Glaucoma After Corneal Transplantation

Incidence

Improvements in technology, microsurgical techniques, and eye banking have led to greater functional success rates after PK. Yet for many years, little was known about changes in IOP occurring after PK because there was no accurate technique for measuring it. Schiotz readings are inaccurate after PK. Goldmann applanation is difficult in recently grafted eyes because of surface irregularity and fluorescein pooling around corneal sutures, and it may not be reliable in edematous corneas. The development of the electronic applanation tonometer in the 1960s ^, provided a tool to more accurately monitor IOP after keratoplasty. Using the Mackay-Marg tonometer, Irvine and Kaufman in 1969 noted an alarming incidence of elevated IOP in the early post-PK period. Although this study called attention to the frequency of an early rise in IOP after PK, little was known about the true incidence of persistent glaucoma after the first week.

Studies of long-term IOP elevation after PK soon followed, and common themes emerge from these studies. First, short-term IOP elevation was common in post PK eyes, with IOP elevation seen in 15%–46% of eyes postoperatively; most of these patients typically normalize after the first postoperative week. Second, long-term IOP elevation was also commonplace, with an incidence ranging from 27% to 34% of patients. ^, Lastly, eyes with preexisting glaucoma or IOP elevation in the first postoperative week were more likely to develop long-term glaucoma. ^{, ,}

More recent studies found similar incidence of posttransplant IOP elevation. When reporting simple ocular hypertension (defined as IOP > 21 mm Hg), Oruçoglu found that 70 of 146 eyes (47.9%) had at least one period of ocular hypertension. Using a more stringent definition of post-PK glaucoma as “durable elevated IOP ≥ 22 mm Hg at different time points …” in patients without prior glaucoma, and escalation of treatment in patients who had preexisting glaucoma, Huber et al. found that the incidence of post-PK glaucoma was 8.7% in 1848 post-PK eyes. Karadag et al. defined glaucoma as IOP > 21 mm Hg that necessitated the use of medication or glaucoma progression in patients with preexisting glaucoma. They found that in 749 eyes, 16.6% developed glaucoma or glaucoma progression in the postoperative period. Naturally, studies that had more stringent definitions of glaucoma found a lower incidence of glaucoma, accounting for the wide range of incidences that were reported.

Other authors described the incidence of treatment escalation in post-PK IOP elevation. Al-Mohaimeed performed a retrospective review of 715 eyes undergoing PK and found that 89 (12.4%) required glaucoma treatment escalation. Wagoner et al., reviewed 910 post-PK eyes and found similar results: 15.5% of patients required treatment for sustained IOP elevation.

Newer transplantation techniques that replace various layers of the cornea have evolved rapidly over the last decade, and these include Descemet stripping endothelial keratoplasty (DSEK) or Descemet stripping automated endothelial keratoplasty (DSAEK), Descemet membrane endothelial keratoplasty (DMEK), and deep anterior lamellar keratoplasty (DALK). For the purpose of this chapter, the term DSEK will also include the technique DSAEK. DSEK is a technique that replaces the endothelial layer of the cornea along with a small strip of stroma, and it has become the technique of choice for corneal transplantation in eyes with pseudophakic bullous keratopathy and Fuchs corneal dystrophy. The incidence of elevated IOP after DSEK not related to pupillary block has been reported to be 0%–41%. As with studies of post-PK eyes, the majority of studies that comment on pressure-related problems after DSEK focus on IOP rise only and did not report on actual damage to the optic nerve nor visual field loss, and thus a comment on glaucoma as a sequela of DSEK cannot be made accurately.

Lee et al. performed an extensive review of outcomes studies related to DSEK and of 34 articles reviewed, the rate of “iatrogenic glaucoma” was 0%–15% with an average of 3%. In another study of 108 post-DSEK eyes, incidence of ocular hypertension after DSEK was 47.2% with 29.7% of patients needing glaucoma surgery within a 2-year period.

Two studies compared the rate of IOP rise in different cohorts of post-DSEK patients. In one study, which included 805 eyes receiving DSEK, cohorts consisted of eyes with preexisting glaucoma but with no previous glaucoma surgery, eyes with glaucoma with previous glaucoma surgery, and eyes with no preexisting glaucoma. IOP rise was defined as IOP ≥ 24 mm Hg or relative increase in IOP from preoperative value ≥10 mm Hg at any postoperative examination. The rate of IOP rise was equally high in the glaucoma with no prior glaucoma surgery group (45%) as glaucoma with prior glaucoma surgery group (43%) versus the group with no preexisting glaucoma (35%). Subsequent glaucoma surgery was required in 5%, 19%, and 0.3% of patients in each group, respectively. At 1 year, 18% of eyes with no preexisting glaucoma required medications to control IOP. Another study looking at similar cohorts likewise showed that the rate of IOP rise was higher in eyes with preexisting glaucoma (41.3%) versus in eyes without preexisting glaucoma (20%).

For post-DMEK and post-DALK eyes, studies demonstrate a lower incidence of IOP elevation or glaucoma when compared to post-PKP eyes. Maier et al. studied a cohort of 117 post-DMEK eyes and found that the 12-month incidence of IOP elevation (IOP ≥ 22 mm Hg or ≥ 10 mm Hg from preoperative values) was 12.10%. Price et al., in a randomized comparison of post-DMEK regimen of prednisolone acetate 1% versus fluorometholone 0.1%, found that there is a higher incidence of IOP elevation (≥24 mm Hg or ≥10 mm Hg compared to preoperative IOP) in the prednisolone group than the FML group (22% vs. 6% in a 1-year period, respectively).

In post-DALK eyes, a similarly lower incidence of IOP elevation was seen. Huang et al, in a retrospective case series of 122 eyes post-DALK, found that an episode of IOP elevation (IOP > 21 mm Hg) occurred in 36.1% of cases, but only 4.48% developed de novo glaucomatous field defects. In a smaller study comparing 25 DALK eyes versus 100 PK eyes over a 1-year period, the authors found 14 cases of glaucoma in the PK group but none in the DALK group.

In summary, elevated IOP is common both immediately after corneal transplantation and later in the postoperative course. Its incidence does not seem markedly different in post-DSEK eyes in comparison to eyes that have received PK, though it may be lower in post-DALK or post-DMEK eyes. Vigilance in monitoring IOP in the first few weeks after surgery and beyond is crucial. Knowledge of risk factors that predispose an eye to long-term elevated IOP is helpful in identifying which patients may need to be monitored more closely.

Risk Factors

The identification of risk factors for post-PK glaucoma is key to its prevention and management ( Table 117.1 ). Patients with multiple risk factors may benefit from vigilant monitoring and proactive management of elevated IOPs.

Multiple studies have demonstrated that a history of preoperative glaucoma is highly predictive of postkeratoplasty IOP elevation. Al-Mohaimeed et al. found a preoperative history of glaucoma to be the most important risk factor for postoperative IOP elevation and treatment escalation (34.9% of people with preexisting glaucoma versus 9.5% in those who do not). Huber et al. reported that 39% (62/160) of the patients who develop post-PK glaucoma carried a pre-PK diagnosis of glaucoma. In the study by Oruçoglu and coworkers, preexisting glaucoma conferred a sixfold increased risk of developing subsequent post-PK ocular hypertension when compared to eyes without preoperative glaucoma. Karadag et al. also verified that patients with preexisting glaucoma are at higher risk of developing persistently elevated IOP after a PK. In a cohort of 749 eyes, post-PK glaucoma progressed in 59.4% of patients with preexisting glaucoma while 14.6% of eyes without glaucoma subsequently developed glaucoma.

Several studies were able to demonstrate that aphakia was significantly associated with post-PK glaucoma. In the study by Simmons and Stern of 229 eyes, aphakia was found to be a risk factor for developing elevated postoperative IOP. However, when subjects with preexisting glaucoma were excluded from the analysis, aphakia was no longer a significant risk factor. In 2016, Borderie et al. reported risk factors for IOP elevation necessitating treatment in a cohort of 1657 post-EK, post-DALK, and post-PK eyes. In the multivariate analysis, aphakic eyes were at a higher risk of developing postoperative glaucoma, with a hazard ratio (HR) of 2.83 when compared to phakic eyes.

Patients who had keratoconus are at lower risk for postkeratoplasty glaucoma than patients with other corneal diagnoses. ^, Al-Mohaimeed et al. found that patients undergoing keratoplasty for keratoconus or corneal stromal dystrophies had a cumulative glaucoma treatment escalation incidence of only 2.5%, while all other surgical indications for keratoplasty had a cumulative 22.3% prevalence of glaucoma therapy escalation.

The type of surgery that the eye has undergone is also a risk factor for subsequent development of glaucoma. In the Borderie study, IOL exchange or removal during surgery had a higher risk for postoperative glaucoma. Additionally, post-PK eyes are also found to be a risk factor for IOP elevation relative to other types or keratoplasties (adjusted HR vs. ALK, 1.12 and adjusted HR vs. EK, 1.10). Nguyen et al. examined 534 eyes with a mean follow-up of 2.7 years. In this study, eyes undergoing PK in conjunction with secondary IOL implantation or exchange and eyes with PK after previous cataract surgery had a higher incidence of postoperative glaucoma (21.4% and 18.7%, respectively).

With these studies in mind, the clinician should carefully monitor IOP in postkeratoplasty patients that have a history of preexisting glaucoma, aphakia, and multiple procedures performed at the time of surgery. For patients with preexisting glaucoma with marginal IOP control or for patients on multiple antiglaucoma medications, surgical management of glaucoma prior to or at the time of PK is recommended—glaucoma in these eyes will usually be made worse after PK without definitive and aggressive treatment, placing the graft and the patient’s vision at long-term risk.

The Prekeratoplasty Evaluation

Many of the risk factors outlined above can be identified preoperatively with a careful, glaucoma-oriented ophthalmic examination and history taking. Tonometry and a careful pupillary examination can often identify previously unsuspected glaucoma despite cloudy or opaque media. Where the media permit, careful gonioscopy can provide information crucial to surgical planning. In a patient with a significant proportion of the angle obstructed by anterior synechiae, postkeratoplasty glaucoma is a virtual certainty, and appropriate surgical options should be considered prior to surgery. When the media do not permit adequate gonioscopy, anterior segment optical coherence tomography (OCT) and/or ultrasound biomicroscopy (UBM) can help determine the configuration of the anterior chamber angle. This is particularly true in postkeratoplasty eyes with opaque media being evaluated for further grafting. Using UBM, Dada and coworkers found that 97% of eyes with elevated IOP had significant peripheral anterior synechiae.

Staging of optic nerve damage is crucial in the patient with established glaucoma. The presence of a large afferent pupillary defect is an ominous clinical sign. Visual fields are frequently unreliable in the patient with cloudy media and impossible to perform in the presence of opaque media. In evaluating patients with opaque media following trauma, Fuller and Hutton found that flash visual evoked potential (flash VEP) was the single best predictor of postoperative vision, followed by bright flash electroretinogram and ultrasonography. In the keratoplasty patient with known or suspected glaucoma, careful examination and documentation of the afferent visual system by pupillary response, subjective brightness sense, or flash VEP is essential to determine what levels of IOP are to be tolerated postoperatively, as well as in counseling the patient against unrealistic expectations.

In the patient with medically controlled glaucoma, a review of the patient’s regimen is necessary to know what medical options may be available postoperatively. A patient well-controlled on a prostaglandin analog (PGA) but who is asthmatic and allergic to α-agonists and topical carbonic anhydrase inhibitors (CAIs) may have few additional options for medical control after keratoplasty, and it is best that this be recognized prior to surgery.

In addition, knowledge of whether a patient is predisposed to IOP rise from steroid use is helpful to know prior to surgery because of the long term postoperative need for topical corticosteroid treatment. For patients who are “steroid responders,” concurrent glaucoma surgery may be necessary, especially for those who have known glaucomatous damage.

Clinical Presentation

Perhaps the most important factor in managing glaucoma after keratoplasty is its detection. Focusing only on the clarity of the corneal graft can delay recognition of glaucoma. The clinician should not forget simple observation of the optic nerve at each postoperative visit when media permit. Documentation of the optic nerve with photography is recommended as a baseline against which structural change can be assessed.

Interestingly, post-PK glaucoma usually presents in the early postoperative period with a clear graft that is thinner than a graft in which the IOP is normal. IOP may be markedly elevated to levels of 40–50 mm Hg; the corneal stroma is thinned by the compressing effect of increased IOP, and epithelial edema may or may not be present—grafts may remain clear even with chronically elevated IOP. The accurate measurement of IOP in this setting is problematic: astigmatism and alterations in corneal thickness can influence the accuracy of applanation measurement techniques, with thinner corneas underreading “true” IOP. We recommend estimating IOP with a variety of tonometers, taking several measurements over both the graft and the host cornea to measure a range of IOPs. The pneumatonometer, Tono-Pen, iCare Rebound Tonometer, and the dynamic contour tonometer all provide clinically acceptable (but not interchangeable) results in postkeratoplasty eyes.

Glaucoma and Graft Failure

Several studies indicate that glaucoma is a risk factor for corneal graft failure whether it is preexisting or develops after PK. Williams et al. reported PK graft survival probabilities by risk factor from data gleaned from the Australian Corneal Graft Registry. In 116 eyes with a history of glaucoma, graft survival probabilities using Kaplan-Meier analysis were 82% at 1 year and 66% at 2 years versus 93% at 1 year and 87% at 2 years in 796 eyes without a history of glaucoma before PK. A history of elevated IOP was a statistically significant risk factor for subsequent graft failure in univariate but not multivariate models. The authors suggested that glaucoma may only indirectly affect graft clarity through other risk factors with which it is associated. In contrast, the large (1090 subjects) and prospective Cornea Donor Study recently showed that a history of glaucoma, and in particular a history of prior glaucoma surgery, was a significant risk factor for graft failure.

Glaucoma may cause graft failure by several mechanisms. Animal studies have shown that high IOP causes damage to endothelial cells. Svedbergh studied moderate IOP elevations in the vervet monkey perfused in vivo for 3 hours and reported a wide spectrum of morphologic lesions of the corneal endothelium, from flattening and unevenness of the cell surface, vacuolization, blebbing, and disruption of the plasma membrane to frank cell loss.

Clinical studies are less conclusive. Among nonkeratoplasty patients, endothelial cell count in patients with ocular hypertension (IOP controlled and uncontrolled), primary open-angle glaucoma, and a control group were not found to be different. Bertelmann et al. performed noncontact endothelial cell counts every 6 months in 293 consecutive patients undergoing PK with a mean follow-up of 36 months. Patients with preoperative glaucoma had a significantly increased rate of endothelial cell loss. In contrast, Nguyen et al. found that endothelial cell density did not decrease in relation to high IOP after PK.

A special area of interest in DSEK is the impact of prior glaucoma surgery and glaucoma on graft failure ( Fig. 117.1 ). Several studies show that the graft failure rate after DSEK in eyes with prior glaucoma surgery (especially glaucoma drainage devices [GDDs]) is higher than that in eyes without a history of glaucoma. ^{, ,} For example, the rates of secondary graft failure (defined as irreversible corneal edema after the first postoperative month) were significantly higher in the glaucoma surgery group (15.9% vs. 3.2% in nonglaucoma surgery eyes) at a mean follow-up of 20.7 months . When comparing DSEK and DSAEK graft failure in medically controlled glaucoma versus eyes without glaucoma, studies with short-term follow-up (average 2 years) did not seem to show a difference. However, with longer-term follow-up (5 years) the rate of graft failure appeared to be significantly higher in eyes with medically controlled glaucoma than in eyes without glaucoma. More recent studies show similar relationships between graft failure, a history of glaucoma, and DSAEK. Pedersen et al. showed that the graft failure rate was higher in “high-risk eyes” (history of corneal neovascularization, glaucoma) after DSAEK than in eyes without these risk factors.

The etiology for higher failure rates over the years remains unclear. It has been reported that eyes with tube shunts have differential expression of aqueous humor proteins as compared to eyes without, including plasma proteins and complement C4a, thereby suggesting that the blood–aqueous barrier becomes relatively compromised. However, it is unknown whether this may differ for patients who have undergone trabeculectomy or if more prolonged exposure of endothelium to such proteins is a contributing factor to the increased rate of graft failure over time.

Price and colleagues suggested that eyes undergoing DSEK with preexisting glaucoma or IOP elevation due to steroid response exhibit a relative risk of 1.8 for graft rejection as compared to nonglaucomatous eyes and eyes without steroid response. This may in part be due to adjustments in corticosteroid regimen in response to IOP elevation and consequent immunologic rejection or from endothelial cell loss from elevated IOP.

An important cause of graft failure in post-DSEK eyes is dehiscence or dislocation. The presence of a prior trabeculectomy or tube shunt may allow the air bubble to escape into the subconjunctival space leading to graft dislocation (see Fig. 117.1 ). However, the literature on prior glaucoma surgery as a risk factor for graft dislocation is conflicting. One study found a significantly higher rate of graft dislocation in eyes with prior glaucoma surgery (trabeculectomy and GDD) versus a control group without prior surgery and dislocation was strongly associated with hypotony. Decroos et al. found a higher rate of graft dislocation in eyes with prior GDD but not trabeculectomy in comparison to eyes without prior surgery. In a multivariate analysis, eyes with prior trabeculectomy was an independent risk factor for graft dislocation. However, other studies show no difference in the graft dislocation rate between eyes with and without prior glaucoma surgery. ^{, , ,}

The preservative benzalkonium chloride (BAK), found in most glaucoma medications, has been implicated as a cause of graft failure. BAK has been related to the accumulation of inflammatory cells and mediators in conjunctival tissue, ^, and researchers have postulated that this increases the likelihood of immunologic graft rejection.