Ophthalmic Surgery, Lasers and Imaging Retina

Clinical Science 

Influence of Retinal Pathology on the Reliability of Macular Thickness Measurement: A Comparison Between Optical Coherence Tomography Devices

Bobak Bahrami, MBBS; Shaun Y. P. Ewe, MBBS; Meidong Zhu, MBBS, PhD; Thomas Hong, PhD; Germane Ong; Kehui Luo, PhD; Andrew Chang, MBBS (Hons), PhD, FRACS, FRANZCO

Abstract

BACKGROUND AND OBJECTIVE:

To evaluate the repeatability, reliability, and comparability of macular thickness measurements between three optical coherence tomography (OCT) machines in healthy eyes, eyes with diabetic macular edema (DME), and eyes with neovascular age-related macular degeneration (nAMD).

PATIENTS AND METHODS:

Twenty-three eyes with DME, 26 eyes with nAMD, and 24 healthy eyes as controls were evaluated. Scans were performed using the swept-source Triton (Topcon, Tokyo, Japan), the spectral-domain Cirrus (Carl Zeiss Meditec, Dublin, CA), and the Spectralis (Heidelberg Engineering, Heidelberg, Germany) machines. Scans were evaluated for central macular thickness (CMT), presence of segmentation and fixation imaging artifacts (IA), re-scan reliability, and agreement between machines and groups.

RESULTS:

Mean CMT was significantly different between all OCT machines in all groups (P < .01 for all comparisons). Manually correcting IA did not alter these results. There was good scan repeatability among healthy and DME eyes for each machine, but poor repeatability among the nAMD group with the Spectralis (P = .038). IA were significantly increased in the presence of pathology.

CONCLUSIONS:

There is poor agreement of CMT measurement between OCT machines in healthy eyes and those with DME and nAMD. DME and nAMD have a significant effect on the rate of IA in scans. Care is required when interpreting measurements from different OCT devices in clinical practice and research settings.

[Ophthalmic Surg Lasers Imaging Retina. 2017;48:319–325.]

Abstract

BACKGROUND AND OBJECTIVE:

To evaluate the repeatability, reliability, and comparability of macular thickness measurements between three optical coherence tomography (OCT) machines in healthy eyes, eyes with diabetic macular edema (DME), and eyes with neovascular age-related macular degeneration (nAMD).

PATIENTS AND METHODS:

Twenty-three eyes with DME, 26 eyes with nAMD, and 24 healthy eyes as controls were evaluated. Scans were performed using the swept-source Triton (Topcon, Tokyo, Japan), the spectral-domain Cirrus (Carl Zeiss Meditec, Dublin, CA), and the Spectralis (Heidelberg Engineering, Heidelberg, Germany) machines. Scans were evaluated for central macular thickness (CMT), presence of segmentation and fixation imaging artifacts (IA), re-scan reliability, and agreement between machines and groups.

RESULTS:

Mean CMT was significantly different between all OCT machines in all groups (P < .01 for all comparisons). Manually correcting IA did not alter these results. There was good scan repeatability among healthy and DME eyes for each machine, but poor repeatability among the nAMD group with the Spectralis (P = .038). IA were significantly increased in the presence of pathology.

CONCLUSIONS:

There is poor agreement of CMT measurement between OCT machines in healthy eyes and those with DME and nAMD. DME and nAMD have a significant effect on the rate of IA in scans. Care is required when interpreting measurements from different OCT devices in clinical practice and research settings.

[Ophthalmic Surg Lasers Imaging Retina. 2017;48:319–325.]

Introduction

The development of optical coherence tomography (OCT) has revolutionized the diagnosis and management of retinal pathology. OCT is an increasingly accessible technology, with more than 35 device manufacturers producing machines for commercial and research use.1 These machines utilize different software and hardware to analyze obtained images. Because of these variations, retinal thickness measurements obtained from different OCT machines are not comparable. Additionally, readings from individual machines may be unreliable due to image segmentation errors and artifacts. These differences are apparent in both healthy eyes and in those with pathology.2–13

Reliable measurements of central macular thickness (CMT) are important in common vision-threatening conditions such as neovascular age-related macular degeneration (nAMD) and diabetic macular edema (DME). This information can be indicative of disease activity and is an outcome measure in many clinical trials of these conditions.14–19 With the ubiquity of OCT, it is not uncommon for a patient to have multiple retinal scans performed by different practitioners on different machines at different times. Thus, interpreting quantitative data obtained from different machines poses a challenge.

In this study, we evaluate the reliability and comparability of CMT measurements and rates of imaging artifacts on two commonly used spectral-domain (SD) machines and a newer swept-source (SS)-OCT machine. We also evaluate the influence of DME and nAMD on the OCT scan performance.

Patients and Methods

Patients and volunteers from a single tertiary referral center were recruited for this cross-sectional study. Participants' eyes were divided into three groups: eyes with no previous ocular history used as control, eyes diagnosed with DME, and eyes diagnosed with nAMD. All study assessments were performed after obtaining informed consent from all participants and were conducted in accordance with the tenets of the Declaration of Helsinki.

Recruited patients had a best-corrected visual acuity (BCVA) of 55 ETDRS letters (Snellen equivalent: 20/80) or better in the study eye and were able to fixate on machine-generated targets. Tropicamide 1% was used for pupillary dilation prior to posterior segment examination and scanning.

All OCT scans were performed by four technicians with previous clinical trial scanning experience. Scans were performed using the swept-source SS Triton DRI-OCT (software version 10.0; Topcon, Tokyo, Japan), the Spectralis SD-OCT (software version 6.4.8.0; Heidelberg Engineering, Heidelberg, Germany), and the Cirrus SD-OCT (software version 6.0.2.81; Carl Zeiss Meditec, Dublin, CA). For each participant, the same OCT operator performed all OCT scans on each of the machines on the same day. Two replicate scans were performed for each equivalent scanning protocol on all three OCT machines (Triton: 3D Macular Scan; Cirrus: Macular Cube 512 × 128; Spectralis: Dense Scan [49 lines]).

Outcome measures were CMT, defined as the mean thickness from Bruch's membrane to the inner retinal border within the central 1-mm circle of the ETDRS grid; presence of imaging artifacts (IA); re-scan repeatability; and agreement between machines and groups.

IA were classified as related to segmentation or fixation errors. Segmentation artifacts were due to inappropriate automated segmentation of retinal layers resulting in inaccurate retinal thickness measurements (Figure 1). Fixation artifacts were associated with inappropriate identification of the fovea (Figure 2). IA were subsequently corrected manually and analyses were redone.


(A) Segmentation error with the Triton device due to retinal hard exudate in an eye with diabetic macular edema. (B) Segmentation error with the Cirrus device due to pigment epithelium detachment in an eye with neovascular age-related macular degeneration (nAMD). (C) Segmentation error with the Spectralis device due to geographic atrophy in an eye with nAMD.

Figure 1.

(A) Segmentation error with the Triton device due to retinal hard exudate in an eye with diabetic macular edema. (B) Segmentation error with the Cirrus device due to pigment epithelium detachment in an eye with neovascular age-related macular degeneration (nAMD). (C) Segmentation error with the Spectralis device due to geographic atrophy in an eye with nAMD.


Fixation error on the Triton device resulting in incorrect ETDRS grid placement and calculation of central macular thickness.

Figure 2.

Fixation error on the Triton device resulting in incorrect ETDRS grid placement and calculation of central macular thickness.

Results were analyzed using IBM SPSS (version 21; SPSS, Chicago, IL). Paired t-tests were used to investigate differences between two repeated scans for reliability on each OCT machine and between each pair of OCT machines for comparability. Mean measurements of the two replicate scans were used when evaluating the comparability between each pair of the three OCT machines. For the analyses presented, all eyes studied were considered independent from one another. A P value less than .05 was considered to be a significant difference.

Results

Subjects

A total of 73 eyes from 45 participants were enrolled during the study period between June and August 2015. The control group consisted of 24 healthy eyes of 12 participants with no history of ocular disease, the nAMD group consisted of 26 eyes of 17 participants, and the DME group consisted of 23 eyes of 16 participants. Both the nAMD and DME groups were undergoing treatment with anti-vascular endothelial growth factor (VEGF) drugs. Baseline characteristics of the entire cohort are summarized in Table 1.


Baseline Characteristics of Study Population

Table 1:

Baseline Characteristics of Study Population

Repeatability of Macular Scans

All test-retest repeated CMT measurements performed for the entire cohort demonstrated reliability in each machine with no significant differences between the first and second scans (Spectralis 3.0 μm [95% CI: −1.1 μm to 7.2 μm; P = .15], Cirrus −0.4 μm [95% CI: −2.8 μm to 2.0 μm; P = .74], Triton −0.4 μm (95% CI: −3.3 μm to 2.4 μm; P = .77]).

Subgroup analysis showed that the control and DME groups maintained repeatability of scans across all three machines. The nAMD group, however, showed relatively poor repeatability, with a significant difference identified on CMT measurements obtained from the Spectralis (P = .038), but not from the Triton or the Cirrus (Table 2).


Mean Difference Between Two Consecutive Scans Stratified by Subgroup Before and After Resegmentation to Correct for Imaging Artifacts

Table 2:

Mean Difference Between Two Consecutive Scans Stratified by Subgroup Before and After Resegmentation to Correct for Imaging Artifacts

Imaging Artifacts

There was a low occurrence of imaging artifacts (IA) in the control group (4.2%, 8.3%, and 8.3% in Spectralis, Triton, Cirrus, respectively). Higher rates of artifacts occurred in the DME group (47.8%, 52.2%, and 34.8% in Spectralis, Triton, and Cirrus, respectively; P < .05 compared to controls) and in the nAMD group (84.6%, 50%, and 42.3% in Spectralis, Triton, and Cirrus, respectively; P < .05 compared to controls). When comparing machines, the incidence of IA in nAMD eyes was significantly higher on the Spectralis compared to Triton (P = .02) and Cirrus (P < .01). The prevalence of IA was otherwise similar in the control group and the DME group across all three machines.

Poor reliability of repeated measures were associated with IA. This was supported by the overall scan-rescan reliability when manual resegmentation corrected these errors (Table 2). The incidence of IA and the breakdown of pathological lesions found in the groups are summarized in Table 3.


Incidence of Artifacts by Optical Coherence Tomography Machine and Subgroup

Table 3:

Incidence of Artifacts by Optical Coherence Tomography Machine and Subgroup

Comparability Between OCT Machines

Significant differences were seen in CMT across all three machines amongst the whole cohort, as well as in each subgroup. The CMT mean measurements consistently measured highest in Spectralis, followed by the Triton and then the Cirrus, showing low agreement between machines (P < .05 in all group comparisons) (Figure 3). These differences remained evident even after manual resegmentation of scans to correct for IA.


Average central macular thickness (CMT) for different groups and machines following manual resegmentation. A consistent difference in measurements was found across all three optical coherence tomography machines, with the Spectralis device consistently recording the highest CMT values, followed by the Cirrus and the Triton devices, respectively, in all three groups.

Figure 3.

Average central macular thickness (CMT) for different groups and machines following manual resegmentation. A consistent difference in measurements was found across all three optical coherence tomography machines, with the Spectralis device consistently recording the highest CMT values, followed by the Cirrus and the Triton devices, respectively, in all three groups.

Discussion

In this study, there was good repeatability of OCT scans in both healthy eyes and in those with nAMD and DME. However, there was a significantly higher rate of IA in eyes with pathology, reducing the reliability of scans in these settings. Additionally, CMT measurements were not comparable between these three machines, with poor agreement of values in healthy eyes and those with pathology. Recorded CMT was highest in the Spectralis, followed by the Cirrus and then the Triton in all groups.

Similarly, previous studies have demonstrated good reliability and reproducibility for a number of SD-OCT machines in calculating CMT in healthy eyes, including the Spectralis and Cirrus.3–5,8,11 Additionally, differences have been demonstrated between OCT machines in both healthy eyes as well as those with pathology.2–5,7–11,13 However, these studies included eyes with only one form of pathology or mixed pathology with small sample sizes and no subgroup analyses. Only one study by Ho et al. performed a subgroup analysis across differing pathology, corrected for IA and re-performed analyses finding a similar effect of pathology on IA.2

Outer retinal pathology such as subretinal fibrosis, drusen, and geographic atrophy encountered in nAMD were more likely to result in segmentation IA. This was most evident in the Spectralis, which generated differing automated segmentation on repeated scans in this group. Eyes with gross central macular thickening were more likely to have fixation IA, likely due to loss of detection of the normal foveal contour.

Variances in CMT measured may be partly explained by different hardware. The Triton utilizes SS-OCT to obtain images, compared with the Cirrus and Spectralis which both use SD-OCT. The differences in technology between these systems are reviewed elsewhere.20

Additionally, each of these machines utilizes different software to analyze obtained images. Identification of the outer retinal border differs between machines as a line above, through, or below the retinal pigment epithelium (RPE) thus affecting the CMT measurement.11 This can also lead to differing rates of IA, especially in nAMD, which affects the Bruch's-RPE complex.21 Scale calibration may also be an issue with differing standards for what defines 1 μm on a scan. This can be critical as prognosis, classification and treatment guidelines for a wide range of retinal pathology may depend on these measurements.14,22 Fovea finding algorithms, as well as eye tracking software also differs, affecting rates of fixation IA.23

Updates in software may reduce IA but make comparisons between different versions unreliable. An industry standard for measurement formulas and calibrated segmentation algorithms across all available OCT devices would help address these issues.2

There may be a role for a conversion formula to allow for better comparability of macular thickness measurements across different OCT for different pathologies, as has been described by others.24 Indeed, a conversion formula was utilized in the Diabetic Retinopathy Clinical Research Network Protocol T study to compare the time domain and SD-OCT devices used in the study.14 However, these formulae would need to be verified with each update in imaging software to ensure accuracy and validity.

The strengths of this study are that all scans for each patient were formed on the same day, eliminating the effects of temporal changes on pathology. The data were collected prospectively with a defined protocol and evaluates the new Triton SS-OCT, which has not been studied for these purposes.

The pathologies evaluated in this study are limited to nAMD and DME. A more complete comparison, including vitreomacular interface pathology, high myopia, retinal degenerations, and segmentation of individual retinal layers, would be valuable. Although including patients able to fixate on a target minimized the effects of fixation errors, this may be a source of selection bias, limiting the application of these findings.

These study findings validate the presence of interdevice measurement variability and how reliability can be further affected by retinal pathology. Consequently, care should be exercised when interpreting measurements from different OCT devices in clinical practice and research settings. Investigators should be aware of the high rates of IA that occur in DME and nAMD and consider manual correction of these errors in order to accurately report quantitative measurements. Further in-depth studies are required to evaluate the reliability of OCT machines when performing scans in participants with poorer vision and other ocular and retinal pathology.

References

  1. Swanson EA. OCT Technology Transfer and the OCT Market. In: Drexler W, Fujimoto JG, eds. Optical Coherence Tomography: Technology and Applications. 2nd ed. Berlin: Springer; 2015. doi:10.1007/978-3-319-06419-2_86 [CrossRef]
  2. Ho J, Sull AC, Vuong LN, et al. Assessment of artifacts and reproducibility across spectral- and time-domain optical coherence tomography devices. Ophthalmology. 2009;116(10):1960–1970. doi:10.1016/j.ophtha.2009.03.034 [CrossRef]
  3. Pierro L, Giatsidis SM, Mantovani E, Gagliardi M. Macular thickness interoperator and intraoperator reproducibility in healthy eyes using 7 optical coherence tomography instruments. Am J Ophthalmol. 2010;150(2):199–204.e1. doi:10.1016/j.ajo.2010.03.015 [CrossRef]
  4. Wolf-Schnurrbusch UE, Ceklic L, Brinkmann CK, et al. Macular thickness measurements in healthy eyes using six different optical coherence tomography instruments. Invest Ophthalmol Vis Sci. 2009;50(7):3432–3437. doi:10.1167/iovs.08-2970 [CrossRef]
  5. Giani A, Cigada M, Choudhry N, et al. Reproducibility of retinal thickness measurements on normal and pathologic eyes by different optical coherence tomography instruments. Am J Ophthalmol. 2010;150(6):815–824. doi:10.1016/j.ajo.2010.06.025 [CrossRef]
  6. Bressler SB, Edwards AR, Andreoli CM, et al. Reproducibility of Optovue RTVue optical coherence tomography retinal thickness measurements and conversion to equivalent Zeiss Stratus metrics in diabetic macular edema. Transl Vis Sci Technol. 2015;4(1):5. doi:10.1167/tvst.4.1.5 [CrossRef]
  7. Forooghian F, Cukras C, Meyerle CB, Chew EY, Wong WT. Evaluation of time domain and spectral domain optical coherence tomography in the measurement of diabetic macular edema. Invest Ophthalmol Vis Sci. 2008;49(10):4290–4296. doi:10.1167/iovs.08-2113 [CrossRef]
  8. Han IC, Jaffe GJ. Comparison of spectral- and time-domain optical coherence tomography for retinal thickness measurements in healthy and diseased eyes. Am J Ophthalmol. 2009;147(5):847–858.e1. doi:10.1016/j.ajo.2008.11.019 [CrossRef]
  9. Kakinoki M, Miyake T, Sawada O, Sawada T, Kawamura H, Ohji M. Comparison of macular thickness in diabetic macular edema using spectral-domain optical coherence tomography and time-domain optical coherence tomography. J Ophthalmol. 2012;2012:959721. doi:10.1155/2012/959721 [CrossRef]
  10. Leung CK, Cheung CY, Weinreb RN, et al. Comparison of macular thickness measurements between time domain and spectral domain optical coherence tomography. Invest Ophthalmol Vis Sci. 2008;49(11):4893–4897. doi:10.1167/iovs.07-1326 [CrossRef]
  11. Tan CS, Chan JC, Cheong KX, Ngo WK, Sadda SR. Comparison of retinal thicknesses measured using swept-source and spectral-domain optical coherence tomography devices. Ophthalmic Surg Lasers Imaging Retina. 2015;46(2):172–179. doi:10.3928/23258160-20150213-23 [CrossRef]
  12. Comyn O, Heng LZ, Ikeji F, et al. Repeatability of Spectralis OCT measurements of macular thickness and volume in diabetic macular edema. Invest Ophthalmol Vis Sci. 2012;53(12):7754–7759. doi:10.1167/iovs.12-10895 [CrossRef]
  13. Mylonas G, Ahlers C, Malamos P, et al. Comparison of retinal thickness measurements and segmentation performance of four different spectral and time domain OCT devices in neovascular age-related macular degeneration. Br J Ophthalmol. 2009;93(11):1453–1460. doi:10.1136/bjo.2008.153643 [CrossRef]
  14. Wells JA, Glassman AR, Diabetic Retinopathy Clinical Research N et al. Aflibercept, bevacizumab, or ranibizumab for diabetic macular edema. N Engl J Med. 2015;372(13):1193–1203. doi:10.1056/NEJMoa1414264 [CrossRef]
  15. Rofagha S, Bhisitkul RB, Boyer DS, Sadda SR, Zhang KSEVEN-UP Study Group. Seven-year outcomes in ranibizumab-treated patients in ANCHOR, MARINA, and HORIZON: A multicenter cohort study (SEVEN-UP). Ophthalmology. 2013;120(11):2292–2299. doi:10.1016/j.ophtha.2013.03.046 [CrossRef]
  16. Elman MJ, Aiello LP, Diabetic Retinopathy Clinical Research Network et al. Randomized trial evaluating ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology. 2010;117(6):1064–1077.e35. doi:10.1016/j.ophtha.2010.02.031 [CrossRef]
  17. Do DV, Schmidt-Erfurth U, Gonzalez VH, et al. The DA VINCI Study: Phase 2 primary results of VEGF Trap-Eye in patients with diabetic macular edema. Ophthalmology. 2011;118(9):1819–1826. doi:10.1016/j.ophtha.2011.02.018 [CrossRef]
  18. Korobelnik JF, Do DV, Schmidt-Erfurth U, et al. Intravitreal aflibercept for diabetic macular edema. Ophthalmology. 2014;121(11):2247–2254. doi:10.1016/j.ophtha.2014.05.006 [CrossRef]
  19. Nguyen QD, Brown DM, Marcus DM, et al. Ranibizumab for diabetic macular edema: results from 2 phase III randomized trials: RISE and RIDE. Ophthalmology. 2012;119(4):789–801. doi:10.1016/j.ophtha.2011.12.039 [CrossRef]
  20. Podoleanu AG. Optical coherence tomography. J Microsc. 2012;247(3):209–219. doi:10.1111/j.1365-2818.2012.03619.x [CrossRef]
  21. Patel PJ, Chen FK, Ikeji F, et al. Repeatability of stratus optical coherence tomography measures in neovascular age-related macular degeneration. Invest Ophthalmol Vis Sci. 2008;49(3):1084–1088. doi:10.1167/iovs.07-1203 [CrossRef]
  22. Haller JA, Stalmans P, Benz MS, et al. Efficacy of intravitreal ocriplasmin for treatment of vitreomacular adhesion: subgroup analyses from two randomized trials. Ophthalmology. 2015;122(1):117–122. doi:10.1016/j.ophtha.2014.07.045 [CrossRef]
  23. Han IC, Jaffe GJ. Evaluation of artifacts associated with macular spectral-domain optical coherence tomography. Ophthalmology. 2010;117(6):1177–1189.e4. doi:10.1016/j.ophtha.2009.10.029 [CrossRef]
  24. Krebs I, Hagen S, Smretschnig E, Womastek I, Brannath W, Binder S. Conversion of Stratus optical coherence tomography (OCT) retinal thickness to Cirrus OCT values in age-related macular degeneration. Br J Ophthalmol. 2011;95:1552–1554. doi:10.1136/bjo.2010.194670 [CrossRef]

Baseline Characteristics of Study Population

CharacteristicsControl GroupDME GroupnAMD Group
Participants (n)121617
Eyes (n)242326
Female (n)9811
Age, mean years (SD)42 (14.1)62.1 (8.3)83.5 (6.0)
BCVA, mean letters (SD)55.8 (1.9)47.4 (9.3)42.2 (11.2)
Phakic (n)242011

Spectralis OCT
CMT (μm), mean (SD)265.0 (14.5)337.7 (65.9)286.3 (72.9)

Cirrus OCT
CMT (μm), mean (SD)245.9 (14.1)319.8 (66.5)229.6 (70.6)

Triton OCT
CMT (μm), mean (SD)231.5 (11.8)287.4 (69.9)213.2 (66.7)

Mean Difference Between Two Consecutive Scans Stratified by Subgroup Before and After Resegmentation to Correct for Imaging Artifacts

Before ResegmentationAfter Resegmentation
SubgroupOCT MachineMean CMT (μm)95% CIP ValueMean CMT (μm)95% CIP Value
ControlSpectralis−0.88−4.7 to 2.9.640.79−0.6 to 2.2.26
Triton−0.17−1.1 to 0.8.72−0.17−1.1 to 0.8.72
Cirrus−2.38−5.7 to 0.9.15−1.29−4.1 to 1.5.36

DMESpectralis−1.83−5.9 to 2.2.36−3.26−7.7 to 1.1.14
Triton−1.65−6.6 to 3.3.490.61−4.3 to 5.5.80
Cirrus0.13−5.2 to 5.5.96−0.74−7.1 to 5.7.81

nAMDSpectralis10.960.7 to 21.2.04−2.54−7.8 to 2.7.33
Triton0.50−4.9 to 5.9.853.65−1.9 to 9.2.19
Cirrus0.88−5.3 to 7.0.770.73−2.2 to 3.7.62

OverallSpectralis3.04−1.0 to7.1.15−1.67−3.9 to 0.6.15
Triton−0.39−2.7 to 1.9.74−0.40−2.7 to 1.9.74
Cirrus−0.42−3.2 to 2.3.771.41−1.7 to 4.5.77

Incidence of Artifacts by Optical Coherence Tomography Machine and Subgroup

Total Artifact, n (%)Artifact TypeP Value*

OCT MachineSubgroupSegmentationFixation
SpectralisControl1/24 (4.2%)01
DME11/23 (47.8%)56< .01
nAMD22/26 (84.6%)1912< .01

TritonControl2/24 (8.3%20
DME12/23 (52.2%)57< .01
nAMD13/26 (50%)103< .01

CirrusControl2/24 (8.3%)20
DME8/23 (34.8%)32.04
nAMD11/26 (42.3%)111.01
Authors

From Sydney Institute of Vision Science, Sydney (BB, SYPE, TH, MZ, AC); Save Sight Institute, University of Sydney, Sydney (BB, AC); University of New South Wales, Sydney (GO); and Macquarie University, Sydney (KL).

Findings from the study were presented at the Royal Australian and New Zealand College of Ophthalmologists Congress, November 1–4, 2015, in Wellington, New Zealand.

Dr. Chang has acted as a consultant for Bayer, Alcon, and Novartis and received grants from Bayer outside the submitted work. The remaining authors report no relevant financial disclosures.

Address correspondence to Andrew Chang, MBBS (Hons), PhD, FRACS, FRANZCO, Sydney Retina Clinic and Day Surgery, 13/187 Macquarie Street, Sydney, New South Wales 2000, Australia; email: achang@sydneyretina.com.au.

Received: October 04, 2016
Accepted: February 08, 2017

10.3928/23258160-20170329-06

Sign up to receive

Journal E-contents