Abstract
Purpose.:
To evaluate the ability of Fourier-domain (FD) optical coherence tomography (OCT) to detect retinal nerve fiber layer (RNFL) atrophy in multiple sclerosis (MS) patients. To test the intrasession reproducibility of RNFL thickness measurements in MS and healthy subjects using Cirrus (Carl Zeiss Meditec, Dublin, CA) and Spectralis (Heidelberg Engineering, Heidelberg, Germany) OCT.
Methods.:
Two hundred twenty-two eyes of 111 subjects (50 MS patients and 61 healthy subjects) underwent three 360° circular scans centered on the optic disc by the same experienced examiner using the Cirrus and Spectralis OCT instruments. Differences between healthy and MS eyes were compared. The relationship between average thicknesses with each OCT was evaluated. Repeatability was studied by intraclass correlation coefficients and coefficients of variation (COV).
Results.:
RNFL atrophy was detected in the MS eyes for all OCT parameters (P < 0.05). Cirrus and Spectralis showed an RNFL average thickness of 99.4 and 102.5 μm, respectively, in healthy subjects, and 86.0 and 90.4 μm in the MS eyes. RNFL average thickness in the MS eyes determined by both OCTs correlated (r = 0.812; P < 0.001), but were significantly different (P < 0.001). Reproducibility was good. In the MS eyes, Cirrus measurements showed a mean COV of 5.85%, Spectralis 6.80%, and Spectralis with a progression feature 4.16%. Intraclass correlation coefficients were higher than 0.840. RNFL average thickness correlated with disease duration and an optic neuritis antecedent.
Conclusions.:
There are significant differences in RNFL thickness measurements between Cirrus and Spectralis despite a high correlation of measurement between the two instruments. Fourier-domain OCT can be considered a valid device for detecting RNFL atrophy in MS patients.
Axonal loss has been detected in multiple sclerosis (MS) patients and is considered to be the main cause of disability in this disease. Several studies have reported a correlation between axonal loss in the optic nerve and the extent of functional disability of MS patients.
1 –3
Loss of ganglion cells can be detected by means of ocular imaging technologies such as optical coherence tomography (OCT) and scanning laser polarimetry, providing noninvasive, objective, and reproducible methods of evaluating the retinal nerve fiber layer (RNFL). OCT allows cross-sectional imaging of the retina and the optic disc based on interference patterns produced by low-coherence light reflected from retinal tissues.
Classic OCTs, such as the Stratus instrument (Carl Zeiss Meditec, Dublin, CA), employ a time-domain (TD) technique that requires long acquisition times and provides axial and lateral resolutions on the order of 15 μm. TD-OCT uses a scanning interferometer and an 820-nm infrared light source to generate a cross-sectional image of the tested tissue.
4 Recently, improvements in OCT technology have been introduced, including three-dimensional high-resolution OCT, which uses Fourier-domain (FD) detection to provide increased resolution in relation to TD-OCT.
5 Many imaging companies have marketed FD-OCT machines; these include the RTVue (Optovue, Fremont, CA), Spectralis (Heidelberg Engineering Inc, Heidelberg, Germany), SOCT Copernicus (Optopol Technology, Zawiercie, Poland), Cirrus HD-OCT (Carl Zeiss Meditec Inc, Dublin, CA), and 3D OCT-1000 (Topcon, Paramus, NJ). All these devices are capable of measuring macular and RNFL thickness.
The main difference between TD- and FD-OCT (also called spectral-domain [SD]-OCT) is the manner in which information is processed. TD-OCT uses a point detector or photodetector in the detector arm, whereas FD-OCT uses a spectrometer that is composed of a transmission grating and an air-spaced focusing lens. In this device, depth information is acquired by analyzing the interference patterns in a spectrum of mixed reflected lights. To achieve ultrahigh-resolution images, TD-OCT requires increased acquisition times, whereas FD-OCT obtains 2- to 3-μm axial resolution images, with an increase in acquisition speed and without a reduction in image quality.
6,7
Several studies have reported the importance of RNFL thickness determination in the early diagnosis and managing of optic nerve pathologies such as glaucoma,
8,9 band atrophy, with or without chiasmal compression, demyelinating diseases,
2,10,11 and optic neuritis.
12,13 These findings based on OCT can reveal changes in RNFL thickness before visual field defects appear.
10 It is important that these new techniques designed to quantify structural alterations are capable of making accurate, reliable, and reproducible measurements, because the results of macular and RNFL thickness evaluations may vary widely according to the devices used.
Given the value of RNFL examination as a method of detecting MS progression,
12 the purpose of this study was to evaluate the ability of two FD-OCT devices to detect RNFL atrophy in MS patients and the effect of some factors, such as disease duration or history of optic neuritis. To the best of our knowledge, only one article has been published concerning RNFL measurements in MS by FD-OCT.
14 We also assessed the reproducibility of RNFL thickness measurements in healthy and MS subjects of the two most available FD-OCT machines (Cirrus and Spectralis) and compared the results between the two machines. FD-OCT reproducibility in eyes of MS patients has not been reported. In addition, we analyzed the correlations between eyes with higher variability and fixation difficulties.
This was an observational, prospective, cross-sectional study. Fifty patients with a diagnosis of defined MS (17 men, 33 women; 20–72 years of age) and 61 healthy controls (16 men, 45 women; 21–70 years of age) were enrolled.
The diagnosis of MS was based on standard clinical and neuroimaging criteria.
15 Related medical records were carefully reviewed, including the duration of the disease, the Expanded Disability Status Scale (EDSS), treatment, and the presence of prior episodes of optic neuritis (ON) as reported by the treating neurologist and the patient. The EDSS was scored at the time of a routine 6-month clinical visit by a neurologist who was experienced in the diagnosis and treatment of MS.
The disease-free controls were age- and sex-matched to the MS patients and were recruited from hospital staff and family members of patients with no evidence of disease, including neurologic disorders, of any nature. Controls had no history of ocular or neurologic disease. Their best-corrected visual acuity (BCVA) was 20/30 or better, according to the Snellen scale, and they had a normal result in a visual field (VF) examination. Both eyes of each subject were included in the study.
Exclusion criteria were the presence of significant refractive errors (>5 D of spherical equivalent refraction or 3 D of astigmatism), intraocular pressure of 21 mm Hg or higher, media opacification, systemic conditions that could affect the visual system, a history of ocular trauma or concomitant ocular diseases, including a history of retinal disease, glaucoma, laser therapy, or ocular pathologies affecting the cornea, lens, retina, or optic nerve. Eyes with an episode of ON in the 6 months before the study inclusion time point were excluded because, several authors have reported that eyes need at least 6 months' recovery time from the last ON episode, to allow for retrograde degeneration of the RNFL.
16
All procedures adhered to the tenets of the Declaration of Helsinki, and the experimental protocol was approved by the local Ethics Committee. All subjects gave informed consent to participate in the study and underwent a complete neuro-ophthalmic evaluation that included pupillary, anterior segment, and funduscopic examinations; assessment of BCVA relative to the Snellen scale; color vision (Ishihara pseudoisochromatic plates) and VF examination; and an OCT scan with the Cirrus HD-OCT (Carl Zeiss Meditec, Inc.) and the Spectralis OCT (Heidelberg Engineering, Inc.),. Each eye was considered separately.
The VF was assessed with a Humphrey Field Analyzer (Carl-Zeiss Meditec). The SITA Standard strategy (program 30-2) was used to decrease the duration of the examination. The parameter evaluated was mean deviation (MD, in decibels).
The OCT tests obtained measurements of the peripapillary RNFL with the Cirrus and Spectralis OCT devices, used in random order to prevent any effect of fatigue bias. All the scans were performed by the same experienced operator. Between scan acquisitions, there was a time delay and subject position and focus were randomly disrupted, meaning that alignment parameters had to be newly adjusted at the start of each image acquisition. No manual correction was applied to the OCT output. An internal fixation target was used, because it has been shown to give the highest reproducibility.
17 The quality of the scans was assessed before the analysis, and poor-quality scans were rejected. Cirrus OCT determines the quality of images using a signal strength measurement that combines signal-to-noise ratio with the uniformity of the signal within a scan and is measured on a scale of 1 to 10, where 1 is categorized as poor image quality and 10 as excellent image quality. Only images with a score higher than 7 were evaluated in our study. Spectralis OCT uses a blue quality bar in the image to indicate the signal strength. The quality score range is 0 (poor quality) to 40 (excellent quality). Only images with a score higher than 25 were analyzed. Three series of good-quality scans were obtained for each option. Only one patient was excluded, because a centered scan could not be acquired due to her poor fixation. Eighteen images with artifacts, missing parts, or showing seemingly distorted anatomy were excluded.
18 Obtaining good-quality, centered images necessitated repeating scan acquisition in 10 eyes with the Cirrus OCT and in 7 with the Spectralis OCT.
Three repetitions of the scan of the optic disc 200 × 200 cube in each eye were performed with the Cirrus HD OCT. After the recommended procedure for scan acquisition, the subject's pupil was first centered and focused in an iris-viewing camera on the system data acquisition screen, and then the system's line-scanning ophthalmoscope was used to optimize the view of the retina. The OCT scan was aligned to the proper depth and the patient's fixation and the system's polarization were optimized, to maximize the OCT signal. The Cirrus OCT optic disc protocol generates 200 × 200 cube images with 200 linear scans that are performed by 200 A-scans. This option analyses a 6 × 6-mm cube around the optic nerve. In each series of scans, average RNFL thickness, quadrant RNFL thickness (superior, inferior, temporal, and nasal) and 12 clock hours of 30° RNFL thickness were analyzed. The numeration of the hour sectors was assigned from positions H1 to H12 in clockwise direction for the right eye and in counterclockwise direction for the left eye. The Cirrus software version was 3.0.
Three image acquisitions obtained by circular peripapillary Spectralis OCT scans were performed in all subjects (RNFL protocol). The Spectralis OCT system simultaneously captures infrared fundus and SD-OCT images at 40,000 A-scans per second. A real-time eye-tracking system measures eye movements and provides feedback to the scanning mechanism, to stabilize the retinal position of the B-scan. This system thus enables sweep-averaging at each B-scan location to reduce speckle noise. The average number of scans to produce each circular B scan was nine. The RNFL Spectralis protocol generates a map showing the average thickness and six sector thicknesses (superonasal, nasal, inferonasal, inferotemporal, temporal, and superotemporal in the clockwise direction for the right eye and counterclockwise for the left eye). This sequence was redone by same observer using the TruTrack eye-tracking technology (Heidelberg Engineering), which recognizes, locks onto, and follows the patient's retina during scanning, and automatically places follow-up scans to ensure accurate monitoring of disease progression. The Spectralis software version was 3.2.
The presence of defects in the RNFL can be detected with both devices (Cirrus and Spectralis OCT) and is provided by the comparison of measurements from each patient with the normative database of each instrument.
The Kolmogorov-Smirnov test was used to assess sample distribution. RNFL thicknesses were compared between patients and healthy controls by means of a Student's t-test, given a normal distribution. Box plots were constructed to represent the differences observed in the RNFL thickness between healthy subjects and MS patients. A Student's t-test for paired data was used in analyzing the differences between the Cirrus and Spectralis RNFL measurements in each group. Values of P < 0.05 were considered to be indicative of statistically significant differences. RNFL thicknesses were compared between 25 eyes with previous ON and 75 eyes without this antecedent by means of a Student's t-test (all analyses: SPSS 15.0, SPSS Inc., Chicago, IL).
The relationship between the measurements obtained using each OCT was evaluated with the Pearson correlation analysis. Average thicknesses obtained with both devices were compared by Student's t-test. Other sectors, quadrants, or RNFL areas could not be compared, because they do not yield equivalent values according to the two tomographies. Regression analysis was used to identify variables that were predictors of axonal damage in MS patients.
For each parameter, the coefficient of variation (COV) was calculated as the standard deviation (SD) divided by the average of the measurement value and expressed as a percentage. Most authors regard devices with a COV < 10% as having high reproducibility; a COV < 5% indicates very high reproducibility.
9 To assess the reliability of the repeated measurements, the intraclass correlation coefficients (ICCs) for absolute agreement were calculated. They measure the concordance of continuous variables and correct correlations for systematic bias. The ICC interpretation that we used was slight reliability (for values between 0 and 0.2), fair reliability (from 0.21 to 0.4), moderate reliability (values between 0.41 and 0.6), substantial reliability (values from 0.61 to 0.8), and almost perfect reliability (for values higher than 0.81). Bland-Altman plots were used to assess agreement. Pearson analysis was performed to test whether measurements in eyes with higher reproducibility with the Cirrus OCT would correlate with those obtained with the Spectralis OCT.
RNFL thickness measurements showed good COVs and ICCs in the MS patients (
Table 4). The results obtained with Cirrus OCT were highly reproducible in all quadrants and sectors of the MS eyes, with a mean COV of 5.85% ± 6.13% (range, 2.70%–9.27%) and an ICC higher than 0.840. Average thickness was the parameter with least variability (COV = 2.70%; ICC = 0.967).
Table 4. Correlation of Repeated RNFL Thickness Measurements
Table 4. Correlation of Repeated RNFL Thickness Measurements
| MS Eyes | Healthy Eyes |
| COV | ICC | COV | ICC |
| Cirrus OCT | | | | |
| Average thickness | 2.70 | 0.967 | 1.36 | 0.988 |
| Superior quadrant | 5.14 | 0.867 | 3.38 | 0.949 |
| Nasal quadrant | 6.69 | 0.878 | 4.07 | 0.961 |
| Inferior quadrant | 4.38 | 0.945 | 3.38 | 0.945 |
| Temporal quadrant | 5.14 | 0.931 | 2.87 | 0.962 |
| C1 hour sector | 6.86 | 0.954 | 5.01 | 0.974 |
| C2 hour sector | 9.27 | 0.906 | 5.49 | 0.956 |
| C3 hour sector | 6.06 | 0.907 | 4.35 | 0.953 |
| C4 hour sector | 6.35 | 0.934 | 5.02 | 0.961 |
| C5 hour sector | 7.19 | 0.938 | 5.49 | 0.956 |
| C6 hour sector | 4.96 | 0.977 | 5.37 | 0.931 |
| C7 hour sector | 6.39 | 0.936 | 3.61 | 0.973 |
| C8 hour sector | 6.61 | 0.980 | 5.34 | 0.924 |
| C9 hour sector | 5.31 | 0.970 | 2.75 | 0.969 |
| C10 hour sector | 4.73 | 0.980 | 3.38 | 0.975 |
| C11 hour sector | 6.06 | 0.840 | 3.41 | 0.978 |
| C12 hour sector | 5.62 | 0.958 | 6.03 | 0.945 |
| Spectralis OCT | | | | |
| Average thickness | 3.70 | 0.907 | 1.78 | 0.980 |
| Superonasal area | 7.0 | 0.885 | 5.65 | 0.974 |
| Nasal area | 7.56 | 0.887 | 6.18 | 0.939 |
| Inferonasal area | 9.92 | 0.854 | 4.60 | 0.965 |
| Inferotemporal area | 5.18 | 0.947 | 2.70 | 0.953 |
| Temporal area | 8.06 | 0.915 | 3.78 | 0.960 |
| Superotemporal area | 5.89 | 0.848 | 3.66 | 0.936 |
| Spectralis OCT with TruTrack | | | | |
| Average thickness | 2.59 | 0.956 | 1.31 | 0.987 |
| Superonasal area | 4.54 | 0.937 | 3.24 | 0.954 |
| Nasal area | 4.01 | 0.921 | 4.04 | 0.946 |
| Inferonasal area | 5.24 | 0.872 | 3.65 | 0.888 |
| Inferotemporal area | 3.51 | 0.968 | 2.67 | 0.979 |
| Temporal area | 5.76 | 0.945 | 3.02 | 0.956 |
| Superotemporal area | 3.45 | 0.977 | 2.97 | 0.984 |
Measurements taken with the Spectralis without TruTrack eye-tracking technology were also highly reproducible, with a mean COV of 6.80% ± 7.47% (range: 3.70%–9.92%) and an ICC higher than 0.937. The parameter with the least variability was again average thickness (COV = 3.70% and ICC = 0.907).
The results obtained with the Spectralis TruTrack progression feature showed higher reproducibility in all quadrants and sectors in the MS and healthy subjects, with mean COVs of 4.16% (range, 2.59%–5.76%) and 2.98% (range, 1.31%–4.04%), respectively. The least variability was also found in average thickness (COV = 2.59% in MS patients).
Figure 3 shows Bland-Altman plots of RNFL average thickness reproducibility between different measurements obtained with the Cirrus (
Fig. 3A) and Spectralis (
Fig. 3B) OCT devices. We found more measurement variability in patients with fixation or ocular movement alterations, even when the quality of the tests was good. Eyes with previous ON and lowest BCVA showed higher variations between successive scans with both devices, but a significant correlation was not found between the eyes with high variability in Cirrus and Spectralis OCT (
P = 0.318).
Almost all parameters showed less variability in healthy eyes, as can be seen from the results presented in
Table 4.
Several studies have suggested that FD-OCT has better reproducibility and can detect retinal pathologies more readily than can conventional TD-OCT.
14 Various reports on FD-OCT RNFL thickness measurements in healthy subjects have been published. Recently, a study of the reproducibility of high-resolution OCT in MS
14 and some comparisons between RNFL measurements with FD and TD-OCT in healthy eyes
19,20 have been published. Only one study comparing the Spectralis and Cirrus OCT devices has been reported, in glaucomatous eyes of four patients.
21 Moreover, comparison between FD-OCT measurements in eyes of MS patients has not been reported.
Our study has demonstrated that the new FD-OCT technologies, which include the Cirrus and Spectralis OCTs, could be useful in the detection of axonal defects in MS patients, regardless of the presence of a previous episode of ON.
In recent years, many new instruments have been introduced to quantify retinal ganglion cells, leading some authors to suggest that changes in the RNFL may reflect similar pathologic changes taking place in the brain.
2,11 Ocular imaging technologies, such as OCT, scanning laser polarimetry (GDx; Carl Zeiss Meditec, Inc.) or confocal scanning laser ophthalmoscopy (HRT; Heidelberg Engineering), provide an opportunity for clinical observation of the axonal constituents of the anterior visual pathway, thereby allowing direct visualization of part of the central nervous system. As the RNFL is composed only of unmyelinated axons, measuring RNFL thickness may be a method of monitoring axonal loss in MS patients.
2,12,22,23 A good correlation between RNFL thickness and magnetic resonance imaging (MRI) brain measurements, such as parenchymal fraction and brain volume, has been described,
2,24 and a strong association between average RNFL thickness and normalized brain volume has been found.
25
New FD-OCT systems measure similar properties of RNFL thickness, as does TD-OCT, but they provide an improved image with less artifact and better resolution, image quality and reproducibility. In the coming years, FD-OCT will no doubt replace TD-OCT in the in vivo analysis of optic nerve fiber integrity and origin in the retinal ganglion cells in neurologic diseases such as MS.
5
In our study, the MS eyes exhibited thinning of the RNFL compared with the healthy eyes, even if the analysis was done on only non-ON eyes. Several reports have suggested that eyes without a history of ON have subclinical axonal loss associated with the disease, but RNFL atrophy is greater in MS eyes with ON.
1,3,12,26,27 In a previous study with a 1-year follow-up, we detected a statistically significant decrease in the RNFL thickness in MS patients, either in eyes with a prior ON or in those with no known prior episode.
12
Most MS studies published to date have included both eyes of each MS patient, because they can be individually evaluated and do not necessarily follow the same disease course. If one eye had presented with ON and the other one had MS without previous ON, the two cases could be used to compare the effect of inflammation in the optic nerve.
1,12,25,28 We found that all OCT-measured thicknesses were higher in healthy eyes than in MS eyes, but the atrophy was significant higher in eyes with an ON antecedent. Our previous results indicate a significant linear decrease in mean RNFL thickness with age in healthy and MS subjects, with negative slopes of 1.40 and 2.66 μm/year, respectively (Garcia Martin E, et al., manuscript submitted).
The RNFL thickness map is different for the Cirrus and Spectralis instruments and so the only parameter that we could compare between both tomographies was the average thickness. Cirrus OCT evaluates the thicknesses of four quadrants and 12 one-hour clock sectors, while Spectralis divides the RNFL thickness analysis into six areas that have no relation to the Cirrus sectors or quadrants. RNFL thicknesses measured by Cirrus OCT tend to be less than those recorded with Spectralis OCT. The Cirrus OCT considers the RNFL anterior limit to be the internal limiting membrane and the posterior border to be the posterior RNFL limit. In contrast, thickness measurements using Spectralis OCT are derived from delineation of the anterior (internal limiting membrane) and posterior borders along a single A-scan at the appropriate eccentricity within each radial B-scan. This eccentricity was determined to be equivalent to 1400 μm from the center of the ON head, as indicated by a ruler within the OCT visualization software. In converting angular span to linear distance, the Spectralis instrument assumes an emmetropic human eye with average axial length.
29 Based on our results, the measurements of Cirrus and Spectralis OCTs were not equivalent, meaning that the same tomography device should be used to evaluate the RNFL of patients when attempting to detect progression or changes in disease, as several authors have previously suggested for other tomography devices.
30 –33 These findings are particularly relevant when an individual undergoes longitudinal follow-up with different OCTs.
32 The difference in RNFL thickness between two machines is most likely due to algorithm differences between the two manufacturers for determining the inner and outer border of the RNFL.
Despite the fundamental differences in these technologies, the generated RNFL average thickness for Cirrus OCT correlated significantly with the Stratus OCT measurement.
33,34 Several authors have found a strong correlation between FD- and TD-OCT in RNFL protocols, but there have been no published results of the correlation between two FD instruments in MS patients.
30,32 –34 We found a good correlation between the Cirrus and Spectralis OCT measurements of RNFL thickness.
In recent months, several reports have been published on RNFL reproducibility with new FD-OCT machines. These show excellent intraobserver reproducibility with virtually identical results between these instruments and retinal thickness measurements performed by Spectralis OCT,
19,35 –38 but normal eyes presented less variability than pathologic eyes. Lee et al.
36 reported a higher COV in the glaucoma group (ranged from 2% to 5.3%) than in healthy subjects (1.9%–5%). Correlations between RNFL measurements by OCT and stereoscopic photographs, between OCT and VF, or between OCT and GDx measurements have been reported.
39 TruTrack technology, we report herein, seems to improve measurement reproducibility.
We have observed higher variance using the Spectralis OCT compared with the Cirrus OCT, but the best reproducibility was found when using the TruTrack system of Spectralis OCT. We think that the fixation system may explain this difference. Complaints about the fixation light are common in patients examined with the Spectralis OCT. They report that the fixation point is easier to follow in the Cirrus OCT. Some of our patients had difficulty in maintaining eye fixation because they had prior episodes of ON or eye movement alterations. When the TruTrack system is activated, Spectralis locks onto and follows the patient's retina and optic nerve during the scan, independent of eye fixation, so that reproducibility is improved.
We found an average RNFL mean variability of 2.99 μm using FD-OCT in the MS eyes, which is enough to detect the degree of change that can be expected from the progression of MS in 1 year (4.50 μm), as we reported with Stratus OCT in our previous study.
12 However, we consider that 1 year is too short a time for detecting significant RNFL changes in a chronic disease such as MS, especially if only those patients without MS relapses are included, as in our longitudinal study. We propose that the results showing RNFL changes would be greater if all patients (with and without relapses) were included. We also suggest that OCT is a useful technique for detecting axonal loss in MS patients since we found decreased RNFL thicknesses in patients compared with those in healthy subjects in both studies. We observed a clear tendency for thickness reduction in all RNFL parameters of MS patients, and we found statistically significant differences between healthy and MS eyes. Although OCT variability was low, any reduction in RNFL thickness should be carefully evaluated because it may be caused by device variability rather than MS progression. This limitation may be reduced by using serial scans and periodic RNFL evaluations.
Several authors have reported, as we have described, that the nasal sector is the most variable in OCT.
8 When using the Cirrus OCT, we found that the least reproducible quadrant was the nasal quadrant and the most reproducible value was the average thickness. Similar results were found with the Spectralis OCT: The least variability was observed in the average thickness value, whereas nasal sectors showed the most variability (inferonasal sector in the MS eyes and nasal sector in healthy eyes). Blumenthal et al.
26 obtained similar results when they evaluated the reproducibility of RNFL measurements using the commercially available OCT 2000 instrument (Humphrey Systems; Carl Zeiss Meditec, Inc.) in normal and glaucomatous eyes. They reported COVs of 18% in the nasal quadrant of healthy eyes and 28% in pathologic eyes, whereas COVs in the inferior quadrant were 8% and 12%, respectively. They also reported that the parameter with the most repeatability was average thickness (COV, 7% in healthy eyes). Previous studies have reported more variability in the nasal quadrant of healthy and glaucomatous eyes,
9,17,26 which authors attributed to the difficulty in using the measurement algorithm to calculate RNFL thickness in the nasal quadrant.
38,40,41 Knighton and Qian
40 have suggested that this problem arises because the angle of incidence of the illuminating beam makes the RNFL image on the nasal side dimmer.
We found lower reproducibility in Spectralis measurements, perhaps because subjects find it easier to see the fixation point with the Cirrus OCT during the scanning procedure.
In recent years, several authors have suggested that OCT variability increases when the thickness measurement is larger, but this concept has not been demonstrated. Blumenthal et al.,
26 and Pueyo et al.
9 reported that RNFL variability is higher in pathologic eyes that have a lower RNFL thickness. A statistical analysis to evaluate whether the variability was higher when RNFL thickness increased was performed in the present study. We found that the variability tended to rise with increased thickness, but the results were not significant. Most of our parameters showed less variability in normal eyes.
Our results suggest that FD-OCTs, such as the Cirrus and Spectralis systems, are able to detect axonal atrophy in the RNFL of MS eyes and show a good reproducibility of RNFL thickness measurements in healthy and MS eyes. Repeat scans showed low intraobserver variation. The best reproducibility was found with the TruTrack eye-tracking technology of the Spectralis OCT. The repeatability of RNFL measurements, with a COV < 5%, is adequate from a clinical perspective, but one must keep in mind that differences less than 2.99 μm in repeat OCT scans could be caused by device variability in the testing of axonal atrophy in MS patients.
Longer prospective studies using FD-OCT may elucidate the ability of these devices to detect axonal degeneration caused by the progression of MS.
Supported in part by Grant PI080976 from the Instituto de Salud Carlos III.
Disclosure:
E. Garcia-Martin, None;
V. Pueyo, None;
I. Pinilla, None;
J.-R. Ara, None;
J. Martin, None;
J. Fernandez, None
The authors thank Antonio Ferreras and Luis E. Pablo for helping in the acquisition of the new tomographers.