An 82-year-old black male with advanced open-angle glaucoma presented for evaluation of a trabeculectomy. He was pseudophakic bilaterally, and his past ocular history was significant for pseudophakic bullous keratopathy in his left eye. The patient was functionally monocular with a UCVA of 20/60 OD and light perception OS. His medications included TIMOPTIC (Merck & Co., Inc., West Point, PA), Xalatan (Pfizer Inc., New York, NY), Alphagan (Allergan, Inc., Irvine, CA), and pilocarpine. His IOPs were in the midteens, higher than the target pressure of 12 mm Hg, and he had a 0.95 C/D OD. The patient's visual field loss was threatening fixation in his right eye. He completed informed consent for a trabeculectomy in June 2003.
Intraoperatively, the scleral tunnel created during the patient's previous cataract surgery made performing the trabeculectomy challenging. Both his cornea and sclera had thinned superiorly. We placed mitomycin C 0.3 mg/mL on the scleral bed for 3 minutes, made a triangular flap, and closed the scleral flap with three 10-0 nylon sutures. The patient began taking atropine b.i.d., Ciloxan (Alcon Laboratories, Inc., Fort Worth, TX) q.i.d., and Pred Forte (Allergan, Inc.) q.h.
On the first postoperative day, the patient's UCVA was 1/200, his IOP was 1 mm Hg, and he had both a shallow anterior chamber and non-appositional choroidal effusions (Figure 1). Additionally, although we had noted no Descemet's membrane detachment intraoperatively, we now observed a large one. We thought it probable that intrastromal corneal penetration with the superblade had occurred upon creating the sclerostomy
Figure 1. On postoperative day 1, diffuse illumination reveals diffuse corneal edema and a large Descemet's membrane detachment (A). Optic section demonstrates the extent of the Descemet's membrane detachment (B).
HOW WOULD YOU PROCEED?
1. Would you observe this patient?
2. Inject sulfur hexafluoride (SF6) gas in order to tamponade the Descemet's membrane detachment? If so, what effect would a trabeculectomy have on gas tamponade?
3. Would you inject a viscoelastic substance?
SURGICAL COURSE
Due to the patient's monocular status, we considered using SF6 gas. Because this was not a closed system, however, gas would have been able to escape through the sclerostomy.
In the minor room, we used two 27-gauge needles to simultaneously enter the eye at the 3- and 9-o'clock positions away from the sclerostomy located at the 1-o'clock position. We injected 1.5 cc of 20% SF6 gas into the anterior chamber by means of the 27-gauge needle with 60% to 70% fill. Aqueous egressed through the second needle. The patient was admitted to the hospital, was put on strict bed rest, and received Pred Forte q1h. After 1.5 hours, we examined the patient on the ward and observed 30% fill with the gas migrating through the sclerostomy into the bleb.
OUTCOME
On the second postoperative day, the patient's visual acuity was hand motions at 4 feet, and his IOP measured 6 mm Hg with a tonopen. His pachymetry reading was 1,012 µm (Figure 2). Owing to the 30% fill, much of the inferior Descemet's membrane remained detached. As a result, we instructed the patient to assume Trendelenburg's position.
Figure 2. After injection, there is a 30% fill, and the SF6 gas has tracked into the bleb temporally (A). The gas bubble tamponades the Descemet's membrane detachment (B).
From the fourth through the sixth postoperative days, the patient's IOP ranged between 4 and 5 mm Hg, and his anterior chamber was shallowed from three corneal thicknesses to one. There was overfiltration through his thin scleral flap. We attempted a trial of pressure patching and continued cycloplegia. On the next day, the patient's anterior chamber shallowed to a depth of one corneal thickness with no lens-corneal touch, still Seidel negative. We injected viscoelastic through the paracentesis to deepen the anterior chamber. We had chosen not to inject a viscoelastic agent earlier in the case, because it would not have resulted in a long-lasting tamponade.
At the final follow-up visit, the patient's UCVA was 20/200, his IOP was 11 mm Hg, his cornea was clear, and Descemet's membrane was attached. The bleb was elevated and filtering (Figure 3).
Figure 3. Diffuse illumination reveals a clear cornea with resolution of edema and the Descemet's membrane detachment (A). Optic section shows the reapposition of Descemet's membrane (B).
DISCUSSION
A Descemet's membrane detachment after a trabeculectomy is a rare complication, one to which this patient's prior surgical history predisposed him. When such detachments occur after cataract surgery, they often become reattached on their own and may improve with observation alone. This patient presented with a large Descemet's membrane detachment, and he was at risk of developing permanent corneal decompensation that would potentially necessitate corneal transplantation. Tamponade of Descemet's membrane detachments with intracameral SF6 is gaining greater acceptance among ophthalmologists.1
We based our decision to proceed with gas tamponade on the patient's monocular status. Because the affected eye possessed substantially better BCVA than his other eye, we pursued an aggressive treatment option in order to give the patient his best chance at visual recovery. The situation was unique in that this was no longer a closed system. Within 1.5 hours, half of the gas had escaped into the bleb.
We admitted the patient because he was functionally monocular and we wished to limit him to strict bed rest in order to allow the depleting gas to tamponade the cornea. Additionally, we were able to manipulate the gas for further inferior tamponade with Trendelenburg positioning.
The authors wish to thank Joseph Halabis, OD.
Nouman Siddiqui, MD, practices at the Duke University Eye Center in Durham, North Carolina. He holds no financial interest in the products and companies mentioned herein. Dr. Siddiqui may be reached at (919) 684-6611; siddi002@mc.duke.edu.
Leon W. Herndon, MD, is Associate Professor of Ophthalmology at the Duke University Eye Center in Durham, North Carolina. He holds no financial interest in the products and companies mentioned herein. Dr. Herndon may be reached at (919) 684-6622; hernd012@mc.duke.edu.
1. Kim T, Hasan SA. A new technique for repairing descemet membrane detachments using intracameral gas injection. Arch Ophthalmol. 2002;120:181-183.
