Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

44th Annual Meeting of the European Association for the Study of Diabetes

Rome, Italy / September 6-11, 2008

The much anticipated results of the DIRECT (Diabetic Retinopathy Candesartan Trials) programme provide support for the ability of an angiotensin receptor blocker (ARB) to prevent retinopathy even though the primary end point of the programme was not met. According to several experts, the preponderance of data supported the basic underlying theories that led to the study. Even an expert who advocated caution because the predefined primary end points were not reached acknowledged that the overall direction of the data supported inhibition of the renin-angiotensin system (RAS) to prevent microvascular diabetic complications. Dr. Kristian F. Hansen, University of Oslo, Norway, indicated that DIRECT added to other data that RAS is a targetable factor in the pathophysiology of progressive diabetes. “One of the take-home messages is that an ARB and probably ACE inhibitors are indicated in diabetes patients with risk of progression to retinopathy,” he stated. However, this was one of the more cautious views. Others did not express the belief that the benefits observed in DIRECT could be extended to ACE inhibitors or even to ARBs other than candesartan. Differences in relative clinical effects have been observed between ARBs, and the DIRECT trials were conducted over a period of more than four years, during which time even relatively modest differences between agents may have a significant cumulative impact. The importance of indefinite and probably lifetime treatment was underscored by the fact that many of the advantages of candesartan did not accrue until after more than three years on treatment.

Objectives of the DIRECT Trials: DIRECT-Prevent 1

In the DIRECT trials programme, 5231 patients at 309 sites in 30 countries were randomized to candesartan or placebo. The primary prevention trial was called DIRECT-Prevent 1 and included 1421 patients with type 1 diabetes (DM1). The primary end point was new retinopathy, defined as 2-step increase in the Early Treatment Diabetic Retinopathy Study (ETDRS) scale. Of the secondary prevention trials, the one conducted in patients with DM1 was called DIRECT- Protect 1, while the other, conducted in patients with type 2 diabetes (DM2), was called DIRECT-Protect 2. Both of the secondary prevention trials randomized 1905 patients each. The primary end point for both was a reduction in the progression of retinopathy, defined as ³3-step increase in ETDRS. A variety of secondary end points were explored for all three studies.

The premise of the DIRECT programme is that upregulation of the RAS plays an important role in the pathophysiology of retinopathy independent of its role in raising blood pressure (BP). In addition to impairment of endothelial function, upregulation of RAS leads to hypertrophy and vascular remodelling. It is also implicated in oxidative stress and contributes to a pro-inflammatory environment that exacerbates vasculopathy. Along with neuronal damage, these processes are all implicated in the visual loss and blindness that characterize retinopathy. Several previous studies have suggested that RAS inhibition can protect against microvascular complications including retinopathy, but DIRECT is the largest and most rigorous study to test this hypothesis.

In the primary prevention trial, DIRECT-Prevent 1, there was an 18% reduction in the primary end point of new retinopathy, which fell just short of statistical significance (P=0.0508). However, the incidence of new retinopathy in the DM1 population over the course of four years of follow-up was lower than pre-study estimates. As the incidence of retinopathy increased toward the end of follow-up, the advantage of candesartan became more pronounced, suggesting a longer follow-up was needed for this end point. When the end point was changed in a post-hoc analysis to a 3-step change on the ETDRS scale, the advantage of active therapy over placebo jumped to 35%, which was highly statistically significant (P=0.003).

“Subgroup analyses were consistent across patient groups,” reported Dr. Nish Chaturvedi, National Heart and Lung Institute, Imperial College, London, UK, who acknowledged that study assumptions about the risk of new retinopathy were not consistent with observed rates in the study. “There is still compelling evidence from the study that candesartan is active.”

DIRECT-Protect 1 and 2 Findings

The results from the other two DIRECT studies reinforce this message, providing the basis for considering the data supportive of the premise that the ARB was active against retinopathy despite the lack of significance on the primary end point. In endorsing RAS inhibition as a standard approach to preventing complications from diabetes, Dr. Hansen noted, “The level of retinopathy was more favourably affected on candesartan at the end of all three studies compared to placebo.”

In the secondary prevention trials, DIRECT-Protect 1 and DIRECT-Protect 2, there was little or no evidence of an effect on progression, but there was evidence of an increased probability of regression in the DM2 study. Specifically, there was no trend towards a reduction in progression in DM1 on candesartan and the 13% reduction in progression in DM2 fell short of significance (P=0.2), but the 34% increase in probability of regression was highly statistically significant (P=0.009). In the DM1 secondary prevention study, the final ETDRS score was more likely to have improved on candesartan than on placebo (P=0.0264).

Retinopathy

The DIRECT trial is one of the most important studies to test protection against microvascular complications independent of glucose or BP control. Its goal of demonstrating protection against progressive retinopathy before it begins or at early stages was ambitious because of the challenge of documenting protection in a slowly progressive disease. Although the mean follow-up in the DIRECT studies was 4.7 years, the incidence of any retinopathy remains less than 20% over the first five years of DM1 before climbing steeply with persistent disease duration, reaching 80% after 30 years. Proliferative retinopathy, the most serious form and the type on which RAS inhibition may have the greatest impact, is even rarer within the first five years after diagnosis but reaches a 40% prevalence by 30 years. In DM2, the timeline of The DIRECT trial is one of the most important studies to test protection against microvascular complications independent of glucose or BP control. Its goal of demonstrating protection against progressive retinopathy before it begins or at early stages was ambitious because of the challenge of documenting protection in a slowly progressive disease. Although the mean follow-up in the DIRECT studies was 4.7 years, the incidence of any retinopathy remains less than 20% over the first five years of DM1 before climbing steeply with persistent disease duration, reaching 80% after 30 years. Proliferative retinopathy, the most serious form and the type on which RAS inhibition may have the greatest impact, is even rarer within the first five years after diagnosis but reaches a 40% prevalence by 30 years. In DM2, the timeline of retinopathy is more condensed. Within five years, the incidence of any retinopathy has reached 30% and sight-threatening proliferative retinopathy has already reached 5%. Both continue to climb in prevalence over the course of the disease, particularly when there is suboptimal glycemic control.

The need for additional therapeutic approaches even in patients who are optimally managed is suggested by the very studies that have demonstrated tight glycemic control reduces the risk of retinopathy. In the landmark UKPDS (United Kingdom Prospective Diabetes Study), tight glycemic control in DM2 reduced the risk of microvascular complications, such as retinopathy, by 37% (P=0.0092) but the absolute reduction was approximately 7%, leaving more than 10% of patients with microvascular events despite tight BP control (UKPDS Group. BMJ 1998;317:703-13). Similarly, in the DCCT (Diabetes Control and Complications Trial), intensive glycemic control was effective for both primary and secondary prevention of retinopathy, but retinopathy progressed in a substantial proportion of both groups receiving tight glycemic control (DCCT Research Group. N Engl J Med 1993;329:977-6).

Tight BP control, another risk factor for retinopathy, also reduces but does not eliminate the risk of this complication over time. This persistent risk of retinopathy despite optimal management has even greater implications in the context of data that less than half of diabetic patients are optimally controlled for either blood glucose or BP. In the US, only about one-half of patients are at or below the HbA1c target of 7.0%. An even smaller proportion are at the BP target of <130/80 mm Hg (Resnick et al. Diabetes Care 2006;29:531-7).

The Role of RAS Inhibition

The evidence that RAS inhibition offers BP-independent benefits was generated by a series of studies completed before the DIRECT trial was launched. One of the most significant of these was the EUCLID (Eurodiab Controlled Trial of Lisinopril in Insulin-Dependent Diabetes), a secondary prevention trial that associated an ACE inhibitor with a significant reduction in one level of progression of visual loss as well as significant protection against progression to proliferative retinopathy. Although three other studies produced only non-significant trends for protection against retinopathy, all of these studies were conducted with ACE inhibitors.

There are a number of theoretical reasons to predict that ARBs might do better. Beyond the practical advantage of greater tolerability with the implications for greater compliance over time, ARBs do not inhibit production of angiotensin but block their activity at the angiotensin II type 1 (AT1) receptor, increasing angiotensin stimulation of the AT2 receptor, which is associated with vasodilatory and antiproliferative effects. This is potentially important throughout the vasculature, but RAS has been documented as a mediator of pathophysiology of the eye, including in diabetic patients. Most notably, in a study of angiotensin II levels in human vitreous solution of non-diabetic patients, diabetic patients without retinopathy and in diabetic patients with retinopathy showed a dramatic increase in those with retinopathy.

There is also good evidence that RAS participates in neovascularization, which is the characteristic manifestation of proliferative retinopathy. Specifically, the dysregulation of vascular tone induced by upregulated RAS in the presence of diabetes leads to compromised blood flow, increasing hypoxia in peripheral tissues, including the eye. In a state of hypoxia, growth factors, including vascular endothelial growth factor (VEGF), stimulate formation of new blood vessels which disrupt the retina and lead to visual loss and blindness. While upregulated RAS plays a role in the impaired vascular function that initiates this process, it is also implicated in the molecular signalling that signals endothelial cell growth formation that is important to neovascularization. In animal models, candesartan has been associated with a reduction in VEGF expression and in neovascularization.

Figure 1. DIRECT-Protect 2 in DM2: Cumulative Proportion of Patients with Regression of Retinopathy*

The need to consider more aggressive approaches to prevent the complications of diabetes is acute. Diabetes is considered an epidemic in many western countries and the prevalence of DM2, which is age-related, will continue to climb steeply both from the rising prevalence of obesity and from the growing average age of the general population. Retinopathy is already the leading cause of blindness among working-age individuals. While glycemic and BP control are important, the investigators did not believe that the reduction in the progression of retinopathy could be solely attributed to the modest reductions in BP among those randomized to candesartan. For example, average BP in DIRECT-Prevent 1 were only 2.6/2.7 mm Hg lower on candesartan than on placebo. Differences were similarly modest in the other DIRECT studies. The largest difference was an average reduction of 4.3/2.5 mm Hg on candesartan relative to placebo in DIRECT-Protect 2 within the subgroup that were taking other antihypertensive therapies. In DIRECT-Protect 2 patients who did not receive other BP lowering drugs, the average reduction on candesartan relative to placebo was 2.9/1.3 mm Hg.

“Diabetic retinopathy is one of the most feared and common complications of diabetes, making this [results of DIRECT] an important clinical finding, as the reversal of this vision-threatening complication to diabetes has not been reported before in large-scale clinical trials,” declared one of the senior DIRECT investigators, Dr. Anne Katri Sjølie, Clinical Professor of Ophthalmology, Odense University, Denmark. Speaking of patients with DM2, she concluded from the DIRECT results that candesartan “could induce improvement in retinopathy.”

RAS inhibition and ARBs in particular have already demonstrated BP-independent benefits in DM2. In a series of studies conducted in patients with microalbuminuria, ARBs have demonstrated protection from end-stage renal disease. Although no effect on the incidence of microalbuminuria was observed among DM2 patients in DIRECT over the course of the study, there was a significant attenuation of the rate of increase in urinary albumin excretion rate, particularly early in the study (P=0.02 in the first year). RAS inhibitors, including ARBs, are also associated with BP-independent reductions in cardiovascular events in high-risk patients, which includes patients with diabetes.

The results of DIRECT are considered important. Although the primary end points were not met, the results overall are consistent with the underlying theory of the study and with the known pathophysiology of retinopathy. The modest early benefit of candesartan in this study was on a trajectory that would be expected to provide large and clinically meaningful protection against retinopathy over the longer periods of treatment that would be expected in routine clinical practice. The benefits would be expected to be additional to tight BP and glycemic control.

Summary

The DIRECT programme addressed important scientific and clinical questions by evaluating whether RAS is involved in the pathophysiology of retinopathy and is an appropriate clinical target. Although the study programme did not meet its primary end points, it did provide substantial evidence that the ARB candesartan has a favourable impact on this microvascular complication independent of its BP-lowering effects. The lack of significant protection against new diagnosis of retinopathy in DM1 may be the result of a lower-than-expected incidence of this complication. There was, however, highly significant protection against several secondary end points that provide support for the basic hypothesis of the study. As the primary end points were not met, the study does not provide an evidence-based endorsement of candesartan for preventing retinopathy in DM1, but it does contribute to other evidence that upregulation of RAS may be a targetable risk for microvascular progression in both DM1 and DM2.