PT - JOURNAL ARTICLE AU - Ramsha Iqbal AU - Gem M. Kramer AU - Eline E. Verwer AU - Marc C. Huisman AU - Adrianus J. de Langen AU - Idris Bahce AU - Floris H.P. van Velden AU - Albert D. Windhorst AU - Adriaan A. Lammertsma AU - Otto S. Hoekstra AU - Ronald Boellaard TI - Multiparametric Analysis of the Relationship Between Tumor Hypoxia and Perfusion with <sup>18</sup>F-Fluoroazomycin Arabinoside and <sup>15</sup>O-H<sub>2</sub>O PET AID - 10.2967/jnumed.115.166579 DP - 2016 Apr 01 TA - Journal of Nuclear Medicine PG - 530--535 VI - 57 IP - 4 4099 - http://jnm.snmjournals.org/content/57/4/530.short 4100 - http://jnm.snmjournals.org/content/57/4/530.full SO - J Nucl Med2016 Apr 01; 57 AB - 18F-fluoroazomycin arabinoside (18F-FAZA) is a PET tracer of tumor hypoxia. However, as hypoxia often is associated with decreased perfusion, the delivery of 18F-FAZA may be compromised, potentially disturbing the association between tissue hypoxia and 18F-FAZA uptake. The aim of this study was to gain insight into the relationship between tumor perfusion and 18F-FAZA uptake. Methods: Ten patients diagnosed with advanced non–small cell lung cancer underwent subsequent dynamic 15O-H2O and 18F-FAZA PET scans with arterial sampling. Parametric images of both 15O-H2O–derived perfusion (tumor blood flow [TBF]) and volume of distribution (VT) of 18F-FAZA were generated. Next, multiparametric classification was performed using lesional and global thresholds. Voxels were classified as low or high TBF and 18F-FAZA VT, respectively. Finally, by combining these initial classifications, voxels were allocated to 4 categories: lowTBF–lowVT, lowTBF–highVT, highTBF–lowVT, and highTBF–highVT. Results: A total of 13 malignant lesions were identified in the 10 patients. The TBF and 18F-FAZA VT values (average ± SD) across all lesions were 0.45 ± 0.20 mL·cm−3·min−1 and 0.94 ± 0.31 mL·cm−3, respectively. The averages of all lesional median values for TBF and 18F-FAZA VT were 0.37 ± 0.15 mL·cm−3·min−1 and 0.85 ± 0.18 mL·cm−3, respectively. Multiparametric analysis showed that classified voxels were clustered rather than randomly distributed. Several intralesion areas were identified where 18F-FAZA VT was inversely related to TBF. On the other hand, there were also distinct areas where TBF as well as 18F-FAZA VT were decreased or increased. Conclusion: The present data indicate that spatial variation of 18F-FAZA uptake is not necessarily inversely related to TBF. This suggests that decreased TBF may result in flow-limited delivery of 18F-FAZA. Areas with both high 18F-FAZA uptake and high TBF values suggest that high 18F-FAZA uptake, possibly suggesting hypoxia, may occur despite high TBF values. In conclusion, multiparametric evaluation of the spatial distributions of both TBF and 18F-FAZA uptake may be helpful for understanding the 18F-FAZA signal.