Afamiliar situation to any ophthalmologist is when the treatment of one condition leads to the development of another. A classic example is steroid-induced glaucoma after corneal transplantation, as this case illustrates.

OPHTHALMIC HISTORY
A 47-year-old white man with a history of keratoconus and high myopia underwent repeat penetrating keratoplasty (PKP) in his right eye. Incidentally, 2 weeks later, he was also treated successfully with a single injection of bevacizumab for a choroidal neovascular membrane in his right central macula. Soon after the second PKP, the patient's IOP rose into the high 20s and low 30s, for which he was treated with brimonidine tartrate 0.1% and timolol maleate 0.5%. Despite therapy, 2 months after PKP, his IOP spiked to 40 mm Hg, presumably due to steroid response. His medication was switched from prednisolone acetate 1% to loteprednol etabonate 0.5%. Therapy appeared to effectively curb the elevation in IOP until the frequency of the steroid's dosing had to be increased to six times daily in response to signs of possible early rejection.

After tapering the loteprednol, the patient's IOP was once again controlled and remained in the teens over the course of a year, with low dosing frequency of the loteprednol and continued use of the brimonidine and timolol. An episode of rejection again required more frequent administration of the loteprednol and triggered a steroid response to nearly 30 mm Hg. Afterward, the steroid was tapered to once daily, but the elevated IOP persisted, despite the addition of travoprost 0.004%. The patient was referred to the glaucoma clinic for management, 6 months after the most recent episode of rejection and 2 years after the repeat PKP.

GLAUCOMA EVALUATION
The patient's vision was 20/40 OD and 20/30 OS (corrected by a hybrid contact lens in his right eye and a rigid gas permeable contact lens in his left eye, spherical equivalent of -12.00 D OD and -10.00 D OS). His IOP measured 26 mm Hg OD and 15 mm Hg OS by applanation tonometry. His right eye had a clear corneal graft with a few buried sutures, a deep and quiet anterior chamber, a normal iris, and a clear lens. The conjunctiva was slightly injected but mobile. The posterior segment exhibited vitreous syneresis, myopic changes in the fundus (including a Fuchs' spot), and a slightly large disc with peripapillary atrophy and a shallow cup (Figure 1). The cup-to-disc ratio was in the range of 0.6. The patient's left eye demonstrated keratoconus but otherwise appeared similar to the right eye.

Gonioscopy revealed lightly pigmented angles that were open to the ciliary body band. Pachymetry measured 600 µm OD and 432 µm OS. Structural imaging of the optic discs yielded poor results due to corneal irregularity bilaterally, and visual field testing revealed superior depression in his right eye (Figure 2). Although it was difficult to accurately assess the extent of his glaucoma, my colleagues and I felt that the patient suffered only mild-to-moderate damage but that his IOP would need to be lowered to preserve the graft's survival and prevent glaucomatous progression.

DISCUSSION
Current Plan
Occurring in up to one-third of patients, secondary glaucoma after PKP is a well-known phenomenon.1 Although a number of factors could be involved (including changes in the trabecular meshwork's anatomy, inflammation, and peripheral anterior synechiae), in this case, steroid response was clearly implicated as the major cause, based on the patient's well-documented history and a wide open angle on gonioscopy. As with any case of steroid-induced glaucoma, appropriate attempts had been made to taper the causative agent to the extent possible (while avoiding graft rejection) and switch to a less potent steroid, but even milder agents can be problematic. This patient was also placed on nearly maximal glaucoma medical therapy, excluding carbonic anhydrase inhibitors. These agents may interfere with corneal endothelial function, and in eyes with prior endothelial cellular loss or dysfunction, they may induce mild corneal thickening, if not outright decompensation on rare occasions.2,3 For the time being, my colleagues and I have elected to avoid this class of medications unless surgery is the only option.

Instead, we plan a trial of laser trabeculoplasty. Despite limited reports on the use of trabeculoplasty in the setting of steroid-induced glaucoma,4 this form of treatment is an option in this case. The patient's cornea is clear, and his angle is open. Moreover, we want to avoid an invasive procedure that has a high risk of causing graft failure. Although it remains to be seen how our patient will respond to laser trabeculoplasty, it is instructive to contemplate the possible next step, should he require further intervention.

Further Options
Trabeculectomy
Despite a high rate of bleb failure in the setting of PKP, the use of mitomycin C (MMC) during trabeculectomy has improved outcomes.5 MMC does not seem to cause the epithelial toxicity observed with 5-fluorouracil, which is best avoided after PKP. In this case, the conjunctiva is relatively quiet and mobile with no signs of scarring, which makes trabeculectomy technically feasible. The patient is heavily dependent on his contact lenses, however, which is a relative contraindication for trabeculectomy, because the procedure may predispose patients to infection and the threat of endophthalmitis. Another consideration is the importance of avoiding hypotony and a shallow anterior chamber, which could damage the graft. Like any intraocular surgery, even uncomplicated trabeculectomy can lead to graft failure.

Glaucoma Drainage Devices Drainage devices are increasingly used to treat glaucoma after PKP and provide fairly good IOP control, but it is now well known that tube shunt surgery results in an even higher rate of graft failure than trabeculectomy.6,7 Possible reasons may include mechanical forces such as direct tube-cornea touch or the creation of an open channel of inflammatory mediators between the anterior chamber and the subconjunctival space. Inserting the tube into the vitreous cavity, however, is associated with a reduced risk of graft failure compared with its insertion in the anterior chamber.8 Keeping the tip of the tube away from the cornea (eg, with placement in the sulcus) is a frequent practice at our institution. Because the eye in this case is phakic with a clear lens, the only reasonable option would be to place the tube shunt deep in the anterior chamber, close to the iris. A valved implant might be an appropriate choice to minimize the risk of hypotony.

Cyclophotocoagulation
Often regarded as a last option for many forms of glaucoma, cyclophotocoagulation (CPC) should be considered in the setting of corneal grafts. Like incisional surgery, CPC can effectively lower IOP (repeat treatments possibly being necessary),9 and it is associated with a high rate of graft failure, up to 44%.10 The known potential complications such as hypotony and even phthisis, however, may not be as easily and predictably avoided with CPC as with surgery. Nevertheless, in most cases, conservative treatments coupled with the adequate use of steroids can prevent significant inflammation and should minimize the risk of graft failure. Because this patient is phakic, a transscleral approach would be required.

Our Choice
The patient's high myopia, dependence on contact lenses, and native crystalline lens status pose limitations. Otherwise, trabeculectomy with MMC might be the best option for lowering his IOP and minimizing the risk of graft failure should laser trabeculoplasty not remedy the problem. In this case, placing a glaucoma drainage device in the anterior chamber and transscleral CPC are more viable choices, although each may carry an increased risk of graft failure in an eye that has already undergone two PKP procedures. Alternatively, although no published results are available, newer glaucoma procedures such as ab interno trabeculotomy and canaloplasty should not be dismissed as possibilities in these challenging situations.

Eiyass Albeiruti, MD, is an assistant professor of ophthalmology at the University of Pittsburgh in Pennsylvania. Dr. Albeiruti may be reached at (412) 647-8199; albeirutie2@upmc.edu.