- Email: [email protected]
Krupin Eye Valve with Disk for Filtration Surgery
Krupin Eye Valve with Disk for Filtration Surgery The Krupin Eye Valve Filtering Surgery Study Group" Purpose: The authors evaluate a long posterior tube shunt device with a pressure sensitive valve for filtration surgery in eyes with recalcitrant glaucoma. Methods: The device consisted of an anterior chamber tube connected to an oval (13 X 18 mm) episcleral explant. The explant was designed to maximize the area of surrounding encapsulation while still allowing implantation within one quadrant. A pressure-sensitive and unidirectional slit valve in the tube provided resistance to aqueous humor flow. One-stage implantation without the use of restrictive sutures was performed in 50 eyes with various types of glaucoma unresponsive to prior glaucoma surgery. Results: Mean (± standard error of the mean) preoperative intraocular pressure (lOP) of 36.4 ± 1.6 mmHg was reduced Significantly (P < 0.001) to 8.3 ± 1.3 mmHg on the first postoperative day. Mean anterior chamber depth (scale, 0-4+) was 3.4 ± 0.1. Mean lOP 1 month after surgery was 14.1 ± 1.3 mmHg. The implant was removed from four eyes due to lOP failure (1 eye), external erosion (2 eyes), or endophthalmitis (1 eye). A suprachoroidal hemorrhage occurred in one eye on the first postoperative day. Diplopia developed in one eye after surgery. Mean lOP at last follow-up examination (mean, 25.4 ± 2.4 months; range, 16-36 months) was 13.1 ± 1.3 mmHg. Intraocular pressure was 19 mmHg or lower in 80% of the eyes, 59% of which were without adjunctive antiglaucoma medications. Conclusions: Design features of the Krupin Eye Valve with Disk result in a large area of encapsulation in a single ocular quadrant which functions as an external reservoir for passage of aqueous humor. The valve portion facilitates maintenance of anterior chamber depth during the early postoperative interval. This new therapeutic device can be effective in the long-term control of lOP in glaucomatous eyes not responsive to prior filtration surgery with adjunctive antimetabolite therapy. Ophthalmology 1994;101:651-658
Filtration surgery lowers intraocular pressure (lOP) to less than 21 mmHg in 75% to 95% of eyes when performed as a primary surgical procedure for uncomplicated open-
* Members of the Krupin Eye Valve Filtering Surgery Study Group are listed in the Appendix at the end of this article. Presented in part as a poster at the American Academy of Ophthalmology Annual Meeting, Dallas, November 1992. Supported in part by unrestricted grants to the Departments of Ophthalmology, Northwestern University and Mt. Sinai School of Medicine, and from Research to Prevent Blindness Inc, New York, New York. None of the authors have proprietary interests in Hood Laboratories, Inc. Reprint requests to Theodore Krupin, MD, Department of Ophthalmology, Ward Building 2-186,303 East Chicago Ave, Chicago, IL 60611.
angle or chronic closed-angle glaucoma. 1-4 Adjunctive use of subconjunctival injections of 5-fluorouracil (5_FU)5,6 or intraoperative application of mitomycin C7 ,8 increases the success of filtration in eyes with various secondary types of glaucoma (e.g., uveitic and neovascular glaucoma) and in eyes that have undergone previous ocular surgery, including cataract removal or prior filtration surgery. However, surgical failure still occurs despite adjunctive antimetabolite therapy as a result of scar tissue formation within the sclerostomy opening or between the conjunctival, Tenon's, or scleral tissues. Various types of plastic seton devices have been used to enhance success for controlling lOP in eyes with recalcitrant glaucoma. Most of the currently used filtration surgical devices are based on the design originally described by Molteno,9,l0 which incorporates a scleral explant to promote formation
651
Ophthalmology
Volume 101, Number 4, April 1994
Table 1. Currently Used Posterior Tube Seton Implants Implant
Anterior Chamber Tube
Molteno
Silicone (OD 0.63 mm, ID 0.3 mm) unrestricted tube
Schocket
Silastic (OD 0.64 mm, ID 0.3 mm) unrestricted tube
Scleral Explant Polypropylene rigid plate Plate 13-mm diameter Single or double plates Dual-chamber implant Silicone #20 band 360 in length Silicone oval disc 13 X 18 mm, 1.75-mm high contoured to the globe 0
Krupin
Baerveldt OD
=
Silastic (OD 0.58 mm, ID 0.38 mm) unidirectional and pressure sensitive slit valve: Opening presure, 11 mmHg Closing pressure, 9 mmHg Silicone (OD 0.6 mm, ID 0.3 mm) unrestricted tube
outside diameter of the tube; ID
=
inside diameter of the tube.
of a functioning bleb. An open tube placed into the anterior chamber shunts aqueous humor 10 to 12 mm posterior to the limbus into an area of encapsulation around the explant. Reduction of lOP by these devices is due to the passive, pressure-dependent flow of fluid across the capsular wall. 11 •12 The magnitude of pressure reduction depends on the capsular wall resistance to aqueous humor flow (the thinner the capsule, the lower the lOP) and the total surface area of encapsulation (the larger the surface area, the lower the lOP). Currently used posterior tube shunt implants (Table 1) differ regarding the presence or absence of a flow-limiting valve. Such a valve restricts aqueous humor egress from the anterior chamber below a specified lOP level. This is advantageous for minimizing profound hypotony and flat anterior chamber during the early postoperative period before completion of encapsulation around the episcleral explant. Devices without a pressure-sensitive valve require some type of temporary ligature to prevent excessive flow of aqueous humor through the open tube until encapsulation is completed. Implants also vary in the shape and size of the episcleral portion of the device. The glaucoma valve implant l3 •14 has been modified by enlarging the size of the episcleral explant to maximize the surrounding area of encapsulation while still enabling its placement within one ocular quadrant. The implant (Krupin Eye Valve with Disk, Hood Laboratories, Inc, Pembroke, MA) contains a pressure-sensitive, unidirectional slit valve to provide resistance to aqueous humor flow. The design, surgical implantation procedure, rates of success, and associated complications are presented in this article.
Materials and Methods Glaucoma Disk Explant The implant (Fig 1) consists of an open Silastic tube with an outside diameter of 0.58 mm and an inside diameter
652
Barium-impregnated silicone; "pear-shaped" explant; 3 sizes (surface area, 200, 350, 500 mm2 )
of 0.38 mm. This tube is approximately 20 mm long and is trimmed in length before placement into the anterior chamber (see below in Surgical Implantation section). The
Figure 1. Krupin Eye Valve with Disk. The implant consists of (1) an open Silastic tube (outside diameter, 0.58 mm; inside diameter, 0.38 mm) approximately 20 mm long (A). The tube is trimmed before placement into the anterior chamber. (2) An oval epscleral Silastic disk (13 X 18 mm) (B) with a US-mm high side wall (C). The disk is contoured to conform to the curvature of the globe (see cross-sectional view). (3) An unidirectional, pressure-sensitive slit valve (D). (4) A Silastic platform for fixation of the explant to the sclera (E).
Krupin Eye Valve Study Group· Krupin Eye Valve with Disk distal end of the Silastic tube has horizontal and vertical slits which create a unidirectional, pressure-sensitive valve. The manufacturer tests each valve by infusing saline through the tube with the outside of the valve end exposed to air. The manometric calibration criteria are an opening pressure between 10 and 12 mmHg and a closing pressure between 8 and 10 mmHg. The episcleral portion of the device is an oval Silastic disk, 13 X 18 mm, with a 1.75-mm high side wall. The disk is contoured to conform to the curvature of the globe. A Silastic platform for fixation of the explant to the sclera extends from the anterior edge of the disk.
Surgical Implantation Each implant was flash-sterilized before implantation. Autoclaving might cause the valve slits to stick together. Under sterile conditions, balanced salt solution was irrigated through the open end of the anterior chamber tube to ensure fluid flow across the valve end on the surface of the oval disk. In addition, the valve end was compressed with smooth forceps to ensure uninhibited movement of the valve slits. Surgery was performed under local or general anesthesia. The superior or inferior conjunctiva was incised in one quadrant for 90° to 110°, using either a fornixbased or limbus-based incision. A limbus-based flap (57 mm posterior to the limbus) permitted a smaller conjunctival incision with improved posterior scleral exposure. A fornix-based flap often required radial relaxing incisions on one or both sides to facilitate adequate rectus muscle isolation for placement of the explant. The conjunctiva was dissected posteriorly and two adjacent rectus muscles were isolated using either muscle hooks or traction sutures under the muscle insertions. Adequate scleral exposure was obtained to permit placement of the disk posterior to the insertions of the rectus muscles with the anterior edge of the disk at least 10 mm posterior to the limbus. The disk was inserted with its curved surface against the sclera. The explant was introduced initially with its long axis directed toward the apex of the orbit and was then rotated to a horizontal orientation, with the tube pointing straight toward the anterior chamber. Proper placement of the Silastic disk was posterior to the rectus muscle insertions. Visualization ensured that the disk did not contact the rectus muscles and that there was no pressure from an excessively tight Tenon's tissue that might push the disk anteriorly. The disk was fixated to the sclera with nonabsorbable sutures (5o to 8-0) passed through the fixation platform on both sides of the Silastic tube. These sutures were used to prevent migration of the disk (and therefore the tube). The long anterior chamber tube was placed across the cornea to determine the site of limbal entry into the anterior chamber. This incision was either full-thickness or within the bed of a lamellar scleral flap, 3 to 5 mm wide extending 2 to 5 mm posterior to the limbus. The flap was approximately 50% of scleral thickness and was dissected into clear cornea, thereby improving visualization of the limbal anatomy and allowing more accurate entry
into the anterior chamber. The tract into the anterior chamber was made with a 22- or 23-gauge hypodermic needle, creating a tight entry wound. This tract was made parallel to the plane of the iris within the cornea and approximately 2 to 3 mm in length. A corneal paracentesis track was performed away from the surgical site. Balanced salt solution or a viscoelastic agent was used to deepen a shallow anterior chamber. If vitreous was present in the anterior chamber, an automated vitrectomy was performed before insertion of the tube via a separate entry site. The open end of the Silastic tube was trimmed bevel up to extend 2 to 3 mm into the anterior chamber, the length estimated by laying the tube across the cornea. The tube was placed carefully into the anterior chamber with smooth forceps. The anterior chamber then was reformed with balanced salt solution or viscoelastic through the paracentesis incision. Restrictive sutures were not placed around or within the tube. Tube contact to the iris or lens posteriorly or to the cornea anteriorly required repositioning of the tube through a new limbal entry. A tube that was too long was removed, trimmed, and reinserted. Reformation of the anterior chamber through the paracentesis tract also ensured a snug limbal entry. If fluid leaked around the tube, then sutures were placed adjacent to the tube to close the wound. Episcleral sutures (10-0 nylon) were tied over the tube near the limbus and along the course of the tube to reduce tube movement and to maintain tube contact with the sclera. Suture knots were rotated into the sclera. If a scleral lamellar flap was used, it was re-approximated, with the corners recessed inward to prevent tension on the tube. A donor scleral patch graft (at least 5 X 5mm) was sutured (8-0 to 10-0 nylon) over the limbal portion of the tube (and scleral flap if it was used). The knots of these interrupted nylon sutures were rotated into the episclera. These maneuvers were done to reduce the possibility of postoperative erosion of the tube. Fornix-based flaps were closed with wing and running sutures to close radial relaxing incisions. Separate Tenon's tissue and conjunctiva closures were performed for limbus-based flaps. Absorbable suture (e.g., 8-0 polyglactin) was used for either type of conjunctival closure. The anterior chamber was reformed with balanced salt solution or viscoelastic through the paracentesis tract to ensure a fluid-tight conjunctival closure. Subconjunctival injections of corticosteroids (either short-acting or depot, such as triamcinolone diacetate) and gentamycin sulfate were administered at the conclusion of surgery. Postoperative topical therapy included I %atropine sulfate twice daily, corticosteroid (0.1 % dexamethasone phosphate or 1 % prednisolone phosphate) every 1 to 3 hours, and antibiotic (gentamycin sulfate or tobramycin sulfate) three to four times daily. Corticosteroid therapy was reduced slowly as the anterior chamber inflammatory response subsided.
Patient Selection Patients undergoing implantation of the glaucoma valve disk implant from April I, 1990, to March 30, 1991, were
653
Ophthalmology
Volume 101, Number 4, April 1994
Table 2. Glaucoma Valve Disk Implant: Results during the First Postoperative Month lOP (mrnHg)* Mean ± SEM Range Anterior chamber depth t Mean ± SEM Range lOP
=
intraocular pressure; SEM
=
Day 1
Week 1
Week 2
Month 1
8.3 ± 1.3 0-46
9.5 ± 1.4 2-42
8.8 ± 1.0 3-38
14.1 ± 1.3
3.4 ± 0.1 0-4
3.5 ± 0.1 1-4
3.6 ± 0.1 1-4
3.7 ± 0.1 2-4
5-38
standard error of the mean .
• There were 50 eyes for each time interval.
t
Graded on a scale of 0 to 4+ based on iris/corneal contact, where 0
=
no anterior chamber over the pupil; 1+
= complete iris/corneal contact except over the pupil; 2+ = iris/corneal contact to pupil ruff; 3+ = peripheral
iris/corneal contact; and 4+
=
no iris/corneal contact.
included in this study. Patient informed consent and institutional committee approval were obtained. Surgery was performed on one eye of 50 patients during this intervaL The mean (± standard error of the mean) patient age was 57.2 ± 2.9 years, with an equal number of female and male patients. Thirty patients were white, 13 were African-American, 4 were Hispanic, and 3 were Asian. These patients had various types of medically uncontrolled glaucoma, including neovascular (15 eyes), openangle (16 eyes), chronic closed-angle (12 eyes), and uveitic (7 eyes) glaucoma. Thirteen eyes were phakic, 15 eyes were aphakic, and 22 eyes were pseudophakic (5 anterior chamber and 17 posterior chamber lens implants). These eyes had a mean of2.1 ± 0.2 (range, 1-4) prior surgical glaucoma procedures, including the adjunctive use of subconjunctival 5-FU injections. Seven eyes had a failed Molteno implant, and three eyes had a cyclodestructive procedure. Seven eyes had previous penetrating keratoplasties with clear grafts, and three eyes had prior repair of a retinal detachment. Mean preoperative lOP was 36.4 ± 1.6 mmHg (range, 25-78 mmHg). Patients were using a mean of 2.8 ± 0.1 (range, 1-4) antiglaucoma medications. Sixty percent of patients were treated with a systemic carbonic anhydrase inhibitor. All patients were followed for a minimum of 1 year after surgery.
Results All surgeons had prior surgical experience with posterior tube shunt implants. Surgery was performed in the superior quadrant in 40 eyes (24 eyes temporally and 16 eyes nasally) and in the inferior quadrant in 10 eyes (6 eyes temporally and 4 eyes nasally). Placement of the large disk explant was achieved in all patients. However, in one eye the sclera and retina were perforated during passage of the fixation suture. In this eye, cryotherapy was applied before fixation of the disk explant. An equal number of eyes had the anterior chamber entry wound full-thickness or under a lamellar scleral flap. Technical difficulty was not encountered with placement of the anterior chamber tube.
654
Early Postoperative Intraocular Pressure Control The mean lOP on the first postoperative day was 8.3 ± 1.3 mmHg (Table 2). On the first day, lOP was elevated (> 30 mmHg) in four eyes due to anterior chamber tube complications: in one eye, the tube was intracorneal, one eye had a blood clot within the tube, and in two eyes the internal opening of the tube was occluded by iris. Twelve eyes had an lOP less than 5 mmHg. Anterior chamber depth was graded on a scale of 0 to 4+ based on iris/ corneal contact (Table 2). Mean anterior chamber depth on the first postoperative day was 3.4 ± 0.1. Two eyes had a flat anterior chamber on the first postoperative day: one eye had aqueous humor leakage around the tube with a limbal bleb and one phakic eye had ciliary block (malignant) glaucoma. The latter eye underwent vitreous aspiration and reformation of the anterior chamber. The lOP in this eye was less than 15 mmHg during subsequent examinations. A nonexpulsive suprachoroidal hemorrhage occurred in one eye on the first postoperative day. Although this resolved, the eye lost all vision, and the surgery was considered a failure. One week after surgery, the mean lOP was 9.5 ± 1.4 mmHg. Intraocular pressure remained elevated (>27 mmHg) in the four eyes with anterior chamber tube complications noted on the first postoperative day. The eye with the complicated intracorneal tube underwent surgical revision. Thirteen (26%) eyes had an lOP less than 5 mmHg. The mean anterior chamber depth was 3.5 ± 0.1. Serous choroidal detachment (1-4 quadrants) was present in 17 (34%) eyes. The mean lOP 2 weeks after surgery was 8.8 ± 1.0 mmHg. The anterior chamber tube was open in the eye that earlier had a blood clot within the tube, and its lOP was 15 mmHg. In two eyes, lOP was elevated (>30 mmHg). One eye with an open anterior chamber tube continued to have elevated lOP and was a long-term failure. The implant was removed from this eye. In the second eye, which previously had a low lOP and a well positioned tube, lOP was elevated due to tube migration out of the anterior chamber. The tube was kinked and therefore did not follow a straight path to the episcleral explant. At the
Krupin Eye Valve Study Group . Krupin Eye Valve with Disk time of surgical revision, the tube was surrounded by fibrous tissue posterior to the limbus. This tissue was removed, the tube was reinserted into the chamber, and the external portion of the tube was sutured to episcleral tissue. In this eye, lOP remained less than 15 mmHg. The mean anterior chamber depth at 2 weeks postoperatively was 3.6 ± 0.1. Serous choroidal detachment was present in 16 (32%) eyes. One eye underwent drainage of choroidal fluid. One month after surgery, the mean lOP was 14.1 ± 1.3 mmHg. All eyes had an lOP of 5 mmHg or greater. The lOP was greater than 20 mmHg in eight eyes. One eye underwent surgical revision with excision of dense fibrous tissue over the surface of the episcleral explant. Long-term lOP control in this eye was less than 13 mmHg. Five eyes required anti-glaucoma medical therapy to lower lOP to less than 20 mmHg. Medical therapy was required for long-term lOP lowering in two of these eyes. Two eyes had persistent elevation of lOP despite medical therapy and were considered failures. The mean anterior chamber depth 1 month after surgery was 3.7 ± 0.1. Six eyes still had residual serous choroidal fluid.
Late Intraocular Pressure Control The implant was removed from three eyes after the third postoperative month (Table 3). Erosion of the tissue over the explant occurred in one eye that had had a prior scleral buckling procedure. A second eye had erosion and melting of the conjunctiva at the junction of the tube with the oval disk, resulting in aqueous humor leakage. Attempts at repair were unsuccessful. A limbus-based conjunctival flap had been used in these two eyes with explant erosion. In a third eye, traumatic endophthalmitis developed 8 months after surgery. The implant was removed at the time of vitrectomy in this eye. Excluding the four eyes in which the explant was removed and the eye with a nonexpulsive hemorrhage on the first postoperative day, the mean lOP (45 eyes) at last follow-up examination (mean, 25.4 ± 2.4 months; range, 16-36 months) was 13.1 ± 1.3 mmHg. The glaucoma
valve disk implant was successful in lowering lOP to 19 mmHg or less in 40 (80%) of the original 50 eyes. In 59% of these eyes, adjunctive antiglaucoma therapy was not required. Eleven (73%) of 15 eyes with neovascular glaucoma and 29 (83%) of 35 eyes with other types of glaucoma had an lOP of 19 mmHg or lower (Fisher's exact test; P > 0.5).
Visual Acuity Preoperative visual acuity was 20/100 or better in 20 eyes, 20/200 to 20/400 in 10 eyes, and counting fingers to hand motions in 20 eyes. Visual acuity at last follow-up was 20/100 or better in 18 eyes, 20/200 to 20/400 in 14 eyes, counting fingers to hand motions in 16 eyes, and no light perception in two eyes (1 eye with the suprachoroidal hemorrhage and 1 eye with endophthalmitis). Long-term visual acuity was unchanged (2 or more Snellen acuity lines) in 19 eyes, decreased in 15 eyes, and improved in 16 eyes. Three eyes had decreased vision due to corneal decompensation: two eyes with clear corneal grafts preoperatively and one eye with iridocorneal endothelial syndrome. One eye with corneal graft failure and the eye with iridocorneal endothelial syndrome underwent repeat corneal transplantation with subsequent improved visual acuity. The lOP in these two eyes was less than 15 mmHg before and after corneal transplantation.
Diplopia Postoperative diplopia occurred in one eye. This was the eye that had an inadvertent scleral perforation (and cryotherapy) at the time of inferonasal implant surgery. In this eye, the preoperative visual acuity was 20/100 and the lOP was 30 mmHg. Hypertropia with restriction of motility inferiorly and nasally was noted 3 months after surgery. One year after surgery, the patient still had diplopia in primary gaze with a visual acuity of 20/60. Intraocular pressure was 16 mmHg.
Table 3. Glaucoma Valve Disk Implant: Long-term Results
No. of eyes· lOP (mmHg) Mean ± SEM ,,;19 mmHg (%) No. of medications Mean ± SEM % without therapy
Month 6
Year 1
Last Follow-up (mos) (25.4 ± 2.4)
46
45
45
14.8 ± 0.8 84
12.9 ± 0.8 80
13.1 ± 1.3 80
0.6 ± 0.1
0.7 ± 0.2
0.8 ± 0.3
69
62
59
lOP = intraocular pressure; SEM = standard error of the mean . • Six-month period excludes one eye with suprachoroidal hemorrhage on the first postoperative day and three eyes with removal of the implant within 3 months after surgery; I-year period also excludes one eye with implant removal 8 months after surgery (see text).
655
Ophthalmology
Volume 101, Number 4, April 1994
Discussion The design of the glaucoma disk implant (Fig 1) is based on clinical observations of currently used posterior tube shunt implants. The open anterior chamber tube, a feature common to all these devices, functions as the sclerostomy. The tube has an advantage over a surgical sclerostomy, in that it can be inserted within the anterior chamber beyond areas of peripheral anterior synechiae. In addition, the tube prevents sclerostomy closure and functions as a conduit for aqueous humor flow from the anterior chamber to the attached episcleral explant. The episcleral explant stimulates a fibrovascular reaction that requires several days for complete encapsulation of the plate. 12, 15 The capsule develops as a separate layer from Tenon's tissue. The internal surface of the encapsulation is an open, collagenous meshwork that is not lined by a double layer of cuboidal epithelium and therefore, is not a true cyst. Studies have shown that latex microspheres (as large as 0.2 JLm) and horseradish peroxidase (molecular weight, 50,000) pass freely through the capsular wall and reach the capillaries in the vascular layer. 11 ,12 Posterior tube shunt procedures reduce lOP via a passive, pressure-dependent flow of fluid across the capsular wall. The postoperative lOP relates partially to the surface area of the encapsulated bleb. The Molteno double-plate implant produces lower lOP than the single-plate implant. 16 Design of the disk implant incorporates an enlarged oval plate, in contrast to a Silastic encircling band used in the prior version of the glaucoma valve implant, to maximize surrounding encapsulation. The oval disk has a 1.75-mm high sidewall to expand the volume of encapsulation. In addition, the size (13 X 18 mm) of the oval plate minimizes the extent of surgical dissection compared with a two-plate Molteno implant, by placement of the disk implant in one ocular quadrant between two rectus muscles. Although the Silastic disc can be trimmed with scissors if it is too large, this was not necessary in the current series. Insertion of an open tube into the anterior chamber before completion of fibrous encapsulation around the episcleral explant is associated with excessive flow of aqueous humor, resulting in hypotony, flat anterior chamber, and choroidal detachment. 15 To avoid these complications, Molteno recommends a two-stage implantation to allow encapsulation of the episcleral explant before inserting the tube into the anterior chamber. During the first stage, the explant is sutured in place without inserting the tube into the anterior chamber. Because this operation does not reduce lOP, a trabeculectomy frequently is performed for interim lOP control at a different site from where the tube will be inserted later into the anterior chamber. Postoperative antiglaucoma therapy may be required during this interim period. The second operative stage in which the tube is inserted into the anterior chamber is performed 2 to 6 weeks later. Flow of aqueous humor through the tube creates a posterior filtration bleb by dissecting the fibrous encapsulation off the surface of the episcleral plate. During this interval, lOP
656
frequently is elevated to preoperative (or even higher) levels and marked inflammation is observed in the region of the posterior bleb. Antiglaucoma medications and topical corticosteroids are administered during this transient hypertensive phase which may last several weeks. 16,17 Gradually, lOP declines and the bleb inflammation resolves, permitting withdrawal of antiglaucoma medications. The magnitude of the hypertensive phase is reduced with the double-plate Molteno system. 16 One-stage implantation of an unrestricted posterior tube shunt implant can be performed by temporarily occluding the flow of aqueous humor through the anterior chamber tube until encapsulation around the episcleral plate is completed. 18-21 One technique uses a suture ligature around the tube posterior to the scleral flap or donor graft. This suture is released (or an absorbable suture is allowed to degrade) 2 to 4 weeks after surgery, using either the argon laser or through an overlying conjunctival incision. The ligature also can be applied with a releasable knot nylon suture that is exposed externally. Alternatively, a polypropylene suture is tied to the wall of the anterior chamber portion of the tube and then around the tube as a ligature before insertion of the tube into the anterior chamber. The argon laser is used to cut the suture, opening the tube but stillieaving the suture attached to its wall. Finally, the internal lumen of the tube can be occluded with a 3-0 or 4-0 suture which is either buried subconjunctivally or exposed externally by passing the suture through the conjunctiva. The suture is removed at a later date by pulling the suture via a conjunctival incision or pulling the externalized suture. The design of the Krupin Eye Valve with Disk includes a pressure-sensitive and unidirectional slit valve that provides resistance to aqueous humor flow, which is especially important before there is completion of the fibrous encapsulation around the episcleral explant. This permits implantation of the episcleral disk and insertion of the anterior chamber tube during the same procedure, without the need for restrictive tube sutures. However, functioning of the slit valve requires sufficient aqueous humor production to achieve an lOP similar to the opening pressure of the device: at a lower lOP «8-10 mmHg), there will be no fluid flow across the slit valve. Although hypotony (lOP <5 mmHg) was present in 12 eyes on the first postoperative day, this was associated with anterior chamber shallowing in only two eyes (1 eye with ciliary block glaucoma). Early postoperative hypotony may be attributed to ciliary body hyposecretion or transient aqueous humor leakage around the tube. In addition, increased aqueous humor flow, which will occur across the one-way slit valve during transient intervals of external pressure on the globe (e.g., eyelid squeezing, blinking), may account for the early postoperative hypotony. Potential advantages of a one-stage implantation of the disk implant include elimination of a second operation, more rapid control ofIOP by eliminating the period of uncontrolled glaucoma during the interval of the two-stage technique, and reduction of the magnitude of the hypertensive
Krupin Eye Valve Study Group . Krupin Eye Valve with Disk phase which can be observed after completion of the two-stage technique. However, one-stage implantation of the disk implant, or other posterior tube implant devices using techniques described above, still is associated with an increased occurrence of hypotony and choroidal effusions compared with a planned two-stage implantation technique. These complications also are more frequent with implantation of the double-plate Molteno device. 16 A transient hypertensive phase was observed 6 to 8 weeks after implantation of the disk implant in three (6%) eyes. These eyes responded to temporary medical treatment with aqueous humor suppressant therapy (topical beta-adrenergic antagonist or systemic carbonic anhydrase inhibitor). In these eyes, long-term lOP was less than 20 mmHg without antiglaucoma therapy. This phenomenon may be similar to the bleb remodeling, "high bleb phase," observed after trabeculectomy.22 It is difficult to compare accurately the success rates among the current long tube implants. Success with respect to control ofIOP with these devices depends on the type of glaucoma. The highest success occurs in eyes with open- or closed-angle glaucoma, either phakic or aphakic/pseudophakic, and the lowest success occurs in eyes with neovascular glaucoma. 23 In the current series, lOP results are similar in eyes with or without neovascular glaucoma. Although the lOP results with the Krupin Eye Valve with Disk and other posterior tube shunt implants are encouraging, these devices should be reserved for eyes with extremely poor surgical prognosis or prior failure of filtration procedures with adjunctive antimetabolites. However, primary implantation of a shunt device may be advised in eyes with neovascular glaucoma or closedangle glaucoma with anteriorly located synechiae which prevent a limbal sclerostomy from communicating with the anterior chamber. Posterior tube shunt implant surgery frequently is associated with a higher rate of operative and postoperative complications than trabeculectomy. In the current series, complications unique to seton surgery include blockage of the anterior chamber tube, migration of the tube due to movement of the episcleral plate, erosion of the tube and episcleral plate, and extraocular muscle imbalance. Postoperative diplopia is more frequent when the episcleral plate is placed beneath the rectus muscles. 24 The design features of the Krupin Eye Valve with Disk maximize the size of the scleral explant to optimize the volume of surrounding encapsulation while still permitting implantation within a single ocular quadrant between the rectus muscles. The device provides resistance to aqueous humor flow, thereby allowing primary (i-stage) implantation without the need for restrictive ligatures to achieve safe, early postoperative lowering of intraocular pressure. The large surface area of encapsulation, which develops with the presence of aqueous humor flow onto the plate, results in a large external reservoir for the passage of aqueous humor and a high filtration success in eyes with recalcitrant glaucoma.
Appendix Members of the Krupin Eye Valve Filtering Surgery Study Group: Northwestern University, Chicago, IL-Theodore Krupin, MD, Lisa F. Rosenberg, MD, Jon M. Ruderman, MD; Geisinger Medical Center, Danville, PA-Marianne E. Feitl, MD; Washington University, St. Louis, MOMichael A. Kass, MD, Allan E. Kolker, MD, Martin B. Wax, MD; University ofNebraska, Omaha, NE-Carl B. Camras, MD; New York Eye & Ear Infirmary, New York, NY-Jeffrey M. Liebmann, MD, Robert Ritch, MD; Mt. Sinai School of Medicine, New York, NY-Steven M. Podos, MD, Janet B. Serle, MD, Robert A. Schumer, MD; University of Wisconsin, Madison, WI-Paul L. Kaufman, MD, Todd W. Perkins, MD.
References 1. Watkins PH Jr, Brubaker RF. Comparison of partial-thick-
2. 3. 4. 5. 6. 7.
8.
9. 10. 11.
12. 13. 14.
ness and full-thickness filtration procedures in open-angle glaucoma. Am J Ophthalmol 1978;86:756-61. Spaeth GL, Poryzees E. A comparison between peripheral iridectomy with thermal sclerostomy and trabeculectomy: a controlled study. Br J Ophthalmol 1981;65:783-9. Lewis RA, Phelps CD. Trabeculectomy v thermosclerostomy. A five-year follow-up. Arch Ophthalmol 1984;102: 533-6. Lamping KA, Bellows AR, Hutchinson BT, Afran SI. Longterm evaluation of initial filtration surgery. Ophthalmology 1986;93:91-10 1. The fluorouracil Filtering Surgery Study Group. fluorouracil filtering surgery study one-year follow-up. Am J Ophthalmol 1989; 108:625-35. Weinreb RN. Adjusting the dose of 5-fluorouracil after filtration surgery to minimize side effects. Ophthalmology 1987;94:564-70. Chen C-W. Enhanced intraocular pressure controlling effectiveness. of trabeculectomy by local application of mitomycin C. Trans Asia-Pacific Acad Ophthalmol 1983;9: 172-7. Skuta GL, Beeson CC, Higginbotham EJ, et al. Intraoperative mitomycin versus postoperative 5-fluorouracil in highrisk glaucoma filtering surgery. Ophthalmology 1992;99: 438-44. Molteno ACB. New implant for drainage in glaucoma. Clinical trial. Br J Ophthalmol 1969;53:606-15. Molteno ACB, Straughan JL, Ancker E. Long tube implants in the management of glaucoma. S Afr Med J 1976;50: 10626. Schocket SS. Investigations of the reasons for success and failure in the anterior shunt-to-the-encircling-band procedure in the treatment of refractory glaucoma. Trans Am Ophthalmol Soc 1986;84:743-98. Minckler DS, Shammas A, Wilcox M, Ogden TE. Experimental studies of aqueous filtration using the Molteno implant. Trans Am Ophthalmol Soc 1987;85:368-92. Krupin T, Kaufman P, Mandell A, et ai. Filtering valve implant surgery for eyes with neovascular glaucoma. Am J Ophthalmol 1980;89:338-43. Krupin T, Kaufman P, Mandell AI, et ai. Long-term results of valve implants in filtering surgery for eyes with neovascular glaucoma. Am J Ophthalmol 1983;95:775-82.
657
Ophthalmology
Volume 101, Number 4, April 1994
15. Molteno ACB. Use of Molteno implants to treat secondary glaucoma. In: Cairns JE, ed. Glaucoma. Vol. 1. London: Grone & Stratton, 1986;211-38. 16. Heuer DK, Lloyd MA, Abrams DA, et al. Which is better? One or two? A randomized clinical trial of single-plate versus double-plate Molteno implantation for glaucomas in aphakia and pseudophakia. Ophthalmology 1992;99: 151219. 17. Minckler DS, Heuer DK, Hasty B, et al. Clinical experience with the single-plate Molteno implant in complicated glaucomas. Ophthalmology 1988;95:1181-8. 18. Egbert PR, Lieberman MF. Internal suture occlusion of the Molteno glaucoma implant for the prevention of postoperative hypotony. Ophthalmic Surg 1989;20:53-6. 19. Price FW Jr, Whitson WE. Polypropylene ligatures as a
658
20. 21. 22. 23. 24.
means of controlling intraocular pressure with Molteno implants. Ophthalmic Surg 1989;20:781-3. Cohen JS, Osher RH. Releasable scleral flap suture. Ophthalmol Clin North Am 1988;1:187-97. Ball SF, Loftfie1d K, Scharfenberg J. Molteno rip-cord suture hypopyon. Ophthalmic Surg 1990;21 :407 -12. Scott DR, Quigley HA. Medical management of a high bleb phase after trabeculectomies. Ophthalmology 1988;95: 1169-73. Mermoud A, Salmon JF, Alexander P, et al. Molteno tube implantation for neovascular glaucoma. Ophthalmology 1993; 100:897-902. Smith SL, Starita RJ, Fellman RL, Lynn JR. Early clinical experience with the Baerveldt 350-mm2 glaucoma implant and associated extraocular muscle imbalance. Ophthalmology 1993;100:914-18.
Filtration surgery lowers intraocular pressure (lOP) to less than 21 mmHg in 75% to 95% of eyes when performed as a primary surgical procedure for uncomplicated open-
* Members of the Krupin Eye Valve Filtering Surgery Study Group are listed in the Appendix at the end of this article. Presented in part as a poster at the American Academy of Ophthalmology Annual Meeting, Dallas, November 1992. Supported in part by unrestricted grants to the Departments of Ophthalmology, Northwestern University and Mt. Sinai School of Medicine, and from Research to Prevent Blindness Inc, New York, New York. None of the authors have proprietary interests in Hood Laboratories, Inc. Reprint requests to Theodore Krupin, MD, Department of Ophthalmology, Ward Building 2-186,303 East Chicago Ave, Chicago, IL 60611.
angle or chronic closed-angle glaucoma. 1-4 Adjunctive use of subconjunctival injections of 5-fluorouracil (5_FU)5,6 or intraoperative application of mitomycin C7 ,8 increases the success of filtration in eyes with various secondary types of glaucoma (e.g., uveitic and neovascular glaucoma) and in eyes that have undergone previous ocular surgery, including cataract removal or prior filtration surgery. However, surgical failure still occurs despite adjunctive antimetabolite therapy as a result of scar tissue formation within the sclerostomy opening or between the conjunctival, Tenon's, or scleral tissues. Various types of plastic seton devices have been used to enhance success for controlling lOP in eyes with recalcitrant glaucoma. Most of the currently used filtration surgical devices are based on the design originally described by Molteno,9,l0 which incorporates a scleral explant to promote formation
651
Ophthalmology
Volume 101, Number 4, April 1994
Table 1. Currently Used Posterior Tube Seton Implants Implant
Anterior Chamber Tube
Molteno
Silicone (OD 0.63 mm, ID 0.3 mm) unrestricted tube
Schocket
Silastic (OD 0.64 mm, ID 0.3 mm) unrestricted tube
Scleral Explant Polypropylene rigid plate Plate 13-mm diameter Single or double plates Dual-chamber implant Silicone #20 band 360 in length Silicone oval disc 13 X 18 mm, 1.75-mm high contoured to the globe 0
Krupin
Baerveldt OD
=
Silastic (OD 0.58 mm, ID 0.38 mm) unidirectional and pressure sensitive slit valve: Opening presure, 11 mmHg Closing pressure, 9 mmHg Silicone (OD 0.6 mm, ID 0.3 mm) unrestricted tube
outside diameter of the tube; ID
=
inside diameter of the tube.
of a functioning bleb. An open tube placed into the anterior chamber shunts aqueous humor 10 to 12 mm posterior to the limbus into an area of encapsulation around the explant. Reduction of lOP by these devices is due to the passive, pressure-dependent flow of fluid across the capsular wall. 11 •12 The magnitude of pressure reduction depends on the capsular wall resistance to aqueous humor flow (the thinner the capsule, the lower the lOP) and the total surface area of encapsulation (the larger the surface area, the lower the lOP). Currently used posterior tube shunt implants (Table 1) differ regarding the presence or absence of a flow-limiting valve. Such a valve restricts aqueous humor egress from the anterior chamber below a specified lOP level. This is advantageous for minimizing profound hypotony and flat anterior chamber during the early postoperative period before completion of encapsulation around the episcleral explant. Devices without a pressure-sensitive valve require some type of temporary ligature to prevent excessive flow of aqueous humor through the open tube until encapsulation is completed. Implants also vary in the shape and size of the episcleral portion of the device. The glaucoma valve implant l3 •14 has been modified by enlarging the size of the episcleral explant to maximize the surrounding area of encapsulation while still enabling its placement within one ocular quadrant. The implant (Krupin Eye Valve with Disk, Hood Laboratories, Inc, Pembroke, MA) contains a pressure-sensitive, unidirectional slit valve to provide resistance to aqueous humor flow. The design, surgical implantation procedure, rates of success, and associated complications are presented in this article.
Materials and Methods Glaucoma Disk Explant The implant (Fig 1) consists of an open Silastic tube with an outside diameter of 0.58 mm and an inside diameter
652
Barium-impregnated silicone; "pear-shaped" explant; 3 sizes (surface area, 200, 350, 500 mm2 )
of 0.38 mm. This tube is approximately 20 mm long and is trimmed in length before placement into the anterior chamber (see below in Surgical Implantation section). The
Figure 1. Krupin Eye Valve with Disk. The implant consists of (1) an open Silastic tube (outside diameter, 0.58 mm; inside diameter, 0.38 mm) approximately 20 mm long (A). The tube is trimmed before placement into the anterior chamber. (2) An oval epscleral Silastic disk (13 X 18 mm) (B) with a US-mm high side wall (C). The disk is contoured to conform to the curvature of the globe (see cross-sectional view). (3) An unidirectional, pressure-sensitive slit valve (D). (4) A Silastic platform for fixation of the explant to the sclera (E).
Krupin Eye Valve Study Group· Krupin Eye Valve with Disk distal end of the Silastic tube has horizontal and vertical slits which create a unidirectional, pressure-sensitive valve. The manufacturer tests each valve by infusing saline through the tube with the outside of the valve end exposed to air. The manometric calibration criteria are an opening pressure between 10 and 12 mmHg and a closing pressure between 8 and 10 mmHg. The episcleral portion of the device is an oval Silastic disk, 13 X 18 mm, with a 1.75-mm high side wall. The disk is contoured to conform to the curvature of the globe. A Silastic platform for fixation of the explant to the sclera extends from the anterior edge of the disk.
Surgical Implantation Each implant was flash-sterilized before implantation. Autoclaving might cause the valve slits to stick together. Under sterile conditions, balanced salt solution was irrigated through the open end of the anterior chamber tube to ensure fluid flow across the valve end on the surface of the oval disk. In addition, the valve end was compressed with smooth forceps to ensure uninhibited movement of the valve slits. Surgery was performed under local or general anesthesia. The superior or inferior conjunctiva was incised in one quadrant for 90° to 110°, using either a fornixbased or limbus-based incision. A limbus-based flap (57 mm posterior to the limbus) permitted a smaller conjunctival incision with improved posterior scleral exposure. A fornix-based flap often required radial relaxing incisions on one or both sides to facilitate adequate rectus muscle isolation for placement of the explant. The conjunctiva was dissected posteriorly and two adjacent rectus muscles were isolated using either muscle hooks or traction sutures under the muscle insertions. Adequate scleral exposure was obtained to permit placement of the disk posterior to the insertions of the rectus muscles with the anterior edge of the disk at least 10 mm posterior to the limbus. The disk was inserted with its curved surface against the sclera. The explant was introduced initially with its long axis directed toward the apex of the orbit and was then rotated to a horizontal orientation, with the tube pointing straight toward the anterior chamber. Proper placement of the Silastic disk was posterior to the rectus muscle insertions. Visualization ensured that the disk did not contact the rectus muscles and that there was no pressure from an excessively tight Tenon's tissue that might push the disk anteriorly. The disk was fixated to the sclera with nonabsorbable sutures (5o to 8-0) passed through the fixation platform on both sides of the Silastic tube. These sutures were used to prevent migration of the disk (and therefore the tube). The long anterior chamber tube was placed across the cornea to determine the site of limbal entry into the anterior chamber. This incision was either full-thickness or within the bed of a lamellar scleral flap, 3 to 5 mm wide extending 2 to 5 mm posterior to the limbus. The flap was approximately 50% of scleral thickness and was dissected into clear cornea, thereby improving visualization of the limbal anatomy and allowing more accurate entry
into the anterior chamber. The tract into the anterior chamber was made with a 22- or 23-gauge hypodermic needle, creating a tight entry wound. This tract was made parallel to the plane of the iris within the cornea and approximately 2 to 3 mm in length. A corneal paracentesis track was performed away from the surgical site. Balanced salt solution or a viscoelastic agent was used to deepen a shallow anterior chamber. If vitreous was present in the anterior chamber, an automated vitrectomy was performed before insertion of the tube via a separate entry site. The open end of the Silastic tube was trimmed bevel up to extend 2 to 3 mm into the anterior chamber, the length estimated by laying the tube across the cornea. The tube was placed carefully into the anterior chamber with smooth forceps. The anterior chamber then was reformed with balanced salt solution or viscoelastic through the paracentesis incision. Restrictive sutures were not placed around or within the tube. Tube contact to the iris or lens posteriorly or to the cornea anteriorly required repositioning of the tube through a new limbal entry. A tube that was too long was removed, trimmed, and reinserted. Reformation of the anterior chamber through the paracentesis tract also ensured a snug limbal entry. If fluid leaked around the tube, then sutures were placed adjacent to the tube to close the wound. Episcleral sutures (10-0 nylon) were tied over the tube near the limbus and along the course of the tube to reduce tube movement and to maintain tube contact with the sclera. Suture knots were rotated into the sclera. If a scleral lamellar flap was used, it was re-approximated, with the corners recessed inward to prevent tension on the tube. A donor scleral patch graft (at least 5 X 5mm) was sutured (8-0 to 10-0 nylon) over the limbal portion of the tube (and scleral flap if it was used). The knots of these interrupted nylon sutures were rotated into the episclera. These maneuvers were done to reduce the possibility of postoperative erosion of the tube. Fornix-based flaps were closed with wing and running sutures to close radial relaxing incisions. Separate Tenon's tissue and conjunctiva closures were performed for limbus-based flaps. Absorbable suture (e.g., 8-0 polyglactin) was used for either type of conjunctival closure. The anterior chamber was reformed with balanced salt solution or viscoelastic through the paracentesis tract to ensure a fluid-tight conjunctival closure. Subconjunctival injections of corticosteroids (either short-acting or depot, such as triamcinolone diacetate) and gentamycin sulfate were administered at the conclusion of surgery. Postoperative topical therapy included I %atropine sulfate twice daily, corticosteroid (0.1 % dexamethasone phosphate or 1 % prednisolone phosphate) every 1 to 3 hours, and antibiotic (gentamycin sulfate or tobramycin sulfate) three to four times daily. Corticosteroid therapy was reduced slowly as the anterior chamber inflammatory response subsided.
Patient Selection Patients undergoing implantation of the glaucoma valve disk implant from April I, 1990, to March 30, 1991, were
653
Ophthalmology
Volume 101, Number 4, April 1994
Table 2. Glaucoma Valve Disk Implant: Results during the First Postoperative Month lOP (mrnHg)* Mean ± SEM Range Anterior chamber depth t Mean ± SEM Range lOP
=
intraocular pressure; SEM
=
Day 1
Week 1
Week 2
Month 1
8.3 ± 1.3 0-46
9.5 ± 1.4 2-42
8.8 ± 1.0 3-38
14.1 ± 1.3
3.4 ± 0.1 0-4
3.5 ± 0.1 1-4
3.6 ± 0.1 1-4
3.7 ± 0.1 2-4
5-38
standard error of the mean .
• There were 50 eyes for each time interval.
t
Graded on a scale of 0 to 4+ based on iris/corneal contact, where 0
=
no anterior chamber over the pupil; 1+
= complete iris/corneal contact except over the pupil; 2+ = iris/corneal contact to pupil ruff; 3+ = peripheral
iris/corneal contact; and 4+
=
no iris/corneal contact.
included in this study. Patient informed consent and institutional committee approval were obtained. Surgery was performed on one eye of 50 patients during this intervaL The mean (± standard error of the mean) patient age was 57.2 ± 2.9 years, with an equal number of female and male patients. Thirty patients were white, 13 were African-American, 4 were Hispanic, and 3 were Asian. These patients had various types of medically uncontrolled glaucoma, including neovascular (15 eyes), openangle (16 eyes), chronic closed-angle (12 eyes), and uveitic (7 eyes) glaucoma. Thirteen eyes were phakic, 15 eyes were aphakic, and 22 eyes were pseudophakic (5 anterior chamber and 17 posterior chamber lens implants). These eyes had a mean of2.1 ± 0.2 (range, 1-4) prior surgical glaucoma procedures, including the adjunctive use of subconjunctival 5-FU injections. Seven eyes had a failed Molteno implant, and three eyes had a cyclodestructive procedure. Seven eyes had previous penetrating keratoplasties with clear grafts, and three eyes had prior repair of a retinal detachment. Mean preoperative lOP was 36.4 ± 1.6 mmHg (range, 25-78 mmHg). Patients were using a mean of 2.8 ± 0.1 (range, 1-4) antiglaucoma medications. Sixty percent of patients were treated with a systemic carbonic anhydrase inhibitor. All patients were followed for a minimum of 1 year after surgery.
Results All surgeons had prior surgical experience with posterior tube shunt implants. Surgery was performed in the superior quadrant in 40 eyes (24 eyes temporally and 16 eyes nasally) and in the inferior quadrant in 10 eyes (6 eyes temporally and 4 eyes nasally). Placement of the large disk explant was achieved in all patients. However, in one eye the sclera and retina were perforated during passage of the fixation suture. In this eye, cryotherapy was applied before fixation of the disk explant. An equal number of eyes had the anterior chamber entry wound full-thickness or under a lamellar scleral flap. Technical difficulty was not encountered with placement of the anterior chamber tube.
654
Early Postoperative Intraocular Pressure Control The mean lOP on the first postoperative day was 8.3 ± 1.3 mmHg (Table 2). On the first day, lOP was elevated (> 30 mmHg) in four eyes due to anterior chamber tube complications: in one eye, the tube was intracorneal, one eye had a blood clot within the tube, and in two eyes the internal opening of the tube was occluded by iris. Twelve eyes had an lOP less than 5 mmHg. Anterior chamber depth was graded on a scale of 0 to 4+ based on iris/ corneal contact (Table 2). Mean anterior chamber depth on the first postoperative day was 3.4 ± 0.1. Two eyes had a flat anterior chamber on the first postoperative day: one eye had aqueous humor leakage around the tube with a limbal bleb and one phakic eye had ciliary block (malignant) glaucoma. The latter eye underwent vitreous aspiration and reformation of the anterior chamber. The lOP in this eye was less than 15 mmHg during subsequent examinations. A nonexpulsive suprachoroidal hemorrhage occurred in one eye on the first postoperative day. Although this resolved, the eye lost all vision, and the surgery was considered a failure. One week after surgery, the mean lOP was 9.5 ± 1.4 mmHg. Intraocular pressure remained elevated (>27 mmHg) in the four eyes with anterior chamber tube complications noted on the first postoperative day. The eye with the complicated intracorneal tube underwent surgical revision. Thirteen (26%) eyes had an lOP less than 5 mmHg. The mean anterior chamber depth was 3.5 ± 0.1. Serous choroidal detachment (1-4 quadrants) was present in 17 (34%) eyes. The mean lOP 2 weeks after surgery was 8.8 ± 1.0 mmHg. The anterior chamber tube was open in the eye that earlier had a blood clot within the tube, and its lOP was 15 mmHg. In two eyes, lOP was elevated (>30 mmHg). One eye with an open anterior chamber tube continued to have elevated lOP and was a long-term failure. The implant was removed from this eye. In the second eye, which previously had a low lOP and a well positioned tube, lOP was elevated due to tube migration out of the anterior chamber. The tube was kinked and therefore did not follow a straight path to the episcleral explant. At the
Krupin Eye Valve Study Group . Krupin Eye Valve with Disk time of surgical revision, the tube was surrounded by fibrous tissue posterior to the limbus. This tissue was removed, the tube was reinserted into the chamber, and the external portion of the tube was sutured to episcleral tissue. In this eye, lOP remained less than 15 mmHg. The mean anterior chamber depth at 2 weeks postoperatively was 3.6 ± 0.1. Serous choroidal detachment was present in 16 (32%) eyes. One eye underwent drainage of choroidal fluid. One month after surgery, the mean lOP was 14.1 ± 1.3 mmHg. All eyes had an lOP of 5 mmHg or greater. The lOP was greater than 20 mmHg in eight eyes. One eye underwent surgical revision with excision of dense fibrous tissue over the surface of the episcleral explant. Long-term lOP control in this eye was less than 13 mmHg. Five eyes required anti-glaucoma medical therapy to lower lOP to less than 20 mmHg. Medical therapy was required for long-term lOP lowering in two of these eyes. Two eyes had persistent elevation of lOP despite medical therapy and were considered failures. The mean anterior chamber depth 1 month after surgery was 3.7 ± 0.1. Six eyes still had residual serous choroidal fluid.
Late Intraocular Pressure Control The implant was removed from three eyes after the third postoperative month (Table 3). Erosion of the tissue over the explant occurred in one eye that had had a prior scleral buckling procedure. A second eye had erosion and melting of the conjunctiva at the junction of the tube with the oval disk, resulting in aqueous humor leakage. Attempts at repair were unsuccessful. A limbus-based conjunctival flap had been used in these two eyes with explant erosion. In a third eye, traumatic endophthalmitis developed 8 months after surgery. The implant was removed at the time of vitrectomy in this eye. Excluding the four eyes in which the explant was removed and the eye with a nonexpulsive hemorrhage on the first postoperative day, the mean lOP (45 eyes) at last follow-up examination (mean, 25.4 ± 2.4 months; range, 16-36 months) was 13.1 ± 1.3 mmHg. The glaucoma
valve disk implant was successful in lowering lOP to 19 mmHg or less in 40 (80%) of the original 50 eyes. In 59% of these eyes, adjunctive antiglaucoma therapy was not required. Eleven (73%) of 15 eyes with neovascular glaucoma and 29 (83%) of 35 eyes with other types of glaucoma had an lOP of 19 mmHg or lower (Fisher's exact test; P > 0.5).
Visual Acuity Preoperative visual acuity was 20/100 or better in 20 eyes, 20/200 to 20/400 in 10 eyes, and counting fingers to hand motions in 20 eyes. Visual acuity at last follow-up was 20/100 or better in 18 eyes, 20/200 to 20/400 in 14 eyes, counting fingers to hand motions in 16 eyes, and no light perception in two eyes (1 eye with the suprachoroidal hemorrhage and 1 eye with endophthalmitis). Long-term visual acuity was unchanged (2 or more Snellen acuity lines) in 19 eyes, decreased in 15 eyes, and improved in 16 eyes. Three eyes had decreased vision due to corneal decompensation: two eyes with clear corneal grafts preoperatively and one eye with iridocorneal endothelial syndrome. One eye with corneal graft failure and the eye with iridocorneal endothelial syndrome underwent repeat corneal transplantation with subsequent improved visual acuity. The lOP in these two eyes was less than 15 mmHg before and after corneal transplantation.
Diplopia Postoperative diplopia occurred in one eye. This was the eye that had an inadvertent scleral perforation (and cryotherapy) at the time of inferonasal implant surgery. In this eye, the preoperative visual acuity was 20/100 and the lOP was 30 mmHg. Hypertropia with restriction of motility inferiorly and nasally was noted 3 months after surgery. One year after surgery, the patient still had diplopia in primary gaze with a visual acuity of 20/60. Intraocular pressure was 16 mmHg.
Table 3. Glaucoma Valve Disk Implant: Long-term Results
No. of eyes· lOP (mmHg) Mean ± SEM ,,;19 mmHg (%) No. of medications Mean ± SEM % without therapy
Month 6
Year 1
Last Follow-up (mos) (25.4 ± 2.4)
46
45
45
14.8 ± 0.8 84
12.9 ± 0.8 80
13.1 ± 1.3 80
0.6 ± 0.1
0.7 ± 0.2
0.8 ± 0.3
69
62
59
lOP = intraocular pressure; SEM = standard error of the mean . • Six-month period excludes one eye with suprachoroidal hemorrhage on the first postoperative day and three eyes with removal of the implant within 3 months after surgery; I-year period also excludes one eye with implant removal 8 months after surgery (see text).
655
Ophthalmology
Volume 101, Number 4, April 1994
Discussion The design of the glaucoma disk implant (Fig 1) is based on clinical observations of currently used posterior tube shunt implants. The open anterior chamber tube, a feature common to all these devices, functions as the sclerostomy. The tube has an advantage over a surgical sclerostomy, in that it can be inserted within the anterior chamber beyond areas of peripheral anterior synechiae. In addition, the tube prevents sclerostomy closure and functions as a conduit for aqueous humor flow from the anterior chamber to the attached episcleral explant. The episcleral explant stimulates a fibrovascular reaction that requires several days for complete encapsulation of the plate. 12, 15 The capsule develops as a separate layer from Tenon's tissue. The internal surface of the encapsulation is an open, collagenous meshwork that is not lined by a double layer of cuboidal epithelium and therefore, is not a true cyst. Studies have shown that latex microspheres (as large as 0.2 JLm) and horseradish peroxidase (molecular weight, 50,000) pass freely through the capsular wall and reach the capillaries in the vascular layer. 11 ,12 Posterior tube shunt procedures reduce lOP via a passive, pressure-dependent flow of fluid across the capsular wall. The postoperative lOP relates partially to the surface area of the encapsulated bleb. The Molteno double-plate implant produces lower lOP than the single-plate implant. 16 Design of the disk implant incorporates an enlarged oval plate, in contrast to a Silastic encircling band used in the prior version of the glaucoma valve implant, to maximize surrounding encapsulation. The oval disk has a 1.75-mm high sidewall to expand the volume of encapsulation. In addition, the size (13 X 18 mm) of the oval plate minimizes the extent of surgical dissection compared with a two-plate Molteno implant, by placement of the disk implant in one ocular quadrant between two rectus muscles. Although the Silastic disc can be trimmed with scissors if it is too large, this was not necessary in the current series. Insertion of an open tube into the anterior chamber before completion of fibrous encapsulation around the episcleral explant is associated with excessive flow of aqueous humor, resulting in hypotony, flat anterior chamber, and choroidal detachment. 15 To avoid these complications, Molteno recommends a two-stage implantation to allow encapsulation of the episcleral explant before inserting the tube into the anterior chamber. During the first stage, the explant is sutured in place without inserting the tube into the anterior chamber. Because this operation does not reduce lOP, a trabeculectomy frequently is performed for interim lOP control at a different site from where the tube will be inserted later into the anterior chamber. Postoperative antiglaucoma therapy may be required during this interim period. The second operative stage in which the tube is inserted into the anterior chamber is performed 2 to 6 weeks later. Flow of aqueous humor through the tube creates a posterior filtration bleb by dissecting the fibrous encapsulation off the surface of the episcleral plate. During this interval, lOP
656
frequently is elevated to preoperative (or even higher) levels and marked inflammation is observed in the region of the posterior bleb. Antiglaucoma medications and topical corticosteroids are administered during this transient hypertensive phase which may last several weeks. 16,17 Gradually, lOP declines and the bleb inflammation resolves, permitting withdrawal of antiglaucoma medications. The magnitude of the hypertensive phase is reduced with the double-plate Molteno system. 16 One-stage implantation of an unrestricted posterior tube shunt implant can be performed by temporarily occluding the flow of aqueous humor through the anterior chamber tube until encapsulation around the episcleral plate is completed. 18-21 One technique uses a suture ligature around the tube posterior to the scleral flap or donor graft. This suture is released (or an absorbable suture is allowed to degrade) 2 to 4 weeks after surgery, using either the argon laser or through an overlying conjunctival incision. The ligature also can be applied with a releasable knot nylon suture that is exposed externally. Alternatively, a polypropylene suture is tied to the wall of the anterior chamber portion of the tube and then around the tube as a ligature before insertion of the tube into the anterior chamber. The argon laser is used to cut the suture, opening the tube but stillieaving the suture attached to its wall. Finally, the internal lumen of the tube can be occluded with a 3-0 or 4-0 suture which is either buried subconjunctivally or exposed externally by passing the suture through the conjunctiva. The suture is removed at a later date by pulling the suture via a conjunctival incision or pulling the externalized suture. The design of the Krupin Eye Valve with Disk includes a pressure-sensitive and unidirectional slit valve that provides resistance to aqueous humor flow, which is especially important before there is completion of the fibrous encapsulation around the episcleral explant. This permits implantation of the episcleral disk and insertion of the anterior chamber tube during the same procedure, without the need for restrictive tube sutures. However, functioning of the slit valve requires sufficient aqueous humor production to achieve an lOP similar to the opening pressure of the device: at a lower lOP «8-10 mmHg), there will be no fluid flow across the slit valve. Although hypotony (lOP <5 mmHg) was present in 12 eyes on the first postoperative day, this was associated with anterior chamber shallowing in only two eyes (1 eye with ciliary block glaucoma). Early postoperative hypotony may be attributed to ciliary body hyposecretion or transient aqueous humor leakage around the tube. In addition, increased aqueous humor flow, which will occur across the one-way slit valve during transient intervals of external pressure on the globe (e.g., eyelid squeezing, blinking), may account for the early postoperative hypotony. Potential advantages of a one-stage implantation of the disk implant include elimination of a second operation, more rapid control ofIOP by eliminating the period of uncontrolled glaucoma during the interval of the two-stage technique, and reduction of the magnitude of the hypertensive
Krupin Eye Valve Study Group . Krupin Eye Valve with Disk phase which can be observed after completion of the two-stage technique. However, one-stage implantation of the disk implant, or other posterior tube implant devices using techniques described above, still is associated with an increased occurrence of hypotony and choroidal effusions compared with a planned two-stage implantation technique. These complications also are more frequent with implantation of the double-plate Molteno device. 16 A transient hypertensive phase was observed 6 to 8 weeks after implantation of the disk implant in three (6%) eyes. These eyes responded to temporary medical treatment with aqueous humor suppressant therapy (topical beta-adrenergic antagonist or systemic carbonic anhydrase inhibitor). In these eyes, long-term lOP was less than 20 mmHg without antiglaucoma therapy. This phenomenon may be similar to the bleb remodeling, "high bleb phase," observed after trabeculectomy.22 It is difficult to compare accurately the success rates among the current long tube implants. Success with respect to control ofIOP with these devices depends on the type of glaucoma. The highest success occurs in eyes with open- or closed-angle glaucoma, either phakic or aphakic/pseudophakic, and the lowest success occurs in eyes with neovascular glaucoma. 23 In the current series, lOP results are similar in eyes with or without neovascular glaucoma. Although the lOP results with the Krupin Eye Valve with Disk and other posterior tube shunt implants are encouraging, these devices should be reserved for eyes with extremely poor surgical prognosis or prior failure of filtration procedures with adjunctive antimetabolites. However, primary implantation of a shunt device may be advised in eyes with neovascular glaucoma or closedangle glaucoma with anteriorly located synechiae which prevent a limbal sclerostomy from communicating with the anterior chamber. Posterior tube shunt implant surgery frequently is associated with a higher rate of operative and postoperative complications than trabeculectomy. In the current series, complications unique to seton surgery include blockage of the anterior chamber tube, migration of the tube due to movement of the episcleral plate, erosion of the tube and episcleral plate, and extraocular muscle imbalance. Postoperative diplopia is more frequent when the episcleral plate is placed beneath the rectus muscles. 24 The design features of the Krupin Eye Valve with Disk maximize the size of the scleral explant to optimize the volume of surrounding encapsulation while still permitting implantation within a single ocular quadrant between the rectus muscles. The device provides resistance to aqueous humor flow, thereby allowing primary (i-stage) implantation without the need for restrictive ligatures to achieve safe, early postoperative lowering of intraocular pressure. The large surface area of encapsulation, which develops with the presence of aqueous humor flow onto the plate, results in a large external reservoir for the passage of aqueous humor and a high filtration success in eyes with recalcitrant glaucoma.
Appendix Members of the Krupin Eye Valve Filtering Surgery Study Group: Northwestern University, Chicago, IL-Theodore Krupin, MD, Lisa F. Rosenberg, MD, Jon M. Ruderman, MD; Geisinger Medical Center, Danville, PA-Marianne E. Feitl, MD; Washington University, St. Louis, MOMichael A. Kass, MD, Allan E. Kolker, MD, Martin B. Wax, MD; University ofNebraska, Omaha, NE-Carl B. Camras, MD; New York Eye & Ear Infirmary, New York, NY-Jeffrey M. Liebmann, MD, Robert Ritch, MD; Mt. Sinai School of Medicine, New York, NY-Steven M. Podos, MD, Janet B. Serle, MD, Robert A. Schumer, MD; University of Wisconsin, Madison, WI-Paul L. Kaufman, MD, Todd W. Perkins, MD.
References 1. Watkins PH Jr, Brubaker RF. Comparison of partial-thick-
2. 3. 4. 5. 6. 7.
8.
9. 10. 11.
12. 13. 14.
ness and full-thickness filtration procedures in open-angle glaucoma. Am J Ophthalmol 1978;86:756-61. Spaeth GL, Poryzees E. A comparison between peripheral iridectomy with thermal sclerostomy and trabeculectomy: a controlled study. Br J Ophthalmol 1981;65:783-9. Lewis RA, Phelps CD. Trabeculectomy v thermosclerostomy. A five-year follow-up. Arch Ophthalmol 1984;102: 533-6. Lamping KA, Bellows AR, Hutchinson BT, Afran SI. Longterm evaluation of initial filtration surgery. Ophthalmology 1986;93:91-10 1. The fluorouracil Filtering Surgery Study Group. fluorouracil filtering surgery study one-year follow-up. Am J Ophthalmol 1989; 108:625-35. Weinreb RN. Adjusting the dose of 5-fluorouracil after filtration surgery to minimize side effects. Ophthalmology 1987;94:564-70. Chen C-W. Enhanced intraocular pressure controlling effectiveness. of trabeculectomy by local application of mitomycin C. Trans Asia-Pacific Acad Ophthalmol 1983;9: 172-7. Skuta GL, Beeson CC, Higginbotham EJ, et al. Intraoperative mitomycin versus postoperative 5-fluorouracil in highrisk glaucoma filtering surgery. Ophthalmology 1992;99: 438-44. Molteno ACB. New implant for drainage in glaucoma. Clinical trial. Br J Ophthalmol 1969;53:606-15. Molteno ACB, Straughan JL, Ancker E. Long tube implants in the management of glaucoma. S Afr Med J 1976;50: 10626. Schocket SS. Investigations of the reasons for success and failure in the anterior shunt-to-the-encircling-band procedure in the treatment of refractory glaucoma. Trans Am Ophthalmol Soc 1986;84:743-98. Minckler DS, Shammas A, Wilcox M, Ogden TE. Experimental studies of aqueous filtration using the Molteno implant. Trans Am Ophthalmol Soc 1987;85:368-92. Krupin T, Kaufman P, Mandell A, et ai. Filtering valve implant surgery for eyes with neovascular glaucoma. Am J Ophthalmol 1980;89:338-43. Krupin T, Kaufman P, Mandell AI, et ai. Long-term results of valve implants in filtering surgery for eyes with neovascular glaucoma. Am J Ophthalmol 1983;95:775-82.
657
Ophthalmology
Volume 101, Number 4, April 1994
15. Molteno ACB. Use of Molteno implants to treat secondary glaucoma. In: Cairns JE, ed. Glaucoma. Vol. 1. London: Grone & Stratton, 1986;211-38. 16. Heuer DK, Lloyd MA, Abrams DA, et al. Which is better? One or two? A randomized clinical trial of single-plate versus double-plate Molteno implantation for glaucomas in aphakia and pseudophakia. Ophthalmology 1992;99: 151219. 17. Minckler DS, Heuer DK, Hasty B, et al. Clinical experience with the single-plate Molteno implant in complicated glaucomas. Ophthalmology 1988;95:1181-8. 18. Egbert PR, Lieberman MF. Internal suture occlusion of the Molteno glaucoma implant for the prevention of postoperative hypotony. Ophthalmic Surg 1989;20:53-6. 19. Price FW Jr, Whitson WE. Polypropylene ligatures as a
658
20. 21. 22. 23. 24.
means of controlling intraocular pressure with Molteno implants. Ophthalmic Surg 1989;20:781-3. Cohen JS, Osher RH. Releasable scleral flap suture. Ophthalmol Clin North Am 1988;1:187-97. Ball SF, Loftfie1d K, Scharfenberg J. Molteno rip-cord suture hypopyon. Ophthalmic Surg 1990;21 :407 -12. Scott DR, Quigley HA. Medical management of a high bleb phase after trabeculectomies. Ophthalmology 1988;95: 1169-73. Mermoud A, Salmon JF, Alexander P, et al. Molteno tube implantation for neovascular glaucoma. Ophthalmology 1993; 100:897-902. Smith SL, Starita RJ, Fellman RL, Lynn JR. Early clinical experience with the Baerveldt 350-mm2 glaucoma implant and associated extraocular muscle imbalance. Ophthalmology 1993;100:914-18.