Population Pharmacokinetic and Pharmacokinetic/Pharmacodynamic Analyses of Pegcetacoplan in Patients with Paroxysmal Nocturnal Hemoglobinuria Authors: Ryan L Crass, Brandon Smith, Sven Adriaens, Sunny Chapel, Grant Langdon Published: 2024 ABSTRACT AbstractBackground and ObjectiveParoxysmal nocturnal hemoglobinuria is a rare blood disorder characterized by life-threatening hemolysis and thrombosis. Complement C5 inhibitor therapy improves symptoms and life prognosis; however, it can result in insufficient hemolysis control, with residual intravascular hemolysis and extravascular hemolysis in some patients. Pegcetacoplan, the first complement C3 inhibitor approved for patients with paroxysmal nocturnal hemoglobinuria, targets both intravascular and extravascular hemolysis. This analysis evaluated population pharmacokinetic/pharmacodynamic profiles of pegcetacoplan.MethodsPooled clinical study data were used to predict pegcetacoplan concentrations and biomarker responses indicative of hemolysis (hemoglobin and lactate dehydrogenase) over time, including the impact of patient characteristics and prior or concurrent complement C5 inhibitor treatment, to support the approved dose of subcutaneous pegcetacoplan 1080 mg twice weekly.ResultsThe population pharmacokinetoc analysis included 284 subjects, and the pharmacokinetic/pharmacodynamic analysis included 165 subjects. Subcutaneous pegcetacoplan 1080 mg twice weekly resulted in rapid serum exposures and robust biomarker response within 4 weeks after treatment initiation. Steady-state serum concentrations demonstrated consistent exposure (median ≥ 600 µg/mL) with minimal peak-to-trough variation. The median effective half-life was 8.6 days in patients with paroxysmal nocturnal hemoglobinuria. Body weight significantly impacted pegcetacoplan exposure, and other covariates impacted hemoglobin (sex and creatinine clearance) or lactate dehydrogenase (prior or concurrent complement C5 inhibitor treatment); however, effects were not clinically meaningful.ConclusionsThe approved dose of pegcetacoplan is predicted to produce rapid and sustained exposure and robust hemoglobin and lactate dehydrogenase responses in adult patients with paroxysmal nocturnal hemoglobinuria, with no initial dose adjustments required for any specific patient population.Supplementary InformationThe online version contains supplementary material available at 10.1007/s40268-024-00500-7. Key Points As paroxysmal nocturnal hemoglobinuria is rare, this model predicted how subcutaneous pegcetacoplan (1080 mg twice weekly) would affect a large group of people with varying characteristics.Pegcetacoplan remained in the bloodstream at consistent levels required for clinical impact (n = 284).Pegcetacoplan resulted in biomarker changes denoting clinical improvement (hemoglobin/lactate dehydrogenase) in patients with paroxysmal nocturnal hemoglobinuria, with no dose adjustments needed for weight or prior treatment (n = 165). Introduction Paroxysmal nocturnal hemoglobinuria (PNH) is a rare, acquired, clonal, non-malignant hematologic disease characterized by complement-mediated hemolysis (with or without hemoglobinuria), fatigue, increased susceptibility to thrombosis, and some degree of bone marrow dysfunction [1]. There are an estimated 1–1.5 PNH cases per million individuals worldwide, though incidence is not clearly established and may vary depending on the region (i.e., greater frequency in countries in Asia vs the USA, Spain, and the UK) [1]. The cellular abnormality in this life-threatening disease stems from a mutation in the phosphatidylinositol glycan class A (PIG-A) gene, resulting in deficiency of glycosylphosphatidylinositol-anchored complement regulatory proteins, including CD55 and CD59, on the surface of red blood cells [2]. Complement dysregulation and increased complement sensitivity of PNH cells can lead to chronic hemolysis and thrombosis that can result in target organ damage and death [2, 3]. In addition, potential symptoms linked to hemolysis in patients with PNH can include anemia, fatigue, dyspnea, abdominal pain, erectile dysfunction, and dysphagia, which can impair overall quality of life [4, 5]. The standard of care with complement C5 inhibitors, such as eculizumab and ravulizumab, has offered improvement in symptom management and overall survival by controlling intravascular hemolysis and reducing thrombosis in patients with PNH; however, several challenges in symptom management have persisted. Eculizumab, with a half-life of approximately 11 days, requires indefinite intravenous (IV) dosing every 2 weeks, which can be burdensome for patients [6]. Likewise, approximately 11–27% of patients may experience breakthrough hemolysis with approved eculizumab dosing [7–9]. Further, despite eculizumab treatment, up to 72% of patients continue to experience residual anemia, and up to 36% of patients require more than one red blood cell transfusion per year [10]. Ravulizumab and crovalimab, both monoclonal antibodies targeting C5, are intended to provide the same benefits as eculizumab but with a more convenient dosing schedule (i.e., a longer IV dosing interval or subcutaneous [SC] dosing) [11–13]. Pegcetacoplan is a PEGylated peptide that provides a novel approach for preventing complement-mediated hemolysis by targeting proximal complement C3, which is upstream of C5 [3, 14]. Pegcetacoplan was designed for optimal complement inhibition via two C3 binding peptide domains linked by PEG in a single pegcetacoplan molecule, allowing each pegcetacoplan molecule to inhibit two molecules of C3 [15]. The PEGylated feature of pegcetacoplan is intended to allow for two C3 binding sites, but also optimize C3 binding affinity, prolonging the biological half-life of the molecule, extending its duration of activity, and producing consistent effects in hemolysis prevention [16]. Further, pegcetacoplan activity on the complement pathway allows for inhibition of C3 and therefore, prevention of downstream C5 activation. This in turn targets both extravascular and intravascular hemolysis, which may provide more complete proximal and terminal control of complement-mediated immune responses in patients with PNH [17]. Pegcetacoplan is the first C3-targeted therapeutic approved by the US Food and Drug Administration for the treatment of adult patients with PNH (May 2021) and by the European Medicines Agency for the treatment of adult patients with PNH who have hemolytic anemia (October 2024) [18, 19]. Phase I and II trials provided evidence for improved hemoglobin (Hb), reticulocyte count, bilirubin, and Functional Assessment of Chronic Illness Therapy-Fatigue scores for up to 1–2 years with pegcetacoplan treatment in patients with PNH (PADDOCK, NCT02588833; PALOMINO, NCT03593200; PHAROAH, NCT02264639) [17, 20, 21]. Phase III trials demonstrated that pegcetacoplan was superior to both eculizumab and supportive care in the improvement of hematological outcomes and was well tolerated (PEGASUS, NCT03500549; PRINCE, NCT04085601) [2, 22, 23]. The interim analysis of Study 307, an open-label extension study of patients previously enrolled in pegcetacoplan clinical trials (NCT03531255), demonstrated durability of improved hematological outcomes and a favorable safety profile with pegcetacoplan in patients with PNH over 48 weeks [24]. Given that PNH is a rare disease, a pooled assessment of the pharmacokinetic (PK) profile of pegcetacoplan in patients with PNH across studies is useful to provide a robust estimate of exposure in this population, including potential intrinsic and extrinsic patient factors that may account for inter-patient variability in exposure. In addition, characterizing the relationship between pegcetacoplan serum concentration and markers of disease (i.e., Hb and lactate dehydrogenase [LDH]) may provide insights on the dose–exposure–response relationship in this population. A pooled model-based assessment facilitates estimation of pegcetacoplan pharmacokinetics and identification of factors that may impact pegcetacoplan pharmacokinetics in a wider range of patients than can be determined from a stand-alone clinical study. Population PK (pop PK) and PK/pharmacodynamic (PD) analyses were conducted to predict estimated pegcetacoplan concentration–time profiles in a broad range of patients in real-world clinical practice settings to support the approved dose of SC pegcetacoplan 1080 mg twice weekly. The pop PK analysis evaluated the pegcetacoplan concentration–time profile following various dosing regimens evaluated in clinical studies, including the approved dose of SC 1080 mg twice weekly, and assessed the impact of patient characteristics on pegcetacoplan serum concentrations. The PK/PD analysis assessed the relationships between pegcetacoplan serum concentration and clinical markers of disease (i.e., Hb and LDH). Materials and Methods Data from Clinical Studies The pop PK model was designed to predict pegcetacoplan serum concentrations over time in patients with PNH to support approved initial dose selection in this population based on intrinsic and extrinsic patient factors present at treatment initiation (baseline). The pop PK model included data derived from 11 completed clinical studies, which enrolled patients with PNH, healthy adults, and adults with renal impairment (Table 1). The PK/PD models aimed to evaluate the magnitude of change of clinical markers of disease (i.e., Hb and LDH) with pegcetacoplan exposure to provide additional supportive evidence for the approved dose. The PK/PD model analyzed data derived from five completed clinical studies of pegcetacoplan in adult patients with PNH who had received C5 inhibitor treatment (PHAROAH, PEGASUS) or were naive to C5 inhibitors (PADDOCK, PALOMINO, PRINCE) (Table 1). Dosing regimens, patient population, and PK sampling information for each study are summarized in Table 1. Table 1Pegcetacoplan studies included in analysesStudyDesignPatientsNDosing regimenRoute | dose (mg) frequency | durationCP0713-1Ph1, double-blind, placebo-controlled, SADHealthy24SC | 45, 90, 180, 360, 720, 1440 x 1 dose | NACP1014Ph1, double-blind, randomized, MADHealthy16SC | 30, 90, 180, 270 once daily | 4 wk101Ph1, double-blind, randomized, multiple dosingHealthy36SC | 360 once daily, 1300 twice weekly, 2600 once weekly, 1080 twice weekly | 4 wk102Ph1, double-blind, randomized, SADHealthy Japanese16SC | 180, 360, 720, 1440 x 1 dose | NA401 [29]Ph1, double-blind, randomized vs control, SADHealthy16IV | 200, 600, 1500, 2300 x 1 dose | NAAIRISPh1, single-dose, open-labelRenal-impaired; healthy16SC | 270 x 1 dose | NAPHAROAH (NCT02264639) [21]Ph1, open-label, SAD and MADPNH, anemic on eculizumab9SC | 25, 50, 180–360, 270 once daily | 4 wk–1 yPADDOCK (NCT02588833) [20]Ph1b, open-label, MADPNH, C5i naive22SC | 180, 270, 360 once daily | 4 wk–1 yPALOMINO (NCT03593200) [20]Ph2a, open-label, MADPNH, C5i naive4SC | 270–360 once daily | 4 wk–1 yPEGASUS (NCT03500549) [22, 23]Ph3, randomized vs eculizumab (RCP), OLPaPNH, anemic on eculizumab80SC | 1080 twice weekly | 16–48 wkPRINCE (NCT04085601) [2]Ph3, randomized vs controlbPNH, C5i naive45cSC | 1080 twice weekly | 26 wkTotal284dC5i complement component 5 inhibitor, IV intravenous, LDH lactate dehydrogenase, MAD multiple ascending doses, NA not applicable, OLP open-label period, Ph phase, PNH paroxysmal nocturnal hemoglobinuria, RCP randomized controlled phase, SAD single ascending dose, SC subcutaneous, wk week, y yearaThe PEGASUS trial had several phases, including 4 weeks of run-in with dual therapy (eculizumab + pegcetacoplan), 16 weeks of randomized controlled monotherapy with either pegcetacoplan or eculizumab (RCP), and 32 weeks of OLP with continued pegcetacoplan monotherapy (for those who had been randomized to receive pegcetacoplan) or 4 weeks of transition to dual therapy followed by 28 weeks of pegcetacoplan monotherapy (for those who had been randomized to receive eculizumab)bSupportive treatment included blood transfusions, anticoagulants, corticosteroids, and supplements (iron, folate, and vitamin B12). Patients assigned to the control group could escape to pegcetacoplan therapy if they demonstrated hemoglobin levels ≥ 2 g/dL below their baseline measurement or presented with a qualifying thromboembolic event secondary to PNHcAll patients who received pegcetacoplan, including 35 randomized to pegcetacoplan and the 11 from the control group who switched to pegcetacoplan. The analysis dataset included 46 patients in PRINCE; however, 1 of them had no quantifiable samples and was excluded from data summariesdIn total, 165 patients with PNH were included in the PK/PD analysis, and 4498 Hb samples and 3423 LDH samples were analyzed Pegcetacoplan studies included in analyses C5i complement component 5 inhibitor, IV intravenous, LDH lactate dehydrogenase, MAD multiple ascending doses, NA not applicable, OLP open-label period, Ph phase, PNH paroxysmal nocturnal hemoglobinuria, RCP randomized controlled phase, SAD single ascending dose, SC subcutaneous, wk week, y year aThe PEGASUS trial had several phases, including 4 weeks of run-in with dual therapy (eculizumab + pegcetacoplan), 16 weeks of randomized controlled monotherapy with either pegcetacoplan or eculizumab (RCP), and 32 weeks of OLP with continued pegcetacoplan monotherapy (for those who had been randomized to receive pegcetacoplan) or 4 weeks of transition to dual therapy followed by 28 weeks of pegcetacoplan monotherapy (for those who had been randomized to receive eculizumab) bSupportive treatment included blood transfusions, anticoagulants, corticosteroids, and supplements (iron, folate, and vitamin B12). Patients assigned to the control group could escape to pegcetacoplan therapy if they demonstrated hemoglobin levels ≥ 2 g/dL below their baseline measurement or presented with a qualifying thromboembolic event secondary to PNH cAll patients who received pegcetacoplan, including 35 randomized to pegcetacoplan and the 11 from the control group who switched to pegcetacoplan. The analysis dataset included 46 patients in PRINCE; however, 1 of them had no quantifiable samples and was excluded from data summaries dIn total, 165 patients with PNH were included in the PK/PD analysis, and 4498 Hb samples and 3423 LDH samples were analyzed Model Development and Evaluation Analysis dataset assembly, exploratory graphical analysis, and pre-/post-processing of model outputs were performed using R (version 3.6.3 or above; R Core Team, Vienna, Austria). Model development and simulations were performed using NONMEM® (version 7.3 or above; ICON, Hanover, MS, USA). The first-order conditional estimation with interaction method was used for all analyses. Fitting of the PK/PD models followed a sequential approach conditioned on individual empirical Bayesian estimates of PK parameters from the pop PK model. Model development occurred in the same sequence for both pop PK and PK/PD analyses. First, a base structural model was identified, including both fixed (population central tendency) and random (variability) effect parameters. Following identification of the base structural model, covariate models were developed to assess the impact of intrinsic and extrinsic patient factors, with statistical selection performed to identify a parsimonious final model. Covariate models were developed using a full model approach including all covariate-parameter relationships of interest simultaneously, followed by statistical selection using a backward elimination procedure (α = 0.001) to identify parsimonious final models. Covariates evaluated in the pop PK model included race, sex, eculizumab treatment status at baseline, body weight, age, creatinine clearance (CrCl), aspartate aminotransferase, alanine aminotransferase, bilirubin, albumin, and body mass index. Covariates evaluated in the PK/PD models included race, sex, eculizumab treatment status at baseline, weight, age, CrCl, and baseline C3. Model selection was based on the following criteria: successful minimization and completion of covariance steps in NONMEM® (Table 1 of the Electronic Supplementary Material [ESM]), assessment of standard goodness-of-fit plots (Figs. 1 and 2 of the ESM), reductions in NONMEM® objective function value for hierarchical models, and reductions in residual variability. Model evaluation was performed using visual predictive checks [25]. Model Applications The identified pop PK and PK/PD models were applied to generate predictions to answer questions relevant to posology of pegcetacoplan in patients with PNH as follows: (1) what pegcetacoplan exposure is predicted from the first dose to steady state in patients with PNH under a given dosing regimen? (2) Do any intrinsic or extrinsic patient factors influence pegcetacoplan exposure or biomarker response to pegcetacoplan treatment? And (3) what biomarker response is predicted with pegcetacoplan treatment in patients with PNH with a given dosing regimen? Pegcetacoplan Exposure Predictions Stochastic simulations were performed to predict key pegcetacoplan PK parameters, concentration–time profiles, and exposure measures for healthy adults and adult patients with PNH. Individual PK parameter estimates and pegcetacoplan serum concentration–time profiles were simulated for 1000 healthy adults and 1000 adult patients with PNH using the pop PK model. Exposure metrics were derived from the simulated concentration–time profiles over 1-week intervals and included maximum concentration and average concentration. Effective half-life was calculated from the accumulation ratio for simulated exposure measures from week 1 to steady state [26]. Evaluation of Intrinsic and Extrinsic Patient Factors The potential impact, or lack thereof, of intrinsic and extrinsic patient factors on pegcetacoplan exposure evaluated in the full covariate model was visualized using forest plots. The ratio (test-reference) and 90% confidence interval (CI) for each covariate was derived from a simulation incorporating the uncertainty in parameter estimation. Body weight was the only covariate retained in the final model with a potential impact on posology; therefore, the effect of body weight on steady-state pegcetacoplan exposure was further evaluated across the full body weight range via a simulation. In the simulation, 1000 simulated steady-state pegcetacoplan concentration–time profiles were generated for patients with PNH at discrete body weight values spanning the range of body weights in the observed dataset. Pegcetacoplan serum concentrations associated with 90% of the maximum achievable biomarker response with pegcetacoplan treatment derived from the PK/PD models were overlaid on the simulated exposure by weight to support an integrated assessment of exposure and response. Biomarker Response Predictions Simulations were performed to predict Hb and LDH response to pegcetacoplan treatment at steady state, based on Hb and LDH measurements from study central laboratories using standard methodologies. The PK/PD relationship was displayed graphically with reference values of predicted pegcetacoplan average concentration at steady state over 1-week intervals from patients with PNH receiving SC pegcetacoplan 1080 mg twice weekly overlaid to facilitate a visual assessment of biomarker exposure response at this dose. Results Patient Population Population Pharmacokinetics In total, 284 individuals (Table 1) were included, and 5195 PK samples were analyzed, including 4737 (91%) quantifiable samples and 458 (9%) samples below the lower limit of quantification. Slightly more than half of the subjects were White (54%), though patients from PRINCE were primarily Asian (32/45, 71%), and the median age was 36 years (range 19–81 years) (Table 2 of the ESM). Subjects were evenly distributed across sex (44% female; 56% male) and patient status (44% healthy subjects; 56% patients with PNH). The median body weight for all subjects was 70.0 kg (range 41–156 kg), though the median body weight for patients from PRINCE was lower than median body weight for patients from PEGASUS (61.8 kg [range 41–95 kg] vs 72.4 kg [range 51–156 kg]). Median baseline C3 level across all subjects was 1.00 g/L (range 0.47–1.64 g/L). Pharmacokinetics/Pharmacodynamics In total, 165 patients with PNH were included, and 4498 Hb samples and 3423 LDH samples were analyzed. Patients were most frequently White (39%) or Asian (38%). As with the pop PK analysis, patients from PRINCE were primarily Asian (36/50; 72%). The median age was 45 years (range 19–81 years). Patients were evenly distributed across sex (56% female; 44% male) and baseline eculizumab treatment status (54% eculizumab treated; 46% eculizumab naive). Median baseline C3 level was 0.910 g/L across studies of patients with PNH (range 0.47–1.64 g/L). Population Pharmacokinetics and PK/PD models Population Pharmacokinetics The final model consisted of one central disposition compartment, a SC depot compartment, one transit compartment for absorption via the SC route, direct administration into the central compartment for the IV route, and first-order elimination from the central compartment (Fig. 3 of the ESM). Parameters were well estimated and model fit was adequate for the pegcetacoplan concentration–time data across studies (Table 3 and Fig. 4 of the ESM). Pharmacokinetics/Pharmacodynamics Sigmoidal maximum effect (Emax) models adequately described Hb and LDH biomarker concentrations as direct functions of pegcetacoplan concentration: Estimates of key structural parameters in the Hb and LDH PK/PD models are summarized in Table 2, and additional details can be found in Tables 4 and 5 and Figs. 5 and 6 of the ESM. In the Hb PK/PD model, a typical Hb baseline concentration was 8.74 g/dL in all patients with PNH. The pegcetacoplan concentration eliciting 50% of the maximal response was 337 μg/mL for Hb. The Emax was a 51% increase from baseline for all patients with PNH, regardless of prior C5 inhibitor treatment. In the LDH PK/PD model, a typical LDH baseline concentration was greater in eculizumab-naive patients (1920 U/L) than eculizumab-treated patients (249 U/L). The concentration eliciting 50% of the maximal response was 187 μg/mL for LDH. The Emax was a 92% decrease from baseline in eculizumab-naive patients and a 20% decrease from baseline in eculizumab-treated patients. Table 2Model structural parameters for the pharmacokinetic/pharmacodynamic analysisParameterEstimate (95% CI)bHbaLDHBaseline concentration Eculizumab naive8.74 (8.55–8.93) g/dL1920 (1740–2140) U/L Eculizumab treated8.74 (8.55–8.93) g/dL249 (226–273) U/LEmax, % change from baseline Eculizumab-naive51 (43–59)−92 (−93 to −90) Eculizumab-treated51 (43–59)−20 (−25 to −16) EC50, μg/mL pegcetacoplan337 (302–372)187 (166–211) Hill sigmoidicity coefficient4.66 (3.81–5.52)3.42 (2.84–4.00)CI confidence interval, CrCl creatinine clearance, EC50 pegcetacoplan concentration eliciting 50% of maximal effect, Emax maximum proportional drug effect, Hb hemoglobin, LDH lactate dehydrogenaseaValues at the reference CrCl of 120 mL/minbParameters estimated in the log or logit domains in the model are back-transformed to their natural units in the table. Parameters estimated as proportions are presented as percentages Model structural parameters for the pharmacokinetic/pharmacodynamic analysis CI confidence interval, CrCl creatinine clearance, EC50 pegcetacoplan concentration eliciting 50% of maximal effect, Emax maximum proportional drug effect, Hb hemoglobin, LDH lactate dehydrogenase aValues at the reference CrCl of 120 mL/min bParameters estimated in the log or logit domains in the model are back-transformed to their natural units in the table. Parameters estimated as proportions are presented as percentages Pegcetacoplan Exposure Predicted geometric mean (coefficient of variation) clearance was 0.356 L/day (30%) in patients with PNH versus 0.286 L/day (25%) in healthy patients (Table 3). Predicted median (interquartile range [IQR]) effective elimination half-life was 8.6 days (6.9–11.0) in patients with PNH versus 11.0 days (8.70–14.2) in healthy subjects (Table 3). Subcutaneous pegcetacoplan 1080 mg twice weekly was predicted to result in consistent exposure (median ≥ 600 µg/mL) with minimal peak-to-trough variation across all timepoints from 4 weeks to steady state in patients with PNH (Fig. 1). The model predicted that steady-state pegcetacoplan concentration exceeded 90% of the maximal response (EC90) for Hb and LDH responses in approximately 75% of patients with PNH (Fig. 1) with the median pegcetacoplan concentration exceeding these thresholds by week 4 of treatment. The model-predicted median (90% prediction interval [5th to 95th percentiles]) average concentration at steady state with SC pegcetacoplan 1080 mg twice weekly was 667 (423–994) µg/mL. Table 3Pegcetacoplan pharmacokinetics in healthy adults and adults with PNHPatientsSummary statisticClearance, L/dayCentral volume of distribution, LEffective half-life, daysHealthy adultsGM (CV%)0.286 (25)4.04 (27)11.6 (46)Median(IQR) [5th; 95th percentiles]0.285(0.246–0.334) [0.198; 0.417]4.05(3.42–4.72) [2.78; 5.95]11.0(8.70–14.2) [7.00; 24.7]Adults with PNHGM (CV%)0.356 (30)3.98 (32)9.00 (43)Median(IQR) [5th; 95th percentiles]0.351(0.299–0.424) [0.235; 0.553]3.95(3.27–4.79) [2.56; 6.34]8.60(6.90–11.0) [5.40; 17.5]CV coefficient of variation, GM geometric mean, IQR interquartile range, PNH paroxysmal nocturnal hemoglobinuria Pegcetacoplan pharmacokinetics in healthy adults and adults with PNH Median (IQR) [5th; 95th percentiles] 0.285 (0.246–0.334) [0.198; 0.417] 4.05 (3.42–4.72) [2.78; 5.95] 11.0 (8.70–14.2) [7.00; 24.7] Median (IQR) [5th; 95th percentiles] 0.351 (0.299–0.424) [0.235; 0.553] 3.95 (3.27–4.79) [2.56; 6.34] 8.60 (6.90–11.0) [5.40; 17.5] CV coefficient of variation, GM geometric mean, IQR interquartile range, PNH paroxysmal nocturnal hemoglobinuria Predicted exposure among adults with paroxysmal nocturnal hemoglobinuria receiving subcutaneous pegcetacoplan 1080 mg twice weekly. EC90 pegcetacoplan concentration producing 90% of the maximal response, Hb hemoglobin, IQR interquartile range, LDH lactate dehydrogenase Intrinsic and Extrinsic Patient Factors Population Pharmacokinetics Overall, the 90% CIs of the test-reference ratios for all covariate effects fell within or overlapped the 0.80–1.25 reference range for average concentration at steady state (Fig. 2), suggesting no clinically meaningful impact on pegcetacoplan exposure. Body weight was a statistically significant predictor of pegcetacoplan exposure; however, the 90% CIs of the test-reference ratios for the effect of body weight fell within or overlapped the 0.80–1.25 reference range for average concentration at steady state from the 5th to 95th percentiles of individual subject body weights in the observed data (54–95 kg) (Fig. 2). Effects are expected to be greater beyond these limits; however, predicted steady-state pegcetacoplan exposure exceeded the EC90 thresholds associated with robust biomarker responses across body weight groupings from 40 to 160 kg (Fig. 3A), though most individuals fell within the 54–95 kg range (Fig. 3B). Impact of intrinsic and extrinsic factors on pegcetacoplan exposure. The orange circles show the median ratio of the exposure metric under the test conditions compared with the reference. The blue line segments represent the corresponding 90% confidence interval. Vertical dashed lines indicate the reference interval of 0.80–1.25. ALT alanine aminotransferase, AST aspartate aminotransferase, BMI body mass index, C5 complement component 5, Cavg average pegcetacoplan concentration, CrCL creatinine clearance Impact of body weight on pegcetacoplan exposure (A) and body weight ranges (B). Boxes represent the interquartile range (IQR) with the median depicted as a solid line inside the box (C). The whiskers represent 1.5-fold below the first quartile or above the third quartile, with open circles indicating outliers beyond this range. Cavg,ss average pegcetacoplan concentration over 1 week at steady state, EC90 pegcetacoplan concentration producing 90% of the maximal response, Hb hemoglobin, LDH lactate dehydrogenase, PNH paroxysmal nocturnal hemoglobinuria, SC subcutaneous Pharmacokinetics/Pharmacodynamics In the Hb model, female sex was associated with a lower maximal response (33.8% increase from baseline) compared with male sex (51.0% increase from baseline) at the reference CrCl of 120 mL/min after pegcetacoplan dosing. Though sex influenced the maximal Hb response, the magnitude of difference was not anticipated to be clinically meaningful, as the ratio of female-to-male Hb achieved at steady state fell within the 0.8–1.25 reference range (0.890; 90% CI 0.850–0.929) (Fig. 4). Impact of sex and renal function on the predicted hemoglobin (Hb) level. The orange circles show the median ratio of the exposure metric under the test conditions compared with the reference. The blue line segments represent the corresponding 90% confidence interval. Vertical dashed lines indicate the reference interval of 0.80–1.25. CrCl creatinine clearance Lower baseline CrCl was also associated with a reduction in the maximal Hb response, such that a 25.2% maximal increase from baseline is predicted for male individuals at the 5th percentile of CrCl (40 mL/min) compared to a maximal increase of 50.1% at the approximate median observed CrCl (120 mL/min). Though baseline CrCl had an effect on Hb Emax, the magnitude of difference is not anticipated to be meaningful (Fig. 4). Steady-state Hb was predicted to be 0.841-fold (90% CI 0.806–0.879) lower at the 5th percentile of baseline CrCl (40 mL/min) and 1.16-fold (90% CI 1.10–1.22) greater at the 95th percentile (191 mL/min) relative to the approximate median baseline CrCl of 120 mL/min. There were no meaningful effects on Hb with all other covariates (i.e., prior eculizumab treatment, race, age, body weight, and baseline C3 level). Eculizumab treatment status affected baseline LDH and Emax, consistent with the known effect of C5 inhibition on this biomarker. The Emax values of pegcetacoplan were estimated at a 91.7% decrease in LDH from a baseline of 1920 U/L for eculizumab-naive patients and a 20.0% decrease in LDH from a baseline of 249 U/L for eculizumab-experienced patients. However, despite the large difference in baseline LDH values, the steady-state LDH ratio was predicted to be 1.09 (90% CI 0.980–1.19) for eculizumab-experienced patients relative to eculizumab-naive patients at a SC pegcetacoplan dose of 1080 mg twice weekly, which is within the 0.80–1.25 reference range (Fig. 5). There was a predicted increase in maximal response during eculizumab co-treatment, resulting in a 35.0% decrease from baseline compared to a 20.0% decrease in baseline for the eculizumab-to-pegcetacoplan monotherapy switch. Despite these covariate effects, prior eculizumab or continued co-treatment with eculizumab did not result in a clinically meaningful difference in steady-state LDH with all conditions falling inside the 0.80–1.25 range compared with a treatment-naive patient receiving pegcetacoplan monotherapy. All other factors (i.e., sex, baseline CrCl, race, age, body weight, and baseline C3 level) did not have meaningful effects on the LDH response. Impact of prior eculizumab treatment on the predicted lactate dehydrogenase (LDH) level. The orange circles show the median ratio of the exposure metric under the test conditions compared with the reference. The blue line segments represent the corresponding 90% confidence interval. Vertical dashed lines indicate the reference interval of 0.80–1.25 Biomarker Response to Pegcetacoplan The relationship between pegcetacoplan and biomarker response followed a sigmoidal, rather than simple, Emax curve with hill coefficients of 4.66 and 3.42 for Hb and LDH, respectively (Table 2; Fig. 6). The median (IQR) predicted Hb level with steady-state pegcetacoplan was 11.7 (10.5–13.3) g/dL for all patients. The median (IQR) LDH levels with steady-state pegcetacoplan were 217 (170–292) U/L for eculizumab-naive patients and 192 (152–239) U/L for eculizumab-treated patients. The probability of reaching LDH levels < 1.5 × ULN (226 U/L from reference laboratory for phase III studies) was 84% for eculizumab-naive patients and 97% for eculizumab-treated patients. The median (IQR) steady-state pegcetacoplan concentration of 667 (551–782) μg/mL associated with SC 1080 mg twice-weekly dosing was estimated to achieve ≥ 95% of the maximal effect for Hb and LDH, regardless of eculizumab treatment history (Fig. 6). Maximal effect on hemoglobin and lactate dehydrogenase (LDH) with subcutaneous pegcetacoplan 1080 mg twice weekly in patients with paroxysmal nocturnal hemoglobinuria. The dashed vertical lines indicates the pegcetacoplan concentration eliciting 50% of maximal effect for hemoglobin (yellow) and LDH (maroon). The solid vertical line indicates the median pegcetacoplan concentration following subcutaneous pegcetacoplan 1080 mg twice-weekly dosing Discussion Pegcetacoplan displayed a long half-life of 8.60 days (IQR 6.90–11.0) in patients with PNH versus 11.0 days (8.70–14.2) in healthy subjects, with minimal peak to trough variation at steady state in patients with PNH, leading to consistent pegcetacoplan exposure. Patients receiving SC pegcetacoplan experience approximately constant serum concentrations at steady state, which differs from the large peak-to-trough concentration fluctuations observed with the IV C5 inhibitors [6]. Covariates of age, sex, race/ethnicity, renal function, hepatic function, and eculizumab treatment had no clinically meaningful impact on exposure to pegcetacoplan or biomarker levels following pegcetacoplan administration; therefore, no adjustment to initial pegcetacoplan dosing is needed for any specific patient populations based on characteristics present at treatment initiation. Weight impacted pegcetacoplan exposure; however, the magnitude of impact is modest over the range of body weight values encompassing 90% of subjects in the observed dataset (54–95 kg). Steady-state pegcetacoplan concentration was predicted to effectively increase Hb levels and suppress LDH levels both across body weight and regardless of C5 inhibitor treatment history in patients with PNH. No adjustment to initial dosing is needed with regard to body weight given the robust biomarker response predicted, even at extremes of body weight (< 53 kg or > 94 kg) in the analysis dataset; however, close clinical monitoring is warranted in patients at extremes of body size, as it is with all patients with PNH. In the Hb model, sex and CrCl were associated with effects on Hb Emax after pegcetacoplan dosing; however, the magnitude of difference for each was not anticipated to be meaningful. The reduced Hb maximal response after pegcetacoplan for female versus male individuals (33.8% vs 51.0%) was expected given that circulating Hb is known to be reduced in female versus male individuals [27]. Prior or co-treatment of eculizumab impacted LDH Emax; however, the effects were not clinically meaningful as the test-reference ratios fell within the reference range. Model-based predictions suggest that, despite large differences in baseline LDH levels, SC pegcetacoplan 1080 mg twice weekly is equally effective at achieving LDH control as initial monotherapy or sequential therapy with a switch from a C5 inhibitor. The pop PK and PK/PD analyses support the intended use of pegcetacoplan in improving Hb and LDH as markers of hemolysis at the SC 1080-mg twice-weekly dosage. Several limitations of the study should be taken into consideration when interpreting the results. The studies in patients with PNH used sparse PK sampling designs, which have limited information to evaluate patient-specific effects on absorption and distribution phase parameters. An empirical modeling approach was employed to facilitate pooling of data across diverse clinical trials and support a robust assessment of baseline intrinsic and extrinsic factors for the purpose of assessing support for the approved dose of SC pegcetacoplan 1080 mg twice weekly in the PNH population. The model does not account for target-mediated disposition because of high affinity binding to the C3 target. Additionally, the model does not account for dynamic complement activating conditions (i.e., infections, surgery) and breakthrough hemolysis episodes, which may alter pegcetacoplan and/or C3 disposition; therefore, it is not directly applicable to projecting dose adjustments for such scenarios without modification. Finally, indirect PK/PD models of Hb and LDH biomarkers incorporating a time delay between pegcetacoplan concentrations and subsequent biomarker response could not be identified, potentially owing to limited sampling immediately following treatment initiation; therefore, the models are best fit-for-purpose of evaluated steady-state dosing conditions. Future studies could include an assessment of a semi-mechanistic modeling framework incorporating target-mediated drug disposition to allow prediction of target engagement. Furthermore, additional research is needed on the modeling of dynamic states to assess dosing in acute clinical scenarios, such as in patients with breakthrough hemolysis [28]. Last, PK/PD modeling of other markers of complement activity would add to the understanding of appropriate pegcetacoplan dosing and subsequent clinical effects. Conclusions Pegcetacoplan displayed consistent pharmacokinetics, with steady-state exposure exceeding thresholds associated with robust biomarker responses across body weight categories and regardless of eculizumab treatment history. Based on covariate analyses, no pegcetacoplan initial dose adjustments were needed because of differences in age, sex, weight, race/ethnicity, renal function, potential hepatic damage, baseline C3, or prior eculizumab. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (PDF 1995 kb) Acknowledgements The authors acknowledge Jessica Savage and Francois Gaudreault for their contributions to the study. Medical writing and editorial support was provided by Caryne Craige, Ph.D., Kay Square Scientific, Newtown Square, PA, USA, which was funded by Apellis Pharmaceuticals, Inc., and Swedish Orphan Biovitrum AB. Declarations Funding The study was funded by Apellis Pharmaceuticals, Inc., and Swedish Orphan Biovitrum AB. Conflict of interest Ryan L. Crass was an employee at Apellis Pharmaceuticals at the time of the study. Grant Langdon is a consultant to Swedish Orphan Biovitrum and does not own stock and/or stock options. Brandon Smith, Sven Adriaens, and Sunny Chapel have no conflicts that are directly relevant to the content of this article. Ethics approval Not applicable. Consent to participate Not applicable. Consent for publication Not applicable. Availability of data and material The datasets generated and/or analyzed during the current study are not publicly available. Data used for this study are from Apellis Pharmaceuticals, Inc., and were used under license for the current study. Therefore, restrictions apply to the availability of these data, which cannot be made publicly available. Access to this dataset may be granted to other interested parties by Apellis Pharmaceuticals, Inc., but may be subject to limitations based on how the original clinical trial data were acquired. Code availability Not applicable. Authors’ contributions All authors were involved in the design of the model, preparation of the manuscript, and review and approval of the manuscript. References Associated Data Supplementary Materials Supplementary file1 (PDF 1995 kb)