Hrev_master [page 34] [Eye Reports 2011; 1:e11] The use of ocular coherence tomography in evaluating optic nerve health in eyes with large disc size Guy A. Weiss,1 Gadi Wollstein,2 Lilly Naveh,3 Maya Landoy-Kalev,3 Dan Ga’aton,4 Zvia Burgansky-Eliash3 1Department of Medicine, University at Buffalo, Buffalo, NY; 2University of Pittsburgh Medical Center Eye Center, Eye and Ear Institute, Ophthalmology and Visual Science Research Center, Department of Ophthalmology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA; 3Edith Wolfson Medical Center; 4Rabin Center, Department of Ophthalmology, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel Abstract Large discs are often associated with large cups; in order to exclude glaucomatous cupping a good objective tool is needed. The purpose of this study is to evaluate ocular coherence tomography (OCT) optic nerve head (ONH) parameters as indicators of ocular health in subjects with large discs. Eighty-one eyes of 53 healthy patients were evaluated; 46 eyes had large discs (disc area ≥2.6 mm2) and 35 eyes had regular size discs (disc area <2.6 mm2). All subjects underwent OCT. All ONH parameters were documented, including vertical integrated rim area (VIRA), horizontal integrated rim width (HIRW), rim area, cup area, cup-to-disc (CD) area ratio, horizontal cup to disc ratio (HCDR), vertical cup to disc ratio (VCDR), cup area topography, and cup volume. In addition, OCT retinal nerve fiber layer (RNFL) global mean thickness and four quadrants mean thick- nesses were analyzed. All cup parameters were significantly higher in the large disc group com- pared to the normal disc group. The parameters estimating the rim varied between the groups: in the large disc group VIRA was significantly lower while HIRW was significantly higher, compared to the control group. Rim area was the only parameter with similar values in both groups (1.52±0.24 mm2 and 1.6±0.3 mm2 in the large and regular disc groups, respectively). Correlation analysis revealed significant posi- tive association between disc area and cup parameters in the large disc group. In contrast, in the regular disc group, disc area was positive- ly associated with rim parameters. Rim area might serve as an indicator for ocular health in large discs with large cups. Introduction Ocular coherence tomography (OCT) is a reliable tool for diagnosing and following patients with glaucoma.1-3 Glaucoma is charac- terized by a gradual loss of retinal ganglion cells and thinning of the retinal nerve fiber layer4,5 leading to a typical increase in the cup- to-disc (CD) ratio. Peripapillary circumferen- tial retinal nerve fiber layer (RNFL) scans and radial optic nerve head (ONH) scans have sim- ilar accuracy for glaucoma detection. The best ONH parameters for discriminating between healthy and glaucomatous discs in different studies were: rim area,3 vertical cup to disc ratio (VCDR),1 and CD area ratio (please see Appendix for definitions of cup and disc meas- urement terminology).2 Unfortunately, there is no available normative database for the ONH scans in the commonly used StratusOCT instrument (Carl Zeiss Meditec, Inc. Dublin, CA, USA); therefore it is not possible to assess individual patients’ values. Large discs might have large cups and increased CD area ratio without an accompa- nying thinning in RNFL. Moreover, histological studies in primates and humans demonstrated an increase in the number of nerve fibers with the increase in disc area.6,7 Yet in humans, the larger disc size was associated with a decrease in nerve fiber density per disc area, since the fibers have a larger area through which to cross.7 The effect of large ONH on RNFL thickness is controversial. It has been suggested that the peripapillary RNFL thickness is greater closer to the ONH.8 When measuring RNFL around a large disc with fixed-diameter OCT protocol, the RNFL thickness is overestimated and pathological thinning might be overlooked. Savini, et al. demonstrated that an increase in disc size is associated with an increase in RNFL measurements using a fixed-diameter OCT protocol scan.9 RNFL thickness decrease with increasing scan radii was also demon- strated by Carpineto, et al.10 Another study using RNFL in a constant distance from the ONH rim, and not the standard fixed-diameter scan, found thinner RNFL in larger discs.11 Other studies did not find such a correlation between the circumpapillary RNFL measure- ment placement and disc size.12 In this study we aim to evaluate OCT ONH parameters as indicators of eye health in sub- jects with large discs. Materials and Methods Eighty-one eyes of 54 healthy subjects were evaluated in this retrospective cross- sectional study in two separate cohorts. Forty-six eyes of consecutive eligible subjects with large discs imaged at the Edith Wolfson Medical Center (Holon, Israel) were recruited to the large disc size group. Thirty-five eyes of consecutive eli- gible subjects with normal disc size imaged at the University of Pittsburgh Medical Eye Center (Pittsburgh, PA, USA) were recruited to the control group. The definition of a large disc was established by the disc area measurement obtained by OCT according to previously pub- lished data in healthy subjects.13-15 All discs with area larger than 2.6 mm2 (mean area +1 standard deviation) were considered to be large discs, while the control group had disc areas under 2.6 mm2 This study was approved by the Institutional Review Board (IRB) / Ethics Committee at both institutions and adhered to the Declaration of Helsinki and Health Insurance Portability and Accountability Act (HIPAA) regulations. Informed consent was obtained from all partic- ipants. In both groups, all subjects denied ocular complaints or diseases, had intraocular pres- sure of less than 22 mmHg, normal anterior segments on slit lamp exam, normal discs and maculae, reliable normal visual fields in both eyes, and normal mean RNFL thicknesses in both eyes. Exclusion criteria were refractive error > ± 6 D (spherical equivalent), peripap- illary atrophy, and any eye pathology or previ- ous eye surgery except for uncomplicated cataract extraction. All subjects were scanned after pupillary dilatation with the time-domain OCT, the StratusOCT (software version 4.0; Carl Zeiss Meditec, Inc. Dublin, CA, USA), by experienced operators. OCT measurements of the ONH were generated with the fast optic disc acqui- sition protocol of six total radial scans 6 mm in length in a spoke pattern configuration cen- Eye Reports 2011; volume 1:e11 Correspondence: Zvia Burgansky - Eliash, Department of Ophthalmology, Wolfson Medical Center, P.O. Box 5, Holon 58100, Israel. Tel. +972.3.5028469 - Fax: +972.3.5028133. E-mail: zviaeb@gmail.com Key words: ocular coherence tomography, optic nerve head, large discs. Received for publication: 11 May 2011. Revision received: 5 October 2011. Accepted for publication: 18 October 2011. This work is licensed under a Creative Commons Attribution NonCommercial 3.0 License (CC BY- NC 3.0). ©Copyright G.A. Weiss et al., 2011 Licensee PAGEPress, Italy Eye Reports 2011; 1:e11 doi:10.4081/eye.2011.e11 Non -co mmerc ial us e o nly [Eye Reports 2011; 1:e11] [page 35] tered on the ONH, evenly spaced 30 degrees apart. Each radial scan included 128 A-scans. Qualified scans had signal strength of at least 6. The OCT defined the ONH margin automat- ically as the end of the retinal pigment epithe- lium. The generated image was manually cor- rected in case the OCT did not recognize accu- rately the ONH margin. The rate of images that were manually adjusted was not recorded; pre- vious studies reported manual adjustment of disc margin in 50-60% of cases.9,16 All ONH parameters were documented, including VIRA, HIRW, rim area, cup area, CD area ratio, HCDR, VCDR, cup area topography (using an offset of 150 microns), and cup volume (using an offset of 150 microns). In addition, OCT RNFL global mean thickness and four quadrant mean thicknesses were analyzed. Statistical analysis was performed to deter- mine the differences in mean values of ONH parameters between the two groups, using the general linear model (GLM). The number of eyes was included in the linear model in order to control for the use of both eyes in some patients. In each model, diagnosis and sex were included as fixed factors. Correlation analysis was used to determine the association between rim and cup parameters to disc area, in both groups. Significance was set as a P value of less than 0.05. Results The study population characteristics are summarized in Table 1. There were no age dif- ferences between the two study groups. All subjects were Caucasians. There were more female subjects in the large disc group than in the control group (P=0.03). The average disc area in the large disc group was 3.01 mm2 (range 2.6-3.7) while the average disc area in the control group was 2.05 mm2 (range 1.5-2.5, P<0.01). The various average ONH parameters pro- vided by the OCT in the two groups are shown in Table 2. All cup parameters were significant- ly higher in the large disc group compared to the control group. The parameters estimating the rim varied between the groups: in the large disc group, VIRA was significantly lower while HIRW was significantly higher, compared to the control group. The only parameter that did not vary between the groups was the rim area, with a value of 1.52 ± 0.24 mm2 in the large disc group and 1.6 ± 0.3 mm2 in the control group (P=0.15). In the large disc group, a significant correla- tion was demonstrated between all cup param- eters and disc area (cup area: r=0.75, P<0.01, CD area ratio: r=0.48, P=0.001, HCDR: r=0.34, P=0.02, VCDR: r=0.495, P<0.01, cup area topography: r=0.56, P<0.01, and cup volume: r=0.36, P=0.001, Figure 1). On the other hand, in this group, no significant correlation was demonstrated between any of the rim parame- ters and the disc area (Figure 2). In the regular disc size group, a significant association with disc area was evident only in cup area (r=0.34, P=0.04) and cup volume (r=0.35, P=0.04) but not in the other cup parameters (Figure 3). In this group, 2 of 3 rim parameters were significantly correlated with disc area: HIRW (r=0.57, P<0.01) and rim area r=0.54, P=0.001 (Figure 4). We found a trend of correlation (P=0.07, r=0.3) between the disc area and the mean RNFL in the control group but not in the large disc group or the two groups combined. The quadrant RNFL data did not correlate to the disc area in any of the groups. Discussion Disc size varies significantly in healthy sub- jects with ophthalmoscopically-measured disc area ranging between 0.8 and 6 mm2.17 Subjects with large discs and large cups are often referred to glaucoma clinics with sus- pected glaucoma by optic nerve appearance. It is known that in a large disc, a large cup is also expected.6,18 Nevertheless, a good objective tool is needed to differentiate between physiologi- cal large cups in large discs and glaucomatous large optic nerves. Measuring circumferential RNFL with fixed-diameter OCT protocol may be misleading in these cases because of closer approximation to the ONH edge, which gives artificially higher values.8 In this study we found that while all ONH parameters varied between subjects with regular disc size and those with large disc and large cups, rim area had similar values in both groups. There are three different rim parameters derived from the ONH scan: VIRA, HIRW and rim area, each behaved differently when the two study groups were compared. VIRA esti- mates the total volume of rim tissue, as defined previously. In this study we found a significantly lower value in the large disc group, suggesting the discs in this group were shallower. HIRW estimates the total area of rim tissue, as defined previously. Interestingly, the value of HIRW was significantly higher in the large disc group compared to the controls, implying that the rim in this group is thicker, as was found in histology studies.7,18 In con- trast to the VIRA and HIRW, the rim area parameter, as defined previously, had similar values in both groups. We did find that in the large discs, all 6 parameters describing the cup and CD area ratio are larger, as described by previous histology studies that found larger Article Table 2. Optic nerve head mean parameters. Large disc Regular disc P* (mean ± SD) (mean ± SD) Vertical integrated rim area (VIRA, mm3) 0.25±0.01 0.41±0.18 <0.0001 Horizontal integrated rim width (HIRW, mm2) 1.76±0.16 1.64±0.19 0.003 Rim area (mm2) 1.52±0.24 1.60±0.30 0.15 Cup area (mm2) 1.49±0.37 0.45±0.27 <0.0001 CD area ratio 0.49±0.01 0.21±0.12 <0.0001 Horizontal CD ratio (HCDR) 0.72±0.008 0.46±0.14 <0.0001 Vertical CD ratio (VCDR) 0.66±0.01 0.43±0.13 <0.0001 Cup area topography (mm2) 1.32±0.37 0.76±0.62 <0.0001 Cup volume (mm3) 0.33±0.14 0.009±0.01 <0.0001 The mean value of the ONH parameters in each group and the P value of comparison. ONH, optic nerve head; CD, cup to disc; *Independent- sample two-tailed t-test. Table 1. Demographic and clinical characteristics of the study population. Large disc Regular disc P N eyes (pts) 46 (25) 35 (28) Age (yrs) 55.1±12.6 54.4±19.4 0.86* Female / Male 31/15 15/20 0.03° Disc area (mm2) 3.02±0.26 2.05±0.26 <0.0001* RNFL (µm) 99.09±9.1 96.24±9.73 0.18* Superior RNFL (mm) 121±11.8 117±17.3 0.26* Nasal RNFL (mm) 80±13.2 79±14.5 0.89* Inferior RNFL (mm) 126±16.8 120±16.1 0.11* Temporal RNFL (mm) 70±12.3 69±11.5 0.62* A comparison of patients’ characteristics between the study groups. RNFL, retinal nerve fiber layer; *Independent-sample two-tailed t-test; °Chi-square tests. Non -co mmerc ial us e o nly [page 36] [Eye Reports 2011; 1:e11] cup area in larger discs.6,18 Moreover, in large discs, we found a positive correlation between the cup parameters and disc area, while such association was not found with rim parameters. Thus, when pass- ing a certain threshold of disc size, further enlargement of the disc directly influences the cup size while rim magnitude does not increase. On the other hand, in disc area range under 2.6 mm2, as the disc is getting bigger, rim size increases while most cup dimensions are not influenced. Indeed, a recent large scale study with subjects who had a wide range of disc sizes (average 2.27 mm2) found correla- tions between disc area and the same parame- ters as seen in the regular disc group in our study.14 Rim area was found to be similar in the entire cohort of healthy discs in our study, whether large or normal in size (1.52±0.24 mm2 and 1.6±0.3 mm2, respectively) as was previously demonstrated using the Heidelberg retinal tomography, HRT (Heidelberg Engineering, Inc., Carlsbad, CA, USA).19 This parameter was also found to be a reliable indi- cator of glaucomatous optic neuropathy. The area under the receiver operating curve (AROC) for discrimination between healthy and glaucomatous eye in different publications was 0.92 to 0.97 with average values of 0.84±0.31 mm2 in Wollstein, et al.3 and 0.89±0.34 mm2 in Naithani, et al.13 in the glau- comatous population. This data from the liter- ature further supports the role of rim area parameter as an important indicator for disc health. The lack of correlation between RNFL meas- urements and disc size in the large disc group may have resulted from inhomogeneity of the discs characteristics. Two possible and contra- dicting factors may influence the RNFL thick- ness measurement with a fixed-diameter scan circle in large discs. The first factor is the shorter distance from the ONH rim which increases the thickness,9,10 and the second fac- tor is the thinner overall RNFL thickness in these discs.7,11 As these factors are opposed, the net correlation in a group may vary. One potential limitation of this study is the exclusive inclusion of Caucasian subjects. Subjects from African ancestry present higher prevalence of large discs, when compared with Caucasian patients,14 therefore further investi- gation is warranted in this population. Another possible limitation results from a higher preva- lence of female participants in the large disc group, although we controlled for sex bias in the statistical analysis we conducted. We did not perform sample size calculations, thus the lack of difference in rim area could result from insufficient study cohort, although differences in other ONH parameters were highly signifi- cant. Another limitation is the lack of axial length comparison between the groups, Article Figure 1. Correlation of cup parameters with disc area in the large disc group. In the large disc group, a significant correlation was demonstrated between all cup parameters and disc area. Figure 2. Correlation of rim parameters with disc area in the large disc group. In the large disc group, no significant correlation was demonstrated between any of the rim parame- ters and the disc area. Figure 3. Correlation of cup parameters with disc area in the control group. In the regu- lar disc group, a significant association with disc area was not evident in most cup param- eters except for the cup area and cup volume. Figure 4. Correlation of rim parameters with disc area in the control group. In the regu- lar disc group, 2 of 3 rim parameters were significantly correlated with disc area (HIRW and rim area). Non -co mmerc ial us e o nly [Eye Reports 2011; 1:e11] [page 37] despite exclusion of subjects with high refrac- tive errors. The axial length was found to be inversely correlated to the disc size20 and should be included in future studies. A possi- ble bias is a selection bias of large cups among the large disc cohort, since these subjects were referred to the clinic with suspicious discs, yet this served to answer the clinical question. Our findings may have significant implica- tions on the clinical use of OCT. Rim area might serve as an indicator for disc health in large discs. These findings need to be further explored in subjects with large discs who have glaucoma. Appendix The StratusOCT provides several calculated values. The vertical integrated rim area (VIRA) is a measure of the total volume of the retinal nerve fiber layer within the rim that is calcu- lated by the StratusOCT using the product of the disc circumference and average of 6 indi- vidually-calculated sectional rim areas (in each of the 6 spokes). The horizontal integrat- ed rim width (HIRW) is a measure of the total rim area that is calculated by the StratusOCT using the product of the disc circumference and average of 6 individually-calculated sec- tional rim widths (in each of the 6 spokes) rather than rim areas. The rim area is the area of the cup subtracted from the area of the disc, calculated from a flattened planimetric image of the disc in axis with the pupil. The cup area also is calculated from a flattened planimetric image of the disc in axis with the pupil. The horizontal cup to disc ratio (HCDR) is the ratio of longest horizontal line across the flattened planimetric ring image of the cup calculated from cross-sectional 3-dimentional data along 6 axes (in each of the 6 spokes) to the longest horizontal line across the flattened planimetric image of the disc calculated from cross-sec- tional 3-dimentional data along 6 axes (in each of the 6 spokes), whether or not these two hor- izontal lines overlap. The vertical cup to disc ratio (VCDR) is the ratio of longest vertical line across the flattened planimetric image of the cup calculated from cross-sectional 3- dimentional data along 6 axes (in each of the 6 spokes) to the longest vertical line across the flattened planimetric image of the disc calcu- lated from cross-sectional 3-dimentional data along 6 axes (in each of the 6 spokes), whether or not these two vertical lines overlap. The cup area topography is the calculated cup area based on the 3-dimentional topography of the cup obtained along 6 axes (in each of the 6 spokes) using a fixed standard offset which is often 150 microns. The cup volume is the cal- culated cup area based on the 3-dimentional topography of the cup obtained along 6 axes (in each of the 6 spokes) using a fixed stan- dard offset, which is often 150 microns. References 1. Manassakorn A, Nouri-Mahdavi K, Caprioli J. Comparison of retinal nerve fiber layer thickness and optic disk algo- rithms with optical coherence tomography to detect glaucoma. Am J Ophthalmol 2006;141:105-15. 2. Medeiros FA, Zangwill LM, Bowd C, et al. Use of progressive glaucomatous optic disk change as the reference standard for evaluation of diagnostic tests in glaucoma. Am J Ophthalmol 2005;139:1010-8. 3. Wollstein G, Ishikawa H, Wang J, et al. Comparison of three optical coherence tomography scanning areas for detection of glaucomatous damage. Am J Ophthalmol 2005;139:39-43. 4. Sommer A, Miller NR, Pollack I, et al. The nerve fiber layer in the diagnosis of glau- coma. Arch Ophthalmol 1977;95:2149-56. 5. Harwerth RS, Carter-Dawson L, Shen F, et al. Ganglion cell losses underlying visual field defects from experimental glaucoma. Invest Ophthalmol Vis Sci 1999;40:2242- 50. 6. Quigley HA, Coleman AL, Dormanpease ME. Larger Optic-Nerve Heads Have More Nerve-Fibers in Normal Monkey Eyes. Arch Ophthalmol 1991;109:1441-3. 7. Jonas JB, Schmidt AM, Muller-Bergh JA, et al. Human optic nerve fiber count and optic disc size. Invest Ophthalmol Vis Sci 1992;33:2012-8. 8. Gabriele ML, Ishikawa H, Wollstein G, et al. Peripapillary nerve fiber layer thickness profile determined with high speed, ultra- high resolution optical coherence tomog- raphy high-density scanning. Invest Ophthalmol Vis Sci 2007;48:3154-60. 9. Savini G, Zanini M, Carelli V, et al. Correlation between retinal nerve fibre layer thickness and optic nerve head size: an optical coherence tomography study. Br J Ophthalmol 2005;89:489-92. 10. Carpineto P, Ciancaglini M, Aharrh-Gnama A, et al. Custom measurement of retinal nerve fiber layer thickness using STRA- TUS OCT in normal eyes. Eur J Ophthalmol 2005;15:360-6. 11. Savini G, Barboni P, Carbonelli M, Zanini M. The effect of scan diameter on retinal nerve fiber layer thickness measurement using stratus optic coherence tomography. Arch Ophthalmol 2007;125:901-5. 12. Kaushik S, Pandav SS, Ichhpujani P, Gupta A. Fixed-diameter scan protocol preferable for retinal nerve fibre layer measurement by optical coherence tomography in all sizes of optic discs. Br J Ophthalmol 2009;93:895-900. 13. Naithani P, Sihota R, Sony P, et al. Evaluation of optical coherence tomogra- phy and heidelberg retinal tomography parameters in detecting early and moder- ate glaucoma. Invest Ophthalmol Vis Sci 2007;48:3138-45. 14. Marsh BC, Cantor LB, Wudunn D, et al. Optic Nerve Head (ONH) Topographic Analysis by Stratus OCT in Normal Subjects: Correlation to Disc Size, Age, and Ethnicity. J Glaucoma 2010;19:310-8. 15. Schuman JS, Wollstein G, Farra T, et al. Comparison of optic nerve head measure- ments obtained by optical coherence tomography and confocal scanning laser ophthalmoscopy. Am J Ophthalmol 2003; 135:504-12. 16. Iliev ME, Meyenberg A, Garweg JG. Morphometric assessment of normal, sus- pect and glaucomatous optic discs with Stratus OCT and HRT II. Eye (Lond) 2006;20:1288-99. 17. Jonas JB, Budde WM, Panda-Jonas S. Ophthalmoscopic evaluation of the optic nerve head. Surv Ophthalmol 1999;43:293- 320. 18. Jonas JB, Gusek GC, Naumann GO. Optic disc, cup and neuroretinal rim size, config- uration and correlations in normal eyes. Invest Ophthalmol Vis Sci 1988;29:1151-8. 19. Min KH, Seong GJ, Hong YJ, Kim CY. Optic nerve head topographic measurements and retinal nerve fiber layer thickness in physiologic large cups. Korean J Ophthal- mol 2005;19:189-94. 20. Savini G, Barboni P, Parisi V, Carbonelli M. The influence of axial length on retinal nerve fibre layer thickness and optic-disc size measurements by spectral-domain OCT. Br J Ophthalmol 2011. [Epub ahead of print]. Article Non -co mmerc ial us e o nly