L.S. Murray, S. Senan1, J. Cassidy(1), P. Workman(1), S.B. Kaye(1) and M.A. Graham(1)
Department of Medicine & Therapeutics, University of Glasgow and (1)Clinical Pharmacology and New Drug Development Group, CRC Department of Medical Oncology, University of Glasgow.
Eight males and four females, aged 38-61 years, were treated with the novel hypoxic cell cytotoxic agent, tirapazamine, in a phase I dose ranging study. Cohorts of three patients were treated at each of 36, 72, 120 and 180 mg/m2 by an intravenous infusion given at 3.5mg/min. Concentrations were measured mid-infusion and at 0, 5, 10, 15, 20, 30, 45, 60, 120, 240 and 360 minutes after the end of the infusion. The data were then analysed by NONMEM. A 2-compartment model with proportional residual variance gave a significantly better fit than a 1-compartment model (chi2 = 36, 2df, p<0.001). The mean parameter estimates (standard deviation) were:- clearance = .77(.20)l/min, Vc = 25.3(10.4)1, Q = .21(.23)l/min and Vp = 18.0 (16.3)1. Modelling clearance as a linear function of body surface area did not improve the fit. The effect of other covariates on the parameter estimates needs to be investigated further.
The dose ranging study aims to identify a range of concentrations where toxicity to the drug is tolerable and therapeutic effect maximised. In therapeutic use it is impractical, unethical and costly to sample frequently. Thus a further aim of this research is to use the population model parameters to derive a limited sampling strategy for use in phase II investigations and beyond. The measured concentrations from such a strategy can then be used to adaptively control serum levels to achieve target AUCS or concentrations with subsequent courses of the drug; In addition correlations between concentrations and therapeutic effects may then be made in large scale studies.
Reference: PAGE 2 (1993) Abstr 908 [www.page-meeting.org/?abstract=908]
Poster: oral presentation