Population Pharmacokinetics of Humanized Monoclonal Antibody HuCC49ΔCH2 and Murine Antibody CC49 in Colorectal Cancer Patients

Abstract

To predict the optimal time for surgery after antibody administration, the population pharmacokinetics of (125)I-HuCC49deltaCH2 and (125)I-CC49 were characterized in 55 patients with colorectal cancers. A 2-compartment linear model was used to fit the pharmacokinetic data. Model stability and performance were assessed using a visual predictive check procedure. Different clinical trial designs were evaluated by simulation in combination with Bayesian estimation method to predict the optimal time for surgery. The results showed that HuCC49deltaCH2 had 65% faster clearance from blood circulation and 24% shorter mean residence time than CC49. Population pharmacokinetic analysis identified body weight as the only covariate to explain between-subject variability in clearance, intercompartmental flow rate, and volume of distribution. Model predictions indicated a wide interval for the optimal time of surgery, suggesting that it would be beneficial to individualize the time of surgery for each patient by measurement of antibody disposition. Clinical trial designs with at least 3 measurements of antibody disposition were found to be better than an empirical direct observation method for the optimal prediction of surgery time.

Population pharmacokinetic data analysis of three phase I studies of matuzumab, a humanised anti-EGFR monoclonal antibody in clinical cancer development

“…With the relatively small residual variability (13.4% CV for the proportional part and a fixed additive error of 0.312 mg l À1 ), it can be suggested that the developed model possess reasonably high predictability. IOV has been rarely investigated in mAb research, but in more recent population PK analysis, it has been included to improve the model (Kloft et al, 2004;Fang et al, 2007). The estimated IOV of matuzumab (23% CV; RSE, 13%) was in the range or slightly higher than that for other immunologicals: sibrotuzumab, humanised antibody HuCC49CH2 and etanercept showed 13, 11 and 28% CV, as well as RSE imprecisions of 25 and 102% (not reported for etanercept), respectively Kloft et al, 2004;Fang et al, 2007).…”

“…IOV has been rarely investigated in mAb research, but in more recent population PK analysis, it has been included to improve the model (Kloft et al, 2004;Fang et al, 2007). The estimated IOV of matuzumab (23% CV; RSE, 13%) was in the range or slightly higher than that for other immunologicals: sibrotuzumab, humanised antibody HuCC49CH2 and etanercept showed 13, 11 and 28% CV, as well as RSE imprecisions of 25 and 102% (not reported for etanercept), respectively Kloft et al, 2004;Fang et al, 2007). The importance of implementing IOV in population PK analysis has been demonstrated (Karlsson and Sheiner, 1993) and investigation of IOV avoids biased population parameter estimates.…”

“…Comparing a modified therapeutic antibody (different Fc part) with its original antibody, the resulting half-life for the modified antibody was shorter. It could be assumed that the modified antibody might less be able to bind to the FcRn and thus, was less protected from elimination (Fang et al, 2007).…”

Population Pharmacokinetics of Therapeutic Monoclonal Antibodies: Examples and Estimation Method Performance Differences

“…In some cases a population PK/PD analysis was performed, and these are noted in Table 1-1. The only non-therapeutic mAb included in the tables is HuCC49∆CH2, which was investigated in patients undergoing radioimmunoguided surgery (17). The structural characteristics and pharmacologic target of the different mAbs are listed in Table 1-1, as well as the patient population the study was conducted in, and the model used to describe the PK of the mAb.…”

“…Other software and methods used in a couple of studies (21,22) included a conditional first-order method implemented in Winnonmix (Pharsight corporation, Mountain View, NC), and the Monte Carlo parametric expectation maximization (MCPEM) method implemented in an augmented version of ADAPT II (Biomedical Simulations Resource, University of Southern California, Los Angeles, CA). (26) Chimeric mouse/human IgG1 CD25 Renal transplant 2C L Basiliximab (27) Chimeric mouse/human IgG1 CD25 Liver transplant 2C L Bevacizumab (28) CDR-grafted mouse/human IgG1 VEGF Various solid tumors and colorectal, 2C L non-small cell lung, and breast cancer Cetuximab (29) Chimeric mouse/ (17) CDR-grafted mouse/human TAG-72 Colorectal cancer 2C L IgG1-CH2 domain deletion Infliximab (36) Chimeric mouse/human IgG1 TNF-α Ankylosing spondylitis 2C L Infliximab (37) Chimeric mouse/human IgG1 TNF-α Ulcerative colitis 2C L Inolimomab (38) †…”