• Question: What is the comparative effectiveness and safety of complement inhibitors in improving geographic atrophy (GA) in patients with age-related macular degeneration (AMD)?

Age-related macular degeneration (AMD), as the most common form of maculopathy, is a major cause of vision loss in elderly people and shows a steep increase in prevalence after the age of 50 years (Jonas et al., 2017; Wong et al., 2014; Vangsted et al., 2023; Vujosevic et al., 2023). Geographic atrophy (GA) is the advanced stage of AMD, which can lead to progressive and irreversible loss of visual function due to loss of the retinal photoreceptors, retinal pigment epithelium (RPE), and choriocapillaris (Nittala et al., 2022). In the absence of treatment, 66% of eyes with GA may lose vision or become severely visually impaired during a patients’ lifetime (Colijn et al., 2021). It is estimated that the number of cases with late AMD in 2050 would be 6.41 million (95% CrI: 3.37–13.22) worldwide (Wang et al., 2022).

For a long time, there were no effective drugs for treating GA secondary to AMD. The American Academy of Ophthalmology guidelines recommended that patients with advanced AMD consider using antioxidant, vitamin, and mineral supplements (Flaxel et al., 2020). However, recent clinical trials have shown promising advancement in treatment options, including diet therapy, antibody therapy, gene therapy, cell therapy, visual cycle modulators, photobiomodulation, and laser therapy. In February 2023, pegcetacoplan, a complement C3 cyclic peptide inhibitor, was approved by the Food and Drug Administration (FDA) as the first-line treatment to treat GA (Schachar, 2023). More recently, another complement inhibitor avacincaptad pegol, a C5 inhibitor, was also approved by the FDA for the treatment of GA secondary to AMD (Kang, 2023).

Numerous genetic and molecular studies have confirmed the significant role of the complement system in AMD, including genetic variants, overactivation of alternative pathway, inflammation, oxidative stress, lipid accumulation, and energy metabolism (Qin et al., 2021; Riley-Gillis et al., 2023). Potential therapeutic targets within the complement system include C1q, C3, C5, complement factors (B, D, H, and I), as well as membrane attack complex and properdin (Patel et al., 2022; Lambris et al., 1996; Mannes et al., 2020). Several randomized controlled trials (RCTs) have been conducted to evaluate the efficacy of complement inhibitors in treating GA. The aim of this network meta-analysis (NMA) was to compare and rank the different complement inhibitors in the treatment of GA secondary to AMD.

The network relationships between different outcomes in this study are illustrated in Figure 2 (network plot). All included RCTs that assessed changes in GA lesion size, BCVA, and SAE involved five drugs, nine interventions, and a total of 4,405 patients. This analysis resulted in 13 direct comparisons and four closed loops, as depicted in Figures 2A–C. Among these, the largest number of studies compared lampalizumab and pegcetacoplan with the sham (three studies). Additionally, there were two trials comparing changes in GA lesion size between avacincaptad pegol and sham groups, and one trial each involving CLG561 and LFG316. Excluding NCT02515942, nine registered RCTs reported on MNV. These studies included 4,292 patients and four drugs, resulting in 10 direct comparisons and three closed loops, as shown in Figure 2D. The intervention characteristic graph for outcomes is provided in Supplementary eAppendix S4, S1.

This Bayesian NMA is the first technique to quantitatively evaluate the comparative effectiveness and safety of different complement inhibitors for improving GA in patients with AMD. We included a total of ten RCTs with 4,405 participants, covering four types of complement inhibitors: C3, C5, Factor D, and properdin inhibitors. The NMA results revealed that, among the five different complement inhibitors (pegcetacoplan, avacincaptad pegol, lampalizumab, CLG561, and LFG316), pegcetacoplan and avacincaptad pegol demonstrated clinically significant effects in reducing GA progression. SUCRA analysis results indicated that avacincaptad pegol 2 mg had the highest efficacy in treating GA, followed by pegcetacoplan monthly and pegcetacoplan EOM. While pegcetacoplan monthly was associated with a high risk of MNV, avacincaptad pegol 2 mg showed favorable outcomes in terms of SAE and MNV.

The primary findings of this study indicate that, among the existing complement inhibitors, avacincaptad pegol may offer greater efficacy and safety in reducing progression of GA in patients with AMD. The complement system is a highly regulated protein network that can be activated in a cascaded manner and operates at the interface of innate and adaptive immunity (Reis et al., 2019; Whitmore et al., 2015). It comprises three interconnected pathways: the classical pathway, alternative pathway, and lectin pathway. These pathways converge at the cleavage of complement C3 and C5, leading to the formation of the membrane attack complex, which is crucial for pathogen cell death (Anderson et al., 2010). Genetic studies have identified significant associations between AMD and variants of several complement-associated genes, including complement factor H, complement factor B, and complement component 3, indicating local complement protein expression within the eye (Riley-Gillis et al., 2023; Anderson et al., 2010; Kim et al., 2022). The complement system plays a key role in the pathogenesis of GA (Boyer et al., 2017; Zheng et al., 2022). Macular drusen, characteristic of early and intermediate stages of AMD, can progress to advanced AMD, which is associated with substantial vision loss. Some complement components, such as fragments of C3 and C5, have been found within drusen (Mullins et al., 2000; Schramm et al., 2014). GA represents the advanced stage of AMD, involving atrophy of the photoreceptors, RPE, and choriocapillaris. On average, patients exhibit an annual absolute GA progression rate of 1.59 mm²/y (95% CI, 1.46–1.71), corresponding to a square root-transformed GA progression rate of 0.26 mm/y (95% CI, 0.24–0.27) (Pfau et al., 2020).

Previous reviews have highlighted the therapeutic potential of C3 and C5 inhibitors in treating GA (Kim et al., 2022; Kim et al., 2021; Mastellos et al., 2023; Rathi et al., 2023). Pegcetacoplan, targeting C3 and C3b, was the first complement inhibitor to receive FDA approval for marketing (Heier et al., 2023). In our study, the paired meta-analysis results of pegcetacoplan EOM versus sham indicated an MD of −0.29 [95CrI%: −0.44 to −0.13]. Similarly, the results for pegcetacoplan administered monthly versus sham indicated an MD of −0.38 [95CrI%: −0.57 to −0.19]. These findings are consistent with those of research conducted by Tzoumas et al. (2023). Subsequently, avacincaptad pegol, targeting C5, also received marketing approval following favorable Phase III results. Our paired meta-analysis indicated that avacincaptad pegol 2 mg versus sham resulted in an MD of −0.58 [95CrI%: −0.96 to −0.20] (see Supplementary eAppendix S7, S1). Although no statistical difference were found in the indirect comparisons of these three interventions, the SUSAR ranking indicated that avacincaptad pegol 2 mg as the most effective, followed by pegcetacoplan monthly and pegcetacoplan EOM. This suggests that avacincaptad pegol may be more potential than pegcetacoplan in improving outcomes related to GA progression. Furthermore, indirect comparison between avacincaptad pegol and pegcetacoplan and other therapies also reported significant differences in efficacy. Research has indicated that intravitreal injections with higher volume may lead to increased intraocular pressure (IOP) (Meyer et al., 2023; El Chehab et al., 2016). In our study, the rate of increased IOP increased with complement inhibitors compared to sham, with avacincaptad pegol 4 mg showing the highest rate of increased IOP at 22.8%.

None of the complement inhibitors demonstrated a statistically significant difference in improving BCVA compared to sham at 1 year, which is consistent with the findings of Anubhav Garg et al. (2023). Visual function is crucial to both patients and physicians, and BCVA is a fundamental measure visual function. However, it does not entirely capture meaningful functional visual decline throughout the macula and does not show a good correlation with patient’s symptoms (Sadda et al., 2016). Important visual function results also include contrast sensitivity, color, depth, and motion, as well as field of view (Bennett et al., 2019). The FDA encourages the development of new clinical endpoints to measure clinically significant effects in patients with retinal diseases (US Department of Health and Human Services FaDA, 2020), especially for studies focusing on the recovery or improvement of vision in patients with severe visual impairment (Bailey and Lovie-Kitchin, 2013). Dolly S Chang‘s study found that perilesional or responding macular sensitivity measured by microperimetry was a more sensitive endpoint than mean macular sensitivity for detecting functional deterioration (Chang et al., 2024). Additionally, the Multi-Luminance Mobility Test is considered to evaluate an individual's functional vision and ability to perform visually independent daily activities, providing standardized and quantifiable assessments for clinical trials (Chung et al., 2018; Maguire et al., 2019; Girgis and Lee, 2023). Several studies have reported emerging endpoints such as low luminance visual acuity (Liao et al., 2020; Holz et al., 2018) and National Eye Institute Visual Functioning Questionnaire 25-item Version Composite Score (Heier et al., 2023; Holz et al., 2018). However, since most studies have not reported these, we did not consider them in this analysis.

Direct and indirect comparisons of all interventions did not reveal statistically significant differences in SAEs, which is consistent with findings from other studies on intravitreal drugs (Garg et al., 2023). Additionally, our study found that pegcetacoplan monthly was associated with a high risk of MNV (OR: 4.30, 95% CrI: 1.48–16.72). In addition to pegcetacoplan monthly, Tzoumas et al., (2023) also suggested that avacincaptad pegol might substantially increase the risk of MNV or exudative AMD compared to sham over a year. Despite the fact that there is much hope and marketing about tissue preservation with pegcetacoplan, there are many safety and efficacy concerns that may interrupt widespread adoption by clinicians (Schachar, 2023; Nadeem et al., 2023). In contrast, avacincaptad pegol 2 mg showed favorable outcomes in terms of SAEs and MNV. Overall, based on both efficacy and safety, avacincaptad pegol 2 mg appears to be the most promising choice among the complement inhibitors for treating GA secondary to AMD.

Although avacincaptad pegol and pegcetacoplan have received FDA approval for the treatment of GA, this approve is primarily based on their efficacy in slowing GA progression, as measured by anatomical endpoints. Phase 3 clinical trials, however, have not demonstrated any significant functional improvement after 1 or 2 years of treatment (Khanani et al., 2023; Heier et al., 2023). Do these drugs represent “clinically relevant outcomes”? Given that visual loss typically manifests only in the advanced stages of GA, utilizing visual acuity as the primary outcome measure in GA trials would necessitate study durations extending well beyond those of most large-scale randomized clinical trials. This makes it challenging to identify practical and clinically meaningful endpoints for GA. While FAF imaging serves as the standard for FDA assessment of increase in GA size, the correlation between functional changes and FAF at the GA boundary remains unclear. Consequently, there is a notable disconnect between anatomical and functional outcomes in existing clinical trials. Csaky et al., (2024) suggested that future clinical trials should combine FAF and optical coherence tomography imaging, while quantifying GA boundaries and incorporating functional endpoints, in order to develop more effective treatment strategies.

This research has several limitations. First, the analysis is based on a limited number of studies comparing sham versus five complement inhibitors. The included studies differed in patient populations, design, and statistical methods, contributing to potential heterogeneity that may affect the precision of our estimates. Factors such as patient age and lesion location could further contribute to this heterogeneity. Additionally, the relatively small number of randomized controlled trials (RCTs) may impact the robustness of some comparisons. Second, using a 1-year endpoint may underestimate the long-term benefits or risks associated with complement inhibitors. Third, the lack of statistically significant differences in changes in BCVA suggests that the chosen outcome measure might lack sensitivity. Given the multifactorial nature of visual function, a more nuanced evaluation that includes patient-reported outcomes and functional vision assessments is needed to fully capture the impact of complement inhibitors on visual acuity. Despite these limitations, significant progress is being made in the development and research of complement inhibitors for treating GA. More large-scale real-world studies are necessary to confirm these findings in the future.

The results of this network meta-analysis indicate that avacincaptad pegol and pegcetacoplan have clinically significant effects in reducing the progression of GA in patients with AMD. The Bayesian NMA indicates that avacincaptad pegol 2 mg is the most effective complement inhibitor with better safety for improving GA in AMD.