Introduction

Various methods are available that assess the three-dimensional structure of the ocular fundus, including the topography of the optic disc. Since its introduction in 1989, the Heidelberg Retina Tomograph (HRT) [24] has been employed to quantitatively assess the topography of optic discs in patients with glaucoma using computerized parametric analysis. Primary open-angle glaucoma (POAG) is a disorder that demonstrates typical structural changes in the optic disc along with visual field defects related to an abnormal elevation of intraocular pressure (IOP), while normal-tension glaucoma (NTG) is a type of glaucoma that shares clinical features and mechanisms with POAG, except for the abnormal elevation of IOP. Some studies have noted that evaluations of the optic disc and nerve fiber layer are important for diagnosis of glaucoma, due to the close correlation between visual field loss and optic disc deterioration or retinal ganglion cell atrophy [7, 13, 23], while others have observed that the progression of optic disc damage can precede visual field loss in early glaucoma [7, 14, 17, 23]. Further, several authors have made comparisons between the topographic parameters of optic discs among patients with glaucoma, individuals with ocular hypertension (OH), and normal controls [8, 11, 19, 20, 22]. However, there are few reports of studies that have compared those parameters among POAG, NTG, and OH [8, 20]. In the present study, we measured topographic parameters of optic discs in patients with POAG, NTG, and OH using an HRT in order to determine whether such parameters could be used to differentiate these conditions.

Patients and methods

Our cross-sectional study investigated 17 eyes in 17 patients with POAG, 23 eyes in 23 patients with NTG, and 15 eyes in 15 patients with OH. Those with excessive refractive error of more than +6 diopters or less than –6 diopters, or with any history of surgical treatment, were excluded. IOP and visual field were examined, and patients were classified into three groups by their mean IOP after more than one measurement. Those with POAG were defined as having a glaucomatous visual field defect with an IOP elevation of 21 mmHg or more, while NTG was a glaucomatous visual field defect with IOP less than 21 mmHg and OH was a normal visual field with IOP of 21 mmHg or more. We based our definition of a glaucomatous standard visual field defect on reliable (fewer than 33% fixation losses, false-negative and -positive rate) Humphrey Field Analyzer (30-2 or 24-2) standard visual fields. A glaucomatous visual field defect was defined as that with an MD value equal to –2 dB or less along with one of the following criteria: (a) at least three adjacent test points greater than or equal to 5 dB lower than the age-matched controls and one point greater than 10 dB lower, (b) at least two adjacent test points greater than or equal to 10 dB, or (c) at least three adjacent test points greater than or equal to 5 dB abutting the nasal horizontal median. Those with an MD value less than –10 dB were excluded from the POAG and NTG groups. Patients were enrolled in the study regardless of whether their optic disc showed glaucomatous changes. We examined them using an HRT (Heidelberg Engineering, Heidelberg, Germany; software: FR1-V2.01) during outpatient visits for treatment or observation. Parameters provided by the HRT were measured relative to a standard reference plane defined as 50 μm below the mean height between 350 and 356 deg along a user-drawn contour line placed along the margin of the optic disc (software version 2.01). Three 10-deg field-of-view scans that were clear and centered on the optic disc were obtained for each test eye. A mean topographic image of these three scans was created with HRT software version 2.01 and used for analysis. The optic disc margin was outlined as a contour line on the mean topographic image by a single doctor while viewing color photographs of the fundus. We compared age, gender, refractive error, and topographic parameters among all of the patients. Statistical analysis of the data was performed using analysis of variance (ANOVA). Correlation analysis was carried out along with regression analysis of major HRT parameters for MD value. A level of P <0.05 was considered significant.

Results

All patients met the entry criteria and there were no significant differences in age, gender, and refractive error among the three groups (Table 1). There were no statistically significant differences in MD between POAG and NTG. Mean IOP levels were 22.4±1.4 mmHg in POAG and 22.2±1.7 mmHg in OH, which was not significantly different (P=0.7113). Table 2 shows a summary of the results of the HRT parameter measurements. The mean values of rim area and rim volume were significantly different between the three groups. The OH group showed the largest rim area (mm2) and rim volume (mm3) values (1.43±0.34 and 0.42±0.20, respectively), followed in order by POAG (1.14±0.35 and 0.30±0.13, respectively) and NTG (0.82±0.21 and 0.18±0.10, respectively). On the other hand, the mean values of cup area (mm2) and cup volume (mm3) were significantly greater in the NTG patients (1.46±0.58 and 0.51±0.35, respectively) than in the POAG patients (1.10±0.55 and 0.29±0.21, respectively) and the patients with OH (0.87±0.36 and 0.30±0.22, respectively), though those parameters showed no statistically significant differences between POAG and OH. Further, the mean values of mean RNFL thickness (mm) and RNFL cross-sectional area (mm2) were significantly smaller in NTG eyes (0.16±0.08 and 0.83±0.41, respectively) than in POAG eyes (0.25±0.11 and 1.31±0.62, respectively) and OH eyes (0.27±0.06 and 1.49±0.42, respectively), though they were not significantly different between POAG and OH. The OH group showed the smallest value for cup shape measure (−0.18±0.05), followed by POAG (−0.09±0.06) and NTG (−0.06±0.05). However, there were no statistical differences among the 3 groups for mean cup depth, maximum cup depth, and height variation contour. Table 3 shows the results of regression analysis of the HRT parameters for MD value. Classification (R 2=0.339, P<0.0001) and cup shape measure (R 2=0.312, P<0.0001) showed a stronger correlation with MD value than rim area (R 2=0.164, P=0.0019) and cup disc area ratio (R 2=0.167, P=0.0018), while the other parameters had a weak correlation with MD value. Further, classification (Fig. 1a) and rim area (Fig. 1c) showed a positive correlation with MD value, whereas cup shape measure (Fig. 1b) and cup area (Fig. 1d) had a negative correlation.

Table 1. A summary of the data of the patients. There were no significant differences in age, gender, and refractive error among the three groups (POAG primary open-angle glaucoma, NTG normal-tension glaucoma, OH ocular hypertension)
Full size table
Table 2. A summary of the results of the HRT parameter measurements
Full size table
Table 3. A summary of the results of regression analysis of the HRT parameters for MD value
Full size table
Fig. 1a–d.
figure 1

Results of regression analysis of major HRT parameters for MD value. Classification (a) and cup shape measure (b) each showed a relatively strong correlation with MD values (R 2=0.339, P<0.0001 and R 2=0.312, P<0.0001, respectively). Rim area (c) also showed a positive correlation with MD, while cup area (d) had a negative correlation (R 2=0.164, P=0.0019 and R 2=0.073, P=0.0464, respectively) POAG Primary open-angle glaucoma, NTG normal-tension glaucoma, OH ocular hypertension

Full size image

Discussion

Since the development of the HRT, many reports have demonstrated its advantages for quantitative assessments of optic disc topography during diagnosis and follow-up of glaucoma patients [5, 6, 7, 8, 10, 11, 12, 15, 18, 19, 21, 22]. As in previous reports [8, 20], disc area showed no significant difference among these disorders in the present study; however, we found significant differences among NTG, OH, and POAG patients for a majority of the other parameters. As a result, even though some parameters did not show significant differences among them, the conditions were distinguishable based on our HRT findings. Our most striking finding was that the mean values of rim area and rim volume are significantly different in the present three clinical entities, whereas the mean values of cup area and cup volume are not.

Our results also suggested that patients with NTG showed significantly larger cupping, smaller rims, and thinner retinal nerve fiber layers than patients with POAG or OH. Yamagami et al. [20] found that the rim area was significantly smaller in NTG than in POAG patients in stereoscopic disc photographs. However, a three-dimensional evaluation of the optic disc using an HRT provides more objective information than photographic findings, and the present results demonstrated that not only the mean value of rim area but also that of rim volume was smaller in NTG eyes than POAG eyes. It was also reported that the spatial correlation between focal optic nerve damage and focal cup deepening may suggest the presence of a pathogenetic aspect in both POAG and NTG [9]. We think that OH and POAG could be disorders in the balance of aqueous production and outflow resulting in ocular hypertension, with or without damage of the optic nerve. On the other hand, we believe that NTG, especially low-tension glaucoma, could be a disorder of the optic nerve. In other words, a fragility of the optic nerve resulting in larger cupping and a thinner optic disc neural rim than in other conditions, could be involved in the pathogenesis of NTG.

The mean values of cup area and cup volume did not show significant differences between our POAG and OH patients, which implies that cupping of the optic disc is variable in those with glaucoma, just as in normal subjects [3]. Cupping of the optic disc is thought to have a relationship with neural fiber loss [6]; however, a decrease in rim area has been shown to be more highly correlated with neural fiber loss [1, 2, 4, 23]. We also found that mean cup depth and maximum cup depth showed no significant differences among POAG, NTG, and OH patients. Rim area and cup area could be the most useful parameters to evaluate glaucomatous damage of the optic disc, as rim area (Fig. 1c) showed a positive correlation with MD, whereas cup area (Fig. 1d) had a negative correlation (R 2=0.164, P=0.0019 and R 2=0.073, P=0.0464, respectively). These results are consistent with our clinical findings. As in previous reports [19], our regression analysis results suggested that classification and cup shape measure each have a relatively strong relationship with MD value. Although the R 2 values were not so high, our data also suggested that rim area and rim volume may have a closer relationship with MD value than cup area or cup volume. Thus, we considered that evaluation of the neural rim would be more critical than evaluation of disc cupping for determining visual function.

It has also been reported that a meticulous evaluation of the optic disc is mandatory for management of early glaucoma [23]. Further investigations are required to clarify the relationship between topographic parameters of the optic disc and visual field defects; however, evaluation of certain topographic parameters, especially cup shape measure and classification, is useful to differentiate patients with early glaucoma from those with a normal visual field.