Abstract Background: HER3 is a member of the Human Epidermal Growth Factor Receptor (HER) family. Although HER3 lacks intrinsic kinase activity, it serves as a scaffold for PI3K/AKT signaling for the HER family via heterodimeric interactions. The goal of this study was to use preclinical modeling to predict a minimal effective dose regimen for objective response using preclinical PK and BxPC3 xenograft mice anti-tumor efficacy and PD data. Materials and Methods: U3-1287 (AMG 888) concentration-time data obtained from mice (0.025 to 2.5 mg/mouse), rats (1 to 100 mg/kg) and monkeys (1 to 200 mg/kg) were combined and analyzed using a target-mediated drug disposition model. Animal PK parameters were scaled based on body weight to predict human PK characteristics. An Emax model was used to relate drug concentration and inhibition of pHER (measure by ELISA) in BxPC3 xenograft tumors. A PK/PD/efficacy model (based on Simeoni et al 2004) was used to analyze tumor growth data from mice bearing BxPC3 pancreatic xenografts (∼200 mm3) treated twice per week at 25, 100, 200, or 500 g for a 1 month. The model was validated with tumor growth data following additional doses of 400 g biweekly and 200 g biweekly, weekly and twice a week. The relationship between drug concentration, the inhibition of pHER3 in animals and interspecies PK scaling was used to select the minimal effective dose for the first in human study. Results: U3-1287 (AMG 888) treatment of BxPC3 xenografts resulted in a statistically significant inhibition of tumor growth and pHER3 levels in a dose and schedule dependent manner (p<0.05). Treatment with U3-1287 (AMG 888) at 400 g/mouse biweekly and 200 g/mouse biweekly, weekly and twice a week resulted in a 50%, 33%, 74% and 70% inhibition of tumor growth (p<0.05), a 70%, 42%, 77% and 80% inhibition of pHER3 (measured by ELISA) versus the IgG control treated group, respectively. Serum concentrations of U3-1287 (AMG 888) at necropsy for the respective dose groups were (mean (SD)) of 2.07 (0.97), 0.45 (0.21), 3.08 (0.82) and 34.9 (9.1) g/mL, respectively. The estimated trough concentration needed to achieve 90% maximal pHER3 inhibition (IC90) was estimated to be ∼3 µg/mL. Based on interspecies scaling of PK parameters, a dose higher than 3 mg/kg given biweekly would achieve a steady-sate trough concentration that is above 3 µg/mL. Conclusion: The anti-tumor efficacy in the BxPC3 pancreatic xenograft model was correlated with an increased serum concentration of U3-1287 (AMG 888) and a decrease in pHER3 levels, allowing for the development of a PK/PD/Efficacy relationship. This relationship combined with interspecies PK scaling was used to determine a dose and schedule for U3-1287 (AMG 888) to investigate in a FIH study. Citation Information: Mol Cancer Ther 2009;8(12 Suppl):B167.
Adalimumab (Humira®) and its approved biosimilar ABP 501 (AMGEVITA®; adalimumab) exhibit nonlinear pharmacokinetics (PK) following a single subcutaneous (SC) dose in healthy volunteers possibly due to target mediated disposition or development of anti-drug antibodies (ADAs). The presence of nonlinear PK leads to low and variable drug levels posing a challenge in assessing PK equivalence of these two agents. To address this we used a population PK modelling approach to assess PK equivalence. A one-compartment linear PK model with first-order absorption was selected to characterise the ABP 501 and adalimumab PK in healthy subjects. The model was parameterised in terms of apparent systemic clearance and apparent central volume of distribution for subjects with or without ADAs. Body weight, albumin and ADA status were evaluated for their potential impact on adalimumab or ABP 501 PK. For subjects who developed ADAs, the model evaluated the inclusion of additional antibody mitigated clearance mechanisms using Michaelis–Menten (MM) type of saturable clearance or time-dependent changes in linear clearance. The effect of treatment on PK parameters was evaluated by comparisons of the empirical Bayes' estimates of the individual PK model parameters. In healthy subjects without ADAs, the ABP 501 and adalimumab PK after SC administration was linear and adequately described by a one-compartment model with first-order absorption and linear clearance from the central compartment. Inclusion of an additive time dependent linear clearance for subjects with ADAs was superior to a model incorporating an additive MM saturable clearance for subjects with ADAs. Graphical and statistical comparisons of empirical Bayes' estimates by treatment demonstrated no difference in any PK model parameters by treatment. The population and inter-individual variability estimates of clearance and volume were 0.399 (33.6%) l/day and 8.94 (23.5%) l, for a 72.7 kg subject (median body weight). The time to onset and magnitude of the additive ADA-related linear clearance was 33.7 (54.4%) days and 1.15 (52.6%) l/day. Diagnostic plots demonstrated good concordance between observed and population or individual predicted concentrations without any bias. Visual predictive checks demonstrated the model described well the variable PK following a single dose of ABP 501 or adalimumab. The model described the PK of each individual well and was used to estimate accurately each individual's area under the concentration curve (time 0 to infinity) for all 203 subjects who completed the study. The present population PK modelling confirms PK similarity of ABP 501 to adalimumab.
Abstract Denosumab (XGEVA®) is a recombinant, fully human IgG 2 monoclonal antibody directed against the receptor activator of nuclear factor kappa‐B ligand (RANKL) that prevents differentiation of osteoclast precursors into mature osteoclasts and acceleration of bone resorption, resulting in the inhibition of osteoclast activation. Denosumab is indicated for the prevention of skeletal‐related events (SREs) in adult patients with bone metastases from solid tumors at the dose of 120 mg administered subcutaneously (SC) every 4 weeks. This review is focused on describing its target‐mediated disposition and direct inhibitory effect on bone resorption, as well as the modeling and simulation techniques used to integrate the PKPD information collected during clinical development of denosumab. In addition, this review further discusses the clinical relevance of patient covariate effects on denosumab systemic exposure, target engagement and downstream pharmacodynamics biomarkers, and the rationale for dosing regimen selection for Phase 3 studies. Phase 3 clinical studies demonstrated that denosumab was superior to zoledronic acid in inhibiting bone resorption and, consequently, delaying the time to first SRE by a median of 8.2 months in patients with bone metastases from solid tumors. Thus, denosumab may be considered a better alternative treatment than zoledronic acid for the prevention of SRE in patients with bone metastases from solid tumors.
2535 Background: R is an investigational, fully human monoclonal antibody to hepatocyte growth factor/scatter factor (HGF/SF), the only known MET receptor ligand. In a double-blind, placebo-controlled, phase 2 trial of 121 pts with G/EGJ cancer, R (7.5 or 15 mg/kg) + ECX showed trends for improved overall survival (OS) and progression-free survival (PFS) compared to ECX alone (placebo, P). Methods: E-R analyses were performed using pharmacokinetic (PK), biomarker, efficacy, and safety data. A population PK model was used to evaluate covariate effects (eg, demographics, ECX, and MET status) on R PK and predict individual R exposure (trough steady-state C min [C]). The effect of C on OS/PFS was evaluated with Kaplan-Meier estimates and Cox proportional hazards models. Covariate effects for OS/PFS were evaluated with multivariate analysis. The relationships between adverse events of interest (AEs), lab values, and C were analyzed with descriptive statistics and linear regression models. Results: R showed linear PKs that were unaffected by ECX or MET levels. Pts who received R + ECX were divided into low C (C L ) and high C (C H ) groups by the median C (94 µg/mL). Median PFS was 4.2, 4.5, and 6.9 months in the P, C L , and C H groups; median OS was 8.9, 9.5, and 13.3 months. Improved OS/PFS in C H and improved PFS in C H and C L groups (vs P) were seen after adjusting for baseline covariates. Pts with tumors that expressed high MET levels (MET High : > 50% cells positive) showed greater improvement in OS in the C H vs C L group. No improvement in OS was seen in MET Low pts (C H and C L groups) compared to placebo (see Table). No relationship was seen between R exposure and AEs or lab values. Conclusions: Higher R exposure was associated with improved OS/PFS in MET High pts without increased toxicity, supporting further study of R + ECX in G/EGJ cancer. [Table: see text]
Rilotumumab is a fully human monoclonal antibody investigated for the treatment of MET-positive gastric cancer. The aim of this study was to evaluate the potential pharmacokinetic (PK)-based drug-drug interaction (DDI) between rilotumumab and epirubicin (E), cisplatin(C) and capecitabine (X).This was a Phase 3 double-blind, placebo-controlled study, in which rilotumumab, epirubicin and cisplatin were administered intravenously at 15 mg kg-1 , 50 mg m-2 , and 60 mg m-2 Q3W, respectively, while capecitabine was given orally at 625 mg m-2 twice daily. Rilotumumab PK samples were taken at pre-dose and at the end-of-infusion from all patients in cycles 1, 3, 5 and 7. ECX PK samples were taken in cycle 3 from patients who participated in the intensive PK assessment. ECX PK was assessed by non-compartmental (NCA) analyses and PK parameters were compared between two arms. Rilotumumab PK was assessed by comparing the observed rilotumumab serum concentrations with model-predicted concentrations using a population PK model developed from previous Phase 1 and Phase 2 studies.The study enrolled 609 patients. ECX plasma concentrations in the presence and absence of rilotumumab were similar, as demonstrated by the geometric mean ratios for Cmax and AUC, which were close to 1.0, suggesting ECX PK was not affected by co-administration of rilotumumab. The observed rilotumumab serum concentrations were similar to the values predicted by population PK modelling on the basis of a prediction-corrected visual predictive check, indicating rilotumumab exposure was not affected by co-administration of ECX.The results suggest lack of PK-based DDI between rilotumumab and ECX.
