Laboratory findings in patients treated with complement factor C3 inhibitor pegcetacoplan

Patients

All blood samples were obtained in routine diagnostics from patients who were treated with pegcetacoplan either with an approved indication or off-label, outside the context of a clinical trial. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). Informed consent was obtained from all individuals included in this study.

Complement factor C3 and C4 levels

The C3 and C4 concentrations were determined in serum by turbidimetry (Cobas 8000 platform, F. Hoffman-La Roche Ltd, Basel, Switzerland). Reference values for C3 are 700–1,500 mg/L and C4 100–400 mg/L.

Classical and alternative pathway activity assay

Classical pathway (CP) and alternative pathway (AP) complement activity analyses were performed as previously described by our group [9] and based on Fredrikson et al. [10] and Seelen et al. [11]. For the CP analysis, a Nunc MaxiSorp ELISA plate was coated with 200 µL 3.1 μg/mL rabbit anti-human IgM (Q0333; Dako Agilent, Santa Clara, CA, USA) in 0.05 m carbonate buffer (pH 9.6) and incubated overnight at 4 °C. For the AP analysis, a Nunc MaxiSorp ELISA plate was coated with 8 μg/mL Salmonella typhosa LPS (Sigma‐Aldrich, St. Louis, MO, USA) (AP) in 120 µL PBS by drying for 2 days at 37 °C. The ELISA plate was blocked for 1 h with 1 % gelatin in Tris‐buffered saline (TBS) and only in case of the CP analysis, followed by incubation with 100 µL 4.5 μg/mL human IgM (31-A135; Fitzgerald-fii, Biosynth, UK). Plates were washed 3× with wash buffer (0.05 % Tween 20 in TBS) using a microplate washer (Tecan, Durham, NC, USA). Serum was diluted several times, 50×/75×/150×/300× (CP) in dilution buffer (10 mm Tris, 84 mm NaCl, 0.5 mm MgCl₂, 2 mm CaCl₂, 0.1 % gelatin) or 5×/7.5×/10×/16×/20× (AP) in dilution buffer (10 mm Tris, 140 mm NaCl, 5 mm MgCl₂, 10 mm EGTA, 0.1 % gelatin), added to the wells in duplicate, and incubated for 60 min (CP) or 75 min (AP) at 37 °C. Plates were washed 3× with wash buffer. Mouse aE11 monoclonal anti‐C5b‐9 (M0777; Dako) was diluted 300× (AP) and 500× (CP) in wash buffer, added to the wells and incubated 1 h at room temperature while shaking at 600 rpm, followed by washing 3× with wash buffer. Alkaline phosphatase‐labelled goat anti‐mouse IgG (610-1519 Rockland Pottstown, PA USA) was diluted 3,000× (AP) and 5,000× (CP) in wash buffer, added to the wells and incubated 1 h at room temperature while shaking at 600 rpm, followed by washing 3× with wash buffer. P‐nitrophenylphosphate (1 mg/mL) in substrate buffer (1 m diethanolamine, 0.5 mm MgCl₂ pH 9.8) was added, incubated at 37 °C and the reaction was stopped after 15–20 min by adding 2 m NaOH. The absorbance was measured at 405 nm using a 96‐well plate reader (Tecan). Pooled human serum was used as control. Normal reference values for AP are 67–133 %, and normal reference values for CP are 67–149 %.

C3d and C3bBbP measurements

The level of C3d was quantified in EDTA plasma using an ELISA, as previously described [12], 13]. To determine C3d values, pooled activated EDTA plasma of healthy blood donors was used and set to 100 %. This value was calibrated to mg/L using purified C3d containing 10 μg/mL protein [14]. The reference value was determined in a group of 20 healthy volunteers (mean age ± SD; 45.8 ± 11.4). The average C3d value in this group was 4.2 mg/L (SD=1.4) and the C3d upper reference value was set at <8.3 mg/L (average plus 3×SD). As the initial C3 concentration may influence the C3d level and because of intra-individual variations in the concentration of C3, the C3d/C3 ratio is used as the parameter for complement activation. In the same cohort of 20 healthy controls the average C3 concentration was 936 mg/L (SD=181). The average C3d/C3 value 4.7 (SD=1.1) and the C3d/C3 upper reference value was set at <8.1 (average plus 3×SD). The complement activation marker C3bBbP (properdin-stabilized C3 convertase complex) was measured in EDTA-plasma by ELISA based on the protocol described previously [15]. The C3bBbP reference value applicable in diagnostics is <26 CAU/mL (average plus 2xSD of 23 healthy controls).

Mass spectrometry-based protein quantitation

EDTA plasma samples were digested using trypsin as described previously with minor modifications [16]. Samples were analyzed by LC-MS/MS in duplicate. Per measurement an equivalent of ∼150 ng total protein was loaded per manufacturer’s instructions. Peptides were loaded on an Evosep One liquid chromatography (LC) system using C18 Evotips (Evosep, Odense, Denmark). Peptide separation was performed on a 15 cm C18 column (Evosep Performance, ReproSil-Pur C18, 150 µm I.D., 1.5 µm particle size) using the Evosep 30SPD method. The eluted peptides were analyzed on a timsTOF Pro 2 mass spectrometer (MS) (Bruker Daltonics, Bremen, Germany) operated in the standard data independent acquisition Parallel Accumulation-Serial Fragmentation (dia-PASEF) mode for long gradients. Protein identification and quantification analysis was done with Bruker Proteoscape (BPS, version 2025b, Bruker Daltonics) using TIMS dia-NN [17]. An in-house generated spectral library for plasma proteins was used, and identifications were re-annotated against the canonical Uniprot human protein database (downloaded on 16-8-2024) plus sequences of known contaminants such as keratin and porcine trypsin. The spectral library was generated using LC-MS analyses of a set of 20 healthy subject plasma samples, and of available recombinant complement proteins (CFB, CFD, CFH, CFI, CFP, C2, C3, C4, C5, C6, C7, C8, C9, CFHR1,2,4,5). These samples were analyzed using an identical instrument setup and LC method, but the MS was operated in data dependent data acquisition mode (dda-PASEF). Database searches to identify peptides were performed using ProLuCID in BPS using carbamidomethylation of cysteine as a fixed modification and methionine oxidation as a variable modification. Precursor and fragment mass tolerance was set to 20 and 15 ppm, respectively. The spectral library was generated by combining search results of these dda-PASEF analyses resulting in a spectral library of 11,641 precursors. dia-NN settings for data processing were: 20 ppm precursor tolerance, 15 ppm fragment ion tolerance, carbamidomethylation as fixed modification, oxidation of Methionine as variable modification, maxLFQ intensities were calculated for relative protein quantitation. Match-between-run (MBR) was performed across all measurements to fill-in missing values with an outlier frequency of 0.2, protein grouping was performed on the protein name level, global normalization was performed. MaxLFQ intensities were used to calculate the fold change in plasma protein levels of a patient compared to the sample before treatment.

Case descriptions

To illustrate the impact of pegcetacoplan treatment on routine laboratory test results, this case-series describes five patients who initiated therapy at our academic center in 2024 or 2025. Table 1 provides an overview of the patient characteristics. Two patients were diagnosed with PNH who initially responded well to eculizumab. Both patients remained anemic despite C5 inhibition and were switched to pegcetacoplan because of persisting anemia caused by C3-mediated extravascular hemolysis.

Three patients received off label use of pegcetacoplan because of treatment resistant dense deposit disease, which is a subtype of C3G. As outlined in Table 1, prior to pegcetacoplan these patients received two to four other lines of treatment.

Laboratory findings

Patients with increased complement activation generally have a low C3 secondary to complement consumption. In response to successful treatment with complement inhibitors, the C3 levels normalize. In all patients who started pegcetacoplan, we indeed observed a fast increase in C3 levels. Strikingly, in all patients, C3 concentrations increased to values far above the reference values (Figure 1). For patients 2 and 4, frequent C3 monitoring was performed in the first months of pegcetacoplan treatment. Based on the slope of C3 levels in these two patients, we conclude that the C3 increase to levels above 3,000 mg/L is observed within 3 weeks after start of pegcetacoplan.

Figure 1:

C3 dynamics in response to pegcetacoplan treatment. Each patient is identified by a unique color. All five patients reach C3 serum concentrations far above the normal reference value (indicated by the grey-shaded area). The green arrow indicates the timepoint in which pegcetacoplan is replaced by methylprednisolone treatment.

In routine diagnostics, it is very rare to measure C3 concentrations substantially above the upper reference value of 1,500 mg/L. More specifically, there are no known complement disorders or other diseases associated with increased C3 values. We performed an in vitro pegcetacoplan spike-in experiment to investigate whether pegcetacoplan-induced C3 increase over time is caused by an actual increase in the concentration of circulating C3 or whether it is a turbidimetric artifact caused by pegcetacoplan-C3 complex formation influencing light scatter. For this, pegcetacoplan was spiked into healthy control serum in concentrations ranging from 0 to 1,000 μg/mL. This concentration-range covers the expected pegcetacoplan levels observed in patients treated with pegcetacoplan [18]. Table 2 shows that the baseline C3 concentration in this sample was 975 mg/L. The measured C3 concentration was not influenced when pegcetacoplan was spiked-in. The minor C3 variation between samples is caused by technical variation of the turbidimetric C3 assay itself. These data clearly demonstrate that pegcetacoplan binding to C3 does not influence the measured C3 concentration.

Functional complement testing showed that spiked-in pegcetacoplan completely abolished the alternative complement pathway (AP). Pegcetacoplan also had a dose-dependent effect on the classical complement pathway (CP). However, even when pegcetacoplan was titrated at 1,000 μg/mL, which is above its physiological concentration [18], the CP is still close to normal. These in vitro experiments are in line with literature regarding the impact of pegcetacoplan on functional complement tests in preclinical studies [8].

Interestingly, the strongly increased C3 values in all patients give rise to an additional band when serum protein electrophoresis was performed (Figure 2A). This unique band must not be misinterpreted as an M-protein. Immunofixation electrophoresis (IFE) in fact demonstrated that the band represented C3 protein (Figure 2B). It is important to note that the C3 band in all five patients migrates in the γ-region because it is bound to pegcetacoplan which affects electrophoretic migration. In individuals not exposed to pegcetacoplan, C3 migrates in the β-region (Figure 2C).

Figure 2:

Electrophoresis data of patients treated with pegcetacoplan. (A) After pegcetacoplan exposure, serum protein electrophoresis shows a distinctive new band in the γ-region in all five patients (illustrated with red asterix). (B) Characterization with immunofixation electrophoresis (IFE) demonstrates that this new band it is not an M-protein. Anti-C3 reagents proof that the band is composed of C3 (IFE of one representative patient is shown). (C) In healthy controls not exposed to pegcetacoplan, C3 migrates in the β-region. The C3 bound to pegcetacoplan affects its electrophoretic migration. (D) IFE of pegcetacoplan spiked in serum further illustrates the shift in migration. The concentrations indicate concentration of pegcetacoplan in each analysis. G, IgG; A, IgA; M, IgM; κ, kappa; λ, lambda; TP, total protein.

The shift in electrophoretic migration of the C3-pegcetacoplan also allows this technique to study the binding kinetics of pegcetacoplan to C3. When pegcetacoplan is spiked-in at a concentration of 200 μg/mL in a healthy control serum, almost all C3 is shifted below the red dotted line, which indicates the migration pattern of unbound circulating C3 (Figure 2D). At a pegcetacoplan concentration of 400 μg/mL, all C3 is bound. This complete binding coincides with a sharp drop in the alternative pathway (AP = <20 %) as shown in Table 2. Remarkably, even when all C3 is bound to pegcetacoplan, the classical pathway is still within normal range (CP=77 %), which suggests that the alternative complement pathway is dependent on unbound C3 while the classical complement pathway is still largely functional with pegcetacoplan bound to C3.

The data of the in vitro experiments mentioned above are in line with the complement pathway test-results obtained during routine diagnostics in the five patients described in this case-series. Pegcetacoplan mainly impacts the alternative pathway and consequently a complete absence of AP activity was observed in most patients (Table 3). The impact on the classical pathway was minimal, with CP activity values that were either normal or slightly decreased. Pegcetacoplan further caused a strong inhibition of complement activation which was measured with the complement activation marker C3d. The C3d/C3-ratio is widely used as a more precise indicator of complement activation, because it accounts for interindividual variability in C3 levels. Pegcetacoplan-induced decline of the C3d/C3-ratio is extremely pronounced and is evidently strongly influenced by the high C3 titers.

How to use C3 as biomarker of disease activity in patients treated with pegcetacoplan

Complement C3 levels can be a useful biomarker for monitoring disease relapse. Reduced C3 levels are a surrogate marker for complement activity and can precede disease flare-ups [19]. The clinical value of monitoring C3 in patients treated with pegcetacoplan is more complex because all these patients have high C3 values, the absolute C3 concentration will not easily fall below the C3 reference values. However, an absolute decline of C3 should alert the clinician to consider that the underlying cause could possibly be a disease flare-up or poor medication adherence. In our case-series, patient two experienced a sharp decrease in C3. This decrease in C3 coincided with increasing creatinine levels in a period in which the patient also experienced a viral infection. It turned out that the strong C3 decline was caused by treatment nonadherence of the patient, followed by a rapid increase in C3 while on hospital controlled pegcetacoplan injections. Compared to the kidney biopsy that was taken before the start of pegcetacoplan, a reduced C3 staining intensity (from 3+ to 1–2+) and only minimal endocapillary proliferation was observed. Chronic injury was clearly increased (from 10 % to 30 % sclerosed glomeruli) with 65 % interstitial fibrosis. Electron microscopy showed an unchanged massive amount of dense deposits. Overall, these paghology-findings suggested no signs of C3G disease relapse but did demonstrate chronic renal injury. Pegcetacoplan was discontinued due to end stage kidney failure. It is interesting to look at the C3 kinetics in this patient after pegcetacoplan was stopped. Instead of reaching steady state low C3-levels, the C3 first peaked to levels again far above reference values which can be explained by the long pegcetacoplan half-life of 8.6 days [18]. After pegcetacoplan clearance, a rapid decrease of C3 levels was observed. In patient two, C3 levels decreased to reference value only two months after pegcetacoplan was stopped (Figure 1).

Broad overview of changes in the complement system during pegcetacoplan exposure

To generate an unbiased overview of the effect of pegcetacoplan on other components of the complement system we performed a plasma proteome analysis. Filtering for complement proteins that were detected in all samples as the sole protein in a protein group, allowed us to investigate the dynamics of 36 complement proteins in the blood of three patients before and during exposure to pegcetacoplan. Of all 36 complement proteins, the most affected complement factor is C3, which increased on average more than ten-fold upon pegcetacoplan exposure (Table 4). Strikingly, we also observed a strong increase in circulating properdin. The three-fold increase in properdin is a surprising result since pegcetacoplan does not directly bind properdin. Three months after pegcetacoplan was stopped in patient 2, both C3 and properdin levels normalized to levels observed prior to pegcetacoplan exposure (data not shown). Mass spectrometric proteomics data cannot differentiate between free properdin and properdin as part of the C3bBbP complex. We therefore performed additional C3bBbP analyses and found that C3bBbP increases on average three-fold after patients start pegcetacoplan treatment (Table 3). We did not observe strong changes in any of the other complement factors in the plasma proteome analysis (Table 4).