11. Pires RTF, Tseng SCG, Prabhasawat P, et al. Amniotic membrane
transplantation for symptomatic bullous keratopathy. Arch Ophthalmol 1999; 117: 1291–7.
12. Su CY, Lin CP. Combined use of an amniotic membrane
and tissue adhesive in treating corneal perforation: a case report. Ophthalm Surg Lasers 2000; 31: 151–4.
13. Koizumi N, Inatomi T, Sotozono C, et al. Growth factor mRNA and protein in preserved human amniotic membrane
. Curr Eye Res 2000; 20: 173–7.
14. Meller D, Pires RT, Mack RJ, et al. Amniotic membrane
transplantation for acute chemical or thermal burns. Ophthalmology 2000; 107: 980–90
15. Chen HJ, Pires RT, Tseng SC. Amniotic membrane
transplantation for severe neurotrophic corneal ulcers. Br J Ophthalmol 2000; 84: 826–33.
16. Choi YS, Kim JY, Wee WR, et al. Effect of the application of human amniotic membrane
on rabbit corneal wound healing after excimer laser photorefractive keratectomy. Cornea 1998; 17: 389–95.
17. Park WC, Tseng SCG. Modulation of acute inflammation and keratocyte death by suturing, blood and amniotic membrane
in PRK. Invest Ophthalmol Vis Sci 2000; 41: 2906–14.
18. Wang MX, Gray TB, Park WC, et al. Reduction in corneal haze and apoptosis by amniotic membrane
matrix in excimer laser photoablation in rabbits. J Cataract Refract Surg 2001; 27: 310–9.
Keywords:© 2002 Lippincott Williams & Wilkins, Inc.
Epithelial defect; Persistent epithelial defect; Amniotic membrane; Therapeutic contact lens; Soft contact lens; Bandage contact lens