Radiosynthesis and Evaluation of 11C-Labeled Diaryl-Substituted Imidazole and Indole Derivatives for Mapping Cyclooxygenase-2

11C-labeled analogs of 4-chloro-5-(3-fluoro-4-methoxyphenyl)-1-(4-methylsulfonylphenyl)imidazole ([11C]1), 4-[4-chloro-5-(3-fluoro-4-methoxyphenyl)imidazol-1-yl]benzenesulfonamide ([11C]2) and 2-(4-aminosulfonylphenyl)-3-(4-methoxyphenyl)indole ([11C]3), which have been shown to have excellent potency and high selectivity for cyclooxygenase isoform 2 (COX-2) inhibiting activity, were prepared and evaluated in rats as potential radiopharmaceuticals for imaging the COX-2 enzyme by positron emission tomography. These 11C-labeled COX-2 inhibitors were synthesized in high radiochemical yields by O-[11C]methylation of phenolic precursors with [11C]methyl triflate in acetone containing NaOH as a base. In vivo evaluation in rats bearing AH109A hepatoma showed no specific binding of any tracer to COX-2 in any tissue such as the brain, heart, lung, kidney, and AH109A hepatoma. In ex vivo autoradiography, [11C]1 showed regionally different distribution in the brain, while [11C]2 and [11C]3 were not substantially taken up by the brain. In in vitro monolayer efflux assays, compound 3 was found to be a substrate for the P-glycoprotein (P-gp) efflux pump, but pretreatment of rats with the potent P-gp inhibitor, cyclosporine A, did not have any significant influence on the cerebral uptake of [11C]3. These results indicate that all three tracers were not suitable for in vivo imaging of COX-2. There seem to be some obstacles to finding a useful candidate for in vivo imaging application of COX-2 selective inhibitors only by standard consideration of in vitro affinity and selectivity, and the lipophilicity of the compound.