Retinal microaneurysm count predicts ... - Wiley Online Library

DIABETICMedicine DOI: 10.1111/j.1464-5491.2010.03210.x

Article: Complications Retinal microaneurysm count predicts progression and regression of diabetic retinopathy. Post-hoc results from the DIRECT Programme A. K. Sjølie, R. Klein*, M. Porta†, T. Orchard‡, J. Fuller§, H. H. Parving–**, R. Bilous††, S. Aldington‡‡ and N. Chaturvedi§§ Department of Ophthalmology, Odense University Hospital, Odense, Denmark, *Department of Ophthalmology and Visual Science, University of Wisconsin, Madison, WI, USA, †Department of Internal Medicine, University of Turin, Turin, Italy, ‡Department of Epidemiology, University of Pittsburgh, Pittsburgh, PA, USA, §Department of Epidemiology and Public Health, University College London, London, UK, –Department of Medical Endocrinology, University Hospital of Copenhagen, Copenhagen, **Faculty of Health Science, Aarhus University, Aarhus, Denmark, ††Academic Centre, James Cook University Hospital, Middlesbrough, UK, ‡‡Department of Ophthalmology, Gloucestershire Hospitals, NHS Foundation Trust, Gloucestershire and §§International Centre for Circulatory Health and Lung Institute, Imperial College Healthcare NHS Trust, London, UK Accepted 22 November 2010

Abstract To study the association between baseline retinal microaneurysm score and progression and regression of diabetic retinopathy, and response to treatment with candesartan in people with diabetes.

Objective

Methods This was a multicenter randomized clinical trial. The progression analysis included 893 patients with Type 1 diabetes and 526 patients with Type 2 diabetes with retinal microaneurysms only at baseline. For regression, 438 with Type 1 and 216 with Type 2 diabetes qualified. Microaneurysms were scored from yearly retinal photographs according to the Early Treatment Diabetic Retinopathy Study (ETDRS) protocol. Retinopathy progression and regression was defined as two or more step change on the ETDRS scale from baseline. Patients were normoalbuminuric, and normotensive with Type 1 and Type 2 diabetes or treated hypertensive with Type 2 diabetes. They were randomized to treatment with candesartan 32 mg daily or placebo and followed for 4.6 years.

A higher microaneurysm score at baseline predicted an increased risk of retinopathy progression (HR per microaneurysm score 1.08, P < 0.0001 in Type 1 diabetes; HR 1.07, P = 0.0174 in Type 2 diabetes) and reduced the likelihood of regression (HR 0.79, P < 0.0001 in Type 1 diabetes; HR 0.85, P = 0.0009 in Type 2 diabetes), all adjusted for baseline variables and treatment. Candesartan reduced the risk of microaneurysm score progression. Results

Conclusions Microaneurysm counts are important prognostic indicators for worsening of retinopathy, thus microaneurysms are not benign. Treatment with renin-angiotensin system inhibitors is effective in the early stages and may improve mild diabetic retinopathy. Microaneurysm scores may be useful surrogate endpoints in clinical trials.

Diabet. Med. 28, 345–351 (2011) Keywords

angiotensin receptor blocker, candesartan, diabetic retinopathy, microaneurysms

Abbreviations

DIRECT, DIabetic REtinopathy Candesartan Trials; ETDRS, Early Treatment Diabetic Retinopathy

Study

Introduction Diabetic retinopathy is still an important cause of visual loss, although improvement in the management of diabetes has Correspondence to: Professor A. K. Sjølie, Department of Ophthalmology, Odense University Hospital, DK-5000 Odense, Denmark. E-mail: [email protected] (Clinical Trials Registry Nos; NCT00252733 for DIRECT-Prevent 1, NCT00252720 for DIRECT-Protect 1, NCT00252694 for DIRECT-Protect 2)

ª 2011 The Authors. Diabetic Medicine ª 2011 Diabetes UK

resulted in fewer individuals progressing to advanced stages of eye disease, such as macular oedema and proliferative retinopathy [1–4]. However, it is estimated that the number of people with diabetes will increase from 240 million in 2010 to 340 million in 2025. In addition, with the growing numbers of persons with Type 2 diabetes emerging in youth, the burden of sight-threatening retinopathy is predicted to increase [5]. Early stages of retinopathy predict progression to sight-threatening retinopathy [6] and are indicators of increased risk of both

345

DIABETICMedicine

Association between retinal microaneurysm count and diabetic retinopathy • A. K. Sjølie et al.

macro- and microvascular complications [7]. It is therefore important to reduce risks of complications at an early stage of diabetes. We have recently shown in the DIabetic REtinopathy Candesartan Trials (DIRECT) Programme [8,9] that the angiotensin receptor blocker candesartan may reduce the incidence and progression of diabetic retinopathy. The effect was strongest in those without or with early diabetic retinopathy. In the DIRECT Programme, retinopathy severity was assessed using the Early Treatment Diabetic Retinopathy Study (ETDRS) severity scale. However, the ETDRS scale was developed at a time when the effect of interventions for severe retinopathy was the targeted outcome and is less sensitive at the earliest stages of retinopathy. Others have highlighted the importance of an alternative more sensitive measure, the microaneurysm count, in predicting progression to more severe retinopathy, and have suggested it as an early measure of treatment effectiveness [10–13]. We therefore explored the association between baseline microaneurysm score and firstly, progression and regression of retinopathy, and, secondly, response to treatment in people with Type 1 and Type 2 diabetes in the DIRECT Programme as posthoc analyses.

Subjects and methods The DIRECT study design has been described in detail elsewhere [8,9]. In brief, 1421 patients with Type 1 diabetes without retinopathy at baseline (DIRECT-Prevent 1), 1905 patients with Type 1 diabetes with mild to moderately severe non-proliferative retinopathy (ETDRS levels 20–47) at baseline (DIRECTProtect 1), and 1905 patients with Type 2 diabetes with ETDRS levels 20–47 retinopathy (DIRECT-Protect 2) were randomized to either the angiotensin receptor blocker candesartan 16 mg daily or matching placebo at baseline. The dosage was doubled to 32 mg after 1 month. All patients were normoalbuminuric (albumin excretion rate < 20 lg ⁄ min) at baseline, patients with Type 1 diabetes were also normotensive untreated (blood pressure < 130 ⁄ 85 mmHg), patients with Type 2 diabetes were either normotensive or treated hypertensive (blood pressure < 160 ⁄ 90 mmHg) and treated with agents other than renin–angiotensin system inhibitors. All participants were followed up for at least 4 years. As the patients in DIRECT-Prevent 1 did not have microaneurysms at baseline, they were included only for analysis of microaneurysm score as an endpoint when examining the effect by treatment. Seven-field stereo photographs of both eyes were taken according to the ETDRS protocol. Photographs were evaluated for retinopathy in each eye in a masked fashion by two teams of independent observers, a primary and secondary grader, constituting a grading team, assigned to each patient for the duration of the study at the Retinopathy Grading Centre, Imperial College London. Regular quality assurance procedures to test reliability and repeatability were instituted at the Retinopathy Grading Centre.

346

The ETDRS scale has 11 assignable levels of increasing severity for retinopathy, from none to advanced proliferative diabetic retinopathy. In addition to grading the individual eyes to the ETDRS level of retinopathy, retinal microaneurysms were counted in patients who did not have other retinal lesions in addition to microaneurysms, using the same approach as adopted in the UK Prospective Diabetes Study 42. The microaneurysm grading for each eye of each patient was recorded on an ordinal scale of 0 to 7, with 0 as microaneurysm absent, a grading of 1 to 5 were actual microaneurysm counts, 6 signified 6–10 microaneurysms and 7 equalled 11 or more microaneurysms. To allow interpretation on the scale of actual microaneurysms, grades 6 and 7 were counted as 6 and 11 microaneurysms, respectively. A combined ordinal score over both eyes was then derived. Retinopathy progression was defined as a two-step or more increase on the ETDRS scale for the patient and regression as a two-step decrease.

Statistical methods Statistical analysis

The two main objectives of the post-hoc analyses were to assess whether the microaneurysm scores at baseline were predictive of subsequent progression (or regression) of retinopathy and whether treatment with an angiotensin receptor blocker influenced the change in microaneurysm scores during the study. DIRECT-Prevent 1 was included in this latter analysis, but excluded a priori from the analyses of progression and regression because this population did not have microaneurysms at baseline. Analyses of baseline microaneurysm as a predictor of progression ⁄ regression by end of study

A generalization of the Cox proportional hazards regression model was used to determine time to first occurrence of progression and regression as interval censored observations [14]. The model produced estimates of the HR and 95% CI of a one-point increase in baseline microaneurysm score upon time to progression and time to regression. The following prespecified prognostic factors were also added simultaneously to the models: For DIRECT-Protect 1, randomized study treatment, baseline duration of diabetes (years), HbA1c (%) and mean sitting systolic blood pressure at baseline (mmHg). DIRECT-Protect 2 included the same covariates as for DIRECT-Protect 1, with the addition of baseline urinary albumin excretion rate (lg ⁄ min) and presence of antihypertensive treatment. The effect of baseline microaneurysm score alone was also estimated to assess any confounding with the other potential prognostic factors. To quantify the relationship between microaneurysm score at baseline and the ETDRS category at end of study, a Spearman rank correlation coefficient with associated 95% CI was calculated.


Original article

DIABETICMedicine

sufficient potential to regress, i.e. ETDRS level 20 in both eyes. Baseline characteristics were not significantly different from those observed for the progression analysis. A higher microaneurysm score at baseline predicted an increased risk of retinopathy progression and reduced the likelihood of regression in both Type 1 and Type 2 diabetes (Table 2). These associations were largely unaltered on multivariate adjustment for baseline variables and further for treatment allocation. The relation between microaneurysm scores at baseline and ETDRS level of retinopathy at end of study at 4.6 years is shown in Fig. 1. There was a consistent increase in the ETDRS retinopathy level at the end of the study, with increasing microaneurysm baseline scores both in DIRECT-Protect 1 and DIRECT-Protect 2, and both in the placebo and candesartantreated patients (Spearman rank correlation 0.31, 95% CI 0.24–0.36, P < 0.0001, and 0.29, 95% CI 0.21–0.37, P < 0.0001, for DIRECT-Protect 1 and DIRECT-Protect 2, respectively). We also analysed microaneurysm score as a predictor of development of clinically significant macular oedema and ⁄ or proliferative retinopathy. The HR was 1.12 (P = 0.006) for Protect 1 and 1.21 (P = 0.003) for Protect 2 for each

Analyses of the influence of angiotensin receptor blocker on change in microaneurysm score by end of study

The Wilcoxon–Mann–Whitney hypothesis test was used to assess whether the distribution of the change in microaneurysm scores from baseline to end of study differed by randomized treatment. To quantify any effect, we also calculated the corresponding odds and associated bootstrap 95% CI [15]. Therefore, an odds of < 1 would suggest a benefit in favour of candesarten.

Results Eight hundred and ninety-three patients (454 placebo, 439 candesartan) of the 1905 patients in DIRECT-Protect 1 and 526 patients (264 placebo, 262 candesartan) of the 1905 patients in DIRECT-Protect 2 had microaneurysms only (ETDRS levels 20 ⁄ 20 or 20 ⁄ 10, with level 20 meaning microaneurysms only, and level 10 meaning retinopathy absent) and were included in the present analyses. Baseline characteristics for the groups regarding progression analyses of the three studies are shown in Table 1 according to treatment allocation. Regression analysis was restricted to those 438 patients in DIRECT-Protect 1 and 216 patients in DIRECT-Protect 2 with

Table 1 Baseline characteristics of patients for progression analysis, according to study and treatment Study DIRECT-Prevent 1 n = 1322

Microaneurysm score Men (%) Age (years) Duration of diabetes (years) HbA1c (%) (mmol ⁄ mol) BMI (kg ⁄ m2) Total cholesterol (mmol ⁄ l) Treated hypertension No SBP (mmHg) DBP (mmHg) Yes SBP (mmHg) DBP (mmHg) Urinary albumin excretion rate (lg ⁄ min) No anti-hypertensive treatment Smoking (ex-smoker or current)

DIRECT-Protect 1 n = 893

DIRECT-Protect 2 n = 526

Candesartan

Placebo

Candesartan

Placebo

Candesartan

Placebo

n = 656

n = 666

n = 439

n = 454

n = 262

n = 264

—

—

3.3 (3.1)

3.3 (3.0)

2.7 (2.4)

2.6 (2.3)

386 (59) 29.5 (8.1) 6.5 (3.9)

365 (55) 29.9 (8.1) 6.8 (3.9)

236 (54) 31.0 (8.4) 9.7 (4.3)

264 (58) 31.3 (8.4) 10.0 (4.3)

120 (46) 56.1 (7.8) 7.2 (4.6)

133 (50) 56.3 (8.1) 7.2 (4.5)

8.0 64 23.8 4.7

8.2 66 24.1 4.8

8.2 66 24.7 4.7

8.2 66 24.4 4.8

7.9 63 29.7 5.2

7.9 63 29.7 5.2

(1.7) (19) (3.1) (1.0)

116.2 (9.4) 72.3 (6.8) — — 4.5 (3.0, 6.5)

(1.7) (19) (3.5) (1.1)

116.4 (9.7) 72.2 (7.3) — — 4.5 (3.0, 6.5)

(1.4) (15) (3.5) (0.9)

116.6 (9.5) 73.4 (6.7) — — 4.5 (3.0, 6.5)

(1.6) (18) (3.4) (0.9)

116.7 (9.8) 72.8 (7.0) — — 4.5 (3.3, 7.0)

121.7 75.4 136.6 79.4 5.0

(1.6) (18) (4.8) (1.0)

(9.0) (7.0) (12.8) (6.8) (3.5, 8.0)

121.4 74.5 138.9 80.5 4.5

(1.5) (16) (5.4) (1.1)

(8.8) (6.8) (11.7) (6.2) (3.0, 7.0)

656 (100)

666 (100)

439 (100)

454 (100)

95 (36)

105 (40)

205 (31)

221 (33)

141 (32)

140 (31)

64 (24)

71 (26)

Data are mean (sd), number (%) or median (IQR). DBP, diastolic blood pressure; SBP, systolic blood pressure.


347


DIABETICMedicine

Table 2 Risk of two-step progression and regression per unit change in microaneurysm score Study DIRECT-Protect 1

Unadjusted two-step progression Adjusted* two-step progression Unadjusted two-step regression Adjusted* two-step regression

DIRECT-Protect 2

Hazard ratio (95% CI)

P-value

Hazard ratio (95% CI)

P-value

1.08 1.08 0.78 0.79

< < <
placebo Odds 0.89, 95% CI: 0.79–0.99 P = 0.031

50 40 30 20 10.5

10 0

(b)

4.5 4.1

–15 –12

–9

–6

2.6 0.9 1.4 0.5 0.6 0.8 0.8 1.9

12

15

18

21

DIRECT-Protect 1

43.2

Candesartan (n = 438) Placebo (n = 448)

36.8

35

Candesartan change > placebo Odds 0.79, 95% CI: 0.68–0.93 P = 0.003

30 25

21.5

20

17.4

15

16.7 12.6

10 5 0

7.8 0.2 0

0.9 0.5 0 1.3

–15 –12

8.3 5.1

3.7 3.6

–9

2.7

–6

2.3

–3 0 3 6 9 Change from baseline

2.5 2.5 2.9 2.2 2.7 1.8 0.9

12

15

18

21

50 45

42.8

45.1

DIRECT DIRECT-Protect 2

40 Per cent (%)

2.1

50 40

(c)

1.3


45

Per cent (%)

13.2

Candesartan (n = 262) Placebo (n = 264)

35 30

Candesartan change > placebo Odds 0.72, 95% CI: 0.59–0.88 P = 0.001

26.0

25 20

17.4

15

10.3

10 5 0

0 0.4

0.8

–15 –12

3.8 0.4

–9

9.9 9.2

8.7 6.1 2.7

2.7

–6

2.7 3.0


1.2

12

1.5

3.4 0.8

15

0

18

1.5

21

FIGURE 2 Distribution of change in microaneurysm score from baseline to end of study.

The disappearance rate for microaneurysms to no retinopathy in the 10-year follow-up study from the Wisconsin Epidemiologic Study of Diabetic Retinopathy [13] was 5.8% and was more common in eyes with fewer than three microaneurysms than in those with four or more. This is in agreement with our study, where fewer microaneurysms significantly predicted regression of retinopathy in patients with both Type 1 and Type 2 diabetes. The problem of a ‘point of no return’ from which diabetic retinopathy may no longer regress as a result of systemic treatment has been addressed in our primary analysis of DIRECT-Protect 2 [9], in which we found that regression was observed only in patients with very early retinopathy. The results of this present analysis


indicate that such a point may be even earlier. This is further supported by the findings from a small study with losartan in patients with Type 2 diabetes with more advanced maculopathy, where no treatment effect was found [24]. IntheUKProspectiveDiabetesStudy[10],27%ofpatientswith Type 2 diabetes with two to four microaneurysms at baseline developed more severe non-proliferative retinopathy at 6 years, and 4.6% developed vitreous haemorrhage or photocoagulation. We found with shorter observation time that 35.2% (314 ⁄ 893) and 37.8% (199 ⁄ 526) progressed two or more steps on the ETDRS scale, in patients with Type 1 and Type 2 diabetes, respectively, whereas few reachedproliferative disease or macular oedema. Further, the UK Prospective Diabetes Study found that,

349

DIABETICMedicine


in eyes with only one microaneurysm, approximately 50% of microaneurysms had disappeared at 3 years. The present study was not designed to determine the formation and disappearance rate of new microaneurysms, as only the total number of microaneurysms was counted. However, a recent report has demonstrated that high microaneurysm formation rate appears to be a good biomarker for progression to diabetic macular oedema [25], The same group has shown a computerassisted programme to be reliable in detecting microaneurysm turnover, in good agreement with graders [26]. The Wisconsin Epidemiologic Study of Diabetic Retinopathy used seven-field stereo photographs, as we did in DIRECT, whereas in the UK Prospective Diabetes Study only four fields were taken and, as the count was the total count per eye, this may be a concern in comparison between studies. Klein et al. found that only approximately 50% of the microaneurysms were found in the ETDRS fields 1 and 2 [12], which are the retinal fields centred on the optic disc and macula, respectively. The consistency of less progression of number of microaneurysms in patients randomized to candesartan than to placebo in all three studies in the DIRECT Programme is important and may have implications for therapeutic prevention and treatment of mild diabetic retinopathy. This is supported by the results of the primary analyses of the DIRECT Programme [8,9], where we found that, in all three studies, there was a statistically significant overall change towards a more favourable retinopathy outcome on treatment with candesartan than with placebo, when examining the distribution of all steps of the ETDRS levels of retinopathy. It is also in agreement with the recently published results of the Renin–Angiotensin System Study (RASS) study [27], which showed a beneficial effect of renin–angiotensin system blockade with enalapril or losartan on retinopathy progression in patients with Type 1 diabetes. We acknowledge that patients recruited for the DIRECT Programme were either normotensive (Type 1) and normotensive or treated hypertensive (Type 2) and thus comprise a selected group of patients with diabetes. The results may therefore not be applicable to all patients with diabetes.

Conclusions Microaneurysm count is important for prediction of retinopathy progression and regression, in patients with both Type 1 and Type 2 diabetes. The development of microaneurysms is thus not a benign event and represents significant structural damage. Our findings support the use of microaneurysm score as a useful surrogate clinical endpoint for progression and regression of retinopathy in clinical trials and may be more sensitive than the ETDRS in earlier stages. It should be emphasized that this is not applicable to screening for sight-threatening retinopathy in clinical practice. Importantly, we also found that treatment with the angiotensin receptor blocker candesartan had a beneficial effect on early stages of retinopathy, suggesting a potential benefit of renin–angiotensin system blockade in terms of retinopathy

350

progression in people with minimal diabetic retinopathy. Longer-term trials are required to explore the full clinical importance of this finding.

Competing interests All authors have received consultancy fees for attending DIRECT committee meetings. SA was Director of the Retinopathy Grading Centre. AKS, JF, HHP, RB and NC have received honoraria for scientific presentations from AstraZeneca and Takeda. The DIRECT Programme was funded by AstraZeneca and Takeda. Acknowledgements

Nigel Brayshaw, Takeda, London, UK has performed the statistical analyses, for which he is acknowledged.

References 1 Aiello LP, Gardner TW, King GL, Blankenship G, Cavallerano JD, Ferris FL et al. Diabetic retinopathy. Diabetes Care 1998; 21: 143– 156. 2 Saaddine JB, Honeycutt AA, Narayan KM, Zhang X, Klein R, Boyle JP. Projection of diabetic retinopathy and other major eye diseases among people with diabetes mellitus: United States, 2005– 2050. Arch Ophthalmol 2008; 126: 1740–1747. 3 Gæde P, Vedel P, Larsen N, Jensen GV, Parving HH, Pedersen O. Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med 2003; 348: 383–393. 4 Gæde P, Lund-Andersen H, Parving HH, Pedersen O. Effect of a multifactorial intervention on mortality in type 2 diabetes. N Engl J Med 2008; 358: 580–591. 5 World Health Organization. Diabetes fact sheet no. 312. 2008. Available at http://www.who.int/mediacentre/factsheets/fs312/e Last accessed 26 October 2009. 6 Klein R, Klein BE, Moss SE. How many steps of progression of diabetic retinopathy are meaningful? The Wisconsin Epidemiologic Study of Diabetic Retinopathy Arch Ophthalmol 2001; 119: 547– 553. 7 Wong TY, Rosamond W, Chang PP, Couper DJ, Sharrett AR, Hubbard LD et al. Retinopathy and risk of congestive heart failure. J Am Med Assoc 2005; 293: 63–69. 8 Chaturvedi N, Porta M, Klein R, Orchard T, Fuller J, Parving HH et al. Effect of candesartan on prevention (DIRECT-Prevent 1) and progression (DIRECT-Protect 1) of retinopathy in type 1 diabetes: randomised, placebo-controlled trials. Lancet 2008; 372: 1394– 1402. 9 Sjølie AK, Klein R, Porta M, Orchard T, Fuller J, Parving HH et al. Effect of candesartan on progression and regression of retinopathy in type 2 diabetes (DIRECT-Protect 2): a randomised placebocontrolled trial. Lancet 2008; 372: 1385–1393. 10 Kohner EM, Stratton IM, Aldington SJ, Turner RC, Matthews DR. Microaneurysms in the development of diabetic retinopathy (UKPDS 42). UK Prospective Diabetes Study Group. Diabetologia 1999; 42: 1107–1112. 11 Klein BE, Davis MD, Segal P, Long JA, Harris WA, Haug GA et al. Diabetic retinopathy. Assessment of severity and progression. Ophthalmology 1984; 91: 10–17. 12 Klein R, Meuer SM, Moss SE, Klein BE. The relationship of retinal microaneurysm counts to the 4-year progression of diabetic retinopathy. Arch Ophthalmol 1989; 107: 1780–1785.


Original article

13 Klein R, Meuer SM, Moss SE, Klein BE. Retinal microaneurysm counts and 10-year progression of diabetic retinopathy. Arch Ophthalmol 1995; 113: 1386–1391. 14 Alioum A, Commenges D. A proportional hazards model for arbitrarily censored and truncated data. Biometrics 1996; 52: 512– 524. 15 O’Brien RG, Castelloe J Exploiting the link between the Wilcoxon– Mann–Whitney test and a simple odds statistic. Proceedings of the Thirty-First Annual SAS Users Group International Conference, March 26-29, 2006, San Francisco, California March 2006. Available at http://www2.sas.com/proceedings/sugi31/209-31.pdf Last accessed 26 October 2009. 16 Luetscher JA, Kraemer FB, Wilson DM, Schwartz HC, Bryer-Ash M. Increased plasma inactive renin in diabetes mellitus. A marker of microvascular complications. N Engl J Med 1985; 312: 1412–1417. 17 Sarlos S, Wilkinson-Berka JL. The renin–angiotensin system and the developing retinal vasculature. Invest Ophthalmol Vis Sci 2005; 46: 1069–1077. 18 Funatsu H, Yamashita H, Nakanishi Y, Hori S. Angiotensin II and vascular endothelial growth factor in the vitreous fluid of patients with proliferative diabetic retinopathy. Br J Ophthalmol 2002; 86: 311–315. 19 Aiello LP, Avery RL, Arrigg PG, Keyt BA, Jampel HD, Shah ST et al. Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med 1994; 331: 1480–1487. 20 Wagner J, Jan Danser AH, Derkx FH, de Jong TV, Paul M, Mullins JJ et al. Demonstration of renin mRNA, angiotensinogen mRNA, and angiotensin converting enzyme mRNA expression in the human eye: evidence for an intraocular renin–angiotensin system. Br J Ophthalmol 1996; 80: 159–163.


DIABETICMedicine

21 Tanaka Y, Katoh S, Hori S, Miura M, Yamashita H. Vascular endothelial growth factor in diabetic retinopathy. Lancet 1997; 349: 1520. 22 Gardner TW, Antonetti DA, Barber AJ, LaNoue KF, Levison SW. Diabetic retinopathy: more than meets the eye. Surv Ophthalmol 2002; 47: S253–262. 23 Moravski CJ, Kelly DJ, Cooper ME, Gilbert RE, Bertram JF, Shahinfar S et al. Retinal neovascularization is prevented by blockade of the renin–angiotensin system. Hypertension 2000; 36: 1099–1104. 24 Knudsen ST, Bek T, Poulsen PL, Hove MN, Rehling M and Mogensen CE. Effects of losartan on diabetic maculopathy in type 2 diabetic patients: a randomised, double-masked study. J Intern Med 2003; 254: 147–158. 25 Nunes S, Pires I, Rosa A, Duarte L, Bernardes R, Cunha-Vaz J. Microaneurysm turnover is a biomarker for diabetic retinopathy progression to clinically significant macular edema: findings for type 2 diabetics with nonproliferative retinopathy. Ophthalmologica 2009; 223: 292–297. 26 Bernardes R, Nunes S, Pereira I, Torrent T, Rosa A, Coelho D et al. Computer-assisted microaneurysm turnover in the early stages of diabetic retinopathy. Ophthalmologica 2009; 223: 284–291. 27 Mauer M, Zinman B, Gardiner R, Suissa S, Sinaiko A, Strand T et al. Renal and retinal effects of enalapril and losartan in type 1 diabetes. N Engl J Med 2009; 361: 40–51.

351