@article {Zanotti-Fregonara108, author = {Paolo Zanotti-Fregonara and Sami Zoghbi and Jeih-San Liow and Jinsoo Hong and Ronald Boellaard and Victor Pike and Robert Innis and Masahiro Fujita}, title = {Kinetic analysis in human brain of 11C-(R)-rolipram, a PET tracer of the cAMP cascade: Test-retest study and the use of an image-derived input function}, volume = {51}, number = {supplement 2}, pages = {108--108}, year = {2010}, publisher = {Society of Nuclear Medicine}, abstract = {108 Objectives Rolipram inhibits phosphodiesterase 4, an enzyme that metabolizes cAMP. The aims of this study were 1) to measure the reproducibility of 11C-(R)-rolipram binding in healthy human subjects by kinetic analysis using arterial input function and 2) to compare the arterial input (the gold standard) with an image-derived input function. Methods Twelve healthy humans had two scans. Using metabolite-corrected arterial input, total distribution volume (VT) was measured with an unconstrained two-tissue compartment model and the Logan plot. Image input was derived from PET images of the internal carotid artery and from four arterial blood samples at 6, 20, 60 and 90 min. The blood samples were used: 1) to correct the PET images of the carotid for spillover and partial volume errors and 2) to estimate the parent fraction in blood using nonlinear regression. Because k4 was poorly identified with the compartment model, arterial and image inputs were compared using only Logan-derived VT values. Results VT determined from arterial input showed good identifiability with standard errors of ~5\% and fair retest variability of 19\%. In all 24 scans, curves from image input excellently reproduced the shape and concentrations of parent radioligand from arterial inputs. The ratio of Logan-derived VT values from arterial input compared to that from image input was 1.0{\textpm}0.1, and both input functions had similar retest reproducibility (arterial = 15.2\%; image = 15.0\%). Conclusions By using metabolite-corrected arterial input function, 11C-(R)-rolipram binding was measured with modest precision. The curves from the image input closely matched those from arterial input, and values of VT from both inputs had similar accuracy and reproducibility. Assuming that venous blood samples can replace the four arterial samples used in this analysis, image input is expected to replace arterial input for 11C-(R)-rolipram with good fidelity and avoid the need for arterial catheterization}, issn = {0161-5505}, URL = {https://jnm.snmjournals.org/content/51/supplement_2/108}, eprint = {https://jnm.snmjournals.org/content}, journal = {Journal of Nuclear Medicine} }