Estimation of oxygen distribution in RIF-1 tumors by diffusion model-basedinterpretation of pimonidazole hypoxia and Eppendorf measurements

Numerical simulations of oxygen diffusion from the capillaries in tumor tis sue were used to predict the capillary oxygen supply within and near hypoxi c regions of the RIP-1 tumor. A finite element method to simulate the oxyge n distribution from a histology section is presented, along with a method t o iteratively estimate capillary oxygen concentrations. Pathological struct ural data for these simulations came from sections of the tumor stained wit h hematoxylin and eosin and were used to define the capillary positions and shapes, while overlapping regions of low oxygen concentration were defined by the hypoxia marker pimonidazole, These simulations were used to calcula te spatial maps of the oxygen concentration and were tested for their abili ty to reproduce Eppendorf pO(2), histograms from the same tumor line. This simulation study predicted that capillary oxygen concentrations ranged from zero to above 20 muM, With a dominant peak in the hypoxic regions showing 78% of capillaries with less than 1 muM oxygen concentration, compared to o nly 12% in the non-hypoxic regions. The results were not highly sensitive t o the metabolic oxygen consumption rate, within the range of 2 to 16 muM/s. This numerical method for oxygen capillary simulation is readily adaptable to histology sections and provides a method to examine the heterogeneity o f oxygen within the capillaries and throughout the tumor tissue section bei ng examined. (C) 2001 by Radiation Research Society.