ORIGINAL ARTICLE Annals of Nuclear Medicine Vol. 15, No. 2, 1 37-139, 2001 Different sensitivities to competitive inhibition of benzodiazepine receptor binding of 11C-iomazenil and 11C-flumazenil in rhesus monkey brain Osamu INOUE,* Rie HOSOI,* Kaoru KOBAYASHI,* Takashi ITOH,** Antony GEE*** and Kazutoshi SUZUKI**** *School of Allied Health Sciences, Faculty of Medicine, Osaka University **Nippon Medical School, Center for Information Science ***PET Centre, Aarhus University Hospital, Denmark ****National Institute of Radiological Sciences The in vivo binding kinetics of 11C-iomazenil were compared with those of 11C-fiumazenil binding in rhesus monkey brain. The monkey was anesthetized with ketamine and intravenously injected with either 11C-iomazenil or 11C-flumazenil in combination with the coadministration of different doses of non-radioactive flumazenil (O,5 and 20 ug/kg). The regional distribution of 11C-iomazenil in the brain was similar to that of 11C-flumazenil, but the sensitivity of 11C-iomazenil binding to competitive inhibition by non-radioactive flumazenil was much less than that of 11C-flumazenil binding. A significant reduction in 11C-flumazenil binding in the cerebral cortex was observed with 20 ug/kg of flumazenil, whereas a relatively smaller inhibition of 11C-iomazenil binding in the same region was observed with the same dose of flumazenil. These results suggest that 11C-flumazenil may be a superior radiotracer for estimating benzodiazepine receptor occupancy in the intact brain. Key words: 11C-romazeml 11C-flumazeml benzodrazepine receptors rhesus monkey PET INTRODUCTION CARBON- 11 Iabeled flumazenil and 123I-labeled iomazenil have been used as selective radioligands for benzodiazepine (Bz) receptor mapping in emission tomography. As previously reported, a significant difference in apparent Bz receptor occupancy was observed between 125I-iomazenil and 3H-flumazenil in mice brain when the dose of flunitrazepam was varied.1 The main purpose of this experiment was to clarify whether this phenomenon is also observed in primate brains. Received October 12, 2000, revision accepted February 2, 2001 . For reprint contact: Osamu Inoue, M.D., School of Allied Health Sciences, Osaka University Faculty of Medicine, 1-7 Yamadaoka, Suita, Osaka 565-0871 , JAPAN. E-mail: inoue@sahs.med.osaka-u.ac.jp MATERIALS AND METHODS 11C-iomazenil and 11C-flumazenil (specific radioactivity 100 GBq//umol) were synthesized by N-methylation according to a previously described method.2 Briefly, N-desmethyl derivatives of iomazenil and flumazenil were reacted with 11C-methyiodide and purified by liquid chromatography. After evacuation of the solvent, 11C-1abeled ligands were prepared in solution for injection. A rhesus monkey (7 kg body weight) was anesthetized with an intramuscular injection of ketamine (5 mg/kg) and subsequently seated on a modified monkey chair, which maintains the head in a fixed position.3 The experimental protocol was approved by the Committee on the Safety and Ethical Handling Regulations for Laboratory Animal Experiments, National Institute of Radiological Sciences. About 0.4 GBq of 11C-iomazenil or 11C-flumazenil was intravenously injected with different doses of non-radioactive flumazenil (O,5 and 20 ug/kg), and a PET scan was performed for 40 minutes with an animal PET camera (SHR2000, Hamamatsu Photonics, Hamamatsu. Japan) with transaxial resolution of 3.0 mm full-width at half-maximum (FWHM) and a center-to-center distance of 6.5 mm. Regions of interest (ROIs) were outlined by hand on reconstructed PET images. The time courses of radioactivity concentration in each region were then determined. RESULTS Figure 1 shows the resulting PET images for the injection of 11 C-iomazenil or 11 C-fiumazenil and the coadministration of different doses of the competitor flumazenil. When the tracers were injected alone, similar regional distribution of ll C-iomazenil and 11 C-flumazenil was observed in the slice, which contained the cerebral cortex, striatum and thalamus. On the other hand, a significant decrease in the accumulation of 11 C-flumazenil was observed in all ROIs when 5 ug/kg and 20 ug/kg of flumazenil were administered. Considerably smaller decreases in 11 C-iomazenil binding were observed for the same doses of flumazenil, as shown in Figure 1. The time courses of the radioactivity concentration in each region are shown in Figure 2. The most important difference in the binding of 11 C-iomazenil and 11 C-flumazenil is their different kinetic properties. The kinetics of ll C-flumazenil binding in rhesus monkey brain occurred very quickly, and binding within the cerebral cortex reached a maximum 5 minutes after the injection of tracer. With 11 C-iomazenil, however, binding occurred very slowly but increased continuously so that the apparent dissociation rate of 11 C-iomazenil appears to be negligible, and the association process can be measured by using 11 C-iomazenil for a period of at least 40 minutes after injection of the tracer. Another difference in the binding properties of 11C-iomazenil and 11C-flumazenil was that a considerable amount of specific 11C-iomazenil binding was observed in the pons, which has been used as a reference region for the quantitative analysis of 11C-flumazenil binding.4 DISCUSSION The most important finding in this study was that the sensitivity of 11C-flumazenil to competitive inhibition by flumazenil is significantly higher than that of 11C-iomazenil binding. Our rough estimations suggested that a dose of 5 ug/kg of flumazenil resulted in 70% occupancy of Bz receptors when measured with 11C-flumazenil and 45% occupancy when measured with 11C-iomazenil. This difference in receptor occupancy seems to be not due to changes in cerebral blood flow, since flumazenil has no significant pharmacological effect on the central nervous system. The present results were consistent with our previous data for mice with Bz agonist and inverse agonist as competitive inhibitors. It is of interest that Bz antagonist also showed a similar discrepancy in receptor occupancy. The reason for the discrepancy between the two radioligands in the apparent receptor occupancy seems to arise from the difference in their kinetic properties. 11C-iomazenil binding probably reflects the association process because of its slow binding kinetics, and does not reach a state of pseudo-equilibrium within the period of PET measurement. Another possibility is that an increase in ligand concentration surrounding the receptors would increase the apparent association rate constant (kon) of 11C-iomazenil binding (apparent positive cooperativity). 5 Since the environment surrounding the receptors in an intact brain is much more heterogeneous, interfacial or surface effects are more likely to significantly affect the available free ligand concentration. Such effects would probably be related to such physicochemical properties of the ligand as lipophilicity and electric charge.6 In addition, the heterogeneity of brain capillaries should also be considered as a possible cause of the apparent discrepancy between 11C-iomazenil and flumazenil in Bz receptor occupancy, as previously suggested.7 11C-flumazenil seems to be a better radioligand than 11C or 123I-labeled iomazenil for estimating Bz receptor occupancy. Many patients with neurological or psychological disorders, who are candidates for SPECT imaging with 123I-labeled iomazenil are sometimes medicated with benzodiazepines. The competitive inhibition effect of these drugs on SPECT images obtained during the association process of 123I-iomazenil binding could be avoided, but our preliminary animal experiment indicated that Bz receptor occupancy could be estimated by means of a late SPECT image instead of an early SPECT image of 123I-iomazenil binding. Both simulation studies and further animal experiments on the relationship between the kinetic properties of the radioligand and apparent receptor occupancy are in progress. This phenomenon is also observed in estimations of other types of receptor occupancy. For example, 11C-raclopride seems to be better for estimating dopamine D2 receptor occupancy than 11C-N-methylspiperone8 and 11C-FLB457 because these substances have very low dissociation rate constants. REFERENCES 1 . Hosoi R, Kobayashi K, Watanabe Y, Inoue O. Discrepancy of benzodiazepine receptor occupancy between 3H fiumazenil and 125I-iomazenil in intact mouse brain. J Neural Transm 1999; 106: 243-256. 2. Suzuki K, Inoue O, Tamate K, Mikado F. Production of 3-N-[11C]methylspiperone with high specific activity and high radiochemical purity for PET studies: suppression of its radiolysis. Int J Rad Appl Instrum 1990; 41 : 593-599. 3. Onoe H, Inoue O, Suzuki K, Tsukada H, Itoh T, Mataga N, et al . Ketamine increases the striatal N-[11C]methylspiperone binding in vivo: positron emission tomography study using conscious rhesus monkey. Brain Res 1994; 663: 191-198. 4. Persson A, Ehrin E, Eriksson L, Farde L, Hedstrom CG, Litton JE, et al. Imaging of [11C]-labelled Ro 15-1788 binding to benzodiazepine receptors in the human brain by positron emission tomography. J Psychiatr Res 1985; 19: 609-622. 5. Inoue O, Kobayashi K, Takai N, Furusawa Y, Ando K, Nakano T, et al. An increase in [3H]QNB binding by proton-beam irradiation in intact rat brain: an apparent positive cooperativity of binding. Neurosci Lett 1998; 250: 33-36. 6. Inoue O, Kobayashi K, Gee A. Changes in apparent rates of receptor binding in the intact brain in relation to the heterogeneity of reaction environments. Crit Rev Neurobiol 1999; 13: 199-225. 7. Videbaek C, Ott P. Paulson OB, Knudsen GM. Blood-brain barrier transport and protein binding of flumazenil and iomazenil in the rat: implications for neuroreceptor studies. J Cereb Blood Flow Metab 1999; 19: 948-955. 8. Inoue O, Wakahara S, Kobayashi K, Gee A. Enhancement of 3H-N-methylspiperone binding but not 3H-raclopride binding in mouse striatum by MK-801 : evidence that factors other than competition by endogenous dopamine are responsible for changes in D2 receptor binding in vivo. J Neural Transm 1999; 106: 131-137.