The Relationship Between the Sighting Eye and Functional and Structural Asymmetries in Glaucoma

Abstract

Purpose: To determine the relationship between the sighting eye choice and functional and structural asymmetries in open-angle glaucoma (OAG).

Methods: Two hundred patients with bilateral OAG underwent dominant eye testing using a hole-in-a-card test. The paired eyes were divided into worse and better eyes according to the mean deviation (MD) of visual field (VF) results. Based on the results, patients whose sighting eyes corresponded to eyes with a better VF were designated as group 1, and those whose sighting eyes corresponded to eyes with a worse VF as group 2.

Results: The sighting eye corresponded predominantly to eyes with better VFs (P < 0.001). The MD of the worse eye was significantly lower (P = 0.008), and the intereye difference in MD was significantly larger in group 1 than in group 2 (P < 0.001). In a comparison of regional VF threshold values and corresponding retinal nerve fiber layer thickness (RNFLT), the superior and inferior VF threshold values were consistently more deteriorated in the nonsighting worse eyes in group 1 (P = 0.009 and 0.003), compared with the sighting worse eyes in group 2, whereas the corresponding RNFLT was not significantly different between the groups. Multiple linear regression analysis showed that the sighting eye choice (P = 0.047), higher vertical cup-to disc ratio (P = 0.042), older age (P = 0.044), thinner average RNFLT (P = 0.007), and average ganglion cell-inner plexiform layer thickness of worse eyes (P < 0.001) were significantly associated with the MD of the worse eyes.

Conclusions: The sighting eye choice was associated with functional asymmetry between eyes with glaucoma. The functional deterioration was evident in nonsighting eyes with worse VFs.

The human binocular visual system exhibits a feature unifying two separate monocular visions to achieve better perception of depth information by processing of the binocular disparity information.¹ Sighting dominance differs from hemispheric laterality, because the human retina exhibits semi-decussation. However, like other lateral dominances in the body, the visual pathway does not exhibit perfect equality. There is a tendency to prefer visual input from one eye to the other, which is known as ocular dominance or eyedness.² 

Glaucomatous damage also tends to be asymmetric between eyes.³ The key feature of glaucoma involves functional deterioration of the visual field (VF). In glaucoma, retrograde and anterograde degenerations of retinal ganglion cells (RGCs) originating from the optic nerve head occurs.^4,5 Glaucoma pathology has been examined extensively at the level of the retina, optic nerve head, intracranial optic nerves, lateral geniculate body, primary visual cortex, and cerebral nervous system beyond the visual cortex,^6,7 suggesting manifestation of transneuronal degeneration in patients.^4,8 

In glaucoma, consideration of ocular dominance significantly improves the estimation of binocular VF sensitivity, implying the functional relevance of the sighting eye in glaucomatous eyes.⁹ We hypothesized that the sighting eye choice may be associated with functional asymmetry between eyes with glaucoma. In the present study, we first sought to establish whether eyes with better VFs correspond to sighting eyes using paired eyes of open-angle glaucoma (OAG) patients. Next, we determined the relationship of sighting eye with the functional and structural asymmetries between eyes. Finally, we investigated whether the structure–function relationship differed according to the sighting eye choice. 

Methods

Study Subjects

This study was approved by the Review Board of The Catholic University of Korea. All procedures adhered to the tenets of the Declaration of Helsinki. In this cross-sectional study, patients with bilateral OAG who underwent dominant eye testing at St. Vincent's Hospital, College of Medicine, The Catholic University of Korea, were enrolled in the study from March 1, 2016, to September 2016, using the following inclusion and exclusion criteria. At the initial examination, each patient underwent a review of his/her medical history; measurement of visual acuity; slit-lamp examination; gonioscopy; Goldmann applanation tonometry; dilated stereoscopic examination of the optic disc; measurement of the central corneal thickness (CCT); red-free fundus photography (CF-60UD; Canon, Tokyo, Japan); standard automated perimetry (SAP) using the 24-2 SITA program (Humphrey Visual Field Analyzer; Carl Zeiss Meditec, Dublin, CA, USA); and optical coherence tomography (OCT) (Cirrus OCT; Carl Zeiss Meditec). All patients were required to have a best-corrected visual acuity of 20/30 or better to minimize the effect of media opacity, a spherical equivalent within ±10.0 diopters (D), normal anterior chamber angles on slit-lamp biomicroscopy, open angles on gonioscopy, and results of ≥2 on consecutive VF tests. Glaucoma was defined by the presence of glaucomatous optic neuropathy associated with typical reproducible VF defects as determined by SAP. An abnormal glaucomatous VF change was defined as the consistent presence of a cluster of ≥3 nonedge points on the pattern deviation plot, with a probability of occurring in <5% of the normal population (with one of these points having a probability of occurring in <1% of the normal population, a pattern standard deviation with a probability of <5%, or a glaucoma hemifield test result outside normal limits). Glaucomatous eyes with normal Humphrey test results using the SAP 24-2 test were considered to have preperimetric glaucoma. We excluded patients with neurological or intraocular diseases that could cause VF defects; a history of any retinal disease; a history of ocular trauma or surgery, including trabeculectomy or glaucoma drainage device implantation, with the exception of uncomplicated cataract surgery; other optic nerve diseases except for glaucoma; a history of systemic medication use; or a cerebrovascular event that could affect the VF. Patients with consistently unreliable VF results (defined as >25% false-negative results, >25% false-positive results, or >20% fixation losses) were excluded. Also, patients with eyes with worse VFs that could not be determined due to inconsistent SAP results between each SAP test, were also excluded from the study. Both eyes of each patient were required to meet the criteria to be included in this paired eye study. 

OCT Measurements

All patients underwent imaging by spectral-domain OCT (Cirrus HD-OCT; Carl Zeiss Meditec). An optic-disc scan (optic disc cube, 200 × 200 protocol) and a macular scan (macular cube, 512 × 128 protocol) for the retinal nerve fiber layer thickness (RNFLT) and ganglion cell-inner plexiform layer thickness (GCIPLT) measurements, respectively, were acquired by the same operator on the same day. Only well-focused, well-centered images without eye movement and with signal strengths ≥ 6/10 were selected. In an optic disc cube scan, the measurement of the average peripapillary RNFLT, in each of the four quadrants, in each of the 12 clock-hour sectors, and vertical C/D ratios were used. In the macular cube scan, the average GCIPLT and six sectoral (superotemporal, superior, superonasal, inferonasal, inferior, and inferotemporal) GCIPLT in an elliptical annulus were measured.^10,11 

Structure–Function Relationships

Structure–function relationships were analyzed by comparing the mean deviation (MD) and corresponding RNFLT values measured by SAP and OCT parameters, respectively. The VF total deviation (TD) was evaluated using the decibel (dB) [10 × log (1/Lambert)] scale in 52 points. To calculate the mean TD of each sector, the decibel level in each location of the TD field, was converted to a linear scale before averaging the data within each sector, and then the averaged data were converted back to decibel units. The mean TD of the superior VF was calculated as the mean of VF TD value in 26 points in the superior hemifield, and that of the inferior VF as the mean of 26 test points in the inferior hemifield, excluding the blind spot. Superior or inferior RNFLT reflecting an inferior or superior VF were constructed as suggested by Ferraras et al.¹² Clock-hour segments four and nine were excluded, as previously described.¹³ Thus, the superior RNFLT was defined as the average of measurements in clock-hour segments 10, 11, 12, 1, 2, and 3, and the inferior RNFLT as the average of measurements in clock-hour segments 5, 6, 7, and 8. The mean TD of the central cluster VF was defined as the average of 12 central data points. The superior center VF was defined as the average of the superior 6 test points of the 12 central cluster points, and the inferior center VF was defined as the average of the inferior 6 test points of the 12 central cluster points.¹⁴ The superior hemifield GCIPLT reflecting the inferior center VF was defined as the average of measurements in the superonasal, superior, and superotemporal sectors, and the inferior hemifield GCIPLT reflecting the superior center VF was defined as the average of measurements in the inferonasal, inferior, and inferotemporal sectors, as described by Shin et al.¹⁴ 

Patient Grouping According to Sighting Eye Choice

The hole-in-a-card test was used to determine the sighting eye.¹⁵ First, the patient was asked to hold a card with a hole centered in the middle using both hands straight ahead at arm's length, to view a 6-m target through the hole in the card. Each eye was then occluded alternately by the observer to establish which eye was aligned with the hole and the distant target. The selected eye was considered the sighting eye for the first method. The process was repeated. The second time, the patient slowly moved the card toward the face, without losing alignment with the fixation point, until the hole was over an eye. This was considered the sighting eye for the second method. In all subjects, the selected sighting eye was the same for both methods. The paired eyes were divided into worse and better eyes according to the MDs of the SAP VF test performed within 3 months of the OCT examination.¹⁶ Based on the results of the sighting eye test, patients whose sighting eyes corresponded to eyes with better VFs were designated as group 1, and those whose sighting eyes corresponded to eyes with worse VFs as group 2. 

Data Analyses

The comparisons of baseline characteristics were performed between worse and better VF eyes using paired t-tests. Baseline characteristics were compared between groups 1 and 2 using the Student's t-test for continuous parameters and the χ² test for categorical parameters. We conducted the Student's t-tests to assess the difference of various variables between two groups with Bonferroni correction for multiple comparisons (α = 0.013; four comparisons). To analyze factors associated with the intereye difference, the differences (Δ) between the paired eyes were calculated for the measured parameters. The difference was always calculated as “eye with better VF − eye with worse VF.” In each group, the correlations of structural and functional parameters between worse and better eyes were analyzed for each of the following pairs: MD of worse eyes versus MD of better eyes, MD of worse eyes versus MD difference between eyes, average RNFLT of worse eyes, and average RNFL of better eyes. Next, to analyze the differences in the structure–function relationship between each group, the correlations between the VF threshold values and corresponding RNFLT/GCIPLT were evaluated by linear regression analyses, with VF threshold values as the dependent variable and RNFLT/GCIPLT as the independent variable. For any particular regression model, the degree of correlation between two variables is expressed as Pearson's correlation coefficient, the R value. We assessed the significance of differences between any two correlation coefficients, to compare the correlation coefficients between VF threshold values in group 1 and that in group 2.¹⁷ In addition, intergroup comparisons of regional VF threshold values and the corresponding RNFL/GCIPL parameters were made using the Student's t-test. Finally, to find out factors related to the MD of the worse eye, simple and multiple linear regression analyses were conducted. The dependent variables were MD of the worse eye. Age, average RNFLT of the worse eye, average GCIPLT of the worse eye, baseline intraocular pressure (IOP) of the worse eye, vertical cup-to-disc ratio of the worse eye, and the sighting eye choice of better or worse eyes, which showed differences of borderline significances (P < 0.150) in the simple linear regression analyses, were included as the independent variables in the analyses. To develop the final multiple linear regression model, a backward elimination process was used. The parameters, MD of worse eye and better eye, and MD difference between eyes were heteroscedastic. Therefore, for all regressions involving heteroscedastic parameters, weighted regressions were performed. For statistical analyses, SPSS statistical software for Windows, version 20.0 (IBM, Armonk, NY, USA) was used. A P value less than .05 was considered statistically significant. 

Results

In this study, 400 eyes of 200 patients with OAG were included. For all eyes, the mean age was 57 ± 13.9 years, and the mean MD was −7.3 ± 7.0 dB. In the comparisons between paired eyes with worse and better VFs (Table 1), eyes with worse VFs were associated with a significantly higher baseline IOP and with a higher vertical cup-to-disc ratio and thinner average RNFLT than eyes with better VFs (all, P < 0.05, paired t-test). 

Table 1

 

Demographics and Ocular Characteristics of the Study Population

 

Sighting eyes predominantly corresponded to eyes with better VFs (eyes with better VFs versus eyes with worse VFs: 70.0% vs. 30.0%, respectively; P < 0.001, χ² test). In the comparisons of baseline characteristics between groups (Table 2), the MD of eyes with worse VFs were significantly lower (P = 0.008) and the intereye difference of the MD was significantly larger in group 1 than in group 2 (P < 0.001), while the MD of the eyes with better VFs did not differ between groups (P = 0.931, Student's t-test). The average RNFLT of worse eyes and better eyes were not significantly different between groups (P = 0.469 and 0.103, respectively, Student's t-test). 

Table 2

 

Baseline Characteristics of OAG Patients According to the Sighting Eye Choice

 

As the MD of the worse eyes decreased, the MD of the better eyes decreased, and the intereye difference of the MD increased in both groups (Figs. 1A, 1B). However, the correlation of the MD difference with the worsening of the MD in worse eyes was significantly stronger in group 1 than in group 2 (P = 0.004). The average RNFLT of worse eyes was significantly correlated with those of better eyes in both groups (Fig. 1C). In the scatterplots showing the relationship between average RNFL measurements and the MD of paired eyes with OAG in each group, the distributions of worse eyes was separated from better eyes in group 1 (Fig. 2A), whereas the distributions overlapped in group 2 (Fig. 2B). 

Figure 1

 

Correlations of functional and structural parameters in each group. As the MD of the worse eyes decreased, the MD of the better eyes significantly decreased (A) and the intereye difference of the MD increased in both groups (B). However, the MD of better eyes were less correlated with that of worse eyes in group 1 than in group 2 (P = 0.028). The average RNFLTs of the worse eyes were significantly correlated with those of the better eyes in both groups (C). Weighted regressions were done in (A) and (B). Solid lines and black circles: group 1; dashed lines and white circles: group 2.

Figure 2

 

Scatterplots showing the relationship between average RNFL measurements and the MD of paired eyes with OAG in each group. The distribution of eyes with worse VFs are separated from eyes with better VFs in group 1 (A), whereas the distribution of paired eyes overlapped in group 2 (B).

When the regional VF threshold values and corresponding peripapillary RNFLT/GCIPLT were compared between groups, the VF threshold values of nonsighting worse eyes in group 1 were significantly lower in all three regional areas (P = 0.002, 0.003, and <0.001 in the superior VF, inferior VF, and central VF, respectively, Student's t-test), compared with those of sighting worse eyes in group 2 (Fig. 3 and Supplementary Table). However, the VF threshold values of better eyes did not differ between groups and the intergroup comparisons of corresponding peripapillary RNFLT/GCIPLT were not significantly different, either in better or worse eyes. 

Figure 3

 

Comparisons of regional VF threshold values and the corresponding peripapillary RNFL/ GCIPL parameters in each group. The superior, inferior, and central VF threshold values were consistently deteriorated in worse eyes in group 1, compared with those in group 2 (P = 0.009, 0.003, and <0.001, respectively, shown as *). However, the regional VF threshold values of better eyes did not differ between groups. In addition, the intergroup comparisons of corresponding peripapillary RNFLT/GCIPLT were not significantly different, either in better or worse eyes.

Table 3 and Figure 4 show the structure–function relationships between VF threshold values and corresponding peripapillary RNFLT/GCIPLT. In both groups, the global and regional VF threshold values and corresponding OCT thicknesses were significantly correlated in better and worse eyes, while the correlation between the superior hemifield VF and inferior RNFLT was not significant in dominant-worse eyes in group 2 (P = 0.084). 

Table 3

 

Global and Regional Structure–Function Relationships Between VF Threshold Values and Corresponding Cirrus-HD OCT Thicknesses

 

Figure 4

 

Global and regional structure–function relationships between VF threshold values and corresponding peripapillary RNFLT/ GCIPL. (A, B) Average RNFL versus MD of worse eyes and better eyes, (C, D) inferior RNFLT versus superior VF of worse eyes and better eyes, (E, F) superior RNFLT versus inferior VF of worse eyes and better eyes, and (G, H) average GCIPLT versus central VF of worse eyes and better eyes. In both groups, the global and regional VF and corresponding Cirrus-HD OCT thickness was significantly correlated with each other either in better or worse eyes, whereas the correlation between the superior hemifield VF and inferior peripapillary RNFLT was not significant in worse eyes in group 2 (dominant eyes) (P = 0.084). The degree of correlation was expressed as Pearson's correlation coefficient (R); the significance of differences between two correlation coefficients is given.²¹ *P < 0.05 in linear regression analyses. Solid lines and black circles: group 1; dashed lines and white circles: group 2.

Multiple linear regression analyses showed that the sighting eye choice of better or worse eyes (P = 0.047), older age (P = 0.029), higher vertical cup-to disc ratio (P = 0.042), thinner average RNFLT (P = 0.007) and GCIPLT of worse eyes (P < 0.001) were significantly associated with the deterioration of MD of the worse eye (Table 4). 

Table 4

 

Factors Associated With the MD of the Worse Eye Using Simple and Multiple Linear Regression Analysis

 

Discussion

The functional asymmetry of glaucomatous damage has been largely addressed by the structural asymmetry of damage, such as differences in the cup-to-disc ratio and RNFL thickness between eyes.^18,19 It has been reported that intereye differences of IOP, CCT, and optic disc morphology such as disc tilt and peripapillary atrophy, which are characteristic features associated with myopia, affected the asymmetry of glaucomatous damage between eyes.^16,20 In accordance with previous studies, we found that the MD of the worse eyes were associated with structural parameters such as cup-to disc ratio, average RNFLT, and average GCIPLT (Table 4). 

Interestingly, the sighting eye choice was associated with functional asymmetry between paired eyes with glaucoma (Table 2) and increased VF deterioration in the worse eye after adjusting for structural damage (Table 4). The larger functional asymmetry in group 1 was largely addressed by functional deterioration of nonsighting worse eye (Fig. 3) and the differential structure–function relationships, especially in nonsighting worse eyes (Fig. 4). As shown in representative cases in Figure 5, nonsighting worse eyes appeared to show prominent functional deterioration when the eye does not correspond to the sighting eye. 

Figure 5

 

Representative cases showing characteristics according to the sighting eye choice. (A) Images from a 55-year-old male with bilateral OAG whose sighting eye corresponds to the eye with a better VF (right eye), representing group 1. The nonsighting worse eye (left eye) shows prominent VF deterioration and increased intereye asymmetry (the intereye difference of MD: 13.12 dB). (B) Images from a 62-year-old male with bilateral OAG whose sighting eye corresponds to the eye with a worse VF (left eye), representing group 2. In this case, the intereye asymmetry is relatively small (the intereye difference of MD: 1.24 dB).

There are two possibilities on the association of sighting eye choice and functional asymmetry of glaucomatous damage shown in our study. First, sighting eye choice may have influenced the functional asymmetry of glaucoma. The interaction between the retina and brain is bidirectional. It is well-known that monocular deprivation causes the degradation of corresponding brain cortical cells in young animals, and retinal brain-derived neurotrophic factors play a role in ocular dominance plasticity.^21,22 In glaucoma, the key pathogenesis is the degeneration of RGCs caused by retrograde deprivation of neurotrophic factors.²³ The possibility of brain-derived modulation of chronic glaucomatous damage has been suggested by Sponsel et al.,²⁴ who showed that glaucomatous degeneration is a brain-derived integrated process that attempts to maximize the binocular VF, showing interloculation of focally asymmetric defects. 

A second explanation for our finding is that progressive glaucomatous damage may have affected the sighting eye choice. The ocular dominance is characterized by early onset and the absence of developmental trends.¹ However, there is still a possibility that patients have switched their sighting eyes from worse eyes into better eyes with the disease progression. In the scatterplot showing the relationship between average RNFLT and MD, the distribution of eyes with worse VFs was separated from eyes with better VFs in group 1, whereas the distributions overlapped in group 2 (Fig. 2). Considering that this was a cross-sectional study, the causal relationships between the sighting eye choice and intereye asymmetry of glaucomatous damages cannot be determined in this study. Longitudinal changes in the sighting eye with disease progression needs to be determined to address the causal relationship. 

Global and regional structure–function relationships were less steep in group 2 compared with group 1 (Table 3). Especially, the correlations between superior VF and inferior RNFLT were not significant in eyes with worse VFs in group 2. In our recent report, when comparing the RNFL characteristics between dominant and nondominant eyes, the inferior RNFL was thicker than the superior RNFL in dominant eyes in the normal population.²⁵ A relatively thicker inferior RNFL in sighting eyes with worse VFs (group 2) seems to be associated with the less steep structure–function relationships in corresponding areas in these eyes. In this regard, the sighting eye choice can therefore be used as a possible parameter to enhance an understanding of the individual structure–function relationships. 

Several limitations to the current study need to be acknowledged. In this study, the better and worse eyes were decided by comparing MD values from the SAP test. Several studies on the comparison between paired eyes with OAG have included patients with a predetermined difference of the glaucomatous damage.^16,26 However, in the present study, we tried to investigate the association of the sighting eye choice with the degree of intereye asymmetry. Therefore, we used paired eyes with OAG without artificially defining the differences between eyes. To reduce bias from misclassification of eyes, we excluded patients whose eyes with worse VFs showed disagreement between each VF examination. MD value is one of the most frequently used global indices representing the overall depression and it is used in the studies regarding the intereye comparison in glaucoma patients.^16,26,27 However, the functional glaucomatous damage can be defined by several other ways, such as by area, depth of loss, or location of damages, and these localized VF damage might be inadequately described by the MD measurement. In addition, only the sighting dominance test was used to determine ocular dominance due to its structural and functional bases and the simplicity of the test.^28–30 Further studies using other dominance measurement methods, such as the adaptive staircase and the +1.5D blur test for sensory dominance, are required. 

In summary, the sighting eye choice of better or worse eyes was associated with higher functional asymmetry of paired eyes with OAG. The deterioration of VFs was evident, especially in nonsighting eyes with worse VFs. 

Acknowledgments

The authors thank Jung-Min Bae, Department of Dermatology, the Catholic University of Korea, for his valuable assistance with statistic computing. 

Supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute, funded by the Ministry of Health & Welfare, Republic of Korea (No. HC16C2299). 

Disclosure: J.A. Choi, None; I.-Y. Jung, None; D. Jee, None