Frank G. Holz, MD, FEBO, FARVO

University of Bonn

Geographic atrophy (GA) due to age-related macular degeneration (AMD) remains a huge unmet need, particularly after failure of various previous interventional trials. As of today, we have no efficacious treatment for our patients with GA associated with severe visual loss. Following the recent positive phase 2 study data from Apellis for its C3 inhibitor APL-2, it is interesting to now learn about the positive results of the phase 2b study of the C5 inhibitor avacincaptad pegol (Zimura) by Iveric bio. The complement system obviously plays an important role in AMD pathogenesis, as many genetic studies have shown with polymorphisms of the CFH gene and other complement genes impacting the onset of disease. The role of complement in the progression of any phenotype of AMD is less clear. In the Zimura study, reduction in the mean rate of geographic atrophy lesion growth over 1 year was 27.38% for the Zimura 2-mg group compared to the sham control group, and 27.81% for the Zimura 4-mg group compared to the sham group. The differences were reported to be statistically significant. Further details are not yet available, and it would be interesting to learn more about the trial data, including plots of the lesion growth in the respective arms to see at what time point there was a divergence between treatment arms and control group. Compared to the 1-mg dose, there was a suggestion of dose response for the 2-mg group with regard to the treatment effect of slowing lesion expansion. For evaluation of efficacy and safety of the C5 inhibition approach in advanced dry AMD, we would like to see the company move forward to a prospective, controlled phase 3 study.

Srinivas Sadda, MD

Doheny Eye Institute

As a practicing retina specialist, the lack of an effective treatment for GA, has been a source of great frustration. Recent failures in phase 3 trials for potentially promising agents has only added to the disappointment. Consequently, the recent announcement that Zimura met its prespecified primary endpoint in demonstrating a significant reduction in the rate of progression of GA without significant safety concerns, is certainly an encouraging development. While we are all hopeful, as physicians caring for AMD patients, that these promising results are confirmed in future phase 3 trials, there are some aspects of the results that warrant further study. Notably, there appeared to be a higher numerical rate (3-fold) of incident choroidal neovascularization (CNV) in the treated eyes compared to the fellow untreated eyes or the sham control. This may not necessarily be negative, because there are some data to suggest that type 1 CNV may have a protective effective against atrophy — and in some ways, this observation appears similar to the APL-2 phase 2 results. However, it would be important to confirm that there was no baseline imbalance in nonexudative CNV between the treatment arms, which might confound the apparent differences in GA growth rate. Finally, while the results are promising, the fact that the growth rate was only reduced by approximately 28% highlights the fact that there is still more for us to do to save vision for our patients with AMD.

Marco A. Zarbin, MD, PhD

Rutgers New Jersey Medical School

The phase 2b Zimura data are interesting and exciting for 2 reasons. First, the results provide additional evidence that complement inhibition may retard the growth of GA in patients with AMD. The involvement of complement in the pathogenesis of AMD has been established beyond a reasonable doubt through a confluence of evidence derived from genetic studies, histopathological studies of human postmortem eyes, and preclinical models of the disease. The failure of smaller complement inhibition trials in which inhibitors were administered systemically (eg, eculizumab) and larger trials in which they were administered intravitreally (eg, lampalizumab) led many physicians, myself included, to question whether complement inhibition might not work for patients for relatively advanced dry AMD. The results of this large phase 2b study are fully consistent with those of the Apellis trial, which differs from the Zimura trial in that Apellis used the C3 inhibitor APL-2, whereas Zimura is a C5 inhibitor. Both trials are phase 2 studies, so additional confirmation of reproducibility awaits the results of planned and ongoing phase 3 studies, but the similarity in the degree of inhibition of GA progression in the two trials is striking.

The second aspect of the results that I find interesting involves the differences in the pharmacology of the agents used in the 2 trials. In some ways, inhibition of C5 action may be more attractive then inhibition of C3, because it allows some pathways of the innate immune response to remain active, such as C3a and C3b formation. Although it might seem counterintuitive, C3 activation has some salutary benefits. It reduces photoreceptor death, for example, in preclinical models of human retinal degenerative disease. This benefit occurs because C3 activation results in opsonization and phagocytosis of photoreceptor debris and reduces the production of neurotoxic inflammatory mediators by microglia. So, blockade of C5 activation preserves some homeostatic aspects of C3 activation while inhibiting formation of the membrane attack complex that we presume contributes to photoreceptor death in GA. C5 antagonism also helps to inhibit inflammasome activation. Activation of the inflammasome by C5a has been shown to induce RPE degeneration, which is another prominent feature of GA. In contrast, C3 blockade will prevent formation of C3a and C3b as well as C5a and C5b. Whether these differences in pharmacology and pathway inhibition will result in differences in efficacy or safety of the 2 agents remains to be seen. I note, for example, that the incidence of choroidal new vessels in the Zimura trial was much lower than what was reported in the Apellis APL-2 trial, although it was greater than in the sham control group, as was the case for the APL-2. The potentially better safety profile of Zimura may be due to its further downstream blockage of complement cascade. Was the apparent difference in the safety profile just the result of chance, or does it reflect something intrinsic to the different pharmacological approaches employed? Only additional data and follow-up in larger ongoing and upcoming trials will provide a definite answer to this question.

Kourous Rezaei, MD

Chief Medical Officer, Iveric bio

Iveric bio recently reported that Zimura, a novel complement factor C5 inhibitor, met its prespecified primary endpoint in reducing the rate of GA growth in patients with dry AMD in a randomized, double-masked, sham-controlled clinical trial that the company believes can serve as the first of 2 pivotal trials for registration purposes. The reduction in the mean rate of GA growth over 12 months was 27.38% (P=.0072) for the Zimura 2-mg group compared to the corresponding sham control group, and 27.81% (P=.0051) for the Zimura 4-mg group compared to the corresponding sham control group. These data for both dose groups were statistically significant. Although efficacy data from patients receiving Zimura 1 mg was not part of the prespecified statistical analysis, preliminary descriptive analysis indicated that, on average, the percent GA growth from baseline to month 12 for Zimura 1-mg group was less than for the corresponding sham control group. Therefore, the overall data for Zimura suggest a dose response relationship across treatment groups.

Zimura was generally well tolerated over 12 months of administration. Over the first 12 months of the trial, there were no reported Zimura-related adverse events or inflammation, and there have been no discontinuations attributed by investigators to Zimura from the trial. Further, there have been no ocular serious adverse events, no retinal detachments, no Zimura-related increases in intraocular pressure, and no cases of endophthalmitis reported in the study eye in this ongoing clinical trial. During the first 12 months of the trial, the incidence of CNV in the untreated fellow eye was 10 patients (3.5%). In the study eye, incidence of CNV was 3 patients (2.7%) in the sham control group, 6 patients (9.0%) in the Zimura 2-mg group, and 8 patients (9.6%) in the Zimura 4-mg group. We believe the observed CNV incidence rate in the study eye for patients receiving Zimura compared to sham and the untreated fellow eyes during the first 12 months of the trial is within an acceptable range and lower than what has been reported in the literature for a complement inhibitor that is currently in development in GA.

We believe that the superior safety profile of Zimura and lower rate of CNV incidence may potentially be in part due to blocking at the level of C5 and not C3. Therefore, based on the current safety profile of Zimura, we believe we may have an advantageous safety profile over our competitors in the patient population that we are seeking to treat. We have begun the planning of our second pivotal trial and plan to enroll our first patient in the first quarter of 2020. RP

Editor’s note: This article is part of a special edition of Retinal Physician that was supported by REGENXBIO.