- Email: [email protected]
A 10-year follow up of ocular hypertensive patients within the Bolton Corneal Thickness Study
Available online at www.sciencedirect.com
Contact Lens & Anterior Eye 31 (2008) 147–153 www.elsevier.com/locate/clae
A 10-year follow up of ocular hypertensive patients within the Bolton Corneal Thickness Study Can measured factors predict prognostic outcomes?§ Neel Bhatt a, Rajan Bhojwani a,*, Anne Morrison a, Jeff Kwartz a, Mohammad Laiquzzaman b,c, Sunil Shah b,c a
Bolton Royal Infirmary, Minerva Road, Farnworth, Bolton BL4 0JR, United Kingdom Heart of England Foundation Trust, Lode Lane, Solihull, B91 2AW, United Kingdom c Birmingham & Midland Eye Centre, Dudley Road, Birmingham B18 7QH, United Kingdom b
Abstract Objective: This is a case note review of a cohort of patients examining the effects of central corneal thickness (CCT), presenting intraocular pressure (IOP), age and gender on the risk of progression of ocular hypertension (OHT) to primary glaucoma (POAG) over a 10-year period. Design: Cohort study with retrospective case note review. Participants and methods: 58 case notes from a cohort at the Bolton Royal Infirmary initially assessed 10 years ago were reviewed again. Presenting CCT, IOP, age and gender were recorded. Development of POAG was assessed by visual field and/or optic disc changes being present. Main outcome measures: The effects of CCT, IOP, age and gender on the risk of progression of OHT to POAG were analysed in a multivariate logistic regression model following a preliminary univariate analysis. Results: 50 out of 116 eyes developed primary open angle glaucoma over the 10-year period. Thinner CCT (odds ratio 0.985 associated with each 1 mm increase in CCT), higher presenting IOP (odds ratio 1.131 associated with each 1 mmHg increase in IOP) and increasing age (odds ratio 1.062 associated with each 1 year increase in age) were found to be associated with progression to POAG. Conclusion: Patients with a CCT of 579 mm or more, a presenting intraocular pressure of 26 mmHg or less and age 75 years or less had a lower risk of developing POAG within this cohort of patients. # 2008 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved. Keywords: Ocular hypertension; Glaucoma; Central corneal thickness
1. Introduction and aims Measurement of intraocular pressure (IOP) has been used for many years as one of the main ways not only to diagnose glaucoma suspect patients, but also to assess the control of established glaucoma. The importance of central corneal thickness (CCT) in the prediction of IOP has been a subject of interest since the introduction of the applanation tonometer by Goldmann and Schmidt in 1957 [1]. They Accepted at the Association of Research in Vision and Ophthalmology (ARVO) Annual Meeting, May, 2006. * Corresponding author. Tel.: +44 7741065722. E-mail address: [email protected] (R. Bhojwani).
calibrated the applanation tonometer for a standard CCT of 500 mm but emphasised that CCT could, in theory, influence the reading. Since then many studies have been performed to test this theory, particularly since ultrasonic pachymeters have become widely available in the past 10 years. Many of these studies have found that CCT does influence the measurement of IOP by applanation tonometry [2–8], and may therefore have implications in the management of ocular hypertension (OHT) and primary open glaucoma (POAG). Despite this effect, there is, as yet, no widely accepted correction factor for intraocular pressure that takes into account a patient’s central corneal thickness. A number of
1367-0484/$ – see front matter # 2008 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.clae.2007.12.003
148
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153
studies have been performed [4,6,9], but there is some variation in the suggested correction factors. In a metaanalysis, Doughty and Zaman suggested a correction factor of 2.5 mmHg for a 10% deviation from an expected normal central corneal thickness of 535 mm [10]. Several of the recent studies have shown that patients with OHT have thicker corneas on average than patients with POAG [2,3,9,11–15]. Shah et al. [2] in their initial assessment of the cohort presented in this paper, demonstrated that patients with OHT had a mean CCT of 579.5 mm compared with 550.1 mm for eyes with POAG. The aim of this study was to perform a 10 year review, looking at the case notes of the cohort of patients involved in the Bolton corneal thickness study [2], to determine what percentage of patients diagnosed with OHT developed POAG and what predictors of this were available from the original data.
2. Patients and methods The original Bolton study looked at patients attending a district general hospital ophthalmic clinic, who had agreed to have their CCT measured. They had been classified at a previous appointment as having OHT, POAG or normal tension glaucoma. The OHT had been diagnosed on there being an IOP of greater than 22 mmHg in at least one eye, but no visual field loss and healthy optic discs. Their CCT had been measured using the Nidek pachymeter (Nidek Co. Ltd., Aichi, Japan) in the mid-pupillary axis. Repeated sets of five readings were taken until the standard deviation of the five readings was 5 mm or less [2]. Intraocular pressure was measured using the Goldmann Applanation Tonometer. From the original cohort of 116 patients, the case notes of 74 patients were reviewed. 42 sets of notes were not available. Most of these had died and their records had been destroyed. These were excluded as the follow up period needed to be 10 years for statistical review. There were no patients in the original cohort who were found to be using medications known to affect the visual field, or have refractive error greater than conventions for visual field research. None of the patients had any form of intraocular surgery prior to CCT measurements. Of the patients reviewed, a further 16 were excluded as they had developed other ocular pathology that would bias the diagnosis of glaucoma when using the criteria below. These included; retinal vascular occlusion, diabetic eye disease treated with laser, myopic degeneration, secondary glaucoma, e.g. following complicated cataract surgery, neovascular glaucoma. Therefore, 58 patients (116 eyes) were included in this study. 54 out of the 58 patients were Caucasian and 4 patients were Indian. The study adhered to the Royal Bolton Hospital NHS Trust ethical guidelines, which are formulated from national and international guidelines. Baseline CCT (measured by Shah et al. [2]) and IOP off medication was recorded at the time of initial presentation
using the Goldmann applanation tonometer. The time of day that the CCT was measured was not found to be significant in a later study by the same group [16]. We made a diagnosis of progression to POAG if the patients developed any of the following criteria: optic disc changes (increase in cupping, rim pallor or disc haemorrhages), and/or two consecutive reliable abnormal glaucoma Hemifield tests following a normal visual field test, which has been shown to have comparable sensitivity and specificity to other algorithms in the classification of visual field tests as normal or glaucomatous [17]. Visual field tests were conducted using Humphrey 24/2 SITA standard algorithm. A test with less than 3 fixation losses, less than 10% false positive and false negative rates was considered reliable. Abnormal visual fields were identified by the Humphrey GHT field analyser, which compares each of 5 zones in the superior field with their corresponding mirror images in the inferior field and evaluates the severity of the disturbed point of each zone pair relative to a normative database (Humphrey field analyserHumphrey Instruments, San Leandro, CA, USA). An arcuate defect or nasal step defect was identified on the pattern standard deviation plot for each of the abnormal results; subjective analysis was not used to define an abnormal test result. Visual field grading scales were not used as these would not normally be at hand in the outpatient clinic and are therefore not routinely used in the management of our patients. It was noted which patients had been commenced on topical ocular hypotensive treatment during the course of the 10-year review, in order to examine the effect of this on the outcomes of the study. We did not have sufficient data available to include refractive error at the time of the baseline measurements as a variable in this study.
3. Statistical analysis StatsDirect Version 2.4.5 was used to analyse the data, analysing both eyes separately. The data was divided into two groups: those who developed primary open-angle glaucoma and those who did not, in order to use a Mann– Whitney U-test to compare means for the two groups. Onetailed tests were used to look for the presence or absence of predictive power for each parameter. Each of the parameters investigated is listed below: (1) (2) (3) (4)
Baseline CCT. Presenting IOP (off medication). Age to the nearest year at the time of review. Gender (using chi-squared test).
In addition to this, we looked at the effect of gender on CCT, using a two-tailed Mann–Whitney U-test, as well as performing a regression analysis to determine whether age at pachymetry influences CCT. The data was then entered for all significant parameters into a multivariate logistic
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153
149
regression model to establish which parameters could independently predict progression to POAG. This approach would also allow us to analyse all variables together, providing greater statistical power. We then applied the Doughty and Zaman correction factor of 2.5 mmHg for a 10% deviation from an expected normal CCT of 535 mm [10] to the presenting IOP to determine whether correction improves any correlation between the presenting IOP and progression to POAG in simple univariate testing.
4. Results 25 out of 58 patients in the study (43.1%) developed POAG over the 10-year follow up period. Boxplots showing the spread of baseline CCT, presenting IOP and age at follow up according to the diagnosis are given in Figs. 1–3. Fig. 1 shows a large cross over between the two groups, which may indicate a poorer predictive value for CCT. Fig. 2 shows an outlier in the non-glaucomatous group, this patient with an IOP of 40 mmHg did not develop glaucoma as treatment was commenced following this measurement. This value was, however, not excluded as no record of difficulty in measuring IOP was mentioned. Table 1 shows the mean with upper and lower 95% confidence intervals for baseline CCT, presenting IOP, corrected presenting IOP, and age according to current diagnosis.
4.1. CCT in glaucomatous and non-glaucomatous patients
Fig. 2. Boxplot showing the spread of presenting intraocular pressure according to diagnosis.
4.2. CCT in men and women 28 patients were male and 30 female. Men had significantly thinner corneas than women (568.9 mm vs. 587.1 mm, p = 0.007, two-tailed Mann–Whitney U-test). 4.3. CCT and age at pachymetry In our study, we found that age at pachymetry also affected CCT measured. Fig. 4 shows an increase in CCT with age (correlation coefficient = 0.28, p = 0.002). 4.4. IOP in glaucomatous and non-glaucomatous patients
The mean baseline CCT of eyes that developed POAG was thinner than those who did not develop POAG (see Table 1) but this was of borderline significance ( p = 0.075, one-tailed Mann–Whitney U-test).
Patients who developed POAG had higher mean presenting IOP than those who did not develop POAG (see Table 1, p = 0.002, one-tailed Mann–Whitney U-test). When the Doughty and Zaman correction factor [10] for CCT was applied to the presenting IOP, the significance of
Fig. 1. Boxplot showing the spread of baseline central corneal thickness according to diagnosis.
Fig. 3. Boxplot showing the age range according to diagnosis.
150
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153
Table 1 Parameter means (lower 95% confidence interval, upper 95% confidence interval) by diagnosis and p-values (to 3 decimal places) Diagnosis
Baseline central corneal thickness (mm)
Presenting intraocular pressure (mmHg)
Age (nearest year)
Corrected presenting intraocular pressure (mmHg)
Not developed glaucoma Developed glaucoma p-Value
583.5 (573.1–594.0) 571.3 (564.0–578.5) 0.075
25.5 (24.7–26.4) 26.9 (26.0–27.9) 0.002
73.0 (70.1–75.9) 77.9 (75.1–80.7) 0.006
23.3 (22.2–24.3) 25.2 (24.3–26.1) <0.001
4.7. Multivariate analysis In light of the above univariate test results we put the following four predictors into our multivariate logistic regression model: (1) (2) (3) (4)
Fig. 4. Scatter graph showing the relationship between age and central corneal thickness. central corneal thickness ¼ 0:932 age at pachymetry þ 516:326
the difference in mean IOP increased (see Table 1, p < 0.001, one-tailed Mann–Whitney U-test). 4.5. Age at 10-year review in glaucomatous and non-glaucomatous patients
CCT. Presenting IOP. Age at follow up. Gender.
The outcome variable was development of POAG (1 or 0). The results in Table 3 show that age at follow up is the most powerful predictor for development of POAG and CCT is also highly significant. Presenting IOP is also significant but is not as powerful a predictor as both age at follow up and CCT. Gender is not a significant predictor for development of POAG but was included in the model due to the difference in CCT between men and women. From these results we can derive the following formula to estimate the likelihood of a patient developing POAG over a 10-year period: Estimated likelihood ¼ 0:632 0:015 CCT þ 0:123 IOP þ 0:060 age
Patients who developed POAG during the study period were older on average than those who did not (see Table 1, p = 0.006, one-tailed Mann–Whitney U-test). 4.6. Gender in glaucomatous and non-glaucomatous patients Table 2 shows the number of males and number of females in each diagnostic category. A chi-squared analysis on the above data showed that gender was not related to development of glaucoma (chisquared = 0.0027, p = 0.959). Table 2 Cross tabulation of gender and development of glaucoma
Developed glaucoma Not developed glaucoma
No. (of eyes) of males
No. (of eyes) of females
24 32
26 34
where CCT = central corneal thickness; IOP = presenting intraocular pressure; age = age at follow up. 88 of the 116 eyes in the study were started on topical treatment, as decided by a treating clinician during the follow up period. There was, however, no standard set of criteria for treatment applied to all patients at any time. Treatment was instead decided according to the clinical judgement of treating doctors, the needs of the patients, and the available research at Table 3 Multivariate logistic regression results Predictor
Odds ratio per unit increase (95% confidence interval)
p-Value
Central corneal thickness Presenting intraocular pressure Age at follow up Gender
0.985 (0.973–0.998)
0.020
1.131 (1.002–1.276)
0.047
1.062 (1.019–1.106) 0.937 (for F against M) (0.408–2.152)
0.004 0.878
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153
the times that each patient was reviewed in clinic. We could therefore not reliably assess the effect of treatment in our analysis. However, if we were to include ‘‘decision to commence treatment’’ as a variable in the multivariate analysis, it would become by far the most powerful predictor of development of glaucoma, and make all other variables except CCT insignificant (data not shown).
5. Discussion CCT, despite being of borderline significance in univariate testing, was a good predictor of progression to POAG in multivariate testing. Other studies have also reported that CCT is a powerful predictor of development of POAG in OHT [18–20]. This finding is also consistent with a number of studies looking at CCT and progression of POAG [21,22]. However, the Early Manifest Glaucoma trial [23] and a study by Chauhan et al. [24] report no such association. In fact these studies demonstrate conflicting results and this may be due to the difference in recruitment criteria. The two studies recruited patients based on the presence of damage [24,25], so their designs may have favoured those liable to progress, possibly reducing the effect of CCT and IOP. Alternatively, this difference in results may suggest that CCT is valuable in identifying healthy eyes with apparently high IOP due to the effect of a high CCT in initial assessment, but is not a useful predictor of progression once glaucomatous damage is established. If the influence of CCT on glaucomatous damage is due to its effect on IOP measurement, then the recruitment criteria in this study demonstrates the effect of CCT more easily. It has been suggested, however, that the effect of CCT may not be solely due to its effect on IOP due to the lack of any association found by the Ocular Hypertension Treatment Study authors [18]. Corneal pathology that causes thinning is not documented in the literature as being linked to POAG, so it is unlikely that thin corneas themselves indicate factors unrelated to their effect on IOP measurement. The most likely explanation is that other factors confound the measurement of intraocular pressure by affecting the deformability of the cornea. The Goldmann tonometer uses corneal deformability to derive an indirect measurement of IOP and is therefore susceptible to differences in corneal curvature, mix and density of collagen fibres and hydration of the cornea, amongst other factors, that will affect elasticity. In addition to this, scleral rigidity, surrounding structures and the amount of fluorescein may also affect IOP measurement. Although CCT differs between men and women, this study found no significant evidence to suggest that either gender is more likely to develop POAG. The current Royal College of Ophthalmologists’ guidelines state that female gender is a risk factor for the development of ocular hypertension [26]. There is some debate as to whether gender affects CCT with some studies reporting a difference
151
[11] and others not [10]. In addition to suggesting that women have thicker corneas than men, the results of the Ocular Hypertension Treatment Study also showed a slight increase in risk for males of developing POAG. This was, however of borderline significance in multivariate testing despite appearing significant in univariate testing [18]. The Early Manifest Glaucoma trial [23] and another study by Herndon et al. [21] found no association between gender and glaucoma risk. Age was shown to be a significant risk factor for developing primary open-angle glaucoma in this study, and this was consistent with the Ocular Hypertension Treatment Study [18], the Early Manifest Glaucoma trial [23], and the Barbados Eye Studies [27]. Interestingly Herndon et al. [21] found that increasing age was associated specifically with worsened mean deviation of visual fields and increasing number of medications, but not with a larger cup:disc ratio. Although patients with diseases which may confound assessment of visual fields were excluded from this study, the data still suggests that age-related change could, theoretically, be implicated in the worsening of visual fields and this may in turn effect the decision as to how to treat the patient. The effect of age on CCT is far less certain with some studies reporting an age related thinning [11,28,29], and others reporting an increase [30,31]. Doughty and Zaman suggest that there is no substantial change in CCT beyond the infant years in whites, although this may not be true for other ethnic groups, some of whom show an age related thinning of the cornea [10]. This study shows a significant increase in CCT with age in a predominantly white group of patients (although the data is not longitudinal). This could be due to changes in the corneal structure with age or simply an effect of the increased risk of developing POAG with increasing age as this study shows. The latter is perhaps more likely because, if older patients with thinner corneas progress more quickly, fewer would be classified as OHT at any one time. The results of this study demonstrate that presenting IOP is a predictor for the development of POAG both in univariate analysis and multivariate analysis, which is consistent with the other studies [18,21,23]. Despite the lack of any association between CCT and presenting IOP in our data, application of the Doughty and Zaman correction factor [10] to the presenting IOP increases the power of IOP as a predictor of POAG risk ( p = 0.002 uncorrected, p < 0.001 corrected). There is some variation in correction factors suggested to date. A possible source of error in this debate, as pointed out by Doughty and Zaman, is the lack of distinction between corneal thickening and oedema in some studies looking at this association. In the same article they consider whether a correction can be applied generally to all eyes, or whether there is a difference between normal eyes, those with chronic disease, and those with acute disease. They analysed the data from all published studies documenting IOP and CCT with
152
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153
the rationale that if those who are found to have significantly lower IOP than ‘‘normal’’ are consistently found to have lower than ‘‘normal’’ CCT, then the is evidence for an association. In their analysis they found that the association for normal eyes is only just statistically significant, and smaller than that for eyes with chronic disease. The eyes with acute disease proved to be a more variable group, but the association was still present [10]. Other studies have looked at the association between CCT and IOP from a slightly different perspective—the association was investigated by cannulating the eyes of patients who are having surgery, and applying a known hydrostatic pressure to see if the thickness of the patients cornea has an impact on the pressure reading obtained by conventional tonometry from outside the eye. The study by Ehlers et al. [6] suggests a correction factor of 5 mmHg for a 70 mm difference in central corneal thickness, while the study by Whitacre et al. [4] suggests a smaller correction factor of 1 mmHg for a difference of 50 mm that was acknowledged by the authors to be only just statistically significant. These two studies were, however, carried out on very small numbers of patients who either had chronic disease or a history of acute disease in the past. Despite this they do provide valuable evidence through an alternative approach, demonstrating that IOP measurement needs to be corrected for CCT, as does a theoretical model by Orssengo and Pye [32]. The exact nature of a correction factor may however not be straightforward and many believe that such a factor is unlikely to be a simple linear one like those quoted so far. Larger studies need to be carried out to arrive at a formula for the correction of IOP before such a formula can become widely accepted. There is also debate as to whether the association can be linear. The effect of treatment was not included in this study as patients were treated by a number of different doctors in the period of follow up. Their medication was often changed either to suit the emerging needs of the patient or in response to new evidence in what has been an extremely well researched field over the past few years. If the effect of treatment was included for analysis in our multivariate model, it not only becomes by far the most powerful predictor of progression to primary open-angle glaucoma, but it reduces the effects of all other factors except central corneal thickness to the point that they become insignificant (data not given). While initially seeming to contradict all current knowledge as well as going against the very idea for treating patients in the first place, this actually shows that the patients in this study were almost all put on topical treatment prior to the development of primary open-angle glaucoma, suggesting that the patients at risk are mostly being identified, but perhaps not early enough to prevent glaucomatous damage in some cases. Topical ocular hypotensive medications have also been found to affect the corneal thickness [33,34], and it was also not possible to take this into account, however the effects
reported are small and may not have made a significant difference in this study. Due to the retrospective design of our study, we were unable to fix a set of treatment criteria for the patients involved, and as a result, patients with thick corneas or low presenting IOP may have been less aggressively treated, resulting in underestimation of the true effect of these variables. Its retrospective design and small numbers limited this study, however we were still able to show an association of a number of risk factors with the development of primary open-angle glaucoma using a multivariate logistic regression model. We tested the prediction logistic regression equation for estimated likelihood for progression using a cut off of 0.5. This showed that our model has 60% sensitivity and 74.24% specificity using our criteria for diagnosing glaucoma. This is likely to be a slight overestimate as the model has been tested on the data from which it was produced. To obtain a more accurate idea of the model’s sensitivity and specificity, we would need to test the model on a different data set. The cut off that has been used can be adjusted to increase sensitivity at the expense of specificity or vice versa depending on the relative importance of sensitivity and specificity when applying the model. The results of testing the model show that it is limited by the small numbers of patients in the study, but support the need for consideration of a number of the variables that we were unable to include in our study as neither the sensitivity nor the specificity are high enough to put the model into practice as it is.
Financial interest None.
Conflict of interest All authors declare no conflict of interest.
References [1] Goldman H, Schmidt T. Uber Applanationstonometrie. Ophthalmologie 1957;134:221–42. [2] Shah S, Chatterjee A, Mathai M, et al. Relationship between corneal thickness and measured intraocular pressure in a general ophthalmology clinic. Ophthalmology 1999;106:2154–60. [3] Argus WA. Ocular hypertension and central corneal thickness. Ophthalmology 1995;102:1810–2. [4] Whitacre MM, Stein RA, Hassanein K. The effect of corneal thickness on applanation tonometry. Am J Ophthalmol 1975;115:592–6. [5] Johnson M, Kass MA, Moses RA, Grodzki WJ. Increased central corneal thickness stimulating elevated intraocular pressure. Arch Ophthalmol 1978;96:664–5. [6] Ehlers N, Bramsen T, Sperling S. Applanation tonometry and central corneal thickness. Acta Ophthalmol (Copenh) 1975;53:34–43.
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153 [7] Ehlers N, Hansen FK. Central corneal thickness in low-tension glaucoma. Acta Ophthalmol (Copenh) 1974;52:740–6. [8] Hansen FK, Ehlers N. Elevated tonometer readings caused by a thick cornea. Acta Ophthalmol (Copenh) 1971;49:775–8. [9] Wolfs RC, Klaver CC, Vingerling JR, et al. Distribution of central corneal thickness and its association with intraocular pressure: The Rotterdam Study. Am J Ophthalmol 1997;123:767–72. [10] Doughty MJ, Zaman ML. Human corneal thickness and its impact on intraocular pressure measures: a review and meta-analysis approach. Surv Ophthalmol 2000;44:367–408. [11] Brandt JD, et al. Central corneal thickness in the ocular hypertension treatment study (OHTS). Ophthalmology 2001;108:1779–88. [12] Singh RP, Goldberg I, Graham SL, et al. Central corneal thickness, tonometry, and ocular dimensions in glaucoma and ocular hypertension. J Glaucoma 2001;10:206–10. [13] Herman DC, Hodge DO, Bourne WM. Increased corneal thickness in patients with ocular hypertension. Arch Ophthalmol 2001;119:334–6. [14] Ventura AC, Bohnke M, Mojon DS. Central corneal thickness in patients with normal tension glaucoma, primary open angle glaucoma, pseudoexfoliation glaucoma, or ocular hypertension. Br J Ophthalmol 2001;85:792–5. [15] Velten IM, Bergua A, Horn FK, et al. Central corneal thickness in normal eyes, patients with ocular hypertension, normal-pressure and open-angle glaucoma’s—a clinical study. Klin Monatsbl Augenheilkd 2000;217:219–24 [in German]. [16] Shah S, Spedding C, Bhojwani R, et al. Assessment of the diurnal variation in central corneal thickness and intraocular pressure for patients with suspected glaucoma. Ophthalmology 2000;107:1191–3. [17] Katz J, Sommer A, Gaasterland DE, Anderson DR. Comparison of analytic algorithms for detecting glaucomatous visual field loss. Arch Ophthalmol 1991;109:1684–9. [18] Gordon MO, Beiser JA, Brandt JD, et al. The Ocular Hypertension Treatment Study. Baseline factors that predict the onset of primary open-angle glaucoma. Arch Ophthalmol 2002;120:714–20. [19] Medeiros FA, Sample PA, Weinreb RN. Corneal thickness measurements and frequency doubling technology perometry abnormalities in ocular hypertensive eyes. Ophthalmology 2003;110:1903–8. [20] Medeiros FA, Sample PA, Weinreb RN. Corneal thickness measurements and visual function abnormalities in ocular hypertensive eyes. Am J Ophthalmol 2003;135:131–7.
153
[21] Herndon LW, Weizer JS, Stinnett SS. Central corneal thickness as a risk factor for advanced glaucoma damage. Arch Ophthalmol 2004;122:17–21. [22] Kim JW, Chen PP. Central corneal pachymetry and visual field progression in patients with open-angle glaucoma. Ophthalmology 2004;111:2126–32. [23] Leske MC, Heijl A, Hussein M, et al. Factors for glaucoma progression and the effect of treatment. The Early Manifest Glaucoma Trial. Arch Ophthalmol 2003;121:48–56. [24] Chauhan BC, Hutchinson DM, LeBlanc RP, et al. Central corneal thickness and progression of the visual field and optic disc in glaucoma. Br J Ophthalmol 2005;89:1008–12. [25] Leske MC, Heijl A, Hyman L, et al. The Early Manifest Glaucoma Trial. Design and baseline data. Ophthalmology 1999;106:2144–53. [26] The Royal College of Ophthalmologists. Guidelines for the management of open angle glaucoma and ocular hypertension; Copyright# 2004, The Royal College of Ophthalmologists, 17 Cornwall Terrace, London, NW1 4QW [27] Nemsure B, et al. Corneal thickness and intraocular pressure in the Barbados eye studies. Arch Ophthalmol 2003;121:240–4. [28] Alsbirk PH. Corneal Thickness. I. Age variation, sex difference and oculometric correlations. Acta Ophthalmol (Copenh) 1978;56:95– 104. [29] Foster PJ, Baasanhu J, Alsbirk PH, et al. Central corneal thickness and intraocular pressure in a Mongolian population. Ophthalmology 1998;105:969–73. [30] Hansen FK. A clinical study of the normal human central corneal thickness. Acta Ophthalmol (Copenh) 1971;49:82–9. [31] Yee RW, Matsuda M, Schultz RO, Edelhauser HF. Changes in the normal corneal endothelial cellular pattern as a function of age. Curr Eye Res 1985;4:671–8. [32] Orssengo GJ, Pye DC. Determination of the true intraocular pressure and modulus of elasticity of the human cornea in vivo. Bull Math Biol 1999;61(3):551–72. [33] Inoue K, Okugawa K, Oshika T, Amano S. Influence of dorzolamide on corneal endothelium. Jpn J Ophthalmol 2003;47:129–33. [34] Lass JH, Eriksson GL, Osterling L, Simpson CV. Comparison of the corneal effects of latanoprost, fixed combination latanoprost-timolol, and timolol: a double-masked, randomised, one year study. Ophthalmology 2001;108:264–71.
Contact Lens & Anterior Eye 31 (2008) 147–153 www.elsevier.com/locate/clae
A 10-year follow up of ocular hypertensive patients within the Bolton Corneal Thickness Study Can measured factors predict prognostic outcomes?§ Neel Bhatt a, Rajan Bhojwani a,*, Anne Morrison a, Jeff Kwartz a, Mohammad Laiquzzaman b,c, Sunil Shah b,c a
Bolton Royal Infirmary, Minerva Road, Farnworth, Bolton BL4 0JR, United Kingdom Heart of England Foundation Trust, Lode Lane, Solihull, B91 2AW, United Kingdom c Birmingham & Midland Eye Centre, Dudley Road, Birmingham B18 7QH, United Kingdom b
Abstract Objective: This is a case note review of a cohort of patients examining the effects of central corneal thickness (CCT), presenting intraocular pressure (IOP), age and gender on the risk of progression of ocular hypertension (OHT) to primary glaucoma (POAG) over a 10-year period. Design: Cohort study with retrospective case note review. Participants and methods: 58 case notes from a cohort at the Bolton Royal Infirmary initially assessed 10 years ago were reviewed again. Presenting CCT, IOP, age and gender were recorded. Development of POAG was assessed by visual field and/or optic disc changes being present. Main outcome measures: The effects of CCT, IOP, age and gender on the risk of progression of OHT to POAG were analysed in a multivariate logistic regression model following a preliminary univariate analysis. Results: 50 out of 116 eyes developed primary open angle glaucoma over the 10-year period. Thinner CCT (odds ratio 0.985 associated with each 1 mm increase in CCT), higher presenting IOP (odds ratio 1.131 associated with each 1 mmHg increase in IOP) and increasing age (odds ratio 1.062 associated with each 1 year increase in age) were found to be associated with progression to POAG. Conclusion: Patients with a CCT of 579 mm or more, a presenting intraocular pressure of 26 mmHg or less and age 75 years or less had a lower risk of developing POAG within this cohort of patients. # 2008 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved. Keywords: Ocular hypertension; Glaucoma; Central corneal thickness
1. Introduction and aims Measurement of intraocular pressure (IOP) has been used for many years as one of the main ways not only to diagnose glaucoma suspect patients, but also to assess the control of established glaucoma. The importance of central corneal thickness (CCT) in the prediction of IOP has been a subject of interest since the introduction of the applanation tonometer by Goldmann and Schmidt in 1957 [1]. They Accepted at the Association of Research in Vision and Ophthalmology (ARVO) Annual Meeting, May, 2006. * Corresponding author. Tel.: +44 7741065722. E-mail address: [email protected] (R. Bhojwani).
calibrated the applanation tonometer for a standard CCT of 500 mm but emphasised that CCT could, in theory, influence the reading. Since then many studies have been performed to test this theory, particularly since ultrasonic pachymeters have become widely available in the past 10 years. Many of these studies have found that CCT does influence the measurement of IOP by applanation tonometry [2–8], and may therefore have implications in the management of ocular hypertension (OHT) and primary open glaucoma (POAG). Despite this effect, there is, as yet, no widely accepted correction factor for intraocular pressure that takes into account a patient’s central corneal thickness. A number of
1367-0484/$ – see front matter # 2008 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.clae.2007.12.003
148
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153
studies have been performed [4,6,9], but there is some variation in the suggested correction factors. In a metaanalysis, Doughty and Zaman suggested a correction factor of 2.5 mmHg for a 10% deviation from an expected normal central corneal thickness of 535 mm [10]. Several of the recent studies have shown that patients with OHT have thicker corneas on average than patients with POAG [2,3,9,11–15]. Shah et al. [2] in their initial assessment of the cohort presented in this paper, demonstrated that patients with OHT had a mean CCT of 579.5 mm compared with 550.1 mm for eyes with POAG. The aim of this study was to perform a 10 year review, looking at the case notes of the cohort of patients involved in the Bolton corneal thickness study [2], to determine what percentage of patients diagnosed with OHT developed POAG and what predictors of this were available from the original data.
2. Patients and methods The original Bolton study looked at patients attending a district general hospital ophthalmic clinic, who had agreed to have their CCT measured. They had been classified at a previous appointment as having OHT, POAG or normal tension glaucoma. The OHT had been diagnosed on there being an IOP of greater than 22 mmHg in at least one eye, but no visual field loss and healthy optic discs. Their CCT had been measured using the Nidek pachymeter (Nidek Co. Ltd., Aichi, Japan) in the mid-pupillary axis. Repeated sets of five readings were taken until the standard deviation of the five readings was 5 mm or less [2]. Intraocular pressure was measured using the Goldmann Applanation Tonometer. From the original cohort of 116 patients, the case notes of 74 patients were reviewed. 42 sets of notes were not available. Most of these had died and their records had been destroyed. These were excluded as the follow up period needed to be 10 years for statistical review. There were no patients in the original cohort who were found to be using medications known to affect the visual field, or have refractive error greater than conventions for visual field research. None of the patients had any form of intraocular surgery prior to CCT measurements. Of the patients reviewed, a further 16 were excluded as they had developed other ocular pathology that would bias the diagnosis of glaucoma when using the criteria below. These included; retinal vascular occlusion, diabetic eye disease treated with laser, myopic degeneration, secondary glaucoma, e.g. following complicated cataract surgery, neovascular glaucoma. Therefore, 58 patients (116 eyes) were included in this study. 54 out of the 58 patients were Caucasian and 4 patients were Indian. The study adhered to the Royal Bolton Hospital NHS Trust ethical guidelines, which are formulated from national and international guidelines. Baseline CCT (measured by Shah et al. [2]) and IOP off medication was recorded at the time of initial presentation
using the Goldmann applanation tonometer. The time of day that the CCT was measured was not found to be significant in a later study by the same group [16]. We made a diagnosis of progression to POAG if the patients developed any of the following criteria: optic disc changes (increase in cupping, rim pallor or disc haemorrhages), and/or two consecutive reliable abnormal glaucoma Hemifield tests following a normal visual field test, which has been shown to have comparable sensitivity and specificity to other algorithms in the classification of visual field tests as normal or glaucomatous [17]. Visual field tests were conducted using Humphrey 24/2 SITA standard algorithm. A test with less than 3 fixation losses, less than 10% false positive and false negative rates was considered reliable. Abnormal visual fields were identified by the Humphrey GHT field analyser, which compares each of 5 zones in the superior field with their corresponding mirror images in the inferior field and evaluates the severity of the disturbed point of each zone pair relative to a normative database (Humphrey field analyserHumphrey Instruments, San Leandro, CA, USA). An arcuate defect or nasal step defect was identified on the pattern standard deviation plot for each of the abnormal results; subjective analysis was not used to define an abnormal test result. Visual field grading scales were not used as these would not normally be at hand in the outpatient clinic and are therefore not routinely used in the management of our patients. It was noted which patients had been commenced on topical ocular hypotensive treatment during the course of the 10-year review, in order to examine the effect of this on the outcomes of the study. We did not have sufficient data available to include refractive error at the time of the baseline measurements as a variable in this study.
3. Statistical analysis StatsDirect Version 2.4.5 was used to analyse the data, analysing both eyes separately. The data was divided into two groups: those who developed primary open-angle glaucoma and those who did not, in order to use a Mann– Whitney U-test to compare means for the two groups. Onetailed tests were used to look for the presence or absence of predictive power for each parameter. Each of the parameters investigated is listed below: (1) (2) (3) (4)
Baseline CCT. Presenting IOP (off medication). Age to the nearest year at the time of review. Gender (using chi-squared test).
In addition to this, we looked at the effect of gender on CCT, using a two-tailed Mann–Whitney U-test, as well as performing a regression analysis to determine whether age at pachymetry influences CCT. The data was then entered for all significant parameters into a multivariate logistic
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153
149
regression model to establish which parameters could independently predict progression to POAG. This approach would also allow us to analyse all variables together, providing greater statistical power. We then applied the Doughty and Zaman correction factor of 2.5 mmHg for a 10% deviation from an expected normal CCT of 535 mm [10] to the presenting IOP to determine whether correction improves any correlation between the presenting IOP and progression to POAG in simple univariate testing.
4. Results 25 out of 58 patients in the study (43.1%) developed POAG over the 10-year follow up period. Boxplots showing the spread of baseline CCT, presenting IOP and age at follow up according to the diagnosis are given in Figs. 1–3. Fig. 1 shows a large cross over between the two groups, which may indicate a poorer predictive value for CCT. Fig. 2 shows an outlier in the non-glaucomatous group, this patient with an IOP of 40 mmHg did not develop glaucoma as treatment was commenced following this measurement. This value was, however, not excluded as no record of difficulty in measuring IOP was mentioned. Table 1 shows the mean with upper and lower 95% confidence intervals for baseline CCT, presenting IOP, corrected presenting IOP, and age according to current diagnosis.
4.1. CCT in glaucomatous and non-glaucomatous patients
Fig. 2. Boxplot showing the spread of presenting intraocular pressure according to diagnosis.
4.2. CCT in men and women 28 patients were male and 30 female. Men had significantly thinner corneas than women (568.9 mm vs. 587.1 mm, p = 0.007, two-tailed Mann–Whitney U-test). 4.3. CCT and age at pachymetry In our study, we found that age at pachymetry also affected CCT measured. Fig. 4 shows an increase in CCT with age (correlation coefficient = 0.28, p = 0.002). 4.4. IOP in glaucomatous and non-glaucomatous patients
The mean baseline CCT of eyes that developed POAG was thinner than those who did not develop POAG (see Table 1) but this was of borderline significance ( p = 0.075, one-tailed Mann–Whitney U-test).
Patients who developed POAG had higher mean presenting IOP than those who did not develop POAG (see Table 1, p = 0.002, one-tailed Mann–Whitney U-test). When the Doughty and Zaman correction factor [10] for CCT was applied to the presenting IOP, the significance of
Fig. 1. Boxplot showing the spread of baseline central corneal thickness according to diagnosis.
Fig. 3. Boxplot showing the age range according to diagnosis.
150
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153
Table 1 Parameter means (lower 95% confidence interval, upper 95% confidence interval) by diagnosis and p-values (to 3 decimal places) Diagnosis
Baseline central corneal thickness (mm)
Presenting intraocular pressure (mmHg)
Age (nearest year)
Corrected presenting intraocular pressure (mmHg)
Not developed glaucoma Developed glaucoma p-Value
583.5 (573.1–594.0) 571.3 (564.0–578.5) 0.075
25.5 (24.7–26.4) 26.9 (26.0–27.9) 0.002
73.0 (70.1–75.9) 77.9 (75.1–80.7) 0.006
23.3 (22.2–24.3) 25.2 (24.3–26.1) <0.001
4.7. Multivariate analysis In light of the above univariate test results we put the following four predictors into our multivariate logistic regression model: (1) (2) (3) (4)
Fig. 4. Scatter graph showing the relationship between age and central corneal thickness. central corneal thickness ¼ 0:932 age at pachymetry þ 516:326
the difference in mean IOP increased (see Table 1, p < 0.001, one-tailed Mann–Whitney U-test). 4.5. Age at 10-year review in glaucomatous and non-glaucomatous patients
CCT. Presenting IOP. Age at follow up. Gender.
The outcome variable was development of POAG (1 or 0). The results in Table 3 show that age at follow up is the most powerful predictor for development of POAG and CCT is also highly significant. Presenting IOP is also significant but is not as powerful a predictor as both age at follow up and CCT. Gender is not a significant predictor for development of POAG but was included in the model due to the difference in CCT between men and women. From these results we can derive the following formula to estimate the likelihood of a patient developing POAG over a 10-year period: Estimated likelihood ¼ 0:632 0:015 CCT þ 0:123 IOP þ 0:060 age
Patients who developed POAG during the study period were older on average than those who did not (see Table 1, p = 0.006, one-tailed Mann–Whitney U-test). 4.6. Gender in glaucomatous and non-glaucomatous patients Table 2 shows the number of males and number of females in each diagnostic category. A chi-squared analysis on the above data showed that gender was not related to development of glaucoma (chisquared = 0.0027, p = 0.959). Table 2 Cross tabulation of gender and development of glaucoma
Developed glaucoma Not developed glaucoma
No. (of eyes) of males
No. (of eyes) of females
24 32
26 34
where CCT = central corneal thickness; IOP = presenting intraocular pressure; age = age at follow up. 88 of the 116 eyes in the study were started on topical treatment, as decided by a treating clinician during the follow up period. There was, however, no standard set of criteria for treatment applied to all patients at any time. Treatment was instead decided according to the clinical judgement of treating doctors, the needs of the patients, and the available research at Table 3 Multivariate logistic regression results Predictor
Odds ratio per unit increase (95% confidence interval)
p-Value
Central corneal thickness Presenting intraocular pressure Age at follow up Gender
0.985 (0.973–0.998)
0.020
1.131 (1.002–1.276)
0.047
1.062 (1.019–1.106) 0.937 (for F against M) (0.408–2.152)
0.004 0.878
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153
the times that each patient was reviewed in clinic. We could therefore not reliably assess the effect of treatment in our analysis. However, if we were to include ‘‘decision to commence treatment’’ as a variable in the multivariate analysis, it would become by far the most powerful predictor of development of glaucoma, and make all other variables except CCT insignificant (data not shown).
5. Discussion CCT, despite being of borderline significance in univariate testing, was a good predictor of progression to POAG in multivariate testing. Other studies have also reported that CCT is a powerful predictor of development of POAG in OHT [18–20]. This finding is also consistent with a number of studies looking at CCT and progression of POAG [21,22]. However, the Early Manifest Glaucoma trial [23] and a study by Chauhan et al. [24] report no such association. In fact these studies demonstrate conflicting results and this may be due to the difference in recruitment criteria. The two studies recruited patients based on the presence of damage [24,25], so their designs may have favoured those liable to progress, possibly reducing the effect of CCT and IOP. Alternatively, this difference in results may suggest that CCT is valuable in identifying healthy eyes with apparently high IOP due to the effect of a high CCT in initial assessment, but is not a useful predictor of progression once glaucomatous damage is established. If the influence of CCT on glaucomatous damage is due to its effect on IOP measurement, then the recruitment criteria in this study demonstrates the effect of CCT more easily. It has been suggested, however, that the effect of CCT may not be solely due to its effect on IOP due to the lack of any association found by the Ocular Hypertension Treatment Study authors [18]. Corneal pathology that causes thinning is not documented in the literature as being linked to POAG, so it is unlikely that thin corneas themselves indicate factors unrelated to their effect on IOP measurement. The most likely explanation is that other factors confound the measurement of intraocular pressure by affecting the deformability of the cornea. The Goldmann tonometer uses corneal deformability to derive an indirect measurement of IOP and is therefore susceptible to differences in corneal curvature, mix and density of collagen fibres and hydration of the cornea, amongst other factors, that will affect elasticity. In addition to this, scleral rigidity, surrounding structures and the amount of fluorescein may also affect IOP measurement. Although CCT differs between men and women, this study found no significant evidence to suggest that either gender is more likely to develop POAG. The current Royal College of Ophthalmologists’ guidelines state that female gender is a risk factor for the development of ocular hypertension [26]. There is some debate as to whether gender affects CCT with some studies reporting a difference
151
[11] and others not [10]. In addition to suggesting that women have thicker corneas than men, the results of the Ocular Hypertension Treatment Study also showed a slight increase in risk for males of developing POAG. This was, however of borderline significance in multivariate testing despite appearing significant in univariate testing [18]. The Early Manifest Glaucoma trial [23] and another study by Herndon et al. [21] found no association between gender and glaucoma risk. Age was shown to be a significant risk factor for developing primary open-angle glaucoma in this study, and this was consistent with the Ocular Hypertension Treatment Study [18], the Early Manifest Glaucoma trial [23], and the Barbados Eye Studies [27]. Interestingly Herndon et al. [21] found that increasing age was associated specifically with worsened mean deviation of visual fields and increasing number of medications, but not with a larger cup:disc ratio. Although patients with diseases which may confound assessment of visual fields were excluded from this study, the data still suggests that age-related change could, theoretically, be implicated in the worsening of visual fields and this may in turn effect the decision as to how to treat the patient. The effect of age on CCT is far less certain with some studies reporting an age related thinning [11,28,29], and others reporting an increase [30,31]. Doughty and Zaman suggest that there is no substantial change in CCT beyond the infant years in whites, although this may not be true for other ethnic groups, some of whom show an age related thinning of the cornea [10]. This study shows a significant increase in CCT with age in a predominantly white group of patients (although the data is not longitudinal). This could be due to changes in the corneal structure with age or simply an effect of the increased risk of developing POAG with increasing age as this study shows. The latter is perhaps more likely because, if older patients with thinner corneas progress more quickly, fewer would be classified as OHT at any one time. The results of this study demonstrate that presenting IOP is a predictor for the development of POAG both in univariate analysis and multivariate analysis, which is consistent with the other studies [18,21,23]. Despite the lack of any association between CCT and presenting IOP in our data, application of the Doughty and Zaman correction factor [10] to the presenting IOP increases the power of IOP as a predictor of POAG risk ( p = 0.002 uncorrected, p < 0.001 corrected). There is some variation in correction factors suggested to date. A possible source of error in this debate, as pointed out by Doughty and Zaman, is the lack of distinction between corneal thickening and oedema in some studies looking at this association. In the same article they consider whether a correction can be applied generally to all eyes, or whether there is a difference between normal eyes, those with chronic disease, and those with acute disease. They analysed the data from all published studies documenting IOP and CCT with
152
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153
the rationale that if those who are found to have significantly lower IOP than ‘‘normal’’ are consistently found to have lower than ‘‘normal’’ CCT, then the is evidence for an association. In their analysis they found that the association for normal eyes is only just statistically significant, and smaller than that for eyes with chronic disease. The eyes with acute disease proved to be a more variable group, but the association was still present [10]. Other studies have looked at the association between CCT and IOP from a slightly different perspective—the association was investigated by cannulating the eyes of patients who are having surgery, and applying a known hydrostatic pressure to see if the thickness of the patients cornea has an impact on the pressure reading obtained by conventional tonometry from outside the eye. The study by Ehlers et al. [6] suggests a correction factor of 5 mmHg for a 70 mm difference in central corneal thickness, while the study by Whitacre et al. [4] suggests a smaller correction factor of 1 mmHg for a difference of 50 mm that was acknowledged by the authors to be only just statistically significant. These two studies were, however, carried out on very small numbers of patients who either had chronic disease or a history of acute disease in the past. Despite this they do provide valuable evidence through an alternative approach, demonstrating that IOP measurement needs to be corrected for CCT, as does a theoretical model by Orssengo and Pye [32]. The exact nature of a correction factor may however not be straightforward and many believe that such a factor is unlikely to be a simple linear one like those quoted so far. Larger studies need to be carried out to arrive at a formula for the correction of IOP before such a formula can become widely accepted. There is also debate as to whether the association can be linear. The effect of treatment was not included in this study as patients were treated by a number of different doctors in the period of follow up. Their medication was often changed either to suit the emerging needs of the patient or in response to new evidence in what has been an extremely well researched field over the past few years. If the effect of treatment was included for analysis in our multivariate model, it not only becomes by far the most powerful predictor of progression to primary open-angle glaucoma, but it reduces the effects of all other factors except central corneal thickness to the point that they become insignificant (data not given). While initially seeming to contradict all current knowledge as well as going against the very idea for treating patients in the first place, this actually shows that the patients in this study were almost all put on topical treatment prior to the development of primary open-angle glaucoma, suggesting that the patients at risk are mostly being identified, but perhaps not early enough to prevent glaucomatous damage in some cases. Topical ocular hypotensive medications have also been found to affect the corneal thickness [33,34], and it was also not possible to take this into account, however the effects
reported are small and may not have made a significant difference in this study. Due to the retrospective design of our study, we were unable to fix a set of treatment criteria for the patients involved, and as a result, patients with thick corneas or low presenting IOP may have been less aggressively treated, resulting in underestimation of the true effect of these variables. Its retrospective design and small numbers limited this study, however we were still able to show an association of a number of risk factors with the development of primary open-angle glaucoma using a multivariate logistic regression model. We tested the prediction logistic regression equation for estimated likelihood for progression using a cut off of 0.5. This showed that our model has 60% sensitivity and 74.24% specificity using our criteria for diagnosing glaucoma. This is likely to be a slight overestimate as the model has been tested on the data from which it was produced. To obtain a more accurate idea of the model’s sensitivity and specificity, we would need to test the model on a different data set. The cut off that has been used can be adjusted to increase sensitivity at the expense of specificity or vice versa depending on the relative importance of sensitivity and specificity when applying the model. The results of testing the model show that it is limited by the small numbers of patients in the study, but support the need for consideration of a number of the variables that we were unable to include in our study as neither the sensitivity nor the specificity are high enough to put the model into practice as it is.
Financial interest None.
Conflict of interest All authors declare no conflict of interest.
References [1] Goldman H, Schmidt T. Uber Applanationstonometrie. Ophthalmologie 1957;134:221–42. [2] Shah S, Chatterjee A, Mathai M, et al. Relationship between corneal thickness and measured intraocular pressure in a general ophthalmology clinic. Ophthalmology 1999;106:2154–60. [3] Argus WA. Ocular hypertension and central corneal thickness. Ophthalmology 1995;102:1810–2. [4] Whitacre MM, Stein RA, Hassanein K. The effect of corneal thickness on applanation tonometry. Am J Ophthalmol 1975;115:592–6. [5] Johnson M, Kass MA, Moses RA, Grodzki WJ. Increased central corneal thickness stimulating elevated intraocular pressure. Arch Ophthalmol 1978;96:664–5. [6] Ehlers N, Bramsen T, Sperling S. Applanation tonometry and central corneal thickness. Acta Ophthalmol (Copenh) 1975;53:34–43.
N. Bhatt et al. / Contact Lens & Anterior Eye 31 (2008) 147–153 [7] Ehlers N, Hansen FK. Central corneal thickness in low-tension glaucoma. Acta Ophthalmol (Copenh) 1974;52:740–6. [8] Hansen FK, Ehlers N. Elevated tonometer readings caused by a thick cornea. Acta Ophthalmol (Copenh) 1971;49:775–8. [9] Wolfs RC, Klaver CC, Vingerling JR, et al. Distribution of central corneal thickness and its association with intraocular pressure: The Rotterdam Study. Am J Ophthalmol 1997;123:767–72. [10] Doughty MJ, Zaman ML. Human corneal thickness and its impact on intraocular pressure measures: a review and meta-analysis approach. Surv Ophthalmol 2000;44:367–408. [11] Brandt JD, et al. Central corneal thickness in the ocular hypertension treatment study (OHTS). Ophthalmology 2001;108:1779–88. [12] Singh RP, Goldberg I, Graham SL, et al. Central corneal thickness, tonometry, and ocular dimensions in glaucoma and ocular hypertension. J Glaucoma 2001;10:206–10. [13] Herman DC, Hodge DO, Bourne WM. Increased corneal thickness in patients with ocular hypertension. Arch Ophthalmol 2001;119:334–6. [14] Ventura AC, Bohnke M, Mojon DS. Central corneal thickness in patients with normal tension glaucoma, primary open angle glaucoma, pseudoexfoliation glaucoma, or ocular hypertension. Br J Ophthalmol 2001;85:792–5. [15] Velten IM, Bergua A, Horn FK, et al. Central corneal thickness in normal eyes, patients with ocular hypertension, normal-pressure and open-angle glaucoma’s—a clinical study. Klin Monatsbl Augenheilkd 2000;217:219–24 [in German]. [16] Shah S, Spedding C, Bhojwani R, et al. Assessment of the diurnal variation in central corneal thickness and intraocular pressure for patients with suspected glaucoma. Ophthalmology 2000;107:1191–3. [17] Katz J, Sommer A, Gaasterland DE, Anderson DR. Comparison of analytic algorithms for detecting glaucomatous visual field loss. Arch Ophthalmol 1991;109:1684–9. [18] Gordon MO, Beiser JA, Brandt JD, et al. The Ocular Hypertension Treatment Study. Baseline factors that predict the onset of primary open-angle glaucoma. Arch Ophthalmol 2002;120:714–20. [19] Medeiros FA, Sample PA, Weinreb RN. Corneal thickness measurements and frequency doubling technology perometry abnormalities in ocular hypertensive eyes. Ophthalmology 2003;110:1903–8. [20] Medeiros FA, Sample PA, Weinreb RN. Corneal thickness measurements and visual function abnormalities in ocular hypertensive eyes. Am J Ophthalmol 2003;135:131–7.
153
[21] Herndon LW, Weizer JS, Stinnett SS. Central corneal thickness as a risk factor for advanced glaucoma damage. Arch Ophthalmol 2004;122:17–21. [22] Kim JW, Chen PP. Central corneal pachymetry and visual field progression in patients with open-angle glaucoma. Ophthalmology 2004;111:2126–32. [23] Leske MC, Heijl A, Hussein M, et al. Factors for glaucoma progression and the effect of treatment. The Early Manifest Glaucoma Trial. Arch Ophthalmol 2003;121:48–56. [24] Chauhan BC, Hutchinson DM, LeBlanc RP, et al. Central corneal thickness and progression of the visual field and optic disc in glaucoma. Br J Ophthalmol 2005;89:1008–12. [25] Leske MC, Heijl A, Hyman L, et al. The Early Manifest Glaucoma Trial. Design and baseline data. Ophthalmology 1999;106:2144–53. [26] The Royal College of Ophthalmologists. Guidelines for the management of open angle glaucoma and ocular hypertension; Copyright# 2004, The Royal College of Ophthalmologists, 17 Cornwall Terrace, London, NW1 4QW [27] Nemsure B, et al. Corneal thickness and intraocular pressure in the Barbados eye studies. Arch Ophthalmol 2003;121:240–4. [28] Alsbirk PH. Corneal Thickness. I. Age variation, sex difference and oculometric correlations. Acta Ophthalmol (Copenh) 1978;56:95– 104. [29] Foster PJ, Baasanhu J, Alsbirk PH, et al. Central corneal thickness and intraocular pressure in a Mongolian population. Ophthalmology 1998;105:969–73. [30] Hansen FK. A clinical study of the normal human central corneal thickness. Acta Ophthalmol (Copenh) 1971;49:82–9. [31] Yee RW, Matsuda M, Schultz RO, Edelhauser HF. Changes in the normal corneal endothelial cellular pattern as a function of age. Curr Eye Res 1985;4:671–8. [32] Orssengo GJ, Pye DC. Determination of the true intraocular pressure and modulus of elasticity of the human cornea in vivo. Bull Math Biol 1999;61(3):551–72. [33] Inoue K, Okugawa K, Oshika T, Amano S. Influence of dorzolamide on corneal endothelium. Jpn J Ophthalmol 2003;47:129–33. [34] Lass JH, Eriksson GL, Osterling L, Simpson CV. Comparison of the corneal effects of latanoprost, fixed combination latanoprost-timolol, and timolol: a double-masked, randomised, one year study. Ophthalmology 2001;108:264–71.