Compstatins: The dawn of clinical C3-targeted complement inhibition

Abstract

Despite the growing recognition of the complement system as a major contributor to a variety of clinical conditions, the therapeutic arsenal has remained scarce. The introduction of an anti-C5 antibody in 2007 raised the confidence in complement-targeted therapy. However, it became apparent that inhibition of late-stage effector generation might not be sufficient in multifactorial complement disorders. Upstream intervention at the level of C3 activation has therefore been considered promising. The approval of pegcetacoplan, a C3 inhibitor of the compstatin family, in 2021 served as critical validation of C3-targeted treatment. This review delineates the evolution of the compstatin family from its academic origins to the clinic, and highlights current and potential future applications of this promising drug class in complement diseases.

The complement system: a master of immune surveillance.

The complement system is an ancient part of innate immunity, which evolved to recognize and eliminate danger, such as pathogenic intruders or cellular debris. In recent years it became apparent that it is also involved in developmental processes (e.g., synaptic pruning), tissue repair/regeneration and shaping of adaptive immune responses [1–3]. Whereas complement has been traditionally described as humoral cascade system, newer studies reported expression and secretion of complement components by tissues and infiltrating cells, as well as intracellular complement turnover [4].

Most of complement’s physiological activities can be ascribed to its canonical function as a cascade system that is initiated by three pathways (Figure 1). The ‘classical’ pathway (CP) recognizes immune complexes, which provide the platform for assembly of the C1 complex comprising the pattern recognition receptor C1q and its associated proteases C1r and C1s [5–7]. Similarly, the ‘lectin’ pathway (LP) is triggered by carbohydrate signatures that are recognized by mannose-binding lectin (MBL, see Glossary), ficolins (Fc) and/or collectins (CL-11) in complex with MBL-associated serine proteases (MASPs) [8]. Both pathways confer proteolytic activity to cleave complement proteins C2 and C4, leading to the assembly of C3 convertase complexes (i.e., C4b2b). These convertases activate the abundant plasma protein C3 to release an anaphylatoxin (C3a) and generate the opsonin C3b that is deposited on activating surfaces. The ‘alternative’ pathway (AP) is activated spontaneously at a low rate in solution by hydrolysis of C3 to C3(H₂O), which can be accelerated by contact with different surfaces (tick-over) [9]. The AP also serves as amplifying mechanism, as surface-deposited C3b builds the basis for the formation of AP C3 convertases [10]. Upon binding to C3b, the serine protease factor B (FB) becomes proteolytically activated by protease factor D to result in the C3bBb convertase complex, which binds and activates even more C3. High-density C3b deposition favors the generation of C5 convertases that activate C5 to generate the anaphylatoxin C5a and the C5b fragment, which serves as nucleus for the assembly of the membrane attack complex (MAC; terminal pathway) [11]. Although the lytic activity of MAC is the most recognizable effector function, it only applies to some susceptible cells. In many cases, the attraction and activation of immune cells by anaphylatoxins, immune shuttling via complement receptor 1 (CR1), adaptive immune stimulation via CR2, and phagocytosis of opsonized particles via CR3 and CR4 are more effective in eliminating threats and engaging in crosstalk with other host defense systems (e.g., coagulation, platelet activation, cytokine release) [12].

Figure 1: Schematic overview of the complement system and therapeutic points of intervention.

The complement cascade is initiated either through recognition of immune complexes (classical pathway) or carbohydrate signatures (lectin pathway) by pattern recognition receptors (i.e., C1q) or via fluid-phase formation of C3(H₂O) (alternative pathway). All pathways lead to the formation of C3 convertase (C4b2b or C3bBb), which cleaves the central protein C3 into the small anaphylatoxin C3a and the opsonin C3b. C3b can form new AP convertases with FB and FD, leading to an amplification of complement activation. At a sufficient density of C3b C3 convertases turn into C5 convertases, which cleave C5 into the anaphylatoxin C5a and the larger fragment C5b. C5b is the nucleus for the formation of the membrane attack complex, which is assembled via subsequent binding of C6, C7, C8 and C9. C3a and C5a induce chemotaxis and inflammatory responses, while MAC forms pores in cell membranes, leading to cell lysis. An additional effector function is the induction of phagocytosis via the degraded opsonins iC3b and C3dg.

Complement as an emerging therapeutic target

The flip side of this instant, forceful and broadly applicable defense machinery is that it may cause considerable damage when directed against host tissue or biomedical surfaces. Although our own cells are protected by complement regulators, the complement system may act as initiator, contributor or exacerbator in various disorders [12]. Excessive and/or misguided activation can be induced by a sudden occurrence of danger signals (e.g., sepsis, trauma), the exposure to foreign surfaces (e.g., transplants, biomaterial) or cellular debris (e.g., atherosclerosis), or the erroneous recognition of host cells by autoantibodies. At the same time, diminished regulatory capacities may render host tissues susceptible to complement-mediated damage. For example, the lack of complement regulators on clonal populations of erythrocytes in paroxysmal nocturnal hemoglobinuria (PNH) patients results in intravascular hemolysis [12]. Similarly, the reduced capacity of endothelial cells to tame a complement attack is the driving force behind the renal disorder atypical hemolytic uremic syndrome (aHUS) [12].

Owing to its disease involvement, it is not surprising that the complement system has gained interest as a therapeutic target [12–15]. Pharmacological intervention in any host defense pathway is understandably regarded with considerable skepticism as it may weaken a patient’s capacity to adequately respond to exogenous or endogenous threats. However, in many complementopathies with severe clinical consequences, stopping the damaging effect of complement will be more important than losing an extra layer of host defense, especially since the impact of innate immunity becomes less critical with the development of adaptive immunity.

A first pathway-specific drug was introduced in 2007 with the approval of the anti-C5 antibody eculizumab (Soliris, Alexion) for the treatment of PNH. The clinical and commercial success of eculizumab has profoundly changed the dynamics in complement drug discovery as it validated the feasibility and safety of complement-targeted interventions, with both small biotech and large pharmaceutical companies featuring complement therapeutics in their pipelines. Yet the therapeutics arsenal has remained limited both in numbers and addressed targets.

At the same time, the clear clinical outcome in PNH, an ultra-rare disease with major involvement of a single pathway (AP) and effector (MAC), may also have raised unrealistic expectations for other indications. In most disorders, complement involvement is far more complex with several pathways, effectors and crosstalk routes being at play simultaneously. In such cases, a broader inhibition that acts upstream in the cascade will be more beneficial to prevent complement-induced damage. Even in the case of PNH, ongoing opsonization of erythrocytes may foster extravascular hemolysis and breakthrough events that leave some patients transfusion-dependent [16]. Recent years have therefore seen a broadening of therapeutic targets beyond C5 [17]. Among those, the central protein C3 is of particular interest, since inhibiting C3 activation has a direct and largely initiation-independent impact on opsonization, amplification and effector generation, including C5a and MAC [18].

So far, though, derivatives of the peptidic C3 inhibitor compstatin have remained the only candidates with direct effect on C3 [19]. Similar to the situation prior to the approval of eculizumab, the clinical use of C3 inhibitors was eyed with skepticism due to safety concerns. The recent approval of pegcetacoplan therefore marks a watershed moment in complement therapy and sets milestones as the first-in-class C3 inhibitor, the first phage display-derived macrocycle and the second complement-specific drug class in the clinic. Most importantly, the long-awaited clinical validation of C3-targeted therapies in general, and the compstatin family in particular, will pave the way for an extended use of a promising drug class that was enabled by strong academic-clinical-private partnerships and has meanwhile produced a panel of preclinical and clinical candidates with distinct and enhanced therapeutic profiles.

Design and evolution of the compstatin family

The starting point for the development of the compstatin family had been set in 1996, when a random 27-mer peptide phage display library was screened against C3b. Thereby, a disulfide-containing peptide sequence was identified, which showed affinity to C3, C3b and C3c and efficacy in inhibiting complement activation. The functional region of the peptide has been located to the first 13 amino acids (compstatin, I[CVVQDWGHHRC]T, Figure 2) and the disulfide bridge conferring it to a peptide macrocycle was shown to be essential for its activity, with IC₅₀ values of 63 and 12 μM for the CP and AP, respectively [20]. The low metabolic stability of the free N-terminus in blood was mitigated by acetylation, simultaneously enhancing target affinity [21]. To identify the key amino acids involved in target binding, several systematic approaches have been employed, including alanine [22], D-amino acid [23] and N-methylation [24] scannings, deletion analogs [21] and retro-inverso mimetics [21]. Early studies identified V4 and the stretch of Q6-G9 to be largely responsible for target affinity [22] and suggested improvements at position V5 and H10, resulting in Cp01 (Ac-I[CVWQDWGAHRC]T) as the most active derivative solely composed of natural amino acids [22,23].

Figure 2: Evolution of compstatin development.

a) Peptide-sequences and binding affinity of compstatin family members to their target C3b, measured in an SPR assay. Sequence-differences are highlighted in red and green, respectively. b) Chemical structures of compstatin derivatives underlining major developmental steps.

Subsequent optimization steps were greatly facilitated by the availability of structural information, the assessment with sensitive binding assays and the use of non-proteinogenic amino acids. The solution structure of free compstatin indicated a type I beta-turn of the segment Q6-D7-W8-G9, which is stabilized by the disulfide bridge [22,25]. Intriguingly, the first crystal structure of a compstatin analog with its target (i.e., Cp01-C3c) showed that it undergoes a large conformational change upon target binding [26]. Furthermore, the crystal structure revealed that compstatin binds between domains MG4 and MG5 of the β-chain of C3c, without inducing conformational changes of C3c upon binding.

The introduction of non-proteinogenic amino acids further improved the affinity of compstatin derivatives, with N-methylated Trp showing a particularly strong effect [27]. The marked improvement in the new lead compound Cp05 (Ac-I[CV(1Me)WQDWGAHRC]T, Figure 2) with a particular impact on kinetic dissociation rates and binding entropy indicated an effect of additional hydrophobic contacts or replacement of water [28]. N-methylation scanning identified two positions (G9, I14) that benefited from methylation, leading to the first single-digit nanomolar compstatin derivatives Cp10 and Cp20 (Figure 2) [24]. The impact of N-methylated Gly (i.e. sarcosine, Sar) was unexpected since larger side chains had not been tolerated at this position [29]. It was subsequently shown that Sar increased the rigidity of the peptide backbone in solution, thereby reducing entropic penalty [24,30]. An extension of the N-terminus to address an additional pocket resulted in analog Cp40 (y-I[CV(1Me)WQDWSarAHRC]mI, Figure 2) that featured subnanomolar affinity [30].

The profound achievements in optimizing the efficacy of the compstatin family has recently been rationalized by extensive structure-activity-relationship (SAR) studies based on the co-crystal structure of Cp40 in complex with C3b, combined with molecular dynamics simulations and binding studies of compstatin derivatives [31]. They confirmed the presence of an extended binding site for D-Tyr1 and, more intriguingly, demonstrated that the profound impact of indole-methylation in (1Me)W5 is linked to the shielding of structural water in the binding pocket on C3b. Of note, the SAR studies also revealed that some of the affinity improvements are achieved by intramolecular stabilization rather than target contacts and that the charged residues (D7, R12) not only confer solubility but substantially contribute to target binding. This insight is expected to guide the future development of the compstatin family.

In addition, the study provided a rationale for the narrow species specificity of compstatin for human and non-human primate (NHP) C3, which was shown in early activity studies [29]. Homology modeling of mouse C3b suggested that compstatin may still fit in the binding pocket but cannot engage in the same beneficial contacts as in the case of human/NHP C3b. The study also strengthened our understanding of compstatin’s mode of action. The binding site of compstatin at the MG4/MG5 region of C3 and C3b has earlier been suggested to mediate the binding of the C3 substrate to the C3bBb convertase complex at the cell surface (Figure 3). Indeed, the study showed that Cp40 acts as protein-protein interaction inhibitor by preventing the binding of C3 to the convertase, thereby impairing C3 activation and C3b deposition. Intriguingly, a direct comparison between Cp40 (developed as AMY-101 by Amyndas) and a surrogate of pegcetacoplan suggests that the two compounds may display distinct activity profiles despite the shared targets and mode of action. The monovalent, non-PEGylated Cp40 showed similar and strong binding to soluble and surface-bound C3 targets. In contrast, the pegcetacoplan surrogate, in which two Cp05 moieties are bridged by a 40-kDa PEG, shows generally weaker affinities yet also a notable preference for surface-immobilized targets [31]. This may likely be explained by an avidity effect of the bivalent compstatin derivative binding to two C3b molecules. Whether and how these target binding difference translate into clinical effects remains to be determined.

Figure 3: Structure and mechanism of action of compstatin.

a) Schematic representation of the molecular mechanism of the amplification loop with generation of the AP C3 convertase (C3bBb) by C3b, FB and FD. Native C3 binds to the convertase and gets cleaved to C3a and C3b, with the latter being able to form new C3 convertases, thereby amplifying the complement response. b) By binding to a dimerization site in C3/C3b, compstatins impair the binding of the C3 substrate to the convertase and prevent its activation. The insert shows the co-crystal structure of C3c (α and β-chains in green and blue, respectively) with Cp40 (orange). The molecular basis and key structural determinants of the compstatin-C3 interaction are discussed in greater detail in the text.

Preclinical and clinical development of compstatin family

While peptide macrocycles typically confer benefits regarding target selectivity and tissue penetration, their pharmacokinetic properties are often impacted by rapid elimination and, potentially, metabolism. Although the disulfide bridge may be considered a metabolic liability, cyclic compstatin analogs featured excellent plasma stabilities and a replacement of the disulfide by a reduction-stable thioether bridge did not result in an advantage [30,32]. Due to their small size (<2 kDa), unbound compstatin analogs may be readily excreted by renal filtration. To increase plasma residence, analog Cp20 had been coupled with albumin-binding tags to exploit the high abundance and long circulation half-life of albumin (ABM2-Cp20, Figure 2). This modification indeed increased plasma protein binding, albeit at the cost of solubility, and unexpectedly improved the binding affinity to C3 by 20-fold [33]. Importantly, these studies also indicated that plasma residence is strongly driven by the tight binding of compstatin derivatives to C3, a highly abundant target (~1 mg/ml), which protects the peptide from excretion and metabolism. This target-driven elimination kinetics was supported by studies in NHP, which showed a correlation between binding affinity and half-life [34]. Although the addition of a 40-kDa PEG moiety to Cp40, a common strategy to enhance the half-life of peptides, improved plasma residence substantially, it also led to an increase of plasma C3 levels. Of note, an increase of C3 levels was also reported in all PNH patients treated with pegcetacoplan, which also contains a PEG-40k moiety, during Phase Ib trials [35]. As the levels stabilized after the first month of treatment, albeit at rather high values (300–500 mg/dL), the effect was attributed to successful C3 inhibition. In the case of Cp40/AMY-101, the peptide itself already shows highly beneficial PK properties, providing advantages concerning cost, administration, and tissue penetration. The absence of PEGylation may also confer potential benefits in patients who develop anti-PEG antibodies, which have been shown to affect the pharmacokinetic, immunogenicity and safety profiles of PEGylated biotherapeutics (e.g., through accelerated blood clearance and complement activation-related pseudo allergies), which might compromise efficacy [36]. Although no such adverse reactions have been reported for pegcetacoplan, the possibility that the high plasma concentrations of PEGylated compstatins or their accumulation on C3b-opsonized surfaces may negatively impact immune tolerance in PEG-sensitized patients cannot be ruled out and should be addressed in future studies. Recent studies showed that smaller modifications to the active peptide may have beneficial effect on both solubility and PK properties, as the attachment of charged amino acids (i.e., Lys, Cp40-KKK, Figure 2) or mini-PEG led to peptides with maintained or even increased target affinity, higher solubility at physiological pH (>200-fold compared to Cp40) and prolonged plasma half-life [34].

Pharmacodynamic aspects in the preclinical development of compstatin analogs

Early on during the drug discovery effort of the compstatin family, proof of concept (PoC) was established in a wide spectrum of preclinical disease models linked to insufficiently regulated or excessive C3 activation. The narrow species specificity of compstatins essentially accelerated their preclinical evaluation in NHP models. C3 inhibition by compstatin analogs elicited potent systemic anti-inflammatory and organ-protective effects in NHP models of extracorporeal circuit-induced complement activation, xenotransplantation, hemodialysis-induced inflammation, bacterial sepsis, periodontal inflammation and polytrauma-induced multi-organ failure (these preclinical studies are extensively reviewed in [12,17,37,38]).

The clinical feasibility of C3 inhibition in PNH and the broader therapeutic effect of compstatin-based C3 inhibitors on both intravascular and extravascular hemolysis was first provided in 2014 [39]. Cp40 prevented both intravascular hemolysis and C3b opsonization, as a surrogate marker of extravascular hemolysis, in an ex vivo PNH model. Moreover, repeated SQ dosing of Cp40 in monkeys indicated that prolonged C3 inhibition was feasible as a chronic dosing regimen, achieving both target saturation and sustained pharmacological activity. Recent efforts have uncovered new pathophysiological attributes of C3 activation, including organ transplantation across HLA incompatibility barriers. Cp40 was shown to prevent acute antibody-mediated rejection and to prolong allograft survival in a sensitized NHP model of kidney transplantation [40]. Cp40 treatment not only blunted C3 deposition on the graft but also attenuated B and T cell activation and proliferation, indicating a broader immunomodulatory impact of C3 inhibition, which may have important implications for many complement-mediated diseases of the (auto-) immune spectrum.

Clinical development of compstatin-based therapeutics

Various compstatin-based candidates are being developed clinically for both systemic and local inflammatory indications requiring acute/transient or chronic intervention (reviewed in [17], Figure 4). One of the first indications to be prioritized in the clinical path was age-related macular degeneration (AMD), a prevalent ocular inflammatory disease linked to AP dysregulation [41,42]. In 2006 the second-generation compstatin analog Cp05 was licensed by the University of Pennsylvania to Potentia for clinical development in both dry and neovascular (wet) AMD. Potentia entered a partnership with Alcon to bring the drug candidate (POT-4/APL-1) to Phase II trials in patients with wet AMD. Despite initial setbacks from a lack of efficacy, likely attributable to insufficient dosing, Apellis assumed the development program for ocular indications but selected APL-2/pegcetacoplan, a PEGylated, bivalent version of Cp05 with improved plasma residence, as clinical candidate. APL-2 was successfully evaluated in a Phase II trial in patients with geographic atrophy (GA, an advanced form of AMD) showing a significant reduction of GA lesion growth by 29% and 20% in monthly and bimonthly dosing regimens, respectively [43]. The presence of a 10–20% rate of exudative AMD conversions in this trial raised the possibility that subclinical choroidal neovascularization (CNV) may have gone undetected at baseline or that an increased PEG burden over time may have contributed to de novo CNV conversion in APL-2-treated patients [44,45]. Two multi-center Phase III trials of pegcetacoplan in GA recently reported initial results: while the primary endpoint was met in one of the trials (OAKS), by reducing the rate of GA lesion growth by 22% and 16% compared to sham-controls in a monthly and bimonthly dosing schedule, respectively, the significance mark was narrowly missed in the other Phase III trial (DERBY) [46]. Of note, both trials recorded much lower CNV conversion rates when compared to the Phase II study. While the results from these trials need to be thoroughly reviewed before broader conclusions can be drawn, this development marks an important milestone for C3-targeted therapeutics in the ocular space. Overcoming the clinical setback from the lack of response in the Phase III trials of lampalizumab (anti-FD Fab antibody) in GA [47], this marks the first time that any complement therapeutic shows clinical efficacy in a Phase III trial in patients with GA, an ocular disease with currently no approved therapy at hand. Furthermore, these results suggest new opportunities for highly effective interventions employing PEG-free, fourth generation compstatin derivatives tailored for ocular delivery. Their prolonged intraocular residence and enhanced retinal tissue penetration render such compstatin analogs attractive candidates for therapeutic C3 modulation in chronic ocular diseases such as AMD. Indeed, a recent study demonstrated that the preclinical candidate AMY-106 (Amyndas) could be detected in the eye 3 months after a single intravitreal injection [48], thereby indicating the potential of 5–6-month dosing schemes. The enhanced inhibitory potency and the lack of PEG and its associated risks (e.g., anti-PEG antibodies, tissue vacuolation, CNV conversion) may point to more efficacious and patient-compliant therapeutic options for dry AMD.

Figure 4: Timeline of the major milestones in the clinical development of compstatins.

The scheme shows key achievements and crucial inflection points in the long history of development of the compstatin class of complement inhibitors from identification of the 1^st generation, selection of lead compounds of the 2^nd generation, towards PoC in preclinical models and therapeutic efficacy in clinical trials for various indications. Cited literature in the figure corresponds to references [28, 32, 34–35, 43, 51–52] in the text.

In addition to retinal disease, the inflamed periodontium has garnered considerable attention as a therapeutic opportunity for C3 therapeutics. Defining a milieu where an intricate interplay between excessive C3 activation and oral microbial dysbiosis perpetuates a vicious cycle of non-resolving periodontal inflammation, the periodontal tissue offers a testbed for developing locally delivered C3-targeted therapies that can alleviate the inflammatory burden of periodontal disease [49]. In this regard, Cp40/AMY-101 has shown consistent therapeutic efficacy by attenuating key markers of periodontal inflammation and bone loss in NHP models of naturally-occurring or experimentally-induced periodontitis [38]. Based on these findings, a Phase IIa trial was designed to evaluate the safety and clinical efficacy of AMY-101 in patients with gingival inflammation (ClinicalTrials.gov Identifier: NCT03694444). Local delivery of AMY-101 in patients with gingivitis resulted in significant attenuation of key clinical indices of periodontal inflammation [50]. Of note, the broad and sustained anti-inflammatory effect of AMY-101 was marked by attenuated MMP expression in the gingival crevicular fluid (GCF) and persisted for 3 months after the discontinuation of therapy. These results pave the way for evaluating C3 inhibition in larger multi-center Phase III studies in patients with severe periodontal disease and indicate that AMY-101 could become a ‘first-in class’ host-modulatory therapy for treating periodontal inflammation.

The successful preclinical evaluation of C3 inhibition in PNH led to the first clinical trial of pegcetacoplan/APL-2 in PNH patients who poorly respond to anti-C5 therapy [35]. The clinical efficacy of pegcetacoplan in improving key hematologic markers of disease was consolidated in a multi-center Phase III trial that performed a head-to-head comparison with eculizumab as the standard therapy. In this trial, pegcetacoplan outperformed eculizumab in key hematological and clinical parameters, improving hemoglobin levels and significantly reducing transfusion dependence over a period of 16 weeks [51], leading to the recent approval of pegcetacoplan (Empaveli, Apellis) by the FDA. This marks a landmark decision that not only validates the compstatin technology in the clinical setting but also largely relieves concerns about the safety of prolonged C3 inhibition. Similar to eculizumab, treatment with pegcetacoplan is prescribed under a risk evaluation and mitigation strategy plan that allows for close monitoring of patients during treatment. The approval of pegcetacoplan opens up new directions for applying C3-targeted therapeutics in a wide spectrum of diseases driven by C3 dysregulation that reaches well beyond PNH [52,53].

Complement targeting in neurodegenerative diseases

Therapeutic complement modulation in acute or chronic neurodegenerative conditions has attracted considerable interest owing to the broad involvement of complement in neuroinflammatory processes that exacerbate brain pathology [3,54]. The recent approval of anti-C5 therapy in patients with refractory generalized myasthenia gravis (gMG) and neuromyelitis optica spectrum disorder (NMOSD), two neurological conditions driven by distinct sets of pathogenic autoantibodies, has been a catalyst for developing complement inhibitors in this field [55,56]. Growing evidence over the past 15 years has indicated that C3 inhibition may hold therapeutic promise in chronic neurodegeneration as an effective immunomodulatory approach that can broadly control aberrant glial responses leading to synaptic loss, reactive astrocyte-driven neurotoxicity and axonal degeneration (i.e., demyelination) [57]. Consistent with this notion, a potentially registrational Phase II trial of pegcetacoplan has recently been initiated in patients with sporadic amyotrophic lateral sclerosis (ALS), the most common form of motor neuron disease that leads to progressive muscle weakness and paralysis [58]. C3 inhibitors may offer broader therapeutic range than terminal pathway-targeting therapeutics in such neurological conditions by not only impairing effector generation (MAC, C5a) but also preventing C3b deposition that in turn promotes opsonophagocytic tagging and microglia-mediated destruction of axons, synaptic bodies and nerve cells [57]. By simultaneously preventing C3a and C5a release, C3 inhibitors can attenuate the recruitment of inflammatory cells that mediate CNS injury (i.e., neutrophils, eosinophils, monocytes) and downregulate neurotoxic astrocyte activation. The immunomodulatory impact of compstatin-based C3 therapeutics on B and T cell responses was recently demonstrated [40]. These findings may project to broader therapeutic gains of C3 inhibition in CNS pathologies with (auto-)immune etiology, likely extending to the downregulation of autoantibody production in peripheral lymphoid tissues [59]. Biomarker-guided patient stratification in future clinical trials will ultimately determine the efficacy of C3 inhibitors in neurological diseases with prominent complement involvement.

Concluding remarks and future perspectives

The compstatin family of C3 inhibitors has experienced a remarkable journey during the 25 years since the initial discovery. Starting from an academic project it has produced a clinically approved drug and several (pre)clinical candidates with enhanced therapeutic profiles. Far beyond its origins, the development of compstatin-based therapeutics serves as a formidable example of the impact of strong academic-clinical-private partnerships in advancing a therapeutic concept from bench to bedside. Much of the development efforts and preclinical evaluation studies have been supported by public funding and collaborative studies, which explains the exceptional publication record for this compound class [60]. The extensive use of compstatin analogs in various models of disease not only leveled the grounds for the clinical development but also proved invaluable for shedding light into the disease involvement of complement in general, and C3 in particular, in a wide spectrum of pathologies. The approval of pegcetacoplan for the treatment of PNH serves as critical validation of the treatment strategy yet likely only marks the beginning of C3-targeted therapies as evidenced by the promising data emerging from the ongoing clinical trials in severe COVID-19 (see textbox), periodontal disease and AMD. These current trials underscore the broad applicability of therapeutic C3 inhibition in chronic and acute, local, and systemic disorders, and the expansion into new areas such as neurological diseases may hold great promise. Regardless of compstatin’s remarkable journey and success, a number of questions are still remaining, such as the disease specific way of its delivery and its expanded use in a broader set of diseases (see Outstanding Questions).

Text box:

Compstatin-based drugs in the treatment of COVID-19

Systemic hyperinflammatory syndromes that culminate in pronounced microvascular endothelial injury, thrombosis and multi-organ failure define another space of therapeutic opportunity for C3 inhibitors. C3 activation was recently identified as a converging mechanism that potentiates aberrant thrombogenic responses in severe COVID-19 driven by the platelet, neutrophil, tissue factor/NETs axis [61]. The broad involvement of C3-mediated pathogenic processes in multiple facets of COVID-19-associated thrombo-inflammation has provided justification for evaluating C3 inhibition as a novel therapeutic strategy in patients suffering from severe complications of SARS-CoV-2 infection [62]. In this regard, the compstatin-based clinical candidate AMY-101 was shown to exert rapid and sustained anti-inflammatory effects in severe COVID-19 patients, leading to resolution of COVID-19-associated acute respiratory distress syndrome (ARDS), following i.v. dosing for 14 days under a compassionate use protocol [63]. AMY-101 attenuated NETosis in severe COVID-19 patients to a greater extent than eculizumab [64] and disrupted TF expression in neutrophils, thereby providing a mechanistic basis for a pivotal role of complement and NETs in COVID-19 immuno-thrombosis [61]. These findings led to the design of a randomized, controlled Phase II trial of AMY-101 in patients with severe COVID-19, for which interim results were recently reported [65]. Initial clinical observations and biological insight from the interim analysis point to a sustained and complete inhibition of C3 throughout the treatment regimen with broad impact on thrombo-inflammatory markers. It was also indicated that parallel activation of extrinsic protease-mediated pathways in severe COVID-19 may override the inhibitory effect of AMY-101 in a small fraction of patients that sustain greater thrombo-inflammatory activation [65]. Altogether these results support the potential of compstatin-based therapeutics as promising treatment options in diverse pathologies of the ICU spectrum involving a hyperinflammatory phenotype, such as the cytokine release syndrome triggered by CAR-T cell immunotherapy and non-COVID19-related ARDS.

Outstanding questions:

• Both systemic and local delivery of compstatins has resulted in sustained C3 inhibition and clinical efficacy in certain indications. Would alternative formulations offer also a more patient-compliant route of drug delivery and enable the expanded clinical application of compstatins in more chronic indications, including neurodegenerative diseases?

• A broader immunomodulatory effect of C3 inhibition on B and T cell activation was documented in a sensitized primate model of kidney transplantation. Could this finding signify a similar effect of compstatins in other (auto-)immune-driven clinical conditions entailing prominent B or T cell reactivity? Could C3 inhibition with compstatins beneficially modulate production of autoantibodies in certain pathologies by attenuating autoreactive B cell responses?

• Given the broad inhibitory effect of AMY-101 on COVID-19-associated thromboinflammation, would compstatins constitute a viable treatment option for curbing hyperinflammation in other pathologies that involve a maladaptive host inflammatory response and deregulated innate immune activation?

With several drug candidates targeting pathway-specific initiation (e.g., anti-C1s, anti-MASP2), amplification (e.g., FD and FB inhibitors) and effector functions (e.g., C5aR1 antagonists) in late-stage development or (pre)approval stages, we finally enter a new era in complement-modulating therapies leading to improved accessibility for hitherto neglected markets. Considering the commercial compstatin-drug development pipeline and strong academic effort in elucidating the remaining secrets of the compstatin class and translating the insight into next-generation analogs with improved properties, there is well-founded hope that the compstatin family will play a central role in this new era of complement-targeted therapies.

Highlights.

• The compstatin family of complement C3 inhibitors was identified in 1996 by phage display screening and subsequently optimized towards subnanomolar affinity.

• Compstatin analogs feature a narrow species specificity, excellent plasma stability and favorable pharmacokinetic profiles.

• Early proof-of-concept studies showed potent systemic anti-inflammatory and organ protective effects of compstatins in NHP models of polytrauma-induced hemorrhagic shock, hemodialysis-induced inflammation, xenotransplantation and HLA-incompatible kidney transplantation, among others.

• A PEGylated, second generation compstatin derivative (pegcetacoplan; Empaveli, Apellis) has been approved in 2021 for the treatment of PNH and is evaluated in further indications.

• This approval not only serves as validation of the compstatin technology in a clinical setting but also opens up new opportunities for therapeutic C3 modulation.

• Next-generation derivatives with improved target affinities and PK properties (e.g., AMY-101, Amyndas) are tested in clinical trials for periodontal disease, COVID-19 and other diseases.

• The enhanced intraocular residence and retinal distribution of the latest-generation compstatins may point to more tailored therapeutic solutions for retinal pathologies driven by C3 dysregulation such as AMD.

• Growing evidence over the past 15 years has indicated that C3 inhibition may hold therapeutic promise in chronic neurodegeneration.

Glossary

Age-related macular degeneration

prevalent ocular inflammatory disease, which leads to blurred or loss of vision in the center of the visual field and can be divided in a ‘wet’ and ‘dry’ subtype

Alanine scanning

method to systematically investigate structure-activity relationships. Several peptides are generated where each amino acid position is changed to alanine one-by-one and compared to the original peptide sequence

Amyotrophic lateral sclerosis

progressive neurodegenerative disease with motor neuron loss leading to loss of muscle control

Anaphylatoxin

small effector protein generated during complement activation, which can exert chemotactic, inflammatory, anaphylactic and/or other biological activities

Atypical hemolytic uremic syndrome (aHUS)

a rare renal disorder characterized by microangiopathic hemolytic anemia, thrombocytopenia, and renal failure. It is strongly linked to complement dysregulation mainly fueled by genetic or acquired defects of the complement alternative pathway

Complementopathies

disorders which are driven by an impaired regulation of complement

Convertase complexes

central protease complexes which are able to cleave C3, and C5 respectively. The convertase complexes are comprised of several activated complement proteins, i.e. C2b, C3b, C4b, Bb

Generalized myasthenia gravis

chronic, autoimmune neuromuscular disease with autoantibodies against nicotinic acetylcholine receptors. It leads to skeletal muscle weakness due to reduced signal transduction at the junction between nerve and muscle

Geographic atrophy (GA)

advanced stage of non-exudative age-related macular degeneration (dry AMD) leading to the irreversible loss of central vision. It is characterized by lesions formed by the deterioration of the photoreceptors, retinal pigment epithelium (RPE), and choriocapillaris (network of small vessels supporting the retina)

Host defense pathways

tightly regulated protein networks that protect the body from microbial intruders and other threats that subvert its homeostasis through danger sensing, pro-inflammatory signaling and generation of effector molecules that contribute to leukocyte activation, migration into tissues and pathogen clearance

Mannose-binding lectin

is an oligomeric protein of the C-type lectin family, which serves as pattern recognition receptor binding to carbohydrate signatures found on pathogenic microorganisms and apoptotic cells

Membrane attack complex

terminal multi-protein complex of the complement cascade, composed of the complement components C5b, C6, C7, C8 and multiple copies of C9. It can insert into the cell membrane forming a pore

Metabolic stability

stability of a therapeutic compound against degradation, oxidation and conjugation by metabolic pathways, such as proteases and liver cytochromes

Neuromyelitis optica spectrum disorder

is an ‘umbrella’ term comprising rare and heterogenous syndromes which are characterized by acute inflammation of the optic nerve (optic neuritis) and the spinal cord (myelitis). In the majority of cases this is caused by autoantibodies against aquaporin 4

N-methylation scanning

method to systematically investigate structure-activity relationships. Several peptides are generated where each amino acid position is changed one-by-one to the N-methylated version of the amino acid and compared to the original peptide sequence. It is used as a general strategy to reduce binding-related entropy-loss of peptides by providing local constraints to the peptide backbone

Opsonization

binding of an opsonin (e.g., antibody or complement protein C3b, C4b) to an epitope of a pathogen or onto cells to serve as danger signal and/or trigger phagocytosis

Paroxysmal nocturnal hemoglobinuria (PNH)

a rare acquired hematological disorder in which complement dysregulation on the surface of erythrocytes lacking the glycophosphatidylinositol-anchored complement regulatory proteins CD55 and CD59 leads to chronic intravascular hemolysis, complement-mediated anemia, and transfusion dependency

Retro-inverso mimetic

peptide derivative which consists of an inverted amino acid sequence and D-stereochemistry compared to the original peptide

Xenotransplantation

transplantation of organs or tissues from one to another species