- Email: [email protected]
Baerveldt implant surgery in the treatment of advanced childhood glaucoma
Baerveldt Implant* Surgery in the Treatment of Advanced Childhood Glaucoma Sean P. Donahue, MD, PhD,* Ronald V. Keech, MD, Paul Munden, MD, and William E. Scott, MD
Background:The efficacy of Baerveldt implant (Pharmacia & Upjohn, Inc., Kalamazoo, Mich.) surgery in the treatment of advanced childhood glaucoma is unknown. Methods:We reviewed the results of 23 consecutive 350 mm Baerveldt implants in 20 eyes of 17 children. Results were classified as follows: (1) success; no further reoperation, no decrease in vision, and intraocul'ar pressure at last follow-up less than 21 mm Hg with no medications; (2) qualified success; medication necessary to bring intraocular pressure to less than 21 mm Hg or complication not associated with tube failure; and (3) failure; intraocular pressure >20 mm Hg, tube failure complication or reoperation causing tube removal, phthisis, or enucleation.Resu/ts:Original glaucoma types were bilateral aphakic (five), unilateral aphakic or persistent hyperplastic primaryvitreous (four), primary infantile (four), juvenile (three), secondary (two), Peter syndrome (one), and Lowe syndrome (one). Patients had undergone a mean of 2.8 previous intraocular procedures. Mean preoperative intraocular pressure was 33.6 mm Hg; average number of preoperative glaucoma medications was 3.0. Mean follow-up was 19 months (range, 6 to 32 months). Eight procedures were considered successful (mean intraocular pressure 15.5 mm Hg), six were qualified successes (mean number of medications 0.8; mean intraocular pressure 16 mm Hg), and nine failed. Two eyes in the qualified success group do not have useful vision as a result of complications. Complications included retinal detachment (five), corneal decompensation (five), corneal graft rejection in five of six grafts; dislocated tubes (three), and recurrent uveitis (two). One ofthese eyes is phthisic and one has been enucleated. Onlytwo of nine procedures in eyes with a history of one or no previous intraocula r operations failed, whereas seven of 13 proced ures in eyes with a history ofthree or more previous procedures failed. Onlyseven of 13 procedures in aphakic eyes were successes or qualified successes, whereas seven of 10 procedures in phakic eyes had good results. Conclusion:Baerveldt implants can produce good short-term results, especially in phakic eyes. Aphakic eyes and eyes that have undergone multiple procedures are at a much higher risk for devastating complications. (J AAPOS 1997;1:41-5)
p
atients with primary and secondary childhood glaucomas are notoriously difficult to manage. Once initial procedures, such as trabeculotomy or goniotomy, have failed, a second choice is usually a filtering procedure with or without an antimetabolite. The longterm efficacy of antimetabolites is unclear, and filtering blebs present a potential problem for the aphaldc contact lens wearing patient. Seton surgery is now relatively well accepted for the treatment of advanced glaucoma in adults. 1-3Early reports have demonstrated relatively good success of Molteno
From the University of Iowa Hospitals and Clinics, Iowa City. Address reprint requests to Sean P. Donahue, MD, PhD, Depa*'tment of Ophthalmology & Visual Sdenees, 1215 21st Ave. South, Vanderbilt Medical Center East, Nashville, TN37232-8808. *None of the authors has any financial interest in the Baevveldt shunt implant or any competing produet. Baerveldt implant is a trade name of Pharmacia & Upjobn, Inc., Kalamazoo, Michigan. ¢S. P. D. is now associated with Vanderhilt University Children's Eye Center, Nashville, Tennessee. Copyright © 1997 by the American Association for Pediatric Ophthalmology and Strabismus. 1091-8531/9755.00+ 0 75/1/80309
Journal of AAPOS
implants (IOP, Inc., Costa Mesa, Calif.) in children. 4-8 However, the efficacy of Baerveldt implant (Pharmacia & Upjohn, Inc., Kalamazoo, Mich.) surgery in advanced childhood glaucoma (aphakic patients and patients having undergone several surgical procedures) is unknown, although an early report in seven patients is favorable.~We retrospectively evaluated our experience with 23 Baerveldt implants in 20 eyes of 17 patients. METHODS A surgical review demonstrated that 23 Baerveldt implants had been performed on patients less than 21 years of age at the University of Iowa Hospitals since 1990. These charts were reviewed, and previous procedures, diagnoses, intraocular pressure, and age at the operation were determined. In addition, it was noted whether the patient was receiving any preoperative medications. Outcome endpoints were last intraocular pressure, number of medications, and the presence of any complications. At the University of Iowa, Baerveldt implants are usually placed in a single procedure, as described by Molteno. 9 A 350 mm Baerveldt implant was used in all but one case, in March 1997 41
42
Donahue et al.
which a 200 mm implant was used. To prevent immediate postoperative hypotony, a 8-0 polyglactin suture was placed around the tube before the tube entered the plate as described previously.6,7,9The suture degenerates after 2 to 3 weeks, and the tube opens. In later procedures, the tube was fenestrated to allow some drainage during the time before tube opening. 1° During the period before tube opening, pressure control was achieved with acetazolamide (Diamox) or p-blockers as necessary. Procedures were considered successful if intraocular pressure was less than 21 mm Hg on no medications, and no complications occurred. A qualified success meant that either medications were needed to produce an intraocular pressure (IOP) less than 21 mm Hg or that a complication occurred, but the shunt continued to function. Failures were defined as patients with IOP greater than 21 mm Hg on medication, complications causing failure or removal of the tube, or loss of the eye. RESULTS Patients and their diagnoses are summarized in Table 1. Four eyes among three patients had primary infantile glaucoma, whereas three eyes among two patients had juvenile glaucoma. Nine eyes among eight patients were aphakic, whereas one was pseudophakic. The patient with Lowe syndrome became aphakic with a lensectomy and vitrectomy after the implant procedure in one eye. Age at the operation ranged from 1 to 191 months (15 years of age). Mean preoperative IOP was 33 mm Hg, and patients averaged three mean preoperative glaucoma medications. Eight procedures were successful. Three of these were in eyes of patients with primary infantile glaucoma, four were in patients with aphakic glaucoma, and one was in a patient with Sturge-Weber syndrome. One successful procedure was a repeat Baerveldt implant after the first tube dislocated out of the anterior chamber. Only one successful procedure was associated with a visionthreatening complication. Patient 13 had a tractional retinal detachment from proliferative vitreoretinopathy. However, this was probably a result of the retinal detachment that was surgically treated before the implant was placed. One patient with well-controlled juvenile glaucoma was hit in the affected eye by a tennis ball and suffered a traumatic hyphema, which resolved without affecting the status of the tube. Six procedures were qualified successes. Three eyes (two patients) were considered qualified successes because the pressure was controlled with topical medications. One patient needed an additional procedure to remove vitreous from the tube but is maintaining adequate pressure control. The other two patients suffered corneal decompensation but have maintained adequate pressure control. Nine procedures failed. Three procedures failed and were repeated; one of these repeat procedures also failed. The other five procedures failed because of inadequate pressure control (two), recurrent inflammation necessitating tube removal (two), or retinal
Journal ofAAPOS Volume I Number 1 March 1997
detachment (the tube was removed during the retinal reattachment procedure). The most common complication was retinal detachment (Table 2). This occurred in five eyes among four patients. All five eyes were aphakic at the time of the detachment. In one patient with bilateral aphakia and high myopia (Patient 13), however, bilateral retinal detachments had previously occurred and been treated with scleral buckling in the weeks immediately before the implants were placed. Proliferative vitreoretinopathy and traction retinal detachments eventually developed in both of these eyes. It is thus unclear whether the tractional retinal detachments observed in these two cases were caused by the implant operation or were merely a complication of the previous retinal detachment. Corneal decompensation was also a common complication, occurring in five eyes. Three eyes (two patients) had penetrating keratoplasty before tube placement. In two of these three eyes, the graft failed. All three corneal transplants performed in eyes with functioning shunts failed. Only one of six corneal grafts remained viable. Two eyes had persistent inflammation. In one eye, this led to a tractional retinal detachment, phthisis, and enucleation. A second eye had recurrent sterile uveitis that caused failure of a penetrating keratoplasty. The implant was eventually removed in this patient, and the eye underwent cyclodestructive procedures. One eye, treated previously with lensectomy for a unilateral congenital cataract, eventually had a secondary membrane develop. This was probably not related to the implant. The membrane has not been treated, but the pressure is well controlled. In three eyes the tube became dislocated in the anterior chamber. Analysis of these data demonstrated two trends associated with tube failure: number of previous operations and surgical aphakia. The sample size was too small, however, for these trends to reach statistical significance with Fisher's Exact Test. Of nine procedures in eyes that had undergone one or fewer previous intraocular operations, only two failed (22%). Of the 13 procedures performed in eyes with a history of more than two previous operations, however, seven failed (54%). The presence of multiple previous intraocular surgical procedures should therefore be considered a relative risk factor for failure. Phakic patients did better than aphakic patients in this study. Of the 10 procedures undertaken in phakic eyes, only three failed (30%), and two of these failures occurred in one eye of a patient with Peter syndrome. In comparison, the failure rate for the 13 procedures undertaken in aphakic eyes was 54%. The high failure rate in aphakic eyes might have been caused by the number of previous procedures. Only one of the six aphakic eyes in which lensectomyor vitrectomywas the only previous intraoculaf procedure failed (17%), whereas four of seven aphakic eyes having three or more previous intraocular procedures failed (58%). Although none of these trends reached statistical significance, it appears that the aphakic patient who has had multiple previous procedures is at higher risk for shunt failure.
Journal ofAAPOS Volume 1 Number I March 1997
Donahue et al.
43
TABLE 1. Complications of Baerveldt implant surgery
Pt, no.
Age at operation Eye (mo) Diagnosis
1 0 D
8
1° inf. gl.
Previous procedure(s) Trabeculotomy; goniotomy x2; Mito Tbx Trabeculotomyx2; PK; goniotomy Trabeculetomy;PK; goniotomy
Preoperative lOP/no, reeds
Last lOP/no. reeds
F/U (mo)
36/1
16/0
11
25/3
20/0
33
40/2
--
30
2
OD
53
1° inf. gh
2
OS
57
1° inf. gl.
3
OS
191
1° inf. gl.
Tbx x3; cyclodialysis; Scheie drainage; goniotomy
30/4
16/0
12
4 5
OS OD
114 174
Juv. gl. Juv. gl.
Goniotomy;Mito Tbx None
25/2 21/3
16/2 14/1
11 16
5 6
OS OS
174 5
Juv. gl Cong. Catu
24/3
14/1
15
7 8
OS OS
6 50
Cong. Cat.~ PHPV
None Goniotomyx2; lensectomy; trabeculetomy Lensectomy Lensectomy
35/1 28/2 46/2
10/0 15/0 18/0
18 16 9
9
OS
56
14
OS
78
Lensectomy;lysis of adhesions; 38/4 surgical PI; reformation of AC; membrane discission; previous seton procedure Lensectomy 36/2
15/0
10
Cong. cat.B; chronic angle closure Cong. cat.B
--
16
11
OS
90
PHPV
Lensectomy; membranectomy; Mite Tbx
28/2
--
27
12
OS
169
Cong. Cat.B
Lensectomy
32/5
--
21
13
OD
167
Cong. Cat?
Lensectomy;scleral buckle
47/3
19/1
37
13 14
OS OD
172 96
Cong. cat.B 2 ° glaucoma (traumatic)
Lensectomy;scleral buckle ECCE;PCIOL; PPV;membranectomy
32/2
17/0
20
49/2
--
38
Complications
Outcome
Recurrentuveitis; failed PK; tube removed (15 mo) Traumatichyphema; resolved without incident QS Tenon's cysts, OU; resolved spontaneously Peaked pupil, OS Secondarymembrane Vitreous to tube (PPV and tube repositioned posteriorly) First tube pulled out of AC
Macula off RD; seton removed at 9mo Persistent inflammation; exposed plate; TRD; enucleation at 15 mo Displaced tube and failed 21 mo RD (repeat) with PVR; scleral buckle tube to cornea with ABK; PK (failed) PVR from previous RD Corneal decompensation at 36 rag; tube failed at 38 mo Dislocated tube (replaced); failed PK
QS QS S S QS
F/S
F
F
F QS
S F
Lensectomy; goniotomy; 34/0 14/0 18 F/QS trabeculotomy; trabeculotomytrabeculectomy; failed tube 16 OD 1 Peter syndrome Trabeculetomy40/0 -6 Failed graft (PK); failed F/F trabeculectomy x2; PK; previous tube; total RD; phthisis seton procedure (see text) 17 OD 182 Sturge-Weber None 32/1 14/0 16 syndrome Pt., Patient; no., number; reeds, medications; F/U, follow-up; 1°, primary; inf. gl., infantile glaucoma; 5', success; Mito Tbx, mitomycin C-trabeculotomy; PK, penetrating keratoplasty; F,failure; juv. gL, juvenile glaucoma; QS, qualified success; OU, both eyes; Cong. Cat., congenital cataract; PHPV,persistenthyperplastic primaryvitreous; PPV,pars planavitrectomy; PI, peripheraliridectomy; AC, anterior chamber; RD, retinal detachment; TRD,traction retinal detachment;PVR, proliferative vitreoretinopathy; ABK, aphakic bullous keratOpathy;2°secondary; ECCE,extracapsulescataract extraction; PCIOL, posteriorchamberintraocular lens. BBilateral. uUnilateral. 15
OD
22
Lowe syndrome
44
Journal ofAAPOS Volu#ne 1 Number 1 March 1997
Donahue et aL
TABLE 2. Complications Number Retinal detachment Corneal decompensation Failed corneal graft* Tube dislocation Persistent inflammation Vitreous occluding tube Secondary membrane *Six patients had grafts.
5 5 5/6 3 2 1 1
% 22 22 83 14 9 5 5
DISCUSSION Although the first choice of surgical procedure for primary infantile glaucoma is generally considered goniotomy or trabeculotomy, the procedure of choice when these fail is less clear. 1I Filtration surgery is often complicated by the failure induced by the exuberant wound healing response of young people. Antimetabolites, such as mitomycin, have been efficacious in adults and are being studied in children12.13; however, the long-term results with these agents are unknown. In addition, filtering procedures put a bleb at or near the limbus. This complicates contact lens wearing in aphakic individuals. Argon laser trabeculoplasty has been attempted in children with limited success. ,4 Baerveldt implants were introduced in 1990. The Baerveldt implant has a nonvalved silicone tube that is placed in the anterior chamber. This is connected to barium-impregnated silicone plates of varying sizes, which now are available in 250, 350 and 500 mm 2 sizes. In our study, we used a 350 mm Baerveldt implant in all but one case. The 350 mm size implant has similar surgical success in adults to that of the 500 mm size. ~s The advantage of Baerveldt implants over the Molteno and Schocket designs is that the Baerveldt plate is placed in only one quadrant of the eye, most often superotemporally. It can be implanted relatively easily. At the University of Iowa, the Baerveldt implant is preferred to the Molteno device for use in children. In adults, the two devices have similar success rates. 16This study did not attempt to compare Baerveldt implants with devices of other types. Adults can be adequately treated with the Baerveldt implant.i, is, 16 We undertook this retrospective review to understand better the risk factors for success and the possible complications associated with the use of Baerveldt implants in the pediatric population. Although we included all children in whom the Baerveldt device had been implanted, its use at the University of Iowa is reserved for patients who have not responded to other surgical procedures. This was evidenced by the mean of 3.0 preoperative medications and 2.8 previous surgical procedures among our patients. The results of this study demonstrate that patients who have had few previous operative procedures and are phakic do relatively well with the Baerveldt implant. Many of the procedures in aphakic eyes failed, however. For these eyes, the number of previous operations was the best predictor of
outcome. The relative success of aphakic individuals with only one previous operation (83 %) compared with those having had more than two previous operations (42 %) may suggest that a seton procedure can be considered the first choice for aphakic glaucoma. Alternatively, this retrospective study may have selected out high-risk eyes by those that failed previous surgical procedures. The most common cause of failure in our series was retinal detachment. This occurred in five eyes (22 %), two of which had had previous retinal detachments before the placement of the Baerveldt implant. One additional aphakic patient underwent a second procedure to remove vitreous from the tube. It is possible that vitreous moving into the anterior chamber and blocking the tube may cause both failure of pressure control and also retinal traction and eventual detachment. If this is true, aphakic individuals may benefit from a vitrectomy at the same time as the tube insertion, and we have begun to do this at the University of Iowa and Vanderbilt University Medical Center. Corneal edema was also a common complication, especially in patients with corneal grafts. Three factors probably contribute to corneal edema in the absence of ocular hypertension. The presence of the tube in the anterior chamber probably creates a foreign body response that causes ocular inflammation. These children could not cooperate with frequent detailed slit-lamp examinations, and chronic low-grade inflammation may not have been detected. Likewise, tube movement in the anterior chamber, as well as intermittent tube-corneal touching during vigorous physical activity, probably also contributed to corneal decompensation. The active lifestyle of children predisposes them toward these complications. It is possible that more frequent examination, with sedation if necessary, combined with more aggressive use of topical steroids could decrease the incidence of this complication. It is noteworthy that a previous report documented graft failure in 42% of 26 adults undergoing penetrating keratoptasty and drainage tube operation.17 The length of follow-up in this study is short at present; the average is only 19 months. However, many of our complications occurred within a few weeks or months after the operation. Fifty percent of the failures occurred within 3 months, and 70% occurred within the first 6 months. Three additional procedures failed after 19 months of functioning. We expect that the incidence of complications will be much lower as the follow-up period progresses, but we cannot be sure of this. We attempted life table analysis, but the sample size was too small to give meaningful results. Although many series classify patients into only two categories (success or failure), we chose to add a third category (qualified success). We included six patients in this category: those who required topical medications to maintain pressure control (three patients), those whose tubes continued to function with adequate pressure but have had a complication (two patients), and one patient who required a modification of the tube (vitreous removal) but
Journal ofAAPOS Volume 1 Number 1 March 1997 whose IOP is still adequately controlled. Although some might wish to classify these eyes as failures, we believe that they represent a group distinct from those who had either tube failure or the devastating complications of phthisis or retinal detachment. Nevertheless, the data are presented in a manner that allows the reader to analyze as he or she desires. Most previous reports of seton surgery for treatment of childhood glaucoma concerned the Molteno implant. Hill and coworkers s have reported the results of Molteno implants in 70 young patients, 53 of whom had undergone previous procedures that failed. Although final pressure control was eventually achieved in 62 % of the patients, only one third were controlled after the initial implant, and most underwent further surgical procedures. T h e rate of eventual pressure control in our series (14 of 23; 61%) is very similar to that reported by Hill and coworkers, s Retinal detachment was also common in their series and occurred in 11 of their 70 patients. T e n of these 11 eyes were aphakic. This is similar to that reported by us. In addition, 12 eyes (17%) had corneal edema. This is also similar to the percentage of patients reported byus as having graft failure or corneal decompensation. It is unclear how many patients in the Hill and coworkers s series had a penetrating keratoplasty at the time of their implant procedure. Goldberg 4 has also reported successful results for 15 eyes with congenital or juvenile glaucoma undergoing Molteno implantation. T h e only previous report of Baerveldt implants in children was by Net]and and Walton6; this included both Molteno implants (n = 16) and Baerveldt implants (n = 7). T h e i r patient population was similar to ours in terms of age, preoperative IOP, and previous surgical procedures. T h e i r combined success rate of 80 % was significantly higher than ours. In addition, their complications were not as severe. However, it is unclear whether these eyes were as advanced as the ones in our series. For example, none of their patients in that series had previous retinal detachments (10 % of our eyes did), none had penetrating keratoplasty (20% of our eyes did), and only 10% were aphakic (45% of ours were). Nevertheless, both our study and that o f Netland and Walton 6 demonstrated that many uncontrolled pediatric glaucoma patients can achieve success after Baerveldt
Donahue et al.
45
device implantation, especially phakic patients and aphakic patients who have had no previous glaucoma procedures. References 1. HitchingsRA, Lavin MJ, Calthorpe M. Glaucoma drainage tubes: their role in glaucomamanagement.Int Ophthalmol 1989;13:151-7. 2. SiegnerSW, Netland PA, Urban RCJr, WilliamsAS, RichardsDW, Latina MA, et al. Clinical experiencewith the Baerveldtglaucoma drainage implant. Ophthalmology1995;102:1298-307. 3. Molteno ACB, Ancker E, Van Biljon G. Surgical technique for advancedjuvenileglaucoma.Arch Ophthalmol 1984;102:51-8. 4. Goldberg I. Management of uncontrolled glaucoma with the Molteno system.Aust N Z J Ophthalmol 1987;15:97-107. 5. Hill RA, Heuer DK, Baerveldt G, Minckler DS, Martone JF. Molteno implantationfor glaucomain young patients. Ophthalmology 1991;98:1042-6. 6. Netland PA, Walton DS. Glaucoma drainage implants in pediatric patients. Ophthalmic Surg 1993;24:723-9. 7. Billson F, Thomas R, AylwardW. The use of two-stage Molteno implants in developmental glaucoma. J Pediatr Ophthalmol Strabismus 1989;26:3-8. 8. MunozM, Tomey KF, Traverso C, Day SH, Senft SH. Clinical experience with the Molteno implant in advanced infantile glaucoma.J Pediatr Ophthalmol Strabismus 1991;28:68-72. 9. Molteno AC, Polkinghorne PJ, Bowbyes JA. The Vicryl tie technique for inserting a draining implant in the treatment of secondaryglaucoma.Aust N ZJ Ophthalmol 1986;14:343-54. 10. Brooks SE, Dacey MP, Lee MB, BaerveldtG. Modificationof the glaucoma drainage implant to prevent early postoperative hypertension and hypotony: a laboratory study. Ophthalmic Surg 1994;25:311-6. 11. BeanchampGR, ParksMM. Filteringsurgeryin children:barriersto success. Ophthalmology1979;86:170-80. 12. Whiteside-MichaelJ, LiebmannJM, Ritch R. Initial 5-fluorouracil trabeculectomyin young patients. Ophthalmology1992;99:7-13. 13. Singh G, Kaur J, Dogra A. Intraoperative mitomycin C in complicated glaucomas. IndianJ Ophthalmol 1993;41:78-80. 14. WilenskyJT, Weinreb RN. Early and late failures of argon laser trabeculoplasty. Arch Ophthalmol 1983;101:895-7. 15. Lloyd MA, BaerveldtG, FellenbaumPS, Sidoti PA, Minckler DS, MartoneJF, et al. Intermediate-termresults of a randomizedclinical trial of the 350- versus the 500-mm2 Baerveldt implant. Ophthalmology 1994;101:1456-63. 16. SmithMF, DoyleJW, SherwoodMB. Comparisonof the Baerveldt glaucomaimplant with the double-plateMolteno drainageimplant. Arch Ophthalmol 1995;113:444-7. 17. SherwoodMB, SmithMF, DriebeWTJr, Stern GA, BenekeJA,Zam ZS. Drainage tube implants in the treatment of glaucomafollowing penetratingkeratoplasty.Ophthalmic Surg 1993;24:185-9.
Background:The efficacy of Baerveldt implant (Pharmacia & Upjohn, Inc., Kalamazoo, Mich.) surgery in the treatment of advanced childhood glaucoma is unknown. Methods:We reviewed the results of 23 consecutive 350 mm Baerveldt implants in 20 eyes of 17 children. Results were classified as follows: (1) success; no further reoperation, no decrease in vision, and intraocul'ar pressure at last follow-up less than 21 mm Hg with no medications; (2) qualified success; medication necessary to bring intraocular pressure to less than 21 mm Hg or complication not associated with tube failure; and (3) failure; intraocular pressure >20 mm Hg, tube failure complication or reoperation causing tube removal, phthisis, or enucleation.Resu/ts:Original glaucoma types were bilateral aphakic (five), unilateral aphakic or persistent hyperplastic primaryvitreous (four), primary infantile (four), juvenile (three), secondary (two), Peter syndrome (one), and Lowe syndrome (one). Patients had undergone a mean of 2.8 previous intraocular procedures. Mean preoperative intraocular pressure was 33.6 mm Hg; average number of preoperative glaucoma medications was 3.0. Mean follow-up was 19 months (range, 6 to 32 months). Eight procedures were considered successful (mean intraocular pressure 15.5 mm Hg), six were qualified successes (mean number of medications 0.8; mean intraocular pressure 16 mm Hg), and nine failed. Two eyes in the qualified success group do not have useful vision as a result of complications. Complications included retinal detachment (five), corneal decompensation (five), corneal graft rejection in five of six grafts; dislocated tubes (three), and recurrent uveitis (two). One ofthese eyes is phthisic and one has been enucleated. Onlytwo of nine procedures in eyes with a history of one or no previous intraocula r operations failed, whereas seven of 13 proced ures in eyes with a history ofthree or more previous procedures failed. Onlyseven of 13 procedures in aphakic eyes were successes or qualified successes, whereas seven of 10 procedures in phakic eyes had good results. Conclusion:Baerveldt implants can produce good short-term results, especially in phakic eyes. Aphakic eyes and eyes that have undergone multiple procedures are at a much higher risk for devastating complications. (J AAPOS 1997;1:41-5)
p
atients with primary and secondary childhood glaucomas are notoriously difficult to manage. Once initial procedures, such as trabeculotomy or goniotomy, have failed, a second choice is usually a filtering procedure with or without an antimetabolite. The longterm efficacy of antimetabolites is unclear, and filtering blebs present a potential problem for the aphaldc contact lens wearing patient. Seton surgery is now relatively well accepted for the treatment of advanced glaucoma in adults. 1-3Early reports have demonstrated relatively good success of Molteno
From the University of Iowa Hospitals and Clinics, Iowa City. Address reprint requests to Sean P. Donahue, MD, PhD, Depa*'tment of Ophthalmology & Visual Sdenees, 1215 21st Ave. South, Vanderbilt Medical Center East, Nashville, TN37232-8808. *None of the authors has any financial interest in the Baevveldt shunt implant or any competing produet. Baerveldt implant is a trade name of Pharmacia & Upjobn, Inc., Kalamazoo, Michigan. ¢S. P. D. is now associated with Vanderhilt University Children's Eye Center, Nashville, Tennessee. Copyright © 1997 by the American Association for Pediatric Ophthalmology and Strabismus. 1091-8531/9755.00+ 0 75/1/80309
Journal of AAPOS
implants (IOP, Inc., Costa Mesa, Calif.) in children. 4-8 However, the efficacy of Baerveldt implant (Pharmacia & Upjohn, Inc., Kalamazoo, Mich.) surgery in advanced childhood glaucoma (aphakic patients and patients having undergone several surgical procedures) is unknown, although an early report in seven patients is favorable.~We retrospectively evaluated our experience with 23 Baerveldt implants in 20 eyes of 17 patients. METHODS A surgical review demonstrated that 23 Baerveldt implants had been performed on patients less than 21 years of age at the University of Iowa Hospitals since 1990. These charts were reviewed, and previous procedures, diagnoses, intraocular pressure, and age at the operation were determined. In addition, it was noted whether the patient was receiving any preoperative medications. Outcome endpoints were last intraocular pressure, number of medications, and the presence of any complications. At the University of Iowa, Baerveldt implants are usually placed in a single procedure, as described by Molteno. 9 A 350 mm Baerveldt implant was used in all but one case, in March 1997 41
42
Donahue et al.
which a 200 mm implant was used. To prevent immediate postoperative hypotony, a 8-0 polyglactin suture was placed around the tube before the tube entered the plate as described previously.6,7,9The suture degenerates after 2 to 3 weeks, and the tube opens. In later procedures, the tube was fenestrated to allow some drainage during the time before tube opening. 1° During the period before tube opening, pressure control was achieved with acetazolamide (Diamox) or p-blockers as necessary. Procedures were considered successful if intraocular pressure was less than 21 mm Hg on no medications, and no complications occurred. A qualified success meant that either medications were needed to produce an intraocular pressure (IOP) less than 21 mm Hg or that a complication occurred, but the shunt continued to function. Failures were defined as patients with IOP greater than 21 mm Hg on medication, complications causing failure or removal of the tube, or loss of the eye. RESULTS Patients and their diagnoses are summarized in Table 1. Four eyes among three patients had primary infantile glaucoma, whereas three eyes among two patients had juvenile glaucoma. Nine eyes among eight patients were aphakic, whereas one was pseudophakic. The patient with Lowe syndrome became aphakic with a lensectomy and vitrectomy after the implant procedure in one eye. Age at the operation ranged from 1 to 191 months (15 years of age). Mean preoperative IOP was 33 mm Hg, and patients averaged three mean preoperative glaucoma medications. Eight procedures were successful. Three of these were in eyes of patients with primary infantile glaucoma, four were in patients with aphakic glaucoma, and one was in a patient with Sturge-Weber syndrome. One successful procedure was a repeat Baerveldt implant after the first tube dislocated out of the anterior chamber. Only one successful procedure was associated with a visionthreatening complication. Patient 13 had a tractional retinal detachment from proliferative vitreoretinopathy. However, this was probably a result of the retinal detachment that was surgically treated before the implant was placed. One patient with well-controlled juvenile glaucoma was hit in the affected eye by a tennis ball and suffered a traumatic hyphema, which resolved without affecting the status of the tube. Six procedures were qualified successes. Three eyes (two patients) were considered qualified successes because the pressure was controlled with topical medications. One patient needed an additional procedure to remove vitreous from the tube but is maintaining adequate pressure control. The other two patients suffered corneal decompensation but have maintained adequate pressure control. Nine procedures failed. Three procedures failed and were repeated; one of these repeat procedures also failed. The other five procedures failed because of inadequate pressure control (two), recurrent inflammation necessitating tube removal (two), or retinal
Journal ofAAPOS Volume I Number 1 March 1997
detachment (the tube was removed during the retinal reattachment procedure). The most common complication was retinal detachment (Table 2). This occurred in five eyes among four patients. All five eyes were aphakic at the time of the detachment. In one patient with bilateral aphakia and high myopia (Patient 13), however, bilateral retinal detachments had previously occurred and been treated with scleral buckling in the weeks immediately before the implants were placed. Proliferative vitreoretinopathy and traction retinal detachments eventually developed in both of these eyes. It is thus unclear whether the tractional retinal detachments observed in these two cases were caused by the implant operation or were merely a complication of the previous retinal detachment. Corneal decompensation was also a common complication, occurring in five eyes. Three eyes (two patients) had penetrating keratoplasty before tube placement. In two of these three eyes, the graft failed. All three corneal transplants performed in eyes with functioning shunts failed. Only one of six corneal grafts remained viable. Two eyes had persistent inflammation. In one eye, this led to a tractional retinal detachment, phthisis, and enucleation. A second eye had recurrent sterile uveitis that caused failure of a penetrating keratoplasty. The implant was eventually removed in this patient, and the eye underwent cyclodestructive procedures. One eye, treated previously with lensectomy for a unilateral congenital cataract, eventually had a secondary membrane develop. This was probably not related to the implant. The membrane has not been treated, but the pressure is well controlled. In three eyes the tube became dislocated in the anterior chamber. Analysis of these data demonstrated two trends associated with tube failure: number of previous operations and surgical aphakia. The sample size was too small, however, for these trends to reach statistical significance with Fisher's Exact Test. Of nine procedures in eyes that had undergone one or fewer previous intraocular operations, only two failed (22%). Of the 13 procedures performed in eyes with a history of more than two previous operations, however, seven failed (54%). The presence of multiple previous intraocular surgical procedures should therefore be considered a relative risk factor for failure. Phakic patients did better than aphakic patients in this study. Of the 10 procedures undertaken in phakic eyes, only three failed (30%), and two of these failures occurred in one eye of a patient with Peter syndrome. In comparison, the failure rate for the 13 procedures undertaken in aphakic eyes was 54%. The high failure rate in aphakic eyes might have been caused by the number of previous procedures. Only one of the six aphakic eyes in which lensectomyor vitrectomywas the only previous intraoculaf procedure failed (17%), whereas four of seven aphakic eyes having three or more previous intraocular procedures failed (58%). Although none of these trends reached statistical significance, it appears that the aphakic patient who has had multiple previous procedures is at higher risk for shunt failure.
Journal ofAAPOS Volume 1 Number I March 1997
Donahue et al.
43
TABLE 1. Complications of Baerveldt implant surgery
Pt, no.
Age at operation Eye (mo) Diagnosis
1 0 D
8
1° inf. gl.
Previous procedure(s) Trabeculotomy; goniotomy x2; Mito Tbx Trabeculotomyx2; PK; goniotomy Trabeculetomy;PK; goniotomy
Preoperative lOP/no, reeds
Last lOP/no. reeds
F/U (mo)
36/1
16/0
11
25/3
20/0
33
40/2
--
30
2
OD
53
1° inf. gh
2
OS
57
1° inf. gl.
3
OS
191
1° inf. gl.
Tbx x3; cyclodialysis; Scheie drainage; goniotomy
30/4
16/0
12
4 5
OS OD
114 174
Juv. gl. Juv. gl.
Goniotomy;Mito Tbx None
25/2 21/3
16/2 14/1
11 16
5 6
OS OS
174 5
Juv. gl Cong. Catu
24/3
14/1
15
7 8
OS OS
6 50
Cong. Cat.~ PHPV
None Goniotomyx2; lensectomy; trabeculetomy Lensectomy Lensectomy
35/1 28/2 46/2
10/0 15/0 18/0
18 16 9
9
OS
56
14
OS
78
Lensectomy;lysis of adhesions; 38/4 surgical PI; reformation of AC; membrane discission; previous seton procedure Lensectomy 36/2
15/0
10
Cong. cat.B; chronic angle closure Cong. cat.B
--
16
11
OS
90
PHPV
Lensectomy; membranectomy; Mite Tbx
28/2
--
27
12
OS
169
Cong. Cat.B
Lensectomy
32/5
--
21
13
OD
167
Cong. Cat?
Lensectomy;scleral buckle
47/3
19/1
37
13 14
OS OD
172 96
Cong. cat.B 2 ° glaucoma (traumatic)
Lensectomy;scleral buckle ECCE;PCIOL; PPV;membranectomy
32/2
17/0
20
49/2
--
38
Complications
Outcome
Recurrentuveitis; failed PK; tube removed (15 mo) Traumatichyphema; resolved without incident QS Tenon's cysts, OU; resolved spontaneously Peaked pupil, OS Secondarymembrane Vitreous to tube (PPV and tube repositioned posteriorly) First tube pulled out of AC
Macula off RD; seton removed at 9mo Persistent inflammation; exposed plate; TRD; enucleation at 15 mo Displaced tube and failed 21 mo RD (repeat) with PVR; scleral buckle tube to cornea with ABK; PK (failed) PVR from previous RD Corneal decompensation at 36 rag; tube failed at 38 mo Dislocated tube (replaced); failed PK
QS QS S S QS
F/S
F
F
F QS
S F
Lensectomy; goniotomy; 34/0 14/0 18 F/QS trabeculotomy; trabeculotomytrabeculectomy; failed tube 16 OD 1 Peter syndrome Trabeculetomy40/0 -6 Failed graft (PK); failed F/F trabeculectomy x2; PK; previous tube; total RD; phthisis seton procedure (see text) 17 OD 182 Sturge-Weber None 32/1 14/0 16 syndrome Pt., Patient; no., number; reeds, medications; F/U, follow-up; 1°, primary; inf. gl., infantile glaucoma; 5', success; Mito Tbx, mitomycin C-trabeculotomy; PK, penetrating keratoplasty; F,failure; juv. gL, juvenile glaucoma; QS, qualified success; OU, both eyes; Cong. Cat., congenital cataract; PHPV,persistenthyperplastic primaryvitreous; PPV,pars planavitrectomy; PI, peripheraliridectomy; AC, anterior chamber; RD, retinal detachment; TRD,traction retinal detachment;PVR, proliferative vitreoretinopathy; ABK, aphakic bullous keratOpathy;2°secondary; ECCE,extracapsulescataract extraction; PCIOL, posteriorchamberintraocular lens. BBilateral. uUnilateral. 15
OD
22
Lowe syndrome
44
Journal ofAAPOS Volu#ne 1 Number 1 March 1997
Donahue et aL
TABLE 2. Complications Number Retinal detachment Corneal decompensation Failed corneal graft* Tube dislocation Persistent inflammation Vitreous occluding tube Secondary membrane *Six patients had grafts.
5 5 5/6 3 2 1 1
% 22 22 83 14 9 5 5
DISCUSSION Although the first choice of surgical procedure for primary infantile glaucoma is generally considered goniotomy or trabeculotomy, the procedure of choice when these fail is less clear. 1I Filtration surgery is often complicated by the failure induced by the exuberant wound healing response of young people. Antimetabolites, such as mitomycin, have been efficacious in adults and are being studied in children12.13; however, the long-term results with these agents are unknown. In addition, filtering procedures put a bleb at or near the limbus. This complicates contact lens wearing in aphakic individuals. Argon laser trabeculoplasty has been attempted in children with limited success. ,4 Baerveldt implants were introduced in 1990. The Baerveldt implant has a nonvalved silicone tube that is placed in the anterior chamber. This is connected to barium-impregnated silicone plates of varying sizes, which now are available in 250, 350 and 500 mm 2 sizes. In our study, we used a 350 mm Baerveldt implant in all but one case. The 350 mm size implant has similar surgical success in adults to that of the 500 mm size. ~s The advantage of Baerveldt implants over the Molteno and Schocket designs is that the Baerveldt plate is placed in only one quadrant of the eye, most often superotemporally. It can be implanted relatively easily. At the University of Iowa, the Baerveldt implant is preferred to the Molteno device for use in children. In adults, the two devices have similar success rates. 16This study did not attempt to compare Baerveldt implants with devices of other types. Adults can be adequately treated with the Baerveldt implant.i, is, 16 We undertook this retrospective review to understand better the risk factors for success and the possible complications associated with the use of Baerveldt implants in the pediatric population. Although we included all children in whom the Baerveldt device had been implanted, its use at the University of Iowa is reserved for patients who have not responded to other surgical procedures. This was evidenced by the mean of 3.0 preoperative medications and 2.8 previous surgical procedures among our patients. The results of this study demonstrate that patients who have had few previous operative procedures and are phakic do relatively well with the Baerveldt implant. Many of the procedures in aphakic eyes failed, however. For these eyes, the number of previous operations was the best predictor of
outcome. The relative success of aphakic individuals with only one previous operation (83 %) compared with those having had more than two previous operations (42 %) may suggest that a seton procedure can be considered the first choice for aphakic glaucoma. Alternatively, this retrospective study may have selected out high-risk eyes by those that failed previous surgical procedures. The most common cause of failure in our series was retinal detachment. This occurred in five eyes (22 %), two of which had had previous retinal detachments before the placement of the Baerveldt implant. One additional aphakic patient underwent a second procedure to remove vitreous from the tube. It is possible that vitreous moving into the anterior chamber and blocking the tube may cause both failure of pressure control and also retinal traction and eventual detachment. If this is true, aphakic individuals may benefit from a vitrectomy at the same time as the tube insertion, and we have begun to do this at the University of Iowa and Vanderbilt University Medical Center. Corneal edema was also a common complication, especially in patients with corneal grafts. Three factors probably contribute to corneal edema in the absence of ocular hypertension. The presence of the tube in the anterior chamber probably creates a foreign body response that causes ocular inflammation. These children could not cooperate with frequent detailed slit-lamp examinations, and chronic low-grade inflammation may not have been detected. Likewise, tube movement in the anterior chamber, as well as intermittent tube-corneal touching during vigorous physical activity, probably also contributed to corneal decompensation. The active lifestyle of children predisposes them toward these complications. It is possible that more frequent examination, with sedation if necessary, combined with more aggressive use of topical steroids could decrease the incidence of this complication. It is noteworthy that a previous report documented graft failure in 42% of 26 adults undergoing penetrating keratoptasty and drainage tube operation.17 The length of follow-up in this study is short at present; the average is only 19 months. However, many of our complications occurred within a few weeks or months after the operation. Fifty percent of the failures occurred within 3 months, and 70% occurred within the first 6 months. Three additional procedures failed after 19 months of functioning. We expect that the incidence of complications will be much lower as the follow-up period progresses, but we cannot be sure of this. We attempted life table analysis, but the sample size was too small to give meaningful results. Although many series classify patients into only two categories (success or failure), we chose to add a third category (qualified success). We included six patients in this category: those who required topical medications to maintain pressure control (three patients), those whose tubes continued to function with adequate pressure but have had a complication (two patients), and one patient who required a modification of the tube (vitreous removal) but
Journal ofAAPOS Volume 1 Number 1 March 1997 whose IOP is still adequately controlled. Although some might wish to classify these eyes as failures, we believe that they represent a group distinct from those who had either tube failure or the devastating complications of phthisis or retinal detachment. Nevertheless, the data are presented in a manner that allows the reader to analyze as he or she desires. Most previous reports of seton surgery for treatment of childhood glaucoma concerned the Molteno implant. Hill and coworkers s have reported the results of Molteno implants in 70 young patients, 53 of whom had undergone previous procedures that failed. Although final pressure control was eventually achieved in 62 % of the patients, only one third were controlled after the initial implant, and most underwent further surgical procedures. T h e rate of eventual pressure control in our series (14 of 23; 61%) is very similar to that reported by Hill and coworkers, s Retinal detachment was also common in their series and occurred in 11 of their 70 patients. T e n of these 11 eyes were aphakic. This is similar to that reported by us. In addition, 12 eyes (17%) had corneal edema. This is also similar to the percentage of patients reported byus as having graft failure or corneal decompensation. It is unclear how many patients in the Hill and coworkers s series had a penetrating keratoplasty at the time of their implant procedure. Goldberg 4 has also reported successful results for 15 eyes with congenital or juvenile glaucoma undergoing Molteno implantation. T h e only previous report of Baerveldt implants in children was by Net]and and Walton6; this included both Molteno implants (n = 16) and Baerveldt implants (n = 7). T h e i r patient population was similar to ours in terms of age, preoperative IOP, and previous surgical procedures. T h e i r combined success rate of 80 % was significantly higher than ours. In addition, their complications were not as severe. However, it is unclear whether these eyes were as advanced as the ones in our series. For example, none of their patients in that series had previous retinal detachments (10 % of our eyes did), none had penetrating keratoplasty (20% of our eyes did), and only 10% were aphakic (45% of ours were). Nevertheless, both our study and that o f Netland and Walton 6 demonstrated that many uncontrolled pediatric glaucoma patients can achieve success after Baerveldt
Donahue et al.
45
device implantation, especially phakic patients and aphakic patients who have had no previous glaucoma procedures. References 1. HitchingsRA, Lavin MJ, Calthorpe M. Glaucoma drainage tubes: their role in glaucomamanagement.Int Ophthalmol 1989;13:151-7. 2. SiegnerSW, Netland PA, Urban RCJr, WilliamsAS, RichardsDW, Latina MA, et al. Clinical experiencewith the Baerveldtglaucoma drainage implant. Ophthalmology1995;102:1298-307. 3. Molteno ACB, Ancker E, Van Biljon G. Surgical technique for advancedjuvenileglaucoma.Arch Ophthalmol 1984;102:51-8. 4. Goldberg I. Management of uncontrolled glaucoma with the Molteno system.Aust N Z J Ophthalmol 1987;15:97-107. 5. Hill RA, Heuer DK, Baerveldt G, Minckler DS, Martone JF. Molteno implantationfor glaucomain young patients. Ophthalmology 1991;98:1042-6. 6. Netland PA, Walton DS. Glaucoma drainage implants in pediatric patients. Ophthalmic Surg 1993;24:723-9. 7. Billson F, Thomas R, AylwardW. The use of two-stage Molteno implants in developmental glaucoma. J Pediatr Ophthalmol Strabismus 1989;26:3-8. 8. MunozM, Tomey KF, Traverso C, Day SH, Senft SH. Clinical experience with the Molteno implant in advanced infantile glaucoma.J Pediatr Ophthalmol Strabismus 1991;28:68-72. 9. Molteno AC, Polkinghorne PJ, Bowbyes JA. The Vicryl tie technique for inserting a draining implant in the treatment of secondaryglaucoma.Aust N ZJ Ophthalmol 1986;14:343-54. 10. Brooks SE, Dacey MP, Lee MB, BaerveldtG. Modificationof the glaucoma drainage implant to prevent early postoperative hypertension and hypotony: a laboratory study. Ophthalmic Surg 1994;25:311-6. 11. BeanchampGR, ParksMM. Filteringsurgeryin children:barriersto success. Ophthalmology1979;86:170-80. 12. Whiteside-MichaelJ, LiebmannJM, Ritch R. Initial 5-fluorouracil trabeculectomyin young patients. Ophthalmology1992;99:7-13. 13. Singh G, Kaur J, Dogra A. Intraoperative mitomycin C in complicated glaucomas. IndianJ Ophthalmol 1993;41:78-80. 14. WilenskyJT, Weinreb RN. Early and late failures of argon laser trabeculoplasty. Arch Ophthalmol 1983;101:895-7. 15. Lloyd MA, BaerveldtG, FellenbaumPS, Sidoti PA, Minckler DS, MartoneJF, et al. Intermediate-termresults of a randomizedclinical trial of the 350- versus the 500-mm2 Baerveldt implant. Ophthalmology 1994;101:1456-63. 16. SmithMF, DoyleJW, SherwoodMB. Comparisonof the Baerveldt glaucomaimplant with the double-plateMolteno drainageimplant. Arch Ophthalmol 1995;113:444-7. 17. SherwoodMB, SmithMF, DriebeWTJr, Stern GA, BenekeJA,Zam ZS. Drainage tube implants in the treatment of glaucomafollowing penetratingkeratoplasty.Ophthalmic Surg 1993;24:185-9.