CASE PRESENTATION
A 56-year-old male presented in early 2007 with a history in his left eye of laser trabeculoplasty in 1994, trabeculectomy with 5-fluorouracil in 1997, cataract extraction with the placement of a posterior chamber IOL in 1998, and a second trabeculectomy with mitomycin C (MMC) in 2003. His BCVA was 20/25, and the IOP was 9 mm Hg while he was taking prednisolone acetate 1 gtt daily.
In mid-2007, the patient returned with mildly decreased vision (20/40) and a complaint of a film over his left eye. His IOP was unchanged. The examination was significant for a diffuse superior bleb and trace subepithelial corneal edema. He was diagnosed with pseudophakic bullous keratopathy and instructed to administer saline ointment every night. When he returned 2 weeks later, his visual acuity had worsened to 20/100, the IOP was 3 mm Hg, and the anterior segment showed an unchanged diffuse bleb. There was now stromal edema with Descemet's folds. The frequency of prednisolone acetate was increased to q.i.d., and the patient was referred to a corneal specialist. Two weeks later, the IOP in his left eye was 7 mm Hg by dynamic contour tonometry. Central corneal thickness measured 520 µm OD and 918 µm OS.
The patient underwent Descemet's stripping automated endothelial keratoplasty (DSAEK) on his left eye in January 2008. The button dislocated superiorly and required recentration the next day. The postoperative course was otherwise uneventful and characterized by rapid visual recovery to 20/40 within a couple of weeks. The patient's most recent examination, 9 months postoperatively, showed a BCVA of 20/40 (pinhole to 20/30-). The IOP by applanation tonometry was 4 mm Hg OS. Examination showed an elevated, avascular, diffuse superior bleb, a clear cornea with a clear and centered corneal endothelial graft (Figure 1), and a well-centered posterior chamber IOL. The visual field (Figure 2) had an unchanged superior arcuate defect, split fixation, and an inferior nasal step. There was advanced glaucomatous optic neuropathy (Figure 3). The macula and peripheral fundus were unremarkable. The patient was administering only prednisolone acetate b.i.d. in his left eye.
Comments on Corneal Edema
RNW: Until recently, bullous keratopathy that was unresponsive to medical therapy would have been treated with penetrating keratoplasty (PKP). Not infrequently, the IOP would then increase, and the trabeculectomy might even fail. A new glaucoma drainage procedure might then be required, which would jeopardize the viability of the transplant and initiate a cycle of repeat surgeries. DSAEK offers a new alternative for a patient with corneal edema who has had a trabeculectomy. Let's first review the problems associated with corneal edema in such patients. How accurate is applanation tonometry in the presence of corneal edema?
NAL: Although applanation tonometry normally overestimates the IOP of an eye with a thick, healthy cornea, the softness of an eye with a thick, edematous cornea can cause an underestimation.1 Dynamic contour tonometry is reportedly more accurate in eyes with corneal edema.2 The IOP of this patient measured only 3 mm Hg by applanation tonometry versus 7 mm Hg by dynamic contour tonometry.
DJS: There was no obvious preexisting corneal condition in this patient such as guttata, and the other eye was fine. We do not have a presurgical cell count. Once stromal edema is established, however, it is almost impossible to eliminate effectively. As in this case, the prebulla, subepithelial edema can be treated with concentrated saline but not the stromal edema. Identifying Descemet's folds can indicate whether conservative treatment will be useful or not. The folds occur when the stroma swells and pushes backward due to the presence of a strong anterior Bowman's membrane. The unchanged surface area but reduced radius of Descemet's membrane causes buckling.
Edema that directly follows cataract surgery should be treated conservatively and monitored for several months before a surgical solution is considered. Topical steroids will decrease intraocular inflammation and help to improve tight junctions. The inflammation reduces the activity of the endothelial carbonic anhydrase and Na/K-ATPase that preserve corneal clarity. My colleagues and I also lower the IOP to reduce the transcorneal fluid pressure. Carbonic anhydrase inhibitors, either systemic or topical, should not be used to weaken this critical pump further. Prostaglandin analogs may incite more inflammation.
In most patients, corneal clarity will be lost if the endothelial cell count drops from between 2,400 and 3,200 cells per square millimeter to fewer than 800. Some eyes, however, can tolerate as few as 400 cells per square millimeter. In this case, cataract surgery will have cost the patient 10% of his endothelial cells in addition to the normal attrition with aging. The two trabeculectomies might have contributed to that decrease, although there is less intraocular manipulation and fluid movement compared with cataract surgery. The now standard use of antimetabolites, 5-fluorouracil or MMC, may also be toxic to the endothelium. Perhaps even the reduced flow of nutrients in the aqueous humor into the cornea is a factor, as there is shunting toward the bleb and reduced IOP, resulting in a decreased diffusion pressure into the cornea.
NAL: In some studies, trabeculectomy was shown to cause a 10% loss of endothelial cells, which is in the range of cataract surgery.3 It has been suggested that this decrease can be countered in a susceptible patient by the intracameral injection of a viscoelastic prior to penetration into the anterior chamber.4
Comments on DSAEK
RNW: How is DSAEK performed?
DJS: Video.google.com is a good source for DSAEK videos. My colleagues and I place a 9.0- to 9.5-mm circular mark on the epithelium. After filling the anterior chamber with viscoelastic or using a chamber maintainer, we strip Descemet's membrane. The folded donor button is inserted with a forceps through a 5-mm incision. The button is folded endothelium toward endothelium like a taco with viscoelastic on the inside. After the insertion, the anterior chamber is deepened with balanced salt solution to allow the button to unfold. Injecting an air bubble opens the button further so that it floats against the recipient stromal bed. Placing three to five perpendicular stab incisions within the circular mark allows fluid to escape from the interface. We do not use fixation sutures to prevent the button's dislocation, because doing so can increase the loss of endothelial cells. The patient should remain supine for the first postoperative day and also at least 8 hours per day subsequently until the bubble has disappeared.
NAL: Does the fact that this patient had a decentered donor button have anything to do with the presence of a bleb? Could the air have escaped into the superiorly located bleb when he arose?
DJS: That is possible. We were able to reposition the button noninvasively with a roller designed to massage the button into the right location.
RNW: What are the pros and cons of DSAEK compared with PKP in a patient with previous trabeculectomy?
DJS: PKP requires approximately 60 minutes versus 30 minutes for DSAEK. In PKP, there is a lot of suturing that causes bending, pressure on the lips of the wound, and inflammation. In DSAEK, visual rehabilitation is fast, and the postoperative care is less involved. Visual acuity improves by 2 weeks after surgery and peaks at 3 months. As a result, DSAEK can be performed on eyes with 20/40 to 20/60 BCVA. PKP induces 4.50 D of irregular corneal astigmatism on average. About 3 months postoperatively, removing some stitches reduces the astigmatism. Fitting contact lenses is a challenge because of the sutures, and the bleb in eyes that underwent trabeculectomy makes a fitting almost impossible. In comparison, the button in DSAEK is only 100 µm thick and does not induce astigmatism. There is a slight hyperopic shift of approximately 0.75 D, because the button is somewhat thinner in the center.
Comments on PKP, Tube Shunts, and Trabeculectomy
RNW: What is the impact of PKP on an existing bleb?
DJS: There is a problem with the amount of inflammation and trauma from a flat anterior chamber for almost 20 minutes.
NAL: A study of combined PKP and trabeculectomy with MMC showed that only 50% of eyes had well-controlled IOP and that 60% of corneal grafts were viable 2 years postoperatively.5 Currently, no studies have looked at DSAEK in comparison. Glaucoma after PKP can occur in 20% to 35% of patients6 and is often multifactorial: the use of steroids can induce ocular hypertension, and inflammation results in the formation of peripheral anterior synechiae. Synechiae can also be induced by a narrow anterior chamber angle after surgery.
DJS: Glaucoma drainage devices can also reduce the viability of a corneal graft transplant. The choice of trabeculectomy or a glaucoma drainage device therefore depends on the surgeon's preference and experience.
GV: Is the measurement of IOP valid after DSAEK? Does it change as the button becomes more firmly adherent?
DJS: We measured IOP with various tonometers in 10 of our patients who had undergone DSAEK. Our findings concur with others' that, despite the increase in total thickness by 100 µm, there was little change in IOP.7 The donor button does not contribute to increased rigidity; it merely adheres to the stromal bed.
RNW: We do not know the influence of DSAEK on the visual field. The changes in contrast sensitivity and visual acuity probably require the establishment of a new baseline similarly to cataract surgery with advanced lenticular changes.
CONCLUSION
A patient with a history of two trabeculectomies and cataract surgery in his left eye developed bullous keratopathy that was treated with DSAEK. After surgery, his visual acuity was good, and bleb function continued to be excellent. Unlike PKP, DSAEK provides rapid visual rehabilitation, and it is a less traumatic and inflammatory procedure that may pose a lower risk of bleb failure. IOP measurements are not influenced by the overall increase in corneal thickness after DSAEK. It is advisable to obtain new baseline visual fields postoperatively.
Section editor Robert N. Weinreb, MD, is Distinguished Professor of Ophthalmology and Director of the Hamilton Glaucoma Center, University of California, San Diego. Dr. Weinreb may be reached at weinreb@eyecenter.ucsd.edu.
Nils A. Loewen, MD, PhD, is a Senior Clinical Fellow, Hamilton Glaucoma Center, Department of Ophthalmology, University of California, San Diego. Dr. Loewen may be reached at nloewen@glaucoma.ucsd.edu.
David J. Schanzlin, MD, is Professor of Ophthalmology and Director of Keratorefractive Surgery at the Shiley Eye Center, University of California, San Diego. Gianmarco Vizzeri, MD, is an International Clinical Fellow, Hamilton Glaucoma Center, Department of Ophthalmology, University of California, San Diego.
