Session 13 - Preclinical complement therapeutics

Chairs: Daniel Ricklin & Kate Smith- Jackson

Background: ADX-097 is an antibody fusion protein that targets the alternative pathway (AP) complement negative regulator factor H to diseased tissue. ADX-097 was designed to bind tissue-fixed C3d-containing C3 activation fragments and inhibit complement in diseased tissue while minimizing systemic blockade.

Methods: Having demonstrated C3 activation and C3d deposition in human renal and skin disease biopsies, we generated and characterized ADX-097, a humanized anti-C3d monoclonal antibody linked to five N-terminal repeats of the AP inhibitor factor H (fH1-5). We further evaluated circulating and tissue PK/PD of ADX-097 in mouse and non-human primate (NHP) models of tissue complement activation. Finally, we tested ADX-097 PK/PD and efficacy in the rat Passive Heymann Nephritis (PHN) model of membranous nephropathy.

Results: We characterized in vitro binding of ADX-097 to C3d, demonstrating that at high C3d densities, use of a bivalent anti-C3d antibody yields binding affinities that are ~1000x greater than those of comparable monovalent binders, suggesting a substantial avidity-based targeting advantage. Consistent with this highly efficient targeting, in rodent and NHP models of local complement activation, ADX-097 dosed subcutaneously (SC) as low as 1 mg/kg distributes to C3d+ tissue and inhibits complement activation. In factor H-/- mice, a single 1 mg/kg SC dose achieves >75% glomerular complement C3 inhibition for at least 7 days while avoiding systemic complement blockade. In a rat model, 1 mg/kg dosed SC or IV inhibited glomerular complement activation and significantly reduced urine protein-creatinine ratios without inhibition of systemic complement, and with an efficacy that was equivalent to full systemic complement blockade using daily injections of cobra venom factor. We also show that urinary C5b-9 (uC5b-9) correlates strongly with decreased glomerular C3 activation and is independent of systemic blockade or proteinuria-associated renal leakage, suggesting the utility of uC5b-9 as a biomarker for local complement regulation in the kidney.

Conclusions: These data demonstrate the therapeutic potential of ADX-097 and show that C3d-mediated tissue targeting of fH1-5 in preclinical models results in potent, durable, and efficacious local AP complement blockade without systemic complement inhibition, and that this is an effective strategy for re-regulating tissue complement in disease.

Stefan Wawersik

Background: Epidermolysis Bullosa Acquisita (EBA) is an autoimmune skin blistering disease characterized by strong production of autoantibodies (AAbs) against type VII collagen (COL7) at the dermal epidermal junction (DEJ). Local complement activation drives C5a generation associated with neutrophil recruitment and activation resulting in skin lesions and inflammation. The clinical effector phase of EBA can be mimicked by repeated injection of rabbit COL7-specific IgG into mice. Here we tested the impact of prophylactic or therapeutic C5a or combined C5/C3 targeting on disease development and skin inflammation in this model.

Methods: C57BL/6 mice were immunized s.c. with purified rabbit anti-mouse-COL7 IgG. An anti-mouse C5a/C5adesArg IgG1 mAb (M031), a fusion protein between an anti-mouse C5 IgG1 mAb and FH-SCR1-5 (M014), or an IgG1 isotype control (n=12/group) were injected (i.p.) prophylactically on days -1, 2, 5 and 8 or therapeutically on days 5 and 8. Then, formation of skin lesions was evaluated every other day. DEJ separation was assessed in cryosections from day 12 and stained with H&E. Tissue-bound IgGs AAbs, C3/C3b deposition as well as MPO+ or Ly6G+ neutrophil infiltration was identified by direct immunofluorescence.

Results: Isotype-treated mice developed first skin lesions at day 4 peaking at day 12. Prophylactic treatment with either M031 or M014 markedly reduced the development of skin lesions. Also, dermal/epidermal separation as well as neutrophil infiltration on day 12 were significantly lower upon M031 or M014 treatment. Surprisingly, C5/C3 inhibition by M014 but not C5a inhibition by M031 markedly reduced the development of skin lesions and the dermal/epidermal separation after therapeutic treatment as well as the infiltration with MPOhi neutrophils. IgG and C3/C3b deposition did not differ in the three treatment groups.

Conclusion: Prophylactic C5/C3 and C5a targeting is equally effective in reducing the development of skin lesions and tissue inflammation in a preclinical model of EBA. In contrast, only combined C5/C3 but not C5a inhibition protects from the development of skin lesions and skin inflammation after therapeutic administration at day 5 when first skin lesions become evident. Our findings point toward an important role of C5a as well as other complement effectors during disease development in EBA.

Samyr Kenno

Background: Ischaemia reperfusion injury is an inevitable consequence of transplant with the complement system, in particular the alternative pathway being one of the key drivers of damage. The main regulator of the alternative pathway is factor H. We hypothesised that homodimeric mini-factor H (HDM-FH; PMID:29588430) would protect kidneys when administered during normothermic machine perfusion (NMP) with whole blood which mirrors the clinical setting of transplantation.

Methods: A model of porcine whole blood NMP with extended warm and cold ischemic times was optimised to induce complement activation and ischaemic injury in kidneys so that the full efficacy of HDM-FH could be assessed. Utilising this model, both kidneys were retrieved from female white landrace pigs. One kidney from each pair was randomised to receive 5mg of HDM-FH (~8μg/mL). Kidneys were perfused at 37°C with autologous blood for 6 hours. HDM-FH binding within kidneys was confirmed using immunofluorescence. HDM-FH levels in perfusate and urine were measured using ELISA. Complement activation was measured by quantifying Bb deposition in tissue and C5a levels in urine. Fibrosis, inflammatory cytokines, and apoptosis were measured as indicators of downstream ischaemic injury. ‘Cold binding’ of HDM-FH was assessed by flushing kidneys with 4°C saline containing HDM-FH.

Results: 25 minutes warm ischaemic time followed by 16hrs cold ischaemic time led to an increase in complement activation and markers of ischaemic injury. ~4mgs of HDM-FH bound from perfusate during perfusion at 37°C, with <10% lost in urine suggesting saturation was achieved. HDM-FH localised to glomeruli with deposition increasing during the perfusion. Complement activation was reduced in kidneys receiving HDM-FH as demonstrated by reduced Bb deposition in tissue and reduced C5a levels in urine. Fibrinogen deposition, inflammatory cytokine and apoptosis levels were reduced. HDM-FH inhibits complement activity in serum in a dose-dependent manner. HDM-FH also binds at 4°C, experiments are ongoing using this approach.

Conclusion: Infusion with HDM-FH during simulated kidney transplant conditions reduced complement activation and downstream ischaemic injury to the organ. Therefore, organ perfusion with HDM-FH is highly likely to help prolong graft survival after transplant and this will be assessed in future studies.

Chloe Connelly

BackgroundIptacopan is a selective low-molecular-weight inhibitor of the alternative complement pathway (AP) that is currently in clinical development for multiple complement-mediated diseases, including aHUS. It inhibits the protease activity of C3bBb by binding to its active site within Bb.
Approximately 3-5% of aHUS patients express one of 18 point-mutations in FB. The aim of this study was to determine if Iptacopan inhibits protease activity of these FB variants.

MethodsEight FB variants (K533R, E566A, D279G, F286L, K323Q, K323E, K350N and M458I) were selected based on mutations in proximity to the active site of Bb and their capacity to stimulate the AP. We tested activity of iptacopan on these mutants as follows:

  1. Binding affinity to FB mutants via displacement of an active site binding probe.
  2. Ability to inhibit an activated C3 convertase formed by cobra venom factor (CVF) bound to wtFB or FB mutants.
  3. Activity to inhibit lysis of sheep red blood cells (SRBCs) coated with C3 convertases containing wtFB or FB mutants.
  4. Activity to inhibit zymosan-induced AP activation in Factor B depleted human serum supplemented with wtFB or FB mutants.


ResultsIptacopan bound to all FB mutants and wtFB with comparable affinity. CVF-Bb C3 convertase activity was increased with two of the mutants (FB(K350N) and FB(D279G)) and reduced in FB(F286L) compared to CVF-Bb containing wtFB, but iptacopan inhibited C3 cleavage by all convertases with similar IC50. Cell surface C3/C5 convertase activity in the hemolytic assay was also increased for FB(K350N) and FB(D279G) and decreased for FB(F286L). Iptacopan inhibited wtFB and all FB variants in this assay, albeit less potently for FB(D279G). While serum containing FB(350N) and FB(279G) caused only limited MAC-formation upon zymosan stimulation, Iptacopan showed similar potency in the zymosan assay for wtFB and all mutants.

ConclusionThis in vitro study demonstrates that Iptacopan binds to and inhibits common FB variants identified from aHUS patients. Therefore, patients with FB variants may be included in clinical aHUS trials with iptacopan.

Anna Schubart

Morning coffee


Meeting Awards

2023 Lambris Complement Training Award

Felix Poppelaars (Introduced by Moh Daha)

International Complement Society Early Career Researcher Award

Ben King (Introduced by Anna Blom)

Session 14 - Clinical complement therapeutics

Chairs: David Kavanagh & Christine Lamers

Background: CD55 deficiency with hyperactivation of complement, angiopathic thrombosis, and protein-losing enteropathy (CHAPLE) disease is an ultra-rare autosomal recessive disorder caused by loss-of-function variants of the CD55 gene, leading to overactivation of the terminal complement system.1 Clinical and laboratory features of CHAPLE disease include hypoalbuminemia and edema; hypogammaglobulinemia; and gastrointestinal symptoms such as abdominal pain, loss of appetite, vomiting, and diarrhea. Patients also present with micronutrient deficiency, anemia, and growth retardation.1 Currently, there is no approved treatment for CHAPLE disease. We assessed the efficacy and safety of pozelimab, an investigational anti-C5 antibody, in patients with CHAPLE disease.

Methods: This is an interim report of an open-label, single-arm, historically-controlled study in patients with CHAPLE disease (ClinicalTrials.gov, NCT04209634). Patients received intravenous pozelimab 30 mg/kg followed by subcutaneous, weight-based dosing once weekly. The primary endpoint was the proportion of patients who achieved normalization of serum albumin and demonstrated improvement or no worsening in clinical outcomes (abdominal pain, bowel movement frequency, facial edema, and peripheral edema) at week 24. Secondary/exploratory efficacy endpoints and safety are also reported.

Results: Ten patients were enrolled in the study, and had ≥48 weeks of efficacy measurements as of the cut-off date for this analysis. All 10 patients (100%) experienced serum albumin normalization and improvement/no worsening in clinical outcomes. Over the 48-week treatment period, most patients experienced remarkable catch-up growth. Following treatment with pozelimab, patients had reduced all-cause hospitalization days; the mean number of hospitalization days across all 10 patients decreased from 26.8 days in the 48 weeks prior to treatment to 0 days by week 48. Four patients (40%) started the treatment period on corticosteroids and all were withdrawn as of the cut-off date for this analysis. Complete inhibition of complement activity (CH50) was achieved. Seven patients (70%) experienced adverse events; none were severe, and only one patient experienced adverse events that were considered related to study drug.

Conclusion: Pozelimab inhibits complement overactivation and resolves the clinical and pathophysiological manifestations of CHAPLE disease.

Reference: 1. Ozen A, et al. N Engl J Med. 2017;377:52–61.

Olivier A. Harari

Background: Paroxysmal nocturnal hemoglobinuria (PNH) is a rare hematologic disorder characterized by hemolysis due uncontrolled complement activation. In vitro, factor D (FD) inhibitors block C3 fragment deposition and lysis of erythrocytes collected from patients with PNH, providing evidence for its potential to inhibit both intravascular and extravascular hemolysis (IVH and EVH). Vemircopan, a second-generation oral FD inhibitor with improved potency against FD and a prolonged T1/2, is being investigated as monotherapy for patients with PNH.

Methods: This ongoing, phase 2, open-label proof-of-concept study (NCT04170023) assesses efficacy, safety, and PK/PD of vemircopan monotherapy in patients with PNH. The study comprises a 60-day screening period, 12-week treatment period, and 96-week long-term extension. Key inclusion criteria for treatment-naïve group: lactate dehydrogenase [LDH] ≥1.5 × upper limit of normal [ULN], absolute reticulocyte count ≥100×109/L, and anemia (hemoglobin [Hgb] <10.5 g/dL). Patients receive vemircopan 120 mg BID (can escalate to 180 mg BID per investigator discretion based on protocol). Primary endpoint is change from baseline to week 12 in Hgb. Safety is reported as treatment-emergent adverse events (TEAEs).

Results: At data cutoff, 11 treatment-naïve patients were enrolled (mean [SD] age, 44.4 [17.8] y; 63.6% male; 72.7% Asian, 54.5% received transfusions during screening period). Nine patients completed 12 weeks in the study and were included in this analysis. From baseline to week 12, mean (SD) Hgb increased by 3.9 (1.11) g/dL from 7.9 (1.29) g/dL to 11.8 (1.25) g/dL, LDH decreased by 81% from 7 × ULN to 1.4 × ULN, absolute reticulocyte count decreased from 212.4 (86.47) x 103/uL to 120.0 (51.44) x 103/uL. No patients needed transfusions except 1 case on day 2 owing to low Hgb (5.1 g/dL). Of 31 TEAEs (n=9 patients) most (25/31; 80.6%) were considered unrelated to study drug. No serious TEAEs, grade ≥3 TEAEs, discontinuations, or deaths were reported. There were no thrombotic events, seizures, or meningococcal infections.

Conclusions: This interim analysis of treatment-naïve patients with PNH suggests that vemircopan monotherapy controlled IVH and prevented EVH. No new safety signals were identified during the 12-week evaluation period. This analysis provides proof-of-concept for FD inhibition in PNH.

Anita Hill

Background: Paroxysmal nocturnal hemoglobinuria (PNH) is a rare, acquired disease characterized by chronic intravascular hemolysis. Pozelimab is an investigational monoclonal antibody against complement component C5. Final results from a phase 2 study (NCT03946748) and subsequent open-label extension (OLE; NCT04162470), evaluating long-term pozelimab monotherapy in patients with PNH are presented.

Methods: During the phase 2 study, patients received an intravenous loading dose of pozelimab 30 mg/kg, then weekly subcutaneous pozelimab 800 mg. All 24 patients completed phase 2 and entered the OLE, where patients received weekly subcutaneous pozelimab 800 mg up to Week 104.

Results: Overall, 87.5% of patients were Asian; 54.2% were male. In phase 2, 18/24 (75.0%) patients achieved adequate control of intravascular hemolysis (defined as LDH ≤1.5 x upper limit of normal; 95% confidence interval [CI]: 57.7-92.3%) at every scheduled timepoint between Weeks 4-26 inclusive. During the OLE, 22/23 (95.7%) patients achieved control of intravascular hemolysis at all timepoints through Week 26 (95% CI: 87.3-100.0%); with 15/16 (93.8%; 95% CI: 81.9-100.0%) at Week 78.

In phase 2, 21/24 (87.5%; 95% CI: 74.3-100.0%) patients achieved transfusion avoidance (no red blood cell transfusion). During the OLE, 22/23 (95.7%) patients achieved transfusion avoidance through the initial 26 weeks, and 15/16 (93.8%) through to Week 78. No patients experienced a breakthrough hemolysis event in either trial. Importantly, one patient with a C5 variant resistant to eculizumab/ravulizumab had adequate control of intravascular hemolysis. All patients showed a rapid decrease in total complement hemolytic activity to near zero, maintained to Week 104.

In phase 2, 21 (87.5%) patients experienced 72 treatment-emergent adverse events (TEAEs); two (8.3%) patients experienced severe TEAEs; and 10 (41.7%) patients experienced TEAEs considered related to pozelimab. During the OLE, 15 (62.5%) patients experienced 41 TEAEs; two (8.3%) patients experienced serious TEAEs; two (8.3%) patients experienced severe TEAEs; two (8.3%) patients experienced TEAEs considered related to pozelimab; and no deaths or discontinuations due to TEAEs occurred.

Conclusion: Patients with PNH receiving pozelimab (up to 130 weeks) experienced improvement in intravascular hemolysis control, with no breakthrough hemolysis, and achieved transfusion avoidance. Pozelimab was also generally well tolerated.

Morag Griffin

Background: In complement mediated atypical haemolytic uraemic syndrome (aHUS) dysregulated complement activation occurs on endothelial cell surfaces and results in thrombotic microangiopathy. Historically management comprised predominantly supportive care with or without plasma exchange, and outcomes were poor. Trials of the terminal complement inhibitor eculizumab published in 2013 suggested efficacy and revolutionized management. Given the high cost of the drug a national specialized service was commissioned by NHS England. We report the real life experience of treating individuals with suspected aHUS with eculizumab in a national cohort and compare outcomes with a control cohort.

Methods: 1956 individuals have been referred with suspected aHUS. Of these, 243 were treated with eculizumab. The control cohort comprised 279 individuals referred with suspected aHUS prior to the availability of eculizumab in whom a pathogenic complement gene mutation or autoantibody was identified. Clinical data were collected in collaboration with local clinicians and from RaDaR. Outcomes were compared using Kaplan-Meier analysis (IBM Statistical Package for Social Sciences (SPSS)). Multivariate analysis used the logistic regression model (R).

Results: The 5-year cumulative estimate (Kaplan-Meier) of end stage kidney disease (ESKD) free survival was 39.5% in the control cohort and 85.5% in the eculizumab treated cohort subgroup with a mutation or FHAA; HR 4.95 (95% CI 2.75-8.90), p=0.000, number needed to treat 2.17 (NNT) (95% CI 1.81-2.73). The magnitude of the improvement with eculizumab depends upon mutation type or autoantibody. The relapse rate upon eculizumab withdrawal was 1 per 9.5 person years for those with a pathogenic mutation. No relapses were recorded in 67.3 person years off eculizumab in those with no rare genetic variant.

Conclusion: Eculizumab costs ~£328000 per person (adult) per year and increases the risk of meningococcal sepsis. It was introduced into clinical practice on the basis of small single-arm clinical trials. In this genotype matched study of aHUS we show ESKD-free survival is significantly improved with eculizumab, and we believe that the magnitude of the improvement in outcomes justifies the high cost and potential complications of terminal complement blockade. We also demonstrate that eculizumab withdrawal in aHUS is safe and relapse is predicted by genotype.

Vicky Brocklebank

Prizes /Awards for ICW2023

Claire Harris and Kevin Marchbank

Announcement - ICW2025

Trent Woodruff

Announcement - EMCHD 2024

Jörg Köhl

Lunch - Grab Bag