REvisiting Lipids in REtinal Diseases: A Focused Review on Age-related Macular Degeneration and Diabetic Retinopathy

1. Cholesterol and lipoproteins

Cholesterol plays many diverse roles in the body, including serving as a contributor to the structure of cell membranes, a precursor for steroid hormones, a regulator of gene transcription, and a component involved in the formation of neuronal synapses.6, 7, 8, 9 For lipid transportation through the systemic circulation, lipoproteins are required. Lipoproteins are complex plasma particles that include a core composed of cholesterol esters and triglycerides (TG) and a surface composed of apolipoproteins, phospholipids, and unesterified cholesterol. According to their size, structure, and apolipoprotein content, lipoproteins are classified as chylomicrons, very-LDLs, intermediate-density lipoproteins, LDLs, HDLs, and lipoprotein (a).10

2. Distribution of cholesterol in the retina

The retina consists of 10 layers, from its internal limiting membrane to the retinal pigment epithelium (RPE). In addition to retinal vessels, the choroidal vasculature, which is connected to the systemic circulation, supplies the blood to the retina, especially outer the retina.11 To achieve physiological structure and function, the retina receives cholesterol from endogenous biosynthesis12, 13 and the systemic circulation across the RPE.14, 15

For lipid transportation, the RPE rapidly takes up lipoproteins from the systemic circulation, and it contains receptors for LDL (e.g., LDL-R) and HDL (e.g., SR-BI and SRBII) on the basolateral side, which contacts the choroid.16

1. AMD specific lesions: drusen and basal linear deposits (BLinDs)

AMD is the most common cause of blindness in developed countries and accounts for 8.7% of all cases of blindness worldwide.17, 18 In AMD, central vision is gradually reduced by changes in the macular region of the retina.19 According to the Beckman classification, AMD is classified into early, intermediate, and late AMD. Early AMD is diagnosed based on the presence of medium-sized drusen (>63 and ≤125 μm) and no retinal pigmentary abnormalities (hyperpigmentation or hypopigmentation). Intermediate AMD is defined as the presence of large drusen (>125 μm) and/or any pigmentary abnormalities. Late AMD is defined by geographic atrophy (GA) or neovascular AMD.20 Neovascular AMD is characterized by choroidal neovascularization (CNV), and GA is characterized by a sharply defined area of RPE degeneration in which the choroidal blood vessels are visible.21 Drusen are focal, dome-shaped lesions between the RPE basal lamina and the inner collagenous layer of Bruch's membrane (BrM) and are observed as yellow-white deposits on a fundus examination.22

BrM, which is the basement membrane of the RPE, has unique characteristics in terms of cholesterol content and is relevant to AMD. As aging progresses, BrM thickens and develops basal deposits. Depending on its location, a basal deposit is classified as a basal laminar deposit (BLamD) or a BLinD. BLamD is considered as a stereotypically thickened area of the RPE basal lamina. BLinD is located in the sub-RPE space, bounded internally by the basal lamina and externally by the inner collagenous layer of BrM. Drusen and BLinD are 2 forms of an AMD-specific lesion, in the shape of a lump and thin layer, respectively.23, 24, 25, 26, 27, 28, 29, 30 BLinD and soft drusen are considered to be alternative forms of the same entity because these lesions are located in the same plane and contains the same materials.31, 32 Curcio and Millican32 reported that eyes with AMD were 24 times more likely to have BLinD and large drusen (>125 μm) than age-matched controls.

These lesions contain lipoprotein-derived particles that have the physical forms of cholesterol seen in the core of atherosclerotic plaques,33, 34, 35 a mark of atheroma maturity. The formation of AMD lesions has thus been considered to share mechanisms with the formation of atherosclerotic plaques.36, 37 In atherosclerosis, trapped apolipoprotein B100 lipoproteins within the arterial wall initiate a cascade of pathological events, including innate immune system-mediated inflammation.38 This launches various downstream deleterious events, including macrophage recruitment, cytokine release, and neovascularization.

Similarly, in AMD, lipoprotein particles from multiple resources accumulate during aging. Those particles could fuse to form lipoprotein-derived debris such as soft drusen and BLinD. These processes may be accompanied by oxidative stress and inflammation, which contribute to the progression from non-neovascular AMD to neovascular AMD.39

2. Serum lipids, cholesterol-lowering medication, and AMD

Since the early 1960s, the correlations of AMD with concentrations of plasma cholesterol or apolipoproteins have been evaluated.40 Several clinical studies of the associations between lipids and AMD are summarized in Table 1. Klein et al.41 reported that carotid artery intima-media thickness and carotid plaques had a weak relationship with the incidence of late AMD. The Eye Disease Case Control Study reported a 4-fold increased risk of exudative AMD in patients with high total cholesterol levels (≥6.749 mmol/L).42 Reynolds et al.43 reported lower mean HDL cholesterol levels and higher LDL cholesterol levels in patients with advanced AMD than in controls. In addition, high total cholesterol and LDL cholesterol levels were found to be related to an increased risk of AMD after adjusting for environmental and genetic covariates.

Table 1
Clinical studies of associations between lipid profiles and age-related macular degeneration

• Click for larger image
• Click for full table
• Download as Excel file

Several studies have investigated the relationship between the risk of AMD and a specific lipid profile, such as HDL cholesterol. van Leeuwen et al.44 reported that elevated HDL cholesterol levels increased the risk of AMD. Klein et al.45 reported that high serum HDL cholesterol levels were associated with GA, and that a high total cholesterol/HDL ratio was associated with the incidence of RPE depigmentation and GA. The Pathologies Oculaires Liées à l'Age study reported that high HDL cholesterol and apolipoprotein A1 were associated with an increased risk of soft drusen.46 A large clinical study with 1,235 patients in North America showed an interesting finding that only neovascular AMD was associated with high HDL cholesterol levels, while non-neovascular AMD was unrelated to serum lipid levels.47

However, other epidemiological studies have not reported an association between serum lipid profile and AMD risk.48, 49 The discrepancies in results among previous studies could be due to the different lipid profiles that were analyzed or the diverse clinical characteristics of the enrolled patients, such as different stages of AMD.

Several studies have evaluated the effect of lipid metabolites in an in vivo model of CNV. The ω-3 and ω-6 long-chain polyunsaturated fatty acids (LCPUFAs) are 2 classes of essential fatty acids that have opposing effects. Metabolites generated by the cytochrome P450 (CYP)—epoxygenase pathway are potent modulators of inflammation and angiogenesis. Yanai et al.50 reported that dietary supplementation with ω-3 LCPUFAs decreased CNV lesions in a mouse model of neovascular AMD. Furthermore, mice fed ω-3 LCPUFAs showed suppressed leukocyte recruitment and adhesion molecule expression in CNV.50 Hasegawa et al.51 reported that the CYP-derived lipid metabolites 17,18-epoxyeicosatetraenoic acids and 19,20-epoxydocosapentaenoic acids play a vital role in alleviating CNV severity by regulating leukocyte recruitment and the inflammatory microenvironment in CNV lesions. Recently, the combined dietary intake of ω-3 LCPUFAs and lutein was found to attenuate CNV in an additive manner, which was related to suppression of inflammatory mediator production, reactive oxygen species generation, and NADPH oxidase 4 expression.52

Regarding lipid metabolites from humans, Lains et al.53 evaluated the plasma metabolomics profiles of AMD from 2 cohorts, including Boston in the United States and Coimbra in Portugal. Meta-analyses showed that 28 plasma metabolites differed between patients with AMD and controls and most of the significant metabolites were lipids. In particular, the metabolites mapping to glycerophospholipid pathways were altered in AMD subjects. Li et al.54 also stated that the glycerophospholipid pathway was one of the most significant pathways in their study group of polypoidal choroidal vasculopathy, a subtype of neovascular AMD.

Statins, which are 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors, have been used to control blood cholesterol levels.55 In AMD, statins have several potential effects such as preserving the blood supply to the outer retina, downregulation of LDL cholesterol and peroxidized lipids, and anti-inflammatory properties.56, 57, 58, 59 Previous studies evaluated whether cholesterol-lowering agents have a therapeutic effect on AMD, although the results are inconsistent. Vavvas et al.60 reported that high-dose statins (80 mg of atorvastatin daily) may result in the regression of drusenoid pigment epithelial detachments and a vision gain of 3.3 letters, without progression to advanced AMD such as GA or CNV. Barbosa et al.61 suggested a possible beneficial effect of statin intake for the prevention of AMD in individuals 68 years or older. Guymer et al.62 also reported a decrease in the risk of progression in the simvastatin group, and the most prominent effect was observed in individuals who were homozygous for the at risk C allele at Y402H of the complement factor H gene.

However, the effects of statins remain controversial. van Leeuwen et al.63 and Klein et al.64 did not find an association between statin use and the risk of AMD. In addition, a meta-analysis of 3 clinical trials showed no correlation between statin use and the incidence or progression of AMD.65

1. Serum lipids, cholesterol-lowering medication, and DR

DR, which is one of the most common microvascular complications of diabetes, is characterized by retinal hemorrhage and yellowish retinal exudates.66 DR is broadly classified into nonproliferative DR and proliferative DR (PDR) according to the presence of neovascularization of the disc or elsewhere, or vitreous hemorrhage.67 Diabetic macular edema is the most common cause of vision loss in patients with diabetes,68 and of particular note, the Early Treatment of Diabetic Retinopathy Study introduced the term “clinically significant macular edema” (CSME), which is defined upon slit lamp biomicroscopy as “1) thickening of the retina at or within 500 μm of the center of the macula; 2) hard exudate at or within 500 μm of the center of the macula associated with thickening of adjacent retina; or 3) a zone of retinal thickening 1 disc area or larger, any part of which is within 1 disc diameter of the center of the macula.69”

Table 2 summarizes several studies that have been conducted on the associations between lipids and DR. Many studies reported risk factors for the development of DR, including the duration of diabetes, glucose control, and the presence of hypertension.70, 71, 72, 73, 74 The effect of fat intake on exudative maculopathy has been studied. Ernest et al.75 reported that retinal exudates decreased in 8 diabetic patients after consumption of a low-fat diet for 2–3 years. Several studies have suggested that serum lipids may cause the development of retinal exudates. Patients with diabetes who had severe exudative maculopathy demonstrated higher levels of serum TG than those with nonexudative retinopathy, although serum cholesterol levels did not differ between the 2 groups.76 A case-control study reported that patients with maculopathy tended to have higher mean serum HDL cholesterol and total cholesterol levels over 7 years, but the difference was not significant.77

Table 2
Clinical studies of associations between lipid profiles and DR

• Click for larger image
• Click for full table
• Download as Excel file

Several studies have reported that serum lipid levels were not associated with DR severity or the presence of macular edema. Cetin et al.78 reported that serum lipid levels were not correlated with DR severity despite their correlation with mean blood glucose and hemoglobin A1c levels. Klein et al.79 reported that there was no association between serum total cholesterol or HDL cholesterol levels and the incidence of PDR or macular edema after adjusting for covariates.

Interestingly, however, some studies have reported that serum lipid levels were associated with CSME. Benarous et al.80 stated that although serum lipids were not associated with the presence and severity of DR, levels of serum lipids including total cholesterol, LDL cholesterol, non-HDL cholesterol, the LDL-to-HDL cholesterol ratio, and the total-to-HDL cholesterol ratio (cholesterol ratio) were independently associated with CSME after adjustment for traditional DR risk factors and lipid-lowering medications. Raman et al.81 also reported associations of serum lipids with CSME and non-CSME; specifically, high serum LDL cholesterol levels, non-HDL cholesterol levels, and a high cholesterol ratio were related to non-CSME, and high serum total cholesterol levels were associated with CSME.

In another small cross-sectional study, lipoprotein (a) was associated with the severity of DR in patients with type 1 diabetes, and in particular, lipoprotein (a) levels above 30 mg/dL were related to the risk of developing PDR.82 In addition, Li et al.83 reported that 8 metabolites were potential biomarkers for DR, including β-hydroxybutyric acid, trans-oleic acid, linoleic acid, and arachidonic acid.

Interestingly, before a role of lipids in the retina was established, several clinical trials of cholesterol-lowering agents in DR were performed. Atromid, a combination of ethyl-alpha-p-chlorophenoxyisobutyrate and androsterone, reduced serum cholesterol and TG levels in patients with ischemic heart disease.84 Duncan et al.85 reported that atromid only reduced the rate and extent of new exudate deposition in exudative DR, but did not improve visual acuity due to established exudates. Since then, statins and other lipid-lowering medications have been extensively studied as treatments for DR.

Statins reduced the risk of DR and reduced the number of patients who required laser photocoagulation treatment.86 Furthermore, statins decreased the cumulative incidence of DR.87 Denniston et al.88 reported that statins were associated with lower rate of diabetic vitreous hemorrhage at 1 year. However, Liinamaa and Savolainen89 reported that statins were associated with high levels of vascular endothelial growth factor in patients with PDR. In contrast, Zhang and McGwin90 reported no association between the use of statins and development of DR. Atorvastatin and simvastatin also showed different results depending on the study. Atorvastatin reduced hard exudates and fluorescein leakage in DR91 and simvastatin slowed the progression of DR.92

Fenofibrate, a peroxisome proliferator-activated receptor alpha agonist, lowers plasma TGs and increases HDL, and 2 randomized controlled clinical trials have analyzed its effects on DR. In 2007, the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study reported long-term effects of fenofibrate on DR in patients with type 2 diabetes mellitus.93 A multinational randomized trial of 9,795 patients were randomly assigned to receive fenofibrate (200 mg/day) or placebo. The cumulative percentage of patients who required a first round of laser treatment was lower in the fenofibrate group compared to the placebo group. Moreover, if DR was already present, the progression of retinopathy was higher in the placebo group than in the fenofibrate group. However, there was no difference in plasma concentrations of lipids between groups, so the mechanism of this effect does not seem to be related to plasma concentrations of lipids. The Action to Control Cardiovascular Risk in Diabetes (ACCORD) eye study, published in 2010, compared the progression of DR in a simvastatin plus placebo group and a simvastatin plus fenofibrate (160 mg/day) group. At 4 years, the rate of progression of DR was lower in the fenofibrate group than in the placebo group.94

However, Narang et al.95 reported that subjects with normal serum lipid levels did not show any relationship between atorvastatin use and reduction in hard exudates, macular edema, or visual acuity. The Collaborative Atorvastatin Diabetes Study trial also reported no added benefits of atorvastatin in DR; however, the treatment group needed less laser treatment, although the difference was not significant.96 Fried et al.97 reported that simvastatin medication did not affect the progression of DR.