This study demonstrated that Wide scan measurements from the Triton SS-OCT and Maestro SD-OCT have excellent agreement in both normal and glaucoma eyes. In addition, the repeatability and reproducibility of cpRNFL, macular GCIPL, and macular GCC thickness measurements from the Wide and the Macular/ Optic Disc Cube scans were similar for both the Triton and the Maestro in normal and glaucoma eyes.
Measuring structural changes is essential for diagnosing and monitoring glaucoma,1 and OCT is a well-established method to objectively assess structural changes in eyes with glaucoma.14 Before the introduction of a Wide scan that simultaneously captures peripapillary and macular anatomical structures in one scan, separate Macular and Optic Disc Cube scans were required to quantitively assess macular RGCs and RNFL respectively. However, in clinical practice, often only one OCT scan, typically the Optic Disc Cube scan, is captured due to time limitations.7 Rather than acquiring data with both Macular and Optic Disc Cube scans, the Wide scan and its incorporated automated segmentation software makes it possible to analyze the thickness of peripapillary RNFL and various macular retinal layers simultaneously using data obtained with only a single scan. As found in the current study, the Wide scan and Cube scan measurements are comparable. Moreover, imaging time is reduced with simultaneous imaging of the macula and the peripapillary region relative to the capturing of two scans per eye. This also minimizes the adverse impact on image quality caused by patient fatigue, motion, and alignment errors. In addition, the paracentral fixation target of the Wide scan reduces the fixation errors caused during acquisition of the Optic Disc Cube scan which requires nasal fixation.8,15 Another strength of the Wide scan is that peripapillary RNFL defects at 11 and 12 o’clock (in the right eye orientation) can be missed in the Macular Cube scans, while they are more likely visualized in a Wide scan including both cpRNFL and macular GCC.16 Another study has shown that the thickness map provided for Wide scans detects early structural changes that might not be detected well using peripapillary RNFL or macular GCIPL thickness maps. Furthermore, RNFL defects distant from the optic disc can be more easily visualized with the Wide scan RNFL maps.17
The diagnostic power of the Wide scan has been evaluated by several studies. Yang et al.18 compared the diagnostic ability (healthy vs glaucoma) of RNFL18 and macular GCIPL and macular GCC19 thickness values from a SS-OCT Wide scan (DRI-OCT, Topcon) and Macular/Optic Disc Cube scans of a SD-OCT (Spectralis, Heidelberg Engineering and Cirrus HD-OCT, Carl Zeiss Meditec), and reported that RNFL, macular GCIPL, and macular GCC thickness values from the Wide scan measured by SS-OCT had similar diagnostic accuracy to Macular/ Optic Disc Cube scan measurements obtained by SD-OCT. Similarly, another study showed the diagnostic ability of SS-OCT (DRI-OCT-1 Atlantis, Topcon) Wide scan measurements for distinguishing preperimetric and early glaucoma from healthy eyes was similar to Macular/ Optic Disc Cube scan measurements from SD-OCT (Cirrus HD-OCT, Carl Zeiss Meditec).6 These previous studies compared diagnostic ability of SS-OCT Wide scan with SD-OCT Cube scan, while Hong et al.8 compared glaucoma-discriminating ability of measurements from Wide scans with Macular/ Optic Disc Cube scans of SS-OCT (DRI-OCT-1 Atlantis, Topcon) and reported that they were comparable. Furthermore, Hood et al.7 reported that the report based upon a single Wide scan has the information needed to diagnose early glaucoma with excellent sensitivity and specificity. Thus, it has been suggested that the Wide scan could replace Macular/ Optic Disc Cube scans for diagnosing and screening glaucoma.7,8
Besides discrimination between normal and glaucoma, monitoring patients with glaucoma to detect progression is the mainstay of glaucoma care, which requires reliable measures with good repeatability and reproducibility.20 Studies addressing measurement precision of Wide scans are limited. One study15 using SD-OCT (Canon OCT-HS100, Canon Europe) compared repeatability of measurements from a Wide scan (13mm×10mm) and Cube scans (Macular scan 10mm×10mm, Optic Disc scan 6mm×6mm) in healthy eyes. Different from our results, they found a 2–3 times larger repeatability limit of the Wide scan compared with the Cube scans. The authors attributed their result partially to the scan density in the Wide scan, which is 4.4 times less than for the individual Cube scans. By contrast, the Wide scan and Cube scan in the current study are closer in scan density. It has also been reported previously that the scan direction affects precision, where horizontal scans have better repeatability than vertical scans.21 The Wide scan in the prior study employed vertical B-scans, while the Optic Disc and Macular Cube scans were captured horizontally and vertically, respectively,15 thereby potentially contributing to the varied repeatability between the Wide and Cube scans. With SS-OCT (DRI-OCT-1 Atlantis, Topcon), another study8, found comparably good repeatability of macular GCIPL and macular GCC thickness values from the Wide scan and Cube scans in healthy and glaucoma eyes; the current results are consistent with these earlier ones. The current study expands on that earlier one as it evaluated repeatability and reproducibility in normal and glaucoma eyes. It shows comparable precision of parameters relevant for glaucoma management between the Wide scan and Macular and Optic Disc Cube scans for both the Triton and Maestro OCT devices.
Agreement between SS-OCT and SD-OCT, as well as agreement between the Wide scan and Macular/ Optic Disc Cube scans have been previously studied. Lee SY et al.22 evaluated agreement between SS-OCT and SD-OCT Cube scans, and Lee WJ et al.6 evaluated agreement between SS-OCT Wide scan and SD-OCT Cube scans, respectively, in normal eyes using the same devices (SS-OCT: DRI-OCT-1 Atlantis, Topcon vs. SD-OCT: Cirrus HD-OCT, Carl Zeiss Meditec); while Yang et al.19 evaluated agreement between SS-OCT Wide scan (DRI-OCT, Topcon) and SD-OCT Cube scans (Cirrus HD-OCT, Carl Zeiss Meditec) for healthy and glaucomatous eyes. In all of these studies, the comparison included different manufacturers, algorithms, and measurement positions and grids. In contrast, the Triton and Maestro share segmentation algorithms and measurement locations. That may explain why the differences observed in this study were generally smaller than those in previous studies. Importantly, the differences between Triton and Maestro (mean difference of all measurements < 3 µm) were less than the axial resolution in tissue (Triton axial resolution 8µm, Maestro axial resolution 6 µm23), and smaller than the corresponding reproducibility limits. Therefore, these differences are assumed not to be clinically significant. The reported agreement results were expected based on minor differences in axial resolution, software and algorithm, and a minimal difference in the pixel calibration factor between Triton and Maestro. In addition, Hong et al.8 found excellent agreement for macular GCIPL, macular GCC and peripapillary RNFL measurements between the SS-OCT (DRI-OCT-1 Atlantis, Topcon) Wide scan and Cube scans for healthy and glaucomatous eyes; Dominguez-Vicent et al.15 showed that measurement differences between the Wide and Cube scans for SD-OCT (Canon OCT-HS100, Canon Europe) were mostly lower than the axial resolution of the device for healthy eyes. In summary, these studies suggested that, for glaucoma follow up, consistency of scan type, device, and OCT technology need to be considered. Although the same device and scan type is optimal, this study demonstrates that measurement interchangeability may be expected within certain configurations of scan types and devices, such as the 12mm ×9mm Wide scan of the Triton SS-OCT and the Maestro SD-OCT.
There are several limitations of this study. First, the results of this study were obtained entirely from Caucasian subjects. Although we do not expect ethnicity to directly affect repeatability or reproducibility, additional studies using subjects from different populations would generalize our conclusions. Second, there was a significant difference in age distributions between Normal and Glaucoma groups. The influence of the inter-group age difference on the current study results (from inter-scan type and inter-device analyses) is negligible because all estimates were presented for each single group without inter-group comparison. The range of retinal thickness measurements of normal subjects from this study were highly similar with that in other publications.24–26 Moreover, although the retinal thickness measurements decrease with age (total retina thinning 0.53 µm/year; RNFL thinning 0.44 µm/year),27 there is no evidence that the rate of age-related thinning varies in different age groups. Therefore, even if test interval is long enough to affect the evaluation of repeatability and reproducibility, which is not applicable for the current study, the effect should be equal between the groups. Nevertheless, one should take the applicable age range into consideration when interpreting the values of retinal thickness of each group. Third, the glaucoma patients included in this study had an average MD of -5.30 dB indicating that most subjects had early to moderate glaucoma. Studies with a wider distribution of glaucoma severity are needed to evaluate the utility of the Wide scan in advanced disease. Fourth, the sample size of the current study is relatively small with 25 eyes in each group. The sample size was determined based on the 95% LOA and the ANOVA model for precision and 21 eyes per group were deemed appropriate. Lastly, although this study suggests a potential role of a Wide scan for glaucoma monitoring, this cross-sectional study was unable to evaluate how well the Wide scan measurements of Triton and Maestro can identify glaucomatous progression. Longitudinal studies are needed to further evaluate the clinical utility of Wide scans in monitoring glaucoma progression.
In conclusion, we have demonstrated high and comparable precision of peripapillary and macula thickness measurements from Wide, Macular Cube, and Optic Disc Cube scans of the Triton SS-OCT and the Maestro SD-OCT in normal and glaucoma eyes. Wide scan measurements of the Triton SS-OCT and Maestro SD-OCT were interchangeable with excellent agreement. These findings show the potential for more simultaneous evaluation of both macular and peripapillary retinal anatomy from a single Wide OCT scan rather than the clinical standard of capturing an Optic Disc Cube scan and a Macular Cube scan for glaucoma diagnosis and management.