Population pharmacokinetics and pharmacokinetic‐pharmacodynamic relationships of the novel anticancer agent E7070 in four phase I studies

Abstract

E7070 is a novel chloroindolyl sulphonamide being developed as an anti‐cancer agent for the treatment of solid tumours. E7070 arrests cell cycle progression in the G1 phase and induces apoptosis. In various in vitro and in vivo models, E7070 appeared to be highly active. Four phase I dose‐escalation studies have been conducted to identify the dose‐limiting toxicities (DLTs) and the maximum tolerated dose. Adult patients with advanced solid tumours resistant or not amenable to conventional therapy received E7070 at four different treatment schedules: 1) 1 h i.v. infusion (doses 50–1000 mg m−2), 2) daily times five 1 h i.v. infusion (d×5) (10–200 mg m−2 day−1), 3) weekly times four 1 h infusion (40–500 mg m−2 week−1), 4) 120 h continuous i.v. infusion (civ) (6–200 mg m−2 day−1). DLTs were haematological in all studies. Pharmacokinetic (PK) sampling was performed in 143 patients (corresponding numbers of patients were 40, 35, 43 and 25) during the first cycle. In this study we investigated the PK and its relationship with the pharmacodynamics for all studies. A population PK model was developed to describe the combined data of the four studies. E7070 showed non‐linear PK at higher dose levels. Therefore, a basic 3‐compartmental model was extended with saturable transport to one of the peripheral compartments and parallel linear and a non‐linear pathways of elimination from the central compartment. The model adequately described the data. A validation using a bootstrap procedure indicated that the model was robust. Total clearance (CL) of E7070 was highly dependent on the plasma concentration; shortly after infusion, at the higher concentrations (i.e. 10–200 mg l−1), CL was 0.5 l h−1 and gradually increased to 5.5 l h−1 at the lower concentrations (0.1–0.01 mg l−1). For each individual, Bayesian estimates of AUC (area under the plasma concentration time curve) were obtained using the PK model. The percentage decreases in neutrophils (ANC) and thrombocytes (THR) were plotted against dose, dose level and AUC for each study. In general, AUC was a better predictor for haematological toxicity than dose and dose level, as variability was reduced when related to AUC. The relationship between % decrease in ANC and THR and AUC could be described using a maximum effect (Emax) model. A maximum % decrease in ANC and THR of approximately 95% was reached with all four schedules.In conclusion, the non‐linear PK of E7070 in all treatment schedules was adequately described using the population PK model. The dose‐limiting, haematological toxicity of E7070 is related to the total exposure, expressed as AUC.