Striatal Transporter Imaging in Nonhuman Primates with Iodine-123-IPT SPECT

Robert T. Malison, Janet M. Vessotskie, Mei-Ping Kung, William McElgin, Gemma Romaniello, Hee-Joung Kim, Mark M. Goodman and Hank F. Kung

Department of Psychiatry, Yale University School ofMedicine, New Haven, Connecticut; Department of Radiology, University of Pennsylvania School of Medicine, Philadelphia, Penn.sylvania;and Department ofRadiolo,@jt,University of Tennessee Medical Center, Knoxville, Tennessee

cells and recycling the residual (1). Phar The regionaldistribution,kineticsand pharmacologk@alspecific macological antagonism of the ap tt)fof a new rad@dinated cocaineanalog,N-((E)-3-iodopropen pears to play a central role in both the addictive properties 2-y@-2p-carbomethoxy-3@3-(4-chloropheny@tropane ff@OlP1) of (2) and the therapeutic actions of certain anti were examined in brain SPECT studies (n = 20) of nonhuman depressant medications (3). In vitro studies in postmortem pnrnates. Methods: Radiolabeling and purification of the iodo human brain have also suggested its importance as a destannylated thalkyftin precursor yielded the tracer at greater marker for monitoring and/or understanding the patho than 90% radiochemicalpurityand high (>20,000 CVmmole) physiology of a variety of neuropsychiatric disorders, in specific actMty. Cynomologous monkeys were injected with 7.2 ±1.3 mCi (mean ±s.d.)ofthetracer, and serial 10-mm images cluding cocaine abuse (4), Tourette's syndrome (5) and were acquired (total scan time = 177 ±22 mm). Images were Parkinson's disease (6,7). As a result, much interest exists reconstructed as transaxial slk@es(2 mm) using restorative tech in the in vivo assessment of the dopamine transporter in niques (Wiener prefittenng). Results Radioactivity concentrated clinical populations. In fact, functional radiotracer tech quicklyin stnatal regions (time of peak = 25 ±13 mm)and niques, including PET and SPEC!', have recently demon cleared gradually thereafter(8.8 ±4.6 %/hr). Stnatal-to-cerebel strated dopamine transporter loss in Parkinson's disease lar ratios of 2.6 ±1.5 (n = 19), 6.7 ±3.2 (n = 20), 15.1 ±10.7 (8,9). (n = 10)and22.8 ±11.0(n = 9)wereobset-vedatthetimeof A variety of radiopharmaceuticals for the dopamine peak and at 1, 2 and 3 hr p.i., respectively. In contrast, extras triatal acthiity peaked eatluer and at lower levels, cleared more transporter have been developed as potential PET and rapidly and resembled cerebellar time-activfty curves. Displac SPEC!' imaging agents (10—17). Among these, recently ingdoses ofnonspecificantagonists ofmonoaminetransporters synthesized analogs of cocaine, in which a phenyltropane ( and j3-Cfl) showed that 95% of specific C9IPT moiety substitutes for the benzoyl ester linkage, have been binding was reversible, while selective antagonists (e.g., parox shown to bind to the dopamine transporter with higher etine, and GBR 12909) demonstrated that striatal affinity, lower nonspecific binding and greater metabolic activfty was specifically associated with dopamine transporters. stability than cocaine itself. Such agents include WIN Conclusion: These results indicate that C23IIIPTis a useful 35,428 (also known as CFT) (8,18), @3-C!T(methyl 3@-(4- radioligand for invivo SPECT imaging of striatal dopaminetrans iodophenyl)tropane-2f3-carboxylate) (also known as RT! porters. 55) (19,20) and IP-CIT (also known as RT!-121) (21 ). In Key Words: iodine-123-IPT; dopamine transporters; single their various radiolabeled forms, these newer agents have photon emission computed tomography offered improvements over prior radiopharmaceuticals J NuciMed1995;362290—2297 with respect to their greater target-to-background ratios and their enhanced brain retention in vivo. Despite these improvements, ongoing research has focused on the devel opment of structurally related derivatives with yet improved ocated on the terminal axons of dopamine neurons, signal-to-noise characteristics, more rapid brain kinetics, the dopamine transporter (i.e., site) is a presyn and greater selectivity for binding to the dopamine as op aptic carrier protein which serves to remove neurotransmit posed to the serotonin (5-HT) transporter. ter from the synaptic cleft following its release, thereby Recently, Goodman et al. reported on a phenyltropane terminating the chemical communication between nerve derivative of cocaine in which a nitrogen-substituted iodovi nyl group was exploited as an alternative radiolabeling scheme (22) (Fig. 1). In its radioiodinated form, N-((E)-3- ReceivedNov.2, 1994;revisionacceptedMar.12,1995. ForcorrespondencecontactRobertT.Malison,MD,DepartmentofPsychia [‘@I]iodopropen-2-yl)-213-carbomethoxy-3f3-(4-chlorophe @ try,YaleUniversitySchoolofMedrJne,34P@kSt.,NewHaven, 06519. Forrepnntscontact:HankF.Kung,PhD,UniversityofPennsylvaniaSchoolof nyl) tropane ([125!@!@'J')appeared to be a particularly prom Med@ine,3700MarketSt, Room305,Philadelphsa,PA19104. ising ligand for the dopamine transporter based on in vitro

2290 The Journal of Nuclear Medicine •Vol. 36 •No. 12 •December 1995 and in vivo studies in the rodent. The purpose of the two-dimensional Wiener prefilter) (23). Images were displayed as current investigation was to characterize the regional dis 2.0-mm thick transaxial slices having pixel dimensions of 1.11 x tribution, kinetics and pharmacological specificity of 1.11mm, resultingin a final voxelvolumeof 2.46mm3.Attenua [l23I]Ip9@'brain uptake in nonhuman primates using SPEC!' tion correction was performed by first-order approximation (24), as an initial step towards establishing its utility as an in vivo assuming uniform attenuation (@= 0.10) within elipses fit to brain imaging agent for the dopamine transporter in humans. contours on a slice-by-slicebasis. The regional distribution and temporal kinetics of [123111}Yf MATERIALS AND METhODS brain uptake were established in a series of control experiments (n = 10)after the administrationof radiotraceralone.Concentra Radiolabeling tions of radioactivitywere determined for multiple brain areas High specific activity (>20,000 Ci/mmole) [‘@I]IPTwas pre (e.g., right and left striatum, hypothalamus/midbrain, occipital pared by iododemetalation of the corresponding tributylstannyl cortex and cerebellum)using native PRISM 3000 software.By precursoras previouslydescribed(22). Hydrogenperoxide(50 @d, summing the first six 10-mm image acquisitions, a composite (i.e., 3% w/v), hydrochloricacid(50 @d,0.1 N) and no-carrier-added 60 mm) imagewas obtained on whichregion of interest (ROI) [‘231]sodiumiodide (5 p.1,specific activity 200,000 Ci/mmole) were added to a septum-sealed vial containing the tin substrate (50 templates were created. Striatal and occipital ROIs were placed on i.tg/50 @dethanol). After 30 mm, peroxidation was interrupted by the same transaxial slice as determined by the brightest striatal adding sodium bisulfite (100 @d,300 mg/ml) and adjusting the pH profile. Midbrain and cerebellar templates were positioned at to 8.5 with saturated sodium bicarbonate. The reaction mixture separate brain levels based on maximal activity and brain contour, was extractedwith ethyl acetate (1 X 3 ml), dried by an annhy respectively. Nonidentical, albeit highly uniform ROI areas (mean drous sodium sulfate column (0.2 x 5 cm) and evaporated under ±s.d. pixels) were achieved across studies for the right (98 ±21) a nitrogen stream. The resulting residue was dissolved in ethanol and left striatum (95 ±19),midbrain(55 ±11), occipitalcortex (50 @l,100%) for reverse-phase HPLC purification using an iso (233 ±38) and cerebellum (275 ±80). Templates were then cratic solvent system (93% acetonitrile, 7% 5 mM 3,3-dimethyl appliedto correspondingslicesof individualimageacquisitionsto glutaric acid buffer, pH = 7.0; flow 1.0 ml/min; retention time obtain regionalmeasuresof relative radioactivedensity (counts @ 11—12mm).Combinedethylacetateextractsof theeffluentwere mm pixel ‘)over time. concentrated under nitrogen and reconstituted in ethanol, yielding (50%—60%overall) the final product at >90% radiochemical purity. Iodine-123-IPT was diluted in normal (0.9%) saline, sterile Pharmacologic Displacements filtered and determined to be pyrogen-freeprior to studies in The reversibilityand pharmacologicalspecificityof regional nonhuman primates. [‘@I]IPTbinding was determined by displacement studies (n = 10) in which nonradioactive drugs were administered 90 mm postin Animal Preparation jection of radiotracer. Various agents were examined, including A series of 20 SPEC!' imagingstudieswere conductedin six nonselective monoamine reuptake inhibitors (i.e., mazindol, 4.7 male cynomologous monkeys (4.0—5.6kg). Following an overnight @.tmolekg ‘;j3-CIT, 0.9 @molekg ‘)(25—27);selective antago fast, animals were immobilized with a combined intramuscular injection of (10 mg/kg) and xylazine (2 mg/kg). Anesthe fists of dopamine (GBR 12909,3.4 and 7.6 @molekg ‘)(28), sia was maintained by passive inhalation of 1.5% isoflurane in serotonin (paroxetine, 3.1 @molekg ‘)(29,30) and norepineph oxygen. Endotracheal intubation and subcutaneous glycopyrrolate tine (nisoxetine 3.0 j.@mole kg ‘)(31) transporters, D1 (2 @gfkg),a peripherallyactinganticholinergicagent, minimized (SCH23390,3.7 @molekg') (32) and D@3(raclopride,1.8 @.tmole risks of aspiration. Core body temperature was kept at >37°C,and kg 1) (33) dopamine receptor antagonists and the mixed dopa vital signs (heart rate and respirations) were monitored every 30 mine releaser/reuptake inhibitor (d-, 2.7 .tmole mm by an indwelling esophageal stethoscope. An intravenous line kg 1). The effects of increases in endogenous neurotransmitter in a superficial saphenous vein allowed maintenance hydration (i.e., dopamine) on radiotracer binding were assessed by an infu (0.9% NaCl, 2 cc kg ‘hr ‘)as well as radioisotope and drug sion of L-DOPA (252 @molekg ‘over 15mm) preceded (60 mm) administration. Animals were placed in a cylindrical polycarbonate by a bolus injection of benserazide (5 mg kg ‘or 20.7 @mole “cradle―equipped with an individualized foam head-holder kg ‘),a peripherally acting DOPA decarboxylase inhibitor. (Smithens Medical Products, Inc.; Tailmadge, OH) to enable re producible head fixation for imaging in the canthomeatal plane. SPECT Data Malysis The time needed to reach peak [1231]IPTuptake was deter Experiments were performed on the Picker PRISM 3000 (Bed mined for individual brain regions in both control and pharmaco ford Heights, OH), a triple-head SPEC!' camera equipped with logicaldisplacementexperiments.Striatalactivitywasanalyzedas low-energy, high-resolution (LEHR) fanbeam collimators. In this configuration, the instrument's tomographic in-plane resolution the weighted (by ROI area) average of bilateral count densities. was determined to be 10.95 mm (FWHM) based on the point Ratiosof striatal-to-cerebellaractivitiesprovidedestimatesof the spread function of an 1231line source in scattering medium. Fol ligand's in vivo signal-to-noise properties at early (i.e., peak and lowing intravenous injection of 7.2 ±1.3 mCi [‘@I}IPT,monkeys 1hr p.i.;allstudies)and late (2 and 3 hr p.i.;controlstudiesonly) underwentserial 10-mixSPECTscansfor 177 ±22 mm (range imagingtimes.Basal(i.e.,postpeak)and drug-induced(i.e.,post 120—240)using a symmetrical 20% ‘@Ienergy window (i.e., 159 ± displacer) rates of brain [‘@I]IPTwashout were calculated as the 16 keV), a 128 x 128 matrix, 120projections over 360°and a percent changeper hour (as estimatedby linear regression)rela magnification factor of 1.78. Individual image acquisitions were tive to peak or predisplacer levels, respectively. Values are cited as reconstructed using restorative techniques (i.e., a count-dependent the mean ±s.d.

Dopamine Transporter Imaging with lodine-123-IPT •Malison at al. 2291 A B

N COOCH3

FIGURE1. Thechemicalstructureof [123l]IPT.

RESULTS — Regional Distiibution and Kinetics C D Iodine-123-IPT distributed rapidly and globally through out the brain on the initial (0—10mm) SPEC!' image, consistent with a pattern of cerebral blood flow. By 20—30 mm p.i., imaged activity visibly concentrated in striatal regions bilaterally. Although extrastriatal activity appeared uniform on individual 10-mm scans, composite (i.e., 0—60 ‘S mm summed) images suggested a discrete area of midline [123J]If@!'uptake of similar intensity to occipital and cere bellar regions (Fig. 2). This area of so-called “hypothalam ic/midbrain― activity was located roughly 4—6slices (8—12 mm) below the level of maximal striatal uptake. FiGURE 2@Two pars of transaxialslicestakenfrom composfte Iodine-123-IPT achieved peak concentrations in the stri 0-60 mm ç@and B)and 120-160 mm(Cand D)SPECT images from atum by 25 ±13 mm and cleared gradually thereafter a controlexpenmentina cynomologousmonkeyafterthe injection (8.8 ±4.6 %/hr, n = 10). In contrast, activity in extrastriatal of 5.5 mCi [email protected] in vivodist,ibution of r@I]lPTbrain activity regions, including the hypothalamus/midbrain, occipital at the level of the striatum and occipital cortex (A and C) and the cortex and cerebellum, peaked earlier, attained lower peak hypothalamus/midbrain and cerebellum (Band D)are depicted. The highest concentrations of @23qIp-@-activity were observed in the levels and cleared more quickly (Table 1). The kinetics of striatal regions, with a faint region of early (0-60 mm)midline (des F‘23I]I@@'uptakeandclearanceinextrastriatalregionswereIgnated as hypothalamus/midbrain) activity(B).ExtracranialactMty comparable. In fact, radiotracer binding in hypothalamic/ in the nose was non-displaceable. midbrain and occipital regions closely resembled time-ac tivity curves for the cerebellum (Fig. 3). Despite a steady decline in striatal [‘@I]IPTactivity after 30 mm, striatal-to Pbermacoiogic Displacement Studies cerebellar ratios continued to climb, equaling 2.6 ±1.5 Given the reasonably slow rates of striatal washout (n = 19), 6.7 ±3.2 (n = 20), 15.1 ±10.7 (n = 10) and 22.8 (13.7%/hr), it waspossibleto assessthe in vivo specificity ±11.0 (n = 9) at the time of peak and at 1, 2 and 3 hr p.i., of [1@I]IPTbinding in striatum by measuring the change in respectively. In contrast, ratios for hypothalamus/midbrain washout induced by individual displacing drugs. Similar and occipital cortex remained close to unity for the dura assessments were not possible for midbrain and occipital tion of the control studies (Table 2). regions due to the rapid clearance of tracer at early times

Test-Retest Expetiments TABLE I Four cynomologous monkeys underwent two control Summary of Regionallodine-123-IPTKinetics* studies each to establish the variability in striatal washout •flmeto Peak Peak Activity rates across animals and the test-retest reliability of wash (%/hr)StriatumRegion (Minutes) (% Striatum)Washoutt out within animals. Washout rates averaged 9.4 ±4.7 %/hr (n = 8 experiments; range 2.8—13.7%/hr) in the four 1009±7Hypothalamus/25±13 78±671±12MidbrainOccipital11±3 monkeys (Table 3). A comparison of test and retest scans demonstrated a mean absolute difference in washout rates 13cortexCerebellum 11 ±3 70 ±1472 ± of 2.3 ±2.3 %/hr (range 0.3—4.8%/hr), indicating a lesser degree of variability in striatal [123I]IPT clearance for stud 80±1274±12‘Values10±0 ies performed within an animal. For reasons unclear, re producibility appeared to be better for animals having expresseds.d.for as the mean ± faster washout rates (absolute difference = 0.3—0.4%/hr). the hourfollowingpeakuptake.

2292 The Journal of NuclearMedicine•Vol.36 •No. 12 •December 1995 TABLE 3 Test-Retest Differences in Striatal Iodine-123-IPT Washout (Percent Change/Hour) AbsoluteAnimal —0-—Stñatum No.TestRetestdifference55226.32.83.555218.73.94.8505313213.50.3209313.713.30.4Mean —+-— Midbrain --.-8..-. O@pitaI H —0---- Cerebellum ±s.d.9.4 4.72.3 2.3

120 Time(mEn) Displacements were dramatic, as depicted by changes in FiGURE 3 Regionaltime-activitycurvesfor stnatum,midbrain, striatal time-activity curves following @3-CIT(Fig. 4). In occipital cortex and cerebellum from the experiment depicted in contrast, highly selective , serotonin and do Figure2. Stnatal activity peaked by 20 mm after radiotracer injection and washed out graduallythereafter (13.7 %/hr). Incontrast, extra pammnereuptake inhibitors demonstrated variable effects striatal (i.e.,midbrain, occipital and cerebellar) activitypeaked early (Fig. 5). Nisoxetine, an agent with selective affinity at the (by 10 mm p.i.)at lower levels and cleared much more rapidly. norepinephrine transporter alone (31,34) failed to induce any noticeable change in [‘23I]I@@fwashout (z@%/hr = 0.0). Similarly, the selective serotonin reuptake inhibitor parox (30—90mm)andthelowlevelsof residualactivityatlater etine had little effect on [‘@I]IPTwashout, with postdis times (>90 mm). Given the variability in tracer washout placer rates increasing only marginally (1.0%/br). Values across animals (Table 3) and the impracticality of perform for nisoxetine and paroxetine were well within the range of ing all displacement studies in a single animal, pharmaco values determined from all nondisplacement studies (—1.9 logically-induced changes in the brain clearance of ±3.8 %/hr; range —9.7to 4.1%/br; n = 10) and the range [123J]I@@!'were determined relative to basal clearance rates of control values calculated from a single experiment in within the same experiment. Specifically, the percent each of the five monkeys in whom displacements were change per hour in [123I]I}Yfbinding was calculated for the performed (—0.3±2.9 %fhr; range —3.0to 4.1 %/hr; n = 60 mm preceding (i.e., 30—90mm p.i.) and the 60 mm 5). In stark contrast to nisoxetine and paroxetine, the highly following (i.e., 90—150mm p.i.) the administration of dis selective dopamine transporter antagonist GBR 12909 placing agents. The net change (z@%/hr) in clearance (i.e., caused clear dose-dependent increases in striatal [123I]IPT not transporter occupancy) was obtained by subtracting the clearance or 24.0 and 83.3%/hr for low (3.4 p@molekg ‘) predrug from postdrug washout rates and provided a qual and high (7.6 p@molekg ‘)doses, respectively. As was the itative, albeit objective, means of establishing whether a case for nonselective antagonists (i.e., mazindol and given agent displaced [123I]IJY!'binding. Changes in brain washout were examined over identical intervals for nondis placement (i.e., control) studies as well to assess normal 40. differences in early compared to late washout rates. B-CIT (0.9@smo1/kg) Nonspecific antagonists of monoamine transporters, in / cluding mazindol and a-CIT, produced robust reductions in 30. [‘23I]I@@!'binding as indicated by net changes in striatal —0-- Striatum washout (i.e., L@%/hr) of 47.8 and 83.0%/hr, respectively. I! -0--- Midbrain ----0―-- Cerebellum TABLE 2 Regional Brain Activity Ratios over Time Activityratio(region/cerebellum)* Region p@t 1 hr@ 2 hr@ 3 hr@ 0 60 120 180 Tw@ie(mm) Striatum 2.6 ±1.5 6.7 ±3.2 15.1 ±10.7 22.8 ±11.0 Hypothalamus/ 1.0 ±0.1 1.4 ±0.3 1.1 ±0.3 0.7 ±0.5 FiGURE 4. Pharmacological displacement of striatal [1231]IPTup Midbrain take by the nonselective monoamine reuptake inhibitor(3-CIT(Le., Occipitalcortex 0.9 ±0.1 1.0 ±0.1 1.0 ±0.1 1.1 ±0.2 RTI-55;0.9 @tmoI&kgi.v.).Baseline (30-90 mm) washout rates in striatum(18%/hr)increasedtolOl %/hrafter@3-C1r,anetchange in r@olPTwashout of 83 %/hr. Specffic(i.e.,Striatal-cerebellar) *M@, ±s.d. C@OlPTbinding was reduced by —95%(calculated as [Specific@ tContrc@Jand displacementstudies. Specific180l/Specffic90* 100), suggesting that the vast majorityof @Controlstudies only. r@°olPTbindingisreversiblyboundtomonoaminetransporters.

DopamineTransporter ImagingwithIodine-123-IPT•Malisonat al. 2293 13-CIT), the magnitude of change in washout rates for GBR 12909 easily exceeded the 99% confidence interval (i.e., 3 s.d.) for control values. Although the mixed dopamine releaser/ reuptake inhibitor d-amphetamine produced reductions in [‘@I]WTbinding (35.2%/hr), an infusion of 1-DOPA ap peared to have no effect (0.5%/kr). Nor were significant changes in [‘@I]IPTwashout produced by SCH23390 (—03%/hr)or raclopride (—4.4%/hr),dopamine D1 and D½ receptor antagonists, respectively. Results from these pharma cological displacement studies are summarized in Figure 6.

DISCUSSION This study demonstrates the utility of [‘231]IPTfor imag ing striatal dopamine transporters in nonhuman primates with SPECT. While other iodinated probes for the dopa 30 mine transporter have been described (10,16,26,35,36), [‘23I]I@@!'may provide significantadvantagesover currently available SPEC!' tracers given its superior target-to-back ground ratio, reversible binding, rapid brain kinetics and ap 20 parent pharmacological specificity for dopamine transporter sites in vivo. As such, [‘@I]IPTis a particularly promising candidate for clinical investigations in human subjects. Regional Distribution 10 Iodine-123-IPT attained high concentrations in monkey striatum, while much lower levels of activity were observed in extrastriatal regions (Fig. 2). Ratios of striatal-to-cere bellar activity as measured by our SPEC!' camera were 15.1 0 at 2 hr and 22.8 at 3 hr p.i. Such values are the highest 0 60 120 180 reported to date for any dopamine transporter imaging GBR12909(7.6@tmol/kg) agent, including ratios for [1@I]j3-CIT/RTI-55 (7.3 at 5 hr 30 and 12 at 21 hr p.m.)(19,20), [11C]CFT/WIN35,428 (2.5—3.0 at 2 hr p.i.) (18) and [‘@IJIP-13-CIT/RTI121 (7.0 at 3.2 hr p.i.) (21) in nonhuman primates. The excellent contrast of [‘@I]IPTis consistentwithwork in the rat in whichin vivo 20' ratios of basal ganglia-to-cerebellar activities reached 16:1 by 1 hr after injection (22). Reasons for [‘@I]WT'srelatively enhanced signal-to-noise characteristics are not entirely clear, sinceothertropanecongenerspossesscomparablein vitro 10@ affinity for striatal binding sites in rats (37). Thus, other factors such as ligand lipophilicityand in vivometabolism may help to explain the differences between radiopharmaceuticals. Both affinity and lipophilicity have been demonstrated to influence target-to-background ratios for D½receptor radio 0 60 120 180 ligands (38). Since absolute uptake (percent dose/gram of Time (minutes) tissue) of [‘@I]Iflin rat whole brain (0.13 ±0.02 at 120 mm) FIGURE 5. The effects of selective norepinephrine (NE),seroto and basal ganglia (0.61 ±0.07 at 120 min) was only modest, nm (5-Hi) and dopamine (D,4)transporter antagonists on brain low levels of nonspecific binding may account for the high [‘23@lp@uptake. Nisoxetine(3.0 @imole/kgi.v.),a selectivenorepi striatal-to-cerebellar ratios. Of note, encouraging preliminary nephnne reuptake inhibitor, and paroxetine (3.1 @.tniole/kgi.v.),a results in humans suggest a similarly robust ligand signal (stri highly selective antagonist, failed to increase [123fllp@washoutrates inthe striatum(netchanges = 0 %/hrand 1 atal/cerebellaractivity= 26at 3 hr (39). %/hr, respectively). In contrast, GBR 12909 (7.6 pn,ole/kg i.v.),a Tracer Reversibility highlyselective antagonist at the dopaminetransporter, dramatically decreased specific striatal C23IIIPTbinding @[email protected]@]/ Despite steadily increasing striatal-to-cerebellar ratios @ Specific90* = 94%), suggesting that sttiatal r231]IPTactivityis (Table 2), [‘@I]IPTsatisfies both pharmacologic and ki exclusivelyassociatedwiththedopaminetransporter.Reductionsin netic criteria for a reversibly binding tracer in vivo. Non specific binding were consistent with the robust differences in pre specific antagonists of monoamine transporters (i.e., mazin (0 %/hr)versus postdisplacer (83%/br) washout. dol and 13-CIT) produced dramatic displacements of

2294 The Journal of NuclearMedicine•Vol.36 •No. 12 •December 1995 Controls

Raclopride SCH23390

Nlsoxetlne FIGURE 6. A bar graph summanzing the results of pharmacological displacement ex Paroxetlne penments in five cynomologous monkeys. Horizontalbars indicate the net change (%/hr) MaZIndOI in stnatal [123@lp@clearance, as derived by 6-CIT subtracting pre- (30-90 mm) from postdis placer (90-150 mm) washout rates. Resutts GBR (3.6) from individual displacement studies are GBR (7.6) compared to the mean ±s.d. (—0.3±2.9) values from control experiments (n = 5) inthe same animals. The dashed line denotes a level of change in [123@]lp@washout of 3 s.d. Change In % Washout above that observed for control studies.

[‘@I]IPTbinding from the striatum. Changes in washout with high affinity to transporters for both dopamine (K@= 0.3 were visually unambiguous (Fig. 3) and statistically distinct nM) and serotonin (K@ = 2.1 nM) (M-P Kung, personal (3 s.d.) from control values (Fig. 6). For f3-CIT (Fig. 4), communication). Although the highest concentrations of do specific (i.e., striatal-cerebellar) [‘@I]IPTbinding after dis pamine transporters are found in the striatum (49), modest placement (i.e., 180 mm p.i.) was reduced to —5%of levelsof striatal serotonin reuptake sites also exist (50). Thus, preinjection (i.e., 90 mm p.i.) levels, a value essentially our displacement studies were specificallydesigned to deter equivalent to background (i.e., cerebellar) values. In con mine whether striatal [‘@I]IPTactivity was primarily associ trast, no discernible change in cerebellar activity was ob ated with dopamine or serotonin reuptake sites. served, arguing against a nonspecific influence of displacers As noted above, dramatic displacements of [‘231]IPT on peripheral tracer metabolism. This near complete re binding were observed for mazindol and n-CIT. Both versibility of [123I]IPTbinding indicates that the vast ma agents are nonselective in their actions at monoamine re jority (—95%)of specific striatal activity is associated with uptake sites. Mazindol potently inhibits dopamine, seroto transporter binding sites. In addition, the comparable levels nm and norepinephrine uptake, possessing half-maximum of postdisplacement activity in striatum and cerebellum sug (i.e., IC50) inhibition values of 18, 30 and 0.65 nM, respec gests the suitabilityof using the latter region as a measure of tively (51). Similarly, @3-CIThas been well characterized in nonspecific binding during in vivo imaging studies. vitro and shows nanomolar affinity for both dopamine In addition to pharmacologic reversibility, [1231]IVFbe (1@ihigh 0.2nM; K,@iow = 5.8nM) and serotonin (K@ = 0.2 haves in a kinetically reversible fashion as well. Striatal nM) transporters (27). In contrast to these nonselective [‘23I]I@'factivity peaked by 30 mm in all but two experi agents, the diphenyl-substituted derivative, ments, clearing at a steady rate (—9%per hour) postpeak. GBR 12909, has been shown to have greater selectivityfor The presence of ascending and descending components to inhibiting dopamine as compared to serotonin (‘—170fold) striatal time-activity curves suggests in vivo phases of net and norepinephrine (—440fold) uptake (28). Injections of association and dissociation, respectively (40). Thus, 3.4 and 7.6 @mole/kgproduced dose-dependent increases [‘23I]I@@fmay lend itself to a varietyof model-based meth in striatal [1BI]IPT washout. For the highest dose, GBR ods for measuring dopamine transporter levels in vivo, displaced 94% of specific [123I]I}Y!'binding in striatum (Fig. including tracer kinetic, graphical and equilibrium (e.g., 5), equalling the 95% reduction seen with a-CIT. Consis peak, transient and sustained/constant infusion) approaches tent with near total displacement of [‘23I]IPTby GBR (40). Such versatility allows for the cross-validation of out 12909 was the absence of any effect of serotonin- and come measures obtained by alternate approaches, as prey norepinephrine-selective agents. Paroxetine, a greater than ously demonstrated with the SPEC!' D@ receptor radioligand 10,000 fold more potent serotonin than dopamine reuptake [‘@I]iodobenzofuran(i.e., [‘@I]IBF)(41,42). inhibitor (51), produced no noticeable displacement of Pharmacological Specificity [1BI]IPT binding. Similarly, the highly selective norepi A major aim of this study was to characterizethe phar nephrine transporter antagonist nisoxetine (31,34) showed macologic nature of striatal [‘@I]IPTbinding in vivo. Like no evidence of change in [l23I]I@Yfwashout. Taken to cocaine and various other phenyltropane analogs (e.g., gether, these displacement data strongly suggest that stria 13-Cr!',CFT)(27,43—48),[‘@IJIPTpossessesinvitroaffinity tal [‘@I]IPTactivity is almost exclusively (—95%)associ for both dopamine and serotonin transporters. Prior work in ated with dopamine transporter sites in vivo. rodent tissue homogenates has shown that [‘@I]IPTbinds The apparent in vivo specificity of [‘23I]IPTbinding,

Dopamine Transporter Imaging with lodine-123-IPT •Malison et al. 2295 while unexpected, may be explained by prior work with a higher degree of nonspecific binding cannot be excluded. [‘@I1IPT,cocaine and related phenyltropane imaging Reports on in vivo imaging studies using [11C]13-CITPET agents. First, our imaging data are consistent with in vitro (54) and [‘@IJ13-CITSPEC!' (20) have noted significant studies performed with [‘@I]IPTin the rat. Competition degrees of “thalamic―or “hypothalamic/midbrain―binding, binding data demonstrate a rank order potency for inhib respectively. Pharmacological characterization of this ex iting basal ganglia [‘@I]IPTbinding which parallel antago trastriatal activity by both groups has suggested its associ nist affinity for the dopamine transporter (22). Similarly, in ation with the serotonin transporter. Whether the relatively vitro studies in human (52) and nonhuman (53) primate lower levels of [‘@IIIPTactivity observed in this region are brain suggest that [3Hjcocaine recognition sites in the stri due to its lesser affinity for the serotonin transporter (2.1 atum are principally associated with the dopamine trans versus 0.2 nM in vitro), an increase in partial volume effects porter, despite cocaine's known affinity for other monoam as a result of using smaller primates (20) or a lower reso me reuptake sites (2). Lastly, in vivo imaging studies in lution instrument (54) or other in vivo factors (38,55) monkeys using radiolabelled f3-CIT demonstrate a similar remains to be determined. regional selectivity for the dopamine transporter (20,54). Specifically, a-CIT activity in striatum appeared to be ex CONCLUSION clusively associated with the dopamine transporter despite We have shown that [‘@I]IPTis a promising new SPEC!' clear evidence of serotonin transporter binding in extras ligand for imaging striatal dopamine transporters in non triatal regions. Thus, one possible explanation for the in human primates. Its extremely favorable target-to-back vivo specificity of [‘@I]IPTis a predominance of dopamine ground binding, reversible kinetics and pharmacological over serotonin reuptake sites in the striatum. Alternatively, specificity in vivo make it an attractive addition to currently in vitro measures of ligand affinity may not always reliably available radiopharmaceuticals for the dopamine trans predict transporter binding in vivo (55). porter. Studies in human subjects are underway (39) and Despite [‘@I]IVF'sclear pharmacologic reversibility, it is should help to establish the ligand's ultimate suitability for noteworthy that 1-DOPA administration failed to change [l23I]IJyfwashoutinthestriatum.Thedoseschosenhavestudies of clinical populations. been shown to result in plasma 1-DOPA concentrations of ACKNOWLEDGMENTS 5—15j.tg/ml in baboons (20), levels above those found to The principalinvestigator(RTM)conductedthese studieswith ameliorate Parkinson's disease symptoms (56). Thus, the supportof a Departmentof VeteransAffairsSubstanceAbuse [123I]I@@fbinding may be relatively immune to changes in Research Fellowship, a National Institute on Drug Abuse Institu endogenous dopamine levels. Since the mixed dopamine tional Research Scientist Development Award and a National releaser/transporter antagonist d-amphetamine produced Alliance for Research on Schizophrenia and Depression substantial changes in [‘23I]IPTbinding, we cannot exclude (NARSAD) Young Investigator Award. the possibility that more dramatic elevations in extracell ular dopamine might have an effect. For the dose of REFERENCES d-amphetamine used, microdialysis studies in vervet 1. Rudnick 0, Clark J. From synapse to vesicle: the reuptake and storage of monkeys have suggested changes on the order of 40-fold biogenic amine . Biochim BiophysActa 1993;1144:249—263. (BradberryCW,personalcommunication). 2. RitzM, LambR,GoldbergS,KuharM.Cocainereceptorsondopamine transporters are related to self-administration of cocaine. Science 1987;237: 1219—1223. Exfresthstal A@ 3. Kapur S, Mann J. Role ofthe dopamine system in depression. BiolPsychiatiy Among other brain regions, only activity in the hypothal 1992;32:1—17. 4. Little K, KirkmanJ, Carroll F, Clark 1, Duncan G. Cocaine use increases amus/midbrain attained levels above those observed in the I3HIWIN35428bindingsitesin humanstriatum.BrainRes1993;628:17—25. cerebellum. Ratios were increased transiently (at 60 mm) 5. Singer H, Hahn I, Moran T. burette's syndrome: abnormal dopamine and marginally (1.4 ±0.3) above unity in control studies uptake sites in postmortem striatum from patients with Tourette's syndrome. Ann Neurol 1991;30:558—562. (Table 2). While visual inspection of time-activity curves at 6. Janowsky A, Vocci F, Berger P. et al. (3HIGBR 12935 binding to the times suggested an effect of displacers (Fig. 5), no objective dopamine transporter is decreased in the caudate nucleus in Parkinson's differences in activity ratios were found when displacement disease. I Neurochem 1987;46:617—621. 7. Kaufman M, Madras B. Severe depletion of cocaine recognition sites asso studies were compared to values obtained in control exper ciated with the dopamine transporterin Parkinson's diseased striatum. Syn iments. As a result, the pharmacological nature of [‘@I]IPT apse1991;9:43—49. binding in this region is unclear. Serotonin reuptake sites 8. FrostJ,RosierA, ReichS,et a!.Positronemissiontomographicimagingof the dopamine transporterwith ‘‘C-WIN35,428 reveals marked declines in are widely distributed in the brain, including a variety of mild Parkinson's disease. Ann Neurol 1993;34:423—431. diencephalic and mesencephalic nuclei (e.g., thalamus, hy 9. InnisR, SeibylJ, ScanleyB, et al.Single-photonemissioncomputedtomo pothalamus, substantia nigra, superior and inferior colliculi graphicimagingdemonstrateslossofstriataldopaminetransportersinPar kinson disease. P@c NadAcad Sd USA 1993;90:11965—11969. and the dorsal raphe) (57—61). In addition, considerable 10. Basmadjian0, Mills 5, KanvindeM, BasmadjianN. Structurebiodistribu numbers of dopamine reuptake sites are known to exist in tion relationship of radioiodinated tropeines: search for a molecular probe extrastriatal regions such as the substantia nigra (62). Thus, for the characterization of the cocaine receptor. I Lab Compd Radiopharm 1990;28:801—810. the low levels of activity in this region could reflect binding 11. Ding Y, Fowler J, Volkow N, et al., Synthesis of [“CJdl-threo-methylpheni to either dopamine or serotonin transporters. Alternatively, date: drug pharmacokinetics and binding to the presynaptic neuron. 10th

2296 The Journal of NuclearMedicine•Vol.36 •No. 12 •December 1995 International Symposium on Radiopharmaceutical Chemistry, Kyoto, Japan, 38. Kessler R, Ansari M, de Paulis T, et al High affinity dopamine D2 receptor October 25—28,1993. radioligands. 1. Regional rat brain distribution of iodinated benzamides. I 12. Fowler J, Volkow N, Wolf A, et al. Mapping cocaine binding sites in human Nuc! Med 1991;32:1593—1600. and baboon brain in vivo. Synapse 1989;4:371—377. 39. Mozley P, Kim H, Stubbs J, et al. Brain and cardiac uptake of 1-123 IPT: an 13. Kilbourn M, Haka M, Muiholland 0, Jewett D, Kuhi D. Synthesis of analog of cocaine [Abstract). I Nuc! Med 1994;35:129P. radiolabeled inhibitors of presynaptic monoamine uptake systems: 40. MalisonR, LaruelleM,InnisR. Positronandsingle-photonemissionto [‘8FJGBR13119(DA), [“Cinisoxetine(NE),and[“Cjfluoxetine(5-FiT).J mography:principlesand applications in psychopharmacology.In:Kupfer D, Lab Compd Radiopharm 1989;26:412. Bloom F, eds. Psychopharmaco!ogy: the fourth generation ofprogress. New 14. Leenders K, Salmon E, Tyrrell P, et a!. The nigrostriatal dopaminergic York:Raven;1994:865—879. system assessed in vivo by PET in healthy volunteer subjects and patients 41. Laruelle M, van Dyck C, Abi-Dargham A, et al. Compartmentalmodeling of with Parkinson's disease. Arch Neurol 1990;47:1290—1298. iodine-123-iodobenzofuran binding to dopamine D2 receptors in healthy 15.MullerL, HaildinC,FardeL, etal.Carbon-11-@-CIT,acocaineanalogue subjects. I Nuc! Med 1994;35:743—754. preparation,autoradiographyand preiminaiy PET investigations.NuciMed 42. Kung M-P, Kung H, Billings J, Yang Y, Murphy R, Alavi A. The character Biol 1993;20:249—255. ization of IBF as a new selective dopamine D-2 receptor imaging agent. I 16. Yu D, Gatley 5, Wolf A, et al. Synthesis of carbon-li labeled iodinated Nuc! Med 1990;31:648—654. cocaine derivatives and their distribution in baboon brain measured using 43. Kennedy L, Hanbauer I. Sodium-sensitive cocaine binding to rat striatal positron emission tomography. I Med Chem 1992;35:2178—2183. membrane: possible relationship to dopamine uptake sites. I Neurochem 17. Maziere B, Coenen H, Halidin C, Nagren K, Pike V. PET radioligands for 1983;41:172—178. dopamine receptors and re-uptake sites—chemistsyand biochemistry.Nuci 44. ReithME,SershenH,AllenDL,LajthaA.Aportionof[3H]cocainebinding Med Biol1992;19:497—512. in brain is associated with serotonergic neurons. Mo! Pharmaco! 1983;23: 18.WongDF,YungB,DannalsRF,etal Invivoimagingofbaboonandhuman 600-606. dopamine transporters by positron emission tomography using [‘1CIWIN 45. Baja J, Patel A, Carroll F, et al. Iodine-125-RTI-55: a potent ligand for 35,428. Synapse 1993;15:130—42. dopaminetransporters.Eurl Pharmaco!1991;194:133—134. 19. Shaya E, Scheffel U, Dannals R, et al. In vivo imaging ofdopamine reuptake 46. ScheffelU,DannalsR,ClineE,et al.Iodine-l23fIodine-125-RTI-55,an in sites in the primate brain using single-photon emission computed tomogra vivo label for the serotonin transporter. Synapse 1992;11:134—139. phy (SPED') and iodine-123 labeled RTI-55. Synapse 1992;10:169—172. 47. Staley JK, Basile M, Flynn DD, Mash DC. Visualizing dopamine and sero 20. Laruelle M, Baldwin R, Malison R, et aL SPECT imaging of dopamine and tonin transporters in the human brain with the potent cocaine analogue serotonin transporters with [@I]@-CIT: pharmacological characterization of [‘@‘I]RTI-55:in vitro binding and autoradiographic characterization. J Neu brain uptake in nonhuman primates. Synapse 1993;13:295—309. rochem 1994;62:549—556. 21. Scheffel U, Dannals RF, Wong DF, Yokoi F, Carroll F!, Kuhar MJ. Dopa 48. LittleK, KirkmanJ, CarrollF, BreeseG, Duncan0. Iodine-125-RTI-55 mine transporter imaging with novel, selective cocaine analogs. Neumreport binding to cocaine-sensitive dopaminergic and serotonergic uptake sites in 1992;3:969—72. thehumanbrain.I Neurochem1993;61:1996—2006. 22. Goodman M, Kung M, Kabalka G, Kung H, Switzer R. Synthesis and 49. RichfieldE. Quantitativeautoradiographyof thedopamineuptakecomplex characterization of radioiodinated N-(3-iodopropen-2-yl)-2@-carbomethoxy in ratbrainusing[3H]GBR 12935:bindingcharacteristics.BrainRes1991; 3@3-(4-chlorophenyl)tropanes:potential dopamine reuptake site imaging 541:1—13. agents. I Med Chem 1994;37:1535—1542. 50. Hensler J, Feny R, Labow D, Kovachich G, Frazer A. Quantitative autora 23.ParkerJ.Imagereconstnsctioninmdiologj.BocaRaton,FL@CRCPress;1990. diographyof the serotonintransporterto assessthe distributionof seroton 24. Chang L. A method for attenuation correction in computed tomography. ergicprojectionsfromthedorsalraphenucleus.Synapse1994;17:1—15. IEEE Trans Nuci Sci 1987;25:638—643. 25.JavitchJ, BlausteinR,SnyderS.Tritiatedmazindolbindingassociatedwith 51. Hyttel J. Citalopram: pharmacological profile of a specific serotonin uptake inhibitor with antidepressant activity. Prog Neurop@ychopharmaco!Bio! Psy neuronal dopamine and norepinephrine uptake sites. Mo! Phwmaco! 1984; chiatiy1982;6:277—295. 26:35—44. 52. Biegon A, Dillon K, Volkow N, Hitzemann R, Fowler J, Wolf A. Quantita 26. Neumeyer J, Wang 5, Milius R, et al. [Iodine-123-]2-B-carbomethoxy-3-B- tive autoradiographyof cocaine binding sites in human brain postmortem. (4-iodophenyl)tropane: high-affmity SPEC!' radiotracer of monoamine re Synapse 1992;10:126—130. uptake sites in brain. I Med Chem 1991;34:3144—3146. 27. Boja J, Mitchell W, Patel A, et al. High-affinitybinding of [‘@‘IJRTI-55to 53. Madras B, Fahey M, Bergman J, Canfield D, Spealman R. Effects of cocaine dopamine and serotonin transporters in rat brain. Synapse 1992;12:27—36. and relateddrugsin nonhumanprimates.I. [3HjCocainebindingsites in 28.AndersenP. The dopamineuptakeinhibitorGBR 12909:selectivityand caudate-putamen. I Pharmacol E*p Ther 1989;251:131—141. molecular mechanism of action. EurJ Pharmaco! 1989;166:493—504. 54. FardeL,HalldinC, MullerL,SuharaT, KarlssonP, HallH. PETstudyof 29. Habert E, Graham D, Tahraoui L, aausti@eY, Langer S. Characterlzation of [“C]@-CITbindingto monoaminetransportersin the monkeyandhuman [3Hlparoxetine binding to rat cortical membranes. Eur I Pharmacol 1985; brain. Synapse 1994;16:93—103. 118:107—114. 55. Laruelle M, Giddings 5, Zea-Ponce Y, et al. Methyl 3@-(4-['@'I]iodophenyl) 30. Marcusson J, Bergstrom M, Eriksson K, Ross S. Characterization of [3H]p- tropane-2@-carboxylatein vitro binding to dopamine and serotonin trans aroxetine binding in rat brain.I Neurochem 1986;50:1783—1790. porters under “physiological―conditions. I Neurochem 1994;62:978—986. 31. Tejani-Butt S. [3H]Nisoxetine: a radioligand for quantitation of norepineph 56. BredbergE, TedroffJ, AquiloniusS, PaalzowL. Pharmacokineticsand rine uptake sites by autoradiography or by homogenate binding.JPharinaco! effectsof levodopain advancedParkinson'sdisease.EurI ClinPharmaco! Eq@Ther 1992;260:427—36. 1990;39:358—389. 32. Hyttel J. SCH23390—the first selective dopamine D-1 antagonist. Eur I 57. Cortes R, Soranoi E, Pazos A, Probst A, Palacios J. Autoradiography of Pharmaco! 1983;91:153—154. antidepressant binding sites in the human brain: localization using [3H]imi 33. Landwehrmeyer B, Mengod 0, Palacios JM. Differential visualization of pramineand[3H]paroxetine.Neuroscience1988;27:473—496. dopamineD2 andD3 receptorsitesin ratbrain.A comparath'estudyusing 58. DahlstrômA, Fuxe K. Evidence for the existence of monoamine-containing in situ hybridizationhistochemistryand ligand binding autoradiography.Eur neurons in the central-nervous system. I. Demonstration of monoamines in I Neurosci 1993;5:145—53. the cell bodies of brain stem neurons. Acta Physio!Scand 1964;62:1—55. 34.Tejani-Butt5,BrunswickD,FrazerA.Thtiatednisoxetine:a newradioligand 59. Felten D, Laties A, Carpenter M. Monoamine-containing cell bodies in the for norepinephrineuptake sites in brain.EurlPhannaco! 1990191:239—43. squirrel monkey brain.Am lAna: 1974;139:153—166. 35. Baldwin R, Zea-Ponce Y, Al-Tikriti M, Ct al. Regional brain uptake and 60. LaruelleM, VanisbergM-A,MaloteauxJ-M. Regionaland subcellular pharmacokinetics of [‘@I]N-w-fluoroallcyl-2fi-carboxy-3@-(4-iodophenyl)- localizationinhumanbrainof[3H]paroxetinebinding,a markerofseroto nortropaneesters in baboons.NuciMedBiol 1994;22:211—219. fin uptake sites.Bk,!Psychüwy1988;24:299. 36. Carroll F, Gao Y, Rahman M, Ctal. Synthesis, ligand binding, QSAR and 61. BackstrOmI, BergstromM, MarcussonJ. High affinity[3H1paroxetine COMFAstudy of 3@-(p-substitutedphenyl)[email protected] bindingto serotoninuptakesitesinhumanbraintissue.BrainRes1989;486: methyl esters. I Med Chem 1991;34:2719—2725. 261-268. 37. Carroll F, Lewin A, Boja J, Kuhar M. Cocaine receptor: biochemical char 62.Dc KeyserJ,De BaeckerJ, Ebinger0, VauquelinG. [3H]GBR12935 acterization and structure-activityrelationships of cocaine analogues at the binding to dopamine uptake sites in the human brain. INeurochem 1989;53: dopamine transporter. I Med Chem 1992;35:969—981. 1400-1404.

Dopamine Transporter Imaging with lodine-123-IPT •Malison at al. 2297