Aims To investigate the choroidal thickness changes in eyes with retinal vein occlusion (RVO) and to determine the effect of an intravitreal dexamethasone implant (Ozurdex) on choroidal thickness.
Methods This study comprised 47 patients diagnosed with unilateral, RVO-associated macular oedema (MO) and treated with Ozurdex. Baseline subfoveal choroidal thickness (SFCT) and central macular thickness (CMT) of the RVO eyes were compared with those of the normal contralateral eyes; serial changes in this parameter were evaluated at 1, 3 and 5 months after Ozurdex injection. The correlation between SFCT and CMT and between visual acuity change and SFCT or CMT changes were also investigated.
Results The mean SFCT was significantly higher in RVO eyes at baseline (260.3±71.2 μm) than it was in contralateral eyes (217.6±55.3 μm; p<0.001). While the mean SFCT decreased at 1 and 3 months after Ozurdex injection, it slightly increased at month 5. The change in SFCT was significantly associated with the change in CMT at 3 months (r=0.5073; p=0.0006) and 5 months after Ozurdex injection (r=0.3489; p=0.0274). The improvement in visual acuity was significantly correlated with the decrease in SFCT (r=0.3241; p=0.0413).
Conclusions SFCT in eyes with RVO was significantly greater than it was in normal contralateral eyes. After Ozurdex injection, SFCT decreased significantly. Improvement in visual acuity was correlated with the decrease in SFCT after Ozurdex injection.
- Treatment Medical
Statistics from Altmetric.com
Retinal vein occlusion (RVO) is the second most common cause of retinal vascular disease, diabetic retinopathy being the most common cause.1 ,2 The risk factors for RVO reportedly include hypertension, arteriosclerosis, diabetes, smoking, hyperlipidaemia, inflammatory disease and hypercoagulable states.3 The standard of care for macular oedema (MO) in RVO has been grid laser photocoagulation.4 ,5 Recently, however, intravitreal injections of anti-vascular endothelial growth factor (VEGF) and sustained-release dexamethasone implants (Ozurdex) have shown promising results in the treatment of MO following RVO.6–8
Since the advent of optical coherence tomography (OCT) for choroidal imaging in clinical practice,9 significant advancements have been made in the understanding of the pathophysiology of the choroid in retinochoroidal diseases. Increments in subfoveal choroidal thickness (SFCT) have been reported to occur in patients with central serous chorioretinopathy,10 ,11 Vogt–Koyanagi–Harada disease12 and polypoidal choroidal vasculopathy.13 ,14 Moreover, the increased choroidal thickness was reported to decrease after various treatments.11 ,15 ,16 It was also reported that choroidal thickness was greater in eyes with central RVO (CRVO) than in the normal contralateral eyes.17 In contrast, the results of the Beijing Eye Study18 showed that SFCT in eyes with longstanding RVO did not differ from that in normal contralateral eyes. Therefore, it still remains unclear whether SFCT is subject to change in eyes with acute-onset RVO as well as branch RVO (BRVO).
Ozurdex has been used as a treatment option for MO following RVO. A randomised, sham-controlled trial8 revealed that the percentage of eyes with a ≥15-letter improvement in best corrected visual acuity (BCVA) was significantly higher in the Ozurdex groups compared with that in the sham treatment group at days 30–90. However, there were no previous reports of changes in SFCT after intravitreal dexamethasone implant injections in eyes with RVO.
In this study, we aimed to both evaluate and compare the choroidal thickness between eyes with RVO (including BRVO) and unaffected, normal contralateral eyes, and to determine the effect of intravitreal dexamethasone implants on choroidal thickness.
Materials and methods
This prospective study was conducted on patients who required intravitreal dexamethasone implant (Ozurdex; Allergan, Irvine, California, USA) injections for the treatment of MO associated with RVO at Seoul National University Hospital during the period from February 2012 to February 2013. Informed consent was obtained from each patient prior to enrolment in the study and the Institutional Review Board of Seoul National University Hospital approved the study protocol. This study was conducted in accordance with the tenets of the Declaration of Helsinki.
The diagnosis of RVO was made by the physician via indirect ophthalmoscopic fundus evaluation and fluorescein angiography. MO was defined as a central subfield retinal thickness (central macular thickness, CMT) ≥300 μm using the 512×128 macular cube mode on the OCT (Cirrus high-definition (HD) OCT, Model 4000; Carl Zeiss Ophthalmic Instruments, Dublin, California, USA; 128 lines, 512 A-scans per line, scan area of 6×6 mm). The major study exclusion criteria were as follows: bilateral RVO; diabetic retinopathy; conditions that could affect choroidal thickness (eg, high myopia (spherical equivalent ≥−6.0 dioptres or axial length (AL) ≥26 mm), uveitis and a previous history of central serous chorioretinopathy or photodynamic therapy); media opacities that prevented assessment of the fovea (eg, cataract of more than Emery–Little classification grade III, vitreous haemorrhage and corneal opacity); previous intravitreal injections of triamcinolone acetonide or bevacizumab (Avastin; Genentech, South San Francisco, California, USA) ≤3 months prior to enrolment; and any history of intraocular surgery.
Examinations and data collection
A complete ocular examination, including measurement of BCVA on a Snellen chart, tonometry, AL (Axis II PR; Quantel Medical, Bozeman, Montana, USA) measurement, slit-lamp biomicroscopy, dilated fundus examination, fundus photography (Vx-10; Kowa Optimed, Tokyo, Japan), fluorescein angiography (Vx-10; Kowa Optimed) and OCT imaging, was conducted on each patient. After the Ozurdex injection, measurement of BCVA, tonometry, slit-lamp biomicroscopy, dilated fundus examination and OCT were repeated at each follow-up visit. The first follow-up visit was scheduled at 1 month after the injection and all subsequent visits were scheduled every other month thereafter. Measurements for BCVA were converted to logarithm of the minimum angle of resolution (logMAR) units for statistical analyses.
Choroidal thickness was measured by enhanced depth imaging-OCT using a Cirrus HD-OCT. The scan pattern selected was the 5-line raster, which is a 6 mm line consisting of 20 480 A-scans with an imaging speed of 27 000 A-scans per second, for an average of 20 frames (B-scans). The location of the fovea within each volume scan was identified automatically with the built-in Fovea Finder algorithm of the Cirrus, resulting in good reproducibility. The resultant images were viewed and measured using the built-in software (V.126.96.36.199; Carl Zeiss Ophthalmic Instruments). The choroid was measured from the outer edge of the retinal pigment epithelium (RPE) to the inner edge of the suprachoroidal space below the fovea. Measurements of SFCT were obtained at the subfoveal region via manual measurement using callipers provided by the instrument's software. The retinal specialist who was masked to the clinical data performed the analysis. Comparisons between the eyes were made using the measurements for both eyes in the corresponding visit.
A paired t test was used to compare CMT and SFCT in eyes with RVO and normal contralateral eyes. Serial comparisons of the mean CMT and SFCT were conducted using Wilcoxon signed-rank test. Comparisons of the mean CMT and SFCT between the two subgroups—BRVO and CRVO—were analysed using the Mann–Whitney U test. The correlations between the CMT and SFCT and between the change in visual acuity and changes in CMT or SFCT were analysed using the Pearson's test. The statistical significance level was set as p<0.05. All statistical analyses were performed using SPSS V.21.0 for Windows (SPSS, Chicago, Illinois, USA).
Forty-seven consecutive patients who were treated with Ozurdex for MO in RVO were included. Among these, 5 were excluded due to poor image quality; 42 patients were ultimately included in the study. The total number of female patients was 25 (59.5%) and the mean age was 57±13 years (range, 25–80 years). The mean AL was 24.18±0.67 (range, 23.03–25.29). The types of RVO that presented were BRVO in 21 (50%) patients and CRVO in 21 (50%) patients. Among the 42 patients, 21 (50%) had hypertension and 4 (10%) had diabetes mellitus without retinopathy (table 1). In normal contralateral eyes, the mean AL was 24.39±0.71 (range, 23.18–25.56) and there were no significant interocular differences in AL (p=0.303).
Changes in retinal and choroidal thickness after Ozurdex injection
Compared with unaffected contralateral eyes, eyes with RVO showed a significantly higher baseline CMT (p<0.001). The mean baseline CMT was 486.6±154.4 μm (range, 273–800 μm) in the eyes with RVO and 249.3±21.8 μm (range, 193–305 μm) in the contralateral eyes (figure 1A). The mean baseline SFCT in the eyes with RVO (260.3±71.2 μm (range, 102–355 μm)) was also significantly higher than it was in the contralateral eyes (217.6±55.3 μm (range, 78–303 μm), p<0.001; figure 1B).
While the mean CMT significantly reduced at 1 month after Ozurdex injection (291.3±74.0 μm, p<0.001), it began to gradually increase at months 3 and 5. Nevertheless, the mean CMT values at 3 months (321.5±110.6 μm, p<0.001) and 5 months (373.8±152.5 μm, p<0.001) were significantly lower than the mean CMT at baseline (figure 2A). The mean SFCT also decreased after Ozurdex injection at 1, 3 and 5 months (239.0±73.8, 228.2±63.1 and 232.9±66.9 μm, respectively; all p<0.001). The mean SFCT at 5 months slightly increased from what it was at 3 months (figure 2B). The serial mean values of CMT, SFCT and BCVA over time are presented in table 2. In addition, figure 3 shows a representative case of CRVO with MO treated with an Ozurdex implant.
Despite the differences in measurements between eyes with RVO and normal contralateral eyes, the mean CMT and SFCT (both at baseline and after Ozurdex injection) were not significantly different between eyes with BRVO and eyes with CRVO. At baseline, the mean CMT was 470.4±147.5 μm in eyes with BRVO and 502.9±163.0 μm in eyes with CRVO (p=0.546). The mean CMT measured at 5 months after Ozurdex injection was 366.8±141.1 μm in eyes with BRVO and 380.9±166.3 μm in eyes with CRVO (p=0.660). At baseline, the mean SFCT was 260.0±66.3 μm in eyes with BRVO and 260.6±77.6 μm in eyes with CRVO (p=0.862). The mean SFCT at 5 months after Ozurdex injection was 228.2±71.3 μm in eyes with BRVO and 237.6±63.7 μm in eyes with CRVO (p=0.718).
Correlation of retinal thickness and choroidal thickness
Univariate analyses using Pearson's correlation test were performed to evaluate the correlation between CMT and SFCT. The ratios of the CMT and SFCT of the affected eye to those of the contralateral eye for the corresponding visit were used to standardise the effects of individual differences. The relationships between the SFCT ratio and CMT ratio at baseline (r=0.1487; p=0.3662) and 1 month after Ozurdex injection (r=0.2788; p=0.0815) were not statistically significant. However, the correlations between these two ratios were significant at 3 months (r=0.4562; p=0.0031) and 5 months after Ozurdex injection (r=0.4015; p=0.0102, figure 4). The correlation was most significant at 3 months after injection when the retina and choroid are most compact in the affected eye.
The relationships between the changes in SFCT and CMT from baseline were also evaluated. The correlation was only significant at 3 months (r=0.5073; p=0.0006) and 5 months after Ozurdex injection (r=0.3489; p=0.0274). No significant correlation was found at 1 month after injection (r=0.0114; p=0.9441, figure 5A–C). A subgroup analysis was also performed in a group of eyes that showed a decrease in CMT at both 1 and 3 months compared to baseline values to assess the effects of steroid in the corticosteroid-responder group. The correlations between the change in SFCT and CMT were significant at 3 months (r=0.3868; p=0.0217) and 5 months after Ozurdex injection (r=0.4180; p=0.0155). No significant correlation was found at 1 month after injection (r=−0.0693; p=0.7016, figure 5D–F). There was no significant correlation between the changes in intraocular pressure and SFCT at 1 month (r=−0.1999; p=0.2162), 3 months (r=−0.1402; p=0.3758) and 5 months after Ozurdex injection (r=−0.2850; p=0.0746).
Impact of choroidal thickness on visual acuity
Univariate analysis employed Pearson's correlation test and was performed in an effort to determine the correlation between changes in CMT or SFCT and changes in visual acuity at 5 months after Ozurdex injection. The improvement in visual acuity was significantly correlated with a decrease in SFCT (r=0.3241; p=0.0413). A similar but non-significant trend was seen between the changes in visual acuity and change in CMT (r=0.2126; p=0.0764) (figure 6).
This study demonstrated that choroidal thickness in RVO eyes with MO was higher than it was in normal contralateral eyes. In addition, injection of an intravitreal dexamethasone implant decreased both choroidal thickness and central retinal thickness. After Ozurdex injection, the mean CMT gradually increased from month 3 and the mean SFCT increased from month 5. These results suggest that a change in choroidal thickness follows a change in retinal thickness, and they also suggest that it is a change in the retina that accounts for choroidal thickening in eyes with RVO.
Mean SFCT, as measured by Tsuiki et al17 in eyes with CRVO, was 257.1±83.2 μm, which was greater than that in normal contralateral eyes (222.6±67.8 μm). Mean SFCT decreased from 266.9±79.0 to 227.7±65.1 μm after intravitreal bevacizumab injection. In this study, mean SFCT in eyes with RVO was 260.3±71.2 μm, which was greater than that in normal contralateral eyes (217.6±55.3 μm). After Ozurdex injection, these values decreased to 239.0±73.8 and 228.2±63.1 μm at 1 and 3 months, respectively. This finding, which even included eyes with BRVO, was consistent with the findings of Tsuiki et al.17 In contrast, there is a contradictory report on choroidal thickness in eyes with RVO. The Beijing Eye Study,18 a population-based, cross-sectional study that was conducted in northern China, showed that there was virtually no difference in SFCT in eyes with RVO and normal contralateral eyes. This discrepancy between the Beijing Eye Study and our study may be explained by the difference in baseline patient characteristics. All eyes in our study exhibited recently developed RVO with MO, whereas in the Beijing Eye Study there was no instance of marked cystoid MO on OCT images and no case of recent-onset RVO. However, the Beijing Eye Study and our study did yield the similar finding that the SFCT in eyes with BRVO and eyes with CRVO did not significantly differ. The reason for the finding that no significant difference in SFCT existed between eyes with BRVO and eyes with CRVO is unclear, but it could be attributed to the effects of several relatively severe cases of BRVO with considerable MO in our study.
Although the exact mechanism that causes an increase in choroidal thickness in eyes with RVO and a decrease in choroidal thickness after Ozurdex injection was not determined, we hypothesised the following: The choriocapillaris has fenestrations that allow for the outflow of large molecules and increases the amount of material leaving the capillaries. Soluble VEGF isoforms can increase vascular permeability and induce fenestrations of the choriocapillaris.19 In hypoxic RVO eyes, VEGF expression is increased in retinal endothelial cells, pericytes, RPE, Müller cells, ganglion cells and astrocytes.20 VEGF induces choroidal vascular hyperpermeability,17 which in turn increases choroidal thickness. Choroidal thickening is also mediated by the vascular dilation induced by nitric oxide production, which in turn is triggered by VEGF expression.19 All of these factors can cause vasodilation and oedema that can then lead to choroidal thickening. Meanwhile, McAllister et al20 demonstrated that triamcinolone downregulates VEGF expression and also inhibits an increase in the expression of glial fibrillary acidic protein, a protein associated with retinal vascular permeability, in an animal model of BRVO. Corticosteroids are also known to reduce tissue oedema and inhibit the synthesis of endothelial nitric oxide synthase.21 The mechanism by which the intravitreal dexamethasone implant reduces choroidal thickness in eyes with RVO may involve a similar pathway.
While the correlation between the SFCT and CMT ratios at baseline was not significant, the correlations between these two ratios were significant at 3 and 5 months after Ozurdex injection. Additionally, the changes in SFCT and CMT at 3 and 5 months after Ozurdex injection correlated significantly in this study. Our results suggest that even though the degree of retinal thickening at baseline does not necessarily reflect the choroidal thickening in eyes with acute-onset RVO, the responsiveness of these two layers to corticosteroid do correlate with each other. Our study also showed that the serial changes of retinal and choroidal thicknesses in RVO eyes treated with Ozurdex were different, demonstrating delayed response of choroid to Ozurdex. At this time, it is unclear why the choroidal thickness decreases slower than the retinal thickness after Ozurdex injection. Meanwhile, the slower change of choroidal thickness over 3 months is comparable to that of visual improvement after Ozurdex injection in our study. The GENEVA Study reported that visual acuity continued to improve over 3 months after Ozurdex injection, suggesting that factors other than changes in central retinal thickness may be affecting visual acuity in RVO eyes treated with Ozurdex.8 Choroidal thickness might be considered as one of the factors. Based on the fact that the choroid plays an important role in maintaining the RPE and the outer retina,22 the choroid is considered important for visual acuity. Eyes with a relatively thicker choroid at baseline may have more choroidal blood supply and choriocapillaris, which could increase the possibility of recovery after corticosteroid treatments.
This study has several limitations. First, the sample size was small and the follow-up periods were short. Further studies with increased patient numbers and a longer follow-up duration will be needed to confirm our results. Second, manual measurements of choroidal thickness always have a potential bias. Automated software is required for a more objective and definitive evaluation.
In conclusion, SFCT in eyes with RVO is significantly greater than in normal contralateral eyes. Injection of an intravitreal dexamethasone implant results in reduction of choroidal thickness and also leads to the resolution of MO. The improvement in visual acuity was significantly correlated with a decrease in SFCT after Ozurdex injection.
EKL and JMH contributed equally to the work.
Contributors Conception and design of the study (EKL, JMH, HGY); collection of data (EKL, JMH); management of data (EKL, JMH); analysis and interpretation of data (EKL, JMH); drafting of manuscript (EKL, JMH); review of manuscript (EKL, JMH, JYH, HGY); approval of the manuscript (JYH, HGY).
Funding This work was supported by Grant No. 06-2014-2370 from the Seoul National University Hospital Research Fund. The funder had no involvement in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.
Competing interests None declared.
Patient consent Obtained.
Ethics approval Seoul National University Hospital approved this study.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.