A retina specialist shares his presentation from the 2023 ASRS Annual Meeting, which describes a study comparing the efficacy of 2 FDA-approved treatments for geographic atrophy based on an anchored matching-adjusted indirect comparison of phase 3 trial findings.
Paul Hahn, MD, PhD: Hello, my name is Paul Hahn, [MD, PhD], from NJ Retina. Today, I’m going to be presenting my paper from [the] ASRS [American Society of Retina Specialists Annual Meeting] [in Seattle, Washington] in July 2023 titled “Pegcetacoplan vs Avacincaptad Pegol in Patients With Geographic Atrophy: An Anchored Matching-Adjusted Indirect Comparison of the Phase 3 Trials.”
We currently have 2 complement inhibitors for geographic atrophy: pegcetacoplan and avacincaptad pegol. Pegcetacoplan was the first FDA-approved treatment for GA [geographic atrophy], having been approved in February 2023. Avacincaptad pegol was just approved earlier this month [August, 2023], and the slides as shown here are outdated as a result.
Pegcetacoplan and avacincaptad pegol target complement [proteins] but through slightly different mechanisms. Pegcetacoplan works a little further upstream, targeting C3 and C3b, whereas avacincaptad pegol targets C5.
Both have been shown to slow GA growth in [findings from] their pivotal phase 3 trials. Pegcetacoplan was studied in 2 phase 3 trials called OAKS and DERBY, whereas avacincaptad pegol was studied in a single phase 3 trial called GATHER2 [because] the GATHER1 trial was [a] mixed phase 2 and phase 3 study. Pegcetacoplan [was] looked at [in] both nonsubfoveal and subfoveal GA lesions in their clinical trials, with both monthly and every-other-month dosing, whereas avacincaptad pegol was studied in GATHER2 in only nonsubfoveal GA lesions and only with monthly dosing. Interestingly, pegcetacoplan has longer-term data available, demonstrating increasing effects on slowing down geographic atrophy growth over time.
Shown here is a very familiar graph to us. We see a reduction in GA growth in the combined OAKS and DERBY results, focusing only on nonsubfoveal lesions at 12 months, similar to what was [seen] in the avacincaptad pegol study [findings]. If we look at the top-line data here, it may be tempting to compare the raw numbers between the pegcetacoplan and avacincaptad pegol studies to try to get a sense of the relative efficacies. However, we know that this is not the correct approach. It is dangerous and frankly wrong to compare top-line data [such as] this. But there is a better way to do it, and that way is called the anchored matching-adjusted indirect comparison. This is a well-validated methodology used in multiple therapeutic areas and sometimes required by regulatory agencies for approval. In the absence of a head-to-head study, what this approach does is [try] to compare the relative efficacies of [findings from] 2 different studies in 2 different agents, in this case comparing pegcetacoplan vs avacincaptad pegol [in] patients with nonsubfoveal GA secondary to macular degeneration.
In this approach, the first step is to try to apply the inclusion-to-exclusion criteria from the more restrictive trial to the less restrictive trial to match those inclusion-to-exclusion criteria as best as possible. Specific to this, the inclusion-to-exclusion criteria from GATHER2 were applied to individual patient–level data from OAKS and DERBY to select a similar group of patients, including only those with nonsubfoveal GA, excluding those with fellow eye choroidal neovascularization, including a specific range of vision as specified in GATHER2, etc. Once we’ve matched the inclusion-to-exclusion criteria, the next step is to try to balance those 2 populations. And we do this by propensity score weighting to balance key baseline characteristics, which were selected a priori between the 2 trials. And those include things [such as] age, vision, pressure, as well as GA characteristics such as size, laterality, and focality.
After we apply this methodology, we see the sample sizes here. We go from about 320 patients or about 310 patients treated with pegcetacoplan in OAKS and DERBY for nonsubfoveal lesions to about 220 patients after filtering out and matching those criteria. And then we compare the different trial [findings]. Shown here are some of the actual data, and you can see here we compare [findings from] each of the trials, OAKS vs GATHER2, DERBY vs GATHER2, all anchored to a sham arm.
But I’m going to show you this more graphically, which may be easier to see. In this graph, we compare pegcetacoplan monthly to avacincaptad pegol monthly. If we look at the top row, comparing OAKS vs GATHER2 [findings], we see a 37% greater reduction with pegcetacoplan monthly compared with avacincaptad pegol monthly.
And this difference was statistically significant. If we compare DERBY vs GATHER2 results, we see a 12% greater reduction favoring pegcetacoplan monthly, but this difference was not statistically significant. But if we compare both OAKS and DERBY [findings] to GATHER2 [findings], then we find a 30% greater reduction favoring pegcetacoplan monthly, which was significant. If we also look at the pegcetacoplan every-other-month arm and compare that with more frequent dosing with avacincaptad pegol monthly, we find that in each of the comparisons, the numerical results favored pegcetacoplan every other month, but these were not statistically significant.
Of course, there are a lot of limitations to this type of analysis. This is a post hoc analysis of clinical trials. As such, the generalizability to the population and the conclusions need to be interpreted with caution. As a feature of this type of analysis, individual patient–level data were only available for OAKS and DERBY and not for GATHER2, and we therefore had to limit our baseline matching to characteristics that were reported or published for gather2. And other differences may exist in nonobserved characteristics. There were differences in study design and enrollment criteria. The effective sample sizes are reduced by this analysis, but through this analysis, we suggest that pegcetacoplan may have a better efficacy profile than avacincaptad pegol among patients with nonsubfoveal GA. Specifically, if we look at pegcetacoplan monthly, we saw a statistically significant reduction of greater than 30% in GA lesion growth at month 12 for the combined analysis of OAKS and DERBY compared with avacincaptad pegol at the same monthly frequency. One step further, there was no difference in efficacy between every-other-month pegcetacoplan dosing and monthly avacincaptad pegol dosing.
As we now have 2 different drugs [that are] FDA approved for the [management] of geographic atrophy, it’s going to be important for us as clinicians to try to figure out which of these drugs to use. This provides 1 piece of the data, and each of us will have to make our own decision as we start to incorporate these drugs into our armamentarium.
Transcript is AI generated and edited for clarity and readability.