Prof Chan Kee Park, Department of Ophthalmology and Visual Science, Seoul St. Mary's Hospital, College of Medicine, The Catholic University of Korea #505 Banpo-Dong, Seocho-Gu, Seoul 137-040, Korea. Email: email@example.com
Background: To find out the relationship between laminar displacement and age between patients with primary open-angle glaucoma and normal tension glaucoma.
Design: Retrospective study conducted at a tertiary university hospital
Participants or Samples: Twenty-six eyes of 26 primary open-angle glaucoma patients and 52 eyes of 52 normal tension glaucoma patients.
Methods: Patients were scanned with a Stratus optical coherence tomography apparatus to measure the retinal nerve fibre layer thickness and to visualize the cross-sectional laminar displacement of 12 clock-hour segments, 30 degrees each. Depth1 was defined as the longest distance between the retinal pigment epithelium and the anterior laminar cribrosa surface, which represents the amount of laminar displacement.
Main Outcome Measure: Partial correlation coefficients adjusted by mean deviation and intraocular pressure between (i) retinal nerve fibre layer thickness and age, and (ii) Depth1 and age.
Results: In the primary open-angle glaucoma group, strong negative correlations (approximately −0.343 ∼ −0.738) were found between Depth1 and age. Eight of 12 clock-hour segments' correlations were significant after Bonferroni correction (α = 0.0021; 24 comparisons). However, no significant correlations were found between Depth1 and age in the normal tension glaucoma group. When the correlation coefficients were compared between the two groups, eight clock-hour segments showed significant differences after Bonferroni correction.
Conclusions: The significantly different correlation between laminar displacement and age between primary open-angle glaucoma and normal tension glaucoma patients may suggest a different role of the lamina cribrosa to the disease.
Glaucoma is an optic neuropathy characterized by specific and progressive injury to the optic nerve and retinal nerve fibre layer (RNFL).1 Previous studies have aimed to find structural differences between primary open-angle glaucoma (POAG) and normal tension glaucoma (NTG), including neuroretinal rim size,2,3 optic disc size4–7 and the presence of peripapillary chorioretinal atrophy.8,9 The majority of these studies targeted global disc indices, particularly those obtained en face, and, thus, it was impossible to obtain detailed values on cross-sectional measures. Moreover, there was no consideration for the effect of age-related changes of the lamina cribrosa in vivo.10 Intraocular pressure (IOP) affects both the RNFL and the lamina cribrosa of the optic nerve head (ONH). The usual consequence in the ONH is a combination of loss of neuroretinal tissue and posterior displacement of the lamina cribrosa.10 This laminar displacement is the result of its resilient nature as well as IOP.10 In other words, the deformational susceptibility of the lamina cribrosa would differ if physical properties such as stiffness and resilience differ between POAG and NTG patients.
To measure the displacement of the lamina cribrosa, we designated a new parameter called Depth1, which was defined as the vertical distance from the retinal pigment epithelium (RPE) to the anterior laminar cribrosa surface. We chose the RPE surface as a reference for Depth1 measurement, as RPE boundaries are used as the anatomic markers for measurement of all features of disc anatomy in ONH analysis by stratus optical coherence tomography (OCT).11,12We designed Depth1 to represent the extent of laminar displacement. The present study aimed to clarify the correlation of RNFL thickness or Depth1 with the age by the diagnosis of patients with POAG and NTG and sought to determine whether differences existed in the correlation pattern between the two groups.
This observational, cross-sectional study included 78 eyes of 78 patients. The medical records at Seoul St. Mary Hospital Glaucoma Clinic from January 2006 to September 2008 were reviewed to identify patients who were newly diagnosed with POAG or NTG for the first time. This study was approved by the Institutional Review Board of Seoul St. Mary Hospital.
All subjects underwent a thorough ophthalmological examination. This consisted of the following: ocular and family history, visual acuity, refraction, IOP measurements, gonioscopy, undilated and dilated slit-lamp examination, disc stereophotography, red-free RNFL photography and OCT. IOPs were measured every week for 6 weeks and visual field examination was performed 6 weeks after their initial visit.
Only those subjects with an open anterior chamber angle, a minimum best corrected visual acuity of 20/25 or better, a spherical equivalent refractive error between −6 and +6 diopters, a clear ocular media with no clinically significant cataract, and reliable measurements of visual field (<20% fixation losses, false-positive and false-negative responses) performed with a Humphrey Field Analyzer (Carl Zeiss Meditec, Dublin, CA, USA) were included. The exclusion criteria included those subjects with a history of laser surgery or intraocular surgery except for cataract surgery, secondary causes of elevated IOP, neurological disorders that may have affected the visual field, and under IOP lowering treatments before examination in this clinic. Patients with advanced glaucoma (mean deviation [MD] < −12) were also excluded from the study.
Patients were defined as having POAG or NTG if they had an abnormal visual field (as described below) and an abnormal ONH or RNFL. The abnormal ONH and RNFL classifications were based on the presence of an optic rim notch or diffuse/generalized loss of optic rim tissue, vertical cup-disc diameter asymmetry, disc haemorrhage, or subjectively ascertained localized defect within the RNFL. In each patient, IOP was measured using Goldmann applanation tonometry. Patients whose IOP was less than 21 mmHg at their initial visit were followed to measure their IOP without medication for 6 weeks. Patients with IOP less than 21 mmHg on repeated measurements were defined as having NTG and those with IOP greater than 21 mmHg were defined as having POAG.
Standard automated perimetry
Visual field examinations were performed with the 24-2 full-threshold (FT) program or 24-2 Swedish Interactive Threshold Algorithm. Visual field loss was defined as significant when the following were present: (i) the Glaucoma Hemifield Test was outside normal limits on two consecutive fields; (ii) a cluster of three or more non-edge points occurred in a location typical for glaucoma, all of which were depressed on the pattern deviation plot at a P < 5% level and one of which was depressed at a P < 1% level on two consecutive fields was present; or (iii) corrected pattern standard deviation (CPSD) was significant at the P < 5% level on two consecutive fields. MD values were adopted for statistical analysis.
The Fast RNFL thickness (3.4) mode of a Stratus OCT apparatus (Carl Zeiss Meditec) was used to measure the thickness of the peripapillary RNFL. Three measurements were performed for each eye. OCT parameters including average RNFL thickness, mean RNFL thickness in each quadrant, and mean RNFL thickness in each of 12 clock-hour, 30-degree segments were obtained. Left eye data were converted into right eye format.
The Scan Profile mode of Stratus OCT was used to measure the cross-sectional laminar displacement of 12 clock-hour, 30 degree segments. The parameters of OCT optic disc scans were defined for the evaluation of the detailed segmental laminar depth (Fig. 1). The yellow pointer in the upper left window was moved to the end of the RPE, and the corresponding A-scan appeared at the lower window (Fig. 1a). The green cursor in the lower window was dragged to the point where the peak corresponded to the reflectivity of the RPE (point a in Fig. 1a). In the upper right window, corresponding decibel and A-scan location were obtained. The A-scan location of the lamina cribrosa surface was measured in the same way (point b in Fig. 1b). Depth1 was defined as the largest difference in A-scan location between point a and b (Fig. 2). Depth1 data were obtained for all 12 clock-hour segment locations. The mean of two measurements were regarded as Depth1 at each clock-hour segment locations. The absolute number in the difference of A-scan location (pixels) was converted into mirometre unit for Depth1 measurement. In the OCT program, a difference of 100 of A-scan location was equivalent with 195 µm.
POAG and NTG patients were compared with respect to age, refraction, IOP, central corneal thickness, MD, RNFL thickness and Depth1 using Mann–Whitney U-test. Age distribution and sex ratio were compared between the two groups using χ2 test. Partial correlation coefficients of RNFL, and Depth1 with age adjusted by MD and IOP were acquired. Fisher's Z transformation was used to compare the correlation coefficients between the two groups. Bonferroni correction for multiple comparisons was also applied.
Descriptive statistics of demographic and ophthalmic measurements are presented in Table 1. This observational, cross-sectional study included 26 eyes of 26 POAG patients and 52 eyes of 52 NTG patients. The mean age at diagnosis, age distribution, sex ratio, refraction, central corneal thickness and MD were not different between the two groups. The mean IOP was significantly higher in the POAG group than in the NTG group (23.5 ± 4.3 vs. 15.4 ± 2.7 mmHg, P < 0.001).
Table 1. Patient characteristics of POAG and NTG patients
Table 2 compares RNFL thickness (µm) and cross-sectional Depth1 (µm) between POAG and NTG patients. Significant differences between the POAG and NTG groups in RNFL thickness were found at 11 o'clock (α = 0.05). After the Bonferroni correction (α = 0.0021; 24 comparisons), there were no significant differences between the two groups in each clock-hour RNFL thickness.
Table 2. Comparison of the retinal nerve fibre layer thickness (µm), and cross-sectional Depth1 (µm) between POAG and NTG patients
Table 3 includes partial correlation between OCT parameters (RNFL thickness and Depth1) and age adjusted by IOP and MD value. In the POAG group, partial correlation coefficients between RNFL thickness at 7 and 8 o'clock and age were significant (α = 0.05). Moreover, partial correlation coefficients between Depth1 at 12, 1, 2, 3, 4, 5, 6, 8, 9, 10 and 11 o'clock and age were significant (α = 0.05). After the Bonferroni correction (α = 0.0021; 24 comparisons), no partial correlations were significant between RNFL thickness and age. Still, partial correlation coefficients between Depth1 at 12, 1, 2, 3, 4, 9, 10 and 11 o'clock and age were significant. In the NTG group, there were no significant partial correlation coefficients between any clock-hour RNFL thickness and age. Also, no significant partial correlation was found between Depth1 and age. Significant differences between the POAG and NTG groups in partial correlation coefficients were found in RNFL thickness at 7, 8, 9 and 10 o'clock, and Depth1 at each clock-hour (α = 0.05). After the Bonferroni correction (α = 0.0021), significant differences were found only in Depth1 at 12, 1, 2, 4, 8, 9, 10 and 11 o'clock. For comparison, representative correlations between Depth1 at 12 o'clock and age in the POAG group and NTG group are shown in Figure 3.
Table 3. Partial correlation between optical coherence tomography parameters (RNFL and Depth1) and age adjusted by IOP and MD value
Fisher's Z transformation. IOP, intraocular pressure; MD, mean deviation; NTG, normal tension glaucoma; POAG, primary open-angle glaucoma; RNFL, retinal nerve fibre layer.
This study was designed with the major objective of determining whether the correlation of the amount of laminar displacement with age differs between POAG and NTG patients. To minimize the effects of disease progression, eyes with advanced glaucoma (MD < −12) were excluded from the study and all correlation coefficients were adjusted by MD and IOP. In the present study, there were no significant differences between the two groups in any clock hour of RNFL thickness and Depth1. However, significant differences were evident in partial correlation coefficients of Depth1 with age between the two groups. In the POAG group, Depth1 showed a strong negative correlation with age but no distinct pattern was evident in the NTG group, and these coefficients were significantly different.
Several investigators have reported the age-related changes of the lamina cribrosa in human eyes. Studies have shown that there is an increase in collagen type I in the cribrosal plates and an increase in the total collagen content of the lamina cribrosa with increasing age.13–15 Kotecha et al.16 reported an increase in human lamina cribrosa thickness with increasing age. These changes may explain a stiffer and less resilient lamina cribrosa structure with older age. Albon et al.17 conducted a study that identified the three-dimensional parameters of pressure induced deformation at the level of the human lamina cribrosa. In the study, the mechanical compliance and the resilience of the human lamina cribrosa decreased with age. Burgoyne et al.10 proposed ONH biomechanics to explain the clinical behavior and increased susceptibility of the aged ONH to glaucomatous damage. They were of the opinion that the aged ONH should demonstrate less deformation for a given magnitude of IOP insult owing to the presence of a stiffer lamina and peripapillary sclera.
With respect to cup depth, cup deepening occurs with most forms of glaucomatous disc damage. Usually it is a combination effect of the loss of neuroretinal tissue and posterior displacement of the lamina cribrosa. In other words, cupping can be divided into prelaminar thinning and laminar displacement.10 If the lamina cribrosa is resilient enough to bear the stress of IOP to a considerable extent, it will deform until the glaucomatous damage occurs. However, as the age of a patient increases, the lamina cribrosa may be less likely to deform posteriorly owing to its increased structural stiffness. This is shown as a negative correlation of Depth1 with age at the onset of disease in our results of POAG patients. Recently, lamina cribrosa was reported to be resistant to displacement in glaucoma patients, and after IOP elevation, usually minimal posterior laminar displacement occurred, and in some cases an anterior laminar displacement occurred.18 However, young patients in the study showed more prominent posterior displacement of the lamina cribrosa. This result also shows that ageing makes the lamina cribrosa more stiff and resistant to IOP changes. However, in NTG, we could not observe the effect of ageing on the displacement of the lamina cribrosa.
The mechanism of laminar displacement and the formation of the cup depth by ageing in glaucoma patients are not known. However, our observation about the different correlation between laminar displacement and ageing may give some information to solve this question. The results indicate that POAG is different from NTG in that it has strong negative correlation of Depth1 with age, that is, Depth1 decreased with age. Furthermore, it was obvious for eight clock-hour measurements after Bonferroni correction. It could be a clinical manifestation of the increasing stiffness and decreasing resilience of the lamina cribrosa with ageing in POAG patient. Unlike POAG, we could not find an apparent correlation of Depth1 with age in NTG. This lack of correlation could be explained in two ways. Firstly, the lamina cribrosa was already stiff enough to resist deformation in the NTG group even in a relatively young patient. Secondly, IOP was not high enough to deform the lamina cribrosa. This different correlation pattern cannot deny the importance of IOP in NTG development but it could imply a different role of the lamina cribrosa to the disease. Our study is different from the previous reports in that quantitative measures in laminar depth were tried and the depth was correlated with age.
In OCT measurements, the anterior surface of the lamina cribrosa was identified as high signal peak on the A scan. As the retinal nerve fibres bend towards the lamina cribrosa at the optic nerve head they become less perpendicular to the incident optical beam. Hence, their reflectivity decreases.12 The intrapapillary RNFL, which changes direction at the disc margin to run parallel to the optical axis of the OCT probe beam, has low reflectivity.19 In a recent study,20 polarization sensitive OCT showed high birefringence in the regions that corresponded to the highly reflective layer beneath the optic disc cup in healthy humans. Inoue et al.21 proposed that the highly reflective region underlying the optic disc cup corresponds to the lamina cribrosa; it is well-known that collagen tissue has high birefringence.
This study has various limitations. Firstly, this study had a retrospective design. The age used for correlation was not the age of disease onset but actually the age of disease diagnosis. Also, the disease progression stage was different among patients. In this case, the laminar displacement cannot represent the initial disease state. We adjusted the correlation by IOP and MD and excluded patients with advanced disease to minimize the disease stage-related confusion. In future studies, we would have a better chance of detecting the relationships of the laminar displacement at all ages and at all levels of IOP if we could include a larger number of patients and account for the effects of age and IOP level at the time of damage. Moreover, including normal controls may allow comparison and differentiation of the lamina depth change in glaucoma with normal ageing. At last, this study used time-domain OCT. With the advent of spectral-domain OCT, the visualization of the lamina cribrosa has been possible. Measurement of the laminar thickness or laminar displacement has been tried with spectral-domain OCTs. Even with time-domain OCT, the anterior surface of the lamina cribrosa can be visualized and measurement of the laminar displacement is possible, as seen in our study. However, it needs more validation whether it is the true measures of the laminar thickness or laminar displacement in both time-domain and spectral-domain OCTs.
In conclusion, correlations of laminar displacement with age were significantly different between the POAG and NTG groups, suggesting that the lamina cribrosa of patients with POAG and NTG may have a different role to the disease. Further serial longitudinal study will be needed to prove these findings in our study.