Age-related macular degeneration (AMD) is a leading cause of severe visual loss especially in old age [1-3]. Based on the presence of choroidal neovascularization (CNV), AMD is typically categorized into two types: dry and wet (exudative) [1,2]. Vascular endothelial growth factor (VEGF) plays a fundamental role in CNV ; therefore, intravitreal injection of an anti-VEGF agent is globally accepted as a standard treatment for wet AMD [5-8]. The dosing schedules of anti-VEGF intravitreal injections are diverse. In terms of the increase in visual acuity, the best efficacy has been observed with a monthly injection of anti-VEGF . However, in actual clinical practice, monthly injections are difficult to administer because of the increased medical and economic burden borne by patients and physicians. Thus, alternative dosing regimens, such as pro re nata or treat-and-extend regimens, have been devised. Furthermore, the interval between injections varies depending on the symptoms of the patient or disease activity. For example, if the patient has more than a-five letter decrease in best-corrected visual acuity or if any evidence of disease activity is identified on optical coherence tomography such as a large increase or recurrence of subretinal or intraretinal fluid, the presence of a subfoveal hemorrhage, or large extrafoveal hemorrhage, then the patient requires additional anti-VEGF injections [9,10]. Since the worsening of the disease increases the frequency of intravitreal injections, thereby resulting in increased medical and financial burden, it is important to detect and block the factors that cause disease exacerbation.
In clinical situations, some patients require more injections than others because their response to anti-VEGF is less effective or the disease worsens more frequently. We believe that this could be related to alcohol consumption. Therefore, this study aimed to evaluate the effect of alcohol on VEGF expression in retinal pigment epithelium (RPE) cells and the angiogenesis of vascular endothelial cells. This in turn could influence disease activity of AMD or response of intravitreal anti-VEGF treatment in AMD patients.
Institutional Review Board/Ethics Committee ruled that approval was not required for this study. Two experiments were developed to investigate the possible effect of alcohol on AMD disease activity and the response of intravitreal anti-VEGF treatment. A VEGF expression experiment was first conducted to investigate the direct effect of alcohol on VEGF expression in RPE. Next, a human umbilical vein endothelial cell (HUVEC) capillary tube formation assay was used to determine whether alcohol induces neovascularization. Before the VEGF expression and HUVEC capillary tube formation assay, a cell viability assay determined the optimal concentrations required for bevacizumab and ethanol and their individual effects on VEGF expression in RPE. All experiments were performed independently three times.
Human RPE cells (ARPE-19; ATCC No.CRL-2302) were grown in a humidified incubator at 37°C in Dulbecco’s Modified Eagle Medium and Ham’s F12 media (DMEM/ F-12, Cat.11320; Gibco, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS; Cat.16000; Gibco) and 100 U/mL of penicillin-streptomycin (Cat.15140; Gibco); cultures were maintained in an atmosphere containing 95% air and 5% CO2. The cells between passages 3 and 5 were used for all experiments.
To investigate the optimal concentration of bevacizumab and ethanol, RPE cells were exposed to 0, 0.125, 0.25, and 0.5 mg/mL bevacizumab and 0, 80, 200, and 600 mMol ethanol. Treated cells were then washed with phosphate-buffered saline, and 10 μm of CCK-8 (Cat. CK04; Dojindo Molecular Technologies, Rockville, MD, USA), and 90 μL of FBS free media was added to each well, followed by incubation for 1 hour at 37°C. The optical density was measured by a microplate reader (EL × 8,000) at 450 nm wavelength.
ARPE-19 cells were seeded and cultured in complete media for 24 hours, following which the cells were grown in serum-free media for the next 24 hours. Cells were then exposed to bevacizumab (diluted to 0.5 mg/mL in serum-free media) and ethanol (600 mMol). To study the combined effects of bevacizumab and ethanol, cells were first exposed to 0.5 mg/mL bevacizumab for 6 hours, after which 0.06 mMol of ethanol was added to the same medium. After the appropriate treatment, the cells were incubated for 24 hours.
RPE cells were treated as four groups: control, bevacizumab 0.5 mg/mL, ethanol 600 mMol, and bevacizumab 0.5 mg/mL + ethanol 600 mMol. All treatments were carried out simultaneously, after which the amount of VEGF in cell culture supernates was measured by duplication using a Quantikine Human VEGF Immunoassay (Cat. DVE00; R&D Systems, Minneapolis, MN, USA).
Geltrex (Life Technologies Corporation, Cat. A1413202, Carlsbad, CA, USA) was added to each plate well at a concentration of 100 μL/cm2 in a 96-well plate. The plates were then incubated at 37°C for 30 minutes for solidification of the gel. ARPE-19 cells were seeded at a cell density of 5 × 105 cells per well in complete media in a 6-well plate for 24 hours, followed by starvation in medium 200-PRF for 24 hours. The cells were then exposed to diluted ethanol, bevacizumab, and recombinant human VEGF 165 protein (Cat. 293-VE; R&D Systems) for 24 hours, and the medium was collected from the culture wells. HUVEC (Cat. C-005-5C; Gibco) was seeded on the coated plates at a concentration of 4 × 104 cells per 1 cm2 and cultured using the above-sensitized media. After 3-, 6-, and 24-hour incubation, images were taken using an OLYMPUS CKX41 microscope.
For analysis of the tube formation assay, the number of extremities, branches, and total branch length per mm2 was quantified with the Angiogenesis Analyzer plugin ImageJ .
The mean ± standard deviation values are reported for the results of three independent experiments. SPSS Statistics (ver. 25.0, IBM Co., Armonk, NY, USA) was used to calculate the statistical significance. The Kruskal-Wallis test was applied, and a p-value < 0.05 was considered significant.
RPE cells were treated with 0, 0.125, 0.25, and 0.5 mg/mL bevacizumab and 0, 80, 200, and 600 mMol ethanol, and cell viability was assessed using the CCK-8 assay to determine the optimal concentration of bevacizumab and ethanol. No decrease was observed in the cell viability after exposure to bevacizumab and ethanol, with no statistical significance at all concentrations (Fig. 1). Results of the cell viability test determined the concentrations of bevacizumab (0.5 mg/mL) and ethanol (600 mMol) to be used for further experiments to assess the effect on VEGF expression at maximal concentrations without compromising cell viability.
The amount of VEGF expression was measured after exposure of RPE cells to bevacizumab 0.5 mg/mL, ethanol 600 mMol, and bevacizumab 0.5 mg/mL + ethanol 600 mMol. VEGF expression measured in the control and ethanol 600 mMol-treated groups were 368.14 ± 5.30 pg/mL and 382.70 ± 16.70 pg/mL, respectively, which were significantly not different. VEGF expression was not detected in the bevacizumab 0.5 mg/mL and bevacizumab + ethanol groups. Fig. 2 graphically represents the amount of VEGF expression observed in the four groups.
The results of HUVEC capillary tube formation are presented in Fig. 3. In the tube forming assay, we evaluated the number of extremities and branches, and total branch length. The following data was assessed at 3 hours. The number of extremities observed per mm2 was 1.56 in control, 2.26 in bevacizumab, 2.96 in ethanol, 2.80 in VEGF, 2.64 in bevacizumab + ethanol, and 2.57 in bevacizumab + VEGF groups. The number of branches per mm2 and total branch length per mm2, respectively, were 1.45 and 74.70 in control, 1.64 and 99.09 with bevacizumab, 2.14 and 135.67 with ethanol, 2.21 and 132.82 with VEGF, 1.70 and 104.98 in bevacizumab + ethanol, and 1.62 and 102.53 with bevacizumab + VEGF. A significant increase was observed in the number of extremities after exposure to ethanol (p= 0.007), VEGF (p= 0.007), bevacizumab + ethanol (p= 0.039), and bevacizumab + VEGF (p= 0.047). There were significant increases in the number of branches per mm2 and total branch length per mm2 in the ethanol (p= 0.022, 0.001, respectively) and VEGF (p= 0.009, 0.022, respectively) groups (Fig. 4).
The relationship between alcohol consumption and AMD remains unclear. Although cigarette smoking, and not drinking, is generally considered a risk factor of AMD [12-14], some animal experiments have shown that alcohol can affect AMD, specifically increasing CNV size [15,16]. The current study was undertaken to examine the relationship between alcohol and AMD by considering two aspects: whether alcohol directly increases the VEGF expression in RPE cells, and whether alcohol increases new vessel formation. Our studies in RPE cells reveal that the VEGF expression remained unaffected after treatment with ethanol. However, a significant increase in the number of extremities and branches, as well as total branch length was noted in the tube forming assay. The results of this study indicate that alcohol does not directly induce the expression of VEGF in RPE but increases angiogenesis in vitro.
Previous studies have found that ethanol promotes VEGF expression in tumor angiogenesis [17-19]. Additionally, in several studies, there is a relation between alcohol and CNV in animal models [15,16]. Bora et al.  and Kaliappanet et al.  showed that the size of CNV increases in alcohol-fed rats. Bora et al.  used a laser-induced CNV rat model and fed the rats alcohol for 10 weeks. They showed that heavy and prolonged alcohol consumption was associated with increased ethyl ester accumulation in the choroid and increases in the size of CNV . Kaliappanet et al.  showed that the complement system was related to increasing CNV size and alcohol consumption. The expression of CD59 was reduced in animals fed with alcohol as compared to the control, and this was associated with exacerbation of CNV . Numerous other reports have also demonstrated the relationship between the complement system and CNV development [20-23]. These studies have provided a link between alcohol and CNV or wet AMD, which is difficult to explain solely by increasing VEGF.
In addition to the association of the complement system and CNV, increased vascular permeability resulting from alcohol consumption might also result in increasing CNV. It is well established that ethanol breaks the vascular endothelial barrier, thereby increasing vascular permeability . In a study of vascular injury in the mucosa of the glandular stomach, the injury started within 1 minute after alcohol administration and appeared with enhanced vascular permeability . In particular, this shows that alcohol could have a greater impact on immature vasculature such as the CNV. Since they are immature blood vessels, the leakage of blood or plasma proteins occurs more in CNV than in normal vessels . In the case of immature vessels, the pericytes are less and loosely attached when compared to normal vessels, thereby enhancing the permeability of immature blood vessels [27-30]. From this perspective, the permeability of retinal and choroidal vessels could increase in patients who consume alcohol. The increased vessel permeability could affect the course of AMD, including the amount of subretinal fluid and hemorrhage, and the size of pigmented epithelial detachment or CNV, which are the main causes for increasing the number of anti-VEGF injections. In summary, the following inference can be made. Alcohol induces immature angiogenesis of the retina and choroid, thereby increasing vascular permeability. Increased vascular permeability increases the incidence of subretinal fluid requiring treatment in patients with AMD.
One of the limitations of this study is that it is an in vitro study and therefore difficult to infer directly to real-world conditions. The concentration of ethanol used in this study was 0.06 mMol since it showed a maximum effect in the cell experiment without affecting cell viability. However, this is equivalent to 3.65% of diluted ethanol, which is too high for practical application. The experiments were not carried out with lower concentrations of bevacizumab and ethanol. As this study was an in vitro study, it is difficult to conclude decisively before in vivo studies are performed. For an in vivo study, it should be necessary to adjust the appropriate concentrations of alcohol and bevacizumab. In addition, interpersonal differences, such as the frequency of drinking or the ability to degrade alcohol, should be considered in an in vivo study.
This study originated from clinical experience and has endeavored to show whether alcohol consumption affects CNV or AMD disease activity in vitro. Our results indicate that capillary tube formation is increased in the alcohol-treated group compared with the control group. We, therefore, surmise that alcohol consumption could influence the progression of AMD or anti-VEGF response in patients with AMD. As alcohol consumption could be a reversible risk factor, it is important for physicians to inform patients who receive anti-VEGF therapy for AMD that alcohol consumption could increase the disease activity or decrease the effect of anti-VEGF therapy.
In summary, this in vitro study indicates that ethanol has no effect on the expression of VEGF but increases tube formation. The result implies that alcohol could affect the disease activity of AMD and response to anti-VEGF treatment at the cellular level and is related to immature neovascularization. Subsequent studies that will identify possible causes of new capillary growth and confirm our results in vivo are warranted.
The authors declare no conflicts of interest relevant to this article.