Reproducibility of Macular, Retinal Nerve Fiber Layer, and ONH Measurements by OCT in Rhesus Monkeys: The Beijing Intracranial and Intraocular Pressure (iCOP) Study

Abstract

Purpose.: We evaluated repeatability and reproducibility of measurements of the retinal nerve fiber layer thickness (RNFLT), optic nerve head (ONH), and macular inner retinal layer (MIRL) by RTVue spectral-domain optical coherence tomography (OCT) in normal Rhesus monkeys.

Methods.: The experimental study included 15 adult Rhesus macaque monkeys. RNFLT, ONH parameters (area of disc, cup, and rim; volume of cup and rim; and cup-to-disc ratios), and MIRL thickness were imaged at three separate examinations within one month. Each eye was imaged three times at the first examination, and once at each of the two following examinations. We determined the intra-session and inter-session (inter-visit) within-subject SD (Sw), precision (1.96 × Sw), coefficient of variation (CVw, 100 × Sw/overall mean), and intraclass correlation coefficient (ICC).

Results.: For the intra-session repeatability, ICC values ranged between 0.97 and 1.00 for all ONH parameters, and between 0.80 and 0.98 for all regional RNFLT parameters, with a value of 0.97 for the average RNFLT. For the inter-session (inter-visit) reproducibility, ICC values ranged from 0.93–1.00 for the ONH measurements and from 0.63–0.97 for regional RNFLT parameters, with a value of 0.96 for the average RNFLT. Again, the ICC values were lowest in the temporal inferior and temporal superior regions. The ICC value for the average MIRL thickness was 0.95.

Conclusions.: Repeatability and reproducibility of spectral-domain OCT measurements of the RNFL, ONH, and MIRL were high in normal monkey eyes. Spectral-domain OCT may be suitable to assess changes in follow-up examinations of monkeys with experimental glaucoma.

Introduction

The morphologic diagnosis of glaucomatous optic neuropathy, by morphometrically analyzing the topography of the optic nerve head (ONH), retinal nerve fiber layer thickness (RNFLT), and macular inner retinal layer (MIRL), has been improved markedly by the introduction of optical coherence tomography (OCT) into clinical practice. ^1–5 It raises the question whether the same methodology applied in clinical routine also can be used in laboratory investigations for the detection of glaucomatous changes in the ONH of monkeys with experimentally-induced glaucoma. ^6–9 Since the reproducibility of a technique is of primary importance to address its suitability to detect diseases, ^10–17 and since the reproducibility of OCT measurements in monkeys has not been examined yet to our knowledge, we conducted this study to assess the reproducibility of OCT measurements of the ONH and retinal nerve fiber layer (RNFL) in normal monkeys. 

Methods

The experimental study included 15 adult male Rhesus macaque monkeys (Macaca mulatta) with a mean age of 5.5 ± 0.9 years (range 4.5–7.5 years) and a mean weight of 6.1 ± 0.7 kg (range 5.2–7.5 kg). The study was conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. At examination, all monkeys had not undergone any intervention and had not been involved in any study. Both eyes of the monkey were examined. For all examinations, the animals were anesthetized with an intramuscular injection of ketamine HCL (20 mg/kg) and midazolam (0.2 mg/kg), with repeated injections of ketamine (10 mg/kg) as needed during the examination. The pupils were dilated by topical application of tropicamide (0.5%). A rigid plano contact lens was placed onto the corneal surface during the in vivo imaging. Artificial tears (Refresh Plus, Allergan, Irvine, CA) were used to maintain corneal moisture. 

All eyes were imaged using a spectral-domain OCT device (RTVue-100 Fourier-domain; Optovue Inc., Fremont, CA; software version A6,1,0,4). The basic principles of the instrument have been described in detail previously. ^12–13,15 In brief, the RTVue uses a scanning laser diode to emit a scan beam with a wavelength of 840 ± 10 nm to provide images of ocular microstructures. The RNFL3.45 mode measures the RNFL thickness and ONH parameters along a circle of 3.45 mm in diameter centered on the optic disc. The EMM5 scan covered a scan area with a diameter of 5 mm centered on the fovea. Average, superior, and inferior hemiretinal MIRL data and information from the RNFL3.45 mode of the RNFL thickness parameters and ONH topography were generated automatically. Images with a signal strength index (SSI) less than 40 were excluded from further analysis. All images were acquired by the same well-trained operator (DYY). 

The intra-session repeatability was assessed using the measurements that were repeated three times at the first examination. The inter-session repeatability was examined using the measurements that were obtained at three different visits within one month. 

Statistical analysis was performed using SPSS for Windows, version 19.0 (IBM-SPSS, Chicago, Illinois). The measurements were presented as mean ± SD. The intra-session repeatability was measured with three OCT images obtained during the first visit. We calculated the intra-session inter-subject SD, intra-session precision or repeatability coefficient (calculated as 1.96 × SD), intra-session coefficient of variation (calculated as 100 × SD/overall mean), and intra-session intraclass correlation coefficient (ICC). The inter-session reproducibility was determined with three OCT images obtained during each visit. We then calculated the inter-session inter-subject SD, inter-session precision or repeatability coefficient, inter-session coefficient of variation, and inter-session intraclass correlation coefficient. 

Results

A total of 30 eyes of 15 adult male Rhesus macaque monkeys was enrolled into the study. Both eyes of the monkey were examined. The OCT images of the RNFL displayed a typical double hump pattern, with the peaks located at the temporal superior and temporal inferior regions, and the troughs located at the nasal and temporal regions (see Fig.). This pattern reflects the normal distribution of the RNFL thickness in normal human eyes. ¹⁸  

Figure. 

 

Optical coherence tomographic image of the RNFL of a rhesus monkey, showing the typical double hump configuration (and showing results of 4 time inter-visit).

The data on the intra-session repeatability of the measurements of the ONH parameters, RNFL thickness, and MIRL are presented in Table 1. The ICC values ranged between 0.97 and 1.00 for all ONH parameters. The value of ICC for the average RNFL thickness was 0.97. For regional parameters, the ICC values ranged from 0.80–0.98, with the lowest values in temporal regions (Table 1). For the superior and inferior MIRL, the intra-session coefficients of variations were 1.55% and 0.65%, and the intra-session ICC values were 0.81 and 0.98, respectively; the inter-session coefficients of variation were 2.06% and 1.87%, and the inter-session ICC values were 0.85 and 0.86, respectively (Table 1). 

Table 1. 

 

Intra-Session Repeatability of Spectral-Domain OCT Measurements (RTVue) of the ONH, RNFL, and Macula in 15 Normal Rhesus Monkeys during Three Consecutive Measurements in a Single Visit

Parameters	Overall Mean ± SD	Sw (95% CI)	CVw (95% CI)	Precision (95% CI)	ICC (95% CI)
Optic disc area (mm2)	1.35 ± 0.38	0.01 (0.00–001)	0.47 (0.34–0.59)	0.01 (0.01–0.02)	1.00 (1.00–1.00)
Cup area (mm2)	0.59 ± 0.36	0.04 (0.03–0.05)	11.45 (6.54–16.37)	0.07 (0.05–0.09)	0.99 (0.99–1.00)
Rim area (mm2)	0.76 ± 0.29	0.04 (0.03–0.05)	6.02 (4.07–7.97)	0.07 (0.05–0.09)	0.99 (0.98–1.00)
Rim volume (mm3)	0.06 ± 0.05	0.01 (0.00–0.01)	10.67 (7.49–13.85)	0.01 (0.01–0.01)	0.99 (0.99–1.00)
Nerve head volume (mm3)	0.15 ± 0.09	0.01 (0.01–0.01)	8.13 (6.00–10.26)	0.02 (0.01–0.02)	1.00 (0.99–1.00)
Cup volume (mm3)	0.08 ± 0.08	0.01 (0.00–0.01)	15.92 (10.0–21.79)	0.01 (0.01–0.02)	1.00 (0.99–1.00)
Cup-to-disc area ratio	0.42 ± 0.22	0.03 (0.02–0.04)	11.36 (6.35–16.37)	0.06 (0.04–0.08)	0.99 (0.98–0.99)
Horizontal cup-to-disc diameter ratio	0.83 ± 0.22	0.03 (0.01–0.06)	5.07 (1.69–8.45)	0.07 (0.02–0.11)	0.97 (0.94–0.98)
Vertical cup-to-disc diameter ratio	0.64 ± 0.20	0.03 (0.02–0.03)	5.20 (2.63–7.76)	0.05 (0.03–0.07)	0.99 (0.98–1.00)
Average RNFL thickness (μm)	107.14 ± 6.63	1.60 (1.27–1.93)	1.48 (1.18–1.78)	3.14 (2.48–3.79)	0.97 (0.95–0.99)
Superior hemisphere (μm)	100.40 ± 6.93	1.90 (1.44–2.36)	1.86 (1.43–2.29)	3.72 (2.81–4.63)	0.96 (0.94–0.98)
Inferior hemisphere (μm)	113.88 ± 8.21	2.58 (1.99–3.18)	2.23 (1.74–2.72)	5.06 (3.90–6.22)	0.95 (0.92–0.98)
Temporal (μm)	81.20 ± 7.04	4.29 (3.50–5.09)	5.27 (4.32–6.23)	8.42 (6.86–9.97)	0.84 (0.71–0.92)
Temporal inferior (μm)	85.14 ± 8.74	5.70 (4.29–7.12)	6.55 (5.00–8.09)	11.18 (8.41–13.95)	0.80 (0.63–0.90)
Temporal superior (μm)	77.27 ± 9.42	5.71 (4.23–7.18)	7.44 (5.56–9.32)	11.19 (8.30–14.08)	0.82 (0.67–0.91)
Superior (μm)	123.62 ± 10.32	2.42 (1.87–2.98)	1.92 (1.52–2.33)	4.75 (3.66–5.84)	0.98 (0.96–0.99)
Superior temporal (μm)	133.05 ± 10.81	2.88 (2.05–3.71)	2.12 (1.56–2.68)	5.64 (4.02–7.27)	0.97 (0.94–0.98)
Superior nasal (μm)	114.17 ± 14.96	3.44 (2.48–4.40)	3.02 (2.16–3.88)	6.74 (4.86–8.62)	0.97 (0.95–0.99)
Nasal (μm)	77.53 ± 6.92	2.36 (1.79–2.93)	3.08 (2.35–3.82)	4.62 (3.50–5.74)	0.95 (0.90–0.97)
Nasal superior (μm)	77.09 ± 6.29	2.49 (1.68–3.30)	3.28 (2.21–4.34)	4.89 (3.30–6.48)	0.91 (0.83–0.95)
Nasal inferior (μm)	77.97 ± 8.84	3.09 (2.43–3.75)	4.06 (3.17–4.94)	6.06 (4.77–7.36)	0.95 (0.90–0.97)
Inferior (μm)	146.20 ± 12.08	3.40 (2.53–4.27)	2.31 (1.72–2.91)	6.66 (4.95–8.37)	0.96 (0.93–0.98)
Inferior nasal (μm)	134.72 ± 18.60	4.34 (3.24–5.44)	3.24 (2.40–4.09)	8.51 (6.35–10.67)	0.98 (0.95–0.99)
Inferior temporal (μm)	157.68 ± 14.18	3.79 (2.80–4.78)	2.40 (1.79–3.01)	7.43 (5.49–9.36)	0.96 (0.93–0.98)
Average MIRL (μm)	115.43 ± 6.18	1.16 (0.02–2.29)	0.93 (0.07–1.80)	2.27 (0.04–4.50)	0.91 (0.84–0.95)
Superior MIRL (μm)	113.94 ± 8.19	2.02 (−0.2–4.25)	1.55 (0.01–3.08)	3.97 (−0.39–8.3)	0.81 (0.65–0.90)
Inferior MIRL (μm)	116.88 ± 5.40	0.76 (0.40–1.12)	0.65 (0.34–0.95)	1.49 (0.79–2.19)	0.98 (0.97–0.99)

 Sw, intra-session inter-subject SD; CVw, within-subject coefficient of variation (100 × Sw/overall mean); Precision (repeatability coefficient), 1.96 × Sw.

The data on the inter-session (inter-visit) repeatability of the measurements are shown in Table 2. Again, a relatively high ICC value also was found for all ONH measurements, ranging between 0.93 and 1.00. The ICC value for the average RNFL thickness was 0.96. For regional parameters, the ICC value ranged from 0.63–0.97, again with the temporal regions having the lowest ICC values of 0.84 and 0.63. 

Table 2. 

 

Inter-Session (Inter-Visit) Reproducibility: Measurements Obtained from Three Separate Visits

Parameters	Over Mean ± SD (μm)	Sw (95% CI)	CVw (95% CI)	Precision (95% CI)	ICC (95% CI)
Optic disc area (mm2)	1.35 ± 0.38	0.01 (0.01–0.01)	0.57 (0.42–0.72)	0.02 (0.01–0.02)	1.00 (1.00–1.00)
Cup area (mm2)	0.60 ± 0.36	0.05 (0.04–0.07)	14.95 (9.3–20.59)	0.10 (0.07–0.14)	0.99 (0.98–0.99)
Rim area (mm2)	0.75 ± 0.27	0.06 (0.04–0.07)	8.49 (5.87–11.11)	0.11 (0.07–0.14)	0.98 (0.96–0.99)
Rim volume (mm3)	0.06 ± 0.04	0.01 (0.01–0.01)	14.85 (10.5–19.2)	0.01 (0.01–0.02)	0.99 (0.97–0.99)
Nerve head volume (mm3)	0.15 ± 0.08	0.02 (0.01–0.02)	12.00 (8.29–15.7)	0.03 (0.02–0.04)	0.98 (0.97–0.99)
Cup volume (mm3)	0.08 ± 0.08	0.01 (0.01–0.01)	24.02 (15.2–32.8)	0.02 (0.01–0.02)	0.99 (0.99–1.00)
Cup-to-disc area ratio	0.42 ± 0.22	0.05 (0.03–0.06)	15.53 (9.8–21.24)	0.09 (0.06–0.12)	0.97 (0.95–0.99)
Horizontal cup-to-disc diameter ratio	0.81 ± 0.22	0.06 (0.02–0.09)	7.68 (3.23–12.14)	0.11 (0.05–0.17)	0.93 (0.87–0.96)
Vertical cup-to-disc diameter ratio	0.65 ± 0.18	0.04 (0.03–0.06)	7.40 (5.21–9.59)	0.09 (0.06–0.11)	0.97 (0.94–0.98)
Average RNFL thickness (μm)	107.06 ± 7.64	1.94 (1.27–2.61)	1.86 (1.17–2.56)	3.80 (2.48–5.12)	0.96 (0.93–0.98)
Superior hemisphere (μm)	100.50 ± 6.78	2.91 (2.13–3.69)	2.90 (2.13–3.68)	5.71 (4.18–7.24)	0.91 (0.83–0.95)
Inferior hemisphere (μm)	113.85 ± 9.83	2.75 (1.86–3.65)	2.50 (1.63–3.37)	5.40 (3.65–7.15)	0.95 (0.92–0.98)
Temporal (μm)	81.45 ± 6.09	4.88 (3.84–5.92)	6.02 (4.68–7.36)	9.57 (7.53–11.60)	0.71 (0.47–0.85)
Temporal inferior (μm)	84.95 ± 9.06	5.38 (4.19–6.56)	6.26 (4.95–7.56)	10.54 (8.22–12.86)	0.84 (0.71–0.92)
Temporal superior (μm)	77.82 ± 6.96	6.46 (5.14–7.78)	8.34 (6.67–1.00)	12.67 (10.08–15.25)	0.63 (0.31–0.81)
Superior (μm)	123.59 ± 9.98	3.95 (2.91–4.98)	3.22 (2.38–4.06)	7.73 (5.71–9.76)	0.92 (0.86–0.96)
Superior temporal (μm)	132.18 ± 9.50	5.33 (3.84–6.82)	4.03 (2.91–5.14)	10.44 (7.52–13.36)	0.83 (0.70–0.92)
Superior nasal (μm)	114.99 ± 15.42	4.26 (3.35–5.17)	3.76 (2.94–4.59)	8.35 (6.57–10.12)	0.97 (0.94–0.98)
Nasal (μm)	77.52 ± 6.75	3.14 (2.49–3.79)	4.12 (3.19–5.06)	6.15 (4.87–7.43)	0.90 (0.82–0.95)
Nasal superior (μm)	76.99 ± 6.06	3.73 (3.03–4.42)	4.92 (3.95–5.89)	7.30 (5.94–8.66)	0.84 (0.71–0.92)
Nasal inferior (μm)	78.04 ± 8.82	3.71 (2.95–4.47)	4.90 (3.78–6.01)	7.28 (5.79–8.76)	0.92 (0.86–0.96)
Inferior (μm)	145.69 ± 14.05	4.02 (3.02–5.02)	2.86 (2.04–3.68)	7.88 (5.92–9.83)	0.96 (0.93–0.98)
Inferior nasal (μm)	133.94 ± 19.02	5.95 (5.06–6.85)	4.53 (3.79–5.27)	11.67 (9.91–13.42)	0.97 (0.94–0.98)
Inferior temporal (μm)	157.45 ± 15.74	4.94 (3.57–6.31)	3.30 (2.28–4.31)	9.68 (7.00–12.36)	0.95 (0.91–0.97)
Average MIRL (μm)	114.46 ± 6.53	2.09 (0.71–3.47)	1.90 (0.51–3.29)	4.10 (1.40–6.80)	0.87 (0.75–0.93)
Superior MIRL (μm)	112.59 ± 6.67	2.29 (0.79–3.79)	2.06 (0.66–3.46)	4.50 (1.56–7.44)	0.85 (0.72–0.92)
Inferior MIRL (μm)	116.23 ± 6.69	2.05 (0.60–3.51)	1.87 (0.38–3.36)	4.03 (1.17–6.88)	0.86 (0.74–0.93)

For superior and inferior MIRL, the intra-session coefficients of variations were 1.55% and 0.65%, and the intra-session ICC values were 0.81 and 0.98, respectively; the inter-session coefficients of variation were 2.06% and 1.87%, and the inter-session ICC values were 0.85 and 0.86, respectively (Tables 1, 2). 

Discussion

In clinical routine as well as in experimental settings, knowledge of the repeatability of measurements is of high importance to assess whether observed changes are due to fluctuations in the methods or are valid changes in the structures observed. It also holds true for the measurements of the RNFL and ONH as the principal morphometric parameters to diagnose optic nerve diseases, and to detect early progression of these disorders. Our results indicated that the reproducibility of the measurements of the RNFL thickness and ONH parameters measured in normal monkey eyes were well comparable with the data on the reproducibility of the measurements of the same parameters in patients. In our study, the values of the ICC for re-measurements of all ONH parameters ranged from 0.93–1.00, suggesting a high repeatability and reproducibility. Within the group of the RNFL thickness parameters, the highest ICC was found for the average RNFL thickness, with an ICC value of 0.97 for repeatability and an ICC value of 0.96 for the reproducibility. For superior and inferior MIRL, the intra-session coefficients of variations were 1.55% and 0.65%, and the intra-session ICC values were 0.81 and 0.98, respectively. These findings suggested that the repeatability and reproducibility of the ONH parameters, RNFL thickness, and inner macular thickness measurements with the RTVue-OCT in normal monkey eyes was high and was similar to that reported in human eyes. 

Results similar to our study were obtained in clinical investigations. Garas et al. evaluated the reproducibility of peripapillary RNFLT and ganglion cell complex measurements made with the RTVue-100 OCT device. ¹³ They found that the intra-session ICC for the average and quadrant RNFL thickness parameters, and the ganglion cell complex parameters varied between 93.9% and 99.0%; the intra-session coefficient of variation ranged between 1.95% and 5.69%, and the intra-test variability varied between 3.11 and 9.13 μm. Ho et al. found a high degree of reproducibility in the central foveal region with ICCs of 0.92–0.97. ¹⁴ González-García et al. reported that the ICCs in healthy eyes and patient eyes were 0.97 for the average RNFL thickness, and 0.97 and 0.96, respectively, for the rim area. ¹⁵ In a recent study by Langenegger et al., the ICC (and lower 95% confidence interval [CI]) for the global mean RNFL thickness in normal and glaucomatous eyes was 0.99 (0.98 CI). ¹⁶ In glaucomatous eyes, the coefficient of variation ranged between 1.3% and 3.5%. Mwanza et al, examining glaucomatous eyes, reported on ICCs ranging from 83.9%–99.2% for intra-visit measurements and from 80.8%–99.1% for inter-visit measurements. ¹⁷  

Our findings may serve to show that spectral-domain OCT (RTVue) may be a useful tool, not only for the clinical application for the diagnosis of optic nerve diseases and the detection of their progression in patients, but also for experimental settings in monkeys. Besides the topography of the ONH and the thickness of the RNFL, the assessment of the MIRL may form a third pillar in the assessment of the optic nerve, since previous clinical studies have shown the diagnostic value of examining these structures for the diagnosis of glaucoma, ^19–22 and since our study showed a relatively high reproducibility of the measurements of the MIRL also in monkeys. The relative importance of our study is based on the clinical importance of optical coherence tomography as an imaging technique to examine the ONH, RNFL, and the macula, and on the scientific importance of monkeys in clearing up the pathogenesis and improving the therapy of human diseases. 

Potential limitations of our study should be mentioned. First, the results may be valid primarily for Rhesus macaque monkeys, and it has not been clarified whether and how far the results can be transferred on the reproducibility of the OCT measurements in other species. Second, since our results were obtained on normal monkey eyes, caution must be exerted if the results are transferred on monkey eyes with an optic nerve disease, such as glaucoma. In patients, the reproducibility of a technique usually has been lower than in healthy subjects. Third, again one may be cautious if our results, using the RTVue OCT device, are transferred on morphometric measurements carried out by other OCT devices. 

In summary, there was a high repeatability and reproducibility of spectral-domain OCT (RTVue) assisted measurements of the RNFL, ONH, and MIRL in normal monkey eyes. These results suggested that spectral-domain OCT (RTVue) may be suitable to assess changes in follow-up examinations of monkeys. 

Footnotes

 This work was supported by Beijing Natural Science Foundation (7122038) and an Award of Excellent Doctoral Dissertation of Beijing (YB20091002501) and a Donation by China Health and Medical development Foundation. The sponsor or funding organization had no role in the design or conduct of this research.

Footnotes

²  These authors contributed equally to this work and should be regarded as equivalent authors.

Footnotes

 Disclosure: Z. Zhang, None; D. Yang, None; J. Sang, None; R. Hou, None; K. Liu, None; Z. Li, None; X. Xie, None; J.B. Jonas, None; N. Wang, None.