[The Conundrum of Relationships in the Multivariate Dataset of nvAMD Treatment]. / Die Frage nach Zusammenhängen in multivariaten Datensätzen zur Behandlung der nvAMD.

Background The evaluation of relationships in clinical case series is complicated by statistical pitfalls, one of which is spurious correlation. In this case, one or more potential factors are actually not related to a target measure, yet it may be wrongly inferred that they are. This may be due to either coincidence or the presence of a one or more confounding unregarded factors. In randomised clinical trials, it is attempted to eliminate such confounding factors or to control them by randomisation. However, this approach is not applicable in retrospective case series, and other statistical methods must be applied, such as multivariate analyses, taking into account all known variables. Patients/Material and Methods Statistical evaluation of the impact of different variables on the change in visual acuity after year 2 in a cohort of 1300 cases of nvAMD. The investigated variables are: visual acuity (VA) at baseline, central retinal thickness at baseline, age at baseline, mean interval between visits, number of injections, initial gain in VA after upload, and a change of the anti-VEGF compound. The target outcome measure was the change in VA as letter score at the end of year 2. Results All investigated variables have a marked impact on VA outcome when only single regression analysis is performed. Initial analysis of a switch in therapy gives a seemingly large effect on VA. However, multivariate analysis to avoid spurious correlations shows that only the following variables influence VA outcome with statistical significance: VA at baseline, initial gain in VA after upload, number of injections, mean interval between visits, and central retinal thickness at baseline. A switch in therapy had no statistically significant effect on VA outcome. Conclusion A switch in anti-VEGF therapy is not causally related to better VA outcome. In the statistical evaluation of non-randomised, retrospective case series, special attention must be given to potential confounding variables. When subgroups under investigation are not well balanced for confounding factors, misleading spurious correlations can result, leading to incorrect inferences.

[Real-life Data on the Treatment of Diabetic Macular Oedema in Germany]. / Daten zur Behandlungsrealität des diabetischen Makulaödems mit Ranibizumab.

Background Controlled prospective clinical trials on the treatment of diabetic macular oedema (DME) using anti-VEGF compounds show very good results in visual acuity gain over several years. To date, only limited data are available from comparable studies under real-life settings in clinical routine. However, real-life data from other indications for anti-VEGF treatment suggest that, in clinical routine, gain in visual acuity is less pronounced and cannot be maintained over a longer period of time, which is related to the significantly lower number of injections administered in clinical routine. Here we report a case series from our clinical routine of patients treated with ranibizumab for visual significant DME. Patients/Material and Methods Retrospective case series of 335 cases with follow-up of up to 3 years. All cases in our clinic treated for visual significant DME with at least one injection of ranibizumab were evaluated for the course of visual acuity gain and number of injections received. Results A mean gain in visual acuity (VA) of + 5.6 and + 3.7 letters was found in years one and two, respectively, with a median VA of logMAR 0.52 (decimal 0.3) at baseline. The steepest increase in VA is found during the first 4 months, with a less pronounced increase up to month 16. The mean number of injections was 5.7, 3.2 and 1.1 for years one, two, and three, respectively. Conclusions Results of real life treatment in clinical routine are found to fall behind controlled, prospective trials, as found to the findings in other anti-VEGF indications: mean gain in VA is lower than in prospective trials, and the initial gain cannot be fully maintained over a prolonged time period. Similarly, the number of injections received is markedly lower than in controlled trials. This can partially be explained by differences in study populations and a negative selection bias in longer term results. However, additional barriers that hamper timely treatment tailored to disease activity requirements must be identified and circumvented where possible.

[Time Course of Changes in Visual Acuity After a Single Injection of Aflibercept or Ranibizumab in Neovascular Age-related Macular Degeneration - Analysis of Aggregated Real Life Data]. / Der zeitliche Verlauf der Visusentwicklung nach einer Injektion von Aflibercept oder Ranibizumab bei der neovaskulären altersbedingten Makuladegeneration.

Background Treatment of neovascular age-related macular degeneration (nvAMD) under real life conditions may differ from controlled prospective trials with respect to the number of injections and long term preservation of visual acuity. In many instances, intervals for controls and re-injection cannot be maintained as frequently as required. This case series examines actual intervals for control and re-injection visits, in order to estimate how prolonged intervals have an impact on momentary visual acuity and how long term visual outcome is affected. Patients/Material and Methods Retrospective case series of 1,324 eyes with nvAMD, treated with a total of 8,150 ranibizumab injections (according to the PRN regimen) or 1,725 aflibercept injections (according to a fixed regimen), during the observation period of up to 3 years. The evaluation covered the time interval between visits, impact of this on the course of visual acuity, as well as the number of injections throughout treatment. Results Planned intervals of 4 or 8 weeks between visits were more often exceeded in the PRN regimen than with the fixed regimen. Visual acuity does not peak after 4 weeks, but only between 6 and 8 weeks. No statistically significant difference between aflibercept and ranibizumab was found. If the mean interval for re-injection was maintained at 4 - 6 weeks, this gave the greatest gains in visual acuity at end of years 1 and 2, respectively. Any prolongation of these intervals was accompanied by worse long term visual acuity. The fixed regimen is associated with consistently briefer re-treatment intervals during years 2 and 3, than with the PRN regimen. Conclusion Our data point to the importance of frequent controls accompanied by timely retreatment, as these have a major impact on visual outcome. It therefore appears to be more important to choose a treatment plan that facilitates frequent re-injections than to select either of the two compounds.

Performance optimization of current focusing and virtual electrode strategies in retinal implants.

The electrode configuration in an implanted visual prosthesis array affects the spatial electric field distribution within the retina, contributing to current focusing and virtual electrode (VE) stimulation strategies. In this paper, a finite element model incorporating various electrode configurations was used to study the interaction between electrode size and electrode-to-cell distance in current focusing and VE stimulation paradigms. The electrode array unit comprises an active electrode, six flanking return electrodes and a distant monopolar return. A quasi-monopolar (QMP) fraction is defined as the proportion of current which can be preferentially returned through the distant return, in comparison with the more adjacent flanking electrodes. The simulation results indicate that current focusing and VE strategies can be optimized by tuning the QMP fraction. The QMP fraction is adjusted to optimize the electric field spread based on retinal ganglion cell (RGC) density in the degenerate retina, thereby offsetting the effect of inhomogeneous distribution of surviving RGCs and leading to a uniform stimulation paradigm across electrodes. Importantly, there is negligible difference in functional performance across electrode configurations for distances less than the electrode diameter, implying that the stimulation mode does not significantly affect activation threshold or activated retinal area for electrode diameters greater than the retinal thickness. Furthermore, the QMP fraction has a significant effect on VE performance, defined by activation threshold and activated retinal area, when threshold current is evenly divided between two adjacent active electrodes.