Case 2 in Table 2 and Table 3. A, Silicone oil pupillary-block glaucoma (SOPBG) with silicone oil–filled anterior chamber in a pseudophakic eye. B and C, Note progressive superior silicone oil displacement after inferior argon laser iridotomy (arrows) and iridoplasty.
Zalta AH, Boyle NS, Zalta AK. Silicone Oil Pupillary BlockAn Exception to Combined Argon–Nd:YAG Laser Iridotomy Success in Angle-closure Glaucoma. Arch Ophthalmol. 2007;125(7):883-888. doi:10.1001/archopht.125.7.883
To examine the rate of laser iridotomy failure at the University of Cincinnati Glaucoma Service, Cincinnati, Ohio, during the last 10 years and to evaluate the importance of silicone oil pupillary block glaucoma (SOPBG) as a causal factor.
We retrospectively reviewed the operative records of all 1711 eyes that underwent laser iridotomy for the treatment of pupillary block angle-closure glaucoma between January 1, 1996, and December 31, 2005. The occurrence, etiology, timing, and rate of laser iridotomy failure were assessed with SOPBG cases analyzed separately.
Analyses using the χ2 test demonstrated significantly higher laser iridotomy failure rates for 13 eyes with SOPBG compared with 1698 eyes with non-SOPBG for all 3 timing outcomes (immediate, 15.4% vs 0%; short term, 92.3% vs 2.5%; and long term, 38.5% vs 0.1%; all P < .0001). To achieve long-term patency, SOPBG iridotomy failures required, on average, 2.7 laser iridotomy procedures, 4.1 periocular steroid injections, and 0.7 intracameral tissue plasminogen activator injections.
Eyes with SOPBG require extensive resources to prevent laser iridotomy failure. In managing SOPBG, ophthalmologists should anticipate the need for additional laser treatment and use adjunctive steroids and intracameral tissue plasminogen activator to enhance long-term patency and avert invasive surgical procedures.
Laser iridotomy is a safe and effective procedure that replaced surgical iridectomy in the early 1980s as the preferred treatment for pupillary block angle-closure glaucoma (PBACG). Laser iridotomy failure may take many forms including the inability to create a patent iris hole at the initial procedure, short-term closure of a patent hole usually within days to months, and long-term failure to maintain hole patency after repeat laser treatment. Historically, initial iridotomy creation and subsequent long-term patency can be achieved in 99% to 100% of cases regardless of the type of laser used.1- 8 In contrast, short-term laser iridotomy closure and re-treatment rates vary markedly depending on the type of laser used, for example, 0% to 9% for combined argon–Nd:YAG lasers,1- 3 0% to 11% for Nd:YAG laser alone,4- 7 and 16% to 35% for argon laser alone.5- 8 Some studies have suggested that combined laser iridotomy procedures are associated with higher rates of patency and fewer complications.1- 3
Since the late 1980s, the use of intravitreal silicone oil (SO) in the repair of complex retinal detachments has created a new type of PBACG9- 18 in which laser iridotomy is less effective.16,17 While intravitreal SO produces prolonged retinal tamponade and improves success in these complicated retinal reattachments, glaucoma develops in 6% to 30% of eyes12- 14 by a variety of pathophysiologic mechanisms9- 14 and, more specifically, silicone oil pupillary block glaucoma (SOPBG) develops in 1% to 3% of eyes.13,18 In the early postoperative period, SO may escape from the vitreous cavity, block the pupil, and act as a barrier to forward aqueous movement. Continued production of aqueous pushes silicone through the pupil until it fills the anterior chamber. While aphakic eyes are especially prone to this complication,9- 11 it has also been shown to occur in 6% of pseudophakic and phakic eyes,14 presumably as a result of partial zonulysis.14- 16 A prophylactic inferior surgical iridectomy at the time of SO injection has been recommended in all aphakic and pseudophakic eyes to prevent this complication.9- 11,13- 15,17,18 Despite this precaution, inferior surgical iridectomies close in 11% to 32% of cases.11,13,14,18,19 Reddy and Aylward17 reported that the Nd:YAG laser failed to reopen 78% of closed inferior surgical iridectomies and recommended a large surgical iridectomy if removal of SO is not desired.
Since the late 1980s, iridotomies on the University of Cincinnati Glaucoma Service, Cincinnati, Ohio, have been successfully performed using argon and Nd:YAG lasers sequentially in the same operation. However, the rate of laser iridotomy failure has seemingly increased during the last 5 years with the occurrence of SOPBG. This study was designed to compare the rate of laser iridotomy failure in eyes with SOPBG with the rate of failure due to all other etiologies of PBACG. We also examined the efficacy of various methods used to enhance the success of laser iridotomy in eyes with SOPBG. To our knowledge, this is the largest series of cases to examine the failure of combined argon–Nd:YAG laser iridotomy. There are also, to our knowledge, no published reports on the use of argon laser iridotomy in anterior chambers filled with SO and the adjunctive use of tissue plasminogen activator (tPA) to enhance iridotomy patency in SOPBG.
We retrospectively reviewed the operative records of all 1711 eyes that underwent laser iridotomy for the treatment of PBACG (1698 non-SOPBG and 13 with SOPBG) in 2 private practice locations (A.H.Z.) and 1 indigent clinic between January 1, 1996, and December 31, 2005. All eyes that had laser iridotomy failure were identified and assessed for the timing of iridotomy failure and underlying etiology of PBACG, with separate analysis for eyes with SOPBG.
Three possible timing outcomes were used to evaluate laser iridotomy failure: immediate, short term, and long term. An iridotomy failure was considered immediate if a patent iris hole could not be created at the initial procedure, short term if a patent iris hole closed in 1 day or longer and required additional laser treatment, and long term if iris hole patency could not be maintained after repeat laser treatment and either required surgical iridectomy or was deemed inoperable because of poor prognosis with end-stage disease and no useful visual function. In these eyes, follow-up ended at the time of long-term iridotomy failure. The rates of laser iridotomy failure were compared for cases with and without SO.
All patients except those with SOPBG and isolated cases of acute angle-closure glaucoma underwent a standard combined laser iridotomy with an Abraham iridectomy lens performed or supervised by one of us (A.H.Z.). Argon laser was initially used to create 2 surface iris contractions in the superior peripheral iris between the 11- and 1-o’clock positions (500-μm spot, 0.5 seconds, 200-400 mW) and penetrate the deep iris stroma (pigmented irides: 50-μm spot, 0.02-0.05 seconds, 1.5-2.0 W; nonpigmented irides: 50-μm spot, 0.5 seconds, 1 W). After deep stromal iris penetration or the first plume of iris pigment epithelium release, the Nd:YAG laser (1.0 mJ, single burst, with power increased as needed) was immediately used to achieve iris perforation and remove the iris pigment epithelium (hole size, 250-400 μm). All iridotomies were reevaluated at a minimum of 3 visits during the following postoperative intervals: 1 to 2 weeks, 5 to 7 weeks, and 3½ to 4 months. If there was any evidence of pigment proliferation, membrane development, or decrease in iridotomy size at the final visit, eyes were monitored until stable long-term patency was achieved.
For all eyes with SOPBG, patients had minimally undergone a pars plana vitrectomy with intravitreal SO injection. All laser iridotomies were performed by one of us (A.H.Z) with an Abraham iridectomy lens and were created between the 5- and 7-o’clock positions through the peripheral iris. The type of laser iridotomy performed was determined by the presence or absence of an inferior surgical iridectomy and whether SO filled the anterior chamber. In all eyes with inferior surgical iridectomy closure, Nd:YAG laser treatment alone was used. In eyes without a surgical iridectomy, combined laser treatment was used. If SO did not fill the anterior chamber, a standard combined procedure was performed; if SO filled the anterior chamber, the argon laser was used to penetrate the iris and Nd:YAG laser use was delayed until aqueous displ aced the SO level above the iridotomy plane (Figure). In the event of partial or complete iridotomy closure, reopening procedures were performed primarily with the Nd:YAG laser.
For eyes with SOPBG, pertinent baseline data were collected including preoperative diagnosis, vitreoretinal procedures performed, type of surgery preceding the SOPBG, and lens status. We also examined the interval between vitreoretinal or inciting surgery and SOPBG, preiridotomy intraocular pressure (IOP) by Goldmann applanation tonometry, extent of peripheral anterior synechiae (PAS), and the mode of pupillary block presentation. Postiridotomy factors were assessed including time to iridotomy closure, number of repeat iridotomies performed, periocular or oral steroid use, intracameral injection of tPA, IOP by Goldmann applanation tonometry and extent of PAS within 1 week, and duration of follow-up.
Table 1 presents the number and rates of immediate, short-term, and long-term failure of laser iridotomies for cases with non-SOPBG and SOPBG between January 1, 1996, andDecember 31, 2005. In 1698 eyes with non-SO PBACG, laser iridotomies failed immediately in 0% of cases, in the short term in 2.5% of cases, and in the long term in 0.1% of cases. In 13 eyes with SOPBG, laser iridotomy failed immediately in 15.4% of cases, in the short term in 92.3% of cases, and in the long term in 38.5% of cases. Rates of SOPBG laser iridotomy failure were significantly higher for immediate, short-term, and long-term outcomes (χ2 test, all P<.0001).
Of the 42 non-SO short-term iridotomy failures, 23 (54.8%) occurred in eyes with primary angle-closure glaucoma, 14 (33.3%) in eyes with chronic uveitis and iris bombé, and 5 (11.9%) in eyes with cicatricial proliferative diabetic retinopathy. The mean (±SD) interval to short-term iridotomy failure was 64 (±55) days, and only 0.6% (10 of 1698) of cases closed after the 90-day postoperative period. Thirty-four eyes (81%) remained patent long term after a single additional laser procedure and 8 eyes (19%) underwent multiple repeat laser procedures (range, 2-6 procedures). Ten eyes received 1 form or more of steroids (sub-Tenon's capsule triamcinolone acetonide, 8 eyes; subconjunctival dexamethasone sodium phosphate, 7 eyes; or oral prednisone, 3 eyes) to enhance long-term iridotomy patency. Long-term iridotomy failure occurred in 2 eyes with cicatricial proliferative diabetic retinopathy.
Baseline demographic and clinical characteristics for all 13 SOPBG cases are given in Table 2. Three cases occurred as the result of inferior surgical iridectomy closure and 10 cases occurred in the absence of a surgical iridectomy. The 3 modes of presentation for SOPBG were as follows: SO filled the anterior chamber in 38%, SO created a mushroom-shaped dome into the anterior chamber in 23%, and 360° of posterior synechiae secluded the pupil with SO in the posterior chamber in 38%. Only 31% had an iris bombé configuration. Silicone oil pupillary block glaucoma occurred in 3 phakic, 5 aphakic, and 5 pseudophakic eyes. While the mean (±SD) interval for the postoperative development of SOPBG was 76 (±101) days, 38% of cases occurred within 3 to 8 days after the last surgical procedure. Before iridotomy, the mean (±SD) IOP was 35 (±20) mm Hg and 75% of eyes had an IOP greater than 35 mm Hg; the mean (±SD) extent of PAS was 265° (±107°) and 46% had 360° of PAS.
Iridotomy procedures and outcomes for the 13 SOPBG cases are given in Table 3. The Nd:YAG laser was used alone in 3 eyes and combined lasers were used in 10 eyes. Adjunctive argon laser peripheral iridoplasty was performed in 23% (3 of 13) of cases. Creation of an inferior laser iridotomy at the initial procedure was successful in 85% (11 of 13) of SOPBG cases. Use of the Nd:YAG laser alone successfully reopened 2 of 3 surgical iridectomies closed by inflammatory membranes. In the third case, use of the Nd:YAG laser created more fibrin and blood at the occluded iridectomy site. This immediate failure was successfully treated with intracameral tPA the same day. For 5 eyes with SO-filled anterior chambers, the Nd:YAG laser was ineffective because it created pits or starburst fractures in the SO interface. In such cases, argon laser heat bubbles were used as a condensing lens to focus additional energy and successfully penetrate the iris. Within minutes, aqueous leaked into the anterior chamber and displaced the lighter SO superiorly above the iridotomy plane (Figure). After this delay, the Nd:YAG laser was used to enlarge the iridotomy. Within 1 week after iridotomy, the overall IOP significantly decreased to a mean (±SD) of 12 (±8) mm Hg (paired t test, 5.0; P < .001) and was less than 13 mm Hg in 85% (11 of 13) of eyes; the overall mean (SD) extent of PAS significantly decreased to 142° (112°) (paired t test, 3.9; P < .01).
Twelve of 13 eyes with SOPBG had short-term failure at a mean (±SD) of 16 (±15) days after iridotomy. Short-term failure occurred within 2 days in 31% (4 of 13) of cases and within 1 month in 85% (11 of 13) of cases. Eyes with closed iridotomies underwent additional laser procedures at a mean (±SD) of 2.9 (±2.0) times, and 2 separate iridotomy sites were created in 6 eyes (50%). Five eyes with SOPBG had long-term iridotomy failure at a mean (±SD) of 2.9 (±2.8) months after the initial procedure. In that time, these 5 eyes underwent additional laser procedures a mean (±SD) of 4.6 (±2.1) times despite receiving a total of 24 subconjunctival dexamethasone, 1 sub-Tenon's capsule triamcinolone, and 7 intracameral tPA injections. Eyes with SOPBG were followed up for a mean (±SD) of 7.7 (±6.7) months.
Twelve of 13 eyes (92%) with SOPBG received a mean (±SD) of 4.0 (±3.9) subconjunctival dexamethasone injections (2 or 4 mg), and 58% (7 of 12) of these eyes retained iridotomy patency during the long term. Five of 13 eyes (38%) received a single sub-Tenon's capsule triamcinolone acetonide injection (30 or 40 mg), and 80% (4 of 5) of these eyes retained iridotomy patency during the long term. Five of 13 eyes (38%) received 2 or 3 injections of intracameral tPA (10 μg), and 40% (2 of 5) of these eyes retained iridotomy patency during the long term. Intracameral tPA use caused no hyphemas, recurrent vitreous hemorrhages, or corneal decompensation.
To our knowledge, the current study represents the largest series of combined argon-Nd:YAG laser iridotomies for the treatment of PBACG to date. Of 1698 iridotomies performed in eyes without SO, 100% were immediately successful and 99.9% achieved long-term patency without the need for surgical iridectomy. Our overall short-term failure rate of 2.5% was similar to or better than that reported for procedures using combined argon–Nd:YAG1- 3 (range, 0%-9%), Nd:YAG laser alone4- 7 (range, 0%-11%), and argon laser alone5- 8 (range, 16%-35%). These findings confirm well-established historical observations that virtually 100% of laser iridotomies can achieve long-term patency.1- 8
Silicone oil pupillary block glaucoma, a relatively new disease, is unique in its high rate of iridotomy failure. While our short-term failure rate (92%) was consistent with previous reports (89%-100%),16,17 our long-term failure rate (38%) was much lower (78%-100%).16,17 To achieve long-term patency, SOPBG iridotomy failures required, on average, 2.7 laser iridotomy procedures, 4.1 periocular steroid injections, and 0.7 intracameral tPA injections. In contrast, 81% of non-SOPBG short-term iridotomy failures remained patent during the long term after a single additional laser procedure and only 0.5% (8 of 1698) of eyes required more than 1 repeat laser procedure. To our knowledge, there are no previous reports of the incidence and effect of SOPBG on laser iridotomy failure.
In our cases of SOPBG, inferior iridotomy effectively reduced the mean IOP by 66% and PAS by 46%. Long-term iridotomy patency was achieved in 62% of SOPBG cases and these eyes did not require surgical iridectomy,14,16,17 SO removal,12- 16,18 glaucoma fistulizing or drainage implant surgery,12,13 or cyclodestructive procedures.13,14,18 Short-term iridotomy failure occurred within 2 days in 31% (4 of 13) and within 1 month in 85% (11 of 13) of SOPBG cases. These findings suggest that timely creation of an inferior iridotomy may relieve SOPBG, reverse PAS, and avert the need for invasive surgical procedures.
In eyes with intravitreal SO, SOPBG may occur after any intraocular procedure irrespective of lens status. In our series, SOPBG developed in 38% (5 of 13) of eyes after a variety of anterior segment procedures, and 23% (3 of 13) of eyes were phakic. Disruption of the lens-zonule complex is probably an important contributory factor to anterior migration of SO and development of SOPBG.14- 16 It has been recommended that a prophylactic inferior surgical iridectomy be performed in all aphakic and pseudophakic eyes containing intravitreal SO.9- 11,13- 15,18 We believe that the same consideration applies to phakic eyes in which the anterior hyaloid face is disrupted and zonular weakness might be present.
In the current study, SO filled the posterior chamber and the pupil became secluded by posterior synechiae in 38% (5 of 13) of SOPBG cases. The mechanism for this presentation may be similar to that of inferior surgical iridectomy closure, which occurs in 11% to 32% of SO cases.11,13,14,18,19 In both situations, the SO bubble in the posterior chamber forms a smooth spherical dome that may serve as a platform for the development of a retroiridal membrane. In 23% (3 of 13) of our cases, pupillary block forced SO through the pupil, creating a mushroom-shaped dome into the anterior chamber. In these presentations, a standard combined argon–Nd:YAG laser procedure was effective because aqueous convection currents removed debris and pigment from the inferior iridotomy site. In contrast, in eyes with SO-filled anterior chambers and no surgical iridectomy, Nd:YAG laser treatment was ineffective in creating virgin iridotomies because SO trapped iris debris, blood, and fibrin over the iridotomy site. Argon laser treatment successfully overcame this problem by creating tiny heat bubbles that acted as a condensing lens to focus additional argon energy through the iris stroma. Eyes with SO-filled anterior chambers were particularly difficult to identify. At slitlamp examination, SO creates a membranous glistening sheen or specular light reflex at the iris interface similar to that seen at the retinal interface when SO completely fills the vitreous cavity. Clearly, identification of these varied presentations of SOPBG is critical for selection of proper treatment.
To our knowledge, this retrospective review is the first report of intracameral tPA use after laser iridotomy. In eyes with SOPBG, tPA was effective in clearing blood and fibrin from iridotomy sites without complication and retained long-term patency in 40% (2 of 5) of eyes. Recombinant tPA20 has proved safe and effective in dissolving intraocular blood and fibrin clots after a variety of eye surgical procedures21 and should be considered as an adjunct in the treatment of recurrent fibrin or inflammatory membranes that occlude or threaten to occlude laser iridotomies in eyes with SOPBG.
Conclusions drawn from this study are limited by its retrospective nature, small SOPBG sample, and the inherent selectivity of refractory cases referred to a subspecialty service. Despite these limitations, our findings suggest that identification of SOPBG presentation and creation of a timely inferior laser iridotomy is critical to relieve pupillary block and avert additional invasive surgical procedures. Ophthalmologists should anticipate iridotomy closure, the need for additional laser treatment, and the use of adjunctive steroids and intracameral tPA to enhance long-term patency.
Correspondence: Alan H. Zalta, MD, Department of Ophthalmology, University of Cincinnati College of Medicine, University Medical Arts Bldg, 222 Piedmont Ave, Suite 1700, Cincinnati, OH 45219 (email@example.com).
Submitted for Publication: August 29, 2006; final revision received October 23, 2006; accepted November 29, 2006.
Author Contributions: Dr Alan H. Zalta had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Financial Disclosure: None reported.