Ganglion cell-inner plexiform layer and retinal nerve fiber layer thickness according to myopia and optic disc area: a quantitative and three-dimensional analysis

Subjects

One hundred and sixty-eight healthy subjects, aged between 20 and 32 years, with no other known ocular abnormality and who had undergone health screening, were invited to participate in this study conducted at the Armed Forces Capital Hospital. This study followed the tenets of the Declaration of Helsinki. The Institutional Review Board of the Armed Forces Capital Hospital permitted the review of the subjects’ medical records. The requirement for informed consent by individual patients was waived given the retrospective nature of the study. All data in this study were analyzed anonymously.

All of the subjects underwent a comprehensive ophthalmic examination, including a review of the medical history, measurements of the best corrected visual acuity, manifest refraction, intraocular pressure (IOP) measurement with Goldmann applanation tonometry, gonioscopy, and dilated fundus examination. Refractive error was recorded with an automatic refractometer (RK-F1; Topcon Corporation, Tokyo, Japan). An ocular biometer (Axis II PR; Quantel Medical, Inc., Bozeman, MT, USA) was used to obtain axial length (AL). Photograph of the optic disc and RNFL was obtained with a digital fundus camera (CF-60UVI; Canon, Tokyo, Japan). Confocal scanning laser ophthalmoscopy using the Heidelberg Retina Tomograph II (HRT-II; Heidelberg Engineering, Dossenheim, Germany) was performed to obtain optic disc area. One operator (JHJ) was required to draw a contour line along the disc margin. Heidelberg Eye Explorer (Version 3.1.2), the operating system of the HRT-II, was employed to generate mean topography images and to perform image analysis.

The inclusion criteria were as follows: best-corrected visual acuity of at least 20/40; open anterior chamber angle on gonioscopy; good-quality fundus images; absence of glaucomatous optic neuropathy (GON); absence of RNFL defect according to red-free fundus photography, and reliable normal visual field result. Absence of GON was defined as a cup-to-disc ratio less than 0.6 and an intact neuroretinal rim without optic disc hemorrhages, notches or localized pallor.

The following individuals were excluded: those with a history of ocular trauma or previous ocular surgery, or neurological disease that could have affected the RNFL or GCIPL; those with RNFL defect or disc anomaly; those with any ocular or neurologic disease that could cause visual disturbance; those with a narrow angle on gonioscopic examination; those with clinical features consistent with glaucoma or IOP greater than 21 mmHg in either eye. When both eyes of a candidate subject were eligible for inclusion in the study, one eye was randomly selected.

Optical coherence tomography

After pharmacologic dilation of the pupil, the subject was seated and properly aligned. Each peripapillary scan centered on the optic disc (optic disc 200 x 200 axial scan) and macular scan centered on the fovea (macular 200 x 200 cube scan) was obtained using a Cirrus OCT device with software version 6.0. The OCT was performed by a single experienced examiner. Scans showing a signal strength greater than 6 and good centration on the optic disc were evaluated.

Utilizing the Fast RNFL program, the RNFL thickness was determined at 256 points around a set diameter (3.4 mm) from the center of the optic disc. The obtained values were averaged to yield 12 clock-hour thicknesses, four-quadrant thicknesses, and a global average RNFL thickness measurement. The clock-hour RNFL thicknesses were recorded based on the right-eye orientation. The superior clock hour was 12 o’clock; the others were assigned accordingly, in the clockwise direction for the right eye and counterclockwise for the left.

The GCA algorithm was employed to detect and measure macular GCIPL thickness within a 14.13 mm² elliptical annulus area centered on the fovea. Average, minimum, and sectoral (superonasal, superior, superotemporal, inferotemporal, inferior, and inferonasal) measurements subsequently were analyzed. Additionally, T/N ratio- the ratio of GCIPL thickness in temporal sectors (sum of superotemporal and inferotemporal) to that in nasal sectors (sum of superotemporal and inferotemporal) - were scrutinized.

Data analysis

The refraction data were converted to spherical equivalents (SE), and the subjects were assigned to one of three groups according to their SE values: a low-myopia group (-3.0 diopters < SE < 0.0 diopters), a moderate-myopia group (-6.0 diopters < SE ≤ -3.0 diopters), and a high-myopia group (SE ≤ -6.0 diopters). They were also assigned one of three other groups according to their optic disc area values: a small-disc group (disc area ≤ 2.0 mm²), a medium-disc group (2.0 mm² < disc area ≤ 2.5 mm²), and a large-disc group (disc area > 2.5 mm²). The 360° average, quadrant and clock-hour RNFL thickness values and GCIPL thickness values were compared among the groups using multivariate analysis of covariance (MANCOVA). Significant differences in MANCOVA were followed by post-hoc comparisons.

Linear regression analysis was employed to assess the relationship of the SE and/or optic disc area with RNFL or GCIPL thickness values. A statistical analysis was performed with SPSS (version 21.0, SPSS, Chicago, IL, USA), and P values less than 0.05 were considered statistically significant.

Surface plots in the forms of three-dimensional graphs on x-, y-, and z-axes were constructed using SigmaPlot ® Version 11.0 (Systat Software; Inc, Chicago, IL, USA). The graphs were constructed for each measurement, with axes x = disc area (range: 1.3 – 4.2 mm²), y = SE (range: -0.5 – –14.0 diopters) and z = RNFL thickness or GCIPL thickness. A local smoothing technique was applied to average the values at neighboring points in order to avoid oscillations and spikes in the surface plots. Each of the resultant surface plots indicated the value of RNFL or GCIPL thickness estimated to yield a different disc area at a different SE. As a visual aid to interpretation, graduated shades of color were used to denote the RNFL or GCIPL thickness values (Fig. 3).

Subject characteristics

A total of 168 eyes of 168 subjects were included in the analysis; 92 of these were right eyes. All of the subjects were Korean. The mean subject age was 23.2 ± 2.2 years (range: 20–32 years). The mean SE was -3.58 ± 2.74 diopters (range: -0.50 – -14.00), and the mean axial length was 24.44 ± 1.30 mm (range: 19.88 – 28.07 mm).

OCT measurements according to degree of myopia

The RNFL and GCIPL thickness measurements of the subgroups, as based on the SE values, are presented in Table 1. A MANCOVA showed that the RNFL and GCIPL parameters were statistically significantly different for the subgroups adjusted for age, IOP, central corneal thickness and disc area (F = 5.487, p < 0.001 and F = 2.859, p < 0.001, respectively). The average and quadrant RNFL thicknesses were significantly thinner in high myopia than in low myopia, except for the temporal quadrant (all Ps ≤ 0.003). The RNFL thicknesses of the 1, 2, 5, 6, and 12 o’ clock sectors were significantly thinner in moderate and/or high myopia than in low myopia (all Ps < 0.001). By contrast, moderate and/or high myopia showed significantly thicker RNFL values than did low myopia at the 8, 9, and 10 o’clock sectors (P = 0.001, 0.003, and < 0.001, respectively) (Fig. 1).

	Low myopia (LM) (N = 86)	Moderate myopia (MM) (N = 48)	High myopia (HM) (N = 34)	P Value	Post hoc
Retinal nerve fiber layer
Average	96.85 ± 7.49	94.21 ± 7.11	91.59 ± 6.28	0.003	LM > HM
Temporal quadrant	70.93 ± 8.07	76.21 ± 14.82	82.32 ± 16.28	<0.001	LM < MM/HM
Superior quadrant	122.76 ± 13.57	115.79 ± 19.99	111.82 ± 11.01	0.001	LM > MM/HM
Nasal quadrant	67.09 ± 8.49	64.27 ± 9.59	60.53 ± 8.06	0.001	LM > HM
Inferior quadrant	126.57 ± 14.02	118.08 ± 11.44	111.74 ± 15.13	<0.001	LM > MM/HM
Ganglion cell inner plexiform layer
Average	84.13 ± 4.39	82.23 ± 4.67	78.24 ± 6.21	<0.001	LM/MM > HM
Minimum	81.12 ± 9.75	78.17 ± 12.75	69.12 ± 18.51	<0.001	LM/MM > HM
Superotemporal	83.86 ± 6.52	81.27 ± 7.67	77.62 ± 8.94	0.001	LM > HM
Superior	86.22 ± 6.50	85.27 ± 5.15	78.88 ± 10.44	<0.001	LM/MM > HM
Superonasal	85.51 ± 4.55	83.77 ± 4.12	80.97 ± 7.32	0.001	LM > HM
Inferonasal	84.01 ± 4.84	82.33 ± 4.29	80.38 ± 6.23	0.002	LM > HM
Inferior	84.44 ± 4.95	82.46 ± 4.87	79.15 ± 7.98	<0.001	LM/MM > HM
Inferotemporal	81.38 ± 6.24	78.10 ± 8.46	72.71 ± 8.64	<0.001	LM > MM > HM

P value by multivariate analysis of covariance (MANCOVA) adjusted by age, intraocular pressure, central corneal thikness, optic disc area

Significant values are shown in bold. Bonferroni’s correction for multiple comparisons

The OCT-measured GCIPL thickness measurements decreased with increasing myopia severity (Table 1, Fig. 1). The average, minimum and all-sectors GCIPL were significantly thinner in highly myopic eyes than in moderate and/or low-myopic eyes (all Ps ≤ 0.002). The temporal/nasal ratio (T/N ratio) showed significantly lower values in the highly myopic eyes (p = 0.013), indicating more pronounced GCIPL decreases in the temporal sector.

OCT measurements according to optic disc area

Effects of myopia and optic disc area on OCT measurements

The simple linear regression analysis results revealed average OCT RNFL thickness to be correlated significantly with SE (0.81 μm/diopter, P < 0.001, Fig. 2a), axial length (-1.44 μm/mm, P < 0.001, Fig. 2b), and optic disc area (5.35 μm/mm², P < 0.001, Fig. 2c). Additionally, the linear regression analysis of the OCT GCIPL parameters showed average GCIPL thickness to be in significant correlation with SE (0.84 μm/diopter, P < 0.001, Fig. 2d) and axial length (-1.65 μm/mm, P < 0.001, Fig. 2e). There was, however, no significant correlation between average GCIPL thickness and optic disc area (Fig. 2f).

From the equation, only SE demonstrated significant effects on average GCIPL thickness. Age and disc area did not show significant effects on average GCIPL thickness, but were associated with average RNFL thickness. IOP, axial length and central corneal thickness had no significant effects on RNFL or GCIPL thickness, based on the multiple regression analysis.

Three-dimensional graphs of the OCT RNFL and GCIPL thickness values show the combined influence of myopia and disc size (Fig. 3). Larger optic discs were associated with increasing average RNFL and GCIPL thickness, and the effect of optic disc area was more pronounced in the OCT RNFL parameters than in the OCT GCIPL ones. Also, more-myopic eyes were associated with decreasing average RNFL and GCIPL thickness, and the effect was less pronounced in the OCT RNFL parameters than in the OCT GCIPL ones.