Development of two fluorine-18 labeled PET radioligands targeting PDE10A and in vivo PET evaluation in nonhuman primates
a b s t r a c t
Introduction: Phosphodiesterase 10A (PDE10A) is a member of the PDE enzyme family that degrades cyclic adenosine and guanosine monophosphates (cAMP and cGMP). Based on the successful development of [11C]T- 773 as PDE10A positron emission tomography (PET) radioligand, in this study our aim was to develop and evaluate fluorine-18 analogs of [11C]T-773.Methods: [18F]FM-T-773-d2 and [18F]FE-T-773-d4 were synthesized from the same precursor used for 11C- labeling of T-773 in a two-step approach via 18F-fluoromethylation and 18F-fluoroethylation, respectively, using corresponding deuterated synthons. A total of 12 PET measurements were performed in seven non- human primates. First, baseline PET measurements were performed using High Resolution Research Tomograph system with both [18F]FM-T-773-d2 and [18F]FE-T-773-d4; the uptake in whole brain and separate brain regions, as well as the specific binding and tissue ratio between putamen and cerebellum, was examined. Second, baseline and pretreatment PET measurements using MP-10 as the blocker were performed for [18F]FM-T-773-d2 includ- ing arterial blood sampling with radiometabolite analysis in four NHPs.Results: Both [18F]FM-T-773-d2 and [18F]FE-T-773-d4 were successfully radiolabeled with an average molar activity of 293 ± 114 GBq/μmol (n=8) for [18F]FM-T-773-d2 and 209 ± 26 GBq/μmol (n=4) for [18F]FE-T- 773-d4, and a radiochemical yield of 10% (EOB, n=12, range 3%–16%). Both radioligands displayed high brain uptake (~5.5% of injected radioactivity for [18F]FM-T-773-d2 and ~ 3.5% for [18F]FE-T-773-d4 at the peak) and a fast washout. Specific binding reached maximum within 30 min for [18F]FM-T-773-d2 and after approximately 45 min for [18F]FE-T-773-d4. [18F]FM-T-773-d2 data fitted well with kinetic compartment models. BPND values obtained indirectly through compartment models were correlated well with those obtained by SRTM. BPND calculated with SRTM was 1.0–1.7 in the putamen. The occupancy with 1.8 mg/kg of MP-10 was approximately 60%.
Conclusions: [18F]FM-T-773-d2 and [18F]FE-T-773-d4 were developed as fluorine-18 PET radioligands for PDE10A, with the [18F]FM-T-773-d2 being the more promising PET radioligand warranting further evaluation.
1.Introduction
Phosphodiesterase 10A (PDE10A) is one of the phosphodiesterase enzyme groups, hydrolyzing both adenosine and guanosine mono- phosphates (cAMP and cGMP) [1]. PDE10A is selectively distributed in the striatum and in the substantia nigra in the brain [2]. Given the local- ization and regulatory functions of PDE10A, inhibiting it could be a promising therapeutic approach for multiple central nervous system diseases such as schizophrenia, Huntington’s disease, and Parkinson’s disease [3–6]. Direct inhibition of PDE10A leading to an increase in the cAMP levels might be more desirable than achieving the same effect with dopamine receptor-targeted drugs [7,8]. As an imaging technique and translational tool, positron emission tomography (PET) would allow in vivo quantification of PDE10A levels, providing information re- garding target occupancy by potential drugs, distribution and density of PDE10A in tissues, and the specificity of the drug molecule [9]. In recent years, several PET radioligands for PDE10A, such as [11C]papaverine, [11C]MP-10, [18F]MNI654, [18F]MNI659 [18F]JNJ42259152,[18F] JNJ41510417, [11C]LuAE92686, and [11C]AMG7980, have been reported [10–14]. However, some of them suffer from limitations — for example, formation of lipophilic radiometabolites complicating image analysis or unsuitable kinetic properties, making them less-than-optimal for rou- tine applications [11,12,15]. In our previous work, we have labeled a series of selective PDE10A inhibitors with carbon-11, resulting in the successful development of one carbon-11 radioligand [11C]T-773 for ap- plication in human studies [16–18]. Carbon-11 radioligands have the benefit of high molar activity, offer reduced dose exposure for the patient, and allow for same-day multiple-injection experimental proto- cols; however, the benefits of longer half-life and lower positron energy of fluorine-18 are not to be discounted. In a clinical setting, a fluorine-18 radioligand can be of distinctive advantage, allowing for transport to remote imaging sites.
In addition, some current PDE10A radioligands may have various limitations [11,12,15]. Given the results with carbon-11 labeling of ligands based on 3-(1H-pyrazol-5- yl)pyridazin-4(1H)-one backbone developed by Takeda Pharmaceutical Company Limited (Japan), it was logical to extend that work by taking the most promising compound of that series – [11C]T-773 – and replac- ing the [11C]methyl fragment in it with [18F]fluoromethyl-d2 and 1-[18F] fluoroethyl-d4 synthons via O-alkylation [16,18,19]. The two com- pounds were evaluated for their affinity and selectivity against human PDE10A2 and over other recombinant human PDE family enzymes, in- cluding PDE1A, PDE2A3, PDE3A, PDE4D2, PDE5A1, PDE6AB, PDE7B,PDE8A1, PDE9A2 and PDE11A4 using previously published methodolo- gy [17]. The IC50 values of FM-T-773-d2 and FE-T-773-d4 for PDE10A2 were 1.8 nM and 9.7 nM, respectively. The minimum IC50 value towards other 10 PDE families was 2.1 μM and 2.0 μM, respectively, for the PDE6AB. Therefore, the PDE family selectivity of FM-T-773-d2 and FE- T-773-d4 for recombinant PDE10A2 was more than 1167-fold and 206-fold, respectively.The aim of the study was to develop a fluorine-18 PDE10A radioligand based on 3-(1H-pyrazol-5-yl)pyridazin-4(1H)-one scaffold with suitable characteristics for examination of PDE10A in drug occu- pancy studies in a clinical setting.
2.Materials and methods
The precursors for labeling, two unlabeled reference standards, and MP-10 succinate (MP-10), were synthesized and supplied by Takeda Pharmaceutical Company Limited (Fujisawa, Japan). MP-10 has been reported to be a selective PDE10A inhibitor developed by Pfizer, Inc. [20,21]. All other chemicals were of analytical grade, were obtained from commercial sources, and were used as received. Solid-phase ex- traction cartridges (Oasis HLB 3cc) were obtained from Waters Corp. (Milford, MA, USA), and Millex GV 0.22-μm sterile filters were pur- chased from Millipore (Ireland). Sterile phosphate-buffered saline was purchased from APL (Sweden). Radioligand purification was performed on a semipreparative HPLC system composed of rheodyne-type injector, Ascentis RP-Amide reversed-phase column (250 × 10 mm, 5 μm), a var- iable wavelength UV absorbance detector set to 254 nm (Knauer, Germany) in series with a PIN-diode detector for radioactivity detection and a HPLC pump (Smartline 1000, Knauer, Germany). The radiochem- ical purity of the labeled products was determined with reverse-phase HPLC on a system consisting of Merck-Hitachi L-7100 pump, Merck- Hitachi L-7400 ultraviolet (UV) detector (set to 220 nm), D-7000 inter- face (Merck-Hitachi), a β-flow radiodetector (Beckman) for radioactiv- ity detection, and an Advanced Chromatography Technologies ACE 5 C18-HL reversed-phase HPLC column (250 × 4.6 mm, 5 μm). Identifica- tion of the fluorine-18-labeled compounds was made by co-injecting it with a known unlabeled standard of the compound and comparing the retention time between the UV and radioactive channels on the same HPLC system. Molar activity (MA) was determined using an HPLC system consisting of Ascentis RP-Amide reversed-phase HPLC column (150 × 4.6 mm, 3 μm or 50 × 4.6 mm, 2.7 μm, Sigma-Aldrich), and a Hewlett-Packard Series 1100 HPLC system composed of an autoinjector, variable-length UV detector, pump, and a degasser. The radio- metabolism of each labeled compound in the non-human primates (NHPs) was assessed using plasma samples analyzed on an HPLC system consisting of an interface module (D-7000; Hitachi, Japan), an L-7100 pump (Hitachi), an injector (model 7125, with a 5.0-mL loop; Rheodyne, USA) equipped with a μ-Bondapak C18 column (300 × 7.8 mm, 10 μm; Waters, USA), and a UV absorption detector (L-7400, 254 nm; Hitachi) in series with a dual bismuth germanium oxide coinci- dence radiation detector (S-2493Z; Oyokoken: Fussa, Japan) equipped with a 550-μL flow cell. Acetonitrile (A) and ammonium formate (100 mM) (B) were used as the mobile phase at 6.0 mL/min, according to the following gradient: 0–6 min (A/B), 30:70 → 70:30 v/v; 6–8 min (A/B), 70:30 v/v. The peaks for radioactive compounds eluting from the column were integrated and their areas were expressed as a per- centage of the sum of the areas of all detected radioactive compounds (decay-corrected to the time of injection on the HPLC).
[18F]FM-T-773-d2 (1-(2-fluorophenyl)-5-[18F]fluoromethoxy-d2-3- (1-phenyl-1H-pyrazol-5-yl)pyridazin-4(1H)-one) and [18F]FE-T-773- d4 (1-(2-fluoro-4-(tetrahydro-2H-pyran-4-yl)phenyl)-5-(2-[18F] fluoroethoxy-d4)-3-(1-phenyl-1H-pyrazol-5-yl)pyridazin-4(1H)-one) were synthesized via the two-step O-alkylation approach, previously described by Iwata et al. [22–25]. Aqueous [18F]fluoride was produced via the 18O(p,n)18F nuclear reaction using a General Electric Medical Systems PETtrace cyclotron in a silver fluorine-18 water target. The ra- dionuclide was transferred from the target by means of helium flow (in a 1.5-mL bolus of [18O]H2O) and trapped on a QMA light Sep-Pak car- tridge (bicarbonate form) to remove [18O]H2O. [18F]fluoride was then eluted into the reaction vessel using 2 mL of acetonitrile/water (96/4 v/v) containing 9.8 mg of Kryptofix 2.2.2 and 1.8 mg of potassium carbonate [26]. The solvents were evaporated by heating at 140 °C under a stream of nitrogen (120 mL/min). With the use of this method, azeotropic distillation with additional acetonitrile was not required to produce reactive [18F]fluoride-Kryptofix complex. Following cooling of the reactor vessel, deuterated dibromomethane-d2 or 2- methylbenzenesulfonate ethylbromide-d4 (15 μL) in acetonitrile or 1,2-dichlorobenzene (1,2-DCB) (600 μL, respectively), was added and the mixture was heated at 50 °C for acetonitrile/dibromomethane-d2 or at 120 °C for 1,2-DCM/2-methylbenzenesulfonate ethylbromide-d4 for 10 min. The reaction mixture was then cooled and the resulting corresponding fluorine-18-labeled alkylating agent was transferred by distillation (50–80 °C, nitrogen flow) via series of silica cartridges for dibromomethane-d2 or without them for 2-methylbenzenesulfonate ethylbromide-d4 to a second vial containing anhydrous DMF (500 μL), precursor (desmethyl-T-773) (1.5–2.0 mg) and NaOH (0.5 M, 5 μL) at 0 °C. The reactor was heated at 110 °C for 10 min and then cooled; the reaction mixture was diluted with water, injected onto reversed- phase HPLC column, and eluted with a mixture of acetonitrile and aque- ous triethylamine mixture as described previously. The product fraction from the HPLC was collected into a vial containing 50 mL of sterile water and 70–100 mg of sodium ascorbate. The resulting solution was then pushed through the Oasis HLB 3-cc cartridge, previously conditioned by 5 mL of 99.6% ethanol and 5 mL of sterile water (in that order). After trapping the product, the cartridge was rinsed with 8 mL of sterile water; the product was then eluted with 1 mL of 99.6% ethanol and collected into sterile vial prefilled with 9–10 mL of phosphate- buffered saline. Finally, the product was passed through a 0.22-μm ster- ile filter (Millipore) in a particle-free aseptic environment. The mobile phase composition for purification and analysis, together with yield, radiochemical purity, and molar activity, is summarized in the Results section.
Two sets of a total of 12 PET measurements were carried out in seven NHPs. In the first set of four PET measurements, two radioligands were evaluated at the baseline using three NHPs to select the preferred PET radioligand. In the second set of eight PET measurements, the preferred radioligand [18F]FM-T-773-d2 was evaluated in a pair of baseline and blocking PET measurements in four additional NHPs.NHPs were housed in the Astrid Fagraeus Laboratory of the Swedish Institute for Infectious Disease Control (SMI), Solna, Sweden. The study was approved by the Animal Ethics Committee of the Swedish Animal Welfare Agency (N452/11) and was performed according to “Guide- lines for planning, conducting, and documenting experimental re- search” of the Karolinska Institutet and international guidelines [27]. Anesthesia for monkeys was induced by intramuscular injection of ke- tamine hydrochloride (~ 10 mg/kg, Ketaminol vet.; Intervet, Sweden) and maintained by administering a mixture of sevoflurane (2%–8%, Sevoflurane®; Abbott Scandinavia AB, Sweden), oxygen, and medical air after endotracheal intubation. The head was immobilized with a fix- ation device [28]. Body temperature was maintained by a Bair Hugger device (model 505; Arizant Health Care, USA) and monitored by an oral thermometer. Electrocardiogram, heart rate, respiratory rate, oxy- gen saturation, and arterial blood pressure were continuously moni- tored throughout the experiment.
PET measurements were conducted using the High Resolution Re- search Tomograph system (Siemens Molecular Imaging, USA). Only one PET measurement was performed on each day because of the long half-life of fluorine-18. A 6-min transmission scan using a single 137Cs source was obtained immediately before the radioligand injection. After intravenous administration of the radioligand, list-mode data were acquired for 180 min in the first part of the NHP PET study and 123 min in the second part of the NHP PET study. PET images were re- constructed with a series of frames of increasing duration (5 × 60 s, 5× 180 s, 5 × 360 s, 13 × 600 s for 180 min, and 9 × 20 s, 3 × 60 s, 5 ×180 s, 17 × 360 s for 123 min) using the ordinary Poisson 3- dimensional ordered-subset expectation maximization (OP-3D-OSEM) algorithm, with 10 iterations and 16 subsets, including modeling of the point spread function, after correction for attenuation, random, and scatter. The in-plane resolution of the reconstructed images was ap- proximately 1.5 mm. In the blocking PET measurements, 1.8 mg/kg of MP-10 was administered intravenously approximately 35 min before the radioligand injection. The duration of MP-10 administration was 30 min.In the first part of four NHP PET measurements, venous blood sam- pling (1–2 mL each) was performed at 5 min (for protein binding) and 4, 15, 30, 60, 90, 120, and 180 min (for radiometabolite analysis) after the radioligand injection.In the second part of eight NHP PET measurements, arterial blood sampling was performed continuously for 3 min using an automated blood-sampling system at a speed of 3 mL/min (ABSS; Allog AB, Sweden). Blood samples (1–2 mL) were drawn manually at 4, 15, 30, 60, 90, and 120 min for measurement of radioactivity in whole blood and plasma and metabolite analysis. These data were used in construct- ing the radiometabolite corrected arterial plasma input function.
For the measurement of protein binding of the radioligand, one arterial blood sample was taken at 5 min before the radioligand injection.A reversed-phase HPLC method was used to determine the percent- ages of radioactivity in plasma that corresponded to unchanged radioligand and radiometabolites during the course of a PET measure- ment. The plasma (0.5–1.5 mL) obtained after centrifugation of blood at 2000 g for 2–4 min was mixed with 1.4 times the volume of acetoni- trile. After stirring with a vortex mixer, the sample was centrifuged at 2000 ×g for 2–4 min and 2 mL of water was added to the supernatant plasma–acetonitrile mixture, which was then injected into a radio- HPLC system. Blood (1.0–3.0 mL) and plasma (0.5–1.5 mL) samples were counted in a NaI well counter. The free fraction (fp) of the radioligand in plasma was estimated using an ultrafiltration method as described elsewhere [29]. Plasma (500 μL) or phosphate-buffered saline solution (500 μL) as a control was mixed with the formulation (50 μL, ~ 1 MBq) and incubated at room temperature for 10 min. After the incubation, 200 μL portions of the incubation mixtures were pipetted into ultrafiltration tubes (Centrifree YM-30, molecular weight cutoff, 30,000; Millipore: Billerica, USA) and centrifuged at 1500 ×g for 15 min. Equal aliquots (20 μL) of the ultrafiltrate (Cfree) and of the plasma (Ctotal) were counted for their radioactivity with a NaI well counter. Each determination was per- formed in duplicate. The free fraction was then calculated as fp =Regions of interest (ROIs) for the whole brain were delineated man- ually on the co-registered magnetic resonance imaging (MRI)/PET im- ages for both PET radioligands.
The time activity curves of the ROIs were generated by applying the ROI to the dynamic PET data. The data were expressed as % injected dose (%ID), which is the total uptake (MBq) in the region divided by the injected radioactivity (MBq) × 100. ROIs were also delineated manually on the putamen, caudate, ven- tral striatum (nucleus accumbens), frontal cortex, temporal cortex, thal- amus, and cerebellum on the co-registered MRI images. The time radioactivity curves of the ROIs were generated by applying the ROI to the dynamic PET data. The data were expressed as % standard uptake value (%SUV), which is regional uptake (MBq/cc)/injected radioactivity(MBq) × body weight (g) × 100.For the second part of NHP PET measurements, the total distribution volume (VT) (mL/mL) was calculated with a one-tissue compartment model (1TC) and two-tissue compartment (2TC) model using metabolite-corrected plasma input. Model fitting of 1TC and 2TC was compared using Akaike Information Criterion and Schwarz Criterion. The distribution volume ratio (DVR) was calculated as VT of the target regions divided by VT of the cerebellum. The binding potential (BPND) was calculated as DVR-1. As a reference tissue model, simplified refer- ence tissue model (SRTM) was also evaluated to calculate BPND [30]. Based on the change in BPND, the occupancy of PDE10A was calculated. Correlation of BPND between SRTM and 2TC was also evaluated.
3.Results
[18F]FM-T-773-d2 (Fig. 1) was synthesized from 1-(2-fluoro-4- (tetrahydro-2H-pyran-4-yl)phenyl)-5-hydroxy-3-(1-phenyl-1H-pyrazol-5-yl)pyridazin-4(1H)-one and purified as described above. Radioactivity produced on average was 805 ± 230 MBq and radiochemical purity N 99.8%. MA was 293 ± 114 GBq/μmol (n=8) at the time of administra- tion. QC analysis prior to administration into NHPs revealed no significant UV-adsorbing impurities. [18F]FE-T-773-d4 (Fig. 2) was synthesized from 1-(4-(3,3- dimethyl-2-oxopyrrolidin-1-yl)-2-fluorophenyl)-5-hydroxy-3-(1- phenyl-1H-pyrazol-5-yl)pyridazin-4(1H)-one and purified using methodology identical to the one used for [18F]FM-T-773-d2. Radio- activity produced on average was 546 ± 311 MBq and radiochemical purity N 99%. MA at the time of administration was 209 ± 26 GBq/μmol (n=4). QC analysis prior to administration into NHPs revealed no sig- nificant UV-adsorbing impurities.For the first part of four NHP PET measurements, 148 and 157 MBq of [18F]FM-T-773-d2 and 159 and 163 MBq of [18F]FE-T-773-d4 were ad- ministered to the monkeys, respectively. The summation images of the brain radioactivity show high and heterogeneous uptake of [18F] FM-T-773-d2 and [18F]FE-T-773-d4 (Fig. 3).The peak uptake of the whole brain was approximately 5.5% ID for [18F]FM-T-773-d2 and 3.5% ID for [18F]FE-T-773-d4 (Fig. 4). Washout was fast with uptake half-life of approximately 20 min for [18F]FM-T-773-d2 and approximately 30 min for [18F]FE-T-773-d4. Re- gional time activity curves for both radioligands showed relatively high uptake in the striatal regions (Fig. 5A and B). Specific binding reached equilibrium within 30 min for [18F]FM-T-773-d2 and after approximately 45 min for [18F]FE-T-773-d4 (Fig. 5C). The tissue ratio between the putamen and the cerebellum was approximately three for [18F]FM-T-773-d2 and less than two for [18F]FE-T-773-d4 (Fig. 5D).Radiometabolite analysis showed seven metabolites for [18F]FM-T- 773-d2 and five for [18F]FE-T-773-d4 identifiable through HPLC analysis (Fig. 6). Based on the order of elution from the HPLC column, all radiometabolites, except M7 for [18F]FM-T-773-d2, were less lipophilic than the parent compounds.
Retention time of M7 for [18F]FM-T-773- d2 was 8.3 min, compared with 7.4 min for the parent compound, and the amount of M7 did not exceed 2% of the total radioactivity.Fraction of the parent compound at 90 min after the radioligand in- jection was, on average, 71% for [18F]FM-T-773-d2 and 58% for [18F]FE- T-773-d4. For the 2×2 PET scans performed protein binding was 75.9%± 1.4% and 75.3% ± 3.1% for [18F]FM-T-773-d2 and 71.4% ± 3.0% and68.4% ± 2.1% for [18F]FE-T-773-d4.For occupancy, the second part of NHP PET evaluation, 165.3 ± 8.7 (154–176) MBq of [18F]FM-T-773-d2 was injected into the monkeys. The summation images at baseline and blocking conditions are shown in Fig. 7.Data fit for 1TC and 2TC models for cerebellum (A) and putamen(B) is shown in Fig. 8.AIC and SC showed that there was no significant difference of VT values calculated via the two models. VT values calculated by 2TC model correlated well with the corresponding values obtained using 1TC model — Fig. 9A. Correlation of BPND values calculated using 2TC and SRTM models is shown in Fig. 9B.VT values calculated by 2TC are shown in Table 1. After blocking the binding of [18F]FM-T-773-d2 with 1.8 mg/kg of MP-10, the VT decreased in the caudate and the putamen, while the cerebellum, thalamus and cortex did not show a decrease in VT.At baseline PET, BPND calculated by 2TC was 1.23 ± 0.26 for the pu- tamen and 0.83 ± 0.29 for the caudate while BPND calculated by SRTM was 1.32 ± 0.33 for the putamen and 0.98 ± 0.40 for the caudate. The calculated PDE10A occupancy by 1.8 mg/kg of MP-10 was approximate- ly 60% (Table 2), which was broadly at a level demonstrated earlier for [11C]T-773 [18,31].
4.Discussion
The novel radioligands [18F]FM-T-773-d2 and [18F]FE-T-773-d4 were developed as fluorine-18 PET radioligands for PDE10A. [18F]FM-T-773- d2 shows more promising characteristics as PET radioligand. The data for [18F]FM-T-773-d2 fitted well with the 2TC model. [18F]FM-T-773- d2 and [18F]FE-T-773-d4 can be easily synthesized in high purity and molar activity from a simple O-desmethylated precursor, thus ensuring radioligand dose conditions in the PET measurements for both radioligands. Based on brain uptake, specific binding, and kinetics, it was concluded that [18F]FM-T-773-d2 has better characteristics as PET radioligand than [18F]FE-T-773-d4. The properties of [18F]FM-T-773-d2 closely mimic those of [11C]T-773, a carbon-11-labeled PDE10A- specific ligand described previously and evaluated in monkeys and humans [16–18]. Both radioligands are comparable as far as imaging properties are concerned. The choice of a particular radioligand would depend on the situation and particular study requirements: [11C]T-773 can be used for repeated measurements on the same day for occupancy studies, and is overall slightly easier/faster to produce if appropriate infrastructure is available; while [18F]FM-T-773-d2 can be delivered to more distant imaging facilities that lack on-site cyclotron. The use of deuterated fluorine-18-bearing syntones alleviates the threat of rapid defluorination as observed in some of the cases, in which defluorination has a potential to compromise ligand usefulness via uptake of fluoride-18 in bone [23,24]. Absence of this phenomenon further adds to the usefulness of the radioligand. Given the well- established labeling chemistry, yield, and high molar activity of [18F]FM-T-773-d2, cGMP production of [18F]FM-T-773-d2 is not like- ly to present a significant challenge, thus enabling [18F]FM-T-773-d2 to be used in human studies.
Brain kinetics of [18F]FM-T-773-d2 fit well to 1TC and 2TC models. BPND values obtained via the compartment models correlated well with BPND values from SRTM. Therefore SRTM can be the choice of the quantitative analysis method as long as the assumption that the cere- bellum is used as the reference tissue is valid.There was one radiometabolite, M7, which appeared more lipophilic than the parent compound from the HPLC data. We have not investigat- ed directly whether the radiometabolite entered the brain or not. How- ever, given that the amount was less than 2% of the total radioactivity and the data from the kinetic compartment analysis, it appears that no extra consideration is needed for the role of this particular radiometabolite as the models fit well with the input function of one parent compound.
5.Conclusions
Of two newly developed fluorine-18 PET radioligands, [18F]FM-T- 773-d2 and [18F]FE-T-773-d4, [18F]FM-T-773-d2 showed more promis- ing binding characteristics as a PET radioligand for PDE10A and warrants further PF-2545920 clinical evaluation.