Study Raises Questions About Effects, Risks of Drugs for Geographic Atrophy

DENVER — An analysis of real-world data suggested that use of complement inhibitors to treat geographic atrophy (GA) is associated with an increase in incidence of exudative age-related macular degeneration (eAMD) and progression to subfoveal GA.

Patients treated for a year had a roughly fourfold increased risk of developing eAMD as compared with a control group and more than a twofold higher risk of progression to subfoveal GA. The hazard for low vision or blindness was increased with pegcetacoplan (Syfovre) and avacincaptad pegol (Izervay) but reached statistical significance only for the latter drug. Intraocular inflammation rarely occurred, and the data suggested no increased risk of systemic adverse events.

The findings highlight the need for additional studies to determine more clearly the risks and benefits of the drugs, said David Ramsey, MD, PhD, of the University of Massachusetts Chan Medical School in Burlington, at the Association for Research in Vision and Ophthalmology meeting.

“In the population we studied, there was no evidence of any benefit on progression towards subfoveal GA during the 1-year follow-up period,” said Ramsey.

Based on data from the TriNetX clinical network, the study had multiple limitations, Ramsey acknowledged, including:

• Limitations inherent to the retrospective, observational design
• The potential for residual confounding despite propensity matching
• Potential selection bias for treatment
• Limited patient-level data
• Variable data quality and completeness
• Limited ophthalmic data, such as imaging and other vision assessments

Ramsey also pointed to limitations associated with reliance on diagnostic codes.

“The outcomes rely exclusively upon the coding accuracy of the clinicians who contribute data to this repository, and there are very broad definitions for our outcomes, so with this heterogeneous group we may mis-estimate the outcomes,” he said. “The coded GA population likely underestimates the true disease burden and may not really reflect the full spectrum of GA severity. Treatment patterns were very likely influenced by access and early adoption of the drugs.”

Session moderator Philip Rosenfeld, MD, PhD, of the University of Miami Bascom Palmer Eye Institute, wasted no time in questioning the reliability of the data.

“Can this data be believed?” he asked. “I ask you that because clinicians are notoriously bad in making the diagnosis simply of subfoveal, non-subfoveal. More importantly, you don’t know what they were using to diagnose the wet age-related macular degeneration.”

Rosenfeld referred to an ARVO poster presentation showing that treatment with pegcetacoplan increases exudation “that can be easily misdiagnosed if you’re using an OCT [optical coherence tomography]-only approach to the diagnosis of macular neovascularization. It goes away if you extend the interval from 4 to 8 weeks.”

“So this is pretty suspect,” he added. “The only way you could even attempt to make this relevant to the real world is to see how it compares to clinical trial data. I believe that those numbers [in the study] align very well to published clinical trial data, but we don’t know that because you didn’t show any comparison to published clinical trial data.”

Ramsey agreed that including data from published clinical trials would have been helpful.

Approved in 2023, pegcetacoplan and avacincaptad pegol slow GA lesion growth, but the drugs’ ability to slow vision loss in patients with late-stage AMD has not been established, Ramsey noted in his introduction. Functional outcomes in clinical trials varied. Anatomical benefits may not translate into gains in visual acuity, and “persistent structure-function disconnect has been debated by many,” he said.

Injection burden, need for long-term treatment and follow-up, and need for imaging studies all have implications for patient counseling.

“When we counsel patients about these medications, we need to set expectations,” he said. “There are, of course, issues related to cost, coverage, travel, time burden, and, from the vitreoretinal specialist perspective, clinic flow.”

Ramsey and colleagues wanted to evaluate utilization patterns of the two drugs in a clinical population with GA. Objectives included studying associations between clinical use and progression to subfoveal GA, new-onset low vision, incident eAMD, and other ocular adverse events.

The investigators queried the TriNetX database to identify patients with GA treated or not treated with complement inhibitors. Eligible patients had at least 1 year of follow-up. The investigators identified 576 patients treated with either of the two complement inhibitors and a propensity-matched cohort of 576 patients with GA not treated with either drug. None of the patients had a history of eAMD documented by an ICD code.

The 576 patients treated with a complement inhibitor had a median age of 80.6, and women accounted for 66% of the group. Sixty-eight percent of the patients were treated with pegcetacoplan. Two-thirds of the patients continued treatment, and 3.3% switched drugs during treatment.

Overall, the analysis showed a hazard ratio of 4.05 for eAMD (95% CI 2.04-8.05, P<0.001) versus the control population, 2.49 for progression to subfoveal GA (95% CI 1.45-4.27, P=0.001), and 1.80 for low vision or blindness (95% CI 1.05-3.11, P=0.032). Use of complement inhibitors was associated with an increased risk of uveitis and intraocular inflammation (HR 3.04, 95% CI 1.08-8.55, P=0.027) but not glaucoma/ocular hypertension or retinal detachments/breaks. Endophthalmitis and vascularis occurred infrequently.

Results were similar for the individual drugs with the exception of development of low vision/blindness, which did not reach statistical significance for pegcetacoplan (HR 1.63, 95% CI 0.81-3.25) but did for avacincaptad pegol (HR 6.89, 95% CI 1.96-24.2, P=0.001).

Rates of cardiovascular complications were similar between the study and control groups.