Aflibercept anD navigateD vErsus coNvensional Laser in Diabetic macUlar edeMa

Overview

Diabetic macular edema (DME) in diabetic retinopathy (DR) is the leading cause of visual impairment among the 300,000 Danish patients with diabetes (DM) and will in time affect 29% of patients. Because of DME, 550 intravitreal injections were given at Odense University Hospital in 2014 with medicine expenses approximating 3.3 million DKK. With an increasing prevalence of diabetes, the number is expected to rise significantly for the years to come. The investigators hypothesize that combination therapy with intravitreal aflibercept and a new computer navigated photocoagulations system (Navilas®) leads to a decreased need for intravitreal injections. Further, the investigators wish to identify retinal risk markers for DME treatment outcome to assist individualized treatment planning. The evaluation of the baseline level of macular ischemia as marker of successful treatment outcome is of particular interest as this is still highly debatable and may prove a significant prognostic factor of anatomical and functional outcome to anti-VEGF treatment. The ADDENDUM study (four-year part-time PhD study) is a 12-month prospective randomized 1:1 study to compare intravitreal aflibercept and Navilas® laser (Group A) with intravitreal aflibercept and conventional Pascal laser (Group B) in the treatment of DME. Eligibility criteria: DM, age 18-99, clinically significant macular edema, central retinal thickness \> 300 μm, best corrected visual acuity 35-75 Early Treatment Diabetic Retinopathy Study (ETDRS) letters. The investigators believe that this study holds the potential to set precedent for a new gold standard of DME-treatment with increased treatment effect, reduced risks and a more cost-effective approach.

Introduction: Diabetic retinopathy (DR) is the most common complication among the 300,000 Danish patients with diabetes. Amongst those with diabetes, diabetic macular edema (DME) is a leading cause of visual impairment and will in time affect 29% of all patients. It was established by the Early Treatment Diabetic Retinopathy Study (ETDRS) that focal/grid laser photocoagulation reduces the risk of visual loss in patients with DME but with a small likelihood of visual improvement. In recent years intravitreal anti-VEGF agents like bevacizumab, ranibizumab and aflibercept have consistently demonstrated efficacy and visual improvement in DME-treatment. However, a high number of injections is needed per patient to achieve sustainable visual improvement. Diabetic Retinopathy Clinical Research Network (DRCR.net) showed, that the median number of intravitreal ranibizumab injections was 9 in the first and 3 in the second year. The burden of intravitreal injections is a concern and it is substantial both for patients as well as the health care system. Because of DME, 550 intravitreal injections were given at Odense University Hospital in 2014. The medicine expenses of this approximate 3.3 million DKK and with an increasing prevalence of diabetes the number is expected to rise significantly for the years to come. This calls for initiatives like a more cost-effective treatment and a better understanding of underlying predictors of successful treatment in order to improve and individualize treatment. Navilas®(OD-OS GmbH, Teltow, Germany) is a navigated laser photocoagulation system that combines fluorescein angiography with image stabilization and tracking. This increases the capability to localize and then subsequently treat leaking microaneurysms. The ability to preplan and deliver planned spots in an automatic mode led to the observation that navigated laser treatment reduces the need for intravitreal injections of bevacizumab and ranibizumab. In a 12-month prospective study of patients with center-involving DME Liegl et al compared navigated laser and ranibizumab with ranibizumab monotherapy. All patients initially received 3 monthly ranibizumab injections and additional injections pro re nata. In addition, patients in the combination group received navigated laser photocoagulation after the loading phase. The study demonstrated a similar improvement in best-corrected visual acuity (BCVA) but a significant lower need for additional intravitreal injections (35% vs. 84%, p\<0.001) as well as a lower number of additional injections in the combination group (mean 0.9 vs. 3.9, p\<0.001). Aflibercept replaced ranibizumab as first-choice treatment in AMD in Denmark, while in DME-treatment results indicate a lower number of intravitreal injections required as well as a greater improvement in visual acuity, especially amongst those with worst visual acuity at baseline, as compared to ranibizumab. However, the combination of aflibercept and navigated laser therapy so far has not been tested in DME. In present standard regime of DME-treatment, both aflibercept and ranibizumab are prescribed as drugs of first. Features of the geometrical pattern of retinal vascular system have consistently been correlated to diabetic retinopathy (DR). In earlier studies the investigators demonstrated that retinal vascular diameter and fractal dimension were independently able to predict long-term microvascular complications in diabetes (proliferative DR, nephropathy and neuropathy). In DME, the investigators reported that laser photocoagulation was associated with a lower macular retinal vascular diameter, possibly due to autoregulated vascular narrowing given the lower oxygen demand after laser therapy. The investigators speculate that changes in retinal vascular geometry may be used as a measure of treatment response but this has not been tested in any prospective studies. Retinal vessel oxygen saturation is another non-invasive measurement of retinal function. Oxymap T1 (Oxymap, Reykjavik, Iceland) has made it possible to evaluate the metabolic changes in the retina. It has been demonstrated by Jørgensen et al that patients with DME have a higher oxygen saturation in retinal venules, but the treatment response on retinal vessel oxygen saturation is still unknown. The effect of macular ischemia on DME is still debatable. Macular ischemia and its impact on the functional and anatomical outcome after intravitreal injections of anti-VEGF have only been studied in a few short-term studies. In a case-control study, Douvali et al demonstrated that after six months in patients treated with ranibizumab for DME there was a negative functional outcome for those who had macular ischemia at baseline as compared to those who were non-ischemic. However, the effect of macular ischemia on the functional and anatomical outcome of DME-treatment has not been demonstrated in aflibercept-treatment. Primary goals: In a randomized 12-months prospective 1:1 study of patients with diabetic macular edema, ADDENDUM aims to (1) examine the treatment response of intravitreal aflibercept and navigated laser as compared to intravitreal aflibercept and conventional Pascal laser, (2) identify non-invasive retinal risk factors (retinal vessel diameter, fractal dimension, and oxygen saturation) for successful treatment outcome, and (3) evaluate baseline level of macular ischemia as marker of successful treatment outcome. Hypothesis: * Combination therapy with intravitreal aflibercept and navigated laser leads to a decreased need for intravitreal injections. * Non-invasive retinal markers make it possible to predict treatment response in patients with DME. Laser treatment leads to lower retinal vascular diameters, lower fractal dimensions and higher retinal oxygen saturation in patients with a successful treatment outcome (irrespectively of treatment group). * The level of macular ischemia at baseline is negatively correlated with functional and anatomical outcome in patients treated with aflibercept for DME (irrespectively of treatment group). Trial design: Period: 1 September 2015 - 31 August 2019 Setup: 12-month randomized prospective 1:1 study to compare intravitreal aflibercept and navigated Navilas® laser (group A, n=24) with intravitreal aflibercept and conventional Pascal laser (group B, n=24). Plan: * Patients will be recruited in the Region of Southern Denmark and included in the study based on a clinical examination. Patients will be randomized at baseline (BL). Patients with both eyes eligible will participate with one eye in each group. * Patients will receive intravitreal treatment at month (M) 0, 1 and 2 (see Treatment scheme). At M3 laser treatment will be performed if possible as determined by the examiner. * From M4 patients will then be followed monthly with additional intravitreal injections given pro re nata. * Criteria for re-treatment will be: 1. Central retinal thickness \> 20% from lowest measurement. 2. Loss in BCVA\>5 letters as compared to baseline. * Eyes not suitable for laser treatment by M3 will continue treatment in a monthly pro re nata intravitreal regimen. If laser has not been given by the end of M5 patients will be withdrawn from the study. Examination: * Systemic biomarkers: Age, type of diabetes, duration of diabetes, health history, smoking status, blood pressure, body mass index, hemoglobin A1c, glomerular filtration rate, serum creatinine (BL). * Diabetes therapy (BL, M12). * ETDRS BCVA measurement (BL, M0-M12). * Intraocular pressure (BL, M12). * Cataract LOCS grading. * Heidelberg fluorescein angiography, incl. measurement of area of foveal avascular zone \[FAZ\] (BL, M3, M12). * Topcon 3D OCT-2000 spectral domain OCT (BL, M0-M12). * Retinal vessel geometry measurements (BL, M3, M6, M12). * Oxymap retinal oximetry measurement (BL, M3, M6, M12). * Optos widefield retinal imaging (BL, M3, M6, M12). Primary endpoint: \- Percentage of eyes that received additional aflibercept injections after laser at month 12 in group A and B (Paper I). Secondary endpoints: * Number of additional aflibercept injections after laser at month 12 in group A and B (Paper I). * Change in ETDRS visual acuity from baseline to month 12 in group A and B (Paper I). * Change in global and macular retinal vessel geometric markers (retinal vascular diameter and fractal dimension) from baseline to month 12 in group A and B (Paper II). * Change in global and macular oxygen saturation from baseline to month 12 in group A and B (Paper II). * Evaluation of non-invasive retinal risk factors (retinal vascular geometric markers and oxygen saturation) for successful treatment outcome at month 12 (in multiple logistic regression model adjusted for age, sex, duration of diabetes, smoking, blood pressure, body mass index, hemoglobin A1c and renal markers) (Paper II). * Change in level of macular ischemia (area of FAZ) from baseline to month 12 in group A and B (Paper III). * Evaluation of baseline area of FAZ as marker of successful treatment outcome at month 12 (in multiple logistic regression model adjusted for age, sex, duration of diabetes, smoking, blood pressure, body mass index, hemoglobin A1c and renal markers) (Paper III). Clinical implication and scientific outcome: Combination therapy with intravitreal aflibercept and navigated laser photocoagulation is likely to be the future in DME-treatment. As compared to intravitreal monotherapy, the fewer injections and lower cost meets the demand of the patients as well as those of society. Identification of retinal risk markers for DME treatment outcome offers an attractive assistance in the guidance of individualized treatment planning. The present study will be performed at a PhD Study at University of Southern Denmark. Three papers in high-ranking international peer-reviewed journals have been planned. All results will be published irrespective positive or negative outcome. Facilities: A research laboratory has been established at the Department of Ophthalmology, Odense University Hospital. Necessary equipment for the study as well as office facilities will be available at the department. Oxymap T1, Pascal laser, Navilas® laser and software for retinal vascular measurement analysis (IVAN image analysis software and Singapore Institute Vessel Assessment-Fractal version 1.0) will be available. Rental of a Topcon 3D OCT-2000 will be included in the budget. Statistical calculations: The primary endpoint of the present study will be the percentage of eyes that will need additional intravitreal injections after laser. Sample size calculations were based on the study by Liegl et al that evaluated the effect of combined intravitreal ranibizumab and navigated laser treatment versus ranibizumab monotherapy in DME. In that study additional injections needed was 35% and 85%, respectively. Given a probability level of 0.05 and a statistical power level of 90%, a study size of 19 patients in each group will be needed in the present study. In order to compensate for a 20% drop-out rate, 24 patients will be included in each group.