optic disc area and correlation with central corneal thickness, corneal hysteresis and ocular pulse...

Original Article

Optic disc area and correlation with central cornealthickness, corneal hysteresis and ocular pulseamplitude in glaucoma patients and controlsElizabeth Insull MBChB,1 Simon Nicholas FRANZCO,1,2 Ghee Soon Ang MRCOphth,1,2 Ali Poostchi MBChB,1,2

Kenneth Chan MBChB1 and Anthony Wells FRANZCO1,2

1Wellington Eye Clinic, Capital and Coast District Health Board, and 2Capital Eye Specialistst, Wellington, New Zealand

ABSTRACT

Background: To examine the relationships betweenoptic disc area and parameters measured at thecornea; central corneal thickness (CCT), corneal hys-teresis (CH) and ocular pulse amplitude (OPA) inglaucoma subjects and controls.

Methods: In this prospective experimental study,patients underwent measurement of CCT, OPA, CHand optic disc imaging with the Heidelberg RetinaTomograph II (HRT-II). Pearson’s correlation coeffi-cient was calculated to assess the associationsbetween optic disc area and CCT, OPA and CH.

Results: A total of 100 patients, 38 with glaucoma and62 controls were examined. In a univariate analysisof this group, CCT and CH were significantly lowerin glaucoma patients (P = 0.01). CCT was inverselycorrelated with optic disc surface area (Pearson’s cor-relation coefficient r = -0.200; P = 0.05). This inversecorrelation did not achieve statistical significancewhen glaucoma patients and controls were analysedseparately. There was no statistically significant asso-ciation between optic disc area and OPA or CH.

Conclusions: There was an inverse relationshipbetween CCT and optic disc area in this study group.No association was found between optic disc areaand OPA or CH.

Key words: cornea, corneal thickness, optic disc.

INTRODUCTION

Intraocular pressure (IOP) is the most significant riskfactor for primary open angle glaucoma (POAG) andis the only parameter for which treatment has beenshown to reduce POAG incidence and diseaseprogression.1–3 Central corneal thickness (CCT) hasalso been shown to be a strong risk factor for thedevelopment of glaucoma.2 It is well established thatCCT influences IOP measurements with Goldmannapplanation tonometry (GAT), and cannulationstudies have directly demonstrated that IOP mea-surements can be underestimated in thin corneas andoverestimated in thick corneas.4 Those with thincorneas have higher ‘true’ IOP than that obtained onmeasurement with Goldmann applanation tonom-etry, thereby potentially conferring an increased riskof glaucoma.2 Structural characteristics of the corneo-scleral envelope associated with reduced CCT mayalso impair its ability to cope with IOP-inducedstress. If thinner corneas are associated with struc-tural differences in the optic nerve, this may alsohelp to explain the increased glaucoma risk seenwith reduced CCT.

Attention is also being focused on the exact impactof corneal hysteresis (CH) on glaucoma risk. CHdescribes the viscous dampening properties of thecornea.5 A recent study found that low CH was apredictor of visual field progression.6 Another studyfound that CH was lowest in patients with normaltension glaucoma and higher in those with ocularhypertension.7

These observations would be consistent withthe hypothesis that parameters measured at thecornea, such as CCT and CH, may constitute

� Correspondence: Dr Simon Nicholas, 68 Branch Road, Highlands Park, New Plymouth, New Zealand. Email: [email protected]

Received 14 March 2010; accepted 4 June 2010.

Clinical and Experimental Ophthalmology 2010; 38: 839–844 doi: 10.1111/j.1442-9071.2010.02373.x

© 2010 The AuthorsClinical and Experimental Ophthalmology © 2010 Royal Australian and New Zealand College of Ophthalmologists

pressure-independent risk factors for glaucoma,perhaps related to the biomechanical characteristicsof the corneoscleral envelope. At present, there arerelatively fewer data correlating optic disc param-eters with these cornea-measured indices. In 2007Pakravan et al. noted an inverse correlation betweenCCT and optic disc area in a POAG population.8 Thisrelationship was also noted by Cankaya et al. in2008, who investigated 208 healthy eyes.9 Otherstudies have not confirmed such a relationship.10–13

The present study examines the relationship of CCT,OPA and CH to optic disc area in glaucoma patientsas well as controls.

METHODS

One hundred consecutive patients who attended spe-cialist appointments in the preceding 12 months wereprospectively recruited from general ophthalmologyand glaucoma clinics. Participants in the study wereall over 18 years of age with best-corrected visualacuities of at least 6/12. Patients were assigned toeither the ‘glaucoma’ or ‘control’ group based onclinical findings from previous visits. Those in theglaucoma group had glaucomatous optic neuropathyverified by a glaucoma specialist over two or morevisits. The control group consisted of patients withhealthy eyes and glaucoma suspects without evidenceof glaucomatous optic neuropathy or characteristicglaucomatous visual field defects in either eye.

Patients with known or suspected ocular perfu-sion abnormalities were excluded from the study.Patients were also excluded if, within the preceding3 months, changes had been made to their treatmentregimen or if they had undergone either intraocularsurgery or laser trabeculoplasty. Other exclusion cri-teria included traumatic or neovascular glaucoma,corneal disease, previous laser refractive surgery andinflammatory connective disorders.

More than half of the patients in the control groupwere glaucoma suspects referred to the clinic withdysmorphic but non-glaucomatous disc appearances,IOPs less than 20 mmHg and no documented discchanges on subsequent visits. The rest of the controlgroup consisted of patients with signs that mightincrease the risk of developing glaucoma, such asnarrow angles and pseudoexfoliation, but normaloptic discs during specialist review. Four patients inthe control group were recruited from general oph-thalmology clinics after being seen with problemsunrelated to glaucoma.

The right eye of all patients was imaged using theHeidelberg Retina Tomograph II (HRT-II; Heidel-berg Engineering, Dossenheim, Germany). Threescans were averaged and combined into a singlemean topography image using the standard HRT dataacquisition process. Contour lines were drawn by an

experienced masked operator using the built-inHRT-II software for all scans (100 right eyes). Theoptic disc area was calculated from these contourlines using software built into the HRT-II. The repro-ducibility of contour line and disc assessment wasevaluated by having the same operator draw contourlines for 30 scans, delete these and then re-draw thecontour lines on a different day. The differences inarea for each contour line were recorded for each eye.

Central corneal thickness was measured byultrasound pachymetry (Pachmate DGH55 PortablePachymeter, DGH Technologies Inc., Exton, PA,USA). The average of three measurements was used.

The OPA was measured using a Dynamic ContourTonometer (PASCAL Dynamic Contour Tonometer,Zeimer Opthalmic Systems, Allmendstrasse, Switzer-land) that records haemodynamic IOP fluctuations.The difference between the systolic and diastolicIOP corresponds to the OPA, which is expressedin mmHg. The quality score of the measurements, Q,ranges from excellent (1) to poor (4 or 5). Measure-ments were repeated until a Q score of 1 wasobtained.

Corneal hysteresis was measured using an OcularResponse Analyser (ORA, Reichert Corp., Buffalo,NY, USA). It uses a calibrated air puff to deform thecornea into a slight concavity and an optical sensor tomeasure the pressures at which the cornea flattensinwards and outwards as the pressure rises and falls.The difference between the two applanation pres-sures is termed hysteresis.4 Readings were examinedgraphically and those with poor quality applanationsignals (multiple applanation spikes or asymmetricsignals) were discarded. Four good quality measure-ments were recorded from each eye and the resultsaveraged.

Data were collated using Microsoft Excel (Micro-soft Corporation, Redmond, WA, USA) and analysedusing Excel and SAS statistics (SAS institute Inc.,Cary, NC, USA). Patient characteristics were analysedusing c2-tests and two-tailed student t-tests. Paramet-ric data within each group were tested for relation-ships using Pearson’s correlation coefficient, and non-parametric data were investigated using Spearman’srank correlation coefficient. Multivariate analysis wasperformed using the General Linear Model in SAS.

Approval for the study was obtained from the localresearch ethics committee. All involved patientsprovided informed consent for participation in thisstudy. We certify that all applicable institutional andgovernmental regulations concerning ethical use ofhuman volunteers was followed during this research.

RESULTS

One hundred patients were included in the study; 38had a specialist diagnosis of glaucoma and 62 were

840 Insull et al.


controls. All study participants were Caucasians. Themean age of the group was 60.6 years (range 32–80,SD 11.2); there was no significant difference in theage of subjects in the glaucoma and control groups.There were significantly more men (76.3%) in theglaucoma group and the gender balance was almosteven in the control group. Twenty-eight (73.7%) ofpatients in the glaucoma group were on topical IOP-lowering medications.

In this cohort of patients, those with glaucomahad significantly thinner corneas on pachymetrycompared with controls (mean 532 mm and 550 mm,respectively, P = 0.01). The glaucoma group alsohad significantly reduced CH compared withcontrols (mean 8.8 mmHg and 9.6 mmHg, respec-tively, P = 0.01). There was no significant differencebetween the two groups for optic disc area orocular pulse amplitude (P = 0.52 and P = 0.17,respectively). These results are summarized inTable 1.

Reproducibility of the optic disc areas enclosed bythe contour lines was excellent, with an average dif-ference between separate measures on the same eyeof 0.89% (0.014 mm2), the standard deviation being2.26% (0.04 mm2).

In the combined group (n = 100), Pearson’s corre-lation coefficient showed a statistically significantnegative correlation between CCT and optic disc area(r = -0.200; P = 0.046); patients with thinner corneashad larger disc areas (Table 2). Negative correlationswere also found between CH and OPA versus opticdisc area, but these were not statistically significant(P = 0.07 and P = 0.27, respectively). There was no

significant gender difference in any of the parameterstested.

When glaucoma and control groups were analysedseparately, optic disc area was negatively correlatedwith CCT, OPA and CH (Figs 1–3). However, none ofthese results were statistically significant in eithergroup (Table 2).

DISCUSSION

Intraocular pressure is so far the only treatable riskfactor for POAG. It has been established that esti-mated IOPs are influenced by CCT and perhaps

Dis

c A

rea

(mm

2 )

3

2

1

0450

CCT (μm)

500 550 600 650

ControlGlaucoma

Figure 1. Scatterplot of central corneal thickness versus discarea in POAG (r = -0.145 P = 0.36 n = 41) and controls(r = -0.220 P = 0.09 n = 59). CCT, central corneal thickness;POAG, primary open angle glaucoma.

Table 1. Age, optic disc area, CCT, OPA and CH measurements in control and glaucoma groups

Control group(95% confidence interval)

Glaucoma group(95% confidence interval)

P-value

Age (years) 59.8 (39.6–10.0) 61.8 (36.2–87.4) 0.38Optic disc area (mm2) 1.97 (1.86–2.09) 2.03 (1.89–2.16) 0.52CCT (mm) 550 (540–559) 532 (521–543) 0.01OPA (mmHg) 3.1 (2.67–3.43) 2.7 (2.27–3.06) 0.17CH (mmHg) 9.6 (9.22–1.0) 8.8 (9.3) 0.01

CCT, central corneal thickness; CH, corneal hysteresis; OPA, ocular pulse amplitude.

Table 2. Correlation of optic disc area to CCT, OPA and CH

Parameter

Combined groups Control group Glaucoma group

Difference betweencontrol and

glaucoma group

n r P-value n r P-value n r P-value P-value

CCT 100 -0.20 0.05 59 -0.22 0.09 41 -0.15 0.36 0.71OPA 76 -0.13 0.27 44 -0.08 0.62 32 -0.22 0.23 0.58CH 77 -0.21 0.07 45 -0.21 0.17 32 -0.24 0.19 0.94

CCT, central corneal thickness; CH, corneal hysteresis; OPA, ocular pulse amplitude.

Optic disc and cornea correlations 841


CH.14,15 It is not yet clear whether CCT and CH con-stitute pressure-independent risk factors for thedevelopment of glaucoma or whether this risk ismediated through their influence on IOP assess-ment. The aim of the present study was to investi-gate relationships between optic disc area andparameters measured at the cornea (CCT, OPAand CH).

In this study, the baseline optic disc area as mea-sured using the HRT-II was larger in glaucomapatients in comparison with control patients, but thisdifference was not statistically significant. Previousstudies assessing optic disc size as a risk factor forglaucoma have had conflicting results. In the BlueMountains Eye Study, the optic disc size of 3654subjects was measured using stereodisc photo-graphs; eyes with glaucoma (mean disc diameter

1.556 mm) were found to have larger optic discsthan non-glaucomatous eyes (mean disc diameter1.506 mm; P < 0.05) and those with ocular hyperten-sion (mean disc diameter 1.494 mm; P < 0.05).16

Others have also reported larger optic discs in low-pressure glaucoma compared with normal sub-jects.17,18 However, analysis of 438 patients in theOcular Hypertension Study using confocal scanninglaser ophthalmoscopy did not show any associationbetween disc size and development of POAG.19

In 2007 Pakravan et al. observed an inverse rela-tionship between CCT and optic disc area in a POAGpopulation and suggested that the increased glau-coma risk from low CCT may not only be due to IOPunderestimation but also to larger optic disc area.8 Inthe current study, our results showed a significantinverse relationship between optic disc area andCCT in the combined group of glaucoma patientsand controls (r = -0.2, P = 0.046, n = 100) wherebythose with thinner corneas had larger optic discs.However, this inverse relationship was not signifi-cant when evaluated separately for the glaucoma orcontrol groups. In addition, we found no significantdifference in this inverse relationship between theglaucoma and control groups; thus we were unableto determine whether this was an independent riskfactor for glaucoma. Patients with glaucoma had sig-nificantly thinner corneas compared with controls;this result has previously been described and wasnot unexpected.1

The OPA is the difference between the mean sys-tolic and mean diastolic IOPs, but its clinical role inglaucoma management is yet to be fully elucidated.The literature on this is mixed, particularly on how itis affected by CCT and how it varies with glaucomatype and severity. Although the OPA is thought to beunaffected by CCT,20 a separate study noted a signifi-cant negative association with CCT,21 and yet anotherfound a positive correlation.22 Some have demon-strated an association between reduced OPA andglaucoma diagnosis (when compared with normalcontrols)20 as well as moderate to severe glaucoma-tous visual field loss,23 and others have observedincreased OPA in glaucoma diagnosis (when com-pared with normal controls)21 and in less severeglaucoma based on vertical and horizontal cup:discratios.22 It has also been reported that no significantrelationship exists between OPA and glaucomatousvisual field severity.21 Our results did not demon-strate any significant association between OPA andoptic disc area in patients with and without glau-coma, which to the best of our knowledge, has notbeen previously described. Although this findingwas not entirely unexpected, we can at least surmisethat even if the OPA was linked with glaucoma insome way, it is unlikely for optic disc area to be aconfounding factor.

Dis

c A

rea

(mm

2 )

0 1 2 3 4 5 6 7

3

2

1

0

ControlGlaucoma

OPA (mm Hg)

Figure 2. Scatterplot of ocular pulse amplitude versus opticdisc area in POAG (r = -0.217 P = 0.23 n = 32) and controls(r = -0.077 P = 0.62 n = 44). OPA, ocular pulse amplitude; POAG,primary open angle glaucoma.

Dis

c A

rea

(mm

2 )

5 6 7 8 9 10 11 12 13 14

3

2

1

0

ControlGlaucoma

CH (mm Hg)

Figure 3. Scatterplot of corneal hysteresis versus optic discarea in POAG patients (r = -0.240 P = 0.19 n = 32) and controls(r = -0.208 P = 0.17 n = 45). CH, corneal hysteresis; POAG,primary open angle glaucoma.

842 Insull et al.


Our results show significantly reduced CH inglaucoma patients compared with controls, and thiswas in accordance with findings from previousstudies.6,7 CH is a measure of the energy absorptionduring the loading–unloading stress–strain cycle ofviscoelastic materials. It is thought to be a directmeasure of corneal biomechanical properties andthus may more completely describe the contribu-tion of corneal resistance to IOP measurement thanCCT alone. CH may also serve as an indicator ofocular biomechanics, which may be independent ofIOP. Our group has previously published that inthis cohort of patients, CH but not CCT may becorrelated with optic nerve head surface deform-ability during changes in IOP.24 As far as we areaware, the only other study evaluating the relation-ship between CH and optic disc area found no asso-ciation in normal Singaporean schoolchildren.13

Our data parallel this report, as we were alsounable to demonstrate a significant correlationbetween CH and optic disc area for both glaucomaand control patients. This suggests that althoughCH may affect the structural properties of the cor-neoscleral envelope, this is not manifest in the opticdisc size.

There were several limitations to our study. First,the task of outlining the optic nerve head in the HRTsoftware was subjective. We minimized the impact ofthis by using a relatively large number of subjects. Inaddition, our reproducibility was found to be good.The contour lines were drawn separately to elimi-nate the possibility that errors in the automatedcontour line import function could affect the results.Second, there were incomplete data with regards toCH and OPA; these measurements were not taken inthe earliest patients in the study. As can be seen inTable 2 this involved a proportionate number ofpatients in both groups so it is envisaged that anypossible bias that may be introduced from thiswould be minimal.

In conclusion, this study has confirmed that glau-coma patients have thinner corneas and reducedCH. It has also confirmed a correlation betweenreduced CCT and larger optic disc size. Thesestructural findings were true in general and notsimply for those with glaucoma. Further research isrequired to evaluate whether structural changes ofthe corneoscleral envelope in patients with reducedCCT and CH actually do influence the likelihood oflarger disc sizes and glaucoma development andprogression.

ACKNOWLEDGEMENT

The authors would like to acknowledge the supportof the Capital Vision Research Trust.

REFERENCES

1. Nouri-Mahdavi K, Hoffman D, Coleman AL, Liu G.Predictive factors for glaucomatous visual field pro-gression in the advanced glaucoma intervention study.Ophthalmology 2004; 111: 1627–35.

2. Gordon MO, Beiser JA, Brandt JD et al. the ocularhypertension treatment study: baseline factors thatpredict the onset of primary open-angle glaucoma. ArchOphthalmol 2002; 120: 714–20.

3. Leske MC, Heijl A, Hussein M, Bengtsson B, Hyman L,Komaroff E. Factors for glaucoma progression and theeffect of treatment – the early manifest glaucoma trial.Arch Ophthalmol 2003; 121: 48–56.

4. Doughty MJ, Zaman ML. Human corneal thicknessand its impact on intraocular pressure measures:a review and meta-analysis approach. Surv Ophthalmol2000; 44: 367–408.

5. Luce DA. Determining in vivo biomechanical proper-ties of the cornea with an ocular response analyzer.J Cataract Refract Surg 2005; 31: 156–62.

6. Congdon NG, Broman AT, Bandeen-Roche K, GroverD, Quigley HA. Central corneal thickness and cornealhysteresis associated with glaucoma damage. Am JOphthalmol 2006; 141: 868–75.

7. Shah S, Laiquizzaman M, Mantry S, Cunliffe I. Ocularresponse analyser to assess hysteresis and cornealresistance factor in low tension, open angle glaucomaand ocular hypertension. Clin Experiment Ophthalmol2008; 36: 508–13.

8. Pakravan M, Parsa A, Sanagou M, Parsa CF. Centralcorneal thickness and correlation to optic disc size:a potential link for susceptibility to glaucoma. Br JOphthalmol 2007; 91: 26–8.

9. Cankaya AB, Elgin U, Batman A, Acaroglu G. Rela-tionship between central corneal thickness and param-eters of optic nerve head topography in healthysubjects. Eur J Ophthalmol 2008; 18: 32–8.

10. Gunvant P, Porsia L, Watkins RJ, Bayliss-Brown H,Broadway DC. Relationships between central cornealthickness and optic disc topography in eyes with glau-coma, suspicion of glaucoma, or ocular hypertension.Clin Ophthalmol 2008; 2: 591–9.

11. Tomais G, Georgopoulos G, Koutsandrea C, MoschosM. Correlation of central corneal thickness and axiallength to the optic disc and peripapillary atrophyamong healthy individuals, glaucoma and ocularhypertension patients. Clin Ophthalmol 2008; 2: 981–8.

12. Kourkoutas D, Georgopoulos G, Maragos A et al. Newnonlinear multivariable model shows the relationshipbetween central corneal thickness and HRTII topo-graphic parameters in glaucoma patients. Clin Ophthal-mol 2009; 3:313–23.

13. Lim L, Gazzard G, Chan Y-H et al. Corneal biome-chanics, thickness and optic disc morphology inchildren with optic disc tilt. Br J Ophthalmol 2008; 92:1461–6.

14. Touboul D, Roberts C, Kerautret J et al. Correlationsbetween corneal hysteresis, intraocular pressure, andcorneal central pachymetry. J Cataract Refract Surg 2008;34: 616–22.

Optic disc and cornea correlations 843


15. Hager A, Loge K, Schroeder B, Fullhas MO, WiegandW. Effect of central corneal thickness and corneal hys-teresis on tonometry as measured by dynamic contourtonometry, ocular response analyzer, and Goldmanntonometry in glaucomatous eyes. J Glaucoma 2008; 17:361–5.

16. Healey PR, Mitchell P. Optic disk size in open-angleglaucoma: the blue mountains eye study. Am J Ophthal-mol 1999; 128: 515–17.

17. Jonas JB. Size of glaucomatous optic discs. Ger JOphthalmol 1992; 1: 41–4.

18. Tuulonen A, Airaksinen PJ. Optic disc size in exfolia-tive, primary open angle and low tension glaucoma.Arch Ophthalmol 1992; 110: 211–13.

19. Zangwill LM, Weinreb RN, Beiser JA et al. Baselinetopographic optic disc measurements are associatedwith the development of primary open-angle glau-coma: the confocal scanning laser ophthalmoscopyancillary study to the ocular hypertension treatmentstudy. Arch Ophthalmol 2005; 123: 1188–97.

20. Stalmans I, Harris A, Vanbellinghen V, Zeyen T,Siesky B. Ocular pulse amplitude in normal tension

and primary open angle glaucoma. J Glaucoma 2008;17: 403–307.

21. Punjabi OS, Ho HK, Kniestedt C, Bostrom AG,Stamper RL, Lin SC. Intraocular pressure and ocularpulse amplitude comparisons in different types ofglaucoma using dynamic contour tonometry. Curr EyeRes 2006; 31: 851–62.

22. Weizer JS, Asrani S, Stinnett SS, Herndon LW. Theclinical utility of dynamic contour tonometry andocular pulse amplitude. J Glaucoma 2007; 16: 700–3.

23. Vulsteke C, Stalmans I, Fleuws S, Zeyen T. Correlationbetween ocular pulse amplitude measured by dynamiccontour tonometer and visual field defects. Graefes ArchClin Exp Ophthalmol 2008; 246: 559–65.

24. Wells AP, Garway-Heath DF, Poostchi A, Wong T,Chan KC, Sachdev N. Corneal biomechanics and opticnerve head compliance: corneal hysteresis but notcorneal thickness correlates with optic nerve surfacedeformation in glaucoma patients. Invest Ophthalmol VisSci 2008; 49: 3262–8.

844 Insull et al.


optic disc area and correlation with central corneal thickness, corneal hysteresis and ocular pulse...

Documents