Effect of radiolabeled metabolite elimination from the brain on the accuracy of cerebral enzyme activity estimation using positron emission tomography with substrate tracers.

Cerebral enzyme activity can be quantified using positron emission tomography (PET) in conjunction with a radiolabeled enzyme substrate. We investigated the relationship between the elimination rate (k(el)) of tracer metabolites from the brain and the precision of target enzyme activity estimation (k(3)). An initial simulation study indicated that the precision of k(3) estimates was highly dependent on k(el), and was characterized by several kinetic parameters including the ratio of k(el) and the efflux rate (k(2)) of authentic tracer (β≡k(el)/k(2)). The optimal tracer condition for high sensitivity was found to be β<0.1. To verify the simulation results, we performed a PET study with a single monkey using two PET tracers, N-[(18)F]fluoroethylpiperidin-4-ylmethyl acetate ([(18)F]FEP-4MA) and N-[(11)C]methylpiperidin-4-yl acetate ([(11)C]MP4A). Both of these substrate type tracers were developed for measuring cerebral acetylcholinesterase activity. There was good retention of the radioactive metabolite of [(11)C]MP4A in the brain (k(el)=0.0036±0.0013 min(-1), β=0.028), whereas that of [(18)F]FEP-4MA was eliminated from the brain (k(el)=0.012±0.0010 min(-1), β=0.085). A non-linear least square analysis for simultaneous estimation of all parameters showed that the precision of the k(3) estimate for [(18)F]FEP-4MA was as high (7.4%) as that for [(11)C]MP4A (10%). These results indicate that tracers with metabolites that are eliminated from the brain at a slow rate (β<0.1) may be useful for the quantitative measurement of target enzyme activity.