PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of brjclinpharmLink to Publisher's site
 
Br J Clin Pharmacol. 2016 May; 81(5): 918–928.
Published online 2016 March 4. doi:  10.1111/bcp.12864
PMCID: PMC4834601

Population pharmacokinetic analysis of sifalimumab from a clinical phase IIb trial in systemic lupus erythematosus patients

Abstract

Aims

Sifalimumab, a human immunoglobulin (Ig) G1 monoclonal antibody against INF‐alpha, is being studied as a treatment for systemic lupus erythematosus (SLE). This analysis characterized population pharmacokinetics (PK) of sifalimumab following repeat fixed dose and evaluated the utility of fixed dosing vs. body weight normalized dosing in SLE patients.

Methods

PK data were collected in a phase IIb study where 298 patients received multiple intravenous doses (200–1200 mg) of sifalimumab every 4 weeks for 52 weeks. A population pharmacokinetic model was developed using 3961 quantifiable serum concentrations and the impact of patient demographics, clinical indices and biomarkers on pharmacokinetic parameters was evaluated. The appropriateness of the final model was evaluated using visual predictive check and bootstrap.

Results

A two compartment model with first order elimination adequately described sifalimumab serum PK. The estimated typical clearance (CL) and central volume of distribution (V 1) were 184 ml day–1 and 2.82 l with 24% and 16% between‐subject variability (BSV), respectively. Body weight, dose, 21 INF gene signature baseline and concomitant steroid use were identified as statistically significant covariates for CL and V 1 and accounted for <10% of PK variability in the final model. Typical values and BSV of PK parameters from the current analysis with fixed dosing were similar to previous population PK results with body weight normalized dosing.

Conclusions

The transition from body weight normalized dosing to fixed dosing did not impact sifalimumab PK. These findings support the use of fixed dosing for sifalimumab in future clinical studies evaluating it as a potential treatment for SLE.

Keywords: clinical trial, population pharmacokinetic modelling, sifalimumab, systemic lupus erythematosus

What is Already Known about this Subject

  • Sifalimumab therapy was well tolerated and resulted in a decrease in the expression of IFN‐induced genes in SLE patients in early clinical trials.
  • Body weight did not explain inter‐subject variability in pharmacokinetic (PK) parameters of sifalimumab to any relevant extent (<7%) following body weight normalized dosing regimen in phase Ib trial.

What this Study Adds

  • The effect of clinical factors e.g. dose, body weight, baseline signature gene and baseline steroid use on PK of sifalimumab following fixed dosing regimen is limited and is not expected to be clinically relevant.
  • PK characteristics of sifalimumab are similar following fixed or body weight normalized dosing regimen, supporting fixed dose strategy for future clinical development of sifalimumab.

Introduction

Systemic lupus erythematosus (SLE) is a complex autoimmune disease associated with hormonal, environmental and genetic factors and linked to the tolerance breakdown of B and T cells to self‐antigens. SLE can occur at any age and in men and women, but predominantly affects women of childbearing years 1, 2. Clinical manifestations of SLE include constitutional symptoms, alopecia and rashes, inflammatory arthritis, renal disease, systemic vasculitis, lymphadenopathy, splenomegaly, haemolytic anaemia, cognitive dysfunction and other central nervous system (CNS) involvement. Multiple immune system abnormalities contribute to the pathogenesis of SLE, including abnormal clearance of apoptotic cells and immune complexes, over‐production of type I interferon (IFN), reduced thresholds for B and T lymphocyte activation and production of autoantibodies against self‐antigens 3.

In recent years, a better understanding of the pathogenic mechanisms of SLE has led to new therapeutic targets, which avoid conventional immunosuppressant mechanisms that lead to some undesirable effects. Several attempts have been made to test new immune‐modulating drugs with different targets such as B lymphocyte surface molecules (CD20, CD22, and CD19) and co‐stimulatory molecules (CTLA‐4, CD40/CD40L, ICOS/B7‐H2), as well as extracellular molecules (cytokines, chemokines) 4, 5. An example among those attempts is belimumab, a monoclonal antibody blocking B lymphocyte stimulator (BLyS), which is the first approved drug for SLE treatment since 1957. Other strategies targeting cytokines such as interleukin and IFN‐alpha are also attractive and hold great therapeutic potential as SLE treatment.

IFN‐alpha, which is secreted by plasmacytoid dendritic cells (pDCs), has autoimmunity enhancing effects. It promotes the maturation of dendritic cells, activates T lymphocytes (TL) and intervenes in B lymphocytes (BL) differentiation to auto‐antibody producing dendritic cells. It has been observed that up to 70% of adults and 95% of children with SLE have an overexpression of IFN and IFN‐induced genes in circulating mononuclear cells and peripheral tissues 4. There is substantial evidence that IFN plays a significant role in the pathogenesis of lupus 3, 6. Recently, several IFN‐alpha blockade strategies were tested and are expected to display some benefit in SLE treatment 7.

Sifalimumab is a human immunoglobulin G1 kappa (IgG1κ) monoclonal antibody (mAb) that non‐competitively binds to and neutralizes human IFN‐alpha. Sifalimumab‐specific inhibition of the overexpression of IFN‐inducible mRNAs has been previously reported in the blood of SLE patients from earlier clinical studies 8. A small, robust panel of IFN‐induced mRNAs of 21 genes was chosen as a pharmacodynamic (PD) biomarker to evaluate target inhibition by sifalimumab in SLE patients, because IFN‐inducible mRNAs is concordantly overexpressed in blood and involved tissue in SLE patients, and is easy to measure by either RT‐PCR or microarray‐based technologies 9. Previous phase 1a and 1b clinical studies with body weight normalized intravenous (i.v.) doses of sifalimumab up to 30 mg kg–1 have demonstrated inhibition of IFN‐induced mRNAs and an acceptable safety profile 10, 11, 12.

Besides evaluation of safety and efficacy of sifalimumab in SLE patients, the pharmacokinetics (PK) of sifalimumab were also investigated in single or multiple dose clinical studies with body weight normalized dosing regimen. Previous population PK analysis of phase 1b data found that covariates including body weight did not explain inter‐subject variability in PK parameters to any relevant extent (<7%) 13 and consequently proposed fixed dosing regimens for future clinical development.

The recently completed phase IIb study was therefore initiated with fixed doses of 200, 600 and 1200 mg, to evaluate safety, efficacy and PK of sifalimumab in a large number of chronic, moderately‐to‐severely active SLE patients with an inadequate response to standard of care treatment for SLE (SOC SLE). The purpose of the current analysis was to characterize sifalimumab PK in SLE patients receiving multiple fixed dose (200, 600 or 1200 mg) i.v. infusions of sifalimumab every 4 weeks for 52 weeks, with an additional loading dose on day 15. The impact of body weight, age, treatment, race, gender, baseline gene signature from 21 genes, baseline steroid use, baseline SLE disease activity index and antidrug antibody (ADA) on sifalimumab PK in SLE patients was evaluated. To evaluate the utility of fixed dosing vs. body weight dosing, the results of the current analysis were compared with previous population PK modelling with body weight normalized doses and the recommendation to study fixed doses of sifalimumab.

Methods

Study design and patient population

A total of 431 adult SLE patients were enrolled into this phase IIb, multinational, multicentre, randomized, double‐blind, placebo‐controlled, parallel group study to evaluate the efficacy and safety of three intravenous treatment regimens of sifalimumab. The study was conducted in compliance with the Declaration of Helsinki and the Guidelines for Good Clinical Practice. Written informed consent was obtained from all patients. Approximately 150 international sites from North and South America participated in this study.

Subjects were randomized in a 1 : 1 : 1 : 1 ratio to receive a blinded, fixed i.v. dose of sifalimumab (200, 600 or 1200 mg) or placebo as an i.v. infusion every 4 weeks (days 1, 29, 57, 85, 113, 141, 169, 197, 225, 253, 281, 309 and 337) with an additional dose on day 15, for a total of 14 doses. Subjects were followed for 180 days after the last dose of investigational product on day 337 (week 48) through day 517 (week 74). From signing of the informed consent to the completion of day 365, subjects received stable doses of baseline SOC including OCS, antimalarials and slow‐acting immunosuppressants. A limited number of corticosteroid bursts and tapers with maximum protocol‐defined duration and dosing were permitted for increased SLE disease activity.

Blood sampling schedule

Blood samples for sifalimumab PK assessments were collected pre‐dose on days 1, 15, 85, 169, 253 and 337, end‐of‐infusion on days 1, 15, 29, 169, 253 and 337 and on days 365, 396, 427 and 517. Serum concentrations of sifalimumab were determined using a previously published validated method of ELISA with a lower limit of quantitation of 1.25 μg ml–1 13. Antidrug antibody to sifalimumab (ADA) was measured using a sensitive electrochemiluminescent bridging immunoassay approximately every 84 days on days 1, 85, 169, 253, 337 and on each of the visits during the 180 day follow‐up period on days 365, 396, 427 and 517.

Data handling

Drug concentrations below the lower limit of quantification (BLQ) were considered as missing. Outlier data points and any concentration data point without associated sampling time were excluded from the dataset. Typically, a data point was deemed an outlier if there was an abnormal unexplained spike or drop in concentration compared with the rest of the individual's concentration–time profile. Outlier data points were identified by visual inspection of each individual's concentration–time profile and omitted from the population PK analysis. Over 97% of 4082 PK data points collected from 298 patients were retained in the final dataset.

Population pharmacokinetic analysis

Prior to population PK analysis, non‐compartmental analysis of PK data post the last dose was performed using WinNonlin Phoenix (Pharsight Corporation, Mountain View, CA, USA) to assess PK profile and estimate PK parameters of sifalimumab.

A population PK model was developed to fit serum concentrations of sifalimumab using the first order conditional estimation with interaction method (FOCE‐I) using nonmem, software, version 7.2 14. Appropriateness of the model was evaluated using various goodness‐of‐fit criteria, including diagnostic scatter plot, likelihood ratio tests (LRTs) and measures of model stability and adequacy (successful convergence, significant digits and matrix singularity). The results of LRTs were considered statistically significant if decreases in the objective function value (OFV) of the nested models was more than 3.84 (P < 0.05, one degree of freedom) throughout the model‐building process. Between‐subject variability (BSV) was estimated on all structural model parameters, assuming log‐normal distributions. Since this was a repeat dose, a year long clinical trial, interoccasion variability (IOV) was also examined with an exponential random effects term. An occasion was characterized as the time period from the start of an infusion until the start of the next administration. The inclusion of IOV was limited to seven infusions due to insufficient data. The residual variability, which comprised, but was not limited to, intra‐individual variability, experimental errors, process noise and/or model misspecifications, was modelled using additive, proportional and combined error model structures. All data preparation, summary statistics, graphics, exploratory analyses and post‐processing of nonmem outputs were performed in R (version 3.0.1) 15.

Covariate analysis

Covariate analysis was performed by testing body weight, age, gender, race, region, dose, baseline steroid use (BSTEROID), baseline interferon gene signature from 21 genes (BGENE21) which is IFN signature score calculated from expression level of a mRNA panel of 21 interferon inducible genes, baseline SLE disease activity index (BSLEDAI) score which is a validated assessment of disease activity in lupus and ADA. Preliminary screening of covariates was conducted based on the examination of individual PK parameter predictions (empirical Bayes estimates or EBEs) obtained from the base model. The relationships between EBEs and covariates were evaluated graphically and statistically by linear and non‐linear regression techniques. Covariates for which a significant (P < 0.05) relationship between covariate and EBE individual parameters was evidenced were tested in nonmem, as well as those for which a significant effect was reported in the previous population PK analysis of sifalimumab. Covariate model building in nonmem was a stepwise process, consisting of a forward and a backward selection procedure. A covariate was retained in the model if a reduction in the OFV was ≥3.84 (P < 0.05). After defining the full model, the significance of each covariate was tested individually by removing each one from the full model. A covariate was retained in the model if, upon removal, the OFV increased by more than 6.64 points (P < 0.001).

Model evaluation

General goodness of fit plots such as observed vs. model predicted concentrations, as well as conditional weighted residuals (CWRES) were utilized in assessing model appropriateness. Distributions of BSV were visualized through histograms to verify normal distribution. Eigen values were calculated during the covariance step and used to determine condition number. The shrinkage of model parameters was used to assess if the model was over‐parameterized for the PK data. A non‐parametric bootstrap resampling method was used to evaluate the stability and robustness of the final PK model. Resampling with replacement generated 500 bootstrap data sets. The final population PK model was then fitted repeatedly to each of the 500 bootstrap data sets, with re‐estimation of the population PK parameters. The median and 95% confidence intervals of parameters obtained from this step were compared with the final parameter estimates. In addition, the adequacy (or appropriateness) of the final model was evaluated using visual predictive checks (VPC). VPC evaluated whether the model is able to produce simulated data that are similar to the original observed data. Five hundred replicates of the study design were simulated using the final model and 95% confidence intervals for the 5th, 50th (median) and 95th percentiles of the concentrations at each time point were computed. These 95% confidence intervals were then compared graphically with the 5th, 50th and 95th percentiles from observed data.

Results

Subject characteristics

Descriptive statistics of the demographic data of the 298 subjects who received sifalimumab in the trial are shown in Table 1. The majority of the SLE patients were women. The overall age range was 18–73 years with a median of 40 years. The overall range of body weight was 39 to 131 kg with a median body weight of 64.3 kg. There were 273 female and 25 male patients, who were mostly White (59%), Asian (15%), African American subjects (7%), American Indian or Alaskan native (4%) and other race (15%) comprising the rest.

Table 1
Patient demographic characteristics and baseline pharmacodynamic biomarker values

Non‐compartmental pharmacokinetic analysis

The PK profile of sifalimumab in this phase IIb study showed that sifalimumab serum concentrations increased dose‐proportionally from 200 mg to 1200 mg administered as an intravenous infusion in SLE patients. The additional loading dose on day 15 resulted in faster achievement of steady‐state with peak serum concentrations on day 29 in the range of 125–639 μg ml–1 across the three dose cohorts. At steady‐state, peak serum concentrations ranged from 117 to 562 μg ml–1 across all three cohorts. Non‐compartmental analysis estimated steady‐state mean clearance (CL) and half‐life values of 145 ml day–1 and 24 days, respectively. C max, C trough and AUC(0,τ) at steady‐state increased linearly in a dose proportional manner over the dose range of 200 mg to 1200 mg. Following the last dose, sifalimumab serum concentrations declined in a bi‐exponential fashion indicating two compartment linear pharmacokinetics.

Population PK model

Following inspection of the raw data and data cleaning, 3961 out of 4082 observations (97%) were included in the analysis. During model development, a two compartment model with first order elimination provided a reasonable fit of the data with the lowest objective function value (OFV) and biologically reasonable parameter estimates. Between‐subject variability (BSV) was estimated on all base model parameters except volume of the peripheral compartment (V 2) and distribution clearance (Q), because BSV on V 2 was negligible and not estimable and RSE% of BSV on Q was greater than 60%. In addition, IOV was implemented. The best result (lowest OFV, smallest relative standard errors in % (RSE %) was achieved with IOV on central compartment volume (V 1), where every infusion corresponded to one occasion ([increment]OFV ~ −13). Residual variability was best described by a combined additive and proportional error model. The typical value (% BSV) of CL, V 1, V 2 and Q estimates from the base model were 0.19 l day–1 (33%), 2.81 l (23%), 1.89 l and 0.32 l day–1, respectively, and listed in Table 2. BSV in base model was moderate (23–33%) and similar to IOV of 30%. There was no significant correlation between BSV of the different PK parameters.

Table 2
Population pharmacokinetic parameters of sifalimumab from the base and final models

The covariate model evaluated the influence of age, gender, ethnicity, region, body weight (WT), sifalimumab dose, BSTEROID, BSLEDAI, BGENE21 and ADA. The statistically significant covariates impacting sifalimumab CL and V 1 in the final model included WT, BGENE21, dose and BSTEROID. The rest of the covariates did not have a significant impact on either CL or V 1.

Sifalimumab CL increased with WT, BGENE21 score and steroid use while, V 1 increased with WT and sifalimumab dose, and decreased with steroid use. A non‐linear power relationship was used to model the effect of WT, dose and BGENE21 on CL and V 1 and the effect of BSTEROID was modelled as a categorical variable. The final model functions for typical value of CL, V 1, V 2 and distribution clearance (Q) are presented as follows (Equations (1)(4)).

CL=θ1×WTMedianθ5×BGENE21Medianθ6×(1+θ7×BSTEROID)
(1)

V1=θ2×WTMedianθ8×Dose600θ9×(1θ10×BSTEROID)
(2)

V2=θ3
(3)

Q=θ4
(4)

where θ1 is the CL of a typical patient representing median WT of 64.3 kg, median BGENE21 score of 12.04 and who is not on steroids at baseline. θ2 represents V 1 of a typical patient with median WT of 64.3 kg, 600 mg dose and who is not on steroids at baseline. θ5–θ10 are the exponents of covariate effect on respective pharmacokinetic parameters.

Population PK parameter estimates for the final model are shown in Table 2. The estimated values for CL, V 1, V 2, and Q for a typical patient were about 184 ml day–1, 2.82 l, 1.88 l and 340 ml day–1, respectively. The estimates (coefficient of variation) of BSV associated with CL and V 1 were 24 and 16%, respectively. The BSV of Q was still not retained in the final model due to high RSE%. The IOV (coefficient of variation) on V 1 was 27%, which was slightly greater than BSV. All pharmacokinetic parameters were estimated with good precision, as reflected by relative standard errors (RSE%). The 3% and 21% shrinkage rates of CL and V 1, respectively, indicated that data were informative to estimate individual CL and V 1. The 7% EPS‐shrinkage indicated the individual predictions were of value for assessing model adequacy. The condition number was 103, which indicated that the model was not over‐parameterized and that there was no evidence of collinearity 16.

The relationship between significant covariates and PK parameters of each patient are presented in Figure 1. Body weight had the largest effect on CL and V 1 among the statistically significant covariates. Sifalimumab CL and V 1 increased with WT with an exponent of 0.46 (θ5) and 0.36 (θ8), respectively, indicating a less than proportional increase. Over the WT range of 39 kg to 131 kg in the study, CL and V 1 ranged from 147 to 250 ml day–1 (80–137% of typical value) and from 2.2 to 3.3 l (80–118% of typical value), respectively. In comparison with the base model, BSV on CL and V 1 in final model decreased by 9% and 7%, respectively, upon incorporation of WT as a covariate.

Figure 1
Plot of individual CL and V 1 values of sifalimumab over range of statistically significant covariate values. Individual estimations of PK parameters by final model are presented as red circles. The function of covariate on PK parameters (CL and V 1) ...

Sifalimumab CL was higher by 0.11 l day–1 in patients receiving steroid therapy while V 1 was lower by 0.09 l compared with patients not on steroids at baseline. Incorporating BSTEROID as a covariate reduced BSV on CL and V 1 by 2% and 1%, respectively. Sifalimumab CL increased with BGENE21 score with an exponent of 0.09, with CL ranging from 132 to 203 ml day–1 (72–111% of typical value) over the range of BGENE21 scores (0.32–38.59). Incorporation of BGENE21 into the final model reduced BSV on CL by 4%. Sifalimumab V 1 increased slightly (exponent 0.06) with dose from 2.6 to 2.9 l over the dose range of 200 to 1200 mg, respectively, explaining about 5% of BSV in V 1. There was no impact of ADA on sifalimumab pharmacokinetics.

The predicted vs. observed concentration profile plots for representative individuals from 200, 600 and 1200 mg dose groups are presented in Figure 2. Standard diagnostic plots of observed vs. individual and population predicted values (Figure 3) showed good agreement with observed data. There was slight disagreement between predictions and observations at the end of terminal phase after the last dose. This is probably due to sparseness of data at these time points. Residual plots show no major trends in general, indicating no apparent bias with the final model. The final model was validated with bootstrap and VPC. Bootstrap analyses showed a success rate of 71%. The model parameter estimates obtained from the final model were overall in good agreement with the respective bootstrap values (Table 2). VPC stratified by dose (Figure 4) showed that the model adequately described the observed data and had good simulation characteristics and predictability.

Figure 2
Predicted vs. observed PK profiles of representative individuals. The black circles represent observations. The blue and red lines represent individual predictions (IPRED) and population predictions (PRED), respectively. The red arrows on x axis indicate ...
Figure 3
Goodness‐of‐fit plots for the final model. A) Scatterplot of the individual observed plasma concentrations (μg ml–1) vs. the individual concentrations (μg ml–1). B) Scatterplot of the individual ...
Figure 4
Visual predictive checks (VPC) for the final population pharmacokinetic model of sifalimumab. VPC plots stratified by dose were generated. The red lines represent the 10th, 50th and 90th percentiles of the observed data (blue circles), the shaded grey ...

Discussion

This phase IIb study was primarily designed to evaluate further the efficacy and safety of sifalimumab in SLE patients. Based on the recommendation from the previous population analyses, fixed sifalimumab doses of 200, 600, and 1200 mg were administered monthly for 1 year in this phase IIb study. Encouraging data from this phase IIb study demonstrated that sifalimumab resulted in greater efficacy than placebo in the treatment of patients with moderate to severe active SLE with an inadequate response to standard of care. A broad‐based improvement was observed in both SLE composite endpoints (SRI, BICLA) and individual organ systems (CLASI, joint counts). A greater distinction from placebo response noted in IFN high patients supports the hypothesis that the peripheral blood IFN test status reflects systemic type I IFN activity. These potentially promising results are important in the early development of a drug but are not definitive until prospectively replicated in larger studies. Adverse events occurred with similar frequencies in the sifalimumab and placebo groups except for Herpes zoster infection, which was more common with sifalimumab. This is consistent with the mechanism of action of sifalimumab and safety results reported from a previous study 11.

In the current population PK analysis, sifalimumab concentration–time data from 298 patients were analyzed using non‐linear mixed effects modelling to characterize sifalimumab PK in SLE patients receiving multiple fixed doses (200, 600 or 1200 mg). Based on results of non‐compartmental analysis, the linear PK of sifalimumab were demonstrated over a wide dose range in this clinical trial. The serum concentration–time profile of sifalimumab post the last dose shows the expected bi‐exponential decline characteristic of a monoclonal antibody exhibiting linear PK. The linear PK of sifalimumab were also supported by the mechanism of action of sifalimumab, which is to block binding between IFN‐alpha and IFN receptor by neutralizing soluble target, IFN‐alpha. Study of fluorescent labeled sifalimumab in an in vitro study demonstrated that the sifalimumab‐IFN complex is not internalized and hence sifalimumab should not exhibit non‐linear PK due to target mediated drug disposition in vivo. Therefore, a linear compartmental PK model was developed and validated in this analysis to describe PK data of this phase IIb study.

The estimates of CL, V 1, V 2 and Q for a typical SLE patient not on steroids, weighing 64.3 kg, with a BGENE21 score of 12.04, receiving a 600 mg dose of sifalimumab were about 184 ml day–1, 2.82 l, 1.88 l and 0.34 l, respectively. BSV estimates of CL and V 1 were about 24% and 16%, respectively. The typical values and BSV of the PK parameters were comparable with those estimated from the phase Ib study where body weight normalized dose was used 13. The model‐estimated PK parameter values are also in good agreement with those of other therapeutic monoclonal antibodies and endogenous IgG immunoglobulins without target mediated drug disposition 17, 18, 19, 20, 21, 22, 23, 24.

The elimination of mAbs exhibiting linear PK is primarily via the neonatal Fc receptor (FcRn) pathway, where mAb bound to FcRn is salvaged, resulting in a long half‐life of up to 4 weeks, while the unbound mAb is degraded. Therefore, the expression level and pattern of FcRn play an important role in regulating elimination of mAbs. In adults, FcRn is expressed on parenchymal cells including hepatocytes, endothelia and epithelia of numerous mucosal surfaces including the intestines and haematopoietic cells 25. The relationship between body weight and FcRn expression in adults is not established in the literature. However, it is well established that CL of mAbs exhibiting linear PK scales allometrically with body weight in adults and this points to higher FcRn expression with increasing body weight. Therefore, body weight normalized dosing regimens have been utilized in clinical trials of many therapeutic mAbs to reduce PK variability. However, Rodriguez et al. reported that body weight normalized dosing did not offer any advantages over fixed dosing in reducing PK variability of many therapeutic mAbs 26. In addition, two reports using 30 biologics with published population PK and/or PD models found that body size normalized dosing and fixed dosing showed similar PK and/or PD variability across the biologics investigated, with fixed dosing being better for some biologics 27, 28. Another published simulation analysis demonstrated that, in general, when the exponents for effect of body weight on CL and V 1 in the population PK model are <0.5, fixed dosing results in less variability and less deviation than body weight‐based dosing. When both exponents are >0.5, body weight normalized dosing results in less variability and less deviation than fixed dosing 29. Given many practical advantages such as ease of preparation and administration, less risk of medical errors, better patient compliance and cost effectiveness, fixed dosing is recommended as the first option in clinical studies of mAbs that have a relatively large therapeutic window, good safety profile and a small contribution of body size to PK variability.

The similarity in parameter values between current and previous PK analyses 13 demonstrates that the PK characteristics of sifalimumab are unchanged with fixed or body weight normalized dosing regimen, and the similarity in BSV values of CL and V 1 of base models confirms minimal impact of body weight on sifalimumab serum exposures. The exponential functions of body weight on CL and V 1 of sifalimumab in the final model of both analyses were <0.5 and resulted in ≤37% change in CL and V 1 values over the body weight range of 39 to 131 kg in this study. Incorporating body weight as a covariate decreased BSV on CL and V 1 by <10% only, indicating limited impact of body weight on inter‐subject PK variability of sifalimumab. Consistent with a previous population PK analysis of a phase Ib study, the current population PK analysis confirmed that a fixed dosing regimen did not increase PK variability compared with body weight adjusted dosing.

IFN‐inducible mRNA is a direct consequence of increased expression of IFN‐alpha proteins, and correlated with disease activity and certain clinical manifestations of SLE. Since it is difficult to measure the target, IFN‐alpha, in the serum of SLE patients, a sensitive IFN‐inducible gene signature based on 21 genes was used as the PD marker to follow sifalimumab inhibition of the target in SLE clinical trials of sifalimumab 8. In agreement with the results of the phase Ib population PK analyses 13, it was found in the current population PK analysis that patients with a higher BGENE21 value had slightly higher CL. Higher BGENE21score is associated with higher expression level of IFN‐alpha, higher level of SLE‐associated auto‐antibody and higher disease activity in SLE patients 30. The basis for the slightly higher CL of sifalimumab in patients with higher BGENE21 and higher IFN level is unclear. Consideration of BGENE21 as a covariate on CL explained only 4% BSV, indicating that the impact of IFN level or disease activity on elimination of sifalimumab is minimal.

Although treatment with long term corticosteroids significantly increases risk of undesirable effects in lupus patients, corticosteroids (e.g. methylprednisolone) are still clinically used in about 80% of lupus patients to control the disease activity of SLE, reduce inflammation and lessen the pain associated with SLE disease, including muscle and joint pain and stiffness 31, 32. As shown in Figure 1, patients who received steroid therapy had 11% higher sifalimumab CL compared with patients not on steroids. This is also consistent with results from a phase Ib population PK analysis where a 19% increase in CL was seen in patients with steroid use. The current population PK analysis also identified a 9% lower V 1 value in patients receiving steroids at baseline and a ≤ 6% change in V 1 over the dose range of 200–1200 mg. It is unclear if these findings are real and supported by biology. Overall, incorporating all of the four statistically significant covariates into the final population model only explained up to 9% of BSV in CL and V 1.

In conclusion, the results of the current population PK analysis demonstrated that PK characteristics of sifalimumab are similar following a body weight normalized or a fixed dosing regimen, and that the fixed dosing regimen does not increase variability in PK in SLE patients. Although clinical factors e.g. dose, body weight, baseline signature gene and baseline steroid use were found to have a statistically significant effect on sifalimumab PK parameters, the magnitude of effect on PK is limited and is not expected to be clinical relevant. These findings are consistent with the result of a previous population analysis, and support fixed dose strategy for future clinical development of sifalimumab.

PI statement

The Principal Investigator of the sifalimumab trial will author the primary manuscript detailing clinical efficacy and safety data (manuscript in preparation) and did not participate in the pharmacokinetic study design, data analysis and interpretation. The authors of the current manuscript were involved in the pharmacokinetic study design, data analysis and interpretation of the data during the sifalimumab trial.

Competing Interests

All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author). The authors have nothing to disclose. This study was sponsored by MedImmune, LLC, which was involved in the study design, data collection, analysis and interpretation, writing of the manuscript and the decision to submit the manuscript for publication. Bo Zheng, Xiang‐Qing Yu, Warren Greth and Gabriel J. Robbie are employees of MedImmune. Xiang‐Qing Yu, Warren Greth and Gabriel J. Robbie own stock in AstraZeneca.

Notes

Zheng B., Yu X.‐Q., Greth W., and Robbie G. J. (2016) Population pharmacokinetic analysis of sifalimumab from a clinical phase IIb trial in systemic lupus erythematosus patients. Br J Clin Pharmacol, 81: 918–928. doi: 10.1111/bcp.12864.

References

1. Cooper GS, Dooley MA, Treadwell EL, St Clair EW, Parks CG, Gilkeson GS. Hormonal, environmental, and infectious risk factors for developing systemic lupus erythematosus. Arthritis Rheum 1998; 41: 1714–24. [PubMed]
2. Lahita RG. The role of sex hormones in systemic lupus erythematosus. Curr Opin Rheumatol 1999; 11: 352–6. [PubMed]
3. Niewold TB. Interferon alpha as a primary pathogenic factor in human lupus. J Interferon Cytokine Res 2011; 31: 887–92. [PubMed]
4. Obermoser G, Pascual V. The interferon‐alpha signature of systemic lupus erythematosus. Lupus 2010; 19: 1012–9. [PubMed]
5. Giraldoa WAS, Villanuevaa MG, Boteanua AL, Iglesiasa AL, Mendozaa A. New therapeutic targets in systemic lupus. Reumatol Clin 2012; 08: 201–7. [PubMed]
6. Choi J, Kim ST, Craft J. The pathogenesis of systemic lupus erythematosus – an update. Curr Opin Immunol 2012; 24: 651–7. [PubMed]
7. Lo MS, Tsokos GC. Treatment of systemic lupus erythematosus: new advances in targeted therapy. Ann N Y Acad Sci 2012; 1247: 138–52. [PubMed]
8. Yao Y, Higgs BW, Richman L, White B, Jallal B. Use of type I interferon‐inducible mRNAs as pharmacodynamic markers and potential diagnostic markers in trials with sifalimumab, an anti‐IFNα antibody, in systemic lupus erythematosus. Arthritis Res Ther 2010; 12 (Suppl 1): S6. [PubMed]
9. Yao Y, Richman L, Higgs BW, Morehouse CA, de los Reyes M, Brohawn P, Zhang J, White B, Coyle AJ, Kiener PA, Jallal B. Neutralization of IFN‐α/β‐ inducible genes and downstream effect in a phase I trial of an anti‐IFN‐α monoclonal antibody in SLE. Arthritis Rheum 2009; 60: 1785–96. [PubMed]
10. Merrill JT, Wallace DJ, Petri M, Kirou KA, Yao Y, White WI, Robbie G, Levin R, Berney SM, Chindalore V, Olsen N, Richman L, Le C, Jallal B, White B. Safety profile and clinical activity of sifalimumab, a fully human anti‐interferon α monoclonal antibody, in systemic lupus erythematosus: a phase I, multicentre, double‐blind randomised study. Ann Rheum Dis 2011; 70: 1905–13. [PubMed]
11. Petri M, Wallace DJ, Spindler A, Chindalore V, Kalunian K, Mysler E, Neuwelt CM, Robbie G, White WI, Higgs BW, Yao Y, Wang L, Ethgen D, Greth W. Sifalimumab, a human anti‐interferon‐α monoclonal antibody, in systemic lupus erythematosus: a phase I randomized, controlled, dose‐escalation study. Arthritis Rheum 2013; 65: 1011–21. [PubMed]
12. Higgs BW, Zhu W, Morehouse C, White WI, Brohawn P, Guo X, Rebelatto M, Le C, Amato A, Fiorentino D, Greenberg SA, Drappa J, Richman L, Greth W, Jallal B, Yao Y. A phase 1b clinical trial evaluating sifalimumab, an anti‐IFN‐α monoclonal antibody, shows target neutralisation of a type I IFN signature in blood of dermatomyositis and polymyositis patients. Ann Rheum Dis 2014; 73: 256–62. [PubMed]
13. Narwal R, Roskos LK, Robbie GJ. Population pharmacokinetics of sifalimumab, an investigational anti‐interferon‐a monoclonal antibody, in systemic lupus erythematosus. Clin Pharmacokinet 2013; 52: 1017–27. [PubMed]
14. Beal S, Sheiner LB, Boeckmann A, Bauer RJ. NONMEM user's guides, 1989–2013. Ellicott City: Icon Development Solutions; 2013.
15. R Foundation for Statistical Computing. 2009. R: a language and environment for statistical computing.
16. Bonate P. Pharmacokinetic‐pharmacodynamic modeling and simulation. New York: Springer; 2006.
17. Tabrizi M, Wang B, Lu H, Huang S, Bell G, Schwab G, Roskos L. Population pharmacokinetic evaluation of a fully human IgG monoclonal antibody in patients with inflammatory diseases. Inflamm Allergy Drug Targets 2010; 9: 229–37. [PubMed]
18. Singh D, Kane B, Molfino NA, Faggioni R, Roskos L, Woodcock A. A phase 1 study evaluating the pharmacokinetics, safety and tolerability of repeat dosing with a human IL‐13 antibody (CAT‐354) in subjects with asthma. BMC Pulm Med 2010; 10: 3. [PubMed]
19. Oh CK, Faggioni R, Jin F, Roskos LK, Wang B, Birrell C, Wilson R, Molfino NA. An open‐label, single‐dose bioavailability study of the pharmacokinetics of CAT‐354 after subcutaneous and intravenous administration in healthy males. Br J Clin Pharmacol 2010; 69: 645–55. [PubMed]
20. Lu JF, Bruno R, Eppler S, Novotny W, Lum B, Gaudreault J. Clinical pharmacokinetics of bevacizumab in patients with solid tumors. Cancer Chemother Pharmacol 2008; 62: 779–86. [PubMed]
21. Ng CM, Bruno R, Combs D, Davies B. Population pharmacokinetics of rituximab (anti‐CD20 monoclonal antibody) in rheumatoid arthritis patients during a phase II clinical trial. J Clin Pharmacol 2005; 45: 792–801. [PubMed]
22. Dirks NL, Nolting A, Kovar A, Meibohm B. Population pharmacokinetics of cetuximab in patients with squamous cell carcinoma of the head and neck. J Clin Pharmacol 2008; 48: 267–78. [PubMed]
23. Mould DR, Holford NH, Schellens JH, Beijnen JH, Hutson PR, Rosing H, ten Bokkel Huinink WW, Rowinsky EK, Schiller JH, Russo M, Ross G. Population pharmacokinetic and adverse event analysis of topotecan in patients with solid tumors. Clin Pharmacol Ther 2002; 71: 334–48. [PubMed]
24. White B, Leon F, White W. Two first‐in‐human, open‐label, phase I dose‐escalation safety trials of MEDI‐528, a monoclonal antibody against interleukin‐9, in healthy adult volunteers. Clin Ther 2009; 31: 728–40. [PubMed]
25. Roopenian DC, Akilesh S. FcRn: the neonatal Fc receptor comes of age. Nat Rev Immunol 2007; 7: 715–25. [PubMed]
26. Rodriguez W, Selen A, Avant D, Chaurasia C, Crescenzi T, Gieser G, Di J, Huang SM, Lee P, Mathis L, Murphy D, Murphy S, Roberts R, Sachs HC, Suarez S, Tandon V, Uppoor RS. Improving pediatric dosing through pediatric initiatives: what we have learned. Pediatrics 2008; 121: 530–9. [PubMed]
27. Wang DD, Zhang S, Zhao H, Men AY, Parivar K. Fixed dosing versus body size‐based dosing of monoclonal antibodies in adult clinical trials. J Clin Pharmacol 2009; 49: 1012–24. [PubMed]
28. Zhang S, Shi R, Li C, Parivar K, Wang DD. Fixed dosing versus body size‐based dosing of therapeutic peptides and proteins in adults. J Clin Pharmacol 2012; 52: 18–28. [PubMed]
29. Bai S, Jorga K, Xin Y, Jin D, Zheng Y, Damico‐Beyer LA, Gupta M, Tang M, Allison DE, Lu D, Zhang Y, Joshi A, Dresser MJ. A guide to rational dosing of monoclonal antibodies. Clin Pharmacokinet 2012; 51: 119–35. [PubMed]
30. Dall'era MC, Cardarelli PM, Preston BT, Witte A, Davis JC Jr Type I interferon correlates with clinical and serologic manifestations of systemic lupus erythematosus. Ann Rheum Dis 2005; 64: 1692–7. [PubMed]
31. Merritt RJ, Hack SL, Kalsch M, Olson D. Corticosteroid therapy‐induced obesity in children. Clin Pediatr (Phila) 1986; 25: 149–52. [PubMed]
32. Oeser A, Chung CP, Asanuma Y, Avalos I, Stein CM. Obesity is an independent contributor to functional capacity and inflammation in systemic lupus erythematosus. Arthritis Rheum 2005; 52: 3651–9. [PubMed]

Articles from British Journal of Clinical Pharmacology are provided here courtesy of British Pharmacological Society