Introduction

The relationship between corneal transplantation and glaucoma is complex. Corneal transplantation may induce glaucoma and the surgical treatment of glaucoma can lead to corneal or graft failure. The successful management of both entities requires close monitoring and coordinated care.

Eduard Zirm is credited with performing the first successful human corneal transplant in 1905¹. Since then, most corneal transplants have been penetrating keratoplasties (PKPs), replacing all layers of a recipient’s cornea even if the pathology was limited to a single layer. With the advent of endothelial keratoplasty, the ‘big bubble’ technique, and femtosecond laser assisted keratoplasty, the selective replacement of layers of the cornea has become increasingly popular².

From 2005 to 2008, the Eye Bank Association of America reported an increase in the number of endothelial transplants from 1429 to 17 468³. During this same period, the number of PKPs declined from 45 821 to 32 524. While the number of corneas distributed for lamellar transplantation increased almost 20% from 2005 to 2008, only 1072 lamellar surgeries were performed in 2008³. The significance of the shift away from PKP to endothelial and lamellar keratoplasties, and its effect on post-keratoplasty glaucoma, has not been well studied; but, may result in lower rates of post-keratoplasty glaucoma.

Keratoprosthesis have also gained greater acceptance. Of these devices, the Boston Keratoprosthesis is the most popular with nearly 5000 having been implanted during the past decade⁴. Outside the United States, the osteo-odonto-keratoprosthesis has gained popularity as a surgical option for those with severe ocular surface diseases. Even in patients who have not had a history of glaucoma, implantation of these devices can induce glaucoma. Management difficulties are made worse because the implanted devices prevent accurate IOP measurements.

Irvine and Kaufman reported increased intraocular pressure (IOP) after PKP in 1969⁵. Angle closure, a common cause of post-PKP glaucoma, is difficult to control medically and frequently requires surgical intervention. Trabeculectomy and glaucoma drainage devices (GDDs) have become the mainstays of post-PKP glaucoma management.

Just as corneal transplantation is a risk factor for the development of glaucoma, glaucoma is associated with an increased risk of corneal failure and the subsequent need for corneal transplantation^6,7. The mechanisms for corneal failure secondary to glaucoma are not clear. Glaucoma surgery, particularly the placement of GDDs, may contribute to corneal failure through peripheral corneal–tube touch, the close proximity of tube tips to the corneal endothelium, and induced anterior chamber turbulence. In patients for whom vision is already compromised by glaucoma, worsening vision may indicate sub-clinical corneal decompensation.

Epidemiology

The incidence of glaucoma after penetrating keratoplasty is linked to the clinical indication for the transplant. Glaucoma occurs in approximately 1–6% of eyes that undergo surgery for keratoconus^8–10, in 3–16% of patients with corneal dystrophies and Fuchs’ endothelial dystrophy^8,10,11, and in 17–44% of eyes with pseudophakic or aphakic bullous keratopathy^8,9,11–13. Eyes with repeated corneal grafts have higher rates of both graft failure and glaucoma.

Pre-existing glaucoma is often more difficult to control after PKP¹⁴. Other factors that contribute to the development of glaucoma after PKP are a history of ocular trauma, pre-existing peripheral anterior synechiae, and combined PKP and cataract surgery^9,10,13,15.

Since endothelial keratoplasty (EK) is a relatively new procedure, the incidence of glaucoma after EK has not yet been well established. Early reports suggest that glaucoma is diagnosed or glaucoma medications are started in 0–18% of EK patients^16–18. In the first year after DSEK, Vajaranant et al. reported starting glaucoma medications in 18% of patients without pre-existing glaucoma and increasing glaucoma therapy in 33% of patients with a history of glaucoma¹⁸. While only 0.3% patients without pre-existing glaucoma went on to require glaucoma surgery, 7 of 85 (8%) patients with a history of glaucoma required additional glaucoma surgery after EK. In this study, median IOP peaked 3 months after EK, which led the authors to speculate that many of the patients were steroid responders. Although no long-term data were presented, Lee et al. reported elevations of IOP to greater than 30 mmHg in 13% of patients in the first 6 months after EK¹⁶. Most IOP elevations occurred during the first week after surgery.

Thirty-six to 76% of patients undergoing keratoprosthesis surgery were treated preoperatively for glaucoma^19,20. The onset or progression of glaucoma as determined by finger tension, visual field progression, or optic nerve head cupping has been reported in 6–28% of cases after keratoprosthesis surgery^19,21.

Diagnosis

The diagnosis of glaucoma after PKP can be made challenging immediately postoperatively, by the presence of large epithelial defects, stromal edema, or irregular astigmatism, all of which can complicate applanation tonometry. Early postoperatively, the Tonopen (Reichert, Depew, New York) or the pneumotonometer may more accurately measure IOP. With high regular astigmatism, IOP is best obtained by averaging two Goldmann applanation measurements taken with the prisms 90° apart²². In some cases, digital palpation may be the best way to obtain an estimate of the IOP.

Difficulty visualizing the optic nerve handicaps diagnosis and monitoring of glaucomatous optic neuropathy. Opaque or edematous corneas may make accurate assessments of the optic nerve head impossible. In post-PKP glaucoma, graft edema, a poor ocular surface, and/or astigmatism may compromise the stereoscopic view, distort photos, or invalidate advanced imaging of the optic nerve.

The replacement of the cornea and ocular surface with a keratoprosthesis makes standard methods of IOP measurement inaccurate. Measurements taken at the limbus, on the sclera, or through conjunctivalized cornea are influenced, for example, by corneal sutures and the proximity of the keratoprosthesis back plate. Finger tension may be the best way to approximate IOP and serial measurements, using the same approach on the same part of the eye, may offer some insight into IOP trends. Compared to PKP, a keratoprosthesis provides an astigmatically neutral and clear visual axis through which to view the optic nerve. The view is similar to that obtained through an undilated pupil. Photographs and advanced imaging of the optic nerve are possible through a Boston keratoprosthesis.

Etiology

The development of glaucoma in eyes with a history of corneal surgery is usually attributed to changes in angle anatomy. Peripheral anterior synechiae (PAS) are present in up to 87% of patients after PKP^10,15. In our experience, PAS caused by corneal transplantation leads to progressive angle closure which is often difficult to manage without surgical intervention. The slowly progressive and recurrent nature of the angle closure prevents goniosynechialysis from being effective: synechiae tend to re-form. Figure 40.1 demonstrates a patient with areas of synechial angle closure.

Fig. 40.1 Synechial angle closure in a pseudophakic patient after penetrating keratoplasty (A and B). Corectopia was caused by nasal and temporal iridocorneal adhesions at the graft–host junction.

Distorted angle anatomy, anterior and posterior to the trabecular meshwork, has been implicated in the development of open angle glaucoma after PKP²³. Anterior to the angle, tight and long sutures may compress the trabecular meshwork²⁴. Posterior to the angle, loss of the support normally provided by the ciliary body–lens system allows for collapse of the trabecular meshwork.

Donor tissue diameter may also contribute to the development of glaucoma post-keratoplasty. Using an oversized graft may decrease the incidence of glaucoma after PKP in the early postoperative period. This effect may persist in the long term, but there is limited supporting data^23–26. In contrast, an undersized graft may decrease angle width and increase the risk for development of PAS with subsequent angle closure.

Immediately post-PKP, elevated IOP may be caused by iritis, hemorrhage, ciliary block glaucoma, and retained viscoelastic. In eyes inflamed preoperatively, the risk of pupillary block due to iritis or hemorrhage may be reduced by an iridotomy. Postoperative steroid response (in 20–32% of patients depending on the definition of steroid induced glaucoma) may elevate IOP in susceptible patients^27,28.

EK involves injecting air into the anterior chamber and keeping the patient supine to assist in graft attachment. Typically this involves an anterior chamber air fill of 10–60 minutes followed by an exchange of approximately half of the air for balanced salt solution. This can lead to complications: immediately postoperatively, patients in whom air bubbles extend beyond the inferior pupillary border in the supine position may experience pupillary block. In Tillett’s original description of posterior lamellar keratoplasty, the cornea was clear on the first postoperative day, but the iris was pushed forward against the graft creating anterior synechiae and angle closure with elevated IOP²⁹. A second mechanism for IOP elevations occurring within one week of EK has been postulated by Lee and co-authors: air displaced behind the iris leads to iridocorneal adhesions and elevated IOP¹⁶.

Postoperatively, steroid response appears to be a common cause of post-EK IOP elevation¹⁸. Fuchs’ endothelial dystrophy, which typically carries a lower risk of glaucoma after PKP, was the most common indication for surgery in the study presented by Vajaranant et al. The median IOP peaked at 3 months postoperatively. Topical steroids were commonly used four times daily for the first 4 months then slowly tapered¹⁸. Using a more rapid steroids taper, Lee et al. reported only one patient with IOP elevation after 6 months follow-up¹⁶.

Medical treatment of glaucoma after PKP

In patients with a history of corneal surgery, topical medications should be used cautiously with awareness of certain risks and benefits. Most topical medications contain preservatives which can be detrimental to the ocular surface: many glaucoma medications contain benzalkonium chloride. If preservatives in topical medications are harmful to the cornea, alternate products such as brimonidine and travoprost should be considered as they have been formulated with Purite® and Sofzia® preservatives, respectively. Timolol and pilocarpine are available in preservative-free forms.

While prostaglandin analogs are effective in lowering IOP, increased risks of intraocular inflammation and cystoid macular edema have been associated with their use. Increased intraocular inflammation can compromise long-term graft survival. As latanoprost has been associated with recurrent herpetic keratitis, prostaglandin agonists should be used with caution in patients with a history of herpetic eye disease³⁰. We use prostaglandin agonists in patients with post-keratoplasty glaucoma, if the alternative is surgery for IOP control. Many of the reported side effects of prostaglandin analoge have been based upon case reports and no controlled or prospective studies have demonstrated any deleterious effects of prostaglandin analoge on graft survival.

Miotics are usually ineffective and not recommended in the early postoperative period; pilocarpine breaks down the blood–aqueous barrier and increases intraocular inflammation. The risk of shallowing of the anterior chamber with the potential for PAS limits the use of miotics after keratoplasty.

Carbonic anhydrase inhibitors (CAIs) can be used topically or systemically. Although the carbonic anhydrase enzyme is expressed by the corneal endothelium and is associated with endothelial pump function, the effect of CAIs on the graft is likely to be clinically significant only if the graft is near failure. We treat glaucoma in the presence of a graft as we do in patients who have not had a history of corneal surgery.

Indications for surgery

Surgical intervention is indicated when medical treatment has failed. Medications fail when IOP is uncontrolled despite the use of medications from all available and tolerated classes. Frequently one or more medication classes cannot be tolerated due to side effects or allergy. Non- or poor compliance may be an indication for glaucoma surgery. When corneal transplantation becomes necessary in a glaucoma patient, we assume that there will be greater difficulty in controlling the IOP after the PKP than before. Patients managed on maximal tolerable medical therapy and undergoing PKP should be considered for combined cornea and glaucoma procedures, or for glaucoma surgery prior to the corneal surgery. Patients should be advised of the complicated nature of their surgeries and the long-term implications for its success.

Preoperative assessment