Figure 6
From: Quantifying Reoxygenation in Pancreatic Cancer During Stereotactic Body Radiotherapy
Measured changes in hypoxic fraction (post SBRT minus pre SBRT) versus changes in volume fraction \({v}_{u}\) of unperfused tissue using two thresholds, TBR > 1.4 (a) and TBR > 2.4 (b) used to determine PET hypoxic fraction for the 9 mice (6 irradiated, 3 control) that underwent both DCE-MR and FAZA-PET imaging. \({v}_{u}\) decreases during SBRT in irradiated tumours, indicating a drop in perfusion heterogeneity. In turn, hypoxic fraction is reduced by the increased oxygenation. (c,d) Measured changes in hypoxic fraction versus the values predicted from baseline HF values and pre- and post-SBRT \({v}_{u}\) values for the thresholds. The dashed lines in the bottom plots are equal to unity (i.e., lines of perfect prediction) and are not fits. The fact that the apparent “slope” of the measured vs predicted HF data for the threshold of 1.4 is less than one suggests that for this relatively low threshold, a nonzero value of \({{\rm{HF}}}_{0}\) (see Methods) would be a better approximation. Conversely, for the predicted HF values corresponding to the higher threshold, the assumption of \({{\rm{HF}}}_{0}\approx 0\) is evidently good (even though correlations are incrementally smaller).
