| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
BIOLOGICS |
Monoclonal Antibody, in Subjects With Rheumatoid ArthritisFrom Centocor Research and Development, Inc, Malvern, Pennsylvania (Dr Zhou, Dr Jang, Dr Bouman-Thio, Dr Xu, Dr Marini, Dr Pendley, Q. Jiao, Dr Shankar, S. J. Marciniak, Dr Rahman, Dr Baker, Dr Mascelli, Dr Davis, Dr Everitt) and Radiant Research, Dallas, Texas (Dr Fleischmann, Dr Cohen).
Address for reprints: Address for correspondence: Honghui Zhou, PhD, FCP, Clinical Pharmacology & Experimental Medicine, Centocor Research & Development, 200 Great Valley Parkway, Malvern, PA 19355; e-mail: hzhou2{at}cntus.jnj.com.
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
) monoclonal antibody that is being developed for intravenous and subcutaneous administration. To assess the pharmacokinetics and safety of the intravenous formulation of golimumab, 36 adult subjects with rheumatoid arthritis were randomly assigned to receive a single infusion of placebo or golimumab (0.1, 0.3, 1, 3, 6, or 10 mg/kg). Serum concentrations of golimumab were determined using a validated enzyme-linked immunosorbent assay method. In addition to the noncompartmental analysis and compartmental modeling, a population pharmacokinetics analysis using NONMEM was also conducted. Both the maximum serum concentration and the area under the serum concentrationtime curve appeared to increase in a dose-proportional manner. The median half-life ranged from 7 to 20 days. A 2-compartment population pharmacokinetic model adequately described the pharmacokinetics of golimumab. The following pharmacokinetic parameters (typical value [% coefficient of variation]) were estimated from the population pharmacokinetic model: clearance (CL: 0.40 [10.1%] L/d), volume of distribution in the central compartment (Vc: 3.07 [6.4%] L), intercompartmental clearance (Q: 0.42 [15.5%] L/d), and volume of distribution in the peripheral compartment (Vp: 3.68 [11.8%] L). Interindividual variability of the pharmacokinetic parameters was quantified for CL (44.3%), Vc (25.5%), Q (44.6%), and Vp (44.6%). Residual variability was estimated to be 15.0%. Body weight was found to be an important covariate on Vc. Golimumab was generally well tolerated. The pharmacokinetics of golimumab appeared to be linear over the dose range evaluated in this study.
Key Words: rheumatoid arthritis • human anti-TNF- monoclonal antibody (mAb) • golimumab • pharmacokinetics
) is a key inflammatory mediator that exhibits a wide variety of functional activities.1 TNF- has been implicated in the pathology of a variety of diseases, such as rheumatoid arthritis (RA),2 uveitis,3 Crohn's disease,4 asthma,5 multiple sclerosis, psoriasis, diabetes mellitus, human immunodeficiency virus (HIV), septic shock, graft versus host disease, allograft rejection, and cerebral malaria, among others. Binding of an antibody to TNF- can inhibit or prevent the interactions of this cytokine with its cellular receptors and may prevent the deleterious effects caused by excessive TNF-.
TNF- inhibitors have been used successfully in a wide variety of immune-mediated anti-inflammatory diseases. Golimumab is a human monoclonal antibody (mAb) to TNF- that is being developed for either subcutaneous or intravenous (IV) administration. The constant regions of the heavy and light chains of this monoclonal antibody are identical in amino acid sequence to the corresponding constant regions of the human/mouse chimeric mAb, infliximab. However, in contrast to infliximab, the heavy and light variable regions of golimumab are human sequence.
The objective of this first-in-human study was to evaluate the pharmacokinetics and safety of golimumab following single ascending IV infusions in subjects with RA.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
The study was conducted at 2 sites in the United States. All patients signed an informed consent document before participating in study-related procedures. Three institutional review boards of both participating centers approved the protocol and informed consent documents. These 3 institutional review boards were the University of Texas Southwestern Institutional Review Board, Dallas, Texas; Western Institutional Review Board, Olympia, Washington; and St Luke's Health System Institutional Review Board, Phoenix, Arizona. The study was conducted in accordance with the Declaration of Helsinki and the regulations established in the United States for the protection of human subjects (US Code of Federal Regulations Title 21, Part 56).
Subjects
Subjects older than 18 years of age were considered eligible if they had a diagnosis of RA according to American College of Rheumatology (ACR) criteria and a duration of active disease of more than 3 months from the onset of persistent synovitis. At baseline, subjects were required to have 8 or more tender joints (out of 68 joints examined) and 6 or more swollen joints (out of 66 joints examined) and either serum C-reactive protein (CRP) level at least 20% above the laboratory reference or erythrocyte sedimentation rate (ESR) of at least 28 mm/h. Subjects were required to be naive to therapy with TNF-blocking agents; however, subjects could have received no more than 3 doses of infliximab, with the last dose not less than 4 months prior to study agent administration. Subjects were allowed to be on a stable therapeutic regimen of up to 2 disease-modifying antirheumatic drugs (DMARDs) (eg, methotrexate, leflunomide, sulfasalazine, plaquenil, intramuscular gold injections), corticosteroids 
10 mg/d prednisone equivalent, and/or nonsteroidal anti-inflammatory drugs (NSAIDs) for the treatment of their arthritis. Intramuscular, intra-articular, and IV steroids were prohibited. All subjects were carefully screened for tuberculosis with both chest radiography and purified protein derivative tuberculin skin testing. Those with any findings indicating the presence of tuberculosis were excluded from participation.
Immune Response Assessment
The development of antibodies to golimumab was determined from blood samples obtained prior to administration of the study agent and following administration of study agent at weeks 2, 4, 8, 12, and 16. Antibodies to golimumab were determined using a validated antigen bridging enzyme immunoassay. First, the sera were screened, and the optical densities (ODs) of the test samples were compared with an assay cut-off OD. In addition, the baseline and posttreatment sample ODs were compared for a 2-fold OD increase. Potentially positive samples and that subject's pretreatment sample were tested for titer, and the specificity of the response for golimumab was assessed by the addition of soluble golimumab. Binding was determined to be specific for golimumab if the sample OD was reduced by at least 50% following preincubation with soluble golimumab. Each assay run also included positive and negative control samples. Samples were determined to be immune response negative when the result was below the assay cut-off OD at the screening dilution with no detectable golimumab. Samples were to be designated immune response inconclusive when the result at the screening dilution was the below assay cut-off OD and the presence of golimumab was detected in the sample.
Blood samples for the measurement of antinuclear antibodies (ANA) and anti-dsDNA (dsDNA) antibodies were collected predose and at weeks 2 and 16. Cellular immune response was evaluated by measuring delayed-type hypersensitivity (DTH) responses to the Candida albicans skin test antigen for cellular hypersensitivity (CANDIN, Allermed Laboratories, Inc, San Diego, Calif) and the mumps skin test antigen, USP (Connaught Laboratories, Inc, Swiftwater, Pa). Subjects were inoculated with the test antigens 24 hours after infusion of the study agent. Skin tests were read 48 hours following panel administration, and an induration of 
5 mm was considered to be a positive response. The ability of subjects to generate a naive antibody response was evaluated in all subjects not inoculated with pneumococcal vaccine within the prior 5 years. Subjects who had not received the vaccine within the prior 5 years were inoculated with the 23-valent pneumococcal vaccine (PNEUMOVAX-23, Merck & Company, Inc, Whitehouse Station, NJ) according to the manufacturer's instructions 72 hours after the infusion of study agent. Immune response titers were measured 4 weeks after antigen challenge. Subjects with a 2-fold increase over baseline titers against >6 of the 12 elevated pneumococcal serotypes were considered to be responders.
Safety Assessments
Subjects were monitored for 16 weeks after the single dose of study agent. All treated subjects were observed in a hospital setting for 24 hours after the infusion and underwent follow-up evaluations at 3 and 7 days and at 2, 4, 8, 12, and 16 weeks following randomization, regardless of whether the administration of study agent was prematurely discontinued. Every attempt was made to keep background medication constant throughout the 16-week study period. Under no circumstances were subjects to receive open-label infliximab or etanercept during the 16 weeks following treatment. Safety assessments included the incidence of all adverse events (AEs) and changes from baseline in vital signs, electrocardiograms, and laboratory parameters. Adverse events were monitored through 16 weeks after administration of study agent.
Samples for Pharmacokinetic Analysis
For the measurement of golimumab in the serum, blood samples were obtained prior to administration of the study agent, immediately following the completion of 2-hour infusion, and at 1, 4, 8, and 24 hours after infusion of the study agent. In addition, blood samples were obtained at day 3 and at weeks 1, 2, 4, 8, 12, and 16 after the infusion of the study agent.
Serum golimumab concentrations were determined at Centocor Research and Development, Inc using a validated enzyme-linked immunosorbent assay (ELISA) method with a lower limit of quantification of 0.078 µg/mL at the minimum 1:10 sample dilution required. The upper limit of quantification was 0.25 µg/mL. The intra-assay and interassay coefficient of variances for serum diluted in fetal bovine serum were 12.7% and 18.3%, respectively. Serum samples may be subjected to 3 freeze/thaw cycles without affecting measurable concentrations of golimumab. Samples from the 0.1- and 0.3-mg/kg dose groups were assayed at the minimally required assay dilution only. Samples from the 0.6-, 1-, 3-, 6-, and 10-mg/kg dose groups were assayed at 1:140 and 1:2000 sample dilutions only.
Noncompartmental Pharmacokinetic Analysis
Noncompartmental analysis (NCA) was employed to determine the pharmacokinetic parameters of golimumab following a single IV infusion using Win-Nonlin (Version 3.3, Pharsight Corporation, Mountain View, Calif). The maximum serum concentration (Cmax) was obtained from inspection of the individual serum concentration versus time data. The terminal elimination rate constant (
z) was determined by least squares regression analysis of the log-linear portion of the terminal phase. The area under the serum concentration-time curve (AUC) from time 0 to infinity was determined as AUC (0-tz) + Cz/z, where Cz is the last measurable serum concentration. The terminal elimination half-life (t
) was calculated from the ratio of 0.693 and z. The total body clearance (CL) was calculated by dividing the dose by the AUC.
Compartmental Pharmacokinetic Modeling
Traditional compartmental pharmacokinetic modeling was also employed to determine the parametric pharmacokinetic parameters of golimumab following a single IV infusion using WinNonlin (Version 4.01a, Pharsight Corporation). Pharmacokinetic parameters estimated by the model included the following: clearance (CL), volume of distribution in the central compartment (Vc), intercompartmental clearance (Q), and volume of distribution in the peripheral compartment (Vp).
Population Pharmacokinetic Modeling
Base Model Development
A base population pharmacokinetic model was built using the pharmacokinetic data (0.3-10 mg/kg) in this study. A total of 198 golimumab serum concentration values from 21 subjects with RA were included in the analysis. The inclusion of 0.1-mg/kg data in the data set did not result in model convergence. It was hypothesized that the lack of convergence with inclusion of 0.1-mg/kg data could be due to potentially significant impact of target (TNF)-mediated disposition of golimumab at a very low dose.6
The initial model was developed by comparing 1- and 2-compartment models without covariates by use of the NONMEM subroutines ADVAN1 TRANS2 and ADVAN3 TRANS4, respectively, with first-order conditional estimation method and interaction. The optimal structural model was selected based on the findings from this comparison in diagnostic plots and the objective function value (OFV) between the 2 models.
Modeling Stochastic Variability
During the population pharmacokinetic analysis, the jth observation (golimumab serum concentration) of the ith individual, Concij, measured at time (t) was assumed by the following equation:
 |
where f represents the structural population model, 
i is the vector of the pharmacokinetic parameters for the ith individual, and 
ij denotes the residual error from the model prediction. It was assumed that ij is symmetrically distributed around mean 0, with variance denoted by 
2.
For im, the mth element of the ith individual's parameter vector, the following model was assumed:
 |
where pop,m is the typical (mean) population parameter of the mth element, and 
im represents the shift of the parameter of the ith individual from the population mean. The parameter im was also assumed to have an independent, multivariate, normal distribution, with a mean of 0 and variance-covariance matrix of 
with diagonal elements (
, 
,..., 
) such that the 
m is approximately the coefficient of variance of the mth parameter with respect to the typical value, pop,m.
For residual error in the population pharmacokinetic analysis, additive, proportional, and combined additive and proportional random error models were examined.
Covariate Model Development
The apparent influence of subject covariates on golimumab disposition was analyzed by use of the individual Bayesian pharmacokinetic parameter estimates. Visual inspection of any potential correlation was the precursor to the formal statistical testing of a certain covariate in the model. The optimal pharmacokinetic model was chosen based on goodness of fit, population-predicted concentrations versus observed concentrations, decrease in interindividual variability (IIV), and decrease in the value of the objective function value (OFV). The OFV was used for hypothesis testing to discriminate among alternative hierarchical models. The influence of a certain covariate was considered significant only if the difference in the OFV between 2 hierarchical models was greater than 11 (P < .001, 
2, 1 degree of freedom), which is aligned with that recommended in the guidance for population pharmacokinetics issued by the Food and Drug Administration in 1999.7
The influence of demographic covariates on the pharmacokinetics of golimumab was explored with the use of locally weighted regression smoothing (loess regression) and by conventional linear regression to test for nonlinear and linear trends, respectively.
The predictive performance of both the structuralbased model and the final covariate model was validated using the bootstrapping approach. The bootstrapping method is a validation tool for estimating generalization error based on a "resampling" technique. The model was fitted to the 1500 bootstrapped samples to evaluate its stability and performance.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
|
Among all randomized subjects, the median numbers of tender and swollen joints at baseline were 25 and 19, respectively. The median baseline scores for pain, the evaluator's global assessment of disease activity, and the patient's global assessment of disease activity were 5.2, 6.5, and 5.4 cm, respectively, on a visual analog scale. The median baseline Health Assessment Questionnaire disability index was 1.188, the median baseline CRP level was 1.5 mg/dL, and the median baseline ESR was 36 mm/h.
|
of 19.3 days for the 10-mg/kg dose group. The median half-life appeared to increase with an increase in dose, with a median t of 6.6 days for 0.1 mg/kg, 7.3 days for 0.3 mg/kg, 8.6 days for 1 mg/kg, 11.2 days for 3 mg/kg, 14.5 days for 6 mg/kg, and 19.3 days for 10 mg/kg, as shown in Table II and illustrated in Figure 1. The shorter t values observed in the lower dose groups were the result of the majority of the terminal phases being below the assay detection limit (lower limit of quantification of 0.078 µg/mL). As a result, the t values estimated from the lower dose groups were not representative of the true t value. In view of that, population pharmacokinetic modeling was conducted to obtain the typical population pharmacokinetic parameters (including t).
|
Table II summarizes the pharmacokinetic parameters that were derived using the noncompartmental pharmacokinetic analysis approach.
The mean Cmax and AUC increased in a doseproportional manner. A 10-fold increase in dose resulted in an approximately 8- to 11-fold increase in mean Cmax and an approximately 8- to 13-fold increase in mean AUC, as illustrated in Figure 2. The mean CL after IV administration of golimumab ranged from 4.89 to 6.72 mL/d/kg; the CL of golimumab appeared to be independent of dose at a range of 0.1 to 10 mg/kg.
|
Population Pharmacokinetic Analysis
Base Model Development
The 2-compartment model with first-order elimination and zero-order input (IV) was superior to the 1-compartment model, as evidenced by a significant drop in OFV (
OFV = -260.9; 2, P < .0001), whereas analysis on the basis of a 3-compartment model did not improve the OFV any further. Hence, the 2-compartment model was selected for further model building.
Figure 3 shows the goodness-of-fit plots of predicted versus observed data over the whole data set, regardless of time of sampling. A good fit was shown between the observed and the population or the individual predicted serum golimumab concentrations (Figure 3, top panels). Also, no apparent systemic bias was found in the weighted residual plots (Figure 3, bottom panels), which confirmed that the base model adequately described the data.
|
|
Base Model Validation Using the Bootstrapping Approach
A bootstrap approach was used to validate the developed model. The typical values of the population pharmacokinetic parameter estimates of the developed model all fell within the 2.5th and 97.5th percentiles of the respective values generated by bootstrapping and were very close to the 50th percentile, indicating that the predictive performance of the developed base population pharmacokinetic model of golimumab was sufficient (Table III).
Covariate Model Development
After we established the base structural pharmacokinetic model, we evaluated the impact of body weight and age on the pharmacokinetic parameters (CL or Vc). Because the majority of the subjects were white women, race and sex were not included in the covariate analysis. Upon visual inspection, it appeared that there was no apparent correlation between the Bayesian-estimated CL or Vc values and age. However, there appeared to be positive relationships between Bayesian-estimated Vc or CL values and weight. The incorporation of weight in Vc significantly improved the model with a decrease of 11.0 in OFV (OFV = -11.0; 2, P < .001, 1 degree of freedom), a decrease in the IIV of Vc from 33.5% in the base model to 25.5%, and a decrease in the IIV of Q from 50.2% in the base model to 44.6%. However, the incorporation of weight in Vc did not further decrease the intraindividual residual variability.
The incorporation of weight in CL only slightly improved the model, with a decrease of 4.28 in OFV (OFV = -4.28; 2, P < .05, 1 degree of freedom). Therefore, the relationship between weight and CL was not retained in the covariate model. The model that incorporated weight in Vc was selected as the covariate model.
Covariate Model Validation Using the Bootstrapping Approach
Similarly, the bootstrapping approach was used to validate the developed covariate model. The typical values of the population pharmacokinetic parameter estimates of the developed model all fell within the 2.5th and 97.5th percentiles of the respective values generated by bootstrapping and were very close to the 50th percentile, indicating that the predictive performance of the developed covariate population pharmacokinetic model of golimumab was sufficient (Table IV).
|
The diagnostic plots, as shown in Figure 4, also demonstrated that the current covariate model adequately described the data. Hence, this model was selected as the final covariate model for golimumab based on the pharmacokinetic data from the current study.
|
|
One subject in the 3-mg/kg dose cohort experienced 6 serious adverse events (hypesthesia, arterial occlusion [3 incidences], infection at an old surgical site, and claudication) during the study. This 67-year-old Caucasian woman had a 4-year history of RA. Other relevant medical history at baseline included intermittent claudication, which was treated with cilostazol. On day 24 of the study, the subject presented with symptoms of claudication involving the right leg and foot. An arteriogram revealed preocclusive stenosis of the right common iliac artery, for which angioplasty with stent placement was performed. On day 34 of the study, the subject complained of hypesthesia, this time of the left leg. She was diagnosed with an occlusion of the left iliac artery, possibly related to proximal positioning and/or proximal migration of the right common iliac stent. This was treated with thrombolysis (tPA) and insertion of a stent in the left common iliac artery. At routine checkup, on day 51 of the study, the subject presented with a reocclusion at the origin of the left iliac artery. The stent was found to be blocked, and a stent replacement in the left common iliac artery with angioplasty was performed. On day 76 of the study, the subject developed subsequent occlusion of the right common iliac artery, which required thrombolysis and dilatation at the site of the previously placed stent. After repeated occlusions of both common iliac arteries, the subject was readmitted to the hospital and finally underwent a femoral-to-femoral bypass graft. The postoperative course was complicated with further occlusion of the left common iliac artery. Therefore, an axillobifemoral component was added to the graft from the right axillary artery. The clot within the graft was sent for analysis and appeared to represent a normal clot without evidence of heparin-induced thrombocytopenia. The subject's hypercoagulability profile was positive for cardiolipin antibody and showed diminished protein S (measured at 50%, with the normal range being 70% to 150%). Her homocysteine was measured at 5.3 µmol/L, which was also slightly below the normal range of 5.4 to 11.9 µmol/L. On day 79 of the study, the subject was readmitted to the hospital with postoperative infection of both groin wounds, which was treated with intravenous imipenem and cilastatin and topical sulfadiazine. The subject withdrew consent for further participation in the study on day 111 because of these serious AEs and was not available for the week 16 final study procedures. The serious AEs were determined by the investigator to be not related to golimumab but more likely related to other concurrent illness. No other subjects reported serious adverse events.
There were no substantial differences in the study between subjects who received placebo and any golimumab in the incidence of infections or infusion reactions or in the hematology, clinical chemistry, vital sign, or electrocardiogram measurements.
Cellular immune response was evaluated by measuring DTH responses to C albicans skin test antigen and mumps skin test antigen. Overall, the numbers of subjects with positive responses to DTH antigens were generally low and similar to those receiving placebo. For subjects receiving placebo, 2 of 10 subjects (20.0%) had a positive response to the C albicans antigen, and 3 of 10 subjects (30.0%) had a positive response to the mumps antigen. Among the golimumab-treated patients, 7 of 26 subjects (27%) showed positive responses to the C albicans antigen, and 6 of 26 subjects (23%) responded to the mumps antigen. Humoral response to polyvalent pneumococcal vaccine was also evaluated. Although the response was low in placebo-treated subjects (9 of 10 subjects [90%]), the response was higher in subjects treated with golimumab. Eleven of 26 (42%) golimumabtreated subjects mounted an effective 2-fold increase in titer to polyvalent pneumococcal vaccine, indicating that IV golimumab did not interfere with B cell-dependent humoral responses. Overall, the subjects enrolled in this study appeared to have compromised cellular and humoral immune responses as measured by DTH testing and changes in titer to pneumococcal vaccine. These low titer immune responses did not appear to be further impaired by treatment with golimumab.
There were slightly higher percentages of subjects with newly positive ANA results among subjects in golimumab dose cohorts (7 of 26 subjects [27%]) than among subjects who received placebo (1 of 10 subjects [10.0%]). There were no subjects in the study who were newly positive for anti-dsDNA antibodies.
All golimumab-treated subjects had appropriate samples taken for antibody response against golimumab (ie, at least 1 sample collected after study agent administration). Antibodies against golimumab were detected in 1 subject from each of the 0.1-, 3-, and 10-mg/kg dose cohorts. Titers were detectable at a 1:40 dilution in the subjects from the 0.1-mg/kg and 3-mg/kg cohorts, as well as at a 1:10 dilution in the subject from the 10-mg/kg cohort. The presence of antibodies to golimumab in these subjects did not correlate with the incidence of any AE or affect the pharmacokinetic profile.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Both placebo and golimumab treatment groups in these study subjects had lower cellular and humoral immune responses, as measured by DTH testing and changes in titer to pneumococcal vaccine. The low incidence of DTH reactivity in these subjects is in marked contrast to studies demonstrating that nearly 100% of healthy volunteers are reactive to 2 or more antigens.8 However, Westhovens et al9 reported low humoral immune response in a cohort of RA subjects treated with IV infliximab. Therefore, the low DTH response rate found in this study may be attributable to widespread concomitant immunosuppressive therapy in patients with RA. Therefore, no meaningful conclusion can be drawn from this small study cohort.
Treatment with anti-TNF- agents may result in the formation of antibodies against the therapeutic agent. In particular, the formation of antibodies may lead to increased clearance or delayed hypersensitivity reactions, which may limit the use of the drug. In this study, 3 of 26 subjects developed a low-titer immune response to antibodies to golimumab. However, the development of these antibodies did not correlate with any clinical adverse event or affect the pharmacokinetic profile. Regardless, we cannot make any firm conclusions about the immunogenicity of golimumab based on the limited number of subjects and the single-dose treatment in this study. Future studies with repeated dosing and administration by other routes will be needed to further investigate the immunogenicity of this agent.
|
Vc was estimated to be 3.07 L for a subject with a body weight of 72.3 kg, which is similar to that of IgG.10 The volume of distribution at steady state (Vss), which was estimated as the sum of Vc and Vp, was 6.91 L, indicating that golimumab resided in both the vascular and extravascular spaces. Body weight was identified as a significant covariate for Vc, which was described by a power function. Body weight was also shown to be a significant covariate for the Vc of adalimumab.11 Although statistically significant, the clinical relevance of this relationship between body weight and Vc is unlikely to be important. This relationship will be further evaluated in the phases II and III studies. No dose adjustment according to body weight appeared to be necessary based on the current finding.
The results from the bootstrapping procedure showed that parameters in the final pharmacokinetic model were estimated with acceptable accuracy, although a high IIV was observed for all pharmacokinetic parameters. However, the high IIV was not unexpected for an IgG-based monoclonal antibody.10,12 The inclusion of weight in Vc decreased the IIV of Vc from 33.5% in the base model to 25.5%. The developed model had a relatively small residual variability of approximately 15%, suggesting that the developed models (both base and covariate models) possessed reasonably high predictability.
Population modeling approach was originally developed to process the sparse pharmacokinetic data collected during routine clinical visits in observational/longitudinal (phase III or IV) clinical studies. Nevertheless, due to its flexibility and robustness in its algorism, this approach has been applied far beyond its original scope. For example, it has also been used to describe rich data from phase I or phase IIa studies.13-17 To substantiate the population pharmacokinetic modeling results in the current study, traditional individual compartmental modeling was also conducted. The individual compartmental modeling yielded results similar to those via the population modeling. The intersubject variability, as represented by CV%, however, appeared to be significantly bigger for the compartmental modeling: CL (46.0%), Vc (37.3%), and Vp (67.0%), compared to the population modeling: CL (10.0%), Vc (7.5%), and Vp (11.8%). This finding was not surprising because the population modeling approach partitioned the total error into intersubject error and intrasubject error (residual error), whereas the traditional compartmental modeling approach did not. The population modeling approach, which is population based, shows superiority over traditional compartmental modeling, which is individual based, if Monte Carlo simulations need to be performed to support further clinical development.
Detectable concentrations of golimumab were observed in the majority of subjects through week 4, even those from the lower dose groups (0.1-1 mg/kg). Detectable concentrations of golimumab were still observed at week 16 (the last pharmacokinetic sampling time point) in 1 of 5 subjects in the 6-mg/kg dose cohort and in 2 of 5 subjects in the 10-mg/kg dose cohort. Based on the noncompartmental analysis, the median t ranged from 6.56 to 19.25 days after a single IV infusion over the dose range of 0.1 to 10 mg/kg. The median t appeared to be shorter (6.56-8.58 days) in subjects receiving the lower doses of 0.1 to 1.0 mg/kg compared with those of the subjects receiving the higher doses of 3 to 10 mg/kg (11.20-19.25 days). This observation was likely the result of an insufficient characterization of the terminal elimination phase of the lower dose group. Drug exposure (Cmax and AUC) increased in an approximately dose-proportional manner over the dose range studied, with large variability observed. The Bayesian estimates of t (25th to 75th percentiles: 12.6-21.0 days), based on the population pharmacokinetic model, also suggested that the t of golimumab ranged from 2 to 3 weeks. This may allow for infrequent dosing of golimumab in further clinical testing.
In summary, the pharmacokinetic characteristics of golimumab, combined with the safety profile observed in this single-dose IV study in subjects with RA, suggest that this fully human anti-TNF- monoclonal antibody may be a therapeutic agent for further development in subjects who are candidates for anti-TNF- therapy.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Financial disclosure: This study was funded by Centocor Research and Development, Inc.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
1. Beutler B. Tumor Necrosis Factors: The Molecules and Their Emerging Role in Medicine. New York: Raven Press; 1992.
2. Brennan FM, Maini RN, Feldmann M. TNF alpha—a pivotal role in rheumatoid arthritis? Br J Rheumatol. 1992;31: 293-298.
3. Santos Lacomba M, Marcos Martin C, Gallardo Galera JM, et al. Aqueous humor and serum tumor necrosis factor-alpha in clinical uveitis. Ophthalmic Res. 2001;33: 251-255.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
4. Braegger CP, Nicholls S, Murch SH, Stephens S, MacDonald TT. Tumour necrosis factor alpha in stool as a marker of intestinal inflammation. Lancet. 1992;339: 89-91.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
5. Russo C, Polosa R. TNF-alpha as a promising therapeutic target in chronic asthma: a lesson from rheumatoid arthritis. Clin Sci (Lond). 2005;109: 135-142.[Medline] [Order article via Infotrieve]
6. Levy G. Pharmacologic target-mediated drug disposition. Clin Pharmacol Ther. 1994;56: 248-252.[Web of Science][Medline] [Order article via Infotrieve]
7. US Department of Health and Human Services FaDA, Center for Drug Evaluation and Research (CDER), Center for Biologics Evaluation and Research (CBER). Guidance for Industry: Population Pharmacokinetics. Rockville, Md: US Department of Health and Human Services; 1999.
8. Birx DL, Brundage J, Larson K, et al. The prognostic utility of delayed-type hypersensitivity skin testing in the evaluation of HIV-infected patients. Military Medical Consortium for Applied Retroviral Research. J Acquir Immune Defic Syndr. 1993;6: 1248-1257.
9. Westhovens R, Houssiau F, Joly J, et al. A phase I study assessing the safety, clinical response, and pharmacokinetics of an experimental infliximab formulation for subcutaneous or intramuscular administration in patients with rheumatoid arthritis. J Rheumatol. 2006;33: 847-853.
10. Koleba T, Ensom MH. Pharmacokinetics of intravenous immunoglobulin: a systematic review. Pharmacotherapy. 2006; 26: 813-827.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
11. Velagapudi RB, Noertersheuser PA, Awni WM, et al. Effect of methotrexate (mtx) coadministration on the pharmacokinetics (pk) of adalimumab (Humira, Abbott) following a single intravenous (iv) injection. Arthritis Rheum. 2003;48: S141.[CrossRef]
12. Lobo ED, Hansen RJ, Balthasar JP. Antibody pharmacokinetics and pharmacodynamics. J Pharm Sci. 2004;93: 2645-2668.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
13. Roshammar D, Hai TN, Friberg Hietala S, Van Huong N, Ashton M. Pharmacokinetics of piperaquine after repeat oral administration of the antimalarial combination CV8 in 12 healthy male subjects. Eur J Clin Pharmacol. 2006;62: 335-341.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
14. Takama H, Tanaka H, Sudo T, Tamura T, Tanigawara Y. Population pharmacokinetic modeling and model validation of a spicamycin derivative, KRN5500, in phase I study. Cancer Chemother Pharmacol. 2001;47: 404-410.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
15. Zhou H, Khalilieh S, Lau H, et al. Effect of meal timing not critical for the pharmacokinetics of tegaserod (HTF 919). J Clin Pharmacol. 1999;39: 911-919.[Abstract]
16. Gisleskog PO, Hermann D, Hammarlund-Udenaes M, Karlsson MO. The pharmacokinetic modeling of GI198745 (dutasteride), a compound with parallel linear and nonlinear elimination. Br J Clin Pharmacol. 1999;47: 53-58.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
17. Launay-Iliadis MC, Bruno R, Cosson V, et al. Population pharmacokinetics of docetaxel during phase I studies using nonlinear mixed-effect modeling and nonparametric maximum-likelihood estimation. Cancer Chemother Pharmacol. 1995;37: 47-54.[Web of Science][Medline]
[Order article via Infotrieve]
CiteULike 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. E. Wenzel, P. J. Barnes, E. R. Bleecker, J. Bousquet, W. Busse, S.-E. Dahlen, S. T. Holgate, D. A. Meyers, K. F. Rabe, A. Antczak, et al. A Randomized, Double-blind, Placebo-controlled Study of Tumor Necrosis Factor-{alpha} Blockade in Severe Persistent Asthma Am. J. Respir. Crit. Care Med., April 1, 2009; 179(7): 549 - 558. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Xu, K. Seitz, A. Fasanmade, J. Ford, P. Williamson, W. Xu, H. M. Davis, and H. Zhou Population Pharmacokinetics of Infliximab in Patients With Ankylosing Spondylitis J. Clin. Pharmacol., June 1, 2008; 48(6): 681 - 695. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
