- Email: [email protected]
Long-term outcome of aqueous shunt surgery in ten patients with iridocorneal endothelial syndrome1
Long-term Outcome of Aqueous Shunt Surgery in Ten Patients with Iridocorneal Endothelial Syndrome David K. Kim, MD,1 Ioannis M. Aslanides, MD,2 Courtland M. Schmidt, Jr., MD,2 George L. Spaeth, MD,2 Richard P. Wilson, MD,2 James J. Augsburger, MD3 Purpose: To report the long-term outcome of ten patients with iridocorneal endothelial (ICE) syndrome who underwent aqueous shunt surgery for uncontrolled glaucoma. Design: Noncomparative, retrospective case series. Participants: The authors reviewed charts of ten patients with ICE syndrome-related glaucoma who underwent aqueous shunt surgery at one institution between 1987 and 1996. Main Outcome Measures: Intraocular pressure (IOP), number of glaucoma medications, and further surgical interventions were measured. Results: With a median follow-up of 55 months, four eyes had adequate IOP control (IOP ⬍ 21 mmHg) with one or two medications after the initial aqueous shunt surgery. An additional three eyes achieved adequate IOP control after one or more tube repositionings or revisions of the initial aqueous shunt. In this series, the aqueous shunt surgery most often failed because of blocking of the tube ostium by iris, ICE membrane, or membraneinduced tube migration. Conclusion: Aqueous shunt surgery appears to be an effective method for IOP lowering in some eyes with ICE syndrome-related glaucoma when medical treatment or conventional filtration surgeries fail, but additional glaucoma procedures and/or aqueous shunt revisions and tube repositionings are not uncommon. Ophthalmology 1999;106:1030 –1034 Iridocorneal endothelial (ICE) syndrome is a spectrum of ocular diseases (essential iris atrophy, Chandler syndrome, and Cogan–Reese syndrome) characterized by corneal endothelial abnormalities, unilateral glaucoma, and iris stromal abnormalities.1,2 Chronic corneal edema results from the decrease in endothelial density as the corneal endothelial cells migrate to the periphery, no longer confined to the posterior corneal surface by “contact inhibition.”3 Dysfunctional endothelial cells also migrate over the trabecular meshwork and anterior portion of the iris, resulting in peripheral anterior synechiae, ectropion uvea, correctopia, and iris hole formation.4 Approximately half of all patients with the ICE syndrome develop glaucoma.5 The patients with essential iris
Originally received: June 24, 1998. Revision accepted: January 14, 1999. Manuscript no. 98330. 1 Glaucoma Service, University of Florida, Jacksonville, Florida. 2 Glaucoma Service, Wills Eye Hospital, Jefferson Medical College, Thomas Jefferson University, Philadelphia, Pennsylvania. 3 Retina Service, Wills Eye Hospital, Jefferson Medical College, Thomas Jefferson University, Philadelphia, Pennsylvania. Supported in part by the Glaucoma Service Foundation to Prevent Blindness, Philadelphia, Pennsylvania. The authors have no proprietary interest in any of the materials used in this study. Reprint requests to Courtland M. Schmidt, Jr., MD, Glaucoma Service, Wills Eye Hospital, 900 Walnut St, Philadelphia, PA 19107-5599.
1030
atrophy tend to develop a more severe form of glaucoma compared to that of Chandler syndrome.6 Presumably, obstruction to aqueous outflow results from either the membrane covering the trabecular meshwork or synechial closure of the anterior chamber angle.1,4,7–9 Glaucoma in the ICE syndrome is often difficult to control. Medical management is usually restricted to aqueous suppressants and is often unsuccessful. Laser trabeculoplasty is ineffective. Filtering surgery often succeeds for a short while but frequently fails, presumably because of continued growth of the endothelial membrane over the surgical fistula.10 The effectiveness of aqueous shunts in controlling glaucoma in ICE syndrome has not been reported. We present long-term outcomes of ten patients with uncontrolled glaucoma associated with the ICE syndrome who underwent an aqueous shunt procedure.
Materials and Methods We reviewed the charts of ten patients with ICE syndrome-related glaucoma who underwent aqueous shunt surgery by a member of the Glaucoma Service of Wills Eye Hospital between 1987 and 1996. Complete success was defined as a postoperative intraocular pressure (IOP) of 21 mmHg or lower, with or without glaucoma medications, for the entire course of available follow-up. Qualified success was defined as IOP of 21 mmHg or lower, with or without medications, at most recent follow-up in eyes that had undergone
Kim et al 䡠 Aqueous Shunt Surgery Outcome in Patients with ICE Table 1. Summary of the Preoperative Characteristics of Ten Patients (10 Eyes) with ICE Syndrome Case No.
Age (yrs)
Eye (OD or OS)
1 2 3
40 30 35
OS OS OS
4 5
46 50
OS OD
6
51
OD
7 8 9
38 30 47
OD OS OD
10
75
OD
Diagnosis
Previous Surgery
Preop IOP (mmHg)
Preop Visual Acuity
No. of Preop Medications
Chandler syndrome Chandler syndrome Essential iris atrophy Iris nevus syndrome Chandler syndrome
Trab, Trab with 5-FU PKP, Trab ⫻ 2 Trab with 5-FU
41 45 29
HM 20/70 20/30
4 3 2
None Trab ⫻ 2, cyclodialysis Trab with 5-FU
34 65
CF 20/100
2 2
30
20/30
2
Trab with 5-FU Trab with 5-FU None
27 40 30
20/20 20/60 20/30
3 2 3
Trab with 5-FU
30
20/400
3
Essential iris atrophy Chandler syndrome Chandler syndrome Essential iris atrophy Essential iris atrophy
Trab ⫽ trabeculectomy; 5-FU ⫽ 5-fluorouracil; PKP ⫽ penetrating keratoplasty; HM ⫽ hand motion; CF ⫽ count fingers; Preop ⫽ preoperative.
one or more tube repositions or revisions of the initial aqueous shunt surgery during available follow-up. Failure was defined as performance of or recommendation of further glaucoma surgery (other than tube repositioning or shunt revision) because of inadequate IOP control (i.e., IOP ⬎ 21 mmHg), loss of vision to no light perception, or phthisis bulbi. Kaplan–Meier survival curves were computed for this group of patients.
Results The clinical characteristics of the ten patients are summarized in Table 1. All ten patients were female. The mean age, mean preoperative IOP, and median number of preoperative glaucoma medications were 44.2 ⫾ 13.3 years, 37.1 ⫾ 11.5 mmHg, and 2.6 ⫾ 0.7 medications, respectively. Baseline visual acuity ranged from 20/20 to hand motion (median ⫽ 20/65). All eyes were phakic. Eight of the ten eyes had undergone previous trabeculectomy, six of which were supplemented with 5-fluorouracil (5-FU). Two eyes had not undergone previous ocular surgery. Five eyes had Chandler syndrome, four had essential iris atrophy, and one had Cogan–Reese syndrome. The implant used was a Molteno double-plate in six eyes, a Schocket shunt in three eyes, and a Baerveldt shunt in one eye. Median follow-up after aqueous shunt surgery was 55 months (minimum, 12 months; maximum, 127 months). Four eyes (cases 3, 8, 9, and 10) had adequate IOP control (IOP ⬍ 21 mmHg) on one or two medications for their complete course of available follow-up. Six eyes (cases 1, 2, 4, 5, 6, and 7) required one or more further glaucoma procedures to control IOP. The following paragraphs describe the detailed clinical courses of these six eyes. The summary of the surgeries performed subsequent to the initial aqueous shunt is presented in Table 2. Three of these six eyes (cases 1, 4, and 7) were categorized as qualified successes in terms of their most recent recorded status because their IOP returned to lower than 21 mmHg after one or more repositionings or revisions of the initial aqueous shunt plus ocular antihypertensive medications. Case 1 underwent two surgical revisions of the initial aqueous shunt to lower elevated IOP, which entailed excision of an excessively thick fibrous capsule around the scleral implant at 7 and 38 months after the initial aqueous shunt procedure. Combined pen-
etrating keratoplasty and cataract surgery was performed after 66 months. The graft remained clear for approximately 3 years, then slowly decompensated over the next 2 years. On the last follow-up visit at 127 months, the patient had an IOP of 20 mmHg taking one medication and a visual acuity of hand motion because of a failed graft. In case 4, neodymium:YAG (Nd:YAG) membranectomy effectively opened the tube ostium, which was plugged by the iris and peripheral anterior synechiae 3 months after the aqueous shunt. However, after 23 months, the same material plugged the tube ostium again and the IOP rose to 28 mmHg, which necessitated surgical repositioning of the tube. On the last follow-up visit at 60 months, the patient had an IOP of 17 mmHg on a beta-blocker and a visual acuity of light perception (the patient had a preoperative visual acuity of count fingers secondary to advanced glaucoma damage, cup:disc ratio of 0.99, and 4⫹ nerve pallor). In case 7, Nd:YAG membranectomy was unsuccessful in opening a blocked tube ostium plugged with iris at 3 months, and the patient required a surgical repositioning of the tube to achieve adequate IOP control. Then, the excision of thick fibrous capsule around the scleral implant at 11 months was needed to lower the IOP from 25 mmHg to 18 mmHg. A second repositioning of the tube was needed when the tube became covered with an ICE membrane and the IOP rose to 33 mmHg at 15 months. On the last follow-up visit at 52 months, the patient had an IOP of 15 mmHg taking three medications and a visual acuity of hand motion with diffuse corneal edema. The remaining three eyes (cases 2, 5, and 6) were regarded as complete failures because they required an additional glaucoma surgical procedure other than revision/repositioning of the original aqueous shunt. In case 2, the IOP remained under good control with a beta-blocker for 8 years. The previous corneal transplant failed as the IOP rose to 25 mmHg with two glaucoma medications. The graft had been clear for approximately 9 years. The patient underwent a combined trabeculectomy with penetrating keratoplasty at 97 months. On the last follow-up visit at 102 months, the patient had an IOP of 17 mmHg with a beta-blocker and a visual acuity of 20/200 with a clear graft. For almost 3 years, the IOP of the patient in case 6 remained under control with a beta-blocker and an oral carbonic anhydrase inhibitor. Then, the IOP rose to 33 mmHg despite what appeared to be a functional Molteno implant with high diffuse blebs over the
1031
Ophthalmology Volume 106, Number 5, May 1999 Table 2. Clinical Course of Aqueous Shunts of Ten ICE Patients IOP (mmHg) at Most Recent Follow-up
Visual Acuity at Most Recent Follow-up
Corneal Status at Most Recent Follow-up
No. of Medications Used at Most Recent Follow-up
Total Duration of Follow-up (mos)
Case No.
Type of Aqueous Shunt
Clinical Course Intervention during Follow-up (months after initial aqueous shunt)
Outcome (success or failure)
1
Schocket
HM
Failed graft
1
127
Schocket
Qualified success Failure
20
2
17
20/200
Clear graft
1
102
3 4
Schocket Molteno DP Molteno DP
Revision (7), Revision (38), PKP and Phaco/PCIOL (66) PKP with trabeculectomy with mitomycin-C (97) None Reposition (23)
Success Qualified success Failure
17 17
20/100 LP
Diffuse edema* Diffuse edema*
2 1
24 60
10
HM
Opaque cornea*
3
96
Failure
12
20/80
Clear graft
1
65
Qualified success Success
15
HM
Diffuse edema*
3
52
10
HM†
Hazy cornea
1
44
Success Success
18 17
20/400 CF
Hazy cornea Diffuse edema*
1 2
32 12
5
6
Molteno DP
7
Molteno DP Molteno DP Baerveldt Molteno DP
8 9 10
Reposition (3), YAG CPC (13), CE/ IOL ⫹ Revision (14), Revision ⫹ Reposition to pars plana (23), YAG CPC (25) CE/IOL ⫹ Inferior Baerveldt (33), PKP ⫹ Reposition (40), PKP ⫹ Reposition (63) Reposition (4), Revision (11), Reposition (15) Phaco/PCIOL (36) None None
Molteno DP ⫽ Molteno double-plate implant; CF ⫽ count fingers; HM ⫽ hand motion; LP ⫽ light perception; AC ⫽ anterior chamber; PKP ⫽ penetrating keratoplasty; CPC ⫽ cyclophotocoagulation; Phaco/PCIOL ⫽ phacoemulsification/posterior chamber intraocular lens implantation; CE/IOL ⫽ cataract extraction/intraocular lens implantation. * Diffuse corneal edema needing penetrating keratoplasty. † Visual acuity limited due to dense pupillary ICE membrane.
scleral plates. The patient underwent the placement of an inferotemporal Baerveldt shunt with mitomycin C and cataract extraction at 33 months. An ICE membrane over the tube ostium was opened with Nd:YAG laser at 35 months; the tube became plugged with a similar membrane again within 1 month. The cornea decompensated gradually, and a penetrating keratoplasty was performed in conjunction with the repositioning of the tube in the anterior chamber at 40 months. The tube slowly migrated anteriorly as peripheral anterior synechiae were formed, leading to tube-corneal touch, and a repeat penetrating keratoplasty with tube repositioning at 63 months was performed. On the last follow-up visit at 65 months, the patient had an IOP of 12 mmHg with one medication and a visual acuity of 20/80 with a clear corneal graft. The last of the three patients who did not respond (case 5) had the most difficult IOP to control. The patient had had two previous trabeculectomies and a cyclodialysis procedure before the Molteno implant. She underwent a surgical release of anterior peripheral synechiae, but these recurred, pulling the tip of the tube up against the corneal endothelium at 3 months. The YAG cyclophotocoagulation (CPC) was performed at 13 months for an IOP of 30 mmHg with maximally tolerated medications. After YAG CPC, the eye became hypotonous with a flat anterior chamber, and a dense anterior chamber membrane formed. This required anterior chamber reformation, excision of the anterior chamber membrane, and extracapsular cataract extraction without intraocular lens implantation. Once again, the iris completely surrounded and occluded the tube, causing the IOP to rise to 27 mmHg and requiring the tube to be repositioned in the vitreous cavity at 23 months. The IOP remained in the upper 20s until repeat YAG CPC was performed at 25 months. On the last follow-up visit at 96 months, the patient had an IOP of 10 mmHg on a beta-blocker and a topical carbonic anhydrase inhibitor with a visual acuity of hand motion and an opaque decompensated cornea.
1032
Figure 1 shows the Kaplan–Meier survival curves for the ten study eyes. The solid line represents the probability of the initial aqueous shunt survival without any additional glaucoma surgical procedures. The estimated probabilities of the aqueous shunt survival at 1 year and 5 years were 70% and 40%, respectively. The dotted line represents the survival curve of all successes (i.e., both qualified and complete successes). The estimated probabilities of
Figure 1. Kaplan–Meier estimated probability of aqueous shunt survival. Solid line represents the probability of the initial aqueous shunt survival without any additional glaucoma surgical procedures. Dotted line represents the survival rate with aqueous shunt revisions or tube repositionings.
Kim et al 䡠 Aqueous Shunt Surgery Outcome in Patients with ICE the aqueous shunt survival for this group were 90% at 1 year and 75% at 5 years. This means that no additional glaucoma surgical procedures other than revision or repositioning of the initial aqueous shunt were required to achieve adequate IOP control in 90% of the eyes at 1 year and 75% at 5 years. The postoperative visual acuity ranged from 20/80 to light perception (median ⫽ count fingers) compared to the baseline visual acuity, which ranged from 20/20 to hand motion (median ⫽ 20/65). The exact cause for the worsened visual function is difficult to ascertain. However, the postoperative visual acuity was limited in at least seven of ten eyes because of either corneal decompensation (cases 1, 3–5, 7, and 10) or pupillary ICE membrane (case 8), both of which made optic nerve evaluation for glaucomatous progression difficult. After the aqueous shunt surgery, the cornea remained relatively clear during available follow-up period in two eyes (cases 8 and 9). Three eyes (cases 1, 2, and 6) had postshunt corneal decompensation and underwent penetrating keratoplasty subsequent to the tube shunt surgery; at the time of most recent follow-up, the graft remained clear in two of three eyes (cases 2 and 6) but was decompensated in the remaining eye (case 1). Five eyes (cases 3, 4, 5, 7, and 10) had a totally decompensated cornea at most recent follow-up; none of these six patients had plans to undergo any future corneal surgery.
Discussion Aqueous shunt surgery is usually performed when conventional filtration surgeries fail or are expected to fail. Patients with ICE syndrome undergoing filtering procedures are believed to have a lower success rate than do patients with most other types of glaucoma.11 Laganowski et al5 reported that glaucoma occurred in 33 (50%) of 66 patients with the ICE syndrome. Of the 25 patients with glaucoma whose detailed records were reviewed, 22 required surgical intervention. Twenty of 22 patients had trabeculectomy without antimetabolite. The failure rates of first operations at 1 and 5 years were 40% and 79%, respectively. Ten (45%) patients required more than one procedure. The use of 5-FU with trabeculectomy in eyes with ICE syndrome that had previous trabeculectomy resulted in a better success rate, albeit lower than the success rate of repeat filtering surgery in other cases in the Fluorouracil Filtering Surgery Study.12 Kidd and colleagues10 reported three eyes with ICE syndrome that had a successful trabeculectomy with 5-FU after failing routine trabeculectomy. In another study, four of nine eyes with previously unsuccessful trabeculectomies had a successful outcome after trabeculectomy with postoperative subconjunctival 5-FU injections with a mean follow-up of 25.3 months.12 The median follow-up time was not reported. All five eyes that failed received a Molteno implant. However, no long-term follow-up of these patients was reported. The use of mitomycin C during trabeculectomy in this population has not been reported to date. In our study, with a median follow-up time of 55 months, four of ten eyes had completely successful IOP control (IOP ⬍ 21 mmHg) over their entire course. The median follow-up for the four complete successes was 28 months. An additional three patients (cases 1, 4, and 7)
achieved adequate IOP control after one or more tube repositionings or revisions of the initial aqueous shunt. In our small study, the complete success rate of aqueous shunt surgery was similar to that reported for trabeculectomy with 5-FU by Wright and colleagues.12 The optimal positioning of a tube in a phakic eye, especially one with an anteriorly displaced iris due to contraction of an ICE membrane and a decompensated cornea, can be challenging. In our series, the initial postoperative position of the tube was proper in all cases. However, subsequently, four eyes (cases 4 –7) required one or more repositionings of the distal tube because of either a blocked tube ostium by iris (cases 4 –7) and/or ICE membrane (cases 6 and 7) or because of anterior migration of the tube resulting in tube-corneal touch secondary to the ICE membrane (cases 5 and 6). To minimize the risk of tube occlusion or anterior migration, surgical techniques of aqueous shunt implantation can be modified. To facilitate future tube repositioning, the distal tube should be cut slightly longer than usual before anterior chamber insertion. This allows future tube repositions at different sites to be feasible and technically easier. The course of the tube from the reservoir to the anterior segment can be slightly curved to avoid a long intracameral portion of the tube. If the patient is pseudophakic and has high peripheral anterior synechiae, placement of the tube through the pars plana should be considered. Another option would be performing a surgical iridectomy and placing the tube tip through the iridectomy into the space posterior to the iris plane. Combined cataract extraction and aqueous shunt implantation may be ideal in patients with cataract and may allow positioning of the tube more posteriorly away from the ICE membrane and peripheral anterior synechiae. In summary, aqueous shunt surgery, in conjunction with medical treatment, appears to be effective for IOP lowering in some patients in whom surgery is performed for ICE syndrome-related glaucoma, when medical treatment and conventional filtration surgeries have failed. Unfortunately, some eyes experience failure of IOP control after aqueous shunt surgery and require additional glaucoma procedures and/or aqueous shunt revisions and tube repositionings.
References 1. Eagle RC Jr, Font RL, Yanoff M, Fine BS. Proliferative endotheliopathy with iris abnormalities. The iridocorneal endothelial syndrome. Arch Ophthalmol 1979;97:2104 –11. 2. Shields MB. Progressive essential iris atrophy, Chandler’s syndrome, and the iris nevus (Cogan–Reese) syndrome: a spectrum of disease. Surv Ophthalmol 1979;24:3–20. 3. Alvarado JA, Murphy CG, Maglio M, Hetherington J. Pathogenesis of Chandler’s syndrome, essential iris atrophy and the Cogan–Reese syndrome. I. Alterations of the corneal endothelium. Invest Ophthalmol Vis Sci 1986;27:853–72. 4. Campbell DG, Shields MB, Smith TR. The corneal endothelium and the spectrum of essential iris atrophy. Am J Ophthalmol 1978;86:317–24.
1033
Ophthalmology Volume 106, Number 5, May 1999 5. Laganowski HC, Kerr Muir MG, Hitchings RA. Glaucoma and the iridocorneal endothelial syndrome. Arch Ophthalmol 1992;110:346 –50. 6. Wilson MC, Shields MB. A comparison of the clinical variations of the iridocorneal endothelial syndrome. Arch Ophthalmol 1989;107:1465– 8. 7. Shields MB, McCracken JS, Klintworth GK, Campbell DG. Corneal edema in essential iris atrophy. Ophthalmology 1979; 86:1533–50. 8. Patel A, Kenyon KR, Hirst LW, et al. Clinicopathologic features of Chandler’s syndrome. Surv Ophthalmol 1983;27: 327– 44.
9. Portis JM, Stamper RL, Spencer WH, Webster RG Jr. The corneal endothelium and Descemet’s membrane in the iridocorneal endothelial syndrome. Trans Am Ophthalmol Soc 1985;83:316 –31. 10. Kidd M, Hetherington J, Magee S. Surgical results in iridocorneal endothelial syndrome. Arch Ophthalmol 1988;106: 199 –201. 11. Shields MB, Campbell DG, Simmons RJ. The essential iris atrophies. Am J Ophthalmol 1978;85:749 –59. 12. Wright MM, Grajewski AL, Cristal SM, Parrish RK. 5-fluorouracil after trabeculectomy and the iridocorneal endothelial syndrome. Ophthalmology 1991;98:314 – 6.
Historical Image DeZeng electric retinoscope, patented August 25, 1903.*
* Courtesy of the Museum of Ophthalmology, Foundation of the American Academy of Ophthalmology, San Francisco, California.
1034
Originally received: June 24, 1998. Revision accepted: January 14, 1999. Manuscript no. 98330. 1 Glaucoma Service, University of Florida, Jacksonville, Florida. 2 Glaucoma Service, Wills Eye Hospital, Jefferson Medical College, Thomas Jefferson University, Philadelphia, Pennsylvania. 3 Retina Service, Wills Eye Hospital, Jefferson Medical College, Thomas Jefferson University, Philadelphia, Pennsylvania. Supported in part by the Glaucoma Service Foundation to Prevent Blindness, Philadelphia, Pennsylvania. The authors have no proprietary interest in any of the materials used in this study. Reprint requests to Courtland M. Schmidt, Jr., MD, Glaucoma Service, Wills Eye Hospital, 900 Walnut St, Philadelphia, PA 19107-5599.
1030
atrophy tend to develop a more severe form of glaucoma compared to that of Chandler syndrome.6 Presumably, obstruction to aqueous outflow results from either the membrane covering the trabecular meshwork or synechial closure of the anterior chamber angle.1,4,7–9 Glaucoma in the ICE syndrome is often difficult to control. Medical management is usually restricted to aqueous suppressants and is often unsuccessful. Laser trabeculoplasty is ineffective. Filtering surgery often succeeds for a short while but frequently fails, presumably because of continued growth of the endothelial membrane over the surgical fistula.10 The effectiveness of aqueous shunts in controlling glaucoma in ICE syndrome has not been reported. We present long-term outcomes of ten patients with uncontrolled glaucoma associated with the ICE syndrome who underwent an aqueous shunt procedure.
Materials and Methods We reviewed the charts of ten patients with ICE syndrome-related glaucoma who underwent aqueous shunt surgery by a member of the Glaucoma Service of Wills Eye Hospital between 1987 and 1996. Complete success was defined as a postoperative intraocular pressure (IOP) of 21 mmHg or lower, with or without glaucoma medications, for the entire course of available follow-up. Qualified success was defined as IOP of 21 mmHg or lower, with or without medications, at most recent follow-up in eyes that had undergone
Kim et al 䡠 Aqueous Shunt Surgery Outcome in Patients with ICE Table 1. Summary of the Preoperative Characteristics of Ten Patients (10 Eyes) with ICE Syndrome Case No.
Age (yrs)
Eye (OD or OS)
1 2 3
40 30 35
OS OS OS
4 5
46 50
OS OD
6
51
OD
7 8 9
38 30 47
OD OS OD
10
75
OD
Diagnosis
Previous Surgery
Preop IOP (mmHg)
Preop Visual Acuity
No. of Preop Medications
Chandler syndrome Chandler syndrome Essential iris atrophy Iris nevus syndrome Chandler syndrome
Trab, Trab with 5-FU PKP, Trab ⫻ 2 Trab with 5-FU
41 45 29
HM 20/70 20/30
4 3 2
None Trab ⫻ 2, cyclodialysis Trab with 5-FU
34 65
CF 20/100
2 2
30
20/30
2
Trab with 5-FU Trab with 5-FU None
27 40 30
20/20 20/60 20/30
3 2 3
Trab with 5-FU
30
20/400
3
Essential iris atrophy Chandler syndrome Chandler syndrome Essential iris atrophy Essential iris atrophy
Trab ⫽ trabeculectomy; 5-FU ⫽ 5-fluorouracil; PKP ⫽ penetrating keratoplasty; HM ⫽ hand motion; CF ⫽ count fingers; Preop ⫽ preoperative.
one or more tube repositions or revisions of the initial aqueous shunt surgery during available follow-up. Failure was defined as performance of or recommendation of further glaucoma surgery (other than tube repositioning or shunt revision) because of inadequate IOP control (i.e., IOP ⬎ 21 mmHg), loss of vision to no light perception, or phthisis bulbi. Kaplan–Meier survival curves were computed for this group of patients.
Results The clinical characteristics of the ten patients are summarized in Table 1. All ten patients were female. The mean age, mean preoperative IOP, and median number of preoperative glaucoma medications were 44.2 ⫾ 13.3 years, 37.1 ⫾ 11.5 mmHg, and 2.6 ⫾ 0.7 medications, respectively. Baseline visual acuity ranged from 20/20 to hand motion (median ⫽ 20/65). All eyes were phakic. Eight of the ten eyes had undergone previous trabeculectomy, six of which were supplemented with 5-fluorouracil (5-FU). Two eyes had not undergone previous ocular surgery. Five eyes had Chandler syndrome, four had essential iris atrophy, and one had Cogan–Reese syndrome. The implant used was a Molteno double-plate in six eyes, a Schocket shunt in three eyes, and a Baerveldt shunt in one eye. Median follow-up after aqueous shunt surgery was 55 months (minimum, 12 months; maximum, 127 months). Four eyes (cases 3, 8, 9, and 10) had adequate IOP control (IOP ⬍ 21 mmHg) on one or two medications for their complete course of available follow-up. Six eyes (cases 1, 2, 4, 5, 6, and 7) required one or more further glaucoma procedures to control IOP. The following paragraphs describe the detailed clinical courses of these six eyes. The summary of the surgeries performed subsequent to the initial aqueous shunt is presented in Table 2. Three of these six eyes (cases 1, 4, and 7) were categorized as qualified successes in terms of their most recent recorded status because their IOP returned to lower than 21 mmHg after one or more repositionings or revisions of the initial aqueous shunt plus ocular antihypertensive medications. Case 1 underwent two surgical revisions of the initial aqueous shunt to lower elevated IOP, which entailed excision of an excessively thick fibrous capsule around the scleral implant at 7 and 38 months after the initial aqueous shunt procedure. Combined pen-
etrating keratoplasty and cataract surgery was performed after 66 months. The graft remained clear for approximately 3 years, then slowly decompensated over the next 2 years. On the last follow-up visit at 127 months, the patient had an IOP of 20 mmHg taking one medication and a visual acuity of hand motion because of a failed graft. In case 4, neodymium:YAG (Nd:YAG) membranectomy effectively opened the tube ostium, which was plugged by the iris and peripheral anterior synechiae 3 months after the aqueous shunt. However, after 23 months, the same material plugged the tube ostium again and the IOP rose to 28 mmHg, which necessitated surgical repositioning of the tube. On the last follow-up visit at 60 months, the patient had an IOP of 17 mmHg on a beta-blocker and a visual acuity of light perception (the patient had a preoperative visual acuity of count fingers secondary to advanced glaucoma damage, cup:disc ratio of 0.99, and 4⫹ nerve pallor). In case 7, Nd:YAG membranectomy was unsuccessful in opening a blocked tube ostium plugged with iris at 3 months, and the patient required a surgical repositioning of the tube to achieve adequate IOP control. Then, the excision of thick fibrous capsule around the scleral implant at 11 months was needed to lower the IOP from 25 mmHg to 18 mmHg. A second repositioning of the tube was needed when the tube became covered with an ICE membrane and the IOP rose to 33 mmHg at 15 months. On the last follow-up visit at 52 months, the patient had an IOP of 15 mmHg taking three medications and a visual acuity of hand motion with diffuse corneal edema. The remaining three eyes (cases 2, 5, and 6) were regarded as complete failures because they required an additional glaucoma surgical procedure other than revision/repositioning of the original aqueous shunt. In case 2, the IOP remained under good control with a beta-blocker for 8 years. The previous corneal transplant failed as the IOP rose to 25 mmHg with two glaucoma medications. The graft had been clear for approximately 9 years. The patient underwent a combined trabeculectomy with penetrating keratoplasty at 97 months. On the last follow-up visit at 102 months, the patient had an IOP of 17 mmHg with a beta-blocker and a visual acuity of 20/200 with a clear graft. For almost 3 years, the IOP of the patient in case 6 remained under control with a beta-blocker and an oral carbonic anhydrase inhibitor. Then, the IOP rose to 33 mmHg despite what appeared to be a functional Molteno implant with high diffuse blebs over the
1031
Ophthalmology Volume 106, Number 5, May 1999 Table 2. Clinical Course of Aqueous Shunts of Ten ICE Patients IOP (mmHg) at Most Recent Follow-up
Visual Acuity at Most Recent Follow-up
Corneal Status at Most Recent Follow-up
No. of Medications Used at Most Recent Follow-up
Total Duration of Follow-up (mos)
Case No.
Type of Aqueous Shunt
Clinical Course Intervention during Follow-up (months after initial aqueous shunt)
Outcome (success or failure)
1
Schocket
HM
Failed graft
1
127
Schocket
Qualified success Failure
20
2
17
20/200
Clear graft
1
102
3 4
Schocket Molteno DP Molteno DP
Revision (7), Revision (38), PKP and Phaco/PCIOL (66) PKP with trabeculectomy with mitomycin-C (97) None Reposition (23)
Success Qualified success Failure
17 17
20/100 LP
Diffuse edema* Diffuse edema*
2 1
24 60
10
HM
Opaque cornea*
3
96
Failure
12
20/80
Clear graft
1
65
Qualified success Success
15
HM
Diffuse edema*
3
52
10
HM†
Hazy cornea
1
44
Success Success
18 17
20/400 CF
Hazy cornea Diffuse edema*
1 2
32 12
5
6
Molteno DP
7
Molteno DP Molteno DP Baerveldt Molteno DP
8 9 10
Reposition (3), YAG CPC (13), CE/ IOL ⫹ Revision (14), Revision ⫹ Reposition to pars plana (23), YAG CPC (25) CE/IOL ⫹ Inferior Baerveldt (33), PKP ⫹ Reposition (40), PKP ⫹ Reposition (63) Reposition (4), Revision (11), Reposition (15) Phaco/PCIOL (36) None None
Molteno DP ⫽ Molteno double-plate implant; CF ⫽ count fingers; HM ⫽ hand motion; LP ⫽ light perception; AC ⫽ anterior chamber; PKP ⫽ penetrating keratoplasty; CPC ⫽ cyclophotocoagulation; Phaco/PCIOL ⫽ phacoemulsification/posterior chamber intraocular lens implantation; CE/IOL ⫽ cataract extraction/intraocular lens implantation. * Diffuse corneal edema needing penetrating keratoplasty. † Visual acuity limited due to dense pupillary ICE membrane.
scleral plates. The patient underwent the placement of an inferotemporal Baerveldt shunt with mitomycin C and cataract extraction at 33 months. An ICE membrane over the tube ostium was opened with Nd:YAG laser at 35 months; the tube became plugged with a similar membrane again within 1 month. The cornea decompensated gradually, and a penetrating keratoplasty was performed in conjunction with the repositioning of the tube in the anterior chamber at 40 months. The tube slowly migrated anteriorly as peripheral anterior synechiae were formed, leading to tube-corneal touch, and a repeat penetrating keratoplasty with tube repositioning at 63 months was performed. On the last follow-up visit at 65 months, the patient had an IOP of 12 mmHg with one medication and a visual acuity of 20/80 with a clear corneal graft. The last of the three patients who did not respond (case 5) had the most difficult IOP to control. The patient had had two previous trabeculectomies and a cyclodialysis procedure before the Molteno implant. She underwent a surgical release of anterior peripheral synechiae, but these recurred, pulling the tip of the tube up against the corneal endothelium at 3 months. The YAG cyclophotocoagulation (CPC) was performed at 13 months for an IOP of 30 mmHg with maximally tolerated medications. After YAG CPC, the eye became hypotonous with a flat anterior chamber, and a dense anterior chamber membrane formed. This required anterior chamber reformation, excision of the anterior chamber membrane, and extracapsular cataract extraction without intraocular lens implantation. Once again, the iris completely surrounded and occluded the tube, causing the IOP to rise to 27 mmHg and requiring the tube to be repositioned in the vitreous cavity at 23 months. The IOP remained in the upper 20s until repeat YAG CPC was performed at 25 months. On the last follow-up visit at 96 months, the patient had an IOP of 10 mmHg on a beta-blocker and a topical carbonic anhydrase inhibitor with a visual acuity of hand motion and an opaque decompensated cornea.
1032
Figure 1 shows the Kaplan–Meier survival curves for the ten study eyes. The solid line represents the probability of the initial aqueous shunt survival without any additional glaucoma surgical procedures. The estimated probabilities of the aqueous shunt survival at 1 year and 5 years were 70% and 40%, respectively. The dotted line represents the survival curve of all successes (i.e., both qualified and complete successes). The estimated probabilities of
Figure 1. Kaplan–Meier estimated probability of aqueous shunt survival. Solid line represents the probability of the initial aqueous shunt survival without any additional glaucoma surgical procedures. Dotted line represents the survival rate with aqueous shunt revisions or tube repositionings.
Kim et al 䡠 Aqueous Shunt Surgery Outcome in Patients with ICE the aqueous shunt survival for this group were 90% at 1 year and 75% at 5 years. This means that no additional glaucoma surgical procedures other than revision or repositioning of the initial aqueous shunt were required to achieve adequate IOP control in 90% of the eyes at 1 year and 75% at 5 years. The postoperative visual acuity ranged from 20/80 to light perception (median ⫽ count fingers) compared to the baseline visual acuity, which ranged from 20/20 to hand motion (median ⫽ 20/65). The exact cause for the worsened visual function is difficult to ascertain. However, the postoperative visual acuity was limited in at least seven of ten eyes because of either corneal decompensation (cases 1, 3–5, 7, and 10) or pupillary ICE membrane (case 8), both of which made optic nerve evaluation for glaucomatous progression difficult. After the aqueous shunt surgery, the cornea remained relatively clear during available follow-up period in two eyes (cases 8 and 9). Three eyes (cases 1, 2, and 6) had postshunt corneal decompensation and underwent penetrating keratoplasty subsequent to the tube shunt surgery; at the time of most recent follow-up, the graft remained clear in two of three eyes (cases 2 and 6) but was decompensated in the remaining eye (case 1). Five eyes (cases 3, 4, 5, 7, and 10) had a totally decompensated cornea at most recent follow-up; none of these six patients had plans to undergo any future corneal surgery.
Discussion Aqueous shunt surgery is usually performed when conventional filtration surgeries fail or are expected to fail. Patients with ICE syndrome undergoing filtering procedures are believed to have a lower success rate than do patients with most other types of glaucoma.11 Laganowski et al5 reported that glaucoma occurred in 33 (50%) of 66 patients with the ICE syndrome. Of the 25 patients with glaucoma whose detailed records were reviewed, 22 required surgical intervention. Twenty of 22 patients had trabeculectomy without antimetabolite. The failure rates of first operations at 1 and 5 years were 40% and 79%, respectively. Ten (45%) patients required more than one procedure. The use of 5-FU with trabeculectomy in eyes with ICE syndrome that had previous trabeculectomy resulted in a better success rate, albeit lower than the success rate of repeat filtering surgery in other cases in the Fluorouracil Filtering Surgery Study.12 Kidd and colleagues10 reported three eyes with ICE syndrome that had a successful trabeculectomy with 5-FU after failing routine trabeculectomy. In another study, four of nine eyes with previously unsuccessful trabeculectomies had a successful outcome after trabeculectomy with postoperative subconjunctival 5-FU injections with a mean follow-up of 25.3 months.12 The median follow-up time was not reported. All five eyes that failed received a Molteno implant. However, no long-term follow-up of these patients was reported. The use of mitomycin C during trabeculectomy in this population has not been reported to date. In our study, with a median follow-up time of 55 months, four of ten eyes had completely successful IOP control (IOP ⬍ 21 mmHg) over their entire course. The median follow-up for the four complete successes was 28 months. An additional three patients (cases 1, 4, and 7)
achieved adequate IOP control after one or more tube repositionings or revisions of the initial aqueous shunt. In our small study, the complete success rate of aqueous shunt surgery was similar to that reported for trabeculectomy with 5-FU by Wright and colleagues.12 The optimal positioning of a tube in a phakic eye, especially one with an anteriorly displaced iris due to contraction of an ICE membrane and a decompensated cornea, can be challenging. In our series, the initial postoperative position of the tube was proper in all cases. However, subsequently, four eyes (cases 4 –7) required one or more repositionings of the distal tube because of either a blocked tube ostium by iris (cases 4 –7) and/or ICE membrane (cases 6 and 7) or because of anterior migration of the tube resulting in tube-corneal touch secondary to the ICE membrane (cases 5 and 6). To minimize the risk of tube occlusion or anterior migration, surgical techniques of aqueous shunt implantation can be modified. To facilitate future tube repositioning, the distal tube should be cut slightly longer than usual before anterior chamber insertion. This allows future tube repositions at different sites to be feasible and technically easier. The course of the tube from the reservoir to the anterior segment can be slightly curved to avoid a long intracameral portion of the tube. If the patient is pseudophakic and has high peripheral anterior synechiae, placement of the tube through the pars plana should be considered. Another option would be performing a surgical iridectomy and placing the tube tip through the iridectomy into the space posterior to the iris plane. Combined cataract extraction and aqueous shunt implantation may be ideal in patients with cataract and may allow positioning of the tube more posteriorly away from the ICE membrane and peripheral anterior synechiae. In summary, aqueous shunt surgery, in conjunction with medical treatment, appears to be effective for IOP lowering in some patients in whom surgery is performed for ICE syndrome-related glaucoma, when medical treatment and conventional filtration surgeries have failed. Unfortunately, some eyes experience failure of IOP control after aqueous shunt surgery and require additional glaucoma procedures and/or aqueous shunt revisions and tube repositionings.
References 1. Eagle RC Jr, Font RL, Yanoff M, Fine BS. Proliferative endotheliopathy with iris abnormalities. The iridocorneal endothelial syndrome. Arch Ophthalmol 1979;97:2104 –11. 2. Shields MB. Progressive essential iris atrophy, Chandler’s syndrome, and the iris nevus (Cogan–Reese) syndrome: a spectrum of disease. Surv Ophthalmol 1979;24:3–20. 3. Alvarado JA, Murphy CG, Maglio M, Hetherington J. Pathogenesis of Chandler’s syndrome, essential iris atrophy and the Cogan–Reese syndrome. I. Alterations of the corneal endothelium. Invest Ophthalmol Vis Sci 1986;27:853–72. 4. Campbell DG, Shields MB, Smith TR. The corneal endothelium and the spectrum of essential iris atrophy. Am J Ophthalmol 1978;86:317–24.
1033
Ophthalmology Volume 106, Number 5, May 1999 5. Laganowski HC, Kerr Muir MG, Hitchings RA. Glaucoma and the iridocorneal endothelial syndrome. Arch Ophthalmol 1992;110:346 –50. 6. Wilson MC, Shields MB. A comparison of the clinical variations of the iridocorneal endothelial syndrome. Arch Ophthalmol 1989;107:1465– 8. 7. Shields MB, McCracken JS, Klintworth GK, Campbell DG. Corneal edema in essential iris atrophy. Ophthalmology 1979; 86:1533–50. 8. Patel A, Kenyon KR, Hirst LW, et al. Clinicopathologic features of Chandler’s syndrome. Surv Ophthalmol 1983;27: 327– 44.
9. Portis JM, Stamper RL, Spencer WH, Webster RG Jr. The corneal endothelium and Descemet’s membrane in the iridocorneal endothelial syndrome. Trans Am Ophthalmol Soc 1985;83:316 –31. 10. Kidd M, Hetherington J, Magee S. Surgical results in iridocorneal endothelial syndrome. Arch Ophthalmol 1988;106: 199 –201. 11. Shields MB, Campbell DG, Simmons RJ. The essential iris atrophies. Am J Ophthalmol 1978;85:749 –59. 12. Wright MM, Grajewski AL, Cristal SM, Parrish RK. 5-fluorouracil after trabeculectomy and the iridocorneal endothelial syndrome. Ophthalmology 1991;98:314 – 6.
Historical Image DeZeng electric retinoscope, patented August 25, 1903.*
* Courtesy of the Museum of Ophthalmology, Foundation of the American Academy of Ophthalmology, San Francisco, California.
1034