Insight from DRCR.net protocol W and PANORAMA study for non-proliferative diabetic retinopathy
A panoramic picture of Hong Kong
Two recent high quality randomized controlled trials were published to look at the role of anti-VEGF therapy in treating non-proliferative diabetic retinopathy to prevent progression. Here I will summarise the results of these studies and discuss their clinical implications.
PANORAMA study
The PANORAMA study asks if treatment of patients with moderately severe to severe nonproliferative diabetic retinopathy (NPDR) with intravitreal aflibercept (eylea) injections would result in significant clinical improvement.
Patients are include if they are at Diabetic Retinopathy Severity Scale level 47 or 53 with no diabetic macula oedema with visual acuity of 6/12 or better. Overall 400 patients were recruited. They are then randomized into three groups:
Intravitreal injections of aflibercept, 2mg, every 16 weeks after 3 initial monthly doses and one 8-week interval (aflibercept 2q16 group)
Intravitreal injections of aflibercept, 2mg, every 8 weeks after 5 initial monthly doses, with pro re nata (PRN) dosing beginning at week 56 (aflibercept 2q8/PRN group). Retreatment is given in the PRN phase if the DRSS level is worse than 35.
Sham injections (control group)
The outcome measures are proportions of eyes with a 2-step or greater improvement in DRSS level, vision-threatening complications, and center-involved DME from baseline to weeks 24, 52, and 100.
Proportion of eyes with 2-step improvement in DRSS level. From the PANORAMA study.
The results are very promising. More than 50% of patients on the aflibercept groups have a 2-step improvement in DRSS level, and the effect is visible as early as 24 weeks.
Treated patients also have less vision threatening events (proliferative diabetic retinopathy and anterior segment neovascularisation) and clinically significant diabetic macula oedema. These are summarised in the Kaplan-Meyer graphs below.
Kaplan-Meyer graph for vision threatening events, from the PANORAMA study
Kaplan-Meyer graph for clinically significant diabetic macula oedema, from the PANORAMA study
The aflibercept 2q8/PRN group demonstrated that by the second year, the requirement for injection is low for most eyes. 33.6% received 0 injections (though 10% of this group developed vision threatening events), 42 eyes (34.4%) received 1 to 2 injections, 25 eyes (20.5%) received 3 to 4 injections, and 14 eyes (11.5%) received 5 to 6 injections. There were no vision threatening events in the eyes that have received one or more injections after week 52.
There is no statistically significance in the visual acuity outcome between all three groups in the PANORAMA study.
DRCR.net protocol W
The DRCR.net protocol W study asks a similar question as the PANORAMA study - if treatment of patients with moderately severe to severe nonproliferative diabetic retinopathy (NPDR) with intravitreal aflibercept (eylea) injections would prevent progression to clinically significant diabetic macula oedema with vision loss and proliferative diabetic retinopathy.
Patients are include if they are at Diabetic Retinopathy Severity Scale level 47 or 53 with no diabetic macula oedema with visual acuity of 6/9 or better. They are then randomized into two groups:
2.0mg of aflibercept injections (n = 200) given at baseline; 1, 2, and 4 months; and every 4 months through 2 years. Between 2 and 4 years, treatment was deferred if the eye had mild NPDR or better.
Sham (n = 199)
Aflibercept was administered in both groups if clinically significant diabetic macula oedema with vision loss (2 lines at 1 visit or 1 line at 2 consecutive visits) or high-risk proliferative diabetic retinopathy (PDR)
The outcome measures are development of clinically significant diabetic macula oedema with vision loss or proliferative diabetic retinopathy.
The result of this study showed that aflibercept is effective in preventing the development of proliferative diabetic retinopathy and clinically significant diabetic macula oedema within this study. The cumulative probability of developing clinically significant diabetic macula oedema with vision loss or PDR within 2 years was 16.3% with aflibercept vs 43.5% with sham. For the development of proliferative diabetic retinopathy irrespective of oedema, the cumulative probability was 13.5% with aflibercept vs 33.2% with sham. The results are summarised in the Kaplan-Meyer graphs below.
Similar to the PANORAMA study, there is no statistically significance in the visual acuity between the two groups in the DRCR.net protocol W study.
Clinical implications from these two studies
Accurate risk of progression of diabetic eye disease with modern diabetes management
The risk of progression to proliferative diabetic retinopathy was established from the ETDRS study published in 1991. Only the 5 year rate is available and this is summarized in the table below. Both the PANORAMA study and the DRCR.net protocol W study reported a similar rate of 30% progression to proliferative diabetic retinopathy over 2 years. This information is valuable for the risk assessment of patients with non-proliferative diabetic retinopathy.
Risk of progression to proliferative diabetic retinopathy from ETDRS
Risk of the development of proliferative diabetic retinopathy from the Wisconsin Epidemiologic Study of Diabetic Retinopathy (from Ryan’s Retina)
The type and duration of diabetes has a strong influence in the development of proliferative diabetic retinopathy. Young patients with insulin dependent diabetes (almost exclusively type I) have a much higher rate of proliferative disease comparing to older patients with insulin dependent diabetes (a mixture of type I and II). The rate of progression is influenced by a number of other risk factors, including blood glucose control, presence of diabetic neuropathy, decreased hematocrit, increased serum triglyceride, and decreased plasma albumin (ETDRS report 18).
I am hopeful that the teams in the PANORAMA study and in the DRCR.net study will publish follow-up analysis on the risk factors for progression to proliferative diabetic retinopathy in the control group to guide therapy.
Patients on anti-VEGF therapy for diabetic macula oedema are unlikely to have progression in their peripheral diabetic retinopathy, but this protective effect is lost once the treatment is stopped
It is clear from this study that, if a patient has moderate to severe non proliferative diabetic retinopathy and is on anti-VEGF therapy for diabetic macula oedema, anti-VEGF therapy is protective against progression to proliferative diabetic retinopathy. Their fundus appearance can also improve due to anti-VEGF therapy. The issue is that if anti-VEGF therapy is stopped due to an improvement of diabetic macula oedema or if the patient is switched to a steroid based agent (triamcinolone or ozurdex), this protective effect is lost.
There are studies looking at the effect of anti-VEGF therapy on the amount of peripheral ischaemia and there are conflicting report if the retina become re-perfused despite an improvement in the DRSS level (summarised here and here). In my personal experience, there has been a number of patients with worsening DRSS level and even progression to proliferative diabetic retinopathy after anti-VEGF therapy has been stopped. There is a signal from the PANORAMA study in the aflibercept 2q8/PRN group regarding this risk - 41 eyes in this group received no injections in the second year, and 4 eyes (9.8%) developed neovascular complications. However, none of the eyes which received 1 or more injections in the second year developed any neovascular complications.
Clinician will need to be careful with monitoring diabetic patients coming off anti-VEGF therapy, and not to be overly reliant on fundal findings and DRSS level in risk assessment. Wide-field Optos fluorescein angiogram or OCT angiography are useful in these cases to give an objective assessment of peripheral ischaemia and the risk of progression.
Early treatment with anti-VEGF maybe justifiable in some patients with moderate to severe non-proliferative diabetic retinopathy
As neither of these studies showed a visual acuity benefit, it is difficult to justify the recommendation of early anti-VEGF therapy to all patients with moderate to severe non proliferative diabetic retinopathy.
It is known from DRCR.net protocol V that patients with diabetic macula oedema and good visual acuity can be safely observed without anti-VEGF therapy. Therefore, it makes little sense to treat patients without oedema with anti-VEGF therapy to prevent diabetic macula oedema.
Traditionally, prophylactic PRP has been used for patients with severe non-proliferative diabetic retinopathy to prevent progression to proliferative disease, but only in cases if ongoing regular follow up is impossible. One cannot make the same argument for the use of anti-VEGF in moderate to severe non-proliferative diabetic retinopathy cases as ongoing follow up is required for the injection therapy. At this stage, the only indication for early anti-VEGF therapy is in patients with poorly controlled diabetes (particularly type I diabetes) with severe non-proliferative diabetic retinopathy, whom are likely to be able to achieve good control quickly. In these patients, a limited course of anti-VEGF therapy will be useful to prevent progression in the short term, and hopefully in the long term there will be some recovery of the retina due to an improvement in diabetes control and thus avoiding laser.