To review the role of intervisit intraocular pressure (IOP) fluctuation as an independent risk factor for glaucoma.
Perspective after literature review.
Analysis of pertinent publications in the peer-reviewed literature.
Disparate findings regarding the role of intervisit IOP variation have been published. IOP variation was a significant risk factor in the Advanced Glaucoma Intervention Study (AGIS), the Collaborative Initial Glaucoma Treatment Study, and other smaller studies. These studies have in common low IOPs (often after surgery) and moderately advanced disease. In the AGIS, when patients were stratified by mean IOP, only those patients with low IOPs showed the detrimental effects of IOP variation. IOP variation was not a significant risk factor in the Early Manifest Glaucoma Treatment Trial, and in 2 separate studies of ocular hypertensives. These studies have in common generally higher IOPs and an earlier stage of glaucoma (or no glaucoma at all). We believe these results are complementary rather than contradictory: existing data suggest that the effects of IOP variation depend on the characteristics of the patient, the baseline IOP, their stage of damage, the type of glaucoma, and other as-yet unknown factors.
Practitioners should consider whether patients who are progressing at low mean IOP may benefit from having IOP variation reduced. Single elevated measures of IOP may not be an anomaly or may not be related to compliance, but may identify patients who are at high risk for progressive glaucomatous damage, and thus should be monitored more carefully and potentially treated more aggressively.
The issue of whether intraocular pressure (IOP) reduction slows glaucomatous damage received a great deal of attention in the late 1980s as a result of a report written by Eddy and Billings. The issue long has been put to rest: a number of well-conducted, prospective, randomized clinical trials have provided overwhelming evidence that IOP reduction effectively slows the rate of development or progression of visual loss caused by glaucoma. The Ocular Hypertension Treatment Study (OHTS) determined whether IOP reduction slowed the rate at which patients with ocular hypertension developed glaucomatous damage, as defined by optic nerve change, visual field loss, or both. Treatment reduced the risk of developing glaucomatous damage from 10% to 5% over 5 years. The Collaborative Normal Tension Glaucoma Trial was the first prospective, randomized, long-term study of glaucoma treatment to demonstrate a reduction in the rate of progression of glaucomatous damage. It is perhaps ironic because it was thought by some that primary open-angle glaucoma in persons with relatively low IOPs (previously called normal-tension glaucoma) may not respond well to IOP reduction; however, the study established that aggressive lowering of IOP slows visual field progression even when the IOP initially is not high. The Advanced Glaucoma Intervention Study (AGIS) showed that, in patients with primary open-angle glaucoma receiving maximal medical therapy, further lowering of IOP with either argon laser trabeculoplasty or trabeculectomy reduced the rate of glaucomatous optic nerve damage as measured by visual field progression. There was a relationship between the magnitude of the IOP reduction and the magnitude of reduction in visual field progression. The Early Manifest Glaucoma Treatment Trial (EMGT) provided conclusive evidence that lowering IOP with laser trabeculoplasty and topical betaxolol reduces the rate of optic nerve damage and visual field progression in patients with newly diagnosed open-angle glaucoma. The risk of progression was approximately halved by treatment. Baseline risk factors for worsening included high IOP, presence of pseudoexfoliation, a worse visual field at diagnosis, presence of an optic disc hemorrhage, and older age. The risk of glaucoma progression was reduced by 10% for each 1-mm Hg reduction in IOP over the course of the study. This latter finding is consistent with the relationship of the magnitude of pressure reduction and progressive damage earlier reported by AGIS. The Collaborative Initial Glaucoma Treatment Study (CIGTS) showed that trabeculectomy or medical treatment as initial therapy for patients with open-angle glaucoma provided considerable pressure reduction over a long period and reduced visual field progression.
IOP reduction having been firmly established as effective treatment for glaucoma, this Perspectives article focuses attention on an aspect of the quality of IOP: its variation. There is accumulating evidence that, at least in some patients, IOP variation is a separate and important risk factor for glaucomatous damage. IOP variation over time may be divided into diurnal fluctuation that occurs on a daily basis, short-term IOP fluctuation that occurs over days to weeks, and long-term IOP fluctuation that occurs over months to years. In practice, IOP variation usually is quantified as intervisit IOP variation, with IOP measurements being made sequentially during office hours over a long period. Although this intervisit IOP variation primarily serves as a surrogate for long-term IOP fluctuation, it likely contains some component of diurnal and short-term fluctuation. We focus in this review on intervisit IOP variation as a measure of IOP fluctuation made in the course of the care of glaucoma patients.
Intraocular Pressure Variation
Several earlier publications have implicated IOP variation as a risk factor for glaucomatous damage. Bergea and associates, in a prospective study of risk factors, found that IOP range, peak, and mean all predicted visual field worsening. Stewart and associates reported in a retrospective, multicenter study that greater long-term standard deviation (SD) of IOP was found in eyes whose visual fields were becoming worse. Similarly, O’Brien and associates reported that the long-term SD and range of IOP measurements both correlated with the rate of glaucomatous visual field loss. IOP variation may be different in subtypes of open-angle glaucoma. For instance, pseudoexfoliation glaucoma may be associated with larger long-term pressure swings, and the relationship between this fluctuation and progression has been reported. Pigmentary glaucoma also may be expected to be associated with high long-term fluctuation, although the relationship between this and progression has not been studied. In a study of predictive factors for glaucomatous visual field progression in AGIS., Nouri-Mahdavi and associates found that IOP variation was a significant and independent predictor of worsening, despite the inclusion of mean IOP and the number of glaucoma interventions as independent covariates in the regression models. The study was limited because it was not designed specifically to evaluate IOP variation as a risk factor. A subsequent analysis reported by Caprioli and Coleman sought to evaluate this potential relationship more rigorously. In this post hoc analysis of the AGIS data, only eyes that had undergone 1 procedure were included, and only IOP measurements until the time of visual field progression were included. There was only a weak correlation between magnitude of IOP variation and mean IOP ( r = 0.03). IOP variation was confirmed as an independent risk factor for visual field progression in a multivariate model, with an odds ratio of 1.39 (95% conference interval, 1.09 to 1.79; P = .009). A subgroup analysis stratified patients by mean IOP: 2 groups were evaluated, 1 with low mean IOP and 1 with high mean IOP. In those with low mean IOP (mean IOP, 10.8 mm Hg), long-term IOP variation was as a significant and independent predictor of visual field progression, whereas IOP variation in the high mean IOP group (mean IOP, 20.6 mm Hg) was not a significant predictor of visual field progression. The authors concluded that IOP variation was important at low mean IOPs (eg, 10 to 12 mm Hg) and that IOP variation was not an important predictor at higher mean IOPs (eg, > 16 mm Hg). They concluded that when IOP was high, mean IOP was the overriding risk factor for visual field loss.
Lee and associates also found a significant association between long-term IOP variation and glaucoma progression. In a review of charts from several practices across the United States, they found that after controlling for age, mean IOP, visual field stage, and other covariates, each unit increase in IOP SD resulted in an approximately a 5 times higher risk of glaucoma. They concluded that IOP variability is an important predictor of glaucoma progression. Hong and associates reported that in patients with low IOP after combined glaucoma and cataract surgery, with mean IOPs of approximately 11 mm Hg in both groups, the group with the larger IOP variation had significantly worse visual field outcomes than those with lower IOP variations. The CIGTS recently confirmed an earlier report that pressure variation was an important, independent risk factor for glaucomatous progression. They found that 3 IOP measures were associated significantly with substantial worsening of the visual field defect and that the effects were similar in both surgical and medical groups. They found that a greater range of IOP ( P < .001), higher peak IOP ( P = .003), and a larger standard deviation of IOP ( P = .006) all were associated significantly with worse visual field outcomes in the 3- to 9-year follow-up. Predictors of higher range, peak, and SD of IOP were black race, higher baseline IOP, and clinical center. The authors concluded that the results supported the consideration of aggressive treatment when either high peak IOPs or a large variation in IOP is observed.
Other well-executed studies have found no relationship between IOP variation and progressive glaucomatous damage. In 2007, results from the EMGT showed that there was a high correlation between mean pressure and IOP variation and that IOP variation was not a separate predictor of visual field progression. Two studies of persons with ocular hypertension also concluded that there was little or no relationship between IOP variation as an independent factor for the prediction of the development of glaucomatous optic disc change or visual field loss. The results from OHTS in this regard have not yet been reported.
To summarize these apparently disparate findings, IOP variation was not a significant risk factor in the EMGT, and in 2 separate studies of ocular hypertensives. These studies have in common generally higher IOPs (which ranged between 16 and 25 mm Hg), an earlier stage of glaucoma (or no glaucoma at all), and modest or no treatment regimens. Surgery was not used in any of these studies to treat patients and EMGT patients were randomized to argon laser trabeculoplasty (ALT) and betaxolol or no treatment. The mean IOP in the treatment arm of the EMGT was approximately 16 mm Hg. IOP variation was a significant risk factor in the AGIS, the CIGTS, and in studies reported by Hong and associates and Lee and associates. These studies have in common low IOPs (often after surgery and ranging from 10 to 14 mm Hg) and moderately advanced disease. In the AGIS, when patients were stratified by mean IOP, only those patients with low IOPs (mean IOP of 11 mm Hg) showed the detrimental effects of IOP variation. To explain these findings, we offer the following hypothesis: when mean IOP is high, it is the strongest and driving risk factor for optic nerve damage; when the mean IOP is low, it may unmask IOP variation as an important contributory factor. Thus, IOP variation would be revealed as a risk factor when its effect is not swamped out by the important mean effect when it is high.
Although on initial observation these data may appear to be contradictory, we believe that in fact they are complementary. Not all primary open-angle glaucoma patients are the same: existing data suggest that the effects of IOP variation depend on the characteristics of the patient, the baseline IOP, their stage of damage, the type of glaucoma, and other as-yet unknown factors.