High purity production and potential applications of copper-60 and copper-61

High Purity Production and Potential Applications ofCopper-60 and Copper-61

Deborah W. McCarthy,1 Laura A. Bass,1 P. Duffy Cutler,1 Ruth E. Shefer,2

Robert E. Klinkowstein,2 Pilar Herrero,1 Jason S. Lewis,1 Cathy S. Cutler,1

Carolyn J. Anderson1 and Michael J. Welch1

1MALLINCKRODT INSTITUTE OF RADIOLOGY, WASHINGTON UNIVERSITY SCHOOL OF MEDICINE, DIVISION OF RADIOLOGICAL

SCIENCES, ST. LOUIS, MISSOURI, USA; AND 2NEWTON SCIENTIFIC, INC., CAMBRIDGE, MASSACHUSETTS, USA

ABSTRACT. Previously we described the high yield production of 64Cu using a target system designedspecifically for low energy, biomedical cyclotrons. In this study, the use of this target system for theproduction of 60Cu and 61Cu is described and the utility of these isotopes in the labeling of biomolecules fortumor and hypoxia imaging is demonstrated. 60Cu and 61Cu were produced by the 60Ni(p,n)60Cu,61Ni(p,n)61Cu, and 60Ni(d,n)61Cu nuclear reactions. The nickel target (>99% enriched or natural nickel)was plated onto a gold disk as described previously (54–225 mm thickness) and irradiated (14.7 MeV protonbeam and 8.1 MeV deuteron beam). The copper isotopes were separated from the nickel via ion exchangechromatography and the radioisotopic purity was assessed by gamma spectroscopy. Yields of up to 865 mCiof 60Cu have been achieved using enriched 60Ni. 61Cu has been produced with a maximum yield of 144 mCiusing enriched 61Ni and 72 mCi using enriched 60Ni. Specific activities (using enriched material) rangedfrom 80 to 300 mCi/mg Cu for 60Cu and from 20 to 81 mCi/mg Cu for 61Cu. Bombardments of natural Nitargets were performed using both protons and deuterons. Yields and radioisotopic impurities weredetermined and compared with that for enriched materials. 60Cu was used to radiolabel diacetyl-bis(N4-methylthiosemicarbazone), ATSM. 60Cu-ATSM was injected into rats that had an occluded left anteriordescending coronary artery. Uptake of 60Cu-ATSM in the hypoxic region of the heart was visualized clearlyusing autoradiography. In addition, 60Cu-ATSM was injected into dogs and excellent images of the heart andheart walls were obtained using positron emission tomography (PET). 61Cu was labeled to 1,4,8,11-tetraazacyclotetradecane-N,N’,N”,N”’-tetraacetic acid-octreotide (TETA-octreotide) and the PET images oftumor-bearing rats were obtained up to 2 h postinjection. After decay of the 61Cu, the same rat was injectedwith 64Cu-TETA-octreotide and the images were compared. The tumor images obtained using 61Cu werefound to be superior to those using 64Cu as predicted based on the larger abundance of positrons emitted by61Cu vs. 64Cu. NUCL MED BIOL 26;4:351–358, 1999. © 1999 Elsevier Science Inc. All rights reserved.

KEY WORDS. Cu-60, Cu-61, Cu-64, Production, PET

INTRODUCTION62Cu, 67Cu, and 64Cu are copper radionuclides already being usedfor several nuclear medicine applications. 62Cu (t1/2 5 9.76 min) isgenerator produced and has been used to label agents to quantifyboth blood flow and hypoxia (5–8, 19). A commercial version ofthe generator is now being investigated (9). The short half-life (9 h)of the 62Zn parent limits the potential usefulness of this nuclide.67Cu is a longer-lived isotope (t1/2 5 61.9 h), which is produced inhigh yield using the high energy proton beam at BrookhavenNational Laboratories (BLIP) Facility and Los Alamos NationalLaboratory. The need for these high energy physics facilities allowsonly limited availability of this radionuclide. 64Cu has an interme-diate half-life of 12.7 h and has applications both for imaging andtherapy. Previously we reported the high yield production of 64Cuusing a target system designed specifically for low energy, biomed-ical cyclotrons (13).

We are interested in developing positron emission tomography(PET) radiopharmaceuticals labeled with other positron-emittingCu radionuclides (see Table 1). 60Cu has a half-life of 23.7 min anddecays 93% of the time by positron emission. 61Cu has a longerhalf-life (3.32 h) and 60% of its emissions are by positron decay. Weare particularly interested in preparing agents for imaging hypoxictissue in tumor, brain, and heart, as well as developing techniquesfor labeling peptides and proteins. In this paper, we demonstrate theutility of these isotopes in the imaging of tumors and hypoxic tissuesat times ranging from tens of minutes to a few hours. If a singletarget and production system can be used to produce these nuclidesas well as 64Cu, copper nuclides would be available for a widevariety of biological studies.

PET studies using compounds labeled with Cu-60 may bepossible, depending on the type of quantitative information neces-sary. Cu-60 emits prompt gamma rays in cascade with each positronin the 1–2-MeV range that are virtually unaffected by the typicallead and lead/tungsten shielding in the PET gantry. These gammasor their secondary radiations contribute substantial noise to the databy means of random coincidences. The thin tungsten septa incommercially available PET scanners are normally effective atcontrolling the randoms fraction by limiting the solid angle seen byeach detector. With this shielding, virtually invisible to the high-

Address correspondence to: Michael J. Welch, Ph.D., Mallinckrodt Instituteof Radiology, Washington University School of Medicine, 510 S. Kings-highway Blvd., St. Louis, MO 63110, USA; e-mail: [email protected].

Received 15 August 1998.Accepted 20 November 1998.

Nuclear Medicine & Biology, Vol. 26, pp. 351–358, 1999 ISSN 0969-8051/99/$–see front matterCopyright © 1999 Elsevier Science Inc. All rights reserved. PII S0969-8051(98)00113-9

energy gammas, any activity inside and within a few centimeters ofthe axial field-of-view will flood all detectors in the tomograph withsingle photon events. This action has the dual effect of adding noisevia increased random coincidences, and reducing good signal viaincreased deadtime. Among the casualties of the higher-energygammas is the deadtime correction. The calculation of an accuratecorrection factor for the data depends on an expected ratio of singleevents to coincident events. Phantom studies have been carried outto determine the magnitude of the effect of the prompt gammaemissions.

MATERIALS AND METHODSNickel Targets Preparations

High purity reagents used in the electroplating and separationstudies were the same as those described previously for the produc-tion of 64Cu (13). Isotopically enriched 60Ni (99.59%) and 61Ni(99.44%) were purchased from Trace Sciences International (Rich-mond Hills, Ontario, Canada). Natural nickel (99.99% purity) waspurchased from Aldrich Chemical Company (Milwaukee, WI). Theelectroplating procedure was an adaptation of the methodologyreported by Piel and co-workers (15). Briefly, appropriate quantitiesof nickel metal were dissolved in 6 N HNO3 and evaporated todryness. The residue was treated with H2SO4, diluted with deion-ized water, and evaporated to almost dryness. The residue wascooled and dissolved with deionized water. The pH was adjusted to9 with concentrated ammonium hydroxide and ammonium sulfateelectrolyte was added. The solution was quantitatively transferred toan electrolytic cell in which the anode was a graphite rod and thecathode was a gold disk (0.06250 thick 3 0.750 diameter). The cellswere typically operated at 2.4–2.6 V and at currents between 10 and25 mA. Electroplating was accomplished in 12–24 h. The diameterof the Ni targets were 0.5 cm unless otherwise indicated.

Isotope Production and Analysis

Targets were irradiated in the previously described holder (13) usingapproximately 14.7 MeV protons and 8.1 MeV deuterons on aCyclotron Corporation CS-15 cyclotron at Washington University.Isotope yield measurements were determined using either a cali-brated Ge detector (Canberra model 1510, Meriden, CT) or aradioisotope dose calibrator (Capintec CRC-10, Pittsburgh, PA) incombination with the Ge detector. The Ge detector was used todetect radionuclidic impurities. Specific activities were determinedas described previously for 64Cu by titration of 64Cu-acetate withthe macrocycle TETA (13). Briefly, aliquots of TETA were added

to 64,61,60Cu and the percent complexation monitored by radio-thinlayer chromatography (radio-TLC) (Bioscan, Washington, DC).The samples were spotted on silica plates and the plates developedusing 1:1 MeOH:10% NH4OAc. Cu acetate remained at the originwhereas complexed copper in the form of Cu(TETA)2- migratedwith Rf 5 0.42. The minimum TETA concentration for which100% labeling occurred was assumed to be equal to the concentra-tion of Cu(II) present.

Radiochemical Synthesis of 61/64Cu-TETA-Octreotide

Octreotide was purchased from DePaul Pharmaceutical (Bridgeton,MO). TETA z 4HCl z 4H2O was purchased from Aldrich.TETA-octreotide was prepared and radiolabeled with copper asdescribed previously (1). For the labeling, 7–15 mCi of 64Cu or61Cu were labeled to 3 mg of TETA-octreotide in 0.1 M ammoniumacetate buffer, pH 5.5 by a 45-min room temperature incubation.Samples were purified by a C-18 SepPakt to remove uncomplexed64/61Cu-acetate. Radiochemical purity was assessed by radio-TLCusing reversed-phase C-18 plates developed in 70:30 methanol:5%NH4OAc. Cu-TETA-octreotide migrates with Rf 5 0.5. Radio-chemical purity of the labeled complexes was typically above 97%.Approximately 300 mCi (0.2 mg) of the Cu-TETA-octreotideradioconjugates were injected into the rats.

FIG. 1. Structure of Cu-diacetyl-bis(N4-methylthiosemicar-bazone) (Cu-ATSM).

TABLE 1. Positron-Emitting Cu Radionuclides

Nuclide Half-life Decay mode Major gamma (keV) Emissions (%) b1max

60Cu 23.7 min 93% b1 1333 80 3.92 MeV, 6%7% EC 1760 52 3.00 MeV, 18%

850 15 2.00 MeV, 69%61Cu 3.32 h 60% b1 284 12 1.22 MeV

40% EC 656 10.762Cu 0.16 h 98% b1 875 0.15 2.91 MeV

2% EC64Cu 12.7 h 19% b1, 1345 0.47 0.656 MeV

43% EC,38% b2

Data from: Table of Isotopes (9a).

352 D. W. McCarthy et al.

TABLE 2. Yields of Copper Isotopes Using 991% Enriched Materials

Isotope Nuclear ReactionTarget thickness

(microns)Maximum EOB yield

(mCi)Typical EOB yield

(mCi)Yield

(mCi/mA z h)Major isotopic

impurities60Cu 60Ni(p,n)60Cu 225 865 450 58 0.05% 61Cu

0.025% 57Co61Cu 61Ni(p,n)61Cu 118 144 100 7.6 0.04% 58Co61Cu 60Ni(d,n)61Cu 122 72 60 2.44 0.04% 58Co64Cu 64Ni(p,n)64Cu 215 921 200 8.0 0.012% 55Co

EOB 5 end of bombardment.

TABLE 3. Natural Ni Bombardment: 14.7 MeV Protons

Target thickness(microns)

EOBsat yield Cu-60(mCi/mA)

% Radionuclidic impurities(relative to Cu-60) EOB

Radionuclidic impurities per 10mCi Cu-60

EOB 45 min post-EOB

116a 10.42 (EOB yield 5 0.875mCi/mA)

Co-55 5 0.65 Co-55 5 0.065 mCi Co-55 5 0.235 mCi

Cu-61 5 0.60 Cu-61 5 0.060 mCi Cu-61 5 0.192 mCi116b 7.86 (EOB yield 5 6.5

mCi/mA)Co-55 5 1.0 Co-55 5 100 mCi Co-55 5 361.8 mCi

Co-57 5 5.3 E-3 Co-57 5 0.53 mCi Co-57 5 1.975 mCiCo-58 5 5.2 E-4 Co-58 5 0.052 mCi Co-58 5 0.1937 mCiNi-57 5 6.7 E-3 Ni-57 5 0.67 mCi Ni-57 5 2.462 mCiCu-64 5 0.38 Cu-64 5 38 mCi Cu-64 5 135.92 mCiCu-61 5 1.12 Cu-61 5 112 mCi Cu-61 5 356.9 mCi

a Sample bombarded at 4 mA for 3 min, 0.6 cm diameter target.b Sample bombarded at 20 mA (avg current) for 1 h.

TABLE 4. Radionuclidic Impurities per 10 mCi 60Cu Produced Using 991% Enriched 60Ni

EOBsat yield Cu-60(mCi/mA)

Radionuclidic impurities per 10 mCi Cu-60

EOB 45 min post-EOB 100 min post-EOB

45 Co-55 5 3.6 E-5 mCi Co-55 5 0.13 mCi Co-55 5 0.627 mCiCo-57 5 0.036 mCi Co-57 5 2.24 mCi Co-57 5 11.16 mCiCu-61 5 5 mCi Cu-61 5 15.93 mCi Cu-61 5 65.6 mCi

EOB 5 end of bombardment. Conditions included using a 125-mm thick target, 14.7 MeV proton beam, ;15 mA avg. current, 44-min bombardment.

TABLE 5. 60Cu production (60Ni(p,n)60Cu) Using 99.59 Enriched 60Ni

Target thickness(microns)

Production parameters(EOB activity, irradiation time,

average beam current)

EOBsat yield

(mCi)Before column

separation

(mCi/mA)Before column

separation

(mCi)After

separation

(mCi/mA)After

separation

68.1 152 mCi, 44 min, 12.68 mA 210 16.56 201 15.754.36 178 mCi, 72 min, 8.5 mA 203 23.9 256 30.1129.33 613 mCi, 49 min, 15.3 mA 805 52.6 458 30.0115.02 309 mCi, 20 min, 14.4 mA 697 48.4 688 47.8125.32 494 mCi, 44 min, 14.86 mA 683 45.93 456 30.67

EOB 5 end of bombardment prior to column separation.Production parameters 5 14.7 MeV protons.

Production and Application of Cu-60 and Cu-61 353

Preparation of 60Cu-ATSM

H2ATSM was synthesized as described by Petering and co-workers(14). All chemicals were purchased from Aldrich, unless statedotherwise. Briefly, 4-methyl-3-thiosemicarbazide (97%) was dis-solved in 5% acetic acid (99.99%) at 50–60°C with stirring.2,3-Butanedione (97%) was taken up in water (distilled, deionized,.18 MV resistivity) and added to the 4-methyl-3-thiosemicarba-zide solution dropwise. The hot solution was filtered through acourse fritted-glass filter and a precipitate was collected. This solidwas washed with H2O and then ethanol and dried at 75°Covernight. The solid was purified by suspension in hot 80% aceticacid under reflux. The mixture was again filtered hot and theundissolved material collected and dried overnight at 75°C. 60Cuwas used to radiolabel ATSM using methods identical to Fujiba-yashi and co-workers (4).

Phantom Studies

To evaluate the quantitative accuracy of imaging Cu-60 with PET,a phantom study was performed where the measured activity wascompared with the true activity. A cylinder was filled with a dilutesolution of 60Cu and scanned over several half-lives using a SiemensECAT Exact PET scanner (12). For comparison, a similar experi-ment was carried out utilizing fluorine-18.

Imaging Studies

All animal studies were performed in compliance with guidelines setforth by the Washington University Animal Studies Committee.Autoradiographic images of tissue slices were obtained utilizing anInstantImagert electronic autoradiography system (Packard Instru-ment Company (Meriden, CT). Images of rats were obtainedutilizing the Siemens 953 B PET scanner at Washington University.PET images of the canine were obtained using PET ElectronicsSP3000E scanner at Washington University. Radiopharmaceuticalsused for all imaging studies were prepared using isotopes producedfrom enriched target materials.

Rat Heart Model Studies60Cu-ATSM (structure shown in Fig. 1) was used to visualizehypoxia in an acute left anterior descending (LAD) coronary arteryoccluded heart model by ex vivo tissue slicing as described previously(4). Briefly, a male Wistar rat (360 g) was anesthetized, intubated,and ventilated with room air. Blood flow was occluded to the LADcoronary artery. Thirteen minutes postocclusion, 270 mCi 60Cu-ATSM were injected through the femoral vein. The animal wassacrificed 10 min postinjection, the heart removed, frozen in

Tissue-Tek embedding medium (Miles Inc., Elkhart, IN), and 1-mmslices cut. The sections were placed on a grid and loaded onto thePackard InstantImager to obtain the 60Cu-ATSM images.

Rat Tumor Imaging Studies

Lewis rats were implanted with somatostatin receptor-positive ratpancreatic tumors (CA20948) in the nape of the neck (10). Theoriginal tumor cells were purchased from the Tumor Bank atBiomeasure, Inc. (Hopkinton, MA). At the time of study, the tumorwas approximately 1,340 mm3 and weighed 1.3 g. For the PETimaging study, the tumor-bearing female rat (170 g) was anesthe-tized and 253 mCi of 61Cu-TETA-octreotide were injected viaintracardiac puncture. Scanning commenced immediately and PETimages were collected for 1 h post injection. After the 61Cu hadbeen allowed to decay (22 h later), 353 mCi of 64Cu-TETA-Octreotide were injected via intracardiac puncture. Scanning beganimmediately and images of the entire animal were collected for 1 h.

PET Dog Study

A 22.73-kg male mongrel dog was fasted 24 h prior to theprocedure. The dog was anesthetized, placed in the SP3000E PETscanner, and administered 19.1 mCi of 11CO. PET scanningcommenced immediately and images were collected for 5 min.Twenty minutes after the 11CO administration, 20 mCi of H2

15Owere injected intravenously and PET data were again collected for5 min. Finally, 10 min after the H2

15O injection, 1 mCi of60Cu-ATSM was injected intravenously and PET data collected for10 min.

TABLE 6. Natural Ni Bombardment: 8.1 MeV Deuterons

Target thickness(microns)

EOB yield Cu-61(mCi/mA z h)

% Radionuclidic impurities(relative to Cu-61) EOB

Radionuclidic impurities per10mCi Cu-61 EOB

117a 0.91 Co-56 5 0.07 Co-56 5 7 mCi130b 0.29 Co-56 5 0.08 Co-56 5 8.0 mCi

Co-58 5 0.05 Co-58 5 5.0 mCiNi-65 5 6.6 Ni-65 5 660 mCi

EOB 5 end of bombardment.a Sample bombarded at 4 mA for 3 min, 0.6-cm diameter target.b Sample bombarded at 10 mA for 2 h. Sample was analyzed 3 h after EOB: short half-lived isotopes not analyzed for.

TABLE 7. Production of 61Cu Using 99.59% Enriched60Niand 99.44% Enriched 61Ni(d,n)61Cu and 61Ni(p,n)61CuNuclear Reactions

Nuclearreaction

Target thickness(microns)

EOB yield(mCi)

EOB yield(mCi/mA z h)

60Ni(d,n)61Cu 120a 56 1.1160Ni(d,n)61Cu 122a 72 2.4460Ni(d,n)61Cu 105 45 1.6261Ni(p,n)61Cu 117 112.4 11.2461Ni(p,n)61Cu 118 144 7.6261Ni(p,n)61Cu 57.8 60.5 3.0261Ni(p,n)61Cu 52.07 98.2 3.0761Ni(p,n)61Cu 116 168 9.6

EOB 5 end of bombardment. 8.1 MeV deuteron beam, 14.7 MeV protonbeam.a 0.6 cm diameter target.

354 D. W. McCarthy et al.

RESULTS AND DISCUSSION64Cu, 61Cu, and 60Cu were produced in high yield using the targetsystem described previously (13) . The maximim production yieldsof the copper radionuclides are shown in Table 2. 61Cu wasproduced by two nuclear reactions 61Ni(p,n)61Cu and 60Ni(d,n)61Cu. Although higher yields of 61Cu are produced using the(p,n) reaction, the target material for this reaction is approximately25 times the cost of the target material for the d,n reaction ($55/mgNi-61 vs $2/mg for Ni-60). None of the proton irradiation targetsused in this work are “thick targets.” Of the targets described inTable 2, the ;220-mm thick targets degrade the 14.7-MeV protonbeam by 5.2 MeV, whereas the ;120-mm thick target degrades thebeam by approx. 2.5 MeV. The cross-sections of the 60Ni/61Ni(p,n)reactions do not decrease dramatically until 6–7 MeV (15), sohigher yields could be obtained at greater target expense withthicker targets (up to 350 mm). The ;120-mm thick target used forthe deuteron irradiation is a “thick target.” The yields obtained aresomewhat variable presumably due to minor alterations in the beamalignment, which affects the 0.5-cm and 0.6-cm diameter targetused. Typical specific activities of these cyclotron-produced Cuisotopes as determined by TETA titrations are: 64Cu 5 40–250mCi/mg Cu, 61Cu 5 20–81 mCi/mg Cu, 60Cu 5 80–300 mCi/mgCu.

Several authors have reported the production of 60Cu on bio-medical cyclotrons using natural Ni. Martin (11) irradiated naturalNi (thick target) with 1 mA of 11-MeV protons for 23 min andreported the following end of bombardment (EOB) yields: 0.01 mCi55Co, 0.01 mCi 57Co, 5.0 mCi 60Cu, 0.07 mCi 61Cu, 3.8 mCi 62Cu,and 0.02 mCi 64Cu. Martin’s EOBsat (end of saturation bombard-ment) yield of Copper-60 was 10 mCi/mA (thick target yield).Similar yields can be obtained using a 14.7-MeV proton beam. OurEOBsat yields using a 14.7-MeV proton beam with natural nickel;120 mm thick) for the 60Ni(p,n)60Cu reaction were 7.86 mCi /mAand 10.42 mCi/mA (see Table 3). The yields obtained were inagreement with those predicted by Piel et al. (15).

Martin (11) described an electrochemical dissolution of thetarget material and separation of the Ni and Co from the copper bysolvent extraction followed by anion exchange separation. Martin’sseparation method removed 99% of the cobalt contaminants and

they were not considered in his dosimetry calculations with Cu-60-PTSM. Using 99.59% enriched material for the 60Ni(p,n)60Cureaction, we produced much less Co radioisotopic impurities thanproduction from natural Ni, even with the higher energy protonsused (see Table 4). This allowed for our separation procedure (anionexchange chromatography) to be less complicated. In addition, the60Cu yield using enriched material was four times higher than usingnatNi, which is important in certain circumstances in whichminimal preparation time is required. Table 5 shows the 60Cu yieldsusing enriched nickel for the 60Ni(p,n)60Cu nuclear reaction. Insome cases, the EOBsat yields of the separated material were lessthan that of the whole target due to loss of 60Cu during theseparation procedure.

Production of 61Ni using natural Ni has also been reported (17).Thick natural nickel foils were bombarded with 6.0 MeV deuterons.A yield of 10 MBq/mA z h (0.27 mCi/mA z h) was reported for 61Cuand the only radioisotopic impurity reported was 62Cu. We observedlow levels of radioisotopic impurities but with higher yields at '8.1MeV deuterons (see Table 6).

Zweit et al. (20) reported cross-section data for 60Ni(d,n)61Cuand 61Ni(d,2n)61Cu; however, values were given only for energiesof 7.2–18.9 MeV. Szelecsenyi et al. (16) irradiated enriched 61Ni-electroplated targets with protons and measured the cross-sectionsof the reactions. Energy range for the production of the 61Cu wasEp 5 12–9 MeV, corresponding to a target thickness of 0.119 g/cm2

(134 mm). The theoretical yield for producing 61Cu with protonsusing 100% 61Ni is 17.5 mCi /mA z h. Our 61Cu yields from theproton bombardment of 112–118-mm thick 61Ni targets were invery good agreement with Szelecsenyi (16) (Table 7).

We conducted a cost comparison (using our production param-eters) for the production of 61Cu using either 60Ni, 61Ni, or natNi asthe target material. For producing 20 mCi of 61Cu, the cost ofproduction using 60Ni or 61Ni was similar ($241 vs. $197 respec-tively). The advantage of using 60Ni (60Ni(d,n)61Cu) as the targetmaterial is its low cost. Process time is reduced if it is not necessaryto recycle the enriched material. Production of 61Cu by deuteronirradiation of natNi was considerably more expensive ($539) due tothe longer bombardment time required (Table 8). If larger quanti-ties of 61Cu are required, i.e., 50 mCi, proton bombardment of 61Ni

TABLE 8. Cost Comparison for Making 20 mCi 61Cu

Nuclearreaction

Yield(mCi/mA z h)

Targetmaterial

Cost of recyclinglossa

# Cyclotronhours

Cyclotron cost(@ $245/hr) Total cost

61Ni(p,n)61Cu 7.62 61Ni $165 0.13 $32 $19760Ni(d,n)61Cu 2.44 60Ni $40 0.82 $201 $24160Ni(d,n)61Cu 0.91 natNi Negligible 2.20 $539 $539

Conditions included 14.7 MeV, 20-mA proton beam, 8.1 MeV, 10 mA deuteron beam.a Assuming 85% recovery of the enriched 61Ni. For the 60Ni, because of the relatively low cost, the material is not recovered.

TABLE 9. Cost Comparison for Making 50 mCi 61Cu

Nuclearreaction

Yield(mCi/mA z h)

Targetmaterial

Cost of recyclinglossa

# Cyclotronhours

Cyclotron cost(@ $245/hr) Total cost

61Ni(p,n)61Cu 7.62 61Ni $165 0.33 $80 $24560Ni(d,n)61Cu 2.44 60Ni $40 2.05 $502 $54260Ni(d,n)61Cu 0.91 natNi Negligible 5.5 $1346 $1346

Conditions included 14.7 MeV, 20 mA proton beam, 8.1 MeV, 10 mA deuteron beam.a Assuming 85% recovery of the enriched 61Ni. For the 60Ni, because of the relatively low cost, the material is not recovered.

Production and Application of Cu-60 and Cu-61 355

(61Ni(p,n)61Cu) is preferable (cost 5 $245 versus $542 for60Ni(d,n)61Cu and $1,346 for deuteron bombardment of natNi)(Table 9). The radionuclide, 60Cu, decays by the emission of highenergy positrons and high energy gamma rays. For this reason, anexperiment was performed in which the measured activity wascompared with the true activity using a cylindrical phantom. It wasdetermined that the 60Cu activity could be quantified to a concen-tration of 0.8 mCi/mL. In human studies using 60Cu-ATSM, weanticipate an activity concentration averaging just 0.18 mCi/mL,and potentially ranging as high as 0.9 mCi/mL. Thus, exceptperhaps at the high end of this range, the effects shown in Figure 2will be small to negligible. With minor adjustments to the deadtimecalculation, calibrated for Cu-60 imaging, we will be able to obtainaccurate parameter estimates over a much wider activity range ifnecessary.

The data obtained for 60Cu were compared with that from 18F inFigure 2. With the standard deadtime correction provided with thePET system, a linear response was obtained for F-18, as expected,but not for Cu-60. The accuracy diverged at a concentration of 0.8mCi/mL. This corresponds to approximately 4 mCi total activity inthe field of view of the tomograph. In a PET study of myocardialperfusion, a relatively small bolus of Cu-60 activity, perhaps 10 mCitotal, would be sufficient to exceed the linear range of thetomograph. In other studies, in which the activity can be infusedover time, or in which the measurement of an input function is notneeded, then a much larger dose could be used while still obtainingaccurate results. This approximates to a uniform distribution ofapproximately 50 mCi in a human. It is highly unlikely that anyradiopharmaceutical would require such a high concentration, solinear data can be obtained utilizing appropriate amounts of thisnuclide. As anticipated, the 60Cu data are affected by the highenergy gammas.

60Cu has been evaluated in two imaging studies. Images wereobtained ex vivo following administration of 60Cu-ATSM into a ratin which the LAD had been tied off (Fig. 3). The images obtainedwith 60Cu were similar to images obtained previously with 64Cu (4).In addition, 60Cu-ATSM, a hypoxic tissue marker, was injected intoa dog under normoxic conditions. 60Cu-ATSM is selectively

trapped in hypoxic tissues but is rapidly washed out of normoxiccells (4). A high quality image of a dog heart was obtained showingsome retention of 60Cu-ATSM in the walls of the dog heart (Fig. 4).The image showed a reconstructed midventricular PET image of theheart from a normal dog obtained 30–640 s after administration of60Cu-ATSM. The image to the left was not corrected for bloodpool, whereas the image on the right was corrected for tracerremaining in the blood pool. Even though Cu-ATSM washes outfrom normoxic tissue (;20% retention), high quality images areobtained in vivo. The image was obtained using 1 mCi of tracer. Ourpreliminary dose estimates allow ;13 mCi of 60Cu-ATSM to beadministered to a human. This method demonstrates the viability ofusing 60Cu-labeled radiopharmaceuticals for PET imaging. Studieson the uptake of Cu-ATSM in ischemic heart are in progress (18).

A potential application of 61Cu is for labeling proteins andpeptides for which maximum uptake occurs from 2 to 6 h postadministration. An agent that falls into this category is 61Cu-TETA-octreotide. Comparison images with 64Cu-TETA-octreotideare shown in Figure 5. These images were obtained using 0.2 mg of61/64Cu-TETA-octreotide. The small amount of injected dose waschosen because it was significantly lower than the amount previ-ously used to block the uptake of 64Cu-TETA-octreotide (250 mg)(2). This small amount is unlikely to cause any decrease in theuptake of Cu-64-TETA-octreotide. Higher quality images wereobtained with 61Cu than with 64Cu, due to the higher positron yield(60% vs. 19%) and the shorter half-life. Because high qualityhuman images (3) can be obtained 4 h postadministration of64Cu-TETA-octreotide, higher quality images could be obtained (atthe same radiation dose) using 61Cu-TETA-octreotide.

CONCLUSION

The cyclotron target described previously for the production of64Cu can be used to produce 60Cu and 61Cu in high yield and athigh specific activitiy. For certain biological applications, theshorter half-life of 60Cu and 61Cu may be preferable to the 12.7-h

FIG. 2. Activity concentration values as measured by a cylindrical phantom. F-18 activity (left) is accurate over the 0–1mCi/mL measured here, and is within 5% out to 5 mCi/mL (not shown). A decaying phantom filled with Cu-60 activity in ionicsolution is shown on the right. The deadtime correction restores data loss up to about 0.8 mCi/mL. Above this level, minormodifications to the deadtime model will be required to account for the effects of Cu-60’s high-energy gammas.

356 D. W. McCarthy et al.

half-life of 64Cu. Despite the high b1 energy and prompt gammaemission of 60Cu, high quality images are obtained both autoradio-graphically (using the Packard InstantImager) and in PET scans of60Cu-ATSM. For studies in which maximum contrast is obtained inless than 1 h, 60Cu will be preferred. If optimal uptake occurs in1–4 h, 61Cu will bepreferred.

This work was supported by research grants from the Department ofEnergy DE-FG02-87ER60512 (M.J.W.) and NIH CA64475(C.J.A.). We thank Bill Margenau, Dave Ficke and Todd Perkins forassistance in isotope production, and Lynne A. Jones, Terry Sharpe andLennis Lich for their contribution to the animal imaging studies.

FIG. 3. Ex vivo imaging of a left anterior descending (LAD) coronary artery occluded heart model. At the top are photographicimages of the heart slices. At the bottom are images obtained on the Packard InstantImager of the distribution ofCu-60-diacetyl-bis(N4-methylthiosemicarbazone) (Cu-60-ATSM) in hypoxic tissue.

FIG. 4. Reconstructed midventricular positron emission tomography (PET) images of the heart from a normal canineobtained 30–640 s after administration of Cu-60-diacetyl-bis(N4-methylthiosemicarbazone) (Cu-60-ATSM). The raw image isshown to the left, and the image to the right represents the raw image after being corrected for tracer remaining in the bloodpool.

Production and Application of Cu-60 and Cu-61 357

References1. Anderson C. J., Jones L. J., Bass L. A., Sherman E. L., McCarthy D. W.,

Cutler P. D., Lanahan M. V., Cristel M. E., Lewis J. S. and SchwarzS. W. (1998) Radiotherapy, toxicity and dosimetry of copper-64-labeled-TETA-octreotide in tumor-bearing rats. J. Nucl. Med. in press.

2. Anderson C. J., Pajeau T. S., Edwards W. B., Sherman E. L. C., RogersB. E. and Welch M. J. (1995) In vitro and in vivo evaluation ofcopper-64-octreotide conjugates. J. Nucl. Med. 36, 2315–2325.

3. Dehdashti F., Anderson C. J., Trask D. D., Bass L. A., Schwarz S. W.,Cutler P. D., McCarthy D. W. and Lanahan M. V. (1997) Initial resultswith PET imaging using Cu-64-labeled TETA-octreotide in patientswith carcinoid tumor. J. Nucl. Med 38, 103.

4. Fujibayashi Y., Cutler C. S., Anderson C. J., McCarthy D. W., Jones L. A.,Sharp T., Yonekura Y. and Welch M. J. (1998) Comparative studies ofCu-64-ATSM and C-11 Acetate in an acute myocardial infarction model:Ex vivo imaging of hypoxia in rats. Nucl. Med. Biol. in press.

5. Fujibayashi Y., Taniuchi H., Yonekura Y., Ohtani H., Konishi J. andYokoyama A. (1997) Copper-62-ATSM: A new hypoxia imaging agentwith high membrane permeability and low redox potential. J. Nucl.Med. 38, 1155–1160.

6. Fujibayashi Y., Wada K., Matsumoto K., Yonekura Y., Konishi J. andYokoyama A. (1991) Retention mechanism of Cu-62-PTSM, a gener-ator-produced positron emitting Cu-62 labeled blood flow tracer.J. Nucl. Med. 32, 974.

7. Green M. A. and Goldenberg D. M. (1987) A potential copper radiophar-maceutical for imaging the heart and brain: Copper-labeled pyruvaldehydebis(N-methylthiosemicarbazone). Nucl. Med. Biol. 14, 59–61.

8. Green M. A., Welch M. J., Mathias C. J., Bergmann S. R., PerlmutterJ. S., Raichle M. E., Rubio F. F. and Janik M. (1990) Copper-62-PTSM:Production and evaluation of a generator produced multi-purpose PETperfusion agent. J. Nucl. Med. 31, 815.

9. Lacy J. L., Haynes N. G., Nayak N., Martin C. S., Dai D. and GreenM. A. (1998) Performance of a Zn-62/Cu-62 generator in clinical trialsof the PET perfusion agent Cu-62-PTSM. J. Nucl. Med. 39, 142.

9a.Lederer C. M. and Shirley V. S. (1978) Table of Isotopes, 7th edn.MacMillan, New York.

10. Longnecker D. S., Lilja H. S., French J., Kuhlmann E. and Noll W.

(1979) Transplantation of azaserine-induced carcinomas of pancreas inrats. Cancer Lett. 7, 197–202.

11. Martin C. C. (1992) The kinetics of copper pyruvaldehyde bis (N4-meth-ylthiosemicarbazone): A blood flow tracer for positron emission tomogra-phy. PhD Dissertation, University of Wisconsin-Madison, Madison.

12. Martin C. C., Christian B. T., Satter M. R., Nickerson L. D. H. andNickles R. J. (1995) Quantitative PET with positron emitters that emitprompt gamma rays. IEEE Trans. Med. Imaging 14, 681–687.

13. McCarthy D. W., Shefer R. E., Klinkowstein R. E., Bass L. A.,Margenau W. H., Cutler C. S., Anderson C. J. and Welch M. J. (1997)Efficient production of high specific activity 64Cu using a biomedicalcyclotron. Nucl. Med. Biol. 24, 35–43.

14. Petering H. G., Buskirk H. H. and Underwood G. E. (1964) Theanti-tumor activity of 2-keto-3-ethoxybutyraldehyde bis(thiosemicarba-zone) and related compounds. Cancer Res. 24, 367–372.

15. Piel H., Qaim S. M. and Stocklin G. (1992) Excitation functions of(p,xn)-reactions on natNi and highly enriched 62Ni: Possibility ofproduction of medically important radioisotope 62Cu at a small cyclo-tron. Radiochim. Acta 57, 1–5.

16. Szelecsenyi F., Blessing G. and Qaim S. M. (1993) Excitation functionsof proton induced nuclear reactions on enriched 61Ni and 64Ni:Possibility of production of no-carrier-added 61Cu and 64Cu at a smallcyclotron. Appl. Radiat. Isot. 44, 575–580.

17. Tolmachev V., Lundqvist H. and Einarsson L. (1998) Production of61Cu from a natural nickel target. Appl. Radiat. Isot. 49, 79–81.

18. Welch M. J., Lewis J. S., McCarthy D. W., Sharp T., Herrero P. andFujibayashi Y. (1998) Evaluation of copper-60-dicaetyl-bis(N9-methyl-thiosemicarbazone)(Cu-ATSM), a hypoxia imaging agent, in caninemodels of ischemia. Eur. J. Nucl. Med. 25, 978.

19. Zweit J., Goodall R., Cox M., Babich J. W., Potter G. A., Sharma H. L.and Ott R. J. (1992) Development of high performance zinc-62/copper-62 radionuclide generator for positron emission tomography.Eur. J. Nucl. Med. 19, 418–425.

20. Zweit J., Smith A. M., Downey S. and Sharma H. L. (1991) Excitationfunctions for deuteron induced reactions in natural nickel: Productionof no-carrier-added 64Cu from enriched 64Ni targets for positronemission tomography. Appl. Radiat. Isot. 42, 193–197.

FIG. 5. Images of tumor-bearing Lewis rats injected with Cu-1,4,8,11-tetraazacyclotetradecane-N,N*,N(,N***-tetraacetic acid(TETA)-octreotide (octreotide is a somatostatin analog, used in imaging tumors that contain somatostatin receptors). Thebottom images were obtained after 353 mCi of 61Cu -TETA-octreotide were injected (;3.6 E, 6 total counts per slice).Twenty-two hours later, 253 mCi of 64Cu-TETA-octreotide was injected into the same rat (;7.8 E, 5 total counts per slice) andpositron emission tomography (PET) images were obtained again (top images).

358 D. W. McCarthy et al.