FormalPara Key Summary Points

Why carry out this study?

The potential correlation between peripheral blood biomarkers and morphological characteristics of retinal imaging has been hypothesized to guide individualized treatment in patients with retinal vein occlusion (RVO).

What was learned from the study?

Specific imaging morphological characteristics were found to be associated with laboratory parameters in patients with RVO.

Eyes with subretinal fluid on optical coherence tomography presented significantly higher neutrophil-to-lymphocyte ratios.

Hyperreflective foci and the presence of cysts on optical coherence tomography were found to be significantly associated with higher triglyceride levels.

Central subfield thickness higher than 464 μm was associated with higher lymphocyte count and higher urea.

Introduction

Retinal vein occlusion (RVO) is the second most common retinal vascular disease after diabetic retinopathy and a significant cause of visual loss if left untreated, especially due to macular edema and retinal ischemia [1,2,3]. In 2015, the global prevalence of RVO was found to be about 0.77% in people aged 30–89 [4]. Today it is estimated that about 16 million people worldwide suffer from RVO [5].

The exact pathogenesis of RVO is not fully understood, but various cardiovascular, thrombophilic, systemic and ocular risk factors predispose to RVO, with the most common being hypertension, atherosclerosis, dyslipidemia, diabetes mellitus, smoking, hyperhomocysteinemia, hypercoagulability, ocular inflammatory diseases and open-angle glaucoma [6, 7]. Inflammation has been also implicated in the pathogenesis of RVO, since concentrations of several inflammatory mediators have been found to be elevated in patients with RVO, showing a correlation with the severity of the disease, as characterized by the presence of macular edema or ischemia [8,9,10,11].

Despite the current understanding of RVO pathogenesis and the data regarding intraocular biomarkers in RVO, literature is scarce about systemic biomarkers and their association with morphological characteristics in retinal imaging. Of note, great developments in retinal imaging are occurring nowadays, especially with the advent of spectral domain-optical coherence tomography (SD-OCT) and swept-source OCT (SS-OCT) [12, 13], as well as OCT angiography [14].

Based on the above, the purpose of the present study was to investigate the potential association between peripheral blood biomarkers and morphological characteristics of retinal imaging in patients with RVO.

Methods

Participants in this observational, cross-sectional study were 65 consecutive patients with treatment-naïve RVO, who were diagnosed and treated at 2nd Department of Ophthalmology, National and Kapodistrian University of Athens, Athens, Greece between 1 September 2019 and 31 March 2021. Diagnosis of RVO was made clinically on the basis of the presence of retinal hemorrhages, retinal vein dilatation, tortuosity, and flame-shaped and dot-blot hemorrhages, and confirmed by retinal imaging, including fundus fluorescein angiography (FFA) and SD-OCT. Patients with retinal diseases other than RVO, corneal disease, uveitis, glaucoma, dry eye disease, trauma or any previous intraocular surgery during the last 6 months were excluded from the study.

All participants underwent a complete ophthalmologic examination, including best-corrected visual acuity (BCVA) measurement by means of Snellen charts (converted to logMAR scale), slit-lamp examination, dilated fundoscopy, SD-OCT and FFA using the Spectralis imaging platform (Spectralis HRA+OCT, Heidelberg Engineering, Heidelberg, Germany), as described previously [15]. SD-OCT was obtained using a standard acquisition protocol; six radial scans 3 mm long were performed at equally spaced angular orientations centered on the foveola. The OCT volume scan was performed on a 20° × 20° cube, and consisted of 49 horizontal B-scans with 20 averaged frames per B-scan centered over the fovea [15]. The following SD-OCT variables were recorded at baseline: central subfield thickness (CST, μm) and presence of intraretinal fluid (IRF), subretinal fluid (SRF), cysts, hyperreflective foci (HF) and epiretinal membrane (ERM), while the ellipsoid zone (EZ) and external limiting membrane (ELM) condition was also assessed. Moreover, macular ischemia was evaluated on FFA and defined as disruption and enlargement of the foveal avascular zone (FAZ).

On the same day and following an 8-hour overnight fast, all patients underwent a forearm venous puncture for peripheral blood extraction and serum was separated. Full blood count was measured on a Sysmex XE-2100 analyzer (Sysmex Corp. Kobe, Japan), while all biochemical analyses were performed on a Roche Cobas 8000 (Roche, Chicago, IL, USA) in the laboratory of Attikon University Hospital [15]. Specifically, we analyzed the following parameters: glucose, glycated Hb (HbA1c), urea, creatinine, uric acid, triglycerides, lipoprotein(a), homocysteine, vitamin D, vitamin B12, folate, ferritin and ferrum. Cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL), apolipoprotein A and apolipoprotein B values were not included in the analysis, since some patients received statins, which could affect the results of the study.

Statistical Analysis

For the description of patients’ characteristics, descriptive statistics were calculated; mean ± standard deviation (SD) was used for continuous variables, while relative frequencies and percentages for categorical variables were reported. All variables were tested for normal distribution with the Shapiro–Wilk test. The associations between laboratory and imaging variables were evaluated with the Mann–Whitney-Wilcoxon test (MWW), as appropriate. Statistical analysis was performed using STATA/SE 13 statistical software (Stata Corporation, College Station, TX, USA). A p value < 0.05 was considered statistically significant [15].

Compliance with Ethics Guidelines

The study protocol adhered to the tenets of the Declaration of Helsinki and was approved by the Institutional Review Board of Attikon University Hospital (Reference number: 699/2019). Written informed consent to participate to the study and for publication of data was obtained by all participants before enrollment in the study.

Results

The demographic and clinical characteristics of the study sample are shown in Table 1. The mean age of patients was 67.5 ± 9.8 years; 53.8% of patients were male and 46.2% female. Forty-seven out of 65 patients (72.3%) had central RVO (CRVO) and 18 patients (27.7%) had branch RVO (BRVO). The mean BCVA was 0.78 ± 0.46 logMAR, while the mean CST was 474.2 ± 76.1 μm.

Table 1 Demographic and clinical characteristics of the study sample (n = 65 patients with retinal vein occlusion)
Full size table

Regarding the potential association between imaging characteristics and laboratory variables, no significant correlations were found for IRF, EZ and ELM condition, or ERM and macular ischemia.

Table 2 shows the comparison of laboratory parameters between eyes with SRF (n = 30) and without SRF (n = 35). Eyes with SRF presented significantly higher neutrophil-to-lymphocyte ratios (NLR) compared to those without SRF (p = 0.028, MWW). There was no statistically significant difference for the other laboratory parameters between the two groups.

Table 2 Association between laboratory variables and subretinal fluid on optical coherence tomography
Full size table

Table 3 shows the comparison of laboratory parameters between eyes with HF (n = 26) and those without HF (n = 39). Eyes with HF presented significantly higher triglyceride levels compared to eyes without HF (p = 0.024, MWW). No statistically significant differences were found for any other laboratory values between the two groups.

Table 3 Association between laboratory variables and hyperreflective foci on optical coherence tomography
Full size table

Table 4 shows the comparison of laboratory parameters between eyes with cysts (n = 57) and those without cysts (n = 8). Eyes with cysts had significantly higher triglyceride levels compared to eyes without cysts (p = 0.010, MWW), while all other laboratory values showed no statistically significant difference between the two groups.

Table 4 Association between laboratory variables and cysts on optical coherence tomography
Full size table

Table 5 shows the comparison of laboratory parameters between eyes with CST above or equal to median (464 μm) and those with CST below median. Eyes with CST ≥ 464 μm presented higher lymphocytes (p = 0.016, MWW) and higher urea (p = 0.015, MWW) compared to eyes with CST < 464 μm. There were no statistically significant differences regarding the other laboratory parameters between the two groups.

Table 5 Association between laboratory variables and central subfield thickness on optical coherence tomography
Full size table

Discussion

The principal message of the study is that most of the imaging morphological characteristics studied herein were not correlated with laboratory parameters in patients with RVO, suggesting that macular edema secondary to RVO may be mainly attributed to a local response more than a systemic effect. However, CST and specific OCT biomarkers, i.e., SRF, HF and cysts, were found to be associated with laboratory findings in patients with RVO, shedding light on the pathophysiology of RVO and its correlation with the development of these specific clinical characteristics.

Firstly, higher serum triglyceride levels were associated with the presence of HF and cysts. Many theories have attempted to explain the pathophysiology of HF, but their precise nature remains elusive. Framme et al. suggested that HF can be leucocytes or retinal pigment epithelium cells, indicating retinal inflammation [16]. Coscas et al. supported this concept and postulated that HF are microglial cells activated by inflammation [17]. Previous studies have found an association between HF and dyslipidemia, especially high levels of triglycerides in patients with diabetic macular edema, supporting the view that inflammatory reaction may be enhanced by dyslipidemia [18]. Moreover, it has been reported that metabolic syndrome and the related dyslipidemia has been involved in the development of microvascular changes in the retina, even in the absence of diabetes [19]. A similar pathogenetic mechanism could be hypothesized in patients with RVO.

Another finding of our study was that patients with SRF on SD-OCT presented with higher NLR. It is worth noting that NLR has been proposed as an indicator of inflammation in several diseases, such as coronary artery disease, non-cardiac diseases, age-related macular degeneration and diabetic retinopathy [20,21,22,23]. Furthermore, NLR has been previously evaluated in patients with RVO and found to be higher compared to controls [24,25,26]. Noma et al. demonstrated that high levels of inflammatory factors (i.e., soluble vascular endothelial growth factor receptor-2 and soluble intercellular adhesion molecule 1) were observed in patients with macular edema and serous retinal detachment secondary to CRVO, suggesting that an inflammatory component may prevail in such cases [27]. The latter can explain the significantly higher NLR in patients with SRF in our study.

In our study, we also found that higher urea was associated with more severe macular edema, with higher CST in patients with RVO. Hwang et al. found that macular edema and CST decreased significantly after initiation of dialysis in patients with chronic renal failure, since improvement in uremia occurs after dialysis [28]. Moreover, renal function has been indicated as a possible predictor for treatment response in certain patients with diabetic macular edema [29]. Finally, in patients with RVO, normal baseline renal parameters (urea and creatinine) predict better visual outcome after treatment and offer additional benefit over and above that obtained with macular edema reduction [30]. Therefore, it could be postulated that high urea is a negative predictor for macular edema in patients with RVO, as was found in our study.

A limitation of the study is that inflammation-related factors, such as C-reactive protein, interleukin and tumor necrosis factor-alpha (TNF-α), were not assessed. Further studies with a larger study sample size are needed, including analysis of these factors.

Conclusion

In conclusion, this study investigated the potential correlation between imaging morphological findings and laboratory biomarkers in patients with RVO. Our results showed that inflammatory biomarkers such as NLR, lymphocytes and urea were associated with more severe disease activity with higher CST and the presence of HF and SRF, while higher triglyceride levels were also found to be associated with the presence of HF and cysts. These findings may shed light on the pathophysiology of RVO, enhancing the view that there is an inflammatory component in RVO patients, especially in cases with specific characteristics such as HF and SRF, allowing an individualized approach in treatment decisions. However, it should be noted that most of the imaging biomarkers studied herein were not correlated with laboratory parameters in patients with RVO, suggesting that RVO may be mainly attributed to a local response more than a systemic effect. Further studies with large sample size are needed to justify our results.