Quantitative assessment of hypoxia kinetic models by a cross-study of dynamic 18F-FAZA and 15O-H2O in head and neck tumors

Kinetic modeling is one important method to assess the underlying physiology behind tracer uptake in molecular imaging. Although there are many well developed models which cover a broad range of applications, it is still challenging to quantitatively assess mathematical models with consideration of clinical applications and their biological nature. Tumor hypoxia is considered as one main resistance factor of standard radiotherapy and some chemotherapy. Hypoxia usually is the result of a decreased oxygen delivery to the cells either by an increase in diffusion distances or a decreased oxygen supply due to an inadequate tumor blood flow (perfusion). Under this assumption, we compared different hypoxia kinetic models. Dynamic PET images of the hypoxia tracer ¹⁸F-FAZA and the perfusion tracer ¹⁵O-H₂O were acquired and the Thorwarth model, the reversible and irreversible two-tissue compartment model, the Logan plot and the Patlak plot were applied to model the process of tracer transport and accumulation under hypoxic condition. With the cross analysis between these two specific tracers, it is shown that hypoxia kinetic modeling delivers significantly different information than static measurements. Different models have a large variation under the same condition and they even can lead to opposite physiological interpretations. Our result shows that the irreversible two compartment model corresponds better to the expectation of a negative (inverse) correlation between hypoxia and perfusion.