Advertisement

Population Pharmacokinetics and Exposure–Response Analyses of Ustekinumab in Patients With Moderately to Severely Active Crohn's Disease

Open AccessPublished:September 21, 2022DOI:https://doi.org/10.1016/j.clinthera.2022.08.010

      Highlights

      • Ustekinumab population PK and E-R relationships in Crohn's disease (CD) are reported.
      • Ustekinumab PK was well described by a 2-compartment model.
      • Ustekinumab disposition is similar between CD and ulcerative colitis populations.
      • No covariates (eg, albumin, body weight) had clinically meaningful impact on exposure.
      • E-R models support recommended adult ustekinumab posology in Crohn's disease.

      ABSTRACT

      Purpose

      Ustekinumab, a fully human immunoglobulin G1κ monoclonal antibody that antagonizes human interleukin-12/23p40, is an effective therapy for several immune-mediated inflammatory diseases, including Crohn's disease (CD). This work characterizes the population pharmacokinetic (PK) and exposure–response (E-R) relationships of ustekinumab in patients with CD using data from four Phase IIb/III clinical studies.

      Methods

      Serum ustekinumab concentration–time data from 1673 patients after IV and/or SC administration of ustekinumab were fitted simultaneously using nonlinear mixed effects modeling to develop a population PK model, which was subsequently used to evaluate simulation scenarios. Logistic regression E-R models were used to assess relationships between serum ustekinumab concentrations and clinical remission after induction (n = 1910) and maintenance (n = 387) treatment.

      Findings

      Ustekinumab PK properties are well described by a two-compartment model with first-order absorption and elimination. Typical values of PK parameters for a 70-kg patient were: clearance, 0.192 L/d; volume of distribution at steady state, 4.62 L; and intercompartmental clearance, 0.287 L/d. Ustekinumab terminal elimination t1/2 was 19 days, and bioavailability after SC administration was 78.3%. Ustekinumab clearance was not affected by coadministration of immunosuppressive agents or corticosteroids. Body weight, serum albumin, and C-reactive protein (CRP) concentrations, tumor necrosis factor (TNF) antagonist failure status, sex, race (Asian vs non-Asian), and anti-ustekinumab antibody status significantly affected ustekinumab disposition; however, the effects of these covariates on ustekinumab exposure were not clinically relevant. The population PK model predicts that a milligram/kilogram dosing approach will result in lower ustekinumab exposure in patients with lower body weight. A positive E-R relationship was established between ustekinumab concentration and efficacy outcomes. The treatment effect of ustekinumab after induction therapy was more pronounced among patients with higher baseline CRP concentrations relative to those with lower values.

      Implications

      In patients with CD, ustekinumab disposition after IV and SC administration was biexponential and consistent with those in patients with ulcerative colitis. Prior treatment with TNF antagonists or the concomitant use of immunosuppressive agents or corticosteroids had no effect on ustekinumab disposition. None of the covariates that affected ustekinumab clearance had a clinically meaningful impact on ustekinumab exposure. E-R models support recommended posology of ustekinumab in adults with CD; however, an ∼6 mg/kg IV induction dose in pediatric patients with lower body weights may not provide exposure that matches that in adult patients. ClinicalTrials.gov identifiers: NCT00771667, NCT01369329, NCT01369342, and NCT01369355.

      Keywords

      Abbreviations:

      %CV (percent coefficient of variation), AZA (azathioprine), CD (Crohn's disease), CDAI (Crohn's disease activity index), CI (confidence interval), CL (clearance), CRP (C-reactive protein), ECLIA (electrochemiluminescence immunoassay), Emax (maximum drug effect), E-R (exposure-response), F (absolute subcutaneous bioavailability), IBD (inflammatory bowel disease), IIV (interindividual variability), IL (interleukin), IOV (interoccasion variability), IV (intravenous(ly)), Ka (first-order absorption rate constant), MSD (Meso Scale Discovery), OFV (objective function value), PK (pharmacokinetics), Q (intercompartmental clearance), q4/8/12w (every 4/8/12 weeks), SC (subcutaneous(ly)), TNF (tumor necrosis factor), UC (ulcerative colitis), Vc (central volume of distribution), Vp (peripheral volume of distribution), Vss (volume of distribution at steady state), WAM (Wald's approximation method)

      Introduction

      Ustekinumab is a fully human immunoglobulin G1κ monoclonal antibody that blocks human interleukin (IL)-12 and IL-23 cytokines through high affinity binding to the p40 subunit common to both. Subsequently, binding of IL-12 and IL-23 to the shared IL-12Rβ1 chain of their respective receptors is inhibited with reduced intracellular signaling. Abnormal regulation of IL-12 and IL-23 is associated with multiple immune-mediated diseases, including inflammatory bowel disease (IBD), and inhibiting IL-12 and IL-23 is effective therapy for Crohn's disease (CD), ulcerative colitis (UC), psoriasis, and psoriatic arthritis.
      Stelara
      INN-Ustekinumab Prescribing Information.
      The pharmacokinetics (PK) of ustekinumab has been previously characterized. Ustekinumab exhibited PK characteristics typical for an immunoglobulin G monoclonal antibody, with no apparent target-mediated disposition. Several factors, including body weight, diabetes comorbidity, the presence of antidrug antibodies, and serum albumin levels, have been associated with the disposition of ustekinumab when used in the treatment of immune-mediated inflammatory disease.
      • Zhu Y.
      • Hu C
      • Lu M
      • Liao S
      • Marini JC
      • Yohrling J
      • et al.
      Population pharmacokinetic modeling of ustekinumab, a human monoclonal antibody targeting IL-12/23p40, in patients with moderate to severe plaque psoriasis.
      ,
      • Wang Z
      • Verstockt B
      • Sabino J
      • Vermeire S
      • Ferrante M
      • Declerck P
      • et al.
      Population pharmacokinetic-pharmacodynamic model-based exploration of alternative ustekinumab dosage regimens for patients with Crohn's disease.
      In patients with CD, median serum ustekinumab concentrations after a single IV administration of ustekinumab ranging from 1 to 6 mg/kg were approximately dose proportional and detectable at all sampling time points through 8 weeks after administration.

      European Medicines Agency. Assessment report: Stelara. 15 September 2016. Accessed April 19, 2022. https://www.ema.europa.eu/en/documents/variation-report/stelara-h-c-000958-x-0049-g-epar-assessment-report-extension_en.pdf.

      During maintenance treatment with ustekinumab 90 mg SC every 8 (q8w) or 12 weeks (q12w), steady state was reached at ∼8 or 12 weeks, respectively, following initiation of maintenance therapy.
      • Adedokun OJ
      • Xu Z
      • Gasink C
      • Jacobstein D
      • Szapary P
      • Johanns J
      • et al.
      Pharmacokinetics and exposure response relationships of ustekinumab in patients with Crohn's disease.
      Based on trend analysis, a higher serum ustekinumab concentration was associated with greater efficacy, as measured by both clinical symptoms and endoscopy.
      • Adedokun OJ
      • Xu Z
      • Gasink C
      • Jacobstein D
      • Szapary P
      • Johanns J
      • et al.
      Pharmacokinetics and exposure response relationships of ustekinumab in patients with Crohn's disease.
      In addition, the serum drug concentration demonstrated a significant inverse correlation with C-reactive protein (CRP) concentrations. No association was observed with concomitant use of immunosuppressive agents (6-mercaptopurine, azathioprine [AZA], or methotrexate). Greater proportions of patients with ustekinumab concentrations between 0.8 and 1.4 µg/mL or greater achieved or maintained clinical remission compared with patients with lower trough concentrations.
      • Adedokun OJ
      • Xu Z
      • Gasink C
      • Jacobstein D
      • Szapary P
      • Johanns J
      • et al.
      Pharmacokinetics and exposure response relationships of ustekinumab in patients with Crohn's disease.
      These and other emerging data showing associations between ustekinumab concentration and efficacy outcomes have led to interest in ustekinumab dose optimization and therapeutic drug-monitoring studies in patients with CD. A prerequisite for these types of “personalized medicine” studies is the availability of a predictive population PK model based on patient and disease characteristics in patients with CD. The predictive population PK model can then be used to simulate the concentration–time profiles of ustekinumab under different dose strategies. In addition, a model linking concentration and efficacy in CD may help identify factors that influence differential response in CD and determine whether a different dose regimen is warranted in a given patient subpopulation to account for such factors.
      Although population PK analyses have been reported for ustekinumab in patients with psoriasis,
      • Zhu Y.
      • Hu C
      • Lu M
      • Liao S
      • Marini JC
      • Yohrling J
      • et al.
      Population pharmacokinetic modeling of ustekinumab, a human monoclonal antibody targeting IL-12/23p40, in patients with moderate to severe plaque psoriasis.
      psoriatic arthritis,
      • Zhu YW
      • Mendelsohn A
      • Pendley C
      • Davis HM
      • Zhou H.
      Population pharmacokinetics of ustekinumab in patients with active psoriatic arthritis.
      and UC,
      • Xu Y
      • Hu C
      • Chen Y
      • Miao X
      • Adedokun OJ
      • Xu Z
      • et al.
      Population pharmacokinetics and exposure-response modeling analyses of ustekinumab in adults with moderately to severely active ulcerative colitis.
      a predictive population PK model in patients with CD is lacking. Notably, the posology of ustekinumab is different in IBD (CD or UC) indications compared with psoriatic disease indications. Specifically, the approved ustekinumab dose regimen in patients with psoriasis or psoriatic arthritis is 45 mg SC at weeks 0 and 4 and then q12w for patients weighing ≤100 kg and 90 mg SC at weeks 0 and 4 and then q12w for patients weighing >100 kg. For patients with CD or UC, the approved dose regimen is an initial body weight–based tiered IV induction dose approximating 6 mg/kg (ie, 260, 390, or 520 mg in patients weighing ≤55 kg, >55 and ≤85 kg, or >85 kg, respectively), referred to hereafter as ∼6 mg/kg. The single IV induction dose is followed by a SC 90-mg maintenance injection 8 weeks later and q8w thereafter (or q12w in some jurisdictions). To characterize the PK variables of ustekinumab after IV and SC administration and to evaluate the impact of CD-related covariates on the ustekinumab PK profile, it was desirable to develop a population PK model in CD that could subsequently be used to support exposure–response (E-R) modeling in this indication.
      Herein we present a population PK model of ustekinumab in CD using data from the pivotal ustekinumab studies in adults with moderately to severely active CD. The results quantitatively describe the important covariates that affect ustekinumab disposition in this patient population. The population PK model was also used to perform simulations to evaluate alternative ustekinumab dose regimens in other CD studies, including pediatric CD. In addition, landmark E-R models linking ustekinumab concentration and efficacy while exploring factors affecting the E-R relationship are presented.

      Participants and Methods

      Clinical Studies

      Data from four Phase IIb/III ustekinumab clinical studies in adult patients with CD were used for the population PK analysis. E-R analyses were based on the Phase III studies. The Institutional Review Board or ethics committee at each study site approved the protocols, and all patients provided written informed consent. Details regarding study designs and patient populations have been published and are summarized in Table I.
      • Sandborn WJ
      • Gasink C
      • Gao L-L
      • Blank MA
      • Johanns J
      • Guzzo C
      • et al.
      Ustekinumab induction and maintenance therapy in refractory Crohn's disease.
      ,
      • Feagan BG
      • Sandborn WJ
      • Gasink C
      • Jacobstein D
      • Lang Y
      • Friedman JR
      • et al.
      Ustekinumab as induction and maintenance therapy for Crohn's disease.
      Table IDescriptive summary of studies.
      Study (No. Treated)DescriptionTreatmentPK Sampling
      C0743T26 (N = 526)Phase IIb, randomized, double-blind, placebo-controlled study in patients with moderate to severe CD who failed or were intolerant to TNF antagonist therapy (NCT00771667)IV: 1 mg/kg, 3 mg/kg, and 6 mg/kg ustekinumab or placebo at week 0

      SC: 90 mg ustekinumab or placebo at weeks 8 and 16; 270 mg ustekinumab at week 8
      Patients who did not respond to IV placebo induction therapy received SC ustekinumab 270 mg at week 8 and 90 mg at week 16.
      Weeks 0, 4, 6, 8, 12, 16, 20, 22, 28, and 36
      Week 0 samples were obtained preadministration and 1 hour postinfusion.
      CNTO1275CRD3001 (N = 740)Phase III, randomized, double-blind, placebo-controlled study of ustekinumab induction in patients with moderate to severe CD who failed or were intolerant to TNF antagonist therapy (NCT01369329)IV: Placebo or 130 mg or ∼6 mg/kg ustekinumab at week 0Weeks 0, 3, 6, and 8
      Week 0 samples were obtained preadministration and 1 hour postinfusion.
      CNTO1275CRD3002 (N = 627)Phase III, randomized, double-blind, placebo-controlled study of ustekinumab induction in patients with moderate to severe CD with elevated CRP or calprotectin or endoscopic evidence of disease, who have failed or are intolerant to corticosteroids, 6-MP, AZA, or MTX, but did not fail and were not intolerant to TNF antagonist therapy (NCT01369342)IV: Placebo or 130 mg or ∼6 mg/kg ustekinumab at week 0Weeks 0, 3, 6, and 8
      Week 0 samples were obtained preadministration and 1 hour postinfusion.
      CNTO1275CRD3003 (N = 1280)Phase III, randomized, double-blind, placebo-controlled study of ustekinumab maintenance therapy in patients with moderate to severe CD from CNTO1275CRD3001 and CNTO1275CRD3002 induction studies who were in clinical response to ustekinumab at induction week 8 or were placebo nonresponders (NCT01369355)SC: Placebo or 90 mg ustekinumab q12w or q8w

      IV: 130 mg ustekinumab at week 0
      Patients who were nonresponders to placebo during CNTO1275CRD3001 or CNTO1275CRD3002 received IV ustekinumab 130 mg at week 0 of the maintenance study.
      Weeks 0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, and 44
      Week 0 samples were obtained preadministration and 1 hour postinfusion.
      ,
      Sample at week 0 of the maintenance study is the same as sample at week 8 of the induction study. Each sampling time for a patient in the maintenance study corresponds to 8 weeks after the start of induction study (eg, week 44 in the maintenance study corresponds to a patient's week 52 sample).
      6-MP = 6-mercaptopurine; AZA = azathioprine; CD = Crohn's disease; CRP = C-reactive protein; MTX = methotrexate; N = patients treated; PK = pharmacokinetic; q8w = administration every 8 weeks; q12w = administration every 12 weeks; TNF = tumor necrosis factor.
      low asterisk Patients who did not respond to IV placebo induction therapy received SC ustekinumab 270 mg at week 8 and 90 mg at week 16.
      Week 0 samples were obtained preadministration and 1 hour postinfusion.
      Patients who were nonresponders to placebo during CNTO1275CRD3001 or CNTO1275CRD3002 received IV ustekinumab 130 mg at week 0 of the maintenance study.
      § Sample at week 0 of the maintenance study is the same as sample at week 8 of the induction study. Each sampling time for a patient in the maintenance study corresponds to 8 weeks after the start of induction study (eg, week 44 in the maintenance study corresponds to a patient's week 52 sample).
      Briefly, C0743T26 was a Phase IIb study of ustekinumab in patients with CD who failed tumor necrosis factor (TNF) antagonist therapy.
      • Sandborn WJ
      • Gasink C
      • Gao L-L
      • Blank MA
      • Johanns J
      • Guzzo C
      • et al.
      Ustekinumab induction and maintenance therapy in refractory Crohn's disease.
      Patients received IV ustekinumab or placebo at week 0 and/or SC ustekinumab or placebo at weeks 8 and 16. CNTO1275CRD3001 and CNTO1275CRD3002 were Phase III studies of IV ustekinumab or placebo induction in patients with CD who had failed TNF antagonist therapy or had failed conventional therapy but were TNF antagonist naive, respectively.
      • Feagan BG
      • Sandborn WJ
      • Gasink C
      • Jacobstein D
      • Lang Y
      • Friedman JR
      • et al.
      Ustekinumab as induction and maintenance therapy for Crohn's disease.
      CNTO1275CRD3003 was a Phase III maintenance study of SC ustekinumab in patients with CD who were enrolled in either of the Phase III induction studies.
      • Feagan BG
      • Sandborn WJ
      • Gasink C
      • Jacobstein D
      • Lang Y
      • Friedman JR
      • et al.
      Ustekinumab as induction and maintenance therapy for Crohn's disease.
      Patients in this study received SC ustekinumab or placebo (q12w or q8w); nonresponders to placebo in the induction studies received IV ustekinumab at week 0.

      Bioanalytical Assessments

      In all 4 studies included in the population PK analysis, serum ustekinumab concentrations were determined by using a validated electrochemiluminescence immunoassay (ECLIA) method on the Meso Scale Discovery (MSD) platform (Gaithersburg, MD, USA). The lowest quantifiable concentration in a sample for the MSD ECLIA method was 0.1688 µg/mL with a calibration range from 0.1688 to 10.8 µg/mL.
      In C0743T26, the presence of antibodies to ustekinumab was determined by using a validated bridging enzyme immunoassay, in which ustekinumab was used to capture, detect, and characterize induced immune responses to ustekinumab. Because the presence of ustekinumab could interfere with the detection of antibodies with this assay, all samples evaluated for antibodies to ustekinumab were also analyzed for ustekinumab concentration. In the three Phase III studies, analyses of antibodies to ustekinumab were performed by using a validated, drug-tolerant, ECLIA method on the MSD platform. This assay incorporates an acid dissociation treatment step to reduce ustekinumab interference when detecting antibodies to ustekinumab in serum samples. With this assay, antidrug antibodies at concentrations of 50 ng/mL were detectable in the presence of up to 100 µg/mL ustekinumab.
      Across all studies included in the population PK analysis, a patient was considered positive for antibodies if treatment-emergent antibodies to ustekinumab were detected in the patient's sample at any time, regardless of the presence or absence of ustekinumab in the sample.

      Population PK Modeling

      All patients who received at least 1 dose of ustekinumab and had at least 1 measurable concentration of ustekinumab in any of the studies were included in the population PK analysis. Serum ustekinumab concentration–time data were fitted simultaneously by using nonlinear mixed effect modeling as implemented in NONMEM 7.3 (ICON Development Solutions, Ellicott City, MD, USA). The stochastic approximation expectation-maximization method in NONMEM was used to estimate the population PK parameters. Postprocessing of data was performed by using SAS 9.3 (SAS Institute, Inc, Cary, NC, USA), S-Plus 8.2 (TIBCO Software, Inc., Boston, MA, USA), or R version 3.1.2 (R Foundation for Statistical Computing, Vienna, Austria). Graphics were generated by using S-Plus or R.
      The population PK analysis proceeded in 3 stages: the development of a base model, a full covariate model, and a final model. For the base model, an appropriate structural PK model, interindividual variability (IIV) model, and residual error model were first determined. Previous population PK analyses of ustekinumab used a one-compartment structural PK model to describe the serum ustekinumab concentration data after SC administration. With the inclusion of PK data from IV administration of ustekinumab in the CD studies, it was anticipated that the PK profile of ustekinumab might be better described by a two-compartment model based on prior experience with IV administered ustekinumab (Figure 1). Both one- and two-compartment structural PK models were tested to confirm that available data supported this assumption.
      Figure 1
      Figure 1Two-compartment pharmacokinetic model after IV and SC ustekinumab administration. CL = clearance; F = absolute SC bioavailability; Ka = first-order absorption rate constant; Q = intercompartmental clearance; Vc = central volume of distribution; Vp = peripheral volume of distribution.
      The IIV of the two-compartment PK disposition parameters (clearance [CL], central volume of distribution [Vc], intercompartmental clearance [Q], and peripheral volume of distribution [Vp]) and the first-order absorption rate constant (Ka) were modeled as follows (using CL as an example):
      CLi=exp(θCL+ηCL,i)
      (1)


      where CLi represents the patient-specific CL, θCL is the log-transformed typical value of the parameter, and ηCL,i is a random effect that denotes the difference between patient i in the population and the typical individual. The log transformation (ie, estimating θCL) allows for unconstrained optimization and is consistent with MU parameterization recommended in the NONMEM user manual.
      The IIV for F, the absolute SC bioavailability, was modeled using the logistic function constraining Fi such that 0< Fi <1:
      Fi=exp(θF+ηFi)1+exp(θF+ηFi)
      (2)


      The vector of random effects ηi was assumed to be distributed as a multivariate normal with a mean vector of 0 and variance-covariance matrix of Ω [ie,ηiN(0,)][ie,ηiN(0,Ω)]. The diagonal elements of Ω are variances and denoted, for example, as ωCL2, the variance of CL. The off-diagonal elements of Ω are covariances and are represented, for example, as ωCL,Vc (the covariance between CL and Vc). The approximate %CV is reported for CL, Vc, Q, Vp, and Ka as, for example, %CV(CL)=ω2CL×100. The SD of F on the logit scale is reported.
      Interoccasion variability (IOV) represents another level of random effects and characterizes random fluctuation in model parameters within an individual at different designated occasions. IOV is useful when there is a high degree of residual error due to lack of reproducibility of a PK profile from 1 occasion to the next within an individual patient. The estimate of the residual (within-patient) variability was small enough that IOV characterization was not considered warranted in the present analysis.
      The residual error, the difference between the observed and individual model-predicted concentration, was modeled by using combined additive and proportional error terms:
      Cpij=Cp^ij×(1+ε1ij)+ε2ij
      (3)


      where Cpijis the observed and Cp^ij is the predicted ustekinumab concentration for patient i at time j. The residual random errors, ε1ij and ε2ij, represent the proportional and additive components, respectively, and are assumed to be independent and normally distributed [ie, ε1ijN(0,σ12) and ε2ijN(0,σ22)]. The %CV was calculated for the proportional component as %CV=σ12×100, which dominates the overall residual error at large concentrations. The residual error near the limit of quantification was dominated by the additive term, which was reported as SD=σ22.
      The final base model selection was based on the objective function value, successful minimization and successful estimation of the covariance matrix of the parameter estimates, and reduction in the residual error. In addition to these statistical considerations, model selection was aided by goodness-of-fit or diagnostic plots.
      After the development of the base model, a full model was constructed that included several covariates selected for evaluation based on prior knowledge of ustekinumab and CD. The continuous covariates examined included body weight, age, serum albumin, CRP, alkaline phosphatase, white blood cell count, fecal calprotectin, fecal lactoferrin, and CD activity index (CDAI) score. The discrete covariates examined included immune response (antibodies to ustekinumab and positive vs negative status), TNF antagonist failure status, TNF antagonist failure versus conventional therapy failure, sex, White versus non-White, concomitant AZA or 6-mercaptopurine use, concomitant methotrexate use, corticosteroid use, and smoking status (current smoker or history of smoking). All covariates were assessed by using baseline values except for immune response, which was evaluated as a patient-specific assignment of positive antibody status if the assay was positive at any postbaseline visit and negative otherwise and as a time-varying covariate based on antibody status at each visit.
      Continuous covariates were included in the full model as follows, using CL as an example:
      CLTV=exp(θCL)(XXREF)θCL,X=exp(θCL+θCL,X[lnXlnXREF])
      (4)


      where CLTV is the typical value of the parameter (at η = 0), X is the continuous covariate, XREF represents a reference value of the covariate (approximate median), θCL denotes the typical value of the log-transformed CL for a subject at the reference value (X = XREF), and θCL,X denotes the effect of the covariate. Discrete covariates were included using the form:
      CLTV=exp(θCL+θCL,X.X)
      (5)


      where X = 1 when the covariate effect is present. θCL denotes the typical value of the log-transformed CL for a subject at the reference value (X = 0). Covariates with multiple categories were expanded using indicator variables to accommodate the effects. For example, 3 indicator variables were created to represent Black (RACE = B), Asian (RACE = A), and other (RACE = O) race categories.
      The quantity expCL,X . X) represents the fractional change in the parameter when the covariate effect is present. This parameterization facilitates MU processing in NONMEM and ensures appropriate CIs, unlike the parameterization (1 + θCL,X . X), which could yield negative CL CI values, for example, depending on the estimate and uncertainty in θCL,X.
      The final population PK model was obtained by applying the Wald approximation method to a full model
      • Kowalski KG
      • Hutmacher MM.
      Efficient screening of covariates in population models using Wald's approximation to the likelihood ratio test.
      and using the Schwarz/Bayesian criteria to select among the top 15 ranked models. The final model was then applied to perform simulations that determined ustekinumab exposures in different clinical studies of interest in CD. Of note, simulations of ustekinumab exposure in pediatric populations were performed with the assumption that age has no significant impact on ustekinumab PK variables once body size is taken into account.

      E-R Modeling

      A logistic regression modeling approach using serum ustekinumab concentrations at specified time points (exposure variable) and clinical remission (response variable) was used to investigate the E-R relationship of ustekinumab in patients with CD during induction and maintenance therapy with ustekinumab. Clinical remission, derived from the CDAI score,
      • Sandborn WJ
      • Feagan BG
      • Hanauer SB
      • Lochs H
      • Lӧfberg R
      • Modigliani R
      • et al.
      A review of activity indices and efficacy endpoints for clinical trials of medical therapy in adults with Crohn's disease.
      which represents the highest bar of efficacy among the prespecified end points, was selected as the efficacy outcome in these analyses because the observed data suggested a clearer E-R trend with these outcomes compared with the clinical response end point.
      In these E-R models, the drug effect for ustekinumab was modeled by using an Emax function; the logit probability of remission was assumed to be the result of additive drug and placebo (β0) effects as follows:
      logit(Pr(Ri=1))=β0+Emax×C0iEC50+C0i
      (6)


      where Ri denotes an indicator variable set to 1 if the ith patient is in remission and 0, otherwise; the probabilities that R = 1 are modeled on the logit scale; that is, logit(p) = log (p) − log (1 − p), β0 is the logit probability for remission (R = 1) for placebo, C0i is the individual predicted trough ustekinumab concentration at the specified time point, Emax is the maximum drug effect on the logit scale, and EC50 is the ustekinumab concentration corresponding to 50% of the maximum drug effect. The covariate effects for TNF antagonist failure status; use of concomitant immunosuppressive agents (AZA, 6-mercaptopurine, or methotrexate); and baseline values of body weight, CRP, and CDAI score were investigated on β0, Emax, and EC50. The covariate effects on β0 were modeled as additive effects while the covariate effects on Emax and EC50 were modeled as exponential effects to ensure that Emax and EC50 remained positive. The evaluation of these baseline covariate effects on the E-R model was expressed as follows:
      βi=β0+(θ1·FTNFi+θ2[WTi68.7]+θ3[CRPi0.93]+θ4[CDAIi307]+θ5·IMMi)
      (7)


      Emaxi=Emax0·exp(θ6·FTNFi+θ7[WTi68.7]+θ8[CRPi0.93]+θ9[CDAIi307]+θ10·IMMi)
      (8)


      EC50i=EC50_0·exp(θ11·FTNFi+θ12[WTi68.7]+θ13[CRPi0.93]+θ14[CDAIi307]+θ15·IMMi)
      (9)


      where β0, Emax0, and EC50,0 are the reference values of these parameters corresponding to the TNF antagonist nonfailure population with baseline body weight of 68.7 kg, baseline CRP of 0.93 mg/dL, baseline CDAI score of 307, and no concomitant immunosuppressive use. The goodness-of-fit for the E-R models was assessed by binning ustekinumab concentrations into quartiles and computing the predicted response rates for each concentration bin. The observed response rates for the placebo group (corresponding to ustekinumab concentration of 0.0 µg/mL) were also included in the goodness-of-fit assessment.

      Results

      Population PK Analysis

      A total of 14,125 serum ustekinumab concentrations from 1673 adult patients with CD were included in the population PK analysis (477 patients from C0743T26, 666 from CNTO1275CRD3001, and 530 from CNTO1275CRD3002). A summary of baseline values for demographic and clinical variables for patients included in the population PK analysis is provided in Table II. Overall, the majority of the patient population was female (56.6%) and White (88.2%), with a median age of 37 (range, 18–76) years, and a median body weight of 69 (range, 35–184) kg.
      Table IIDemographic and baseline characteristics of patients included in the population pharmacokinetic analysis.
      VariableC0743T26 (n = 477)CNTO1275CRD3001 (n = 666)CNTO1275CRD3002 (n = 530)Overall (N = 1673)
      Age, y
       Mean (SD)39.3 (12.8)37.4 (12.1)39.5 (13.5)38.6 (12.8)
       Median (range)38 (18, 76)36 (18, 71)37 (18, 75)37 (18, 76)
      Sex
       Female278 (58.3%)386 (58.0%)283 (53.4%)947 (56.6%)
       Male199 (41.7%)280 (42.0%)247 (46.6%)726 (43.4%)
      Race
       White443 (92.9%)580 (85.2%)452 (85.3%)1475 (88.2%)
       Black14 (2.9%)21 (3.1%)17 (3.2%)52 (3.1%)
       Asian7 (1.5%)55 (8.1%)44 (8.3%)106 (6.3%)
       Other13 (2.7%)10 (1.5%)17 (3.2%)40 (2.4%)
      Body weight, kg
       Mean (SD)73.6 (20.2)69.6 (18.4)73.2 (20.5)71.9 (19.7)
       Median (range)69.6 (36.3, 150.0)66.8 (35.0, 172.8)70.3 (35.0, 184.0)68.7 (35.0, 184.0)
      Prior or current smoker226 (47.4)325 (48.8)270 (50.9)821 (49.1)
      CDAI score
       Mean (SD)324.6 (64.9)321.1 (62.4)301.1 (57.7)315.7 (62.5)
       Median (range)316 (209, 483)316 (198, 512)289 (198, 465)307 (198, 512)
      CRP, mg/L
       Mean (SD)21.9 (27.5)18.5 (23.7)16.0 (21.0)18.7 (24.1)
       Median (range)10.0 (0.1, 199.0)9.9 (0.1, 157.0)8.1 (0.1, 137.0)9.3 (0.1, 199.0)
      Fecal calprotectin, mg/kg
       Mean (SD)1177 (2107)945 (1632)726 (973)942 (1628)
       Median (range)523 (14, 17327)495 (118, 16647)503 (12, 8519)503 (12, 17327)
      Fecal lactoferrin, µg/mL
       Mean (SD)238 (362)233 (290)201 (273)224 (307)
       Median (range)88 (0.2, 2644)92 (0.2, 1002)84 (0.4, 1002)88 (0.2, 2644)
      Alkaline phosphatase, IU/L
       Mean (SD)79.0 (35.7)76.0 (32.4)78.6 (32.3)77.7 (33.4)
       Median (range)73 (27, 422)70 (32, 418)72 (31, 415)71 (27, 422)
      Albumin, g/dL
       Mean (SD)4.0 (0.4)3.4 (0.5)3.6 (0.5)3.6 (0.5)
       Median (range)4.0 (2.8, 5.0)3.5 (1.5, 4.6)3.7 (1.2, 5.1)3.7 (1.2, 5.1)
      Platelet, × 103/µL
       Mean (SD)344.2 (124.0)371.9 (131.8)357.2 (109.2)359.4 (123.3)
       Median (range)327 (85, 1101)338 (101, 925)338 (137, 933)335 (85, 1101)
      White blood cells, × 103/µL
       Mean (SD)9.0 (2.9)8.8 (3.2)8.5 (3.2)8.8 (3.1)
       Median (range)8.6 (2.6, 21.3)8.3 (2.2, 25.3)8.0 (3.0, 22.5)8.3 (2.2, 25.3)
      Anti-ustekinumab antibodies present3 (0.6%)16 (2.4%)11 (2.1%)30 (1.8%)
      Prior TNF antagonist treatment477 (100%)666 (100%)01143 (68.3%)
      Prior TNF antagonist failure425 (89.1%)572 (85.9%)0997 (59.6%)
      Concomitant medications
       AZA or 6-MP74 (15.5%)143 (21.5%)153 (28.9%)370 (22.1%)
       MTX47 (9.9%)60 (9.0%)27 (5.1%)134 (8.0%)
       Corticosteroids202 (42.4%)257 (38.6%)157 (29.6%)616 (36.8%)
      6-MP = 6-mercaptopurine; AZA = azathioprine; CDAI = Crohn's disease activity index; CRP = C-reactive protein; MTX = methotrexate; TNF = tumor necrosis factor.
      A two-compartment structural model parameterized in terms of CL, Vc, Q, Vp, Ka, and F best described the time course of serum ustekinumab concentration in patients with CD and, hence, was chosen for the present analyses. IIV was assessed on all 6 structural PK parameters along with covariances between CL and Vc and CL and Vp. The objective function value for the two-compartment structural model was >8000 points lower than for the one-compartment model, and proportional residual error was smaller for the two-compartment model than for the one-compartment model (17% vs 24%, respectively). Following the application of the Wald approximation method covariate modeling procedure, a final population PK model was obtained. The final population PK model parameter estimates along with estimates of the statistically significant covariate effects on ustekinumab PK parameters are listed in Table III. All parameters were precisely (low relative SE) estimated, with the exception of Q and the effect of antibodies to ustekinumab on clearance. Evaluation of the final models using diagnostic plots (Figure 2) and visual predictive checks (Figure 3) showed no appreciable lack-of-fit or obvious model misspecification and good predictive performance.
      Table IIIUstekinumab population pharmacokinetic parameters and covariate effects from the final model.
      ParameterNONMEM Estimate (ASE)
      Parameter estimates as provided in NONMEM based on MU parameterization
      Transformed Estimate (%RSE)
      Parameter estimates in standard pharmacokinetic units.
      IIV, as %CV (%RSE)
      CL (L/d)–1.65 (0.0162)0.192 (8.4)29.1 (4.5)
       Effect of weight on CL0.481 (0.0384)0.481 (8.0)
       Effect of albumin on CL–0.972 (0.0549)–0.972 (5.6)
       Effect of sex on CL (male)0.161 (0.0164)0.161 (10.2)
       Effect of race on CL (Asian)0.135 (0.0303)0.135 (22.4)
       Effect of TNF antagonist failure status on CL (yes)0.109 (0.0149)0.109 (13.7)
       Effect of C-reactive protein on CL0.0705 (0.00542)0.0705 (7.7)
       Effect of anti-drug antibody status on CL (positive)0.125 (0.124)0.125 (99.2)
      Vc (L)1.01 (0.00852)2.75 (0.3)14.7 (8.9)
       Effect of weight on Vc0.402 (0.0254)0.402 (6.3)
       Effect of albumin on Vc–0.208 (0.0397)–0.208 (19.1)
       Effect of sex on Vc (male)0.155 (0.0125)0.155 (8.1)
      Q (L/d)–1.25 (0.26)0.287 (90.7)75.6 (64.3)
       Effect of weight on Q0.793 (0.289)0.793 (36.4)
      Vp (L)0.63 (0.064)1.88 (3.4)32.2 (54.2)
       Effect of weight on Vp0.705 (0.0932)0.705 (13.2)
      Absorption rate constant (1/d)–1.71 (0.0498)0.181 (27.5)72.1 (19.8)
      Absolute SC bioavailability1.28 (0.0687)0.782 (8.8)102.5 (14.8)
      Correlation between CL and Vc0.5100.510
      Correlation between CL and Vp–0.528–0.528
      Proportional residual error0.0287 (0.000861)16.9 (3.0)
      Proportional residual error in %CV.
      Additive residual error0.00558 (0.00055)0.075 (9.9)
      Additive residual error in µg/mL.
      ASE = approximate SE; CL = clearance; IIV = interindividual variability; Q = intercompartmental clearance between the central compartment and a peripheral compartment; RSE = relative SE; TNF = tumor necrosis factor; Vc = central volume of distribution; Vp = peripheral volume of distribution.
      low asterisk Parameter estimates as provided in NONMEM based on MU parameterization
      Parameter estimates in standard pharmacokinetic units.
      Proportional residual error in %CV.
      § Additive residual error in µg/mL.
      Figure 2
      Figure 2Goodness-of-fit plots for the final ustekinumab population pharmacokinetic model. Observed versus population-predicted concentrations (A), observed versus individual-predicted concentrations (B), conditional-weighted residuals versus population-predicted concentrations (C), and conditional-weighted residuals versus time after dose (D). Broken lines denote the lines of identity in panels A and B and the zero line in panels C and D.
      Figure 3
      Figure 3Prediction-corrected visual predictive checks at steady state in patients with Crohn's disease using the final population pharmacokinetic model. The solid magenta line and dashed lines represent the median and the 5th/95th percentiles of the observations. The black solid and dashed lines represent the median and the 5th/95th percentiles of model predictions. The shaded regions represent the 90% prediction interval for the simulated median and 5th/95th percentiles, respectively. TSLD = time since last dose.
      In a reference patient with CD weighing ∼70 kg, the typical estimate of ustekinumab CL was 0.192 L/d and Vc was 2.75 L. The IIV expressed in terms of the %CV was 29.1% for CL and 14.7% for Vc, with corresponding shrinkage estimates of 5.2% and 34.1%, respectively. The estimate of the volume of distribution at steady state (Vss), which is the sum of Vc and Vp, was 4.62 L, which is comparable to the volume of the vascular system (5 L in a 70-kg individual), suggesting that ustekinumab is primarily distributed within the vascular system. The estimated bioavailability of ustekinumab after SC administration was 78.3%, and the typical value of the terminal elimination t1/2 was ∼19 days.
      In the final population PK model, the effect of body weight was significant on the CL, Vc, Vp, and Q of ustekinumab. In addition, TNF antagonist failure status, immunogenicity (anti-ustekinumab antibodies), sex, race (Asian vs non-Asian), and serum albumin and CRP levels were found to be significant covariates on CL. The impact of these covariates on CL and Vc in the final model is illustrated in Supplemental Figure 1 (given in the online version at doi:10.1016/j.clinthera.2022.08.010). Baseline CDAI, concomitant use of immunosuppressive agents or corticosteroids, prior use of TNF antagonists, and smoking status had no significant impact on ustekinumab PK. In terms of ustekinumab exposure, patients with albumin at the 10th percentile (3 g/dL) or lower and patients with CRP at the 90th percentile (45.9 mg/L) were predicted to have trough concentrations less than the lower limit of the bioequivalence interval (ie, 80%–125%) around the typical steady-state trough ustekinumab concentration (see Supplemental Figure 1 in the online version at doi:10.1016/j.clinthera.2022.08.010).
      The final population PK model was used to simulate ustekinumab exposure after IV reinduction during steady state and every-4-week (q4w) SC dosing and to evaluate the pediatric dose regimen. After IV reinduction, ustekinumab exposure was comparable to that attained with the initial IV induction dose, and ustekinumab concentrations were predicted to return to steady state ∼16 weeks after the reinduction IV dose (see Supplemental Figure 2 in the online version at doi:10.1016/j.clinthera.2022.08.010). Ustekinumab trough concentrations after q4w SC dosing were predicted to be 4-fold those after the standard of care regimen (q8w) (see Supplemental Figure 3 in the online version at doi:10.1016/j.clinthera.2022.08.010). With respect to the exploration of pediatric posology, a 6 mg/kg IV induction dose was predicted to result in 50% lower exposure among pediatric patients with body weight <40 kg compared with ustekinumab exposure in the typical adult patient receiving the approved ∼6 mg/kg IV induction dose.
      • Rosh JR
      • Turner D
      • Griffiths A
      • Cohen SA
      • Jacobstein D
      • Adedokun OJ
      • et al.
      Ustekinumab in pediatric patients with moderately to severely active Crohn's disease pharmacokinetics, safety, and efficacy results from UniStar, a phase 1 study.

      E-R Modeling

      IV Induction Phase

      For the induction E-R analysis, the relationship between ustekinumab concentration and clinical remission at week 8 was assessed for patients who received induction treatment in studies C0743T26, CNTO1275CRD3001, and CNTO1275CRD3002. Of the 1910 patients who received IV induction (placebo or ustekinumab) at week 0 of the induction studies, 597 received placebo, 130 received ustekinumab 1 mg/kg, 133 received ustekinumab 3 mg/kg, 459 received ustekinumab 130 mg, and 591 received ustekinumab ∼6 mg/kg. The parameters of the induction E-R model are summarized in Table IV. TNF antagonist failure status, baseline CRP, and baseline CDAI had statistically significant effects on the induction E-R model parameters. The covariate effects in Table IV indicate that the placebo effect decreases in patients with TNF antagonist failures, and with increasing baseline CRP and baseline CDAI score. In addition, the maximum drug effect increases with increasing baseline CRP, and the EC50 decreases in patients with TNF antagonist failures and increasing baseline CDAI score. Taken together, the model suggests that, at a given ustekinumab concentration, patients who previously failed a TNF antagonist or had a higher baseline CRP or CDAI score tended to have lower clinical remission rates at week 8. The goodness-of-fit plot for the induction E-R model is shown in Figure 4A.
      Table IVParameter estimates for the induction and maintenance exposure–response models.
      ParameterEstimate (%RSE)
      Induction exposure–response model
      Placebo effects
      β0–1.600 (9.4)
       TNF antagonist failure status–0.996 (16.9)
       Baseline CRP–0.094 (53.2)
       Baseline CDAI score–0.012 (12.2)
      Emax
       Emax01.260 (14.3)
       Baseline CRP0.089 (22.6)
      EC50
       EC50,00.712 (69.1)
       TNF antagonist failure status–1.980 (44.4)
       Baseline CDAI score–0.022 (50.9)
      Maintenance exposure–response model
      Placebo effects
       β0–0.185 (127.6)
       TNF antagonist failure status–0.423 (67.4)
      Emax
       Emax0 non-TNF antagonist failure0.992 (52.5)
       Emax0 TNF antagonist failure0
      EC50
       EC50,0 Non-TNF antagonist failure0.873 (139.5)
      β0 = logit probability that R = 1 for placebo; CDAI = Crohn's disease activity index; CRP = C-reactive protein; EC50 = half-maximal effective drug concentration; EC50,0 = the EC50 for the reference patient who is a TNF non-failure patient and has a median CDAI score of 307; Emax = maximum drug effect; Emax0 = the Emax for the reference patient with a median baseline CRP of 0.93 mg/dL; RSE = relative SE; TNF = tumor necrosis factor.
      Figure 4
      Figure 4Goodness-of-fit plots for the induction (A) and maintenance (B) exposure–response models. The observed clinical remission rates were determined according to bins of the predicted ustekinumab concentrations at week 8 corresponding to the placebo group (0.0 µg/mL) and the quartiles of serum ustekinumab concentrations for patients receiving ustekinumab treatment for panel A and according to bins of the predicted steady-state ustekinumab trough concentration (before week 44) corresponding to the quartiles of serum ustekinumab concentrations for panel B. The points were plotted at the median concentration within each quartile bin. A 95% CI around the observed data was computed based on the normal approximation for proportions. The line segments at the bottom of the chart show the range of predicted concentrations at week 8 (median [solid circle], and the 10th and 90th percentiles) for the 130-mg IV dose (light purple) and the ∼6 mg/kg IV dose (dark purple) for panel A and at steady-state trough (median [solid circle], and the 10th and 90th percentiles) for the 90-mg administration every 8 weeks (q8w) SC dose regimen (dark purple) and the 90-mg administration every 12 weeks (q12w) SC dose regimen (light purple) for panel B. TNF = tumor necrosis factor.

      SC Maintenance Phase

      For the maintenance E-R analysis, the relationship between steady-state trough ustekinumab concentration and clinical remission at maintenance week 44 was assessed. Because the ustekinumab dose regimens used in the maintenance study were administered at different intervals (ie, q8w or q12w), the predicted steady-state trough concentration nearest to the time point of the primary maintenance end point of remission was selected as the exposure metric in this analysis. The maintenance E-R analysis data set included 387 patients who were in clinical response to ustekinumab induction during the Phase III induction studies (CNTO1275CRD3001 and CNTO1275CRD3002) and who were randomized to receive SC placebo (n = 128), ustekinumab 90 mg q12w (n = 130), or ustekinumab 90 mg q8w (n = 129) during the CNTO1275CRD3003 study. The parameters of the maintenance E-R model are summarized in Table IV. TNF antagonist failure status was the only variable with a statistically significant effect on the maintenance E-R model parameters. The model suggests that at a given ustekinumab concentration, patients who previously failed a TNF antagonist tend to have lower clinical remission rates. The goodness-of-fit plot for the maintenance E-R model is shown in Figure 4B.

      Discussion

      A robust population PK model was developed for ustekinumab in patients with CD based on a comprehensive data set from 4 late-phase clinical studies. Covariates affecting the disposition of ustekinumab in patients with CD were identified, and their effects on PK parameters were quantified. The PK model was used to simulate concentration–time profiles and predict ustekinumab exposure for new study scenarios and a pediatric CD population. E-R models were also developed and used to support the dose recommendation for the CD indication.
      A two-compartment linear PK model with first-order absorption and elimination adequately described the ustekinumab concentration–time data in CD. The typical value of ustekinumab CL in a patient with CD weighing ∼70 kg was 0.192 L/d, and the Vss was 4.62 L. These parameter estimates were comparable to those in patients with UC,
      • Adedokun OJ
      • Xu Z
      • Gasink C
      • Jacobstein D
      • Szapary P
      • Johanns J
      • et al.
      Pharmacokinetics and exposure response relationships of ustekinumab in patients with Crohn's disease.
      ,
      • Xu Y
      • Hu C
      • Chen Y
      • Miao X
      • Adedokun OJ
      • Xu Z
      • et al.
      Population pharmacokinetics and exposure-response modeling analyses of ustekinumab in adults with moderately to severely active ulcerative colitis.
      , suggesting that the ustekinumab PK profile is similar in CD and UC populations,
      • Adedokun OJ
      • Xu Z
      • Gasink C
      • Jacobstein D
      • Szapary P
      • Johanns J
      • et al.
      Pharmacokinetics and exposure response relationships of ustekinumab in patients with Crohn's disease.
      , which is consistent with what has been reported for other monoclonal antibodies approved for these indications, including infliximab,
      • Adedokun OJ
      • Xu Z
      • Padgett L
      • Blank M
      • Johanns J
      • Griffiths A
      • et al.
      Pharmacokinetics of infliximab in children with moderate-to-severe ulcerative colitis: results from a randomized, multicenter, open-label, phase 3 study.
      ,
      • Fasanmade AA
      • Adedokun OJ
      • Blank M
      • Zhou H
      • Davis HM.
      Pharmacokinetic properties of infliximab in children and adults with Crohn's disease: a retrospective analysis of data from 2 phase III clinical trials.
      adalimumab,

      Humira. Prescribing Information. AbbVie; 2021 Accessed April 21, 2021. https://www.rxabbvie.com/pdf/HUMIRA.pdf.

      and vedolizumab

      Entyvio. Prescribing Information. Takeda; 2021. Accessed April 21, 2021. https://general.takedapharm.com/ENTYVIOPI.

      ,
      • Rosario M
      • Dirks NL
      • Gastonguay MR
      • Fasanmade AA
      • Wyant T
      • Parikh A
      • et al.
      Population pharmacokinetics-pharmacodynamics of vedolizumab in patients with ulcerative colitis and Crohn's disease.
      (see Supplemental Table I in the online version at doi:10.1016/j.clinthera.2022.08.010) and does not support the hypothesis that fundamental PK differences generally exist between UC and CD.
      • Ordás I
      • Mould DR
      • Feagan BG
      • Sandborn WJ.
      Anti-TNF monoclonal antibodies in inflammatory bowel disease: pharmacokinetics-based dosing paradigms.
      The estimate of Vss suggests that ustekinumab is primarily distributed into the vascular system, a finding consistent with its molecular weight (∼149 kDa) as well as the known distribution of endogenous immunoglobulin G and some other monoclonal antibody therapies.
      • Ryman JT
      • Meibohm B.
      Pharmacokinetics of monoclonal antibodies.
      The bioavailability of ustekinumab after SC administration was 78.3%, which is higher than the value reported in psoriasis (57.2%) yet comparable to the estimate reported in UC (87.2%).
      • Xu Y
      • Hu C
      • Chen Y
      • Miao X
      • Adedokun OJ
      • Xu Z
      • et al.
      Population pharmacokinetics and exposure-response modeling analyses of ustekinumab in adults with moderately to severely active ulcerative colitis.
      The difference in these estimates of bioavailability may be that the lower estimate obtained in psoriasis was based on a cross-study comparison of data between two Phase I studies with a limited number of patients receiving a single IV dose (C0379T01) and a single SC dose (C0379T02) of ustekinumab.
      • Gottlieb AB
      • Cooper KD
      • McCormick TS
      • Toichi E
      • Everitt DE
      • Frederick B
      • et al.
      A phase 1, double-blind, placebo-controlled study evaluating single subcutaneous administrations of a human interleukin-12/23 monoclonal antibody in subjects with plaque psoriasis.
      • Kauffman CL
      • Aria N
      • Toichi E
      • McCormick TS
      • Cooper KD
      • Gottlieb AB
      • et al.
      A phase I study evaluating the safety, pharmacokinetics, and clinical response of a human IL-12 p40 antibody in subjects with plaque psoriasis.

      United States Food and Drug Administration, Center for Drug Evaluation and Research. Clinical pharmacology and biopharmaceutics review for ustekinumab single use liquid in vial, 45 mg/0.5 mL and 90 mg/1.0 mL. 2009. Accessed April 21, 2021. https://www.accessdata.fda.gov/drugsatfda_docs/nda/2009/125261s000_ClinPharmR.pdf.

      In contrast, the bioavailability estimates in CD and UC were obtained from respective indication data sets in which patients received both IV and SC ustekinumab in the same study, providing a more robust estimate of SC bioavailability. The estimate of the elimination t1/2 was ∼19 days, consistent with those reported for other ustekinumab indications.
      • Zhu Y.
      • Hu C
      • Lu M
      • Liao S
      • Marini JC
      • Yohrling J
      • et al.
      Population pharmacokinetic modeling of ustekinumab, a human monoclonal antibody targeting IL-12/23p40, in patients with moderate to severe plaque psoriasis.
      ,
      • Zhu YW
      • Mendelsohn A
      • Pendley C
      • Davis HM
      • Zhou H.
      Population pharmacokinetics of ustekinumab in patients with active psoriatic arthritis.
      ,
      • Xu Y
      • Hu C
      • Chen Y
      • Miao X
      • Adedokun OJ
      • Xu Z
      • et al.
      Population pharmacokinetics and exposure-response modeling analyses of ustekinumab in adults with moderately to severely active ulcerative colitis.
      In addition, similar to findings in UC studies,
      • Xu Y
      • Hu C
      • Chen Y
      • Miao X
      • Adedokun OJ
      • Xu Z
      • et al.
      Population pharmacokinetics and exposure-response modeling analyses of ustekinumab in adults with moderately to severely active ulcerative colitis.
      maintenance-phase steady-state ustekinumab concentrations were achieved by the start of the second maintenance dose in the present analysis, with the influence of the single IV induction dose (loading dose) on the concentration profile being washed out by the time of administration of the second SC maintenance dose.
      The higher IIV in CL (29.1%) relative to Vc (14.7%), along with the identification of several covariates influencing ustekinumab CL in the final population PK model, suggest that the CL parameter plays a relatively important role in the variability of systemic exposure and potentially in the variability of response to ustekinumab in CD. Covariate analysis identified body weight as influential on the key structural model parameters of CL, Vc, Vp, and Q, with these parameters increasing nonlinearly with body weight. Although the approved induction doses approximated 6 mg/kg in each body-weight tier (260 mg for patients weighing ≤55 kg, 390 mg for patients weighing >55 to ≤85 kg, and 520 mg for patients weighing >85 kg), the less than dose-proportional increase in CL with body weight in the model explains why serum ustekinumab concentrations increased slightly as absolute dose levels increased. On the other hand, the effect of body weight on PK variables in the model also explains why slightly higher ustekinumab concentrations may be expected in patients with CD with lower body weights when the same flat dose is administered.
      Results from the present analysis showed that CL increased with lower baseline serum albumin levels and higher CRP, consistent with findings in the population PK analysis of some other biologics.
      • Rosario M
      • Dirks NL
      • Gastonguay MR
      • Fasanmade AA
      • Wyant T
      • Parikh A
      • et al.
      Population pharmacokinetics-pharmacodynamics of vedolizumab in patients with ulcerative colitis and Crohn's disease.
      ,
      • Fasanmade AA
      • Adedokun OJ
      • Ford J
      • Hernandez D
      • Johanns J
      • Hu C
      • et al.
      Population pharmacokinetic analysis of infliximab in patients with ulcerative colitis.
      ,
      • Suleiman AA
      • Khatri A
      • Minocha M
      • Othman AA.
      Population pharmacokinetics of the interleukin-23 inhibitor risankizumab in subjects with psoriasis and Crohn's disease: analyses of phase I and II trials.
      Low serum albumin has been associated with rapid clearance of TNF antagonists and poorer efficacy outcomes in patients with IBD.
      • Fasanmade A
      • Adedokun OJ
      • Olson A
      • Strauss R
      • Davis HM.
      Serum albumin concentration: a predictive factor of infliximab pharmacokinetics and clinical response in patients with ulcerative colitis.
      Possible explanations for the correlation between serum albumin concentrations and monoclonal antibody CL include the interaction between monoclonal antibodies and the neonatal fragment crystallizable receptor system, for which albumin is also a substrate.
      • Fasanmade A
      • Adedokun OJ
      • Olson A
      • Strauss R
      • Davis HM.
      Serum albumin concentration: a predictive factor of infliximab pharmacokinetics and clinical response in patients with ulcerative colitis.
      It has also been hypothesized that the increased mucosal permeability in severe IBD may lead to loss of both administered biologic drug and albumin, and this correlation explains the increased drug clearance seen in patients with low albumin levels.
      • Lichtenstein GR
      • McGovern DPB.
      Using markers in IBD to predict disease and treatment outcomes: rationale and a review of current status.
      Although Vc also increased with albumin levels, the effect was much smaller compared with that of CL. CRP elevation is indicative of a higher degree of inflammation that, in turn, correlates with lower serum ustekinumab concentration. Indeed, elevated baseline CRP was identified as an indicator of dose intensification in a recent real-world study of ustekinumab in patients with CD.
      • Sipponen T
      • af Björkesten C-G
      • Hallinen T
      • Ilus T
      • Soini E
      • Eberl A
      • et al.
      A nationwide real-world study on dynamic ustekinumab dosing and concomitant medication use among Crohn's disease patients in Finland.
      CL and Vc in male patients were 17% higher than in female patients. Because body weight was already accounted for in the model, these effects may represent an additional layer of variability. CL was higher (11%) in patients who had previously failed TNF antagonists than in those without this treatment history. Although TNF antagonist failure is predicted by low drug concentrations mediated, in part, by immunogenicity to the TNF antagonist,
      • Kennedy NA
      • Heap GA
      • Green HD
      • Hamilton B
      • Bewshea C
      • Walker GJ
      • et al.
      Predictors of anti-TNF treatment failure in anti-TNF-naive patients with active luminal Crohn's disease: a prospective, multicentre, cohort study.
      it is not clear why patients in this category experienced slightly higher CL. CL was also higher (14%) in Asian versus non-Asian patients. Several studies have shown that observed differences in biologic exposure between Asian and White populations are primarily driven by the difference in body weight and not due to race or ethnic differences.
      • Zhu Y
      • Wang Q
      • Frederick B
      • Bouman-Thio E
      • Marini JC
      • Keen M
      • et al.
      Comparison of the pharmacokinetics of subcutaneous ustekinumab between Chinese and non-Chinese healthy male subjects across two phase 1 studies.
      • Chiba K
      • Yoshitsugu H
      • Kyosaka Y
      • Iida S
      • Yoneyama K
      • Tanigawa T
      • et al.
      A comprehensive review of the pharmacokinetics of approved therapeutic monoclonal antibodies in Japan: are Japanese phase I studies still needed?.
      • Khatri A
      • Eckert D
      • Oberoi R
      • Suleiman A
      • Pang Y
      • Cheng L
      • et al.
      Pharmacokinetics of risankizumab in Asian healthy subjects and patients with moderate to severe plaque psoriasis, generalized pustular psoriasis, and erythrodermic psoriasis.
      • Okamoto H
      • Dirks NL
      • Rosario M
      • Hori T
      • Hibi T.
      Population pharmacokinetics of vedolizumab in Asian and non-Asian patients with ulcerative colitis and Crohn's disease.
      However, the effect of race on CL in our study was observed after accounting for differences in body weight; this finding does not have an obvious physiological explanation. The development of anti-ustekinumab antibodies resulted in a 13% increase in CL; this small impact on CL may explain why there was no demonstrable effect of immunogenicity on the efficacy of ustekinumab in CD,
      • Adedokun OJ
      • Xu Z
      • Gasink C
      • Jacobstein D
      • Szapary P
      • Johanns J
      • et al.
      Pharmacokinetics and exposure response relationships of ustekinumab in patients with Crohn's disease.
      which is in contrast to other biologics where the development of anti-drug antibodies has been shown to adversely affect efficacy.
      • Strand V
      • Balsa A
      • Al-Saleh J
      • Barile-Fabris L
      • Horiuchi T
      • Takeuchi T
      • et al.
      Immunogenicity of biologics in chronic inflammatory diseases: a systematic review.
      None of the other covariates tested influenced ustekinumab PK variables. Notably, coadministration of immunosuppressive agents or corticosteroids did not affect ustekinumab CL in CD, a finding consistent with the UC population
      • Xu Y
      • Hu C
      • Chen Y
      • Miao X
      • Adedokun OJ
      • Xu Z
      • et al.
      Population pharmacokinetics and exposure-response modeling analyses of ustekinumab in adults with moderately to severely active ulcerative colitis.
      but different from that reported for TNF antagonists in patients with rheumatoid arthritis or CD, in whom CL decreased with administration of methotrexate or AZA.
      • Ternant D
      • Ducourau E
      • Perdriger A
      • Corondan A
      • Le Goff B
      • Devauchelle-Pensec V
      • et al.
      Relationship between inflammation and infliximab pharmacokinetics in rheumatoid arthritis.
      ,
      • Colombel J-F
      • Adedokun OJ
      • Gasink C
      • Gao L-L
      • Cornillie FJ
      • D'Haens GR
      • et al.
      Combination therapy with infliximab and azathioprine improves infliximab pharmacokinetic features and efficacy: a post hoc analysis.
      This observation suggests that from a strictly PK exposure perspective, monotherapy with ustekinumab may be as effective as combination therapy with these agents in IBD. Indeed, the consistency in efficacy between ustekinumab monotherapy and combination therapy has been reported in previous analyses of ustekinumab clinical data.
      • Sandborn WJ
      • Gasink C
      • Gao L-L
      • Blank MA
      • Johanns J
      • Guzzo C
      • et al.
      Ustekinumab induction and maintenance therapy in refractory Crohn's disease.
      ,
      • Feagan BG
      • Sandborn WJ
      • Gasink C
      • Jacobstein D
      • Lang Y
      • Friedman JR
      • et al.
      Ustekinumab as induction and maintenance therapy for Crohn's disease.
      In general, the effects of the explanatory covariates were modest in magnitude and consistent with those found in patients with UC; the impact of the covariates on the respective PK parameters was within ±20% when evaluated across a representative range (25th–75th percentiles) of covariate values or categories in the data.
      PK simulations based on the population PK model developed from this work were used to support the dose justification for an IV reinduction study of ustekinumab in patients with secondary loss of response to SC ustekinumab (NCT03782376). The PK model was also used to predict exposures in a study comparing treat-to-target versus routine care maintenance strategies in CD in which dose escalation of the ustekinumab SC maintenance dose from 90 mg q8w to 90 mg q4w was included in the study design (NCT03107793). Furthermore, the model supported the evaluation of a higher milligram per kilogram induction dose in pediatric patients (particularly for those with low body weight) compared with the dose approved in adult patients (NCT02968108).
      The E-R assessments revealed that nonlinear logistic regression models assuming a maximum effect (Emax) E-R relationship adequately fit the concentration-remission data for both induction and maintenance. In the induction E-R model, baseline values of CRP and CDAI along with TNF antagonist failure status influenced the placebo response, indicating that these factors should be considered for stratification when conducting ustekinumab induction studies. Indeed, baseline CDAI was used as a stratification factor in both pivotal induction trials of ustekinumab in CD (CNTO1275CRD3001 and CNTO1275CRD3002), whereas initial response to TNF antagonist was used as a stratification factor in CNTO1275CRD3001.
      • Feagan BG
      • Sandborn WJ
      • Gasink C
      • Jacobstein D
      • Lang Y
      • Friedman JR
      • et al.
      Ustekinumab as induction and maintenance therapy for Crohn's disease.
      These stratification strategies are supported by the induction E-R model. Only baseline CRP was determined to influence Emax with the model, suggesting that the effect of ustekinumab was more pronounced among patients with higher baseline CRP levels. This observation is consistent with earlier reports of greater treatment effects with ustekinumab in patients with higher levels of CRP.
      • Toedter GP
      • Blank M
      • Lang Y
      • Chen D
      • Sandborn WJ
      • de Villiers WJS.
      Relationship of C-reactive protein with clinical response after therapy with ustekinumab in Crohn's disease.
      The concentration of ustekinumab eliciting 50% of the maximum response was lower in patients who had higher baseline CDAI scores or who previously failed a TNF antagonist. This observation implies that, in these subgroups of patients, a higher induction dose, which corresponds to higher ustekinumab concentration, may not be needed to elicit better efficacy outcomes than in patients with lower baseline CDAI scores or in those with no history of TNF antagonist failure. Accordingly, the approval of ustekinumab at the same induction posology regardless of baseline disease severity or history of TNF antagonist failure seems justified.
      In the maintenance E-R analysis, an Emax E-R model could only be supported for the TNF antagonist nonfailure population, probably due to the relatively smaller sample size used to develop the model compared with that used for induction. The flat maintenance E-R curve for the TNF antagonist failure population may have implications for E-R–based extrapolation requirements for pediatric studies in this subgroup of patients with CD. Because pediatric extrapolation is sometimes contingent on the existence of an E-R model in adults,

      European Medicines Agency. Reflection paper on the use of extrapolation in the development of medicines for paediatrics. October 7, 2018. Accessed April 21, 2021. https://www.ema.europa.eu/en/documents/scientific-guideline/adopted-reflection-paper-use-extrapolation-development-medicines-paediatrics-revision-1_en.pdf.

      these maintenance E-R data warrant an alternative approach to extrapolation. Instead of attempting to establish an E-R curve, which may not be feasible given the smaller size of a pediatric study relative to the maintenance adult E-R population, a more reasonable approach would be to determine the ustekinumab exposure (and consequently the dose regimen) in pediatric patients that results in efficacy comparable to that observed in adults.
      Collectively, the general pattern of a positive E-R relationship observed in adult patients with CD confirms the effectiveness of ustekinumab in this indication and also supports the use of a posology consisting of the single ∼6 mg/kg IV induction dose followed by the 90 mg SC q12w or q8w maintenance dose regimen. Although 90 mg q12w might be efficacious in many patients who may require a lower concentration threshold for efficacy, the E-R analysis supports 90 mg SC q8w as the more optimal treatment regimen for the overall CD patient population.
      A limitation of the E-R analyses presented in this report is that the models do not necessarily establish a causal link between ustekinumab exposure and efficacy. Prospective studies that take into account the findings from the E-R model can further support therapeutic strategies to optimize the use of ustekinumab in the treatment of CD.

      Conclusions

      In patients with CD, ustekinumab disposition after IV and SC administration was biexponential and consistent with that in patients with UC. Prior treatment with TNF antagonists or the concomitant use of immunosuppressive agents or corticosteroids had no impact on ustekinumab disposition. Ustekinumab clearance varied with body weight, serum albumin and CRP levels, sex, history of TNF antagonist failure, presence of anti-drug antibodies, and race (Asian vs non-Asian); however, these covariates did not have a clinically meaningful impact on ustekinumab exposure. E-R models support recommended posology of ustekinumab in adults with CD.

      Acknowledgments

      The study was funded by Janssen Research and Development, LLC. Medical writing support was provided by Holly Capasso-Harris, PhD, and Linda J. Cornfield, PhD, of Certara Synchrogenix under the direction of the authors in accordance with Good Publication Practice guidelines.
      Author contributions were as follows: acquisition of data, Mr Adedokun, Dr Xu, and Dr Gasink; and analysis and interpretation of data, Mr Adedokun, Dr Xu, Dr Gasink, Mr Kowalski, Dr Sandborn, and Dr Feagan. All authors contributed to conception and design of the study, manuscript review, and editing for important intellectual content.

      Funding

      The study was funded by Janssen Research and Development, LLC. Employees from the study funder were involved in study design, data analysis, data interpretation, and writing of the report.

      Declaration of Interest

      Omoniyi J. Adedokun and Zhenhua Xu are employees of Janssen Research & Development, LLC., and own stock in Johnson & Johnson, of which Janssen Research & Development, LLC., is a wholly owned subsidiary.
      Christopher Gasink is an employee of Janssen Scientific Affairs, LLC., and owns stock in Johnson & Johnson, of which Janssen Scientific Affairs, LLC., is a wholly owned subsidiary.
      Kenneth G. Kowalski was a paid consultant for Janssen Research & Development, LLC.
      William J. Sandborn has received research grants from AbbVie, Abivax, Arena Pharmaceuticals, Boehringer Ingelheim, Celgene, Genentech, Gilead Sciences, Glaxo Smith Kline, Janssen, Lilly, Pfizer, Prometheus Biosciences, Seres Therapeutics, Shire, Takeda, and Theravance Biopharma; consulting fees from AbbVie, Abivax, Admirx, Alfasigma, Alimentiv (previously Robarts Clinical Trials, owned by Alimentiv Health Trust), Alivio Therapeutics, Allakos, Amgen, Applied Molecular Transport, Arena Pharmaceuticals, Bausch Health (Salix), Beigene, Bellatrix Pharmaceuticals, Boehringer Ingelheim, Boston Pharmaceuticals, Bristol Meyers Squibb, Celgene, Celltrion, Cellularity, Cosmo Pharmaceuticals, Escalier Biosciences, Equillium, Forbion, Genentech/Roche, Gilead Sciences, Glenmark Pharmaceuticals, Gossamer Bio, Immunic (Vital Therapies), Index Pharmaceuticals, Intact Therapeutics, Janssen, Kyverna Therapeutics, Landos Biopharma, Lilly, Oppilan Pharma (acquired by Ventyx Biosciences), Otsuka, Pandion Therapeutics, Pfizer, Progenity, Prometheus Biosciences, Prometheus Laboratories, Protagonists Therapeutics, Provention Bio, Reistone Biopharma, Seres Therapeutics, Shanghai Pharma Biotherapeutics, Shire, Shoreline Biosciences, Sublimity Therapeutics, Surrozen, Takeda, Theravance Biopharma, Thetis Pharmaceuticals, Tillotts Pharma, UCB, Vendata Biosciences, Ventyx Biosciences, Vimalan Biosciences, Vivelix Pharmaceuticals, Vivreon Biosciences, and Zealand Pharma; stock or stock options from Allakos, BeiGene, Gossamer Bio, Oppilan Pharma (acquired by Ventyx Biosciences), Prometheus Biosciences, Prometheus Laboratories Progenity, Shoreline Biosciences, Ventyx Biosciences, Vimalan Biosciences, Vivreon Biosciences; and is an employee at Shoreline Biosciences. Spouse: Iveric Bio - consultant, stock options; Progenity - stock; Oppilan Pharma (acquired by Ventyx Biosciences) - consultant, stock options; Prometheus Biosciences - employee, stock, stock options; Prometheus Laboratories – stock, stock options, consultant; Ventyx Biosciences – stock, stock options; Vimalan Biosciences – stock, stock options.
      Brian Feagan has received grant/research support from AbbVie Inc., Amgen Inc., AstraZeneca/MedImmune Ltd., Atlantic Pharmaceuticals Ltd., Boehringer-Ingelheim, Celgene Corporation, Celltech, Genentech Inc/Hoffmann-La Roche Ltd., Gilead Sciences Inc., GlaxoSmithKline (GSK), Janssen Research & Development LLC., Pfizer Inc., Receptos Inc. / Celgene International, Sanofi, Santarus Inc., Takeda Development Center Americas Inc., Tillotts Pharma AG, and UCB; is a consultant for Abbott/AbbVie, AdMIRx Inc., Akebia Therapeutics,  Allergan, Amgen, Applied Molecular Transport Inc., Aptevo Therapeutics, Asta Pharma, Astra Zeneca, Atlantic Pharma, Avir Pharma, Biogen Idec, BioMx Israel, Boehringer-Ingelheim, Boston Pharmaceuticals, Bristol-Myers Squibb, Calypso Biotech, Celgene, Elan/Biogen, EnGene, Ferring Pharma, Roche/Genentech, Galapagos, Galen/Atlantica, GiCare Pharma, Gilead, Gossamer Pharma, GSK, Inception IBD Inc., Intact Therapeutics, JnJ/Janssen, Kyowa Kakko Kirin Co Ltd., Lexicon, Lilly, Lycera BioTech, Merck, Mesoblast Pharma, Millennium, Nestles, Nextbiotix, Novonordisk, ParImmune, Parvus Therapeutics Inc., Pfizer, Prometheus Therapeutics and Diagnostics, Progenity, Protagonist, Qu Biologics, Rebiotix, Receptos, Salix Pharma, Shire, Sienna Biologics, Sigmoid Pharma, Sterna Biologicals, Synergy Pharma Inc., Takeda, Teva Pharma, TiGenix, Tillotts, UCB Pharma, Vertex Pharma, Vivelix Pharma, VHsquared Ltd., and Zyngenia; is a member of the speakers bureau for Abbott/AbbVie, JnJ/Janssen, Lilly, Takeda, Tillotts, UCB Pharma; a member of the scientific advisory board for Abbott/AbbVie, Allergan, Amgen, Astra Zeneca, Atlantic Pharma, Avaxia Biologics Inc., Boehringer-Ingelheim, Bristol-Myers Squibb, Celgene, Centocor Inc., Elan/Biogen, Galapagos, Genentech/Roche, JnJ/Janssen, Merck, Nestles, Novartis, Novonordisk, Pfizer, Prometheus Laboratories, Protagonist, Salix Pharma, Sterna Biologicals, Takeda, Teva, TiGenix, Tillotts Pharma AG, and UCB Pharma; and is Senior Scientific Officer – Robarts Clinical Trials Inc.
      Employees from the study funder were involved in study design, data analysis, data interpretation, and writing of the report.

      Appendix. Supplementary materials

      References

        • Stelara
        INN-Ustekinumab Prescribing Information.
        European Medicines Agency, 2022 (Updated 04/04/2022. [Accessed 19 April 2022])
        • Zhu Y.
        • Hu C
        • Lu M
        • Liao S
        • Marini JC
        • Yohrling J
        • et al.
        Population pharmacokinetic modeling of ustekinumab, a human monoclonal antibody targeting IL-12/23p40, in patients with moderate to severe plaque psoriasis.
        J Clin Pharmacol. 2009; 49: 162-175
        • Wang Z
        • Verstockt B
        • Sabino J
        • Vermeire S
        • Ferrante M
        • Declerck P
        • et al.
        Population pharmacokinetic-pharmacodynamic model-based exploration of alternative ustekinumab dosage regimens for patients with Crohn's disease.
        Br J Clin Pharmacol. 2022; 88: 323-335
      1. European Medicines Agency. Assessment report: Stelara. 15 September 2016. Accessed April 19, 2022. https://www.ema.europa.eu/en/documents/variation-report/stelara-h-c-000958-x-0049-g-epar-assessment-report-extension_en.pdf.

        • Adedokun OJ
        • Xu Z
        • Gasink C
        • Jacobstein D
        • Szapary P
        • Johanns J
        • et al.
        Pharmacokinetics and exposure response relationships of ustekinumab in patients with Crohn's disease.
        Gastroenterology. 2018; 154: 1660-1671
        • Zhu YW
        • Mendelsohn A
        • Pendley C
        • Davis HM
        • Zhou H.
        Population pharmacokinetics of ustekinumab in patients with active psoriatic arthritis.
        Int J Clin Pharmacol Ther. 2010; 48: 830-846
        • Xu Y
        • Hu C
        • Chen Y
        • Miao X
        • Adedokun OJ
        • Xu Z
        • et al.
        Population pharmacokinetics and exposure-response modeling analyses of ustekinumab in adults with moderately to severely active ulcerative colitis.
        J Clin Pharmacol. 2020; 60: 889-902
      2. Stelara. Prescribing Information. Janssen; 2020. Accessed April 21, 2021. https://www.janssenlabels.com/package-insert/product-monograph/prescribing-information/STELARA-pi.pdf.

        • Sandborn WJ
        • Gasink C
        • Gao L-L
        • Blank MA
        • Johanns J
        • Guzzo C
        • et al.
        Ustekinumab induction and maintenance therapy in refractory Crohn's disease.
        N Engl J Med. 2012; 367: 1519-1528
        • Feagan BG
        • Sandborn WJ
        • Gasink C
        • Jacobstein D
        • Lang Y
        • Friedman JR
        • et al.
        Ustekinumab as induction and maintenance therapy for Crohn's disease.
        N Engl J Med. 2016; 375: 1946-1960
        • Kowalski KG
        • Hutmacher MM.
        Efficient screening of covariates in population models using Wald's approximation to the likelihood ratio test.
        J Pharmacokinet Pharmacodyn. 2001; 28: 253-275
        • Sandborn WJ
        • Feagan BG
        • Hanauer SB
        • Lochs H
        • Lӧfberg R
        • Modigliani R
        • et al.
        A review of activity indices and efficacy endpoints for clinical trials of medical therapy in adults with Crohn's disease.
        Gastroenterology. 2002; 122: 512-530
        • Rosh JR
        • Turner D
        • Griffiths A
        • Cohen SA
        • Jacobstein D
        • Adedokun OJ
        • et al.
        Ustekinumab in pediatric patients with moderately to severely active Crohn's disease pharmacokinetics, safety, and efficacy results from UniStar, a phase 1 study.
        J Crohns Colitis. 2021; 15: 1931-1942
        • Adedokun OJ
        • Xu Z
        • Padgett L
        • Blank M
        • Johanns J
        • Griffiths A
        • et al.
        Pharmacokinetics of infliximab in children with moderate-to-severe ulcerative colitis: results from a randomized, multicenter, open-label, phase 3 study.
        Inflamm Bowel Dis. 2013; 19: 2753-2762
        • Fasanmade AA
        • Adedokun OJ
        • Blank M
        • Zhou H
        • Davis HM.
        Pharmacokinetic properties of infliximab in children and adults with Crohn's disease: a retrospective analysis of data from 2 phase III clinical trials.
        Clin Ther. 2011; 33: 946-964
      3. Humira. Prescribing Information. AbbVie; 2021 Accessed April 21, 2021. https://www.rxabbvie.com/pdf/HUMIRA.pdf.

      4. Entyvio. Prescribing Information. Takeda; 2021. Accessed April 21, 2021. https://general.takedapharm.com/ENTYVIOPI.

        • Rosario M
        • Dirks NL
        • Gastonguay MR
        • Fasanmade AA
        • Wyant T
        • Parikh A
        • et al.
        Population pharmacokinetics-pharmacodynamics of vedolizumab in patients with ulcerative colitis and Crohn's disease.
        Aliment Pharmacol Ther. 2015; 42: 188-202
        • Ordás I
        • Mould DR
        • Feagan BG
        • Sandborn WJ.
        Anti-TNF monoclonal antibodies in inflammatory bowel disease: pharmacokinetics-based dosing paradigms.
        Clin Pharmacol Ther. 2012; 91: 635-646
        • Ryman JT
        • Meibohm B.
        Pharmacokinetics of monoclonal antibodies.
        CPT Pharmacometrics Syst Pharmacol. 2017; 6: 576-588
        • Stelara
        Summary of Product Characteristics.
        Janssen, 2013 (Accessed April 21, 2021)
        • Gottlieb AB
        • Cooper KD
        • McCormick TS
        • Toichi E
        • Everitt DE
        • Frederick B
        • et al.
        A phase 1, double-blind, placebo-controlled study evaluating single subcutaneous administrations of a human interleukin-12/23 monoclonal antibody in subjects with plaque psoriasis.
        Curr Med Res Opin. 2007; 23: 1081-1092
        • Kauffman CL
        • Aria N
        • Toichi E
        • McCormick TS
        • Cooper KD
        • Gottlieb AB
        • et al.
        A phase I study evaluating the safety, pharmacokinetics, and clinical response of a human IL-12 p40 antibody in subjects with plaque psoriasis.
        J Invest Dermatol. 2004; 123: 1037-1044
      5. United States Food and Drug Administration, Center for Drug Evaluation and Research. Clinical pharmacology and biopharmaceutics review for ustekinumab single use liquid in vial, 45 mg/0.5 mL and 90 mg/1.0 mL. 2009. Accessed April 21, 2021. https://www.accessdata.fda.gov/drugsatfda_docs/nda/2009/125261s000_ClinPharmR.pdf.

        • Fasanmade AA
        • Adedokun OJ
        • Ford J
        • Hernandez D
        • Johanns J
        • Hu C
        • et al.
        Population pharmacokinetic analysis of infliximab in patients with ulcerative colitis.
        Eur J Clin Pharmacol. 2009; 65: 1211-1228
        • Suleiman AA
        • Khatri A
        • Minocha M
        • Othman AA.
        Population pharmacokinetics of the interleukin-23 inhibitor risankizumab in subjects with psoriasis and Crohn's disease: analyses of phase I and II trials.
        Clin Pharmacokinet. 2019; 58: 375-387
        • Fasanmade A
        • Adedokun OJ
        • Olson A
        • Strauss R
        • Davis HM.
        Serum albumin concentration: a predictive factor of infliximab pharmacokinetics and clinical response in patients with ulcerative colitis.
        Int J Clin Pharmacol Ther. 2010; 48: 297-308
        • Lichtenstein GR
        • McGovern DPB.
        Using markers in IBD to predict disease and treatment outcomes: rationale and a review of current status.
        Am J Gastroenterol Suppl. 2016; 3: 17-26
        • Sipponen T
        • af Björkesten C-G
        • Hallinen T
        • Ilus T
        • Soini E
        • Eberl A
        • et al.
        A nationwide real-world study on dynamic ustekinumab dosing and concomitant medication use among Crohn's disease patients in Finland.
        Scand J Gastroenterol. 2021; 56: 661-670
        • Kennedy NA
        • Heap GA
        • Green HD
        • Hamilton B
        • Bewshea C
        • Walker GJ
        • et al.
        Predictors of anti-TNF treatment failure in anti-TNF-naive patients with active luminal Crohn's disease: a prospective, multicentre, cohort study.
        Lancet Gastroenterol Hepatol. 2019; 4: 341-353
        • Zhu Y
        • Wang Q
        • Frederick B
        • Bouman-Thio E
        • Marini JC
        • Keen M
        • et al.
        Comparison of the pharmacokinetics of subcutaneous ustekinumab between Chinese and non-Chinese healthy male subjects across two phase 1 studies.
        Clin Drug Investig. 2013; 33: 291-301
        • Chiba K
        • Yoshitsugu H
        • Kyosaka Y
        • Iida S
        • Yoneyama K
        • Tanigawa T
        • et al.
        A comprehensive review of the pharmacokinetics of approved therapeutic monoclonal antibodies in Japan: are Japanese phase I studies still needed?.
        J Clin Pharmacol. 2014; 54: 483-494
        • Khatri A
        • Eckert D
        • Oberoi R
        • Suleiman A
        • Pang Y
        • Cheng L
        • et al.
        Pharmacokinetics of risankizumab in Asian healthy subjects and patients with moderate to severe plaque psoriasis, generalized pustular psoriasis, and erythrodermic psoriasis.
        J Clin Pharmacol. 2019; 59: 1656-1668
        • Okamoto H
        • Dirks NL
        • Rosario M
        • Hori T
        • Hibi T.
        Population pharmacokinetics of vedolizumab in Asian and non-Asian patients with ulcerative colitis and Crohn's disease.
        Intest Res. 2021; 19: 95-105
        • Strand V
        • Balsa A
        • Al-Saleh J
        • Barile-Fabris L
        • Horiuchi T
        • Takeuchi T
        • et al.
        Immunogenicity of biologics in chronic inflammatory diseases: a systematic review.
        BioDrugs. 2017; 31: 299-316
        • Ternant D
        • Ducourau E
        • Perdriger A
        • Corondan A
        • Le Goff B
        • Devauchelle-Pensec V
        • et al.
        Relationship between inflammation and infliximab pharmacokinetics in rheumatoid arthritis.
        Br J Clin Pharmacol. 2014; 78: 118-128
        • Colombel J-F
        • Adedokun OJ
        • Gasink C
        • Gao L-L
        • Cornillie FJ
        • D'Haens GR
        • et al.
        Combination therapy with infliximab and azathioprine improves infliximab pharmacokinetic features and efficacy: a post hoc analysis.
        Clin Gastroenterol Hepatol. 2019; 17: 1525-1532
        • Toedter GP
        • Blank M
        • Lang Y
        • Chen D
        • Sandborn WJ
        • de Villiers WJS.
        Relationship of C-reactive protein with clinical response after therapy with ustekinumab in Crohn's disease.
        Am J Gastroenterol. 2009; 104: 2768-2773
      6. European Medicines Agency. Reflection paper on the use of extrapolation in the development of medicines for paediatrics. October 7, 2018. Accessed April 21, 2021. https://www.ema.europa.eu/en/documents/scientific-guideline/adopted-reflection-paper-use-extrapolation-development-medicines-paediatrics-revision-1_en.pdf.