Preparation and Biological Evaluation of Copper-64–Labeled Tyr³-Octreotate Using a Cross-Bridged Macrocyclic Chelator

Extensive studies have been conducted on radiolabeled somatostatin analogues for both imaging and radiotherapy of somatostatin receptor-positive tumors. Modifications to octreotide, a somatostatin analogue, have been evaluated to improve target tissue uptake in somatostatin receptor subtype 2 (SSTr2)-positive tissues; based on these, the octapeptide Tyr³-octreotate (Y3-TATE) was identified (1, 2). Radiolabeled somatostatin analogues incorporating positron-emitting radionuclides such as ⁶⁴Cu and ⁶⁸Ga have been shown to be equivalent or improved SSTr2 tracers for positron emission tomography (PET) imaging of SSTr2-positive tumors compared with ¹¹¹In-labeled somatostatin analogues for γ scintigraphy (3, 4, 5). Copper-64, ¹⁷⁷Lu, and ⁹⁰Y-radiolabeled somatostatin analogues have also shown promise as radiotherapeutic agents for the treatment of SSTr-positive tumors (6, 7, 8, 9, 10, 11, 12).

Copper-64 (t_1/2 = 12.7 hours) is both a positron (β⁺; 17.4%, E_β+max = 656 keV) and β minus (β⁻; 39%, E_β−max = 573 keV) emitter and can be produced in high specific activity on a small biomedical cyclotron (13, 14). The decay characteristics of ⁶⁴Cu make it suitable for PET imaging as well as radiotherapy when targeted to tumors using monoclonal antibodies or peptides (6, 7, 15, 16, 17, 18, 19). Biodistribution and metabolism studies of ⁶⁷Cu and ⁶⁴Cu-bifunctional chelator-monoclonal antibodies or peptide conjugates have demonstrated significant dissociation from the bifunctional chelator resulting in transchelation to liver superoxide dismutase (20, 21). For the somatostatin analogue ⁶⁴Cu-TETA-octreotide (where TETA is 1,4,8,11-tetraazacyclotetradecane-1,4,8,11-tetraacetic acid), this resulted in retention of ⁶⁴Cu activity in blood, poor liver clearance, and increasing bone marrow uptake over time (7). High blood uptake and retention in the liver reduce imaging sensitivity by increasing background radioactivity levels. High accumulation of ⁶⁴Cu in the bone marrow may limit the therapeutic dose of ^64/67Cu-bifunctional chelator-peptide due to myelotoxicity (22).

Weisman, Wong, and coworkers (23, 24) have synthesized and characterized a series of copper (II) cross-bridged cyclam complexes that demonstrate improved kinetic stability compared with TETA complexes. Sun et al. (25) evaluated the biodistribution of four ⁶⁴Cu-labeled cross-bridged cyclam complexes in rats and found ⁶⁴Cu-CB-TE2A (CB-TC2A-4,11-bis(carboxymethyl)-1,4,8,11-tetraazabicyclo[6.6.2]hexadecane) to have the most improved blood, liver, and kidney clearance compared with ⁶⁴Cu-TETA (26). Transchelation to liver superoxide dismutase was reduced for ⁶⁴Cu-CB-TE2A compared with ⁶⁴Cu-TETA in rat metabolism studies (27). These data suggest that CB-TE2A–peptide conjugates may possess improved biodistribution characteristics, making CB-TE2A an attractive bifunctional chelator for attaching ⁶⁴Cu to peptides for tumor imaging and radiotherapy.

Because Y3-TATE is a well-characterized SSTr2 ligand, this peptide is a good model for evaluating novel bifunctional chelator-peptide conjugates. The AR42J rat pancreatic carcinoma cell line expresses SSTr2 both in vitro and in vivo (28, 29, 30, 31). In these studies we report in vitro and in vivo characterization of ⁶⁴Cu-CB-TE2A-Y3-TATE and ⁶⁴Cu-TETA-Y3-TATE (Fig. 1) in the AR42J tumor model.

Copper-64 was produced on a CS-15 biomedical cyclotron at Washington University School of Medicine according to published procedures (13). Copper chloride (CuCl₂) was purchased from Johnson Matthey (West Deptford, NJ). Trifluoroacetic acid (TFA) was purchased from J.T. Baker (Chicago, IL). 9-Fluorenylmethoxycarbonyl (Fmoc) amino acids were purchased from Novabiochem (San Diego, CA). All of the other chemicals were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). All of the solutions were prepared using ultrapure water (18 MΩ-cm resistivity). TLC was performed using Whatman MKC₁₈F reversed-phase plates with 10% ammonium acetate to methanol (30:70) as the mobile phase. Radio-TLC detection was accomplished using a BIOSCAN AR2000 Imaging Scanner (Washington, DC). Analytical reversed-phase high-performance liquid chromatography (HPLC) was performed on a Waters 600E (Milford, MA) chromatography system with a Waters 991 photodiode array detector and an Ortec Model 661 (EG&G Instruments, Oak Ridge, TN) radioactivity detector. HPLC samples were analyzed on a Vydac diphenyl column (4.6 × 100 mm; Hesperia, CA). The mobile phase was H₂O (0.1% TFA; solvent A) and 90% acetonitrile (ACN; 0.1% TFA; solvent B). The gradient consisted of 5% B to 70% B in 20 minutes (1.0 mL/min flow rate). Radioactive samples were counted using a Beckman 8000 automated well-type gamma counter (Fullerton, CA). Electrospray mass spectrometry was accomplished using a Waters Micromass ZQ (Milford, MA).

Male Lewis rats (21-days old, 40–50 g) were purchased from Charles River Laboratories (Boston, MA). The AR42J tumor line was obtained from American Type Culture Collection (ATCC, Rockville, MD).

1,4,8,11-Tetraazabicyclo[6.6.2]hexadecane was prepared as reported previously (24).

Anhydrous sodium carbonate (1.18 g, 11.1 mmol) and t-butyl bromoacetate (2.17 g, 11.13 mmol) were added to a solution of 1,4,8,11-tetraazabicyclo[6.6.2]hexadecane (1.12 g, 4.95 mmol) in acetonitrile (30 mL). The mixture was stirred under N₂ and heated to 50°C for 48 hours. Solvent was then removed under reduced pressure, and residual material was dissolved in 20% aqueous NaOH (20 mL) at ice bath temperature (0–5°C). The aqueous solution was extracted with cold (0–5°C) toluene (3 × 20 mL), combined extracts were dried (anhydrous Na₂SO₄), and solvent was removed under reduced pressure to give an oil, which solidified on standing (2.08 g, 4.58 mmol, 93%): mp 39–40°C; ¹H NMR (400 MHz, C₆D₆) δ 1.21–1.46 (m, 4H), 1.40 (s, 18H, C(CH₃)₃), 2.22 (ddd, 2H, J = 13.5, 3.9, 1.9 Hz), 2.29 (ddd, 2H, J = 13.1, 4.9, 2.5 Hz), 2.42–2.56 (XX’ of AA’XX’, 2H, NCH_AH_XCH_A’H_X’N cross-bridge), 2.60–2.94 (m, 10H), 3.01 and 3.17 (AB, 4H, J = 16.4 Hz, N-CH_AH_B-CO₂), 3.24–3.40 (AA’ of AA’XX’, 2H, NCH_AH_XCH_A’H_X’N cross-bridge), 3.69 (td, 2H, J = 11.9, 4.2 Hz); ¹³C NMR (100.5 MHz, C₆D₆) δ 28.10 (C(CH₃)₃), 28.20 (CH₂CH₂CH₂), 50.87, 53.32, 55.96, 56.31, 57.24, 60.13, 79.67 (C(CH₃)₃), 171.08 (C =O); IR (KBr) 2976, 2917, 2819, 2784, 1739, and 1728 (C=O), 1477, 1367, 1392, 1323, 1165, 1122 cm⁻¹; HRFABMS, m/z (M+H)⁺ exact mass calculated for C₂₄H₄₇N₄O₄: 455.3597; Found: 455.3600 (error 2.5 ppm).

4,11-Bis-(carbo-tert-butoxymethyl)-1,4,8,11-tetraazabicyclo[6.6.2]hexadecane (1.182 g, 2.600 mmol) was dissolved in a 1:1 (vol:vol) mixture of CF₃CO₂H (TFA) and CH₂Cl₂ (40 mL). The mixture was stirred under N₂ at room temperature for 48 hours. Solvent was removed under reduced pressure to give an oil, which crystallized after removal of residual solvent under vacuum to yield 1.510 g (2.566 mmol, 99%) of product as the di-TFA salt monohydrate: mp 165–169°C; ¹H NMR (400 MHz, CD₃CN, CD₂HCN central peak set to 1.94 ppm) δ 1.64–1.74 (dm, 2H, J = 16.8 Hz, CH₂CH_eqH_axCH₂), 2.18–2.34 (∼qm, 2H, J = ∼13 Hz, CH₂CH_eqH_axCH₂), 2.57–2.67 (dm, 2H, J = 13.6 Hz), 2.89–3.14 (m, 12H), 3.31 (td, 2H, J = 13.0, 3.3 Hz), 3.43–3.48 (AA’BB’, 4H, NCH₂CH₂N cross-bridge), 3.46 (d, 2H, J = 17.0 Hz, NCH_AH_XCO₂), 4.14 (d, 2H, J = 17.0 Hz, NCH_AH_XCO₂), (Note: N-H/CO₂H fast-exchange signal broadened into baseline. In other spectra, broad signals integrating for up to 4H were observed in the range 8.9–12 ppm); ¹³C NMR (100.5 MHz, CD₃CN, CD₃CN central peak set to 1.39 ppm) δ 20.63, 48.48, 48.84, 54.35, 56.28, 59.09, 60.81, 172.69 (C=O); IR (KBr) 3408, 2960, 2886, 2529, 1723 (C=O), 1433, 1198, 1132, 722 cm^-1; Anal Calcd for C₁₆H₃₀N₄O₄·2TFA· H₂O: C, 40.82; H, 5.82; N, 9.52. Found: C, 41.14; H, 5.91; N, 9.62.

Y3-TATE was prepared on solid support by standard Fmoc procedure (32, 33) using an Advanced Chem Tech peptide synthesizer. To facilitate on-board cyclization, acetamidomethyl thiol-protected cysteine was used to introduce the cysteine amino acids (34). After peptide assembly, intramolecular disulfide cyclization was performed on solid support with thallium trifluoroacetate (23 mg, 42 μmol in 2 mL DMF) for 90 minutes, and the resin was thoroughly washed with DMF. The N-terminal Fmoc protecting group was removed from the peptide with 20% piperidine in DMF. The resin was sequentially washed with DCM (2 × 3 mL), methanol (2 × 3 mL), and DCM (2 × 3 mL), and was dried under house vacuum. The peptide-resin was then swollen in DCM (3 mL) for 45 minutes and washed with DMF (3 mL) without drying the resin. CB-TE2A·2TFA salt (39 mg, 68 μmol) was dissolved in DMF (500 μL). To this solution was added diisopropylcarbodiimide (DIC, 8.3 mg, 66 μmol) and DIEA (8.5 mg, 66 μmol), and the solution was stirred for 25 minutes before adding it to the wet peptide-resin. The mixture was gently agitated for 3 hours and filtered. The resin was washed with DCM (95 × 3 mL) and dried under vacuum before cleavage. TETA-Y3-TATE was synthesized as described previously (35).

Both CB-TE2A-TATE and TETA-Y3-TATE were cleaved from the resin with a mixture of TFA/water/thioanisol/phenol (85:5:5:5) for 4 hours and precipitated in cold t-butyl methyl ether. The crude peptide conjugates were purified by semipreparative reverse-phase HPLC using two solvent systems A and B with the gradient elution protocols: 30% to 70% B in 45 minutes at a flow rate of 10 mL/min, where A is 5% CH₃CN in 0.1% aqueous TFA and B is 10% H₂O in 0.1% TFA solution of CH₃CN. Peak detection was at 214 nm with a tunable absorbance detector. The purified compounds (∼10% isolated yield) were lyophilized in H₂O/MeCN (3:2) mixture and characterized by analytical HPLC, electrospray mass spectrometry and also by high-resolution mass spectrometry (matrix-assisted laser desorption/ionization) for CB-TE2A-Y3-TATE [m/z calculated for C₆₅H₉₂N₁₄O₁₅S₂ (M + H)⁺ = 1373.6386; found 1373.6532].

Radiolabeling of TETA-Y3-TATE with ⁶⁴Cu(II) was based on methods reported previously (35); here, 0.1 mol/L ammonium acetate (pH 6.5) was used as the reaction solvent. For CB-TE2A-Y3-TATE, 1–15 mCi (37–555 MBq) of ⁶⁴Cu in 0.1 mol/L ammonium acetate (pH 8.0) was added to 1 to 15 μg of the peptide in 0.1 mol/L ammonium acetate (pH 8.0). The reaction mixture was heated for 1 hour at 95°C. Radiochemical purity for both ⁶⁴Cu-labeled conjugates was determined by radio-TLC using Whatman MKC18F TLC plates developed with 10% NH₄OAc to methanol ratio (30:70) and was confirmed by radio-HPLC.

The affinity of ⁶⁴Cu-TETA-Y3-TATE or ⁶⁴Cu-CB-TE2A-Y3-TATE for SSTr2 was measured in AR42J tumor membranes using methods described previously for homologous competitive binding (7). To reduce nonspecific binding, 0.1% BSA was added to the receptor-binding buffer [50 mmol/L Tris-HCl (pH 7.4), 5.0 mmol/L MgCl₂ 0.5 μg/mL aprotinin, 200 μg/mL bacitracin, 10 μg/mL leupeptin, and 10 μg/mL pepstatin A]. The competing ligands, ^natCu-TETA-Y3-TATE or ^natCu-CB-TE2A-Y3-TATE, were prepared with high-purity natural copper chloride (CuCl₂) using the procedure described above for preparation of the ⁶⁴Cu-labeled peptides. Complex formation was confirmed by HPLC and electrospray mass spectrometry. IC₅₀ values were determined according to published methods (7) using the Millipore MultiScreen Assay system (Bedford, MA). Data analysis was performed using GraphPad PRISM (San Diego, CA) for nonlinear regression of homologous competitive binding curves with ligand depletion.

The procedure for the cellular internalization assay was based on the method published by Wang et al. (36). Forty-eight hours before each assay, AR42J cells (1.9 × 10⁶ cells per well) were seeded in 6-well BD Primaria plates (BD Biosciences, Bedford, MA) containing F12K media supplemented with 2 mmol/L l-glutamine, 1.5 g/L sodium bicarbonate, and 20% fetal bovine serum; cells were incubated at 37°C in a humidified 5% CO₂ atmosphere. Cells were washed twice with HBSS, and then 1 mL of F12K supplemented media was added to each well. To block specific binding, at each time point 3 of 6 wells were incubated 5 minutes at room temperature with 2 μg Y3-TATE. To each well was added ⁶⁴Cu-CB-TE2A-Y3-TATE (specific activity = 0.23 mCi/μg) or ⁶⁴Cu-TETA-Y3-TATE (specific activity = 0.23 mCi/μg) to a final concentration of 4.0 nmol/L of each radiotracer. Cells were incubated at 37°C. At each time point, surface-bound fractions were collected as described by Wang et al. (36). Internalized radioactivity was collected by lysing the cells in 0.5% SDS. Total protein concentration in the cell lysate was determined using the BCA Protein Assay (Pierce Biotechnology, Rockford, IL). Internalized and surface bound fractions were expressed as fmol radiotracer/mg protein.

All of the animal experiments were performed in compliance with the Guidelines for the Care and Use of Research Animals established by Washington University’s Animal Studies Committee. AR42J tumors (Mallinckrodt, Inc., St. Louis, MO, passage 36) were implanted in the left leg of male Lewis rats (21 days old) and maintained by serial passage in animals as described previously (37). Tumors were allowed to grow 10 to 14 days by which time tumors achieved a solid palpable mass.

The ⁶⁴Cu-labeled peptide conjugates (40 μCi, 140 ng ⁶⁴Cu-TETA-Y3-TATE; 50 μCi, 170 ng ⁶⁴Cu-CB-TE2A-Y3-TATE) were injected i.v. via the tail vein into AR42J tumor-bearing rats. Tissue biodistribution data were obtained at 1, 4, and 24 hours after injection according to methods described previously (35).

To examine in vivo uptake specificity, additional AR42J tumor-bearing Lewis rats were injected with ⁶⁴Cu-TETA-Y3-TATE (50 μCi, 170 ng) or ⁶⁴Cu-CB-TE2A-Y3-TATE (10 μCi, 35 ng) via the tail vein with or without coinjection of 150 μg of Y3-TATE (CS Bio, San Carlos, CA). Tissue biodistributions were obtained.

AR42J tumor-bearing rats were anesthetized with 1% to 2% isoflurane for the duration of the study. Catheters were placed in the femoral artery and vein. The femoral artery catheter was used to administer ⁶⁴Cu-TETA-Y3-TATE (140 μCi, 165 ng) or ⁶⁴Cu-CB-TE2A-Y3-TATE (160 μCi, 190 ng). Blood samples (5 μ L) were drawn through the femoral vein catheter in capillary tubes between 15 seconds and 1 hour after injection. At 1 hour, a biodistribution study was performed as described above.

PET imaging was performed on a microPET-R4 scanner (Concorde Microsystems, Knoxville, TN). Imaging studies were carried out on male Lewis rats bearing AR42J tumors. Four rats were i.v. injected via the tail vein with ⁶⁴Cu-TETA-Y3-TATE (0.73 mCi, 2.7 μg) or ⁶⁴Cu-CB-TE2A-Y3-TATE (0.77 mCi, 3.5 μg). At 1, 4, and 24 hours after injection, rats were anesthetized with 1% to 2% isoflurane, positioned supine, immobilized, and imaged in two bed positions. Immediately after imaging, a biodistribution study was performed on each rat as described above. Tumor, kidney, and liver standard uptake values {(nCi/ml) × [weight (g)/injected dose (nCi)]} of ⁶⁴Cu-activity were generated by measuring regions of interest that encompassed the entire organ from the microPET images of AR42J tumor-bearing Lewis rats.

All of the data are presented as mean ± SD or mean (95% confidence interval). For statistical classification, a Student t test (two-tailed, unpaired) was performed using GraphPad PRISM (San Diego, CA). Ps < 0.05 were considered significant.

The chelating agent CB-TE2A was synthesized as shown in Fig. 2. The parent cross-bridged cyclam [1,4,8,11-tetraazabicyclo[6.6.2]-hexadecane] was dialkylated with t-butyl bromoacetate to give the conveniently handled and quantified pendant-arm bis(t-butyl ester) in excellent yield (93%). This synthetic intermediate was cleanly and quantitatively deprotected with trifluoroacetic acid in dichloromethane to give CB-TE2A as the bis(trifluoroacetic acid) salt, which was used for the subsequent bioconjugation. This preparation of CB-TE2A is an improvement over our route published previously involving the intermediacy of the bis(ethyl ester; ref. 24).

The peptide Y3-TATE was first assembled on a Wang resin by standard Fmoc solid phase peptide synthesis. A one-pot coupling method was used to introduce CB-TE2A to the peptide by using DIC, which forms both intramolecular and intermolecular anhydrides. The activated intermediate can react with the free amino group of the peptide to form the amide linkage while simultaneously generating a free unactivated carboxyl group. Additionally, the use of low loading resins (<0.4 mmol reactive function/g resin) prevents the dimerization of possible oligomeric acid anhydrides.

A solid-phase method was used to conjugate TETA to Y3-TATE as a solution phase reaction yielded an intractable mixture of products. To solubilize the chelator, minimize hydrolysis of activated carboxyl function, and prevent complete shrinking of the resin, a mixture of DMSO and water in the ratio of 1:100 (v/v) was used. A low-load Wang resin (<0.4 mmol/g resin) was used for the coupling reaction. N,N,N′,N′-Tetramethyl-O-(1H-benzotriazol-1-yl)uronium hexafluorophosphate was used as the coupling reagent to give the desired compound in ∼ 10% yield (100% HPLC purity).

Initial radiolabeling experiments of CB-TE2A-Y3-TATE with ⁶⁴Cu in NH₄OAc (pH 7.0) at 60°C resulted in <40% radiochemical yield (decay corrected) after 4 hours without purification. Raising the temperature to 95°C reduced the reaction time to 2 hours with >95% radiochemical yield (decay corrected) and eliminated the need for any purification steps. This was desirable as including a SepPak purification step reduced the yield by 40% to 50%. The labeling procedure was additionally modified by raising the reaction pH from 7.0 to 8.0. Changing the pH reduced the reaction time to 1 hour without affecting radiochemical yield and improved maximal specific activity from 1.3 mCi/μg to 5.1 mCi/μg. Radiochemical purity of ⁶⁴Cu-CB-TE2A-Y3-TATE and ⁶⁴Cu-TETA-Y3-TATE was >95% as determined by radio-TLC and radio-HPLC. Radiolabeled peptides were used immediately or stored overnight at 4°C. Radiochemical purity was not reduced by overnight storage as determined by radio-TLC.

A homologous competitive binding assay was performed using AR42J tumor membranes bearing SSTr2. ⁶⁴Cu-CB-TE2A-Y3-TATE and ⁶⁴Cu-TETA-Y3-TATE were challenged with increasing concentrations of their respective natural isotopic abundance copper complexes. The IC₅₀ values were determined to be 0.8 nmol/L (0.6–1.3 nmol/L, 95% confidence interval, CI) and 2.3 nmol/L (1.7–3.2 nmol/L, 95% CI) for ^natCu-TETA-Y3-TATE and ^natCu-CB-TE2A-Y3-TATE, respectively (Fig. 3). k_d was calculated to be 0.7 nmol/L (0.5–1.1 nmol/L, 95% CI) for the TETA conjugate and 1.7 nmol/L (1.2–2.4 nmol/L, 95% CI) for the cross-bridged conjugate. These results suggest that Cu-TETA-Y3-TATE and Cu-CB-TE2A-Y3-TATE have similar binding affinities for SSTr in AR42J cell membranes. B_max was determined to be 192 fmol/mg (123–262 fmol/mg, 95% CI) for ^natCu-TETA-Y3-TATE and 1596 fmol/mg (1088–2103 fmol/mg, 95% CI) for ^natCu-CB-TE2A-Y3-TATE.

AR42J cells in culture demonstrated time-dependent internalization of both ⁶⁴Cu-CB-TE2A-Y3-TATE and ⁶⁴Cu-TETA-Y3-TATE that was readily blocked by Y3-TATE in the media (Fig. 4 A). Near-linear internalization over time was observed for ⁶⁴Cu-CB-TE2A-Y3-TATE to 60 minutes (slope = 21.3 ± 2.0, y-intercept = 516 ± 82 fmol/mg, r² = 0.95); the amount of radiotracer internalized at later time points showed no significant change. The nonzero y-intercept of cross-bridged complex suggests a rapid early internalization phase before 15 minutes followed by a slower internalization phase to 1 hour. ⁶⁴Cu-TETA-Y3-TATE uptake was significantly less than ⁶⁴Cu-CB-TE2A-Y3-TATE at 15 (P = 0.0002), 30 (P < 0.0001), and 60 (P = 0.0025) minutes. ⁶⁴Cu-TETA-Y3-TATE internalization increased linearly to 120 minutes (slope = 18.0 ± 1.0, y-intercept = −47.8 ± 67.6 fmol/mg, r² = 0.97). There was no difference between TETA and CB-TE2A peptide-conjugate internalization at 120 and 240 minutes. By 240 minutes, 48.1 ± 1.4% dose/mg protein ⁶⁴Cu-CB-TE2A-Y3-TATE and 47.1 ± 2.3% dose/mg protein ⁶⁴Cu-TETA-Y3-TATE was internalized.

The surface-bound fraction of radiotracer was also quantified (Fig. 4 B). For both conjugates, cell surface-associated activity was specifically blocked by Y3-TATE. The amount of ⁶⁴Cu-CB-TE2A-Y3-TATE bound to the cell surface was significantly higher than ⁶⁴Cu-TETA-Y3-TATE at all of the time points (P ≤ 0.009) except 120 minutes. For ⁶⁴Cu-CB-TE2A-Y3-TATE, the surface-bound activity fell slightly over time. At 240 minutes, the surface-bound fraction was 7.88 ± 0.02% dose/mg protein for ⁶⁴Cu-CB-TE2A-Y3-TATE and 7.19 ± 0.59% dose/mg protein for ⁶⁴Cu-TETA-Y3-TATE.

The tumor and tissue uptakes of ⁶⁴Cu in AR42J tumor-bearing rats are shown in Fig. 5. Both ⁶⁴Cu-TETA-Y3-TATE and ⁶⁴Cu-CB-TE2A-Y3-TATE demonstrated rapid blood clearance (0.242 ± 0.071%ID/g for ⁶⁴Cu-TETA-Y3-TATE at 1 hour, 0.127 ± 0.015%ID/g for ⁶⁴Cu-CB-TE2A-Y3-TATE at 1 hour). By 24 hours, the ⁶⁴Cu activity in blood had fallen to <60% of the 1-hour activity for ⁶⁴Cu-TETA-Y3-TATE and to 11% of the 1-hour activity for ⁶⁴Cu-CB-TE2A-Y3-TATE resulting in 10-fold higher blood activity at 24 hours for ⁶⁴Cu-TETA-Y3-TATE (P < 0.0001). In addition, liver uptake at all of the time points was significantly higher for ⁶⁴Cu-TETA-Y3-TATE than for ⁶⁴Cu-CB-TE2A-Y3-TATE (P < 0.003). Initially (1 hour) there was 1.8 times greater activity in the kidneys for ⁶⁴Cu-TETA-Y3-TATE than for ⁶⁴Cu-CB-TE2A-Y3-TATE (P < 0.03). However, the ⁶⁴Cu-TETA-Y3-TATE activity in the kidneys was reduced by 74% at 24 hours compared with only a 36% reduction in ⁶⁴Cu-CB-TE2A-Y3-TATE kidney activity over the same period. This resulted in 1.3 times higher uptake of ⁶⁴Cu-CB-TE2A-Y3-TATE at 24 hours compared with ⁶⁴Cu-TETA-Y3-TATE (P < 0.01).

In a separate study, blocking specific uptake at 1 hour after injection by coinjection of Y3-TATE resulted in a significant increase in ⁶⁴Cu activity uptake in nontarget organs for ⁶⁴Cu-CB-TE2A-Y3-TATE (blood = 0.179 ± 0.029% ID/g, blood blocked = 0.382 ± 0.053% ID/g; liver = 0.401 ± 0.065% ID/g, liver blocked = 0.837 ± 0.254% ID/g; kidney = 2.917 ± 0.474% ID/g, kidney blocked = 5.746 ± 0.340% ID/g; P < 0.02). This may be due to a general increase in bioavailability of this ⁶⁴Cu-labeled peptide when coinjected with Y3-TATE to block receptor-mediated uptake. There was no significant increase in nonspecific uptake of ⁶⁴Cu-TETA-Y3-TATE with a blocking dose of Y3-TATE (data not shown).

At 1 hour after injection there was no difference in ⁶⁴Cu uptake in bone marrow uptake between the two ⁶⁴Cu-labeled peptide conjugates (Fig. 5). However, at later time points, the marrow uptake of ⁶⁴Cu-TETA-Y3-TATE increased such that the 24-hour uptake was 212% the 1-hour uptake. In contrast, the marrow activity for ⁶⁴Cu-CB-TE2A-Y3-TATE decreased to 39.5% the 1-hour uptake at 24 hours. Thus, by 24 hours the marrow uptake of ⁶⁴Cu-TETA-Y3-TATE was 4.7-fold higher than ⁶⁴Cu-CB-TE2A-Y3-TATE (P < 0.0001).

The uptakes for ⁶⁴Cu-CB-TE2A-Y3-TATE in SSTr-positive tumor, pancreas, and adrenal gland were significantly higher than for ⁶⁴Cu-TETA-Y3-TATE at all of the time points (P < 0.004; Fig. 5). For both ⁶⁴Cu-labeled conjugates, the activity did not change significantly from 1 to 4 hours after injection, suggesting that ⁶⁴Cu was retained in the tumor for at least 4 hours. Copper-64 activity uptake in the tumor cleared over time (18.9% of 1-hour activity in the tumor at 24 hours for ⁶⁴Cu-CB-TE2A-Y3-TATE; 20.1% of 1-hour activity in the tumor at 24 hours for ⁶⁴Cu-TETA-Y3-TATE). Uptake of ⁶⁴Cu-CB-TE2A-Y3-TATE in adrenals and pancreas was >90% blocked by coinjection of unlabeled Y3-TATE (adrenals = 9.112 ± 2.890% ID/g, adrenals blocked = 0.374 ± 0.056% ID/g; pancreas = 5.841 ± 1.217% ID/g, pancreas blocked = 0.371 ± 0.031) indicating specific targeting to SSTr-positive tissues (P < 0.001). In tumors, specific uptake was only reduced by 66% for ⁶⁴Cu-CB-TE2A-Y3-TATE (tumor = 2.408 ± 0.434% ID/g, tumor blocked = 0.823 ± 0.062% ID/g; P = 0.0004) and 74% for ⁶⁴Cu-TETA-Y3-TATE (tumor = 1.613 ± 0.471% ID/g; tumor blocked = 0.422 ± 0.141% ID/g; P = 0.003). This might be attributable to necrosis in the tumor that could result in some nonspecific uptake due to poor tumor clearance.

The tumor to blood ratio at 4 hours for ⁶⁴Cu-CB-TE2A-Y3-TATE was 156 ± 55; for ⁶⁴Cu-TETA-Y3-TATE the tumor to blood ratio was 8.2 ± 1.6 (P < 0.001). Similarly, the tumor to muscle ratio at 4 hours was significantly higher for ⁶⁴Cu-CB-TE2A-Y3-TATE (243 ± 85) than for ⁶⁴Cu-TETA-Y3-TATE (19.6 ± 4.7; P = 0.0003).

To demonstrate that the higher specific uptake with ⁶⁴Cu-CB-TE2A-Y3-TATE compared with ⁶⁴Cu-TETA-Y3-TATE was not due to increased bioavailability of this compound, blood levels of both radiotracers were monitored for 1 hour after injection (data not shown). For the two compounds, peak blood activity levels (<1 minute after injection) were not significantly different (P = 0.08). By 5 minutes after injection, the blood level for both compounds was reduced to ∼25% of the peak level. At 60 minutes after injection, blood radioactivity uptake had cleared to <5% maximum blood levels for both compounds. Thus, no differences in blood clearance over the first hour were detected for ⁶⁴Cu-CB-TE2A-Y3-TATE and ⁶⁴Cu-TETA-Y3-TATE, and no significant differences in blood radiotracer levels were found at any time point. This suggests equivalent bioavailability for ⁶⁴Cu-CB-TE2A-Y3-TATE and ⁶⁴Cu-TETA-Y3-TATE.

Fig. 6,A shows the projection microPET images of AR42J tumor-bearing rats 4 hours after injection of ⁶⁴Cu-CB-TE2A-Y3-TATE (left) and ⁶⁴Cu-TETA-Y3-TATE (right). Imaging was performed at 1, 4, and 24 hours after injection. Prominent uptake was observed in the liver and kidneys of all of the animals. Measuring standard uptake values allows a quantitative comparison between microPET images. Standard uptake values for liver, kidneys, and tumor were determined at all of the time points (n = 2 rats; Fig. 6,B and C). Liver uptakes were similar for the two radiolabeled conjugates at 1 hour, but diverged at later time points rising to 160% of the 1-hour value at 4 hours for ⁶⁴Cu-TETA-Y3-TATE, whereas liver activity for ⁶⁴Cu-CB-TE2A-Y3-TATE did not change significantly over this period. By 24 hours after injection, the liver uptake of ⁶⁴Cu-TETA-Y3-TATE is 3-fold higher than that of ⁶⁴Cu-CB-TE2A-Y3-TATE (P < 0.02). Renal uptake of ⁶⁴Cu-CB-TE2A-Y3-TATE is only reduced by 41% over 24 hours compared with 76% for ⁶⁴Cu-TETA-Y3-TATE, which resulted in higher uptake of ⁶⁴Cu-CB-TE2A-Y3-TATE at 24 hours than of ⁶⁴Cu-TETA-Y3-TATE. These imaging results are consistent with biodistribution data (Fig. 5 A).

Standard uptake values were also determined for tumors imaged by microPET with the ⁶⁴Cu-radiolabeled conjugates (Fig. 6 C). At all of the time points, ⁶⁴Cu-CB-TE2A-Y3-TATE had a higher tumor uptake than ⁶⁴Cu-TETA-Y3-TATE. No tumor uptake was detectable above background by microPET for ⁶⁴Cu-TETA-Y3-TATE at 24 hours.

Copper-64 has been shown to dissociate from TETA and TETA-conjugates and to undergo transchelation to superoxide dismutase in rat liver in biodistribution and metabolism studies (7, 20, 21, 27). The cross-bridged cyclam, CB-TE2A, has demonstrated improved in vitro kinetic stability as well as a reduction of in vivo transchelation (25, 27, 38). These studies suggest that CB-TE2A is an enhanced chelator compared with TETA for labeling ⁶⁴Cu for in vivo imaging and therapy.

Similar binding affinities for SSTr prepared from AR42J tumor membranes were determined for ⁶⁴Cu-TETA-Y3-TATE and ⁶⁴Cu-CB-TE2A-Y3-TATE (Fig. 3). However, there was a significant difference in the B_max values measured for the TETA (192 fmol/mg) and CB-TE2A (1551 fmol/mg) conjugates. This might be attributed to a broader specificity for ⁶⁴Cu-CB-TE2A-Y3-TATE binding to SSTr than for ⁶⁴Cu-TETA-Y3-TATE. Although AR42J tumors express SSTr2, these tumor cells have also been shown to express SSTr1, SSTr3, and SSTr5 (31). Reubi et al. (39) demonstrated that relatively small changes in a radiolabeled peptide, such as changing the chelator, can markedly change the SSTr binding profile of somatostatin analogues. The data reported here suggest that the change from Cu(II)-TETA to Cu(II)-CB-TE2A, i.e., from trans- to cis-folded coordination geometry (24, 25), does not inhibit peptide interactions with SSTr2 and may improve tumor uptake by binding to other SSTr subtypes. This hypothesis is supported by cell internalization studies, where at early time points, the surface-bound levels of ⁶⁴Cu-CB-TE2A-Y3-TATE were significantly higher than for ⁶⁴Cu-TETA-Y3-TATE. In addition, ⁶⁴Cu-CB-TE2A-Y3-TATE showed more rapid internalization kinetics than ⁶⁴Cu-TETA-Y3-TATE. Thus, ⁶⁴Cu-CB-TE2A-Y3-TATE was predicted to have improved specific uptake in target tissues compared with ⁶⁴Cu-TETA-Y3-TATE.

In vivo, the ⁶⁴Cu-labeled CB-TE2A conjugate demonstrated greater SSTr-mediated uptake in SSTr-positive tissues compared with the TETA conjugate (Fig. 5). Higher uptake in SSTr-specific tissues for CB-TE2A could not be attributed to differences in bioavailability of the two radiotracers as determined by 1-hour blood clearance profiles. However, both adrenals and pancreas have been shown to express multiple SSTr subtypes including SSTr2 (40, 41); broadening of subtype specificity with Cu-CB-TE2A-Y3-TATE could contribute to higher specific uptake in these organs compared with the TETA conjugate. The more rapid internalization kinetics of the cross-bridged conjugate in vitro likely also contributes to the increase in specific uptake observed in vivo.

Similar fractions of 1-hour uptake washed out of AR42J tumors by 24 hours for both compounds despite proposed differences in in vivo chelate stability. After SSTr/somatostatin analogue complexes are endocytosed, both receptors and analogues are recycled to the cell surface, and the analogues can be reinternalized (42). Because both TETA and CB-TE2A conjugates show similar affinity for SSTr from AR42J cells, it is reasonable to assume that both would be recycled to the same extent regardless of chelate stability.

There was greater accumulation of ⁶⁴Cu-TETA-Y3-TATE in nontarget tissues such as liver, blood, and marrow compared with ⁶⁴Cu-CB-TE2A-Y3-TATE; this was increasingly evident at later time points. Higher blood and liver uptake and increasing marrow uptake suggest that ⁶⁴Cu dissociates from the TETA peptide conjugate to a greater extent than from the CB-TE2A peptide conjugate in vivo (21). Differences in nontarget tissue biodistribution likely reflect differences in ⁶⁴Cu-chelate stability and clearance between ⁶⁴Cu-CB-TE2A-Y3-TATE and ⁶⁴Cu-TETA-Y3-TATE.

The in vivo behavior of ⁶⁴Cu-TETA-Y3-TATE in a CA20948 tumor-bearing rat model was described previously by Lewis et al. (1). In the current study we report lower target tissue uptake and greater nontarget organ uptake than was found previously (1). The CA20948 rat pancreatic carcinoma model uses older rats than required for AR42J tumors (1, 37). Immature rats may metabolize ⁶⁴Cu-TETA-Y3-TATE differently from older rats, resulting in the observed differences in biodistribution.

Copper-64 activity was retained in the kidney 24 hours after administration of ⁶⁴Cu-CB-TE2A-Y3-TATE, whereas ⁶⁴Cu-TETA-Y3-TATE showed much greater clearance from the kidney (Fig. 5). Differences in renal uptake and excretion may be due to the charge differences of these two conjugates; Cu(II)-TETA-Y3-TATE has a net charge of −1 on the entire conjugate, whereas Cu(II)-CB-TE2A-Y3-TATE is predicted to have a +1 charge. Several studies have investigated the effect of charge on kidney uptake of both ⁶⁴Cu- and ¹¹¹In-labeled compounds (20, 26, 43). Indium-111-DTPA-conjugated peptides are retained in the kidney after reabsorption by renal tubular cells and lysosomal proteolysis (44, 45). Renal retention of ¹¹¹In activity was greatest for positively charged peptide conjugates and lowest for negatively charged ones (43). A similar charge effect was seen for ⁶⁴Cu-labeled azamacrocycles (26). Blocking cationic binding sites in the kidney with d- or l-lysine reduced kidney uptake of ¹¹¹In-DTPA-d-Phe¹-octreotide (46, 47). We suggest that renal uptake of ⁶⁴Cu-CB-TE2A-Y3-TATE may be blocked by coadministration of lysine. Modification of the CB-TE2A backbone by addition of a carboxylate arm may improve the biodistribution characteristics of ⁶⁴Cu-CB-TE2A-Y3-TATE by changing the net charge from positive to either neutral or negative. In addition, a cross-bridged chelator with two carboxylates available for complexation may increase the in vivo stability of the conjugate relative to ⁶⁴Cu-CB-TE2A-Y3-TATE resulting in additional improvements in the biodistribution.

In summary, ⁶⁴Cu-CB-TE2A-Y3-TATE was evaluated for microPET imaging of SSTr-positive tumors in a tumor-bearing rat model. These data suggest that CB-TE2A is a superior chelator for ⁶⁴Cu compared with TETA, due to improved liver and blood clearance, resulting in higher tumor to tissue ratios. Changing the chelator also increased tumor detection sensitivity by PET for ⁶⁴Cu-CB-TE2A-Y3-TATE compared with ⁶⁴Cu-TETA-Y3-TATE. Modifications to CB-TE2A to reduce the net positive charge of the Cu(II) complex may additionally improve the biodistribution. The development of radiopharmaceuticals radiolabeled with ⁶⁴Cu is a growing area of research. The cross-bridged chelator CB-TE2A represents a highly attractive method for stable complexing of ⁶⁴Cu to a wide variety of biological molecules. ⁶⁴Cu-CB-TE2A-Y3-TATE has great potential as a clinical PET radiopharmaceutical for imaging neuroendocrine tumors.

The authors acknowledge Todd A. Perkins and Deborah Sultan for production of ⁶⁴Cu; Susan Adams, Laura Meyer, Lynne A. Jones, Nicole Fettig, and John Engelbach for their excellent technical assistance; and Jerrel Rutlin and Chih-Hsien Chang for help with microPET data analysis.