• Early tumor drug pharmacokinetics is influenced by tumor perfusion but not plasma drug exposure.

      Saleem, Azeem; Price, Patricia M; Academic Department of Radiation Oncology, The Christie Hospital NHS Foundation Trust, Manchester. azeem.saleem@manchester.ac.uk (2008-12-15)
      PURPOSE: Pharmacokinetic parameters derived from plasma sampling are used as a surrogate of tumor pharmacokinetics. However, pharmacokinetics-modulating strategies do not always result in increased therapeutic efficacy. Nonsurrogacy of plasma kinetics may be due to tissue-specific factors such as tumor perfusion. EXPERIMENTAL DESIGN: To assess the impact of tumor perfusion and plasma drug exposure on tumor pharmacokinetics, positron emission tomography studies were done with oxygen-15 radiolabeled water in 12 patients, with 6 patients undergoing positron emission tomography studies with carbon-11 radiolabeled N-[2-(dimethylamino)ethyl]acridine-4-carboxamide and the other 6 with fluorine-18 radiolabeled 5-fluorouracil. RESULTS: We found that tumor blood flow (mL blood/mL tissue/minute) was significantly correlated to early tumor radiotracer uptake between 4 and 6 minutes [standard uptake value (SUV)4-6; rho = 0.79; P = 0.002], tumor radiotracer exposure over 10 minutes [area under the time-activity curve (AUC)0-10; predominantly parent drug; rho = 0.86; P < 0.001], and tumor radiotracer exposure over 60 minutes (AUC0-60; predominantly radiolabeled metabolites; rho = 0.80; P = 0.002). Similarly, fractional volume of distribution of radiolabeled water in tumor (Vd) was significantly correlated with SUV4-6 (rho = 0.80; P = 0.002), AUC0-10 (rho = 0.85; P < 0.001), and AUC0-60 (rho = 0.66; P = 0.02). In contrast, no correlation was observed between plasma drug or total radiotracer exposure over 60 minutes and tumor drug uptake or exposure. Tumor blood flow was significantly correlated to Vd (rho = 0.69; P = 0.014), underlying the interdependence of tumor perfusion and Vd. CONCLUSIONS: Tumor perfusion is a key factor that influences tumor drug uptake/exposure. Tumor vasculature-targeting strategies may thus result in improved tumor drug exposure and therefore drug efficacy.
    • Plasma pharmacokinetic evaluation of cytotoxic agents radiolabelled with positron emitting radioisotopes.

      Saleem, Azeem; Aboagye, E O; Matthews, Julian C; Price, Patricia M; Academic Department of Radiation Oncology, Christie Hospital NHS Foundation Trust, Wilmslow Road, and The University of Manchester Wolfson Molecular Imaging Centre, Manchester M20 4BX, UK. azeem.saleem@manchester.ac.uk (2008-04)
      PURPOSE: This study aimed to evaluate the utility of plasma pharmacokinetic analyses of anti-cancer agents from data obtained during positron emission tomography (PET) oncology studies of radiolabelled anti-cancer agents. PATIENTS AND METHODS: Thirteen patients were administered fluorine-18 radiolabelled 5-FU ([(18)F]5-FU) admixed with 5-FU, corresponding to a total 5-FU dose of 380-407 mg/m2 (eight patients) and 1 mg/m2 (five patients). Nine patients received 2.2-19.2 microg/m2 of carbon-11 radiolabelled N-[2-(dimethylamino)ethyl]acridine-4-carboxamide ([11C]DACA) at 1/1,000th of phase I dose, as part of phase 0 microdosing study. Radioactivity of parent drug obtained from arterial blood samples, the injected activity of the radiolabelled drug, and the total dose of injected drug were used to obtain plasma drug concentrations. Plasma pharmacokinetic parameters were estimated using model-dependent and model-independent methods. RESULTS: 5-FU plasma concentrations at therapeutic doses were above the Km and a single compartment kinetic model was best used to fit the kinetics, with a mean half-life of 8.6 min. Clearance and volumes of distribution (Vd) obtained using both model-dependent and model-independent methods were similar. Mean (SE) clearance was 1,421(144), ml min(-1) and 1,319 (119) ml min(-1) and the mean (SE) Vd was 17.3 (1.8) l and 16.3 (1.9) l by the model-independent method and model-dependent methods, respectively. In contrast, with 1 mg/m2, plasma concentrations of 5-FU were less than the Km and a two-compartment model was used to best fit the kinetics, with the mean 5-FU half-life of 6.5 min. The mean (SE) clearances obtained by the model-independent method and model-dependent methods were 3,089 (314) ml min(-1) and 2,225 (200) ml min(-1), respectively and the mean (SE) Vd were 27.9 (7.0) l and 2.3 (0.4) l, by the model independent and dependent methods, respectively. Extrapolation of AUC0-Clast to AUC0-infinity was less than 3% in both these cohort of patients. A two-compartment model with a mean half-life of 42.1 min was used to best fit the kinetics of DACA; considerable extrapolation (mean 26%) was required to obtain AUC0-infinity from AUC0-Clast. Mean (SE) clearance of DACA was 1,920 (269) ml min(-1), with the model-independent method and 1,627 (287) ml min(-1) with the model-dependent method. Similarly, Vd [mean (SE)] of DACA with the model-independent and model-dependent methods were 118 (22) l and 50 (15) l, respectively. CONCLUSIONS: Pharmacokinetic parameters can be estimated with confidence from PET studies for agents given at therapeutic doses, whose half-lives are significantly less than the total sampling time during the scan. Tracer studies performed alone, wherein plasma levels below the Km are expected, may also provide valuable information on drug clearance for the entire range of linear kinetics. However, drugs with half-lives longer than the sampling duration are inappropriate for PET plasma pharmacokinetic evaluation.