- Email: [email protected]
Amniotic membrane transplantation for repair of leaking glaucoma filtering blebs
Amniotic Membrane Transplantation for Repair of Leaking Glaucoma Filtering Blebs DONALD L. BUDENZ, MD, KEITH BARTON, MD, AND SCHEFFER C. G. TSENG, MD, PHD
● PURPOSE:
To compare the safety and efficacy of human preserved amniotic membrane transplant with conjunctival advancement for repair of late-onset glaucoma filtering bleb leaks. ● METHOD: A prospective, randomized clinical trial compared amniotic membrane transplant with conjunctival advancement in patients with leaking glaucoma filtering blebs. Intraocular pressure, number of glaucoma medications, and reoperation for glaucoma or persistent or recurrent bleb–leak were compared in the two groups. Patients were followed for a minimum of 1 year. ● RESULTS: Mean intraocular pressure was the same at 6 (amniotic membrane transplant, 15.4 ⴞ 4.4, conjunctival advancement 14.1 ⴞ 6.4, P ⴝ 0.6), 12 (amniotic membrane transplant, 15.0 ⴞ 6.3, conjunctival advancement, 13.2 ⴞ 6.6, P ⴝ 0.5), and 24 (amniotic membrane transplant, 17.2 ⴞ 7.1, conjunctival advancement, 15.0 ⴞ 6.3, P ⴝ 0.6) months. The mean number of glaucoma medications in use was the same in the two groups at all time intervals. After an average follow-up of 19 months, there were seven failures in the amniotic membrane transplant group (two with persistent leaks that were unresponsive to further suturing, two with late-onset leaks, and three who required repeat glaucoma surgery) and none in the conjunctival advancement group. The cumulative survival rate for amniotic membrane transplant was 81% at 6 months, 74% at 1 year, and 46% at 2 years. The cumulative survival rate was 100% for conjunctival advancement throughout followup.
Accepted for publication Jun 7, 2000. From the Bascom Palmer Eye Institute, Department of Ophthalmology, University of Miami School of Medicine, Miami, Florida (Drs Budenz and Tseng), and Moorfields Eye Hospital, London, England (Dr Barton). This study was supported by an unrestricted grant from Research to Prevent Blindness, New York, NY. Presented as a paper at the American Glaucoma Society meeting, March 3, 2000, San Antonio, Texas, and as a paper at the Association for Research in Vision and Ophthalmology, May 2, 2000, Fort Lauderdale, Florida. Dr. Tseng has filed a patent on preparation and clinical uses of amniotic membrane and has a proprietary interest in Bio-Tissue. Correspondence to Donald L. Budenz, MD, Bascom Palmer Eye Institute, 900 N.W. 17th St, Miami, FL 33136; fax (305) 326-6337; e-mail: [email protected]
580
©
2000 BY
● CONCLUSIONS:
Amniotic membrane transplantation does not offer an effective alternative to conjunctival advancement for repair of leaking glaucoma filtering blebs. (Am J Ophthalmol 2000;130:580 –588. © 2000 by Elsevier Science Inc. All rights reserved.)
L
EAKAGE OF AQUEOUS HUMOR FROM CONJUNCTIVAL
filtering blebs may occur as an early or late complication of glaucoma filtration surgery. Early-onset leaks are usually managed successfully with observation, patching, bandage contact lens, direct suturing, or compression stitches.1–3 Late-onset leaks are more common after fullthickness glaucoma filters compared with trabeculectomy without antifibrotic agents.4 Trabeculectomies performed with 5-fluorouracil appear to have a higher rate of lateonset bleb leaks than those done without it,5,6 and trabeculectomies performed with mitomycin-C appear to have an even higher rate of late-onset bleb leaks than those performed with 5-fluorouracil.7,8 The frequency of lateonset filtering bleb leaks seems to be increasing, possibly because of increased use of these antifibrotic agents.1 Although many nonsurgical techniques have been tried to close late-onset bleb leaks, successful closure often requires surgical revision.1 Advancement of conjunctiva from the adjacent conjunctiva and/or Tenon layers9 –16 and free conjunctival autografts17,18 are the two most frequent techniques employed. However, these techniques are technically difficult or impossible in patients with large filtering blebs, and they may be associated with failure of filtration because of external scarring.2,19,20 In addition, conjunctival advancement may predispose to ptosis and diplopia as a result of interference with the levator superioris or superior rectus muscle complex. Preserved human amniotic membrane donor tissue transplantation is useful in the reconstruction of the ocular surface after pterygium excision21 and cicatrizing conjunctival disease.22,23 Amniotic membrane has been safely used to inhibit neovascularization before limbal stem cell transplantation,24 in the treatment of nonhealing corneal epithelial defects,25 and in the treatment of symptomatic bullous keratopathy.26 Amniotic membrane epithelializes relatively rapidly and integrates into the surrounding conjunctiva. In addition, certain properties of amniotic
ELSEVIER SCIENCE INC. ALL
RIGHTS RESERVED.
0002-9394/00/$20.00 PII S0002-9394(00)00600-0
membrane, such as high hydraulic conductivity,27 antifibrosis,28 and low immune response potential (because of lack of any living cells in the tissue), make it an attractive alternative to conjunctiva in the reconstruction of leaking glaucoma filtering blebs. In a recent animal study, amniotic membrane produced lower intraocular pressure than conjunctiva in the construction of filtering blebs.29 This study compared amniotic membrane transplant with conjunctival advancement, our standard technique of surgical repair of late-onset glaucoma filtering bleb leaks, in a prospective, randomized study.
METHODS AFTER APPROVAL BY THE HUMAN SUBJECTS COMMITTEES
of the Institutional Review Boards of the University of Miami School of Medicine and Moorfields Eye Hospital, consecutive patients who required surgical repair of leaking glaucoma filtering blebs beyond the 3-month postoperative period were invited to participate in the study after conservative methods failed to close the leak. After informed consent was obtained, patients were randomized to receive surgery with amnionic membrane transplant or conjunctival advancement. Both the patient and physician were masked to the treatment group until after randomization was assigned. All surgeries were performed by two glaucoma subspecialty trained surgeons (D.L.B., K.B.). The surgical procedure for conjunctival advancement has been described previously.16 The surgical procedure for the amnionic membrane transplant group was as follows. Surgery was performed under local anesthesia with monitored sedation. A paracentesis was performed temporally. A conjunctival peritomy was performed, including 2 to 3 mm of normal conjunctiva adjacent to the existing filtering bleb. The old filtering bleb was completely excised, and the conjunctiva adjacent to the previous bleb was undermined 2 to 3 mm. Sutures were added to the trabeculectomy flap if excessive flow or a shallow anterior chamber was noted. If excessive flow persisted, a donor scleral patch graft was used to further limit flow. Surgeons were required to report the number of sutures added to the trabeculectomy flap and whether a scleral patch was used. The peripheral corneal epithelium was debrided in the area of the conjunctival defect for a distance of 2 to 3 mm with a #64 blade or bipolar cautery to allow better adhesion of the tissue. Preserved human amniotic membrane (Biotissue; Miami, Florida), prepared as previously described,21–23 was cut into a rectangular or square of adequate length and width to cover the defect in the conjunctiva and extend at least 2 mm under the surrounding conjunctiva and onto the peripheral cornea. The membrane was inserted underneath healthy conjunctiva. The inferior two corners of the membrane were sutured to the peripheral cornea, using an interrupted 9-0 nylon suture for fixation during suturing. VOL. 130, NO. 5
An 8-0 polyglactin suture on a vascular needle (Ethicon BV-130; Westwood, Massachusetts) was used to suture the amniotic membrane to the underside of the conjunctiva in a continuous running closure. The anterior edge of the membrane was sutured to the Bowman membrane using several different techniques. The first two patients’ inferior closure was performed with a several 10-0 nylon mattress sutures. However, early leakage through the suture holes in the amniotic membrane (none of which persisted or required resuturing) precipitated a change in technique. Six subsequent cases were performed using a 9-0 nylon compression suture,20 which stretched across the entire anterior edge of the membrane-corneal interface. Seven cases were performed using a partial-thickness corneal groove into which the anterior edge of the membrane was placed. Horizontal mattress sutures were then used to close the cornea over the membrane. The paracentesis tract was used to inject balanced salt solution into the anterior chamber to confirm flow into the newly created filtering bleb and to check for leakage with the Seidel test. All leaks were treated with additional suturing until Seidel testing was negative. Subconjunctival injections of antibiotics and steroids were given in the inferior conjunctiva. The eye was patched until the next morning, after which steroid drops were administered four times per day, followed by a tapering schedule. Antibiotic drops were given four times per day for 1 week. Patients were examined at 1 day, 1 week, 1 month, 3 months, 6 months, and every 6 months after for a minimum follow-up of 1 year for each patient (protocol visits). Additional visits were performed at the discretion of the operating surgeon. Visual acuity, intraocular pressure, and slit-lamp examination with Seidel testing were performed at each protocol visit. Blebs were characterized in terms of bleb vascularity (0 to 4), height (0 to 4), and percent epithelialization after staining with fluorescein. Medications were added if the intraocular pressure was considered too high for the amount of optic nerve damage and/or visual field loss. Repeat glaucoma filtering surgery was performed for patients who failed medical therapy. Attempts were made to resuture patients who had persistent or recurrent bleb leaks. If these attempts were unsuccessful, patients underwent either bleb revision with conjunctival advancement and/or repeat glaucoma surgery. Results are given as mean ⫾ SD, except where indicated. Categorical variables were compared using the 2 or Fisher exact test, as indicated. Continuous variables were compared using the nonpaired Student t test. Differences were considered statistically significant if they reached the P ⬍ .05 level. Kaplan-Meier survival analysis and curves were performed. Success was defined as intraocular pressure of 21 mm Hg with or without glaucoma medications and with no need for glaucoma reoperation. The endpoint was defined as sustained (⬎2 visits) intraocular pressure of more than 21 mm Hg or reoperation for glaucoma or persistent bleb leak. Patients who required minor interven-
BUDENZ-AMNIOTIC MEMBRANE
FOR
LEAKING BLEBS
581
TABLE 1. Demographic and Preoperative Clinical Characteristics
Age (years) Sex (M/F) Race (W/H/B)‡ Antifibrotic agent None 5-Fluorouracil Mitomycin C
Amniotic Membrane (n ⫽ 15)
Conjunctival Advancement (n ⫽ 15)
P Value
61.7 ⫾ 13.9 5/10 4/8/3
61.4 ⫾ 8.5 8/7 3/9/3
0.9* 0.3† 0.9§
1 3 11
0 5 10
TABLE 2. Indications for Revision and Interventions Before Revision
Indications for revision Seidel positive leak Previous bleb-related infection Hypotony (IOP ⬍ 5 mm Hg) Corneal edema Hypotony maculopathy Choroidal effusion Epiphora Dysesthesia Interventions before revision Compression suturing Autologous blood injection Contact lens Tissue adhesive
0.6§
*Unpaired Student t test. † 2 Test of significance. ‡ W/H/B: White/Hispanic/black. § Fisher exact test.
tion for elevated intraocular pressure (laser suture lysis, needling) or persistent or recurrent leakage (suturing) were not considered failures unless major surgical intervention was then required under complete failure. Once patients reached the endpoint, subsequent data collected were disregarded in all analyses, and the follow-up time for each patient stopped at the endpoint or the actual last follow-up visit, whichever came first. Best-corrected visual acuity was converted to a logMar scale for analysis. All statistical analyses were performed using SPSS for Windows (SPSS Inc; Chicago, Illinois, Release 9.0).
15 8
15 5
5
5
2 2 0 1 2
2 1 2 0 2
2 2 1 1
3 0 1 0
Patients were followed for an average of 19.5 ⫾ 6.9 months (range, 12 to 32 months). As shown in Table 3, follow-up time was not statistically different between groups. The mean intraocular pressure and average number of glaucoma medications over time are presented in Tables 3 and 4 and Figures 1 and 2. Although intraocular pressure was lower in the amnionic membrane transplant group than the conjunctival advancement group at weeks 1 (P ⫽ 0.03) and 4 (P ⫽ 0.04), there was no difference in intraocular pressure during the 24-month follow-up period. There were no differences in the number of glaucoma medications in use at any time interval between the two groups. Seven of 15 patients in the amnionic membrane transplant group failed during the study period. Two patients had persistent leakage at the limbus postoperatively because of retraction of the membrane, which was unresponsive to multiple sutures in the minor operating room. In both of these patients, the amniotic membrane was attached to the peripheral cornea using the compression suture technique, and both were from the same donor. These two patients were considered failures on postoperative day 1, because the leak was never closed. One of these two patients had subsequent successful conjunctival advancement. The second underwent conjunctival advancement, which then developed persistent leakage because of retraction of the conjunctiva at the limbus. This patient ultimately required closure of the trabeculectomy with a donor scleral patch graft combined with a glaucoma drainage implant. Two additional patients in the amnionic membrane transplant group developed late-onset bleb leaks that failed to resolve. One patient developed a
A TOTAL OF 30 PATIENTS WERE INCLUDED IN THE STUDY.
Fifteen received amnionic membrane transplant and 15 conjunctival advancement. Table 1 provides demographic and clinical characteristics. There was no difference in antifibrotic agent used at initial trabeculectomy between groups (P ⫽ 0.6). The average time between initial glaucoma surgery and revision was 40.6 ⫾ 22.9 months in the amnionic membrane transplant group and 30.6 ⫾ 10.8 months in the conjunctival advancement group (P ⫽ 0.15). Intraoperatively, five patients in the amnionic membrane transplant group required one to four additional sutures in the trabeculectomy flap, whereas two patients in the conjunctival advancement group required four additional sutures. Two patients in the amnionic membrane transplant group required a donor scleral patch graft at the time of revision, whereas one in the conjunctival advancement group required a scleral patch graft. Table 2 lists the indications for bleb revision and interventions before bleb revision. All patients were followed for a minimum of 1 year. AMERICAN JOURNAL
Conjunctival Advancement
IOP ⫽ Intraocular pressure.
RESULTS
582
Amniotic Membrane
OF
OPHTHALMOLOGY
NOVEMBER 2000
TABLE 3. Intraocular Pressure*
Preoperative
Amniotic membrane transplant Conjunctival advancement P value†
7.2 ⫾ 3.6 (15) 5.6 ⫾ 3.1 (15) 0.2
1 Week
1 Month
3 Months
6 Months
12 Months
18 Months
24 Months
Last Followup
Follow-up (Months)
5.4 ⫾ 2.9 12.1 ⫾ 6.6 17.1 ⫾ 8.0 15.4 ⫾ 4.4 15.0 ⫾ 6.3 13.4 ⫾ 4.8 17.2 ⫾ 7.1 13.9 ⫾ 6.4 20.0 ⫾ 6.4 (13) (13) (12) (12) (11) (10) (5) (15) (15) 12.2 ⫾ 9.8 17.1 ⫾ 5.3 13.5 ⫾ 4.9 14.2 ⫾ 6.4 13.2 ⫾ 6.6 14.0 ⫾ 4.9 15.0 ⫾ 6.3 13.0 ⫾ 5.8 19.2 ⫾ 7.5 (15) (15) (15) (15) (14) (10) (7) (15) (15) 0.03 0.04 0.2 0.6 0.5 0.8 0.6 0.9 0.8
Numbers in parentheses indicate number of patients who had not reached the endpoint or were still available for follow-up at a given time interval. *Values given as mean ⫾ SD. † Unpaired Student t test.
TABLE 4. Number of Glaucoma Medications*
Amniotic membrane transplant Conjunctival advancement P value†
Preoperative
1 Week
1 Month
3 Months
6 Months
12 Months
18 Months
24 Months
Last Follow-up
0 (15) 0 (15) –
0 (13) 0 (15) –
0 (13) 0.3 ⫾ 0.7 (15) 0.1
0.8 ⫾ 1.1 (12) 0.6 ⫾ 0.8 (15) 0.6
0.8 ⫾ 1.3 (12) 0.8 ⫾ 1.2 (15) 0.9
1.3 ⫾ 1.6 (11) 0.6 ⫾ 1.2 (14) 0.2
1.7 ⫾ 1.9 (10) 1.1 ⫾ 1.4 (10) 0.5
2.6 ⫾ 1.7 (5) 1.8 ⫾ 1.6 (7) 0.5
1.5 ⫾ 1.9 (15) 0.8 ⫾ 1.4 (15) 0.3
Numbers in parentheses indicate the number of patients who had not reached endpoint or were still available for follow-up at a given time interval. *Values given as mean ⫾ SD. † Unpaired Student t test.
late-onset bleb leak 17 months after amnionic membrane transplant, which did not respond to aqueous suppressant therapy. This patient underwent closure of the trabeculectomy with a scleral patch graft and a glaucoma drainage implant. The second patient developed a lateonset leak at the limbus 9 months postoperatively, which persisted for 3 months, after which the patient was lost to follow-up. Three patients in the amnionic membrane transplant group required repeat glaucoma surgery for uncontrolled intraocular pressure during the study period. Two patients underwent glaucoma drainage implant surgery (at 2 months and 2 years) and one repeat trabeculectomy with mitomycin C (2 years). All three patients had low, avascular filtering blebs at the time of failure, suggesting that the reason for failure was scarring of the scleral flap rather than external bleb failure from scarring of the amniotic membrane. A single patient in the conjunctival advancement group underwent reoperation for glaucoma during the study period. Nine months postoperatively, the patient developed failure of a previous corneal transplant, which necessitated a repeat penetrating keratoplasty. A glaucoma drainage implant was placed during the penetrating keratoplasty, despite the fact that the patient’s intraocular pressure was 15 mm Hg on three glaucoma VOL. 130, NO. 5
medications. The patient was counted as a success for the purposes of the study, because the glaucoma surgery was performed as part of a nonglaucoma surgery and the patient was successful by intraocular pressure criteria. Subsequent data were censored for the purposes of the study. Figure 3 shows the survival curve for the two groups. The cumulative survival rate for amnionic membrane transplant was 81% at 6 months, 74% at 1 year, and 46% at 2 years. The cumulative survival rate was 100% for conjunctival advancement. The Kaplan-Meier survival analysis was highly significant (P ⬍ .00001). There were no intraoperative complications in either group. Early postoperative complications included transient shallowing of the anterior chamber, (two in the amnionic membrane transplant group and one in the conjunctival advancement group). One patient in the amnionic membrane transplant group also had transient choroidal effusions associated with the shallow anterior chamber. This patient developed a cataract that required removal 4 months postoperatively, then a retinal detachment 8 months postoperatively, which required pars plana vitrectomy and intraocular gas–fluid exchange. Despite these surgical interventions, the intraocular pressure remained 10 mm Hg throughout 2 years of follow-up. A
BUDENZ-AMNIOTIC MEMBRANE
FOR
LEAKING BLEBS
583
FIGURE 1. Plot of intraocular pressure (IOP) over time. Error bars indicate standard error of the mean.
FIGURE 2. Plot of number of glaucoma medications in use over time. Error bars indicate standard error of the mean.
single patient in the amniotic membrane group developed transient hypotony maculopathy, and one patient in the conjunctival advancement group developed aqueous misdirection. One patient in the conjunctival advancement group developed a Tenon cyst that responded to aqueous suppressants. Two patients in the conjunctival advancement group developed ptosis that persisted; one of these patients also developed symptomatic diplopia, which was transient. Four patients in each group were positive by Seidel testing on postoperative day 1, but this resolved in all patients in the conjunctival advancement group by postoperative week 2. Postoperative interventions included laser suture lysis through the amniotic membrane in three cases. Two of these cases had sutures added at the time of revision that required cutting, and one had an autologous scleral patch added that required suture lysis. Three patients in the amnionic membrane transplant group required needling procedures for uncontrolled intraocular pressure. Two of these patients developed high, dome-shaped blebs that resembled Tenon cysts and responded to needling. The third patient developed a low bleb with a trabeculectomy flap that had scarred down. This patient went on to glaucoma drainage implant. Three patients in the amnionic membrane transplant group underwent resuturing at the limbus for persistent postoperative bleb leaks; one of these resolved and two persisted, requiring reoperation. One patient in the conjunctival advance584
AMERICAN JOURNAL
FIGURE 3. Kaplan-Meier survival curves for amniotic membrane transplantation and conjunctival advancement.
ment group required resuturing of a limbal leak 1 month postoperatively, followed by a contact lens, which was successful in resolving the leak. Six patients in the conjunctival advancement group required postoperative subconjunctival injections of 5-fluorouracil, and only one patient in the amnionic membrane transplant group required 5-fluorouracil. OF
OPHTHALMOLOGY
NOVEMBER 2000
TABLE 5. Bleb Height*
Amniotic membrane transplant Conjunctival advancement P value†
1 Week
1 Month
3 Months
6 Months
12 Months
18 Months
24 Months
Last Follow-up
2.1 ⫾ 1.0 (13) 1.3 ⫾ 0.9 (15) 0.04
2.0 ⫾ 0.9 (13) 1.8 ⫾ 0.9 (15) 0.5
1.6 ⫾ 0.8 (12) 1.7 ⫾ 0.8 (15) 0.9
1.8 ⫾ 0.8 (12) 1.8 ⫾ 1.0 (15) 0.9
1.5 ⫾ 0.7 (11) 1.8 ⫾ 1.2 (14) 0.6
1.1 ⫾ 0.7 (10) 2.0 ⫾ 0.8 (10) 0.1
1.3 ⫾ 0.5 (5) 1.7 ⫾ 0.6 (7) 0.4
1.6 ⫾ 0.7 (15) 1.9 ⫾ 1.0 (15) 0.5
*Values given as mean ⫾ SD. † Unpaired Student t test.
TABLE 6. Bleb Vascularity*
Amniotic membrane transplant Conjunctival advancement P value†
1 Week
1 Month
3 Months
6 Months
12 Months
18 Months
24 Months
Last Follow-up
0 (13)
0.3 ⫾ 0.6 (13)
0.7 ⫾ 1.0 (12)
0.6 ⫾ 0.8 (12)
0.8 ⫾ 1.2 (11)
1.9 ⫾ 1.6 (10)
1.3 ⫾ 1.5 (5)
0.9 ⫾ 1.0 (15)
2.3 ⫾ 0.9 (15) ⬍0.001
2.2 ⫾ 0.9 (15) ⬍0.001
1.8 ⫾ 0.8 (15) 0.007
1.5 ⫾ 1.2 (15) 0.05
1.9 ⫾ 0.9 (14) 0.6
1.3 ⫾ 1.5 (10) 0.6
2.0 ⫾ 1.4 (7) 0.6
1.3 ⫾ 1.3 (15) 0.5
*Values given as mean ⫾ SD. † Unpaired Student t test.
Preoperative visual acuity was 0.56 ⫾ 0.63 logMAR units (approximately 20/70 Snellen acuity) in the amnionic membrane transplant group and 0.77 ⫾ 0.73 (approximately 20/100 Snellen acuity) in the conjunctival advancement group (P ⫽ 0.4). At last follow-up, visual acuity was 0.51 ⫾ 0.53 (approximately 20/70) in the amnionic membrane transplant group and 0.72 ⫾ 0.81 (approximately 20/100) in the conjunctival advancement group (P ⫽ 0.4). Seven patients lost 2 or more lines of vision in the amnionic membrane transplant group versus five in the conjunctival advancement group, whereas three patients in each group gained 2 or more lines of vision. By 1 month postoperatively, 11 of 15 amniotic membranes (73%) were completely epithelialized and all membranes were completely epithelialized by 3 months. The conjunctival advancement blebs were completely epithelialized from postoperative day 1 onward. Table 5 demonstrates that the blebs were very similar in height throughout follow-up in the two groups. Blebs in the conjunctival advancement group were significantly more vascular than those in the amnionic membrane transplant group until the 12-month follow-up, when bleb vascularity became similar between groups for the remainder of follow-up (Table 6). Figures 4 and 5 show the clinical appearance of two patients who underwent successful VOL. 130, NO. 5
FIGURE 4. Appearance of successful amniotic membrane transplantation filtering bleb 2 years postoperatively. This eye maintained an ischemic filtering bleb and an intraocular pressure between 15 and 17 mm Hg throughout follow-up on two glaucoma medications.
amnionic membrane transplantation. Figure 6 shows the clinical appearance of a patient who underwent successful conjunctival advancement.
BUDENZ-AMNIOTIC MEMBRANE
FOR
LEAKING BLEBS
585
FIGURE 5. Appearance of successful amniotic membrane transplantation filtering bleb 1 year postoperatively. Patient had undergone argon laser suture lysis 3 months postoperatively for an intraocular pressure of 17 mm Hg, which lowered the intraocular pressure to 10 to 15 mm Hg through 20 months of follow-up on no glaucoma medications.
FIGURE 6. Appearance of successful conjunctival advancement 12 months postoperatively. Note lower, more vascular appearance than amniotic membrane blebs. Despite this, the patient’s intraocular pressure was maintained between 16 and 18 mm Hg on no medications.
DISCUSSION
method and had success in two of three cases, but follow-up time is unspecified in each of these studies. Wilensky14 reviewed outcomes of 14 leaks in 13 eyes of 12 patients; all five blebs that had bleb excision with conjunctival mobilization had resolution of the leak after a follow-up of 2.5 to 13 months, compared with five of nine leaks (56%) managed without surgery. In the most extensive study of conjunctival advancement for late-onset glaucoma filtering bleb leaks, Budenz and associates16 reported successful closure in 24 of 26 patients (92%) and adequate intraocular pressure control in 24 of 26 patients (92%) after an average follow-up of 19.6 months. Buxton and associates17 and Wilson and Kotas-Neumann18 described closure of late-onset bleb leaks using a free conjunctival autograft in three eyes (follow-up, 18 to 96 months) and four eyes (follow-up, 2 to 12 months), respectively. All eyes with leaking trabeculectomy blebs were successfully repaired, and only one eye needed medications. This technique was not assessed in the current study, because of our success with conjunctival advancement.16 Because amniotic membrane epithelializes fairly rapidly, integrates into the surrounding conjunctiva, has high hydraulic conductivity,27 shows antifibrotic properties,28 and has a low immune response potential, it should be an ideal tissue for the repair of leaking glaucoma filtering blebs. Our two patients with persistent bleb leaks after surgery both had compression suture closure of the membrane at the bleb– cornea junction, which is where the retraction of the membrane occurred. This complication was not seen in subsequent cases, which were performed by suturing the membrane within a superficial keratectomy rather than with a compression suture. Perhaps the retraction would have been avoided in these patients if a different surgical technique had been used. Two patients
ALTHOUGH CONSERVATIVE METHODS FOR CLOSING LATE-
onset glaucoma filtering bleb leaks may be tried, many patients require surgical revision.1 To our knowledge, this is the first report on the use of nonocular tissue as a replacement for conjunctiva in filtration surgery. Fujishima and colleagues30 reported their experience using amniotic membrane underneath the scleral flap to inhibit scarring of the trabeculectomy flap to the scleral bed as a substitute for antifibrotic medications. Conjunctiva was then placed over the filtering site as in standard trabeculectomy. Conjunctival or Tenon layer advancement and conjunctival autografts have historically been used for the repair of leaking filtering blebs that do not respond to conservative treatments. Dunnington and Regan9 first described a series of 15 patients with either inadvertent (N ⫽ 13) or intentional (N ⫽ 2) filtering blebs who underwent conjunctival advancement for bleb leaks. Sugar10 used a similar method in five eyes; two eyes required more medications, one eye still had hypotony and cataract progression despite successful repair of the leak, and in one eye the conjunctival flap shrank and the leak recurred. In a later report, Sugar19 had little success using full-thickness conjunctival advancement for repair of acutely leaking blebs (one of seven eyes [14%] successful), compared with the use of a Tenon flap (four of seven eyes [57%] successful). Failure of conjunctival advancement usually occurred because of either retraction of the flap from the tension required for closure or scarring of the bleb with loss of function. Galin and Hung13 used a mattress suture technique to secure the conjunctiva to the peripheral cornea and reported successful repair in seven of eight cases (87.5%), and Cohen and colleagues12 used a similar 586
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
NOVEMBER 2000
developed spontaneous leaks in the amniotic membrane 9 and 17 months postoperatively. Additionally, the single pilot patient in whom amniotic membrane was used to repair a leaking filtering bleb developed a recurrent leak 18 months postoperatively. These failures are not attributable to surgical technique. In 1940, De Roth31 first used live fetal membrane (amnion and chorion) for the repair of conjunctival symblepharon and conjunctival defects. He reported that five of six patients developed either shrinkage or dissolution over time. When used as an alternative to conjunctival flaps for symptomatic bullous keratopathy, Pires and colleagues26 noted either complete or partial dissolution of the membrane after an average follow-up of 34 weeks. Given these results and those of the current study, we are concerned about the long-term integrity of the membrane in glaucoma filtering surgery. Unlike amniotic membrane used for conjunctival surface reconstruction, which is primarily structural, amniotic membrane in glaucoma filtering surgery has a functional role, which is to allow filtration while maintaining a watertight seal. Amniotic membrane may not heal as well as conjunctiva when exposed to aqueous humor contained in filtering blebs. The mature amniotic membrane bleb is typically avascular and similar in appearance to a bleb that has received antimetabolites; therefore, it may be prone to late leakage in the same manner as mitomycin blebs. In addition, this study suggests a higher rate of bleb failure necessitating needling or reoperation for uncontrolled intraocular pressure when bleb revision is performed with amniotic membrane. Six patients in the conjunctival advancement group received 5-fluorouracil postoperatively for signs of early bleb scarring, and only one patient in the amnionic membrane transplant group received 5-fluorouracil. The three bleb failures in the amnionic membrane transplant group occurred late postoperatively (one at 2 months and two at 2 years), and it is unlikely that 5-fluorouracil would have made a difference this late in the postoperative course. These three failures were unrelated to surgical technique or properties of the membrane, because all failed as a result of scarring of the trabeculectomy flap rather than external scarring of the amniotic membrane. These three patients may very well have failed if conjunctival advancement had been performed. The 100% success rate for conjunctival advancement obtained in this study is slightly higher than the 92% success rate reported in the only retrospective analysis of this procedure for a large group of patients with a similar follow-up time.16 When assessing any new therapeutic intervention, it is critical to compare efficacy and complications with a standard procedure. For treating late-onset bleb leaks, we adopted conjunctival advancement as the standard procedure, which, in this study, proved to be superior to the new treatment modality. Despite the relatively small sample size in the current study, the number of failures in the amnionic membrane transplant group suggests that this VOL. 130, NO. 5
material may not be a suitable substitute for conjunctiva in the repair of leaking filtering blebs, at least using the current surgical technique. Further study to determine whether alternative surgical techniques may improve the success rate of amniotic membrane repair of leaking glaucoma filtering blebs may be warranted.
REFERENCES 1. Ritch R, Schuman JS, Belcher CD III. Cases in controversy. Management of the leaking filtration bleb. J Glaucoma 1993;2:114 –118. 2. Azuara-Blanco A, Katz LJ. Dysfunctional filtering blebs. Surv Ophthalmol 1998;43:93–126. 3. Palmberg P. Late complications after glaucoma filtering surgery. In: Leader BJ, Calkwood JC. Proceedings of the 45th Annual Symposium of the New Orleans Academy of Ophthalmology, The Hague/The Netherlands, Kugler Publications, 1998:183–193. 4. Lamping KA, Bellows AR, Hutchinson BT, Afran SI. Longterm evaluation of initial filtration surgery. Ophthalmology 1986;93:91–101. 5. Fluorouracil Filtering Surgery Study Group. Fluorouracil filtering surgery study one year follow-up. Am J Ophthalmol 1989;108:625–35. 6. Liebmann JM, Ritch R, Marmor M, Nunez J, Wolner B. Initial 5-fluorouracil trabeculectomy in uncomplicated glaucoma. Ophthalmology 1991;98:1036 –1041. 7. Skuta GL, Beeson CC, Higginbotham EJ. Intraoperative mitomycin versus postoperative 5-fluorouracil in high-risk glaucoma filtering surgery. Ophthalmology 1992;99:438 – 444. 8. Greenfield DS, Liebmann JM, Jee J, Ritch R. Late-onset bleb leaks after glaucoma filtering surgery. Arch Ophthalmol 1998;116:443– 447. 9. Dunnington JH, Regan EF. Late fistulization of operative wounds. Arch Ophthalmol 1950;43:407– 418. 10. Sugar HS. Complications, repair and reoperation of antiglaucoma filtering blebs. Am J Ophthalmol 1967;63:825– 833. 11. McCulloch C. Surgery of filtering blebs. Int Ophthalmol Clin 1967;1:125–134. 12. Cohen JS, Shaffer RN, Hetherington J, Hoskins D. Revision of filtration surgery. Arch Ophthalmol 1977;95:1612– 1615. 13. Galin MA, Hung PT. Surgical repair of leaking blebs. Am J Ophthalmol 1977;83:328 –333. 14. Wilensky JT. Management of late bleb leaks following glaucoma filtering surgery. Tr Am Ophthalmol Soc 1992;90: 161–168. 15. O’Connor DJ, Tressler CS, Caprioli J. A surgical method to repair leaking filtering blebs. Ophthalmic Surg 1992;23:336 – 338. 16. Budenz DL, Chen PP, Weaver YK. Conjunctival advancement for late-onset filtering bleb leaks: Indications and outcomes. Arch Ophthalmol 1999;117:1014 –1019. 17. Buxton JN, Lavery KT, Liebmann JM, Buxton DF, Ritch R. Reconstruction of filtering blebs with free conjunctival autografts. Ophthalmology 1994;101:635– 639. 18. Wilson MR, Kotas-Neumann R. Free conjunctival patch for repair of persistent late bleb leak. Am J Ophthalmol 1994; 117:569 –574. 19. Sugar HS. Treatment of hypotony following filtering surgery for glaucoma. Am J Ophthalmol 1971;71:1023–33.
BUDENZ-AMNIOTIC MEMBRANE
FOR
LEAKING BLEBS
587
20. Palmberg P. Surgery for Complications. In: Albert DM, editor. Ophthalmic surgery: principles and techniques, vol 1. Malden, MA: Blackwell Science, 1999:476 – 491. 21. Prabhasawat P, Barton K, Burkett G, Tseng SCG. Comparison of conjunctival autografts and amniotic membrane grafts for pterygium excision. Ophthalmology 1997;104: 974 –985. 22. Tsubota K, Satake Y, Ohyama M, et al. Surgical reconstruction of the ocular surface in advanced cicatricial pemphigoid and Stevens-Johnson Syndrome. Am J Ophthalmol 1996; 122:38 –52. 23. Tseng SCG, Prabhasawat P, Lee S. Amniotic membrane transplantation for conjunctival surface reconstruction. Am J Ophthalmol 1997;124:765–774. 24. Tseng SCG, Prabhasawat P, Barton K, Gray TB, Meller D. Amniotic membrane transplantation with or without limbal autografts for corneal surface reconstruction in patients with limbal stem cell deficiency. Arch Ophthalmol 1998;116: 431– 441. 25. Lee S-H, Tseng SCG. Amniotic membrane for persistent
26. 27. 28.
29. 30. 31.
epithelial defects with ulceration. Am J Ophthalmol 1997; 123:303–312. Pires RTF, Tseng SCG, Pabhasawat P, et al. Amniotic membrane transplantation for symptomatic bullous keratopathy. Arch Ophthalmol 1999;117:1291–1297. Lloyd SJ, Garlid KD, Reba RC, Seeds EA. Permeability of different layers of the human placenta to isotopic water. J Appl Physiol 1969;26:274 –276. Tseng SCG, Li D-Q, Ma X. Suppression of TGF-isoforms, TGF-receptor type II and myofibroblast differentiation in cultured human corneal and limbal fibroblasts by amniotic membrane matrix. J Cell Physiol 1999;179:325–335. Barton K, Budenz DL, Khaw PT, Tseng SCG. Amniotic membrane transplantation in glaucoma surgery. Invest Ophthalmol Vis Sci. 38:S473, 1997. Fujishima H, Shimazaki J, Shinozaki N, Tsubota K. Trabeculectomy with the use of amniotic membrane for uncontrolled glaucoma. Ophthalmic Surg Lasers 29:428, 1998. De Roth A. Plastic repair of conjunctival defects with fetal membrane. Arch Ophthalmol 1940;134:833– 845.
The full-text of AJO is now available online at www.ajo.com. Authors Interactive威, currently available in limited form, is undergoing an upgrade.
588
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
NOVEMBER 2000
● PURPOSE:
To compare the safety and efficacy of human preserved amniotic membrane transplant with conjunctival advancement for repair of late-onset glaucoma filtering bleb leaks. ● METHOD: A prospective, randomized clinical trial compared amniotic membrane transplant with conjunctival advancement in patients with leaking glaucoma filtering blebs. Intraocular pressure, number of glaucoma medications, and reoperation for glaucoma or persistent or recurrent bleb–leak were compared in the two groups. Patients were followed for a minimum of 1 year. ● RESULTS: Mean intraocular pressure was the same at 6 (amniotic membrane transplant, 15.4 ⴞ 4.4, conjunctival advancement 14.1 ⴞ 6.4, P ⴝ 0.6), 12 (amniotic membrane transplant, 15.0 ⴞ 6.3, conjunctival advancement, 13.2 ⴞ 6.6, P ⴝ 0.5), and 24 (amniotic membrane transplant, 17.2 ⴞ 7.1, conjunctival advancement, 15.0 ⴞ 6.3, P ⴝ 0.6) months. The mean number of glaucoma medications in use was the same in the two groups at all time intervals. After an average follow-up of 19 months, there were seven failures in the amniotic membrane transplant group (two with persistent leaks that were unresponsive to further suturing, two with late-onset leaks, and three who required repeat glaucoma surgery) and none in the conjunctival advancement group. The cumulative survival rate for amniotic membrane transplant was 81% at 6 months, 74% at 1 year, and 46% at 2 years. The cumulative survival rate was 100% for conjunctival advancement throughout followup.
Accepted for publication Jun 7, 2000. From the Bascom Palmer Eye Institute, Department of Ophthalmology, University of Miami School of Medicine, Miami, Florida (Drs Budenz and Tseng), and Moorfields Eye Hospital, London, England (Dr Barton). This study was supported by an unrestricted grant from Research to Prevent Blindness, New York, NY. Presented as a paper at the American Glaucoma Society meeting, March 3, 2000, San Antonio, Texas, and as a paper at the Association for Research in Vision and Ophthalmology, May 2, 2000, Fort Lauderdale, Florida. Dr. Tseng has filed a patent on preparation and clinical uses of amniotic membrane and has a proprietary interest in Bio-Tissue. Correspondence to Donald L. Budenz, MD, Bascom Palmer Eye Institute, 900 N.W. 17th St, Miami, FL 33136; fax (305) 326-6337; e-mail: [email protected]
580
©
2000 BY
● CONCLUSIONS:
Amniotic membrane transplantation does not offer an effective alternative to conjunctival advancement for repair of leaking glaucoma filtering blebs. (Am J Ophthalmol 2000;130:580 –588. © 2000 by Elsevier Science Inc. All rights reserved.)
L
EAKAGE OF AQUEOUS HUMOR FROM CONJUNCTIVAL
filtering blebs may occur as an early or late complication of glaucoma filtration surgery. Early-onset leaks are usually managed successfully with observation, patching, bandage contact lens, direct suturing, or compression stitches.1–3 Late-onset leaks are more common after fullthickness glaucoma filters compared with trabeculectomy without antifibrotic agents.4 Trabeculectomies performed with 5-fluorouracil appear to have a higher rate of lateonset bleb leaks than those done without it,5,6 and trabeculectomies performed with mitomycin-C appear to have an even higher rate of late-onset bleb leaks than those performed with 5-fluorouracil.7,8 The frequency of lateonset filtering bleb leaks seems to be increasing, possibly because of increased use of these antifibrotic agents.1 Although many nonsurgical techniques have been tried to close late-onset bleb leaks, successful closure often requires surgical revision.1 Advancement of conjunctiva from the adjacent conjunctiva and/or Tenon layers9 –16 and free conjunctival autografts17,18 are the two most frequent techniques employed. However, these techniques are technically difficult or impossible in patients with large filtering blebs, and they may be associated with failure of filtration because of external scarring.2,19,20 In addition, conjunctival advancement may predispose to ptosis and diplopia as a result of interference with the levator superioris or superior rectus muscle complex. Preserved human amniotic membrane donor tissue transplantation is useful in the reconstruction of the ocular surface after pterygium excision21 and cicatrizing conjunctival disease.22,23 Amniotic membrane has been safely used to inhibit neovascularization before limbal stem cell transplantation,24 in the treatment of nonhealing corneal epithelial defects,25 and in the treatment of symptomatic bullous keratopathy.26 Amniotic membrane epithelializes relatively rapidly and integrates into the surrounding conjunctiva. In addition, certain properties of amniotic
ELSEVIER SCIENCE INC. ALL
RIGHTS RESERVED.
0002-9394/00/$20.00 PII S0002-9394(00)00600-0
membrane, such as high hydraulic conductivity,27 antifibrosis,28 and low immune response potential (because of lack of any living cells in the tissue), make it an attractive alternative to conjunctiva in the reconstruction of leaking glaucoma filtering blebs. In a recent animal study, amniotic membrane produced lower intraocular pressure than conjunctiva in the construction of filtering blebs.29 This study compared amniotic membrane transplant with conjunctival advancement, our standard technique of surgical repair of late-onset glaucoma filtering bleb leaks, in a prospective, randomized study.
METHODS AFTER APPROVAL BY THE HUMAN SUBJECTS COMMITTEES
of the Institutional Review Boards of the University of Miami School of Medicine and Moorfields Eye Hospital, consecutive patients who required surgical repair of leaking glaucoma filtering blebs beyond the 3-month postoperative period were invited to participate in the study after conservative methods failed to close the leak. After informed consent was obtained, patients were randomized to receive surgery with amnionic membrane transplant or conjunctival advancement. Both the patient and physician were masked to the treatment group until after randomization was assigned. All surgeries were performed by two glaucoma subspecialty trained surgeons (D.L.B., K.B.). The surgical procedure for conjunctival advancement has been described previously.16 The surgical procedure for the amnionic membrane transplant group was as follows. Surgery was performed under local anesthesia with monitored sedation. A paracentesis was performed temporally. A conjunctival peritomy was performed, including 2 to 3 mm of normal conjunctiva adjacent to the existing filtering bleb. The old filtering bleb was completely excised, and the conjunctiva adjacent to the previous bleb was undermined 2 to 3 mm. Sutures were added to the trabeculectomy flap if excessive flow or a shallow anterior chamber was noted. If excessive flow persisted, a donor scleral patch graft was used to further limit flow. Surgeons were required to report the number of sutures added to the trabeculectomy flap and whether a scleral patch was used. The peripheral corneal epithelium was debrided in the area of the conjunctival defect for a distance of 2 to 3 mm with a #64 blade or bipolar cautery to allow better adhesion of the tissue. Preserved human amniotic membrane (Biotissue; Miami, Florida), prepared as previously described,21–23 was cut into a rectangular or square of adequate length and width to cover the defect in the conjunctiva and extend at least 2 mm under the surrounding conjunctiva and onto the peripheral cornea. The membrane was inserted underneath healthy conjunctiva. The inferior two corners of the membrane were sutured to the peripheral cornea, using an interrupted 9-0 nylon suture for fixation during suturing. VOL. 130, NO. 5
An 8-0 polyglactin suture on a vascular needle (Ethicon BV-130; Westwood, Massachusetts) was used to suture the amniotic membrane to the underside of the conjunctiva in a continuous running closure. The anterior edge of the membrane was sutured to the Bowman membrane using several different techniques. The first two patients’ inferior closure was performed with a several 10-0 nylon mattress sutures. However, early leakage through the suture holes in the amniotic membrane (none of which persisted or required resuturing) precipitated a change in technique. Six subsequent cases were performed using a 9-0 nylon compression suture,20 which stretched across the entire anterior edge of the membrane-corneal interface. Seven cases were performed using a partial-thickness corneal groove into which the anterior edge of the membrane was placed. Horizontal mattress sutures were then used to close the cornea over the membrane. The paracentesis tract was used to inject balanced salt solution into the anterior chamber to confirm flow into the newly created filtering bleb and to check for leakage with the Seidel test. All leaks were treated with additional suturing until Seidel testing was negative. Subconjunctival injections of antibiotics and steroids were given in the inferior conjunctiva. The eye was patched until the next morning, after which steroid drops were administered four times per day, followed by a tapering schedule. Antibiotic drops were given four times per day for 1 week. Patients were examined at 1 day, 1 week, 1 month, 3 months, 6 months, and every 6 months after for a minimum follow-up of 1 year for each patient (protocol visits). Additional visits were performed at the discretion of the operating surgeon. Visual acuity, intraocular pressure, and slit-lamp examination with Seidel testing were performed at each protocol visit. Blebs were characterized in terms of bleb vascularity (0 to 4), height (0 to 4), and percent epithelialization after staining with fluorescein. Medications were added if the intraocular pressure was considered too high for the amount of optic nerve damage and/or visual field loss. Repeat glaucoma filtering surgery was performed for patients who failed medical therapy. Attempts were made to resuture patients who had persistent or recurrent bleb leaks. If these attempts were unsuccessful, patients underwent either bleb revision with conjunctival advancement and/or repeat glaucoma surgery. Results are given as mean ⫾ SD, except where indicated. Categorical variables were compared using the 2 or Fisher exact test, as indicated. Continuous variables were compared using the nonpaired Student t test. Differences were considered statistically significant if they reached the P ⬍ .05 level. Kaplan-Meier survival analysis and curves were performed. Success was defined as intraocular pressure of 21 mm Hg with or without glaucoma medications and with no need for glaucoma reoperation. The endpoint was defined as sustained (⬎2 visits) intraocular pressure of more than 21 mm Hg or reoperation for glaucoma or persistent bleb leak. Patients who required minor interven-
BUDENZ-AMNIOTIC MEMBRANE
FOR
LEAKING BLEBS
581
TABLE 1. Demographic and Preoperative Clinical Characteristics
Age (years) Sex (M/F) Race (W/H/B)‡ Antifibrotic agent None 5-Fluorouracil Mitomycin C
Amniotic Membrane (n ⫽ 15)
Conjunctival Advancement (n ⫽ 15)
P Value
61.7 ⫾ 13.9 5/10 4/8/3
61.4 ⫾ 8.5 8/7 3/9/3
0.9* 0.3† 0.9§
1 3 11
0 5 10
TABLE 2. Indications for Revision and Interventions Before Revision
Indications for revision Seidel positive leak Previous bleb-related infection Hypotony (IOP ⬍ 5 mm Hg) Corneal edema Hypotony maculopathy Choroidal effusion Epiphora Dysesthesia Interventions before revision Compression suturing Autologous blood injection Contact lens Tissue adhesive
0.6§
*Unpaired Student t test. † 2 Test of significance. ‡ W/H/B: White/Hispanic/black. § Fisher exact test.
tion for elevated intraocular pressure (laser suture lysis, needling) or persistent or recurrent leakage (suturing) were not considered failures unless major surgical intervention was then required under complete failure. Once patients reached the endpoint, subsequent data collected were disregarded in all analyses, and the follow-up time for each patient stopped at the endpoint or the actual last follow-up visit, whichever came first. Best-corrected visual acuity was converted to a logMar scale for analysis. All statistical analyses were performed using SPSS for Windows (SPSS Inc; Chicago, Illinois, Release 9.0).
15 8
15 5
5
5
2 2 0 1 2
2 1 2 0 2
2 2 1 1
3 0 1 0
Patients were followed for an average of 19.5 ⫾ 6.9 months (range, 12 to 32 months). As shown in Table 3, follow-up time was not statistically different between groups. The mean intraocular pressure and average number of glaucoma medications over time are presented in Tables 3 and 4 and Figures 1 and 2. Although intraocular pressure was lower in the amnionic membrane transplant group than the conjunctival advancement group at weeks 1 (P ⫽ 0.03) and 4 (P ⫽ 0.04), there was no difference in intraocular pressure during the 24-month follow-up period. There were no differences in the number of glaucoma medications in use at any time interval between the two groups. Seven of 15 patients in the amnionic membrane transplant group failed during the study period. Two patients had persistent leakage at the limbus postoperatively because of retraction of the membrane, which was unresponsive to multiple sutures in the minor operating room. In both of these patients, the amniotic membrane was attached to the peripheral cornea using the compression suture technique, and both were from the same donor. These two patients were considered failures on postoperative day 1, because the leak was never closed. One of these two patients had subsequent successful conjunctival advancement. The second underwent conjunctival advancement, which then developed persistent leakage because of retraction of the conjunctiva at the limbus. This patient ultimately required closure of the trabeculectomy with a donor scleral patch graft combined with a glaucoma drainage implant. Two additional patients in the amnionic membrane transplant group developed late-onset bleb leaks that failed to resolve. One patient developed a
A TOTAL OF 30 PATIENTS WERE INCLUDED IN THE STUDY.
Fifteen received amnionic membrane transplant and 15 conjunctival advancement. Table 1 provides demographic and clinical characteristics. There was no difference in antifibrotic agent used at initial trabeculectomy between groups (P ⫽ 0.6). The average time between initial glaucoma surgery and revision was 40.6 ⫾ 22.9 months in the amnionic membrane transplant group and 30.6 ⫾ 10.8 months in the conjunctival advancement group (P ⫽ 0.15). Intraoperatively, five patients in the amnionic membrane transplant group required one to four additional sutures in the trabeculectomy flap, whereas two patients in the conjunctival advancement group required four additional sutures. Two patients in the amnionic membrane transplant group required a donor scleral patch graft at the time of revision, whereas one in the conjunctival advancement group required a scleral patch graft. Table 2 lists the indications for bleb revision and interventions before bleb revision. All patients were followed for a minimum of 1 year. AMERICAN JOURNAL
Conjunctival Advancement
IOP ⫽ Intraocular pressure.
RESULTS
582
Amniotic Membrane
OF
OPHTHALMOLOGY
NOVEMBER 2000
TABLE 3. Intraocular Pressure*
Preoperative
Amniotic membrane transplant Conjunctival advancement P value†
7.2 ⫾ 3.6 (15) 5.6 ⫾ 3.1 (15) 0.2
1 Week
1 Month
3 Months
6 Months
12 Months
18 Months
24 Months
Last Followup
Follow-up (Months)
5.4 ⫾ 2.9 12.1 ⫾ 6.6 17.1 ⫾ 8.0 15.4 ⫾ 4.4 15.0 ⫾ 6.3 13.4 ⫾ 4.8 17.2 ⫾ 7.1 13.9 ⫾ 6.4 20.0 ⫾ 6.4 (13) (13) (12) (12) (11) (10) (5) (15) (15) 12.2 ⫾ 9.8 17.1 ⫾ 5.3 13.5 ⫾ 4.9 14.2 ⫾ 6.4 13.2 ⫾ 6.6 14.0 ⫾ 4.9 15.0 ⫾ 6.3 13.0 ⫾ 5.8 19.2 ⫾ 7.5 (15) (15) (15) (15) (14) (10) (7) (15) (15) 0.03 0.04 0.2 0.6 0.5 0.8 0.6 0.9 0.8
Numbers in parentheses indicate number of patients who had not reached the endpoint or were still available for follow-up at a given time interval. *Values given as mean ⫾ SD. † Unpaired Student t test.
TABLE 4. Number of Glaucoma Medications*
Amniotic membrane transplant Conjunctival advancement P value†
Preoperative
1 Week
1 Month
3 Months
6 Months
12 Months
18 Months
24 Months
Last Follow-up
0 (15) 0 (15) –
0 (13) 0 (15) –
0 (13) 0.3 ⫾ 0.7 (15) 0.1
0.8 ⫾ 1.1 (12) 0.6 ⫾ 0.8 (15) 0.6
0.8 ⫾ 1.3 (12) 0.8 ⫾ 1.2 (15) 0.9
1.3 ⫾ 1.6 (11) 0.6 ⫾ 1.2 (14) 0.2
1.7 ⫾ 1.9 (10) 1.1 ⫾ 1.4 (10) 0.5
2.6 ⫾ 1.7 (5) 1.8 ⫾ 1.6 (7) 0.5
1.5 ⫾ 1.9 (15) 0.8 ⫾ 1.4 (15) 0.3
Numbers in parentheses indicate the number of patients who had not reached endpoint or were still available for follow-up at a given time interval. *Values given as mean ⫾ SD. † Unpaired Student t test.
late-onset bleb leak 17 months after amnionic membrane transplant, which did not respond to aqueous suppressant therapy. This patient underwent closure of the trabeculectomy with a scleral patch graft and a glaucoma drainage implant. The second patient developed a lateonset leak at the limbus 9 months postoperatively, which persisted for 3 months, after which the patient was lost to follow-up. Three patients in the amnionic membrane transplant group required repeat glaucoma surgery for uncontrolled intraocular pressure during the study period. Two patients underwent glaucoma drainage implant surgery (at 2 months and 2 years) and one repeat trabeculectomy with mitomycin C (2 years). All three patients had low, avascular filtering blebs at the time of failure, suggesting that the reason for failure was scarring of the scleral flap rather than external bleb failure from scarring of the amniotic membrane. A single patient in the conjunctival advancement group underwent reoperation for glaucoma during the study period. Nine months postoperatively, the patient developed failure of a previous corneal transplant, which necessitated a repeat penetrating keratoplasty. A glaucoma drainage implant was placed during the penetrating keratoplasty, despite the fact that the patient’s intraocular pressure was 15 mm Hg on three glaucoma VOL. 130, NO. 5
medications. The patient was counted as a success for the purposes of the study, because the glaucoma surgery was performed as part of a nonglaucoma surgery and the patient was successful by intraocular pressure criteria. Subsequent data were censored for the purposes of the study. Figure 3 shows the survival curve for the two groups. The cumulative survival rate for amnionic membrane transplant was 81% at 6 months, 74% at 1 year, and 46% at 2 years. The cumulative survival rate was 100% for conjunctival advancement. The Kaplan-Meier survival analysis was highly significant (P ⬍ .00001). There were no intraoperative complications in either group. Early postoperative complications included transient shallowing of the anterior chamber, (two in the amnionic membrane transplant group and one in the conjunctival advancement group). One patient in the amnionic membrane transplant group also had transient choroidal effusions associated with the shallow anterior chamber. This patient developed a cataract that required removal 4 months postoperatively, then a retinal detachment 8 months postoperatively, which required pars plana vitrectomy and intraocular gas–fluid exchange. Despite these surgical interventions, the intraocular pressure remained 10 mm Hg throughout 2 years of follow-up. A
BUDENZ-AMNIOTIC MEMBRANE
FOR
LEAKING BLEBS
583
FIGURE 1. Plot of intraocular pressure (IOP) over time. Error bars indicate standard error of the mean.
FIGURE 2. Plot of number of glaucoma medications in use over time. Error bars indicate standard error of the mean.
single patient in the amniotic membrane group developed transient hypotony maculopathy, and one patient in the conjunctival advancement group developed aqueous misdirection. One patient in the conjunctival advancement group developed a Tenon cyst that responded to aqueous suppressants. Two patients in the conjunctival advancement group developed ptosis that persisted; one of these patients also developed symptomatic diplopia, which was transient. Four patients in each group were positive by Seidel testing on postoperative day 1, but this resolved in all patients in the conjunctival advancement group by postoperative week 2. Postoperative interventions included laser suture lysis through the amniotic membrane in three cases. Two of these cases had sutures added at the time of revision that required cutting, and one had an autologous scleral patch added that required suture lysis. Three patients in the amnionic membrane transplant group required needling procedures for uncontrolled intraocular pressure. Two of these patients developed high, dome-shaped blebs that resembled Tenon cysts and responded to needling. The third patient developed a low bleb with a trabeculectomy flap that had scarred down. This patient went on to glaucoma drainage implant. Three patients in the amnionic membrane transplant group underwent resuturing at the limbus for persistent postoperative bleb leaks; one of these resolved and two persisted, requiring reoperation. One patient in the conjunctival advance584
AMERICAN JOURNAL
FIGURE 3. Kaplan-Meier survival curves for amniotic membrane transplantation and conjunctival advancement.
ment group required resuturing of a limbal leak 1 month postoperatively, followed by a contact lens, which was successful in resolving the leak. Six patients in the conjunctival advancement group required postoperative subconjunctival injections of 5-fluorouracil, and only one patient in the amnionic membrane transplant group required 5-fluorouracil. OF
OPHTHALMOLOGY
NOVEMBER 2000
TABLE 5. Bleb Height*
Amniotic membrane transplant Conjunctival advancement P value†
1 Week
1 Month
3 Months
6 Months
12 Months
18 Months
24 Months
Last Follow-up
2.1 ⫾ 1.0 (13) 1.3 ⫾ 0.9 (15) 0.04
2.0 ⫾ 0.9 (13) 1.8 ⫾ 0.9 (15) 0.5
1.6 ⫾ 0.8 (12) 1.7 ⫾ 0.8 (15) 0.9
1.8 ⫾ 0.8 (12) 1.8 ⫾ 1.0 (15) 0.9
1.5 ⫾ 0.7 (11) 1.8 ⫾ 1.2 (14) 0.6
1.1 ⫾ 0.7 (10) 2.0 ⫾ 0.8 (10) 0.1
1.3 ⫾ 0.5 (5) 1.7 ⫾ 0.6 (7) 0.4
1.6 ⫾ 0.7 (15) 1.9 ⫾ 1.0 (15) 0.5
*Values given as mean ⫾ SD. † Unpaired Student t test.
TABLE 6. Bleb Vascularity*
Amniotic membrane transplant Conjunctival advancement P value†
1 Week
1 Month
3 Months
6 Months
12 Months
18 Months
24 Months
Last Follow-up
0 (13)
0.3 ⫾ 0.6 (13)
0.7 ⫾ 1.0 (12)
0.6 ⫾ 0.8 (12)
0.8 ⫾ 1.2 (11)
1.9 ⫾ 1.6 (10)
1.3 ⫾ 1.5 (5)
0.9 ⫾ 1.0 (15)
2.3 ⫾ 0.9 (15) ⬍0.001
2.2 ⫾ 0.9 (15) ⬍0.001
1.8 ⫾ 0.8 (15) 0.007
1.5 ⫾ 1.2 (15) 0.05
1.9 ⫾ 0.9 (14) 0.6
1.3 ⫾ 1.5 (10) 0.6
2.0 ⫾ 1.4 (7) 0.6
1.3 ⫾ 1.3 (15) 0.5
*Values given as mean ⫾ SD. † Unpaired Student t test.
Preoperative visual acuity was 0.56 ⫾ 0.63 logMAR units (approximately 20/70 Snellen acuity) in the amnionic membrane transplant group and 0.77 ⫾ 0.73 (approximately 20/100 Snellen acuity) in the conjunctival advancement group (P ⫽ 0.4). At last follow-up, visual acuity was 0.51 ⫾ 0.53 (approximately 20/70) in the amnionic membrane transplant group and 0.72 ⫾ 0.81 (approximately 20/100) in the conjunctival advancement group (P ⫽ 0.4). Seven patients lost 2 or more lines of vision in the amnionic membrane transplant group versus five in the conjunctival advancement group, whereas three patients in each group gained 2 or more lines of vision. By 1 month postoperatively, 11 of 15 amniotic membranes (73%) were completely epithelialized and all membranes were completely epithelialized by 3 months. The conjunctival advancement blebs were completely epithelialized from postoperative day 1 onward. Table 5 demonstrates that the blebs were very similar in height throughout follow-up in the two groups. Blebs in the conjunctival advancement group were significantly more vascular than those in the amnionic membrane transplant group until the 12-month follow-up, when bleb vascularity became similar between groups for the remainder of follow-up (Table 6). Figures 4 and 5 show the clinical appearance of two patients who underwent successful VOL. 130, NO. 5
FIGURE 4. Appearance of successful amniotic membrane transplantation filtering bleb 2 years postoperatively. This eye maintained an ischemic filtering bleb and an intraocular pressure between 15 and 17 mm Hg throughout follow-up on two glaucoma medications.
amnionic membrane transplantation. Figure 6 shows the clinical appearance of a patient who underwent successful conjunctival advancement.
BUDENZ-AMNIOTIC MEMBRANE
FOR
LEAKING BLEBS
585
FIGURE 5. Appearance of successful amniotic membrane transplantation filtering bleb 1 year postoperatively. Patient had undergone argon laser suture lysis 3 months postoperatively for an intraocular pressure of 17 mm Hg, which lowered the intraocular pressure to 10 to 15 mm Hg through 20 months of follow-up on no glaucoma medications.
FIGURE 6. Appearance of successful conjunctival advancement 12 months postoperatively. Note lower, more vascular appearance than amniotic membrane blebs. Despite this, the patient’s intraocular pressure was maintained between 16 and 18 mm Hg on no medications.
DISCUSSION
method and had success in two of three cases, but follow-up time is unspecified in each of these studies. Wilensky14 reviewed outcomes of 14 leaks in 13 eyes of 12 patients; all five blebs that had bleb excision with conjunctival mobilization had resolution of the leak after a follow-up of 2.5 to 13 months, compared with five of nine leaks (56%) managed without surgery. In the most extensive study of conjunctival advancement for late-onset glaucoma filtering bleb leaks, Budenz and associates16 reported successful closure in 24 of 26 patients (92%) and adequate intraocular pressure control in 24 of 26 patients (92%) after an average follow-up of 19.6 months. Buxton and associates17 and Wilson and Kotas-Neumann18 described closure of late-onset bleb leaks using a free conjunctival autograft in three eyes (follow-up, 18 to 96 months) and four eyes (follow-up, 2 to 12 months), respectively. All eyes with leaking trabeculectomy blebs were successfully repaired, and only one eye needed medications. This technique was not assessed in the current study, because of our success with conjunctival advancement.16 Because amniotic membrane epithelializes fairly rapidly, integrates into the surrounding conjunctiva, has high hydraulic conductivity,27 shows antifibrotic properties,28 and has a low immune response potential, it should be an ideal tissue for the repair of leaking glaucoma filtering blebs. Our two patients with persistent bleb leaks after surgery both had compression suture closure of the membrane at the bleb– cornea junction, which is where the retraction of the membrane occurred. This complication was not seen in subsequent cases, which were performed by suturing the membrane within a superficial keratectomy rather than with a compression suture. Perhaps the retraction would have been avoided in these patients if a different surgical technique had been used. Two patients
ALTHOUGH CONSERVATIVE METHODS FOR CLOSING LATE-
onset glaucoma filtering bleb leaks may be tried, many patients require surgical revision.1 To our knowledge, this is the first report on the use of nonocular tissue as a replacement for conjunctiva in filtration surgery. Fujishima and colleagues30 reported their experience using amniotic membrane underneath the scleral flap to inhibit scarring of the trabeculectomy flap to the scleral bed as a substitute for antifibrotic medications. Conjunctiva was then placed over the filtering site as in standard trabeculectomy. Conjunctival or Tenon layer advancement and conjunctival autografts have historically been used for the repair of leaking filtering blebs that do not respond to conservative treatments. Dunnington and Regan9 first described a series of 15 patients with either inadvertent (N ⫽ 13) or intentional (N ⫽ 2) filtering blebs who underwent conjunctival advancement for bleb leaks. Sugar10 used a similar method in five eyes; two eyes required more medications, one eye still had hypotony and cataract progression despite successful repair of the leak, and in one eye the conjunctival flap shrank and the leak recurred. In a later report, Sugar19 had little success using full-thickness conjunctival advancement for repair of acutely leaking blebs (one of seven eyes [14%] successful), compared with the use of a Tenon flap (four of seven eyes [57%] successful). Failure of conjunctival advancement usually occurred because of either retraction of the flap from the tension required for closure or scarring of the bleb with loss of function. Galin and Hung13 used a mattress suture technique to secure the conjunctiva to the peripheral cornea and reported successful repair in seven of eight cases (87.5%), and Cohen and colleagues12 used a similar 586
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
NOVEMBER 2000
developed spontaneous leaks in the amniotic membrane 9 and 17 months postoperatively. Additionally, the single pilot patient in whom amniotic membrane was used to repair a leaking filtering bleb developed a recurrent leak 18 months postoperatively. These failures are not attributable to surgical technique. In 1940, De Roth31 first used live fetal membrane (amnion and chorion) for the repair of conjunctival symblepharon and conjunctival defects. He reported that five of six patients developed either shrinkage or dissolution over time. When used as an alternative to conjunctival flaps for symptomatic bullous keratopathy, Pires and colleagues26 noted either complete or partial dissolution of the membrane after an average follow-up of 34 weeks. Given these results and those of the current study, we are concerned about the long-term integrity of the membrane in glaucoma filtering surgery. Unlike amniotic membrane used for conjunctival surface reconstruction, which is primarily structural, amniotic membrane in glaucoma filtering surgery has a functional role, which is to allow filtration while maintaining a watertight seal. Amniotic membrane may not heal as well as conjunctiva when exposed to aqueous humor contained in filtering blebs. The mature amniotic membrane bleb is typically avascular and similar in appearance to a bleb that has received antimetabolites; therefore, it may be prone to late leakage in the same manner as mitomycin blebs. In addition, this study suggests a higher rate of bleb failure necessitating needling or reoperation for uncontrolled intraocular pressure when bleb revision is performed with amniotic membrane. Six patients in the conjunctival advancement group received 5-fluorouracil postoperatively for signs of early bleb scarring, and only one patient in the amnionic membrane transplant group received 5-fluorouracil. The three bleb failures in the amnionic membrane transplant group occurred late postoperatively (one at 2 months and two at 2 years), and it is unlikely that 5-fluorouracil would have made a difference this late in the postoperative course. These three failures were unrelated to surgical technique or properties of the membrane, because all failed as a result of scarring of the trabeculectomy flap rather than external scarring of the amniotic membrane. These three patients may very well have failed if conjunctival advancement had been performed. The 100% success rate for conjunctival advancement obtained in this study is slightly higher than the 92% success rate reported in the only retrospective analysis of this procedure for a large group of patients with a similar follow-up time.16 When assessing any new therapeutic intervention, it is critical to compare efficacy and complications with a standard procedure. For treating late-onset bleb leaks, we adopted conjunctival advancement as the standard procedure, which, in this study, proved to be superior to the new treatment modality. Despite the relatively small sample size in the current study, the number of failures in the amnionic membrane transplant group suggests that this VOL. 130, NO. 5
material may not be a suitable substitute for conjunctiva in the repair of leaking filtering blebs, at least using the current surgical technique. Further study to determine whether alternative surgical techniques may improve the success rate of amniotic membrane repair of leaking glaucoma filtering blebs may be warranted.
REFERENCES 1. Ritch R, Schuman JS, Belcher CD III. Cases in controversy. Management of the leaking filtration bleb. J Glaucoma 1993;2:114 –118. 2. Azuara-Blanco A, Katz LJ. Dysfunctional filtering blebs. Surv Ophthalmol 1998;43:93–126. 3. Palmberg P. Late complications after glaucoma filtering surgery. In: Leader BJ, Calkwood JC. Proceedings of the 45th Annual Symposium of the New Orleans Academy of Ophthalmology, The Hague/The Netherlands, Kugler Publications, 1998:183–193. 4. Lamping KA, Bellows AR, Hutchinson BT, Afran SI. Longterm evaluation of initial filtration surgery. Ophthalmology 1986;93:91–101. 5. Fluorouracil Filtering Surgery Study Group. Fluorouracil filtering surgery study one year follow-up. Am J Ophthalmol 1989;108:625–35. 6. Liebmann JM, Ritch R, Marmor M, Nunez J, Wolner B. Initial 5-fluorouracil trabeculectomy in uncomplicated glaucoma. Ophthalmology 1991;98:1036 –1041. 7. Skuta GL, Beeson CC, Higginbotham EJ. Intraoperative mitomycin versus postoperative 5-fluorouracil in high-risk glaucoma filtering surgery. Ophthalmology 1992;99:438 – 444. 8. Greenfield DS, Liebmann JM, Jee J, Ritch R. Late-onset bleb leaks after glaucoma filtering surgery. Arch Ophthalmol 1998;116:443– 447. 9. Dunnington JH, Regan EF. Late fistulization of operative wounds. Arch Ophthalmol 1950;43:407– 418. 10. Sugar HS. Complications, repair and reoperation of antiglaucoma filtering blebs. Am J Ophthalmol 1967;63:825– 833. 11. McCulloch C. Surgery of filtering blebs. Int Ophthalmol Clin 1967;1:125–134. 12. Cohen JS, Shaffer RN, Hetherington J, Hoskins D. Revision of filtration surgery. Arch Ophthalmol 1977;95:1612– 1615. 13. Galin MA, Hung PT. Surgical repair of leaking blebs. Am J Ophthalmol 1977;83:328 –333. 14. Wilensky JT. Management of late bleb leaks following glaucoma filtering surgery. Tr Am Ophthalmol Soc 1992;90: 161–168. 15. O’Connor DJ, Tressler CS, Caprioli J. A surgical method to repair leaking filtering blebs. Ophthalmic Surg 1992;23:336 – 338. 16. Budenz DL, Chen PP, Weaver YK. Conjunctival advancement for late-onset filtering bleb leaks: Indications and outcomes. Arch Ophthalmol 1999;117:1014 –1019. 17. Buxton JN, Lavery KT, Liebmann JM, Buxton DF, Ritch R. Reconstruction of filtering blebs with free conjunctival autografts. Ophthalmology 1994;101:635– 639. 18. Wilson MR, Kotas-Neumann R. Free conjunctival patch for repair of persistent late bleb leak. Am J Ophthalmol 1994; 117:569 –574. 19. Sugar HS. Treatment of hypotony following filtering surgery for glaucoma. Am J Ophthalmol 1971;71:1023–33.
BUDENZ-AMNIOTIC MEMBRANE
FOR
LEAKING BLEBS
587
20. Palmberg P. Surgery for Complications. In: Albert DM, editor. Ophthalmic surgery: principles and techniques, vol 1. Malden, MA: Blackwell Science, 1999:476 – 491. 21. Prabhasawat P, Barton K, Burkett G, Tseng SCG. Comparison of conjunctival autografts and amniotic membrane grafts for pterygium excision. Ophthalmology 1997;104: 974 –985. 22. Tsubota K, Satake Y, Ohyama M, et al. Surgical reconstruction of the ocular surface in advanced cicatricial pemphigoid and Stevens-Johnson Syndrome. Am J Ophthalmol 1996; 122:38 –52. 23. Tseng SCG, Prabhasawat P, Lee S. Amniotic membrane transplantation for conjunctival surface reconstruction. Am J Ophthalmol 1997;124:765–774. 24. Tseng SCG, Prabhasawat P, Barton K, Gray TB, Meller D. Amniotic membrane transplantation with or without limbal autografts for corneal surface reconstruction in patients with limbal stem cell deficiency. Arch Ophthalmol 1998;116: 431– 441. 25. Lee S-H, Tseng SCG. Amniotic membrane for persistent
26. 27. 28.
29. 30. 31.
epithelial defects with ulceration. Am J Ophthalmol 1997; 123:303–312. Pires RTF, Tseng SCG, Pabhasawat P, et al. Amniotic membrane transplantation for symptomatic bullous keratopathy. Arch Ophthalmol 1999;117:1291–1297. Lloyd SJ, Garlid KD, Reba RC, Seeds EA. Permeability of different layers of the human placenta to isotopic water. J Appl Physiol 1969;26:274 –276. Tseng SCG, Li D-Q, Ma X. Suppression of TGF-isoforms, TGF-receptor type II and myofibroblast differentiation in cultured human corneal and limbal fibroblasts by amniotic membrane matrix. J Cell Physiol 1999;179:325–335. Barton K, Budenz DL, Khaw PT, Tseng SCG. Amniotic membrane transplantation in glaucoma surgery. Invest Ophthalmol Vis Sci. 38:S473, 1997. Fujishima H, Shimazaki J, Shinozaki N, Tsubota K. Trabeculectomy with the use of amniotic membrane for uncontrolled glaucoma. Ophthalmic Surg Lasers 29:428, 1998. De Roth A. Plastic repair of conjunctival defects with fetal membrane. Arch Ophthalmol 1940;134:833– 845.
The full-text of AJO is now available online at www.ajo.com. Authors Interactive威, currently available in limited form, is undergoing an upgrade.
588
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
NOVEMBER 2000