Browse the corpus

Diabetic retinopathy (DR) is the most common microvascular complication of diabetes and a leading cause of blindness worldwide and in the US (-). The individual lifetime risk of DR is estimated to be 50–60% in patients with type 2 diabetes and over 90% in patients with type 1 diabetes (). It is one of the most frequent causes of blindness in adults between 20-74 years of age in developed countries (). The same pathologic mechanisms that damage the kidneys and other organs affect the microcirculation of the eye (). With the global epidemic of diabetes, one expects that diabetes will be the leading global cause of vision loss in many countries ( ,). While DR is specific for diabetes, other eye disorders, such as glaucoma and cataracts, occur earlier and more frequently in people with diabetes (). Often, by the time patients seek ophthalmologic examination and treatment, there are significant alterations of the retinal microvasculature. Therefore, it is important for non-ophthalmologists to recognize the importance of eye disease in patients with diabetes so that appropriate referral to eye-care specialists can be a part of their diabetes management program.

In the Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR), the prevalence of DR in patients with type 1 diabetes was 17% in those with less than 5 years of diabetes vs 98% in those with 15 or more years of diabetes (). Proliferative diabetic retinopathy (PDR) was absent in patients with type 1 diabetes of short duration but present in 48% of those with 15 or more years of diabetes. In patients with type 1 diabetes, the 25-year rate of progression of DR was 83%, with progression to PDR occurring in 42% of patients (). Improvement of DR was observed in 18% of patients with type 1 diabetes. In the WESDR, 3.6% of patients with type 1 diabetes were legally blind, and 86% of the blindness was attributable to DR (). The risk of blindness increases with the duration of diabetes.

In the WESDR, patients with type 2 diabetes of less than 5 years had a prevalence of DR of 28%, while in patients with greater than 15 years of diabetes, the prevalence was 78% (). A considerable number of patients with type 2 diabetes (12-19%) have DR at the time of the diagnosis of diabetes (). The prevalence of PDR was relatively low in patients with type 2 diabetes (2%) in patients with less than 5 years duration vs 16% in patients with greater than 15 years duration of diabetes (). The prevalence of DR and PDR was greater in the patients with type 2 diabetes using insulin. In the patients with type 2 diabetes, 1.6% were legally blind, and one-third of cases of legal blindness were due to DR (). Of note, the WESDR cohort is 99% white, and data suggest a higher prevalence of DR in Mexican-Americans and African-Americans with type 2 diabetes (, , ). Asians appear to have the same or lower prevalence of DR (, ). DR occurs in both males and females with diabetes, but males appear to be at a slightly higher risk (). Diabetic macular edema (DME) occurs more commonly in patients with type 2 diabetes, and with the marked increase in the prevalence of type 2 diabetes, DME is becoming more common (). DME is over two times more prevalent than PDR ().

In a pooled analysis of 35 studies between 1980 and 2008, among 22,896 individuals with diabetes, the overall prevalence of DR was 34.6%, PDR 6.96%, DME 6.81%, and vision-threatening DR 10.2% (). The longer the duration of diabetes, the greater the prevalence of all of these diabetic eye manifestations (). Moreover, the prevalence of DR, PDR, and DME was greater in patients with type 1 diabetes (77%, 32%, and 14%) compared to patients with type 2 diabetes (32%, 3, and 6%) (, ).

In developed countries the incidence and the risk of progression of DR have greatly declined in patients with type 1 and type 2 diabetes (, , ). The WESDR showed that from 1980 to 2007, the estimated annual incidence of PDR decreased by 77%, and vision impairment decreased by 57% in patients with type 1 diabetes (). In an analysis of 28 studies with 27,120 patients, the rates of DR and PDR were lower among participants in 1986-2008 than in 1975-1985 (). Thus, patients with recently diagnosed type 1 or type 2 diabetes in developed countries have a much lower risk of PDR, DME, and visual impairment as compared with patients who developed diabetes in the past (, ). This marked decrease in the prevalence and incidence of DR and vision impairment is likely due to improved glycemia control, early screening for eye disease, and the more aggressive treatment of blood pressure (). However, in countries with limited medical resources, this reduced risk of DR and vision impairment is not occurring (). In caring for patients with diabetes, health care providers must bear in mind the substantial risks of developing visual loss that these patients face and the treatments that can reduce this risk. For affected patients, diabetes-related visual loss decreases the quality of life and interferes with the performance of daily activities.

The most important treatable risk factor for the development of DR is hyperglycemia. In patients with both type 1 and type 2 diabetes, elevated HbA1c levels are associated with an increased risk and progression of DR (, , -). Most importantly, randomized controlled trials comparing intensive glycemic control vs. usual care demonstrated a decrease in DR. A meta-analysis of 6 relatively small randomized trials prior to the publication of the Diabetes Control and Complications Trial (DCCT) reported that after 2 to 5 years of intensive therapy the risk of retinopathy progression was significantly reduced (OR 0.49, P = 0.011) (). Intensive therapy significantly retarded retinopathy progression to more severe states such as PDR or changes requiring laser treatment (OR 0.44, P = 0.018) ().

The DCCT was a randomized, controlled study of intensive glycemic control (HbA1c approximately 7%) vs. usual care (HbA1c approximately 9%) in 1,441 patients with type 1 diabetes (). This study found that intensive glucose control reduced the risk of developing retinopathy by 76% compared to usual care. In patients with pre-existing retinopathy, intensive control slowed progression of the DR by 54% (). For every 10% reduction in HbA1c (e.g., 10% to 9% or 9% to 8.1%) the risk of retinopathy progression was reduced on average by 44% (). The DCCT participants were followed in an observational Epidemiology of Diabetes Interventions and Complications (EDIC) study. During the EDIC study, the mean HbA1c levels became very similar in the intensive and usual care group, with the HbA1c of the intensive treatment group increasing to approximately 8% and the usual care group HbA1c decreasing to approximately 8% (). Despite the similar A1c levels in the 2 groups over 30 years there continued to be an approximately 50% risk reduction of further DR progression and the development of PDR and DME in the original intensive control group, a phenomenon termed metabolic memory (). These results indicate the need for early intensive glucose control.

In the Kumamoto study, 110 patients with type 2 diabetes were randomly assigned to a multiple insulin injection treatment group (MIT group) or to a conventional insulin injection treatment group (CIT group) and followed for 6 years (, ). HbA1c levels were 7.1% in the MIT group and 9.4% in the CIT group. Moreover, the development of DR after 6 years was 7.7% for the MIT group and 32.0% for the CIT group in the primary-prevention cohort (no microvascular disease at baseline) (P = 0.039), and progression of DR occurred in 19.2% of the MIT group and 44.0% of the CIT group in the secondary-intervention cohort (microvascular disease at baseline) (P = 0.049). This study demonstrated that improved glycemic control reduced DR in patients with type 2 diabetes.

In the UK Prospective Diabetes Study (UKPDS), 3,867 newly diagnosed patients with type 2 diabetes were randomized to diet therapy alone or to sulfonylureas or insulin with the goal of achieving a fasting glucose of 108 mg/dL (6mMol/L) in those treated with sulfonylureas or insulin (intensive group). Over 10 years, HbA1c levels were approximately 7.0% in the patients treated with sulfonylureas/insulin therapy compared with 7.9% in the diet group. This study found a 25% reduction in the risk of microvascular endpoints, including the need for retinal laser treatment, with intensive glucose control (). A risk reduction of 21% per 1% decrease in HbA1c was observed in this trial. Patients were closely followed after the study ended, and HbA1c levels after one year became similar in the two groups. Similar to the results seen in the DCCT/EDIC study, the benefits on microvascular disease persisted in the intensive control group, confirming the concept of metabolic memory in patients with type 2 diabetes ().

The ACCORD study was a randomized trial that enrolled 10,251 individuals with type 2 diabetes of a mean duration of 10 years who were at high risk for cardiovascular disease to receive either intensive or standard treatment for glycemia (HbA1c 6.4% vs. 7.5%). A subgroup of 2,856 individuals were evaluated for the effects of intensive vs. standard care at 4 years on the progression of diabetic retinopathy by 3 or more steps on the Early Treatment Diabetic Retinopathy Study Severity Scale. After 4 years, the rates of progression of diabetic retinopathy were 7.3% in the intensive group vs.10.4% in the standard therapy group (odds ratio, 0.67; P=0.003) (). It should be noted that in an analysis of the entire ACCORD study cohort, three-line change in visual acuity was reduced in the intensive control group (HR 0.94, CI 0.89-1.00; p=0.05) but no differences in photocoagulation, vitrectomy, or severe visual loss were observed (). Four years after the ACCORD trial ended, DR progressed in 5.8% of the intensive treatment group vs.12.7% in the standard treatment group (odds ratio 0.42, P < 0.0001) (), once again confirming the concept of metabolic memory.

It should be noted that two large cardiovascular outcome trials, the ADVANCE trial and the VADT, failed to demonstrate a benefit of intensive glucose control on diabetic retinopathy (, ). However, these trials enrolled patients who already had diabetes for several years prior to enrollment, so they likely had not had the opportunity to develop the metabolic memory that could yield better retinopathy outcomes. Additionally, a meta-analysis of the four large cardiovascular outcome studies in patients with type 2 diabetes (UKPDS, ACCORD, ADVANCE, and VADT) found that more intensive glucose control resulted in a decrease in HbA1c of -0.90% and a 13% reduction in the need for retinal photocoagulation therapy or vitrectomy, development of PDR, or progression of DR (). Another meta-analysis of 7 trials with 10,793 participants reported a 20% decrease in DR with intensive glycemic control (0.80, 0.67 to 0.94; P=0.009) (). Taken together, these results clearly demonstrate that in patients with both type 1 and type 2 diabetes, improvements in glycemic control will reduce the risk of the development and progression of DR.

Deterioration of DR, upon initiation of intensive diabetes treatment, was described in the 1980s in patients with type 1 diabetes who were treated intensively with continuous subcutaneous insulin infusions (-). In patients with poor glycemic control and DR, rapidly improving glycemic control can worsen DR and, in some instances, result in PDR or DME. This worsening can occur as soon as 3 months after initiating intensive glycemic control. In the DCCT early worsening was observed at the 6- and/or 12-month visit in 13.1% of patients in the intensive treatment group and in 7.6% of patients assigned to conventional treatment (odds ratio, 2.06; P < .001) (). In the DCCT the most important risk factors for early worsening of DR were a higher HbA1c level and reduction of this level during the first 6 months of treatment (). It must be recognized that in the DCCT the long-term ophthalmic outcomes in intensively treated patients who had early worsening were similar to or more favorable than outcomes in conventionally treated patients. This early worsening of DR with improved glycemic control has also been described in patients with type 2 diabetes treated with insulin or GLP-1 agonists, following bariatric surgery, in pregnant women with diabetes, and following pancreatic transplants in patients with type 1 diabetes (). The mechanism(s) leading to early worsening of DR with improvements in glycemic control are unknown (). The FOCUS study (NCT03811561) is evaluating this further by comparing the rate of 3-step ETDRS severity level progression, and other DR-related outcomes, at 5 years in patients on semaglutide versus placebo.

While this worsening is distressing, it must be recognized that the long-term benefits of improving glycemic control on DR greatly outweigh the risks of early worsening.

In the WESDR, blood pressure (BP) was not related to incidence or progression of retinopathy in the patients with type 2 diabetes using insulin or the type 2 patients not using insulin, but in the patients with type 1 diabetes systolic BP was a significant predictor of the incidence of DR (). In contrast, in the UKPDS and other studies high BP in patients with type 2 was associated with the development of DR (, , ). In one prospective study the risk of DR increased by 30% for every 10 mm Hg increase in systolic BP at baseline (). While observational studies can show an association, randomized controlled trials are required to demonstrate causation and the benefits of treatment. A number of studies have examined the effect of lowering BP in patients with hypertension on the development and progression of DR.

The UKPDS examined the effect of tight vs. less tight BP control in 1,148 hypertensive patients with type 2 diabetes (). In the tight BP control group (captopril and atenolol), BP was significantly reduced compared to the less tight group (144/82 mm Hg vs.154/87 mm Hg; (P<0.0001). After nine years the tight BP control group had a 34% reduction in the deterioration of retinopathy (P=0.0004) and a 47% reduced risk (P=0.004) of deterioration in visual acuity. Additionally, patients in the tight BP group were less likely to undergo photocoagulation (RR, 0.65; P = 0.03), a difference primarily due to a decrease in photocoagulation due to maculopathy (RR, 0.58; P = 0.02) (). In contrast to glycemic control, the benefits of lowering BP were not sustained when therapy was discontinued and the differences in blood pressure were not maintained, indicating the absence of metabolic memory ().

The HOPE study was a randomized study that compared ramipril vs. placebo in 3,577 participants with diabetes who had a previous cardiovascular event or at least one other cardiovascular risk factor (). The baseline BP was approximately 142/80 mm Hg, and BP decreased by 1.92/3.3 mm Hg in the ramipril group vs a 0.55 mm Hg increase in systolic BP and 2.30 mm Hg decrease in diastolic BP in the placebo group. This study was not focused on DR but did report that the need for laser was 9.4% in the ramipril group vs. 10.5% in the placebo group (22% decrease; P=0.24). Another ACE inhibitor, lisinopril, was found in a two-year, placebo-controlled trial to be associated with lower rates of retinopathy progression in non-hypertensive patients with type 1 diabetes ().

The ADVANCE study examined the effect of BP control on DR in 1,241 patients with type 2 diabetes (). Patients were randomized to BP-lowering agents (perindopril and indapamide) or placebo and followed for approximately 4-5 years. Baseline BP was approximately 143/79 mmHg. In the group randomized to BP medications, a decrease in systolic BP of 6.1 ± 1.2 mmHg and diastolic BP of 2.3 ± 0.6 mmHg was observed (p < 0.001 for both). Fewer patients on BP lowering therapy experienced new or worsening DR compared with those on placebo (OR 0.78; 95% CI 0.57–1.06; p = 0.12), but the difference was not quite statistically significant; as mentioned above, this could be attributable to the fact that patients had diabetes for a long period prior to enrollment. Certain secondary outcomes were significantly reduced (for example DME) in the BP lowering group, but most other eye end points were not significantly decreased compared to the placebo group. The ACCORD eye study evaluated 2,856 patients with type 2 diabetes for the effect of intensive BP control (BP<120 mm Hg) vs standard BP control (BP<140 mm Hg) on the progression of DR after 4 years of treatment (). Systolic BP was 117 mm Hg in the intensive-therapy group and 133 mm Hg in the standard-therapy group. The progression of DR was 10.4% with intensive blood-pressure therapy vs. 8.8% with standard therapy (adjusted odds ratio, 1.23; P=0.29).

The Appropriate Blood Pressure Control in Diabetes (ABCD2) Trial was a randomized blinded trial that compared the effects of intensive versus moderate BP control in 470 patients with type 2 diabetes and hypertension (). The intensive group was treated with either nisoldipine or enalapril, while the usual care BP group received placebo. The mean blood pressure achieved was 132/78 mm Hg in the intensive group and 138/86 mm Hg in the moderate group. Over the 5-year follow-up period, there was no difference in the progression of DR between the intensive and moderate groups. Thus, in patients with hypertension, randomized trials of lowering BP have not consistently shown beneficial effects on DR.

There are numerous possible explanations for the differences in results between these studies. First, the duration of diabetes prior to enrollment in a diabetes-management trial will impact the subsequent outcomes and risk of developing complications. Second, the severity of the hypertension may be important, with greater responses in individuals with higher BP levels. Third, the magnitude of the reduction in BP may be important, with greater benefit with greater decreases in BP. Fourth, the duration of the study may be an important variable, with the longer the study the greater the chances of benefits. Fifth, the presence of DR at baseline and the severity of DR at baseline may influence the response to BP lowering. Sixth, patient variables such as glycemic control, age, diabetes type, duration of diabetes, etc., may influence results. Finally, the drugs used to lower BP may be a key variable as described below.

Because of the potential benefits of angiotensin converting enzyme inhibitors (ACE inhibitors) and angiotensin receptor inhibitors (ARBs) (Renin-Angiotensin System (RAS) inhibitors) on microvascular disease independent of BP effects, a number of studies have explored the effects of these drugs on DR in patients without elevated BP. Below we briefly describe the largest of these studies. The EUCLID trial was a randomized double-blind placebo-controlled trial in 354 patients with type 1 diabetes who were not hypertensive and were normoalbuminuric (85%) or microalbuminuric (). Study participants were randomized to lisinopril or placebo and followed for 2 years. Systolic BP was 3 mm Hg lower in the lisinopril group than in the placebo group. DR progressed in 23.4% of patients in the placebo group and 13.2% of patients in the lisinopril group (p=0.02). Notably progression to PDR was also reduced in the lisinopril treated group.

The Appropriate Blood Pressure Control in Diabetes (ABCD1) trial was a randomized trial in 480 normotensive type 2 diabetic subjects of more intensive vs. usual BP control (). The intensive group was treated with either nisoldipine or enalapril, while the usual care BP group received placebo. Mean BP in the intensive group was 128/75 mm Hg vs. 137/81 mm Hg in the placebo group (P < 0.0001). After a mean follow-up of 5.4 years, the intensive BP control group demonstrated less progression of diabetic retinopathy (34% vs. 46%, P = 0.019). PDR developed in 0% of patients in the intensive therapy group vs. 3.9% in the placebo group. However, in patients who at baseline did not have DR, the number of patients developing retinopathy was similar in the two groups (39% of patients in the intensive therapy group vs. 42% in the placebo group). The DIRECT- Prevent 1 trial was a randomized, double-blind, placebo-controlled trial in 1,421 normotensive, normoalbuminuric individuals with type 1 diabetes without retinopathy (). Patients were randomized to candesartan or placebo and followed for 4.7 years. Mean systolic and diastolic BP was reduced by 2.6 mm Hg and 2.7 mm Hg, respectively, in the candesartan group vs. the placebo group. DR developed in 25% of the participants in the candesartan group vs. 31% in the placebo group (18% decrease).

The Direct Protect 1 was a randomized, double-blind, placebo-controlled trial in 1,905 normotensive, normoalbuminuric patients with type 1 diabetes with existing retinopathy (). Patients were randomized to candesartan or placebo and followed for 4.7 years. Mean systolic and diastolic BP was reduced by 3.6 mm Hg and 2.5 mm Hg, respectively, in the candesartan group versus the placebo group. There was an identical 13% progression of DR in the placebo and candesartan groups, and progression to the combined secondary endpoint of PDR or clinically significant DME, or both, did not differ between the two groups.

The DIRECT-Protect 2 trial was a randomized, double-blind, placebo-controlled trial in 1,905 normoalbuminuric, normotensive, or treated hypertensive people with type 2 diabetes with mild to moderately severe retinopathy (). Patients were randomized to candesartan or placebo and followed for 4.7 years. The decrease in systolic/diastolic blood pressure was 4.3/2.5 mm Hg greater in the candesartan group than in the placebo group in individuals who were receiving antihypertensive treatment at baseline (p<0·0001 for both), and for those not on anti-hypertensive therapy at baseline the decrease was 2.9/1.3 mm Hg (p=0.0003/p=0.0045). The risk of progression of retinopathy was non-significantly reduced by 13% in patients on candesartan compared to the placebo group (HR 0.87; p=0.20). However, regression on active treatment was increased by 34% (HR 1.34; p=0.009), and overall change towards less severe retinopathy by the end of the trial was observed in the candesartan group (odds 1.17; p=0.003).

The RASS trial was a controlled trial involving 223 normotensive patients with type 1 diabetes and normoalbuminuria and who were randomly assigned to receive losartan, enalapril, or placebo (). The systolic and diastolic BP during the study were lower in the enalapril group (113/66 mm Hg) and the losartan group (115/66 mm Hg) than in the placebo group (117/68 mm Hg) (P<0.001 for the two systolic and P≤0.02 for the two diastolic comparisons, respectively). After 5 years progression in DR occurred in 38% of patients receiving placebo but only 25% of those receiving enalapril (P=0.02) and 21% of those receiving losartan (P=0.008).

Many of the studies described above used either an ACE inhibitor or an ARB with variable results on DR. To better understand the effect of RAS inhibitors on DR, a meta-analysis has extensively examined these studies and a number of other trials (). In 7 studies with 3,705 participants without DR, RAS inhibitors reduced the development of DR by 27% (p= 0.00006). This decrease in the development of DR was seen in patients with both type 1 and type 2 diabetes and patients who were hypertensive or normotensive. In 16 studies with 9,580 participants with pre-existing DR, RAS inhibitors decreased the progression of DR by 13% (p=0.00006). This decrease in progression of DR was seen in patients with both type 1 and type 2 diabetes and patients who were normotensive. In hypertensive patients there was a trend (7% decrease) that was not statistically significant. It should be noted that in the hypertensive patients RAS inhibitors were compared to other hypertensive drugs, and the number of hypertensive participants was relatively small (n=839). Therefore, the absence of a decrease in progression of DR in hypertensive patients is not definitive. Six studies with 2,624 participants examined the effect of RAS inhibitors on inducing regression of DR. RAS inhibitors increased the regression of DR by 39% (p=0.00002), and this beneficial effect was seen in patients with type 1 and type 2 diabetes. ACE inhibitors were more effective in reducing the development, progression, and regression of DR than ARBs. Thus, with the data available, RAS inhibitors appear to have benefits on DR above and beyond their effects on BP control.

Observational studies have shown an association of elevated BP with a higher risk of DR. As should be obvious from the above discussion, the beneficial effects of lowering BP in hypertensive patients on DR have not produced consistent results. Several large carefully carried out studies have failed to demonstrate a beneficial effect of lowering BP on DR (ACCORD, ADVANCE, ABCD2). Potential reasons for this inconsistency were discussed above. It is unlikely that future studies will provide definitive data on this issue, as lowering BP in hypertensive patients with diabetes to prevent cardiovascular disease is essential, and therefore designing clinical trials regarding DR will be very difficult. From the clinician’s viewpoint, treating hypertension in patients with diabetes to prevent cardiovascular disease is standard therapy and may also have beneficial effects DR. Similar to the beneficial effects on renal disease, RAS inhibitors appear to decrease the development and progression of DR, and therefore when treating patients with diabetes who are hypertensive, one should be preferentially consider RAS inhibitors to lower BP in patients with or at high risk of DR. In normotensive patients the available data suggests that RAS inhibition will have beneficial effects on DR, and further studies in this population are possible and would be informative.

Observational studies of the association of plasma lipids with DR have been inconsistent () with some studies reporting an increased risk of DR with elevated lipid levels (-), while other studies have not observed a relationship between lipid levels and DR (10, 38. 58-60). Of note a Mendelian randomization study did not demonstrate a causal role of total cholesterol, LDL cholesterol, HDL cholesterol, or triglycerides on DR (). From the clinician’s point of view the key question is whether lowering lipid levels will have a beneficial effect on DR.

Small studies in the 1960’s presented evidence that treatment with clofibrate improved diabetic retinopathy (, ). Larger randomized studies have confirmed these observations.

The Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study was a randomized trial in patients with Type 2 diabetes. Patients were randomly assigned to receive either fenofibrate 200 mg/day (n=4895) or placebo (n=4900). Laser treatment for retinopathy was significantly lower in the fenofibrate group than in the placebo group (3.4% patients on fenofibrate vs 4.9% on placebo; p=0.0002) (). Fenofibrate therapy reduced the need for laser therapy to a similar extent for maculopathy (31% decrease) and for proliferative retinopathy (30% decrease). In the ophthalmology sub-study (n=1012), the primary endpoint of 2-step progression of retinopathy grade did not differ significantly between the fenofibrate and control groups (9.6% patients on fenofibrate vs 12.3% on placebo; p=0.19). In patients without pre-existing retinopathy there was no difference in progression (11.4% vs 11.7%; p=0.87). However, in patients with pre-existing retinopathy, significantly fewer patients on fenofibrate had a 2-step progression than did those on placebo (3.1% patients vs 14.6%; p=0.004). A composite endpoint of 2-step progression of retinopathy grade, macular edema, or laser treatments was significantly reduced in the fenofibrate group (HR 0.66, 95% CI 0.47-0.94; p=0.022).

In the ACCORD Study a subgroup of participants was evaluated for the progression of diabetic retinopathy by 3 or more steps on the Early Treatment Diabetic Retinopathy Study Severity Scale or the development of diabetic retinopathy necessitating laser photocoagulation or vitrectomy over a four-year period (). At 4 years, the rates of progression of diabetic retinopathy were 6.5% with fenofibrate therapy (n=806) vs. 10.2% with placebo (n=787) (adjusted odds ratio, 0.60; 95% CI, 0.42 to 0.87; P = 0.006). Of note, this reduction in the progression of diabetic retinopathy was of a similar magnitude as intensive glycemic treatment vs. standard therapy. A double-blind, randomized, placebo-controlled study in 296 patients with type 2 diabetes mellitus and DR evaluated the effect of placebo or etofibrate on DR (). After 12 months an improvement in ocular pathology was more frequent in the etofibrate group vs the placebo group ((46% versus 32%; p< 0.001). The MacuFen study was a small double-blind, randomized, placebo-controlled study in 110 subjects with DME who did not require immediate photocoagulation or intraocular treatment. Patients were randomized to fenofibrate for placebo for 1 year. Patients treated with fenofibrate acid had a modest improvement in total macular volume that was not statistically significant compared to the placebo group.