Factors Associated With Long-term Progression or Stability in ExfoliationGlaucoma FREE

MANY INVESTIGATORS believe that exfoliation glaucoma is more severethan primary open-angle glaucoma (POAG).^{1 - 3} Konstasand coworkers^{4 - 5} found that theinitial intraocular pressure (IOP) was higher in patients with exfoliationglaucoma than in patients with POAG. Similar findings have been noted by Tezeland Tezel,⁶ Lindblom and Thorburn,⁷ and Futa and associates,⁸ althoughnot by Linnér and associates.⁹ Further,Teus and coworkers¹⁰ demonstrated that higheruntreated IOP levels were associated with greater visual field loss (meandeviation).

Several authors have noted that over time the visual field loss andoptic nerve damage in exfoliation glaucoma are more rapid and associated witha higher IOP than in POAG.² Accordingly, exfoliationglaucoma is believed by many investigators to be more difficult to manageclinically, with a higher incidence of treatment failure than POAG.^{1 ,11} In a long-term study, Blika and Saunte,¹² showed that exfoliation glaucoma was controlled witha β-blocker in only 8% of patients compared with 33% of patients withPOAG. Additionally, Pohjanpelto¹³ demonstratedthat during approximately 10 years of follow-up, 35% of patients with exfoliationglaucoma required trabeculectomy to control their IOP compared with 18% ofpatients with POAG. Despite the reported incidence of increased progressioncompared with POAG, however, little data exist that evaluate patients withexfoliation glaucoma and the treatment end points that would help preventglaucomatous progression and visual field loss. The purpose of this studywas to retrospectively evaluate patients with exfoliation glaucoma with 5years of follow-up or longer to determine the IOP levels associated with progressionor stable glaucomatous optic disc cupping or visual field findings.

We included patients who met 1 of 2 criteria: stable exfoliation glaucomafor a minimum of 5 years or diagnosis or progression of exfoliation glaucomaduring the follow-up period. Patients were chosen from consecutive medicalrecords at the practices of the study investigators and reviewed alphabetically.Each site was asked to identify 50 consecutive patients with exfoliation glaucoma,of whom only those with glaucomatous damage were included for this analysis.

Data collection began from the patient's initial examination. The initialIOP, however, was excluded to allow for adjustment of therapy for elevatedIOP. Data were recorded from each available visit included in the follow-upperiod. For patients with stable glaucoma, data were collected for as longas records were available. For those with glaucoma progression, data werecollected until the time the glaucoma worsened. Data were not recorded afterthe time of progression so that the information included in this study wouldreflect the ocular condition that caused the progression.

Each patient had been diagnosed by the investigator as having exfoliationglaucoma secondary to glaucomatous optic disc (neural rim thinning or notching,saucerization, thin nasal rim, or total cupping) and/or visual field changes(typical nerve fiber layer changes, including nasal step or paracentral Seidelor arcuate scotoma). Patients demonstrated typical anterior segment findingsof exfoliation syndrome.¹⁴ Excluded from thisstudy were patients with congenital, primary, narrow-angle, or low-tensionglaucoma and patients thought to have progressive nonglaucomatous visual loss.

Data recorded during the follow-up period from each visit included resultsof Goldmann applanation tonometry, evaluation of the ocular adnexa, and slitlampbiomicroscopy. Routine follow-up visits typically were performed every 3 to6 months. Dilated optic disc and visual field examinations generally werecompleted yearly or more frequently if required. At dilated examinations,the optic disc was examined by stereoscopic techniques. The same investigatorevaluated each patient during the follow-up period.

The investigator determined progression clinically. In each case, progressionmust have been noted in the medical record with the associated reason. Generally,the criterion for progression was an increase in thinning of the neural rimor a worsening of glaucomatous visual field loss. In patients with total glaucomatouscupping and diffusely depressed visual fields, worsened visual acuity wasalso used as a sign of progression. Patients without "progression" noted wereassumed to be stable. One of us (W.C.S.) reviewed the suitability of the designationof each patient as stable or progressed in a masked fashion. If the medicalrecord data seemed inconsistent with either progression or stability, theclinical findings were reviewed with the investigator to assure accuracy.

Data collection and statistical analysis between patients who were eitherstable or progressed were performed as follows. All data were 2-sided andunpaired. A P value of .05 was selected as determinatestatistical significance. A t test was used betweengroups to analyze data for patient age, mean and peak IOP, the number of officevisits, the number of medication changes per year, the number of medicationstaken at the end of the study, the study term in years, the number of lasertrabeculoplasties, and the number of trabeculectomies per year.^{15 - 16} Riskfactors for progression were analyzed by both a univariate analysis and amultivariate Spearman ρ correlation analysis. An F test analyzed the differencesbetween the variance of each patient's IOP, measured during the follow-upperiod.¹⁵

A χ² test was used to analyze differences between groupsof nonordered scores, such as left or right eye, sex, and optic disc and visualfield status. A Mann-Whitney U test was used to evaluatevisual acuity.¹⁷ If both eyes of a patientmet the criteria for entrance into the study, only one eye was randomly chosento be analyzed.

We included 167 patients (82 progressed and 85 stable) in this study.Baseline patient characteristics at the beginning of the follow-up periodare presented in Table 1. Thestable glaucoma group had a mean ± SD of 13.8 ± 7.0 visits perpatient during 6.1 ± 2.3 years, and the progressed glaucoma group had8.5 ± 5.1 visits during 3.4 ± 1.7 years (Table 2).

The mean ± SD IOP for the stable glaucoma group was 18.1 ±2.6 mm Hg and for the progressed glaucoma group was 20.1 ± 4.3 mm Hg.A statistical difference existed between groups for mean IOP (P<.001). In Figure 1,the mean IOPs for each patient in both groups are shown. Several differentmean IOP levels were instructive in differentiating the clinical course ofpatients in this study. Eleven (28%) of 40 patients with an IOP 17 mm Hg orlower had their glaucoma progress. In contrast, 29 (43%) of 67 patients withan IOP of 18 to 19 mm Hg had their glaucoma progress. Progression was moremarked when IOPs were 20 mm Hg or higher (42 [70%] of 60 patients) (Figure 1).

The highest mean ± SD recorded peak IOP was 24.1 ± 5.4mm Hg in the stable glaucoma group and 29.2 ± 10.3 mm Hg in the progressedglaucoma group (P<.001). In addition, the meanSD of the IOP for each patient was significantly higher in the progressed(4.6 mm Hg) than the stable (2.9 mm Hg) glaucoma group (P<.001) (Table 2). Figure 2 shows the mean SD of all measuredIOPs for each patient. Generally, patients whose SD was less than 3 mm Hgor lower (37% progressed glaucoma) were more stable than those whose SD was4 mm Hg or higher (71% progressed glaucoma). Figure 3 shows the rate of progression of glaucoma during the first5 years of follow-up.

At baseline, the progressed group had significantly more advanced visualfield, visual acuity, and optic disc changes by univariate analysis (Table 1). During the follow-up period,the progressed glaucoma group had significantly more office visits, lasertrabeculoplasties, and trabeculectomies per year (Table 2). By multivariate analysis, a significant correlation toprogression was observed with the cup-disc ratio at diagnosis, as well asthe number of trabeculectomies, medication changes, and the mean IOP overtime (Table 3). We examined patientswith advanced optic disc damage and those with visual field changes, separately.The IOPs associated with progression or stability of the glaucoma did notchange in this group compared with patients with less advanced glaucoma.

Controversy still exists over the proper treatment end points of patientswith advanced POAG. Several historic and recent studies have demonstratednot only the benefit of IOP reduction in POAG but have indicated specifictarget IOPs that help prevent progressive glaucomatous damage. These reportshave implied that patients with moderate or advanced glaucomatous damage usuallyremain stable, with a mean IOP between 15 and 18 mm Hg for 5 to 10 years.^{18 - 26} Severalstudies, however, have indicated a further benefit in patients with advancedglaucoma when IOPs are as low as 12 to 15 mm Hg.^{20 ,26 - 27} HigherIOPs generally led to greater incidences of progressive glaucomatous damage.For example, in 2 of these reports, patients with slightly higher mean IOPs(19-21 mm Hg) had their glaucoma progress in approximately 50% to 67% of cases,and in almost 100% of patients, progression occurred when IOPs were 22 mmHg or higher.^{23 ,28} In addition,several studies have noted that the extent and rate of glaucomatous damageare worse with higher peak IOPs.^{23 ,28 - 32} However,the level of mean or peak IOPs that would provide safety for all patientshas not yet been defined clearly.²⁶

More data are needed to establish treatment end points that will helppreserve vision in patients with exfoliation glaucoma. Such information isespecially important for these patients for 2 reasons. First, exfoliationglaucoma is a more severe form of glaucoma, more often associated with visionloss than POAG. Second, although not as common as POAG, exfoliation glaucomais probably the most common identifiable entity that leads to glaucoma.¹⁴ Further, in some populations (eg, Baltic and Mediterranean),its incidence may reach a higher percentage of the population than the primaryform of the disease.¹⁴

We evaluated the long-term follow-up in patients with exfoliation glaucomato determine treatment levels that would help prevent progression. Our resultsindicated that patients with an IOP of 17 mm Hg or lower were most likelyto remain stable during 5 to 10 years of follow-up. Twenty-eight percent ofthese patients' glaucoma progressed during this time. However, some patientswith mean IOPs of 13 mm Hg still had their glaucoma progress during the follow-upperiod. Two patients with an IOP of 12 mm Hg or lower remain stable; however,this was too few patients to determine if this was a safe level or not. Patientswith an IOP of 18 to 19 mm Hg demonstrated a 43% chance of progression oftheir glaucoma during the follow-up period. In contrast, patients with a meanIOP of 20 mm Hg or higher had a 70% incidence of progression. In addition,progressed glaucoma patients had a statistically greater peak IOP and variationof IOP over time.

The findings in this study mirror those found previously in patientswith POAG. These studies showed that IOPs of 17 to 18 mm Hg or lower wererequired to help prevent progression of their glaucoma for 5 to 10 years.They, too, noted that elevated peak IOPs more often lead to progression.

The patient characteristic at baseline that was associated with progressionby multivariate analysis was more advanced optic disc damage. However, theIOP levels required to control glaucoma progression in patients with advanceddisc and field damage were not different from those required for patientswith less advanced disease. This finding was consistent with a recent studyperformed in a similar fashion with POAG patients.²⁵ Othercharacteristics of patients whose glaucoma progressed during the follow-upperiod included a greater number of medication changes and trabeculectomiesper year. These findings may indicate that physicians had a more difficulttime controlling the progressed group, allowing for higher IOPs and a greatertendency for glaucomatous damage.

Unfortunately, some patients with reduced IOP did continue to have theirglaucoma progress. It is not clear from the results of this study why IOPreduction does not completely stabilize vision. It could be that other factorsexist, apart from the IOP, in some patients with exfoliation glaucoma thatcause them to be more susceptible to progression. Determining such factorsin the future might lead to new treatments to better stabilize IOPs in patientswith this disease.

Our findings suggest that IOP reduction helps to prevent glaucoma progressionin patients with exfoliation glaucoma, although it does not guarantee theprevention or worsening of the disease. This study did not evaluate patientsin a prospective manner. Such research is needed to confirm these findingsas well as to identify specific risk factors that would help further subclassifypatients with exfoliation glaucoma and potentially lead to new IOP treatmentparameters for these groups to help stabilize this disease. In addition, thisstudy did not evaluate all populations with this disease. Intraocular pressurerequirements may differ in other genetically distinct populations with exfoliationglaucoma.

Corresponding author and reprints: William C. Stewart, MD, PharmaceuticalResearch Network, LLC, 1639 Tatum St, Charleston, SC 29412 (e-mail: prnc@bellsouth.net).

Submitted for publication October 10, 2002; final revision receivedMarch 10, 2003; accepted September 10, 2003.

This study was supported in part by Hungarian National Health Researchgrant ETT 293/2000 (Dr Hollo).