HMR 1826 (N-[4-beta-Glucuronyl-3-nitrobenzyl-oxycarbonyl]doxorubicin) is a nontoxic glucuronide prodrug from which active doxorubicin is released by beta-glucuronidase. Preclinical studies aimed at dose optimization of HMR 1826, based on intratumoral pharmacokinetics, are important to design clinical studies. Using an isolated perfused human lung model, the uptake of doxorubicin into normal tissue and tumors after perfusion with 133 microg/ml (n = 6), 400 microg/ml (n = 10), and 1200 microg/ml (n = 6) HMR 1826 was compared. Extracellular tissue pH was measured, and enzyme kinetic studies were performed in vitro to investigate the effect of pH on the formation of doxorubicin. Extracellular pH was lower in tumors than in healthy tissue (6.46 +/- 0.35, n = 8 versus 7.30 +/- 0.33, n = 10; p < 0.001). In vitro, beta-glucuronidase activity was 10 times higher at pH 6.0 than at neutral pH. After perfusion with HMR 1826, there was a linear relationship between HMR 1826 concentrations in perfusate and normal lung tissue. After perfusion with 133, 400, and 1200 microg/ml HMR 1826, the final doxorubicin concentrations in normal and tumor tissue were 2.7 +/- 0.9, 11.1 +/- 5.4, and 21.8 +/- 8.4 microg/g (p < 0.05 for all comparisons), and 0.7 +/- 0.3, 8.6 +/- 2.0 microg/g (p < 0.01 versus 133 microg/g), and 8.7 +/- 4.9 microg/g, respectively. This agrees with the enzyme kinetic observations of saturation of beta-glucuronidase at 400 microg/ml HMR 1826 in the acidic environment of the tumor. Therefore, the escalation of the HMR 1826 dose most likely results in higher circulating concentrations than 400 microg/ml but does not increase the uptake of doxorubicin into tumors and, subsequently, antitumor efficacy. The isolated perfused human lung is an excellent model for preclinical investigations aimed at optimization of tissue pharmacokinetics of tumor-selective prodrugs.