Frequently asked questions about the results of the DIRECT Programme
Should clinicians consider treating diabetic patients with candesartan?
Were there any side effects of treatment with candesartan in DIRECT?
Why did candesartan have no effect in patients with type 1 diabetes who already had diabetic retinopathy?
Why did the DIRECT programme not run for longer?
Why did you not investigate the effect of candesartan on the incidence of retinopathy in type 2 diabetes?
Was there a difference in the occurrence of sight-threatening retinopathy between the treatment and placebo groups in the progression of retinopathy studies?
Could the results of the DIRECT programme be explained by the blood pressure controlling effect of candesartan?
Why are the sponsoring pharmaceutical companies not applying for approval of candesartan as an indication for regression of retinopathy in type 2 diabetes?
Why did candesartan have no effect on the incidence of microalbuminuria?
Should clinicians consider treating diabetic patients with candesartan?
Candesartan is an efficient treatment for hypertension with few side effects. For patients with type 2 diabetes who need treatment for hypertension and also have established mild to moderate retinopathy, clinicians could consider treating the hypertension with candesartan because it has an additional effect on improvement of retinopathy.
Patients with type 1 diabetes and no retinopathy who have high glucose levels that they have difficulty controlling are at a higher risk of developing retinopathy. Clinicians may wish to consider treating these patients with candesartan to reduce their risk of developing retinopathy.
Were there any side effects of treatment with candesartan in DIRECT?
Candesartan was well tolerated. The patients in DIRECT taking candesartan reported similar experiences of taking the treatment to those taking placebo.
Why did candesartan have no effect in patients with type 1 diabetes who already had diabetic retinopathy?
For patients with type 1 diabetes and retinopathy at baseline we did not see a statistically significant difference between the candesartan group and the placebo group for our pre-specified primary endpoint of a 3-step change in retinopathy severity measured using the Early Treatment of Diabetic Retinopathy Study (ETDRS) scale. One of the other outcomes that we looked at was the overall change in retinopathy severity from the baseline visit to the last visit, in other words, changes of 1 step or more on the ETDRS scale. These results showed that those treated with candesartan were less likely to progress and more likely to regress in terms of retinopathy severity than those on placebo. The trend of these results suggested that a treatment effect might have been seen for the primary endpoint if the trial had run for longer.
Why did the DIRECT programme not run for longer?
We calculated the duration of the programme and the number of participants required for each study to achieve at least 90% statistical power for detecting a treatment effect of candesartan on the incidence and progression of retinopathy. Our calculations were based on expected event rates for retinopathy incidence and progression, which we estimated from studies of glycaemic control that were carried out several years earlier. We also expected that any treatment effect would occur early in the programme, mirroring the early treatment effect seen in blood pressure lowering studies. We have evidence now that glycaemic control has improved in recent years and rates of incidence and progression of retinopathy have fallen. With this knowledge we would have planned a longer duration for the DIRECT programme. Although we realised during the study that the event rates were lower than expected, we had no way of knowing whether this reflected a treatment effect or a decrease in events across the treatment and placebo groups.
Why did you not investigate the effect of candesartan on the incidence of retinopathy in type 2 diabetes?
At the time of diagnosis, nearly 40% of people with type 2 diabetes will already have some degree of retinopathy, another 22% will develop it within 6 years; thus the rate of incidence of retinopathy in people with type 2 diabetes who have no retinopathy at diagnosis is low. To have sufficient statistical power to detect a treatment effect in an investigation of a medicinal product when event rates are low, a large number of participants and a long follow-up time are required. It was not feasible for us to conduct a study with the number of participants and follow-up time needed to investigate incidence of retinopathy in type 2 diabetes.
Was there a difference in the occurrence of sight-threatening retinopathy between the treatment and placebo groups in the progression of retinopathy studies?
Development of proliferative retinopathy or clinically significant macular oedema can cause a deterioration of vision. In type 1 and type 2 diabetes there were no differences in the incidence of proliferative retinopathy, or clinically significant macular oedema, or both, between the candesartan and placebo groups.
Could the results of the DIRECT programme be explained by the blood pressure controlling effect of candesartan?
The results of the three retinopathy studies (DIRECT-Prevent 1, DIRECT-Protect 1 and DIRECT-Protect 2) were adjusted statistically for differences in blood pressure and this did not change the main effects. Improved blood pressure control has an influence on diabetic retinopathy, but the results suggest that there is an additional benefit of candesartan on retinopathy over and above its blood pressure controlling effect.
Why are the sponsoring pharmaceutical companies not applying for approval of candesartan as an indication for regression of retinopathy in type 2 diabetes?
Regression of retinopathy was a pre-specified secondary endpoint in the DIRECT programme. It is only possible to apply for indications for pre-specified primary endpoints of clinical trials.
Why did candesartan have no effect on the incidence of microalbuminuria?
The DIRECT programme was not designed to be statistically powered for detecting a treatment effect on the incidence of microalbuminuria because this was a secondary endpoint. All the patients in DIRECT had little evidence of vasculopathy at baseline; they had very low urinary albumin excretion rates (UAERs) and little established cardiovascular disease. This meant that the patients were at low risk of developing microalbuminuria. The rate of incidence of microalbuminuria that we found in DIRECT was much lower than in previously published studies. These factors could explain why we found no effect of candesartan on the incidence of microalbuminuria.