Differences in Factors Associated With Glaucoma Progression With Lower Normal Intraocular Pressure in Superior and Inferior Halves of the Optic Nerve Head

Abstract

Purpose: The purpose of this study was to investigate risk factors for progression in the superior and inferior hemi-visual fields (hemi-VFs) and the corresponding hemi-disc/retinas in eyes with normal tension glaucoma (NTG).

Methods: A 5-year prospective follow-up of 90 patients with NTG with untreated intraocular pressure (IOP) consistently ≤ 15 mm Hg was conducted. The IOP and Humphrey Perimeter measurements and disc/retina stereo-photographs were taken every 3 and 6 months, respectively. Risk factors for progression in the superior and inferior hemi-VFs and in the superior and inferior hemi-disc/retinas were investigated.

Results: The mean total deviation values decreased at −0.50 ± 0.76 and −0.13 ± 0.34 dB/year in the superior and inferior hemi-VFs, respectively (P < 0.001). In the superior hemi-VF, the risk factor for faster progression was greater long-term IOP fluctuation (P = 0.022). In the inferior hemi-VF, the risk factors were disc hemorrhage (DH), greater myopic refraction, body mass index (BMI), and vertical cup-to-disc ratio (v-C/D; P < 0.05). The progression probability was 47.7 ± 6.0 and 17.7 ± 4.7% at 5 years in the superior and inferior hemi-disc/retinas respectively (P < 0.001), and DH was a risk factor for progression in both (P = 0.001).

Conclusions: In NTG eyes, greater BMI, myopia, and v-C/D are characteristic risk factors for faster progression in the superior half of the optic nerve head (ONH), whereas long-term IOP fluctuation is the significant risk factor in the inferior half of the ONH, whereas DH is a risk factor in both.

Translational Relevance: Different risk factors were identified in superior and inferior hemifields in NTG eyes.

Introduction

Glaucoma is the leading cause of irreversible blindness in the world.¹ The disease damages patients’ visual field (VF) sensitivities and eventually affects their visual acuity (VA) and vision-related quality of life. Intraocular pressure (IOP)-reducing treatments can halt the disease's progression, as verified in several randomized controlled trials.² However, a considerable proportion of patients with normal tension glaucoma (NTG) experience disease progression despite lower normal IOP during follow-up.³ The prevalence of NTG is between 0.2% and 3.9% worldwide, and population-based epidemiological studies have revealed a high prevalence (3.3 or 3.6%) of NTG in the Japanese population.⁴ Therefore, it is important to identify the risk factors associated with the progression of NTG, especially in populations with a high prevalence.⁴ 

It has been suggested that the process of glaucomatous damage in the superior and inferior halves of the optic nerve head (ONH) may differ.^5–10 For example, a population-based study in normal participants found the superior ONH rim width to be correlated with IOP and ocular perfusion pressure, whereas the inferior ONH rim width correlated only with IOP.^5,6 Patients with glaucomatous eyes and systemic circulatory disturbances or normal IOP are reportedly more likely to develop inferior hemi-visual field damage (hemi-VFD).^7–9 Patients with glaucoma with a smoking habit may experience faster progression of inferior hemi-VFD but not superior hemi-VFD.¹⁰ 

The current study aimed to investigate whether the influence of ocular and systemic risk factors on glaucoma differs between the superior and inferior hemi-VFs and between the corresponding inferior and superior half disc/retinas (hemi-disc/retina) in NTG eyes with lower normal IOP. These eyes are presumed to be less exposed to IOP-associated insults or the regional variations in vulnerability to IOP. 

Methods

The Japanese Lower Normal Pressure Glaucoma Study was a multicenter prospective cohort study of the natural course of NTG in unmedicated Japanese patients.³ Patients with NTG and IOP consistently ≤ 15 mm Hg without treatment at baseline were followed up for 5 years. IOP and VF were measured every 3 months. No medical treatment was prescribed during the observation period unless glaucoma progression was confirmed. 

The inclusion and exclusion criteria were as described in Table 1. 

Table 1.

 

Inclusion and Exclusion Criteria

 

IOP measurements were carried out every 3 months using a Goldmann applanation tonometer (Haag-Streit Deutschland, Wedel, Germany). The IOP variation during follow-up was calculated as the standard deviation of the IOP values. The VF tests were performed every 3 months with a Humphrey Field Analyzer (HFA; Carl Zeiss Meditec, Dublin, CA) using the Swedish interactive threshold algorithm standard program (24-2). For unreliable VF data, re-examination was performed on the same day or within 2 weeks. Central corneal thickness (CCT) was measured at registration with an A-Scan Ultrasound Pachymeter (SP-2000; Tomey, Nagoya, Japan), a rotating Scheimpflug camera (Pentacam, Oculus, Optikgeräte, Germany), or a specular microscope (NIDEK, Tokyo, Japan). Data from the various instruments were transformed to A-Scan Ultrasound Pachymeter values. Body weight and blood pressure were measured with a DC-250 instrument (Tanita, Tokyo, Japan), a HEM-7511T device (Omuron, Kyoto, Japan), respectively at baseline. Height was measured and the body mass index (BMI; weight [Kg]/height [m]²) was calculated at baseline. Using image analysis software (JGSTK-DiscAnalysis Soft; Topcon, Tokyo, Japan), the β-zone peripapillary atrophy (PPA-β) area/disc area ratio and the vertical cup-to-disc ratio (v-C/D) were determined from the stereo disc/retina and wide-angle photographs taken at study entry.¹¹ The presence of disc hemorrhage (DH) was defined as a history of DH within 6 months before enrollment or the presence of DH during follow-up (confirmed by fundus photographs every 6 months). 

Determination of Disc/Retina Deterioration

Disc/retina deterioration was determined when the expansion of disc excavation, focal marginal rim narrowing, or appearance or expansion of retinal nerve fiber layer defects were detected. The appearance of DH was not included in the criteria. The determination methods have been described elsewhere.³ A disc/retina was considered as satisfying the deterioration criteria when the three Optic Disc Reading Committee members concurred. When only two concurred, the final decision was reached through discussion by all three members. If there was no consensus, the disc/retina was considered to be unchanged. If deterioration was confirmed in 2 consecutive photographs taken at a 6-month interval, the disc/retina in question was then classified as deteriorated at the first date of committee non-concurrence. These analyses were performed without reviewing VF. 

Statistical Analysis

The mean of the total deviation (mTD) values was calculated for the superior and inferior hemifields, and the progression rates were calculated by linear regression against time. Then, for each hemi-VF, the association between the mTD change rate and 13 ocular and systemic variables at baseline (age, BMI, systolic and diastolic blood pressures [SBP and DBP], mean deviation [MD], pattern standard deviation [PSD], CCT, spherical equivalence, v-C/D ratio, PPA-β/disc ratio, and history of DH in the inferior or superior hemi-disc/retina, mean IOP, and long-term IOP fluctuation during follow-up), was investigated using a linear mixed model. The optimal linear model was selected among all possible combinations of predictors (2ⁿ patterns when there are n explanatory variables) based on the second -order bias-corrected Akaike Information Criterion (AICc) index. This provides a more accurate estimation than AIC, especially with a small sample size. Because the degrees of freedom value in a multivariate regression model decreases with many variables, the use of model selection methods is recommended to improve the model fit by removing redundant variables.¹² Adjustment was made using Bonferroni's method. 

The Cox proportional hazards model was then used to identify risk factors for the deterioration of the inferior or superior half disc/retina, from 11 ocular and systemic baseline factors (age, BMI, SBP and DBP, CCT, spherical equivalence, v-C/D ratio, PPA-β/disc ratio and history of DH in the inferior or superior hemi-disc/retina, mean IOP, and long-term IOP fluctuation during follow-up). The optimal model was also selected among all possible combinations of predictors (2ⁿ patterns when there are n explanatory variables) based on the second-order bias AICc. Adjustment was also made using Bonferroni's method. 

All analyses were performed using the statistical programming language “R” (R version 3.1.3; The Foundation for Statistical Computing, Vienna, Austria). 

This study involves human participants and followed all relevant tenets of the Declaration of Helsinki, and the protocol was approved by the institutional review boards of Tokyo University Hospital (#1655) and each of other facilities included in the Japanese Lower Normal Pressure Glaucoma Study (registered ID: UMIN000001041). Written informed consent was obtained from the patients. 

Results

The patients’ demographics are shown in Table 2. The mean patient age was 53.9 ± 9.8 years (standard deviation) and the IOP during the follow-up was 12.3 ± 1.2 mm Hg. The baseline mTDs in the superior and inferior hemi-VFs were −3.8 ± 5.1 and −1.8 ± 2.6 dB (P = 0.006, Wilcoxon signed-rank test) and the mTD change rates in the superior and inferior hemi-VFs were −0.50 ± 0.76 and −0.13 ± 0.34 dB/year, respectively (P < 0.001). The probability of the deterioration of the inferior and superior hemi-disc/retina at 5 years was 47.7 ± 6.0 (standard error) and 17.7± 4.7%, respectively (P < 0.001, log-rank test). 

Table 2.

 

Demographic Data

 

Table 3 shows the association between the variables and faster progression in the superior and inferior hemi-VFs. In the superior hemi-VF, a higher long-term IOP fluctuation was significantly associated with faster progression (P = 0.022). In the inferior hemi-VF, a higher BMI and v-C/D ratio (superior hemi-disc/retina), a history of DH in the superior hemi-disc/retina, and a more negative spherical equivalence were significantly associated with faster progression (P = 0.034, 0.038, 0.017, and 0.040, respectively). 

Table 3.

 

Association Between Ocular and Systemic Variables and the Progression Rate of the Damage in Superior and Inferior Hemi-VFs

 

Table 4 shows the association between the variables and the deterioration of the inferior and superior hemi-disc/retina. A history of DH was significantly associated with progression both in the inferior and superior hemi-disc/retina (P = 0.001). 

Table 4.

 

Association Between Ocular and Systemic Variables and the Progression of Inferior and Superior Disc/Peripapillary Retina Deterioration

 

Discussion

The results showed that a higher long-term IOP fluctuation and the presence of DH were risk factors for progression of the disease in the superior hemi-VF and corresponding inferior hemi-disc/retina pair (referred to as the “inferior half of the ONH” in the following description), whereas the presence of DH, a higher BMI, a greater v-C/D, and higher myopic power were risk factors for progression in the inferior hemi-VF and corresponding superior hemi-disc/retina pair (referred to as the “superior half of the ONH” in the following description). 

A higher long-term IOP fluctuation as a significant risk factor for faster progression in the inferior half of the ONH in the current study participants was compatible with it being a risk factor in the inferior hemi-disc/retina (P = 0.084). This confirms that a higher long-term IOP fluctuation is a risk factor for further deterioration of the inferior half of the ONH, and supports previous reports that the superior hemi-VF was more likely to be affected in open-angle glaucoma (OAG) eyes with higher pressure.^9,14 Further, a higher long-term IOP fluctuation, rather than mean IOP value, during follow-up was a risk factor for the progression of glaucoma in eyes where IOPs were maintained at normal average IOP value or lower.^3,15,16 The current result that a higher long-term IOP fluctuation was not a significant risk factor in the superior half of the ONH may be partly because of the lower power of statistical detection due to lower progression rates in the superior half of the ONH than in the inferior half of the ONH (−0.13 ± 0.34 vs. −0.50 ± 0.76 dB/year, P < 0.001 and 17.7 ± 4.7 vs. 47.7 ± 6.0%, P < 0.001, respectively). Thus, the current result is not evidence that a higher long-term IOP fluctuation has no association with inferior hemi-VF progression. 

A greater BMI and v-C/D and higher myopic power were found to be risk factors for faster progression only in the superior half of the ONH, whereas the presence of DH was a risk factor for faster progression in the superior and halves of the ONH. As discussed above, the progression rate in the superior half of the ONH (inferior hemi-VF) was significantly lower than that in the inferior half of the ONH (superior hemi-VF), and the probability of damage progression in the superior half of the ONH (superior hemi-disc/retina) was significantly lower than that in the inferior hemi-disc/retina (the inferior half of the ONH). Therefore, the potential risk factors could be more sensitively detected in the inferior than in the superior half of the ONH. In other words, the impacts of a higher BMI, v-C/D, and myopic power as a risk for further progression of glaucomatous damage should be more evident in the superior half of the ONH than in the inferior half ONH. 

A higher BMI is associated with a higher IOP,¹⁷ and obesity is a risk factor for OAG.^18,19 However, studies suggesting a protective effect of higher BMI on glaucomatous damage generally prevailed in Western populations,^20–22 whereas reports suggesting an association between a higher BMI and an increased risk for glaucoma have prevailed in non-Western populations.^23–26 A conclusion from the current study that there is a significant contribution of higher BMI to the progression of glaucomatous damage in the superior half of the ONH (inferior hemi-VF), agrees with studies performed in non-Western populations.^23–26 Although it is difficult to explain the discrepancy regarding the influence of BMI, it may be related to differences between ethnicities²⁷ and the fact that the prevalence of NTG is relatively low in Western populations.⁴ The association between BMI and glaucoma is not straightforward. In the South Korean national health and nutrition survey, the fat mass/weight ratio and the fat mass/muscle mass ratio were found to be negatively associated with glaucoma. However, the muscle mass parameter/BMI ratio in male patients and the fat mass/BMI ratio in female patients were positively related to glaucoma.²⁸ The reason that higher BMI was found to be a risk factor for faster progression only in the superior half of the ONH (inferior hemi-VF) is not entirely clear. The effects of BMI require further study. 

The superior half of the ONH may be more vulnerable to circulatory disturbance. A previous population-based study reported that the superior, not inferior, rim width was significantly correlated with ocular perfusion pressure in normal participants.^5,6 Glaucoma eyes with systemic circulatory disturbances were reported to be more likely to have an inferior hemi-VFD (corresponds to the superior half of the ONH),^7,8 and patients with glaucoma who smoked had faster inferior hemi-VFD (corresponds to the superior half of the ONH) progression than in the superior hemi-VFD (corresponds to the inferior half of the ONH).¹⁰ In addition, the association between the ONH blood flow and glaucomatous damage was suggested to be more obvious in the superior half of the ONH (inferior hemi-VF and superior hemi-disc/retina) than in the inferior half of the ONH (superior hemi-VF and inferior disc/retina).²⁹ A greater v-C/D is a well-known risk factor for glaucoma progression, irrespective of the baseline IOP.² The current result supports previous findings that the superior half of the ONH is more vulnerable than the inferior half to glaucomatous insults related to a greater v-C/D. 

There is a consensus that myopia is a risk factor for developing OAG.³⁰ Conversely, myopia is suggested to be protective against glaucoma progression in treated patients with OAG.^31–34 In the current untreated NTG eyes, a higher degree of myopia was significantly associated with faster progression in the superior half of the ONH (corresponds to the inferior hemi-VF), but not the inferior half of the ONH (corresponds to the superior hemi-VF). A retrospective longitudinal study of non-glaucomatous, highly myopic eyes, excluding those with fundus lesions, suggested that scleral myopic changes themselves might be related to progressive VF damage.³⁵ If myopia is a protective factor for OAG in treated and untreated eyes, it cannot be a risk factor for having OAG, as shown in a systematic review and meta-analysis.³⁰ Thus, the current result that a higher myopic power is a risk factor for faster progression in the superior half of the ONH (inferior hemi-VF) of untreated NTG eyes is interesting, because it can resolve the apparent contradiction between the results of the cross-sectional studies³⁰ and those of the treated OAG eyes.^31–34 The effects of BMI and myopia were not detected as a significant contributor in our previous study, where the progression of glaucoma was determined in terms of changes in the whole VF and disk/retina.³ The current trend-type analysis was thought to be more sensitive in detecting risk factors than the event-type analysis of the previous study.³ 

DH is a well-known risk factor for glaucoma progression.³⁶ It was confirmed in the superior and inferior halves of the ONH (superior and inferior hemi-disc/retina and inferior hemi-VF) of the untreated NTG eyes. Although the exact mechanism of DH occurrence remains unclear, many studies have suggested a relationship between DH and systemic vascular disturbances.³⁶ Several studies have shown that the inferior hemi-VF was more likely to be affected by non-IOP-related risk factors, such as diabetes^7,8 or smoking habits.¹⁰ Taken together, the association between DH and VF deterioration might be more straightforward in the superior half of the ONH corresponds to the inferior hemi-VF) than in the inferior half of the ONH (corresponds to the superior hemi-VF). However, a study with a longer follow-up may confirm the effects of DH in the inferior half of the ONH (superior hemi-VF), since it was found to be significantly associated with both superior and inferior halves of the ONH (superior and inferior hemi-disc/retinal) deterioration. 

The current study has several limitations. The first is the lack of the HFA 10-2 test. A more detailed investigation of the central VF area using the HFA 24-2 and 10-2 tests might have revealed a higher progression rate in the VF and additional risk factors. Second, the data of optical coherence tomography (OCT)-measured parameters were not available in the current study. If the trend-type analysis was applied to the time change in the OCT-measured parameters, such as circumpapillary retinal nerve fiber layer thickness, more risk factors might have been identified. Third, the strict inclusion criteria regarding baseline IOP and ocular and systemic medication, limited the sample size. All these limitations could have caused a lack of statistical power to detect possible risk factors. Conversely, the current results are free of the confounding effects of topical or systemic drugs, since the subjects were free of medication. Furthermore, the subjects’ IOP during the follow-up as low as 12.3 mm Hg might have been advantageous in detecting non-IOP-dependent risk factors. 

In conclusion, the effects of ocular and systemic risk factors on glaucoma progression were investigated, focusing on the superior and inferior hemi-VF and corresponding inferior and superior hemi-disc/retina in NTG eyes whose mean IOP during the follow-up was 12.3 mm Hg without medication. The risk of long-term IOP fluctuation was more prominent in the superior hemi-VF, that is, in the inferior half of the ONH, whereas risks of DH, higher myopic power, and greater v-C/D and BMI were greater in the superior half of the ONH. DH was a risk factor for progression in the superior and inferior halves of the ONH. 

Acknowledgments

Supported by the Japan Glaucoma Society, Tokyo, Japan, and Grants 19H01114, 18KK0253, and 26462679 (R.A.) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. The funding organization had no role in the design or conduct of this research. 

Conflict of Interest: The authors have made the following disclosures: Ryo Asaoka received personal fees from Oculus, Reichert, Nidek, and Kowa, outside the submitted work. Makoto Araie received personal fees from Pfizer, Santen Pharmacy, Topcon Medical System, Otsuka, Senju, Aerie, and Kowa, outside the submitted work. In addition, Makoto Araie has a patent JGSTK-DiscAnalysis licensed to Topcon (No fee). Aiko Iwase received personal fees from Carl Zeiss Meditec, Kowa, Otsuka, Pfizer, Santen, Senju, and Novartis, outside the submitted work. In addition, Aiko Iwase has a patent JGSTK-DiscAnalysis licensed to the Topcon Medical System. 

Disclosure: R. Asaoka, None; R. Sakata, None; T. Yoshitomi, None; A. Iwase, None; C. Matsumoto, None; T. Higashide, None; M. Shirakashi, None; M. Aihara, None; K. Sugiyama, None; M. Araie, None