|Home | About | Journals | Submit | Contact Us | Français|
To evaluate the influence of corneal power on circumpapillary retinal nerve fiber layer (cpRNFL) and optic nerve head (ONH) measurements by spectral-domain optical coherence tomography (SD-OCT).
Twenty-five eyes of 25 healthy participants (mean age 23.6±3.6y) were imaged by SD-OCT using horizontal raster scans. Disposable soft contact lenses of different powers (from −11 to +5 diopters including 0 diopter) were worn to induce 2-diopter changes in corneal power. Differences in the cpRNFL and ONH measurements per diopter of change in corneal power were analyzed.
As corneal power increased by 1 diopter, total and quadrant cpRNFL thicknesses, except for the nasal sector, decreased by −-0.19 to −0.32 µm (P<0.01). Furthermore, the disc, cup, and rim areas decreased by −0.017, −0.007, and −0.015 mm2, respectively (P<0.001); the cup and rim volumes decreased by −0.0013 and −0.006 mm3, respectively (P<0.01); and the vertical and horizontal disc diameters decreased by −0.006 and −0.007 mm, respectively (P<0.001).
For more precise OCT imaging, the ocular magnification should be corrected by considering both the axial length and corneal power. However, the effect of corneal power changes on cpRNFL thickness and ONH topography are small when compare with those of the axial length.
Spectral-domain optical coherence tomography (SD-OCT) enables the detection of slight structural changes before visual field deterioration in early glaucoma–. Such changes are difficult to detect by traditional ophthalmoscopy or fundus photography. However, measurements of structures such as the circumpapillary retinal nerve fiber layer (cpRNFL) and optic nerve head (ONH) are influenced by factors including axial length and high myopia independently of the degree of glaucomatous change,–. Therefore, these measurements should be corrected according to the individual's ocular magnification for accuracy.
Traditionally, Littmann's formula and Bennett et al's modification are used to correct for ocular magnification, as follows: t=p×q×s, where t is the actual fundus dimension, p is the magnification factor of the camera of the imaging system, q is the magnification factor of the individual eye, and s is the value obtained from the imaging device. Factor p is a constant in a telecentric system. Factor q can be determined by the following formula: q=0.01306×(axial length −1.82).
Nevertheless, these formulas do not consider the optical properties of the anterior segment, particularly the corneal power, because the position of the second principal point is assumed constant. Researchers have investigated the influence of corneal power on cpRNFL measurements by SD-OCT–, but their findings are not consistent. In addition, previous studies did not analyze the effect on ONH measurements. In this study, we evaluated the influence of corneal power on cpRNFL and ONH measurements by SD-OCT.
This cross sectional study followed the tenets of the Declaration of Helsinki. Written informed consent was obtained from each participant after approval was received from the Ethics Committee of Kitasato University School of Allied Health Science (No.2015-07). UMIN clinical trials registry (http://www.umin.ac.jp/) under unique trial number UMIN000016698 (date of registration: 03/03/2015).
Twenty-five healthy participants (mean age 23.6±3.6y, 3 males) underwent comprehensive ophthalmic examinations, including noncycloplegic refraction testing, visual acuity testing at 5 m using a Landolt ring chart, intraocular pressure and axial length measurements, and slit-lamp and fundus examinations, by a glaucoma specialist (Shoji N). For each participant, the eye with a corrected visual acuity of 20/20 or better, intraocular pressure of 21 mm Hg or lower, and more normal optic disc appearance was included in the study. If both eyes met these inclusion criteria, the eye with lower astigmatism was included.
The cpRNFL thickness and ONH topography were measured by an SD-OCT system (3D OCT-2000, version 8.1.1; Topcon, Tokyo, Japan) using the 3D optic disc horizontal raster scan mode with a 512×128 scan resolution and 6 mm2 scan area. This device operates at a speed of 50 000 A-scans per second and has a depth and lateral resolution of 6 µm and 20 µm or less, respectively. It requires a pupil size of 2.5 mm or larger for imaging. Although the device can correct for ocular magnification on the basis of Littmann's formula ocular magnification was not corrected in this study.
A single expert examiner (Hirasawa K) performed all of the measurements in the selected eyes without cycloplegia. The participants wore 10 differently powered (from -11 to +5 diopters including plano) disposable soft contact lenses (1-day Acuvue, Johnson & Johnson Vision Care, Inc., New Brunswick, NJ, USA) in random order to change the corneal power, which was measured with an auto kerato-refractometer (KR-8100PA, Topcon) before SD-OCT. When the signal strength was unacceptable by over 40 at each contact lens power or when B-scan line images were absent or deviated because of movement, the imaging was repeated up to twice for each imaging. The following parameters were evaluated: total and quadrant cpRNFL thicknesses, centered on the optic disc; disc, cup, and rim areas; cup and rim volumes; vertical and horizontal disc diameters; and image quality.
All data were analyzed using R software (http://www.R-project.org) and G*Power3 version 3.1.7–. The effect size, α error, power (1-β error), and nonsphericity correction were 0.25 (middle), 0.05, 0.95, and 0.12, respectively, and the required sample size was 11 participants for 10 repeated measurements. Using three sets of measurements obtained with plano contact lenses, the repeatability was calculated by the Bland and Altman method– as 2.77×Sw. Sw is the within-subject standard deviation and formula is as follows:
Where is the standard deviation of measurements on each subject, where n is the number of participants. Intraclass correlation coefficients were also calculated. When the confidence limit on either side of the estimate of Sw was set to 0.20, the required sample size was 24 eyes.
The first set of measurements were obtained with plano contact lenses, and data collected without a contact lens were compared by the paired t-test to analyze the effect of contact lens wearing on cpRNFL and ONH measurements. Differences of cpRNFL thickness and ONH parameter with different powers of contact lenses were analyzed by repeated-measures analysis of variance.
In this study, 15 right and 10 left eyes were imaged. Table 1 shows their initial optical characteristics.
As shown in Table 4, the measured cpRNFL thickness in every region except for the nasal sector, ONH parameters, and image quality significantly differed with varying contact lens powers (repeated-measures analysis of variance, P<0.05). The changes in total cpRNFL thickness with 2-diopter induced increases in corneal power are depicted in Figure 1.
The different colored dots and their approximating lines indicate data from individual participants. The crosses and solid line indicate the mean data of all the participants.
Table 5 shows that the total cpRNFL thickness significantly decreased by −0.26 µm (−0.25%, P<0.001) and the quadrant cpRNFL thickness, with the exception of the nasal sector, significantly decreased by −0.19 to −0.32 µm (−0.17% to −0.25%, all P<0.007) as the corneal power increased by 1 diopter. All ONH measurements also significantly decreased with the 1-diopter-induced increases in corneal power (P<0.001). Only the image quality increased (0.2 or 0.36% per diopter) with increasing corneal power (P=0.007).
This study demonstrated good repeatability of the measurements with and without a contact lens. Therefore, contact lens wearing does not introduce bias in SD-OCT imaging. However, image quality reduces with induced decreases in corneal power, in turn affecting assessment of cpRNFL thickness–. The current data might include bias where image quality is concerned.
The total and quadrant cpRNFL thicknesses, except for nasal region, showed up to 0.3 µm decreases (−0.4%), and ONH area measurements were reduced up to 1.1% per diopter induced increase in corneal power. One study showed that the total cpRNFL thickness measured by time-domain OCT does not significantly differ with varying corneal power, whereas another study demonstrated that cpRNFL thickness measured by SD-OCT decreases by approximately 0.5 µm (−0.5%) per diopter induced increase in corneal power–. Positional variation of the second principal point due to changes in corneal power would affect cpRNFL and ONH measurements.
In Littmann's formula modified by Bennett et al, the second principal point is assumed to be located at 1.82 mm from the corneal surface based on Bennett and Rabbetts' schematic eye. However, its position moves backward and forward when the corneal power becomes steeper and flatter, respectively, because the calculation is based on the principal point of the crystalline lens, corneal power, and total ocular power, as follows:
Where Plens′ is the second principal plane of the crystalline lens, Peye′ is the second principal plane of the eye, is the distance from the second principal plane of the crystalline lens to the second principal plane of the eye, nvitreous is the refractive index of the vitreous body, Pcornes′ Plens′ is the distance from the second principal plane of the cornea to the second principal plane of the crystalline lens, Fec is the equivalent power of the cornea, naqueous is the refractive index of the aqueous humor, Feye is the equivalent power of the eye, Acornea is the anterior surface of the cornea, and is the distance from the anterior surface of the cornea to the second principal plane of the eye (second principal point). By substituting the variation value of corneal power in the current study and other parameters based on Bennett and Rabbetts' schematic eye into these formulas, the second principal point position ranges from 2.67 to 1.46. As a result, the q value in Littmann's formula modified by Bennett et al, which expresses the magnification factor of the individual eye, varies by -0.0048 (-1.6%) to 0.0111 (+3.8%) compared to average axial length of 24.39 mm and the second principal point of 1.82 mm. Therefore, the apparent ONH size on a fundus photograph might be slightly decreased with induced increases in corneal power, decreasing cpRNFL and ONH measurements. However, the second principal point position was calculated with approximate value based on the schematic eyes, not actually measured in each participant. Further study is needed using the actual value in each participant.
A slight difference in cpRNFL thickness was noted between the previous (−0.4 to −0.5 µm/diopter)– and the current (−0.2 to −0.3 µm/diopter) studies. This difference can be attributed to the control of accommodative effects by cycloplegic eye drops. Although cycloplegic eye drops were used to control pupil size and accommodation in previous studies–, the cpRNFL and ONH were imaged without cycloplegia in the current study. The anterior pole of the lens moves anteriorly by 0.05 mm/diopter of accommodation, while the posterior pole moves slightly back by 0.01 mm/diopter; thus, the center of the lens moves forward by 0.02 mm/diopter. This means that 0.24 mm of the 1.2 mm range of the second principal point change may be a direct result of the lens anterior shift as a consequence of the accommodation in this study. Further, the position of the second principal point would have varied slightly due to accommodation that occurred when the corneal power was decreased by using the contact lenses with a high negative power.
Previous studies showed that measured cpRNFL thickness without correction for ocular magnification decreases in the range of −1.8 to −4.8 µm as the 1-mm axial length increases,–,,–,,. These slope values can be converted to −0.6 to −1.6 per diopter using a ratio of 1 mm axial length to 3-diopter refractive error based on a three-surface schematic eye. In addition, the measured disc area without correction for ocular magnification becomes smaller by −0.72 mm2 as myopia increases by 1 diopter. Although the results cannot be directly compared because the previous data are based on interindividual comparisons,–,,–,–, they suggest that the influence of corneal power on cpRNFL and ONH measurements is less than that of axial length.
There was no difference in cpRNFL thickness in the nasal region induced by an increase in corneal power. The magnitude of curvature in this region is generally larger than that of the temporal, superior, or inferior region, especially considering the longer axial length of a myopic eye. The cpRNFL thickness was measured by the same scan circle size. When the fundus image is magnified by the induced increase in corneal power, the scan area at the nasal region is smaller than that of the temporal, superior, or inferior region. No difference in cpRNFL thickness at the nasal region could be attributed to the magnitude of curvature of the fundus since the scan circle is centered on the optic disc.
Research on refractive surgeries for myopia such as laser-assisted in situ keratomileusis–, small incision lenticule extraction–, and phakic intraocular lens implantation– has been performed worldwide. Although these procedures could change the position of the second principal point, a previous report indicated that refractive surgery does not affect the measured cpRNFL thickness–. A reason for this finding is that ocular magnification does not change considerably because the cornea is minimally resected. However, careful attention is required for ocular magnification when the corneal resection volume is large.
In summary, induced changes in corneal power lead to decreased cpRNFL and ONH measurements in SD-OCT. For more precise OCT imaging, the ocular magnification should be corrected by considering the individual axial length and second principal point position. However, the conventional magnification correction based on Littmann's formula modified by Bennett et al is adequate for daily clinical imaging because the apparent changes in cpRNFL thickness and ONH topography due to corneal power changes are small when compared with those due to axial length.
Foundation: Supported by a Research Fund at Kitasato University.
Conflicts of Interest: Hirasawa K, None; Shoji N, None.