The Systemic and Ocular Risk Factors for Retinal Vein Occlusion: A Retrospective Study

BACKGROUND: We sought to evaluate the systemic and ocular risk factors for severity on visual acuity and central retinal thickness in macular edema secondary to retinal vein occlusion (RVO-ME). METHODS: This retrospective study included 46 RVO-ME patients in The First A�liated Hospital of Sun Yat-sen University from January 2015 to November 2019. Systemic examinations include blood pressure, blood glucose, blood lipids, vascular endothelial function, and carotid artery color ultrasound. Ocular examinations include the best-corrected visual acuity (BCVA) and the central retinal thickness (CRT). The integrity of the outer retina was evaluated as well. According to the baseline BCVA and CRT levels, the patients were divided into high vision group and low vision group, high CRT group, and low CRT group. Multivariate logistic regression analyses were performed to analyze the risk factors on baseline BCVA and CRT. RESULTS: We enrolled 19 eyes of CRVO (central retinal vein occlusion) and 27 eyes of BRVO (branch retinal vein occlusion). We identi�ed 31 (67.4%) as high CRT and 23 (50.0%) as poor VA of 46 patients on admission. There were 15 cases of BRVO in the high CRT group (48.4%) and 12 cases in the low CRT group (80.0%). The type of disease (BRVO/CRVO) was an independent factor of baseline CRT (P=0.017). Endothelial dysfunction correlates with baseline BCVA independently (P=0.038). Ellipsoidal zone (EZ) destruction was found in 19 cases (82.6%) in the low vision group and 6 cases (26.1%) in the high vision group. EZ integrity correlates with baseline BCVA independently (P=0.017). CONCLUSION: The central retinal vein occlusion (CRVO) has markedly higher CRT than branch retinal vein occlusion (BRVO). Endothelial dysfunction and disrupted ellipsoidal zone were signi�cantly associated with poor baseline VA on admission.


Background
Retinal vein occlusion (RVO) is the second common retinal vascular disease, secondary to diabetic retinopathy.RVO causes sudden painless visual loss accompanied by retinal hemorrhages, retinal edema, and venous engorgement, and vascular tortuosity.The causes of visual loss in RVO include macular edema (ME), ischemia, and the presence of exudates and hemorrhages [1,2].RVO includes central RVO (CRVO), branch RVO (BRVO), and hemiretinal RVO (HRVO), which are a group of diseases caused by obstruction of venous ow.Because vascular endothelial growth factor (VEGF) has a signi cant role in neovascularization of ME in RVO [3], the injection of anti-VEGF drugs is an effective way of treatment of ME secondary to RVO.Poor nal visual acuity after treatment with anti-VEGF agents in RVOs is related to several predictive factors at baseline, including the presence of intraretinal uid, cystoid type of macular edema among others [4].In patients with CRVO, macular thickness and integrity of the ellipsoidal zone (EZ) have been observed to correlate with baseline visual acuity and prognosis [5].Baseline visual acuity is also a predictor of visual outcome after the resolution of ME in CRVO [6].
RVO also has several systemic risk factors.From RVO's pathogenesis, Virchow's triad for thrombosis plays an important role, including hemodynamic change, degenerative vessel wall, and blood hypercoagulability.Besides, systemic diseases such as hypertension, diabetes mellitus, and hyperlipidemia are strongly associated with the occurrence of RVO [7].Atherosclerotic-associated diseases are also linked to RVO, including carotid plaque, obesity, and cigarette smoking [8].In Grave's disease, thyroid eye disease can be an unusual risk factor for RVO, so as in other retrobulbar compressive pathology [9].However, the relationship between baseline CRT and VA in ME secondary to RVO and these well-known risk factors was not clear.These studies aimed to elucidate the impact of systemic and ocular risk factors on baseline visual acuity and OCT parameters.

Methods
We conducted a retrospective study including 46 patients with treatment-naive RVO-ME referred consecutively from January 2015 to November 2019 in the department of Ophthalmology, the First A liated Hospital of Sun Yat-sen University.The CRVO and BRVO diagnosis were based on the International Classi cation of Diseases, Tenth Revision (ICD-10) code H34.813 and H34.833, respectively [10].This study was approved by the Independent Ethics Committee for Clinical Research and Animal trials of the First A liated Hospital of Sun Yat-sen University.This study followed the tenets of the Declaration of Helsinki.
The inclusion criteria included: Only patients with clinical ndings consistent with a diagnosis of RVO (acute vision loss, diffuse intraretinal hemorrhage, venous tortuosity) and macular edema and age ≥ 18 years old were included in the studies.The exclusion criteria included: patients who received any form of surgical or medical treatment; patients who exhibit other complications such as neovascularization of the iris, neovascularization of the angle, and vitreous hemorrhage; patients who have other retinal diseases such as diabetic retinopathy, hypertensive retinopathy, and retinal detachment.
Patient charts were reviewed to collect the following data: age, sex, history of hypertension, diabetes, hyperlipemia, smoking, operation history, the date of presentation probable duration of RVO (as determined by subtracting the subjective symptom onset from the date of clinical presentation), Snellen best-corrected visual acuity (BCVA) based on subjective refraction on admission.Snellen BCVA was converted to logarithm of the minimum angle of resolution (log MAR): Log MAR = lg (1/Snellen BCVA).The log MAR value for counting ngers visual acuity was assigned as + 2.0 log MAR according to methods published by Holladay [11].We also recorded hematological examination (concentration of blood glucose and blood lipid) and results of carotid artery color ultrasound and brachial artery color ultrasound.The subjects were classi ed as hypertension when systolic blood pressure > 140 mmHg and/or diastolic blood pressure > 90 mmHg according to the guidelines of The European Society of Hypertension/European Society of Cardiology or if they reported taking antihypertensive medication, as veri ed by the interviewer [12].Diabetic subjects were de ned in line with the American Diabetes Association [13] or based on self-reported data.Dyslipidemia was de ned following the criteria of the ATP III Expert Panel of the US National Cholesterol Education Program, as total cholesterol > 240 mg/dL or triglyceride > 200 mg /dL [14], or has been diagnosed as hyperlipidemia and was currently taking hypolipidemic drugs.Smoking history was de ned as smoking ≥ 5 packs per week for more than one year.

Brachial artery color ultrasound
We examined the endothelial function of the brachial artery using an established method [15].In brief, subjects were instructed to rest in a supine position for 15 mins, after which ow-mediated vasodilation (FMD) was assessed in the right arm using high-resolution ultrasonography (Phillips HD15, USA), The brachial artery was scanned laterally, and its diameter at the diastole end (i.e. from the inner border of the adventitia to the inner border of the adventitia) was measured.The cuff was positioned 5 cm proximal to the antecubital fossa.A view of a 5-cm transverse section of the brachial artery was recorded for periods of 30 sec at baseline and during peak reactive hyperemia (up to 3 min after cuff release following de ation of the blood pressure cuff, which had been previously in ated around the forearm to 50mm Hg above systolic blood pressure for 5min).The vessel diameter was calculated automatically using built-in software.FMD was de ned by the following formula: FMD (%) = {(maximal artery lumen diameter after cuff release-artery lumen diameter at baseline)/ artery lumen diameter at baseline} × 100 [16].Endothelial dysfunction was de ned as FMD (%) 10% [17].

Carotid artery color ultrasound
All patients' tests were conducted by one physician from the Ultrasonography Department of The First A liated Hospital of Sun Yat-sen University using diagnostic ultrasound equipment (Phillips HD15, 50-MHz probe, USA).Ultrasound longitudinal images of the common carotid artery were acquired at the end of diastole, in which the far wall intima-media interface was clearly de ned [18].The leading edge of the intima and the media-adventitia interface were traced as continuous lines, and mean IMT values were calculated automatically.A carotid plaque was de ned as a localized protruded lesion with a thickness of IMT of the carotid artery ≥ 1.5 mm [19].
OCT imaging protocol SD-OCT (Spectralis, Heidelberg Engineering Inc, CA, USA) was performed on all patients during the initial examination.The integrity of the outer retinal layers was analyzed in the horizontal and vertical scans centered on the fovea, including the external limiting membrane (ELM), ellipsoid zone (EZ), and retinal pigment epithelium (RPE).For example, ELM was graded as "disturbed" if we were unable to follow the hyperre ective zone of the ELM in an area measuring 200 µm or more, regardless of whether it was in the horizontal or vertical SD-OCT scan [20].Similarly, the integrity of EZ and RPE was assessed (Fig. 1a-c).
Experienced operators manually measured the central retinal thickness (CRT) as the distance between the inner limiting membrane (ILM) of the macular fovea and the hyperre ective inferior limit of retinal pigment epithelium (RPE) on B-scan through the fovea, using a caliper integrated into the device (Fig. 2).
According to the central retinal thickness of the macula, the enrolled eyes were classi ed into the high CRT group (CRT ≥ 440 µm) and the low CRT group (CRT 440µm).According to the degree of impaired vision, the enrolled eyes were classi ed into the good VA group (BCVA ≥ 1.0 Log MAR) and the poor VA group(BCVA 1.0).

Statistical analysis
Statistical analysis was performed using SPSS for Windows software, version 20.0 (SPSS Inc, Chicago, Illinois USA).Analysis of Pearson χ2 test was used to compare the risk factors in different groups.Univariate analyses are followed by multivariate logistic regression analyses, modeling visual condition (BCVA) and morphological condition (CRT) as a binary variable which were performed to determine risk factors for visual loss and macular pathology.A P value of 0.05 or less was considered statistically signi cant.

Discussion
In this study, we evaluated risk factors for macular edema secondary to RVO.RVO type was the strongest independent factor for higher CRT (CRT ≥ 440µm).Besides, disrupted ELM and EZ was found to be a signi cant risk factor by univariate analysis.There was no difference in gender, age, duration, systemic diseases, smoking habits, vascular endothelial function, carotid plaque and disrupted ELM, EZ, and RPE.These results indicate that patients with CRVO may potentially be more likely to suffer more serious macular edema than patients with BRVO.
Prior studies have shown that there are differences in pathological mechanisms between CRVO and BRVO.Rachel et al. reported visual acuity was generally poor at baseline ( 20/40) and decreased further over time in CRVO [21].It has been demonstrated that CRVO eyes had a higher ischemic index and VEGF level compared with BRVO eyes [22].Spaide et al. reported that increased VEGF would induce dilated macular capillaries and hyperpermeability [23], which may explain why patients with CRVO had a thicker central retina than BRVO.
Our study demonstrates that endothelial dysfunction and disrupted EZ associate with worse visual acuity in RVO patients.Spaide et al. reported that endothelial dysfunction is an independent risk factor for BRVO, however, CRVO and smoking patients were not included in their studies [23].The endothelium plays an important role in vascular function and prior studies suggest that abnormal vascular endothelium and arteriosclerosis are risk factors for RVO [24,25].Generated from vascular endothelial cells, Nitric oxide (NO) is an important signal molecule, regulating local blood ow [26].In RVO with endothelial dysfunction, decreased NO level might reduce the retinal blood ow and increase platelet aggregation, which may have negative effects on visual acuity [27].In experimental BRVO eyes, decreased vitreous NO level and narrowing of retinal arteries can be observed, supporting that impairment in the release of NO may contribute to the development of hypoxia and necrosis in the affected retina [28].However, further investigation is needed to demonstrate it.
Our studies also show that disruption of EZ correlated with poorer visual acuity.The EZ is referred to the hyperre ective band between ELM and RPE in OCT, which is used for the evaluation of photoreceptor health.Its disruption correlates with poor vision in various diseases including RVO.It is reported that photoreceptor loss and EZ loss are the predictive factors for poor visual outcome and a large extent of macular edema in RVO [5,29].Touka et al. evaluated quantitative EZ metrics and observed that baseline VA was inversely associated with EZ loss [30].In long-term visits, Chatziralli et al. also reported the association between poor nal visual acuity and photoreceptor disruption [20].Kanakis et al. removed shadowing from the OCT scans and demonstrated the relationship between disruption of EZ and the areas of capillary nonperfusion [31].Therefore, the disruption of EZ is not only caused by photoreceptor loss, but also other interconnected factors, including optical and histologic effects of both edema and ischemia.The association of EZ integrity with baseline vision which is similar to the results of previous studies paves the way for studies using OCT analysis of EZ integrity to evaluate the baseline VA in RVO patients.
Previous studies have shown that hypertension, diabetes mellitus, hyperlipidemia, carotid plaque, and cigarette smoking are associated with an increased risk of RVO [7,8].However, our study has shown that they are not related to the extent of clinical manifestation.The reason for this marked variance from prior studies is uncertain.Evidence supporting the direct impact of systemic disease on visual function and retinal morphology is still not su cient.As it is a retrospective study, potential bias cannot always be eliminated, but it can nevertheless provide important results.
As with any retrospective analysis, the limitations of this study must be considered.This study is small and the results should be considered preliminary.Further assessment and research are needed to better understand the underlying pathophysiology of these retinal changes.Although other risk factors were not associated with VA and CRT at baseline, long-term follow-up and assessment of retinal architecture changes are needed.Finally, the cause of the association between endothelial dysfunction and baseline visual changes remains unknown, and further prospective human studies are required to determine the true nature of these changes.Despite these limitations, our results highlighted some baseline features in patients with RVO-ME, which should help inpatient counseling and planning preventive management.Measuring the FMD of the brachial artery and evaluating the integrity of EZ may help earlier recognition of eyes being prone to have poor vision.Patients with CRVO tend to have more signi cant ME than BRVO.Our study is limited by the small sample size, therefore larger studies are required to con rm these observations.

Declarations Ethics approval and consent to participate
All procedures performed in studies involving human participants were following the ethical standards of the Independent Ethics Committee for Clinical Research and Animal trials of the First A liated Hospital of Sun Yat-sen University.The number of ethical approval is 2020 [362].Informed consent was obtained from all individual participants included in the study.The Approval letter for Research Protocol is attached in supplemental les.Consent for publication

TABLE 1
Demographics and ocular characteristics in RVO eyes (n=46 patients, 46 eyes) **: P 0.01 TABLE 5 Logistic regression analysis for risk factors of visual acuity