Correlations between choroidal thickness and renal function in patients with retinal vein occlusion

The purpose of this study was to analyze the correlation of renal function indices with sub-foveal choroidal thickness (SFChT) in treatment-naïve (naïve) eyes with retinal vein occlusion (RVO) using swept-source optical coherence tomography (SS-OCT) and systemic workup. Retrospective chart review was performed from Dec 2016 to Sep 2019 in patients newly diagnosed with treatment-naïve unilateral RVO. Ocular parameters, including SFChT, and systemic profiles, including renal function indices, were reviewed. Simple and multiple linear regression analyses were performed to check if there was a correlation between renal profiles and SFChT. A total of 56 patients were included in the study; 34 of them were branch RVO and 22 were central RVO patients. Multiple linear regression analysis revealed that SFChT was positively correlated with estimated glomerular filtration rate (eGFR) (P < 0.001). SFChT showed significant correlation with renal function indices. In addition, choroidal thickness may be considered as systemic biomarkers for renal function.


Discussion
In this study, we evaluated the correlation of renal function with choroidal thickness using SS-OCT. This study revealed that choroid thickness was closely correlated with eGFR in patients with RVO, independently of age and other potentially confounding factors.
The correlation of choroidal thickness with renal function has been rarely described with reports only by few studies [10][11][12] . Balmforth et al. reported that chorioretinal thinning in patients with CKD is associated with lower eGFR and greater proteinuria 10 . Mule et al. described a correlation between SFChT and eGFR in patients with HTN 11,12 . In patients with retinal disease, however, a close correlation has not been identified, even though patients with retinal disease frequently suffer from systemic diseases such as HTN or CKD. Moreover, RVO is a representative disease that is affected by the systemic condition. Thus, we investigated the systemic effect on choroidal thickness in eyes with RVO. Furthermore, there have been conflicting results for choroidal thickness in eyes with RVO. Du et al. reported that SFChT was significantly thinner in the RVO group than in the control group 8 . Tang et al. reported that SFChT was significantly thicker in eyes with RVO than in fellow eyes 9 . These www.nature.com/scientificreports/ inconsistencies in SFChT in eyes with RVO suggest that it is affected by some confounding factors 8,9 . Thus, these are the reasons why we have tried to reveal the effects of systemic factors in patients with RVO. BRVO and CRVO are different diseases. However, they have similar phenotypes, similar pathophysiology, and common cardiovascular risk factors 15,16 . The primary purpose of this study is to reveal the association between choroidal thickness in normal contralateral eyes and systemic states because macular edema and RVO affect choroidal thickness. Thus, we considered BRVO and CRVO under the same category. CRVO usually involves the macula, whereas BRVO may not always involve the macula. Therefore, macular involvement by RVO may affect choroidal thickness. In this study, however, there were no BRVO patients in whom the macula was not involved.
Elevated arterial pressure is a major risk factor of both RVO and kidney damage 3,13,17 . The glomerulus is the target structure of hypertensive kidney damage due to its extensive vascular network 18 . The choroid is also a highly vascularized tissue and is very similar to the glomerulus in terms of anatomical structure and physiological function 19,20 . Moreover, both vascular structures are affected by the regulation of localized and systemic renin-angiotensin-aldosterone hormonal cascades 14,21 . The anatomic and physiologic similarity of the tissues may lead to simultaneous damage.
The retina has been considered an important organ for non-invasive evaluation of microvasculature through the diagnosis of diabetic or hypertensive retinopathy 7,22,23 . Our findings suggest that SFChT may be used as a  www.nature.com/scientificreports/ potential biomarker for subclinical or clinical renal dysfunction, even though the parameters would require adjustment for age, axial length, and macular edema.
In addition, we found that SFChT was greater in eyes with RVO than in normal fellow eyes. Several studies reported that the choroid was significantly thicker in eyes with RVO or the occlusive area than in contralateral eyes or the non-occlusive area 9,24-26 . Our results were consistent with these studies. It is known that that the choroid is regulated by autonomic and sensory innervation 27 . Thus, the first possible explanation for thicker choroid in RVO eyes is that local autoregulatory myogenic mechanism (associated with vascular smooth muscles) caused by ischemic retina leads to compensatory increase in choroidal blood flow and consequently to choroidal thickening 27 . The second possible explanation is that the choroid becomes edematous passively by the intravitreal elevation of the vascular endothelial growth factor (VEGF) level. Several studies showed that choroidal thickness in eyes with RVO decreased after anti-VEGF injection 24,28 . From these previous studies, we can hypothesize that the choroid tissue is sensitive to intravitreal VEGF concentration. Accordingly, in this study, we used contralateral eyes without RVO to investigate if there is a correlation between SFChT and renal function. However, in a report by Du et al., SFChT was significantly thinner in the group with RVO than in the group without RVO in a large population-based study 8 . We additionally compared eyes with RVO and normal control eyes. SFChT was greater in eyes with RVO than in normal control eyes. However, the difference was not significant. In addition, SFChT was lesser in contralateral eyes than in normal control eyes. However, the difference was not significant here either. The systemic status of RVO patients may act as a confounding factor and lead to a large standard deviation in SFChT in patients with RVO. Consequently, this may lead to no significant statistics. Thus, in this study, when compared with contralateral eyes, eyes with RVO had a thicker choroid. However, eyes with RVO did not have a thicker choroid than normal control eyes. Interestingly, this discrepancy is similar to the previous conflicting results 8,9 . Given the effect of renal function on the choroid, the SFChT of eyes with RVO should be compared with that of contralateral eyes without RVO.
Additionally, Rayess et al. reported baseline SFChT as a predictor of improvement in visual acuity after anti-VEGF injection in both BRVO and CRVO 29,30 . The baseline SFChT of functional responder (BCVA gain ≥ 2 lines) was greater than that of nonresponders in BRVO and CRVO 29,30 . Although Rayess et al. did not consider renal function, we carefully suggest that functional recovery in patients with RVO is associated with renal function through the results of the present study.
This study has several limitations. First, this was a retrospective observational study. Thus, the timing of OCT was not controlled (diurnal variations were not considered). Second, this study has relatively few patients because systemic evaluation was not routinely performed in patients with RVO. Third, only patients whose systemic workup was performed within a span of four weeks from initial retinal evaluation were included. This interval could affect the systemic status. Fourth, this study did not investigate the relationship between functional outcome and renal dysfunction because of the cross-sectional nature of this study. Fifth, axial length was not considered although myopic and hyperopic eyes were excluded. Finally, retinal hemorrhage and edema may have affected the measurements of SFChT of eyes with RVO. However, multiple regression analysis for our key findings used normal contralateral eyes. A prospective future study should provide stronger evidence for correlation between SFChT and renal dysfunction.
In conclusion, SFChT was positively correlated with eGFR. It is the first glance to reveal the association between choroidal thickness and kidney function in treatment-naïve RVO patients.

Methods
Study design. This retrospective cross-sectional study was approved by the institutional review board committee of Chung-Ang University Hospital in Seoul, South Korea (IRB #1701-004-16029) and adhered to the tenets of the Declaration of Helsinki. Informed consent was waived by the institutional review board committee of Chung-Ang University Hospital in Seoul because of the retrospective nature of the study.

Study subjects.
The medical records of newly diagnosed patients with treatment-naïve unilateral RVO (macular involving) at Chung-Ang University Hospital between Dec 1, 2016, and Sep 30, 2019, were retrospectively reviewed. Among them, only patients whose renal function was evaluated within a span of 4 weeks from retinal evaluation were selected. The following parameters were recorded from medical charts: age, gender, mean arterial pressure (mmHg), BMI (km/m 2 ), HTN duration, serum level of glucose, serum osmolarity [(2 × (sodium + potassium)) + (blood urea nitrogen (BUN)/2.8) + (glucose/18)], and serum level of total cholesterol.
In addition, the following renal parameters were recorded from laboratory examination: BUN, creatinine, BUN-to-creatinine ratio (BUN/Cr), eGFR, phosphorus, and calcium. The exclusion criteria were as follows: other chorioretinal disorders including drusen, age-related macular degeneration, diabetic retinopathy, eyes with treatment history using laser and intravitreal injection, myopic or hyperopic eyes with refractive error > ± 3.0 D, and eyes with low-quality OCT images due to media opacity (i.e., low image quality index < 90). In addition, patients with a history of smoking, glaucoma, ocular trauma, ocular inflammation, ocular ischemic syndrome, or any type of intraocular surgery (except cataract surgery) were excluded as well. The patients with systemic disease including history of angina, acute myocardial infarction, acute coronary syndrome or coronary revascularization (with a stent or coronary artery bypass graft), or significant plaque on coronary angiography, acute or uncontrolled hypertension (systolic BP ≥ 150 mmHg, diastolic BP ≥ 90 mmHg), significant carotid artery stenosis, history of a carotid endarterectomy were excluded as well.
ophthalmic examination and diagnosis of retinal vein occlusion. Patients  optical coherence tomography. OCT imaging was done using Deep Range Imaging (DRI)-OCT Triton (Topcon, Tokyo, Japan). The device provided a 7 μm axial and 20-μm horizontal resolution and a scan speed of 100,000 A-scans per second using a swept wavelength light source. Central RT and ChT were defined and measured using the built-in software as previously described (Fig. 2) 31 . ChT was measured additionally at 1000-μm intervals from the fovea to 3000 μm to nasal and temporal area.
Statistical analysis. Two independent masked observers (OJY and JTK) measured ChT, and the averaged values were used for statistical analyses. Data are presented as means ± standard deviations. Statistical analyses were performed using SPSS version 25.0 software (IBM Corp., Armonk, NY, USA). The baseline characteristics of patients were compared with those of normal subjects using independent t-tests. Ocular parameters of eyes with RVO were compared with those of normal fellow eyes (without RVO) using paired t-test. The correlations of SFChT with renal profile were analyzed using simple and multiple linear regression analyses. P values < 0.05 were considered statistically significant.

Data availability
The datasets during and/or analysed during the current study are available from the corresponding author on reasonable request.