Multicellular resistance to tirapazamine is due to restricted extravascular transport: a...

In common with other bioreductive drugs, metabolic reduction is required for activation of the benzotriazine-di-N-oxide tirapazamine (TPZ) in hypoxic regions of tumors. This same metabolism also consumes the drug as it diffuses, impeding its penetration into hypoxic tissue. In this study, we develop a pharmacokinetic (PK)/pharmacodynamic (PD) model for TPZ that explicitly includes its diffusion characteristics as measured in multicellular layer (MCL) cultures of HT29 colon carcinoma cells. The kinetics of TPZ metabolism to its mono-N-oxide derivative SR 4317, determined by high-performance liquid chromatography using anoxic HT29 single cell suspensions, demonstrated both a first order and saturable (K(m) = 3.6 micro M) component. Cell killing, assessed by clonogenic assay under the same conditions, demonstrated an approximately quadratic concentration dependence and linear time dependence. TPZ transport through MCLs, determined under hyperoxic conditions (95% O(2)) to suppress reductive metabolism, provided a concentration-independent diffusion coefficient of 0.40 x 10(-6) cm(2)s(-1). Under anoxia, this transport was strongly suppressed and was well predicted by the single cell metabolism parameters (scaled to the cell density in MCLs). These PK (transport) and PD (cytotoxicity) parameters were used to calculate cell killing as a function of distance in anoxic HT29 MCLs after the addition of TPZ to both sides of the MCL. The predicted average cell kill was in good agreement with measured values, which showed much less killing than for single cell suspensions under the same conditions. The success of this PK/PD model in predicting response in MCL shows that inefficient transport, rather than changes in intrinsic sensitivity, is responsible for TPZ resistance in these three-dimensional cell cultures and suggests that optimization of transport properties is a high priority in developing second-generation TPZ analogues.