Scott IU, Gedde SJ, Budenz DL, Greenfield DS, Flynn HW, Feuer WJ, Mello MO, Krishna R, Godfrey DG. Baerveldt Drainage Implants in Eyes With a Preexisting Scleral Buckle. Arch Ophthalmol. 2000;118(11):1509-1513. doi:10.1001/archopht.118.11.1509
To describe the surgical insertion of a Baerveldt drainage implant and postoperative visual acuity and intraocular pressure (IOP) outcomes in patients with a preexisting scleral buckle.
Medical records of all patients with a preexisting scleral buckle who underwent insertion of a Baerveldt drainage implant at Bascom Palmer Eye Institute, Miami, Fla, from January 1, 1994, through December 31, 1998, were reviewed. Outcome measures included visual acuity and IOP at 1 year.
At 1 year postoperatively, 14 (88%) of 16 patients had stable or improved visual acuity. Preoperatively, mean IOP was 30.9 mm Hg and the mean number of antiglaucoma medications was 3.4; at 1 year postoperatively, mean IOP was 12.0 mm Hg and the mean number of antiglaucoma medications was 0.8 (P<.001). Nine patients (56%) achieved an IOP of greater than 5 and no greater than 21 mm Hg without medication, and an additional 7 (44%) achieved this level of IOP control with medication. No patient required further surgery for uncontrolled IOP during the follow-up interval, which ranged from 19.1 to 45.5 months.
Baerveldt drainage device insertion behind or over a preexisting encircling band is often successful in managing refractory glaucoma in patients who have undergone previous scleral buckling procedures.
PRIMARY OPEN-ANGLE glaucoma (POAG) occurs more commonly in patients with a rhegmatogenous retinal detachment than in the general population. The prevalence of POAG has been reported as 1.1% to 3.0% in the general population,1- 3 compared with 4.0% to 5.8% in patients with retinal detachments.4,5 The presence of myopia as a common risk factor for POAG and retinal detachment and the reported relationship between miotic therapy and retinal detachment does not fully account for the association between POAG and retinal detachment.5
Glaucoma also may be associated with the treatment of a retinal detachment. Angle-closure glaucoma has been reported in 0.4% to 4.4% of patients after scleral buckling procedures5- 7; this association has been attributed to occlusion of the vortex veins by a posteriorly positioned or high scleral buckle, which causes congestion and forward rotation of the ciliary body with secondary angle closure.8 In addition, postoperative glaucoma has been reported in 2% to 48% of eyes after retinal detachment repair with intraocular gas or silicone oil retinal tamponade.9- 13 In these cases, pressure elevation may be due to pupillary block angle closure, inflammation, preexisting angle abnormality, or the presence of silicone oil bubbles in the angle.
Patients with medically uncontrolled glaucoma who have undergone previous scleral buckling procedures often present a difficult management challenge. Conjunctival scarring and recession caused by previous retinal surgery may decrease the likelihood of successful filtration surgery, even with the adjunctive use of an antimetabolite.14 Cyclodestructive procedures have often been used in this clinical setting. However, because cyclodestructive procedures may be associated with significant complications, including vision loss, choroidal detachment, and hypotony, and because their effect is frequently unpredictable and irreversible, they represent a less-than-ideal management option, especially in eyes with good visual potential.15- 17 Glaucoma drainage implants offer an important alternative surgical approach.
Our purpose is to describe the surgical insertion of a Baerveldt drainage implant and postoperative visual acuity and intraocular pressure (IOP) outcomes in a series of eyes with a preexisting scleral buckle.
Approval for the study was obtained from the School of Medicine Medical Sciences Subcommittee for the Protection of Human Subjects, University of Miami, Miami, Fla (Institutional Review Board protocol 98/378). The medical records were reviewed for all patients with preexisting scleral buckles who underwent insertion of a Baerveldt drainage implant at the Bascom Palmer Eye Institute, Miami, between January 1, 1994, and December 31, 1998. Data collected included demographic information, length of follow-up, operative factors, functionality of the glaucoma drainage implant, retinal status, additional procedures performed, and preoperative and postoperative best-corrected visual acuity, IOP, and number of antiglaucoma medications used.
The surgical procedure performed was similar in all patients. An attempt was made to select the quadrant for Baerveldt implantation with the least amount of episcleral hardware from previous retinal surgery (ie, if a segmental or radial scleral buckle rather than an encircling element was present, then a quadrant without the buckle was selected as the surgical site for Baerveldt drainage implant insertion). A fornix-based flap of conjunctiva and Tenon capsule was raised in this quadrant. In patients with an encircling band, dissection was performed to expose the encapsulated episcleral encircling band and provide adequate space posterior to the band for placement of the Baerveldt implant. The fibrous capsule overlying the encircling band was excised in the quadrant of surgical implantation. A 250 (7 patients [44%]) or 350-mm2 (9 patients [56%]) Baerveldt device was positioned over or behind the encircling band and over the adjacent recti muscles. The implant was sutured to sclera or the encircling band using the fixation holes. The tube was tied off in a watertight fashion using a 7-0 polyglactin suture. Two to 4 fenestrations of the tube (at the surgeon's discretion) were performed anterior to the polyglactin ligature using a TG160 needle (Ethicon; Somerville, NJ). The tube was trimmed to an appropriate length and inserted into the anterior chamber through a 23-gauge needle tract. A glycerin-preserved donor sclera graft was used to cover the limbal portion of the tube. Conjunctiva and Tenon capsule were reapproximated to the limbus using 7-0 polyglactin suture. Subconjunctival injections of cefazolin (50 mg) and dexamethasone sulfate (2 mg) were given at the conclusion of the case.
Sixteen patients (16 eyes) with a preexisting scleral buckle underwent Baerveldt drainage implant insertion during the study period. Mean (SD) interval between scleral buckle procedure and implant placement was 44 months (60 months); median, 20 months (range, 3 weeks to 17 years). Patient characteristics and study results are displayed in Table 1 and Table 2. All but 1 eye had a preexisting encircling element. The scleral buckle in the single remaining eye was a radial element in the superotemporal quadrant; the Baerveldt drainage implant was placed in the superonasal quadrant in this eye. Three additional eyes had an encircling element and a superotemporal radial element; in each of these eyes, the Baerveldt drainage implant was placed in the superonasal quadrant. In all other eyes, the Baerveldt drainage implant was placed in the superotemporal quadrant. The implant was sutured to the scleral buckle in 2 eyes (12%); the sclera in 12 eyes (75%); and both in 2 eyes (12%).
In 14 patients (88%), Snellen visual acuity was stable or improved 1 year postoperatively. The cause of vision loss in the remaining 2 patients (12%) was progression of preexisting glaucoma in one patient and epiretinal membrane formation in the other.
Preoperative IOP was 30.9 mm Hg and the mean number of preoperative antiglaucoma medications was 3.4. At 1 year postoperatively, mean IOP was 12.0 mm Hg and the mean number of antiglaucoma medications was 0.8. The postoperative reductions in IOP and in number of antiglaucoma medications were statistically significant (P<.001, paired t test). Nine patients (56%) achieved successful IOP control (defined as IOP >5 and ≤21 mm Hg at 1 year postoperatively and at final follow-up without antiglaucoma medication or glaucoma reoperation). An additional 7 patients (44%) achieved IOP control with medication at 1 year postoperatively and at final follow-up. Successful IOP control without medication at 1 year was achieved in 2 patients (29%) with a 250- and 7 patients (78%) with a 350-mm2 Baerveldt device (P = .13). Throughout the follow-up interval, which ranged from 19.1 to 45.5 months, all patients maintained an IOP of greater than 5 but no greater than 21 mm Hg (with or without medication) without glaucoma reoperation.
No intraoperative complications occurred in the current series. Postoperative events included choroidal effusions that resolved spontaneously within 2 weeks in 5 patients (31%), hyphema that resolved spontaneously within 2 weeks in 4 patients (25%), mild vitreous hemorrhage that resolved spontaneously within 1 month in 1 patient (6%), limited suprachoroidal hemorrhage that resolved spontaneously within 3 weeks in 1 patient (6%), and recurrent retinal detachment at 2 weeks repaired using pars plana vitrectomy and fluid-gas exchange in 1 patient (6%). The drainage implant remained functional in all patients. Persistent hypotony (IOP ≤5 mm Hg on ≥2 consecutive visits), hypotony maculopathy, tube occlusion, diplopia, scleral buckle migration or exposure, Baerveldt drainage implant migration or exposure, and epithelial ingrowth did not develop in any patient in this series.
Surgical treatment of glaucoma in patients with preexisting scleral buckles may be difficult. The conjunctiva is frequently scarred or recessed owing to previous surgery. Such eyes are at high risk for filtering surgery failure, even if an adjunctive antimetabolite is used.14
Glaucoma drainage implants have been used increasingly in the management of refractory glaucoma.14,18- 27 These devices consist of a silicone tube that is inserted into the eye via a scleral fistula and that functions by shunting aqueous humor from the anterior or posterior chamber to an equatorial explant. Fibrous encapsulation of the explant produces a reservoir into which aqueous flows freely, and aqueous then passes through the capsular wall into the orbital tissues.
The various glaucoma drainage devices differ with regard to the design of their equatorial explant or plate. Several glaucoma drainage implants have used a silicone encircling strip for the explant, which ranges from 90° to 360° in circumferential extent.28- 31 As with other nonvalved glaucoma drainage implants, temporary restriction of aqueous flow through these implants is required until fibrous encapsulation of the explant has occurred, to avoid early postoperative hypotony.
Sidoti et al32 described the use of a silicone tube to shunt aqueous from the anterior segment to the fibrous capsule surrounding a previously placed episcleral encircling element in 13 patients. With this technique, the fibrous capsule already present around the scleral buckle was used as the reservoir for aqueous collection, thereby negating the need to introduce additional hardware into the orbit in the form of an episcleral explant. Because a preformed capsule was already present, temporary restriction of aqueous flow to allow encapsulation was not needed. Successful control of IOP (≤21 mm Hg) with or without medication was achieved in 11 (85%) of 13 eyes. Early postoperative hypotony was observed in 3 (23%) of 13 eyes, presumably due to leakage of aqueous around the tube at its insertion into the band capsule or from the scleral fistula. Obstruction of the distal tube opening by episcleral fibrous tissue was a frequent complication, occurring in 5 (38%) of 13 eyes.
Smith et al33 recently reported 11 eyes that received modified aqueous drainage implants after retinal surgery with scleral band placement. A technique similar to that described by Sidoti et al32 was used in 7 eyes; however, a long-valved (plateless) Krupin-Denver tube was inserted into the encapsulated band instead of simple silicone tubing so that immediate postoperative hypotony could be avoided. In the remaining 4 eyes, a 200- or 250-mm2 Baerveldt implant with the wings trimmed off was inserted underneath the encircling band with an occluding 8-0 supramid stent. Life-table success rates (defined as a final IOP of ≤21 mm Hg without the need for further glaucoma surgery and with no loss of visual acuity) of all eyes with or without medication were 82% and 73% at 1 and 2 years, respectively. No cases of early postoperative hypotony occurred, although 2 (29%) of 7 eyes with long Krupin-Denver–valved implants experienced distal tube occlusion requiring surgical revision.
Our report describes an alternative management approach in eyes with medically uncontrolled glaucoma and preexisting scleral buckles. The Baerveldt implant was positioned over or behind the encircling band, depending on the location of the band, or in the patient with only a radial element, the Baerveldt implant was positioned in a quadrant without retinal hardware. In cases where the band was located anteriorly just behind the insertion of the recti muscles, the implant was inserted behind the band. If the band was more posterior in location, the implant was placed over the band. An effort was made to excise the capsule overlying the band, in the quadrant of surgical implantation. We hypothesize that this allowed contiguous encapsulation of the encircling band and Baerveldt plate. Studies suggest that the extent of IOP reduction with glaucoma drainage implants is directly proportional to the surface area of the fibrous capsule.34,35 Although not a statistically significant difference, perhaps because of the small sample size, a greater percentage of patients with a 350-mm2 Baerveldt device achieved IOP control without medication at 1 year than patients with a 250-mm2 Baerveldt device in our study (78% vs 29%; P = .13). Furthermore, a capsule that encompasses the Baerveldt plate and episcleral band theoretically should offer greater surface area and IOP reduction than a capsule around the Baerveldt plate or episcleral band alone. It remains unclear whether encircling element size influences the degree of IOP reduction. The size of the episcleral encircling element could not be determined by retrospective chart review in most of our patients, as most had undergone their scleral buckling procedures at other institutions before referral.
No attempt was made to dissect underneath the encircling band. The silicone tube was not placed under the band for fear that the band-scleral compression would close off the tube. Sidoti et al36 reported a case of Baerveldt implantation complicated by scleral perforation at the site of severe ectasia underlying a previously placed scleral buckle. Areas of thin sclera may remain hidden underneath an encircling element. Our technique minimizes the risk for scleral perforation by avoiding dissection under the encircling band. In some of our cases, the Baerveldt implant was sutured directly to the buckle, further reducing the risk of inadvertent scleral perforation during fixation of the implant.
Occlusion of the distal tube opening by fibrous tissue is not an infrequent complication after placement of an anterior chamber tube to encircling band. Sidoti et al32 observed this complication in 5 of 13 patients undergoing aqueous tube shunt to a preexisting episcleral band, and Smith et al33 noted it in 3 of 7 patients undergoing long Krupin-Denver–valved implantation to preexisting episcleral bands. Surgical revision was required in all cases to resolve the distal tube occlusion.32,33 This complication is rarely observed after Baerveldt implantation and was not seen in any of our patients. The distal tube opening of the Baerveldt implant lies flush with a ridge at the anterior aspect of the explant, and this configuration at the tube-plate junction appears to be important in maintaining patency during the process of fibrous encapsulation of the device.
Smith et al33 trimmed the wings off 200- or 250-mm2 Baerveldt implants to debulk them before insertion because of concerns that a full-sized implant in the presence of retinal hardware in all 4 quadrants might be unsafe due to a crowding effect. In the current series, there was no difficulty inserting a 250-mm2 implant without trimming the explant, and in several cases a 350-mm2 implant was used. Sufficient posterior dissection between the sclera and Tenon capsule in the quadrant of implantation is necessary to allow the implant to seat comfortably. Furthermore, the plate was positioned over the recti muscles rather than placing the wings underneath the adjacent recti muscles, as is generally recommended with the Baerveldt implant. Previous scleral buckling surgery often produces scarring that may render isolation of the recti muscles difficult. The Baerveldt episcleral plate contains fenestration holes that permit growth of fibrous tissue through the plate, which serves not only to reduce the height of the bleb, but also secures the implant in place. Implant migration or exposure was not observed in any of our patients.
Sidoti et al36 reported 4 cases of epithelial downgrowth after 350-mm2 Baerveldt implantation in eyes with concurrent or recent previous scleral buckling surgery. They suggested that tension on the operative wounds due to placement of multiple episcleral devices caused poor wound closure with imbrication of conjunctival epithelium within the operative wound, allowing epithelial ingrowth. No cases of epithelial ingrowth occurred in our series.
Implantation of a Baerveldt drainage device behind or over a preexisting episcleral encircling band is often a successful management option for refractory glaucoma in patients who have undergone a previous scleral buckling procedure.
Accepted for publication April 28, 2000.
Supported in part by Research to Prevent Blindness Inc, New York, NY.
Corresponding author: Ingrid U. Scott, MD, MPH, Bascom Palmer Eye Institute, PO Box 016880, Miami, FL 33101 (e-mail: firstname.lastname@example.org).