Improving Contrast and Detectability: Imaging with [⁵⁵Co]Co-DOTATATE in Comparison with [⁶⁴Cu]Cu-DOTATATE and [⁶⁸Ga]Ga-DOTATATE

Abstract

PET imaging at late time points after injection may allow tracer clearance from normal tissue and hence improve image contrast and detectability. ⁵⁵Co is a promising isotope with high positron yield and a long half-life suitable for imaging at delayed time points. Here, we compared the 3 radioconjugates [⁶⁸Ga]Ga-DOTATATE, [⁶⁴Cu]Cu-DOTATATE, and [⁵⁵Co]Co-DOTATATE by PET/CT imaging in NOD-SCID mice bearing subcutaneous somatostatin receptor–expressing AR42J tumors. Methods: ⁵⁵Co and ⁶⁴Cu were produced by the ⁵⁴Fe(d,n)⁵⁵Co and ⁶⁴Ni(p,n)⁶⁴Cu nuclear reactions, whereas ⁶⁸Ga was obtained from a ⁶⁸Ge/⁶⁸Ga generator. ⁵⁵Co and ⁶⁴Cu were labeled with DOTATATE by heating in a sodium acetate buffer and 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid buffer, respectively. AR42J tumor–bearing mice were dynamically scanned 0–1 h after injection. For ⁶⁴Cu and ⁵⁵Co, additional imaging was also performed at late time points after 4 and 24 h. Dose calculations were based on a known biodistribution. The cumulated disintegrations in each organ were calculated by integration of a fitted exponential function to the biodistribution of each respective organ. Equivalent doses were calculated by OLINDA/EXM using the MIRD formalism. Results: Tumor uptake was rapid from 0 to 1 h after injection for all 3 radioconjugates. Normal-tissue ratios as represented by tumor-to-liver, tumor-to-kidney, and tumor-to-muscle ratios increased significantly over time, with [⁵⁵Co]Co-DOTATATE reaching the highest ratio of all radioconjugates. For [⁵⁵Co]Co-DOTATATE, the tumor-to-liver ratio increased to 65 ± 16 at 4 h and 50 ± 6 at 24 h, which were 15 (P < 0.001) and 30 (P < 0.001) times higher, respectively, than the corresponding ratios for [⁶⁴Cu]Cu-DOTATATE and 5 (P < 0.001) times higher than that of [⁶⁸Ga]Ga-DOTATATE at 1 h. Correspondingly, tumor-to-kidney and tumor-to-muscle ratios for [⁵⁵Co]Co-DOTATATE were 4 (P < 0.001) and 11 (P < 0.001) times higher than that of [⁶⁴Cu]Cu-DOTATATE at 24 h. An equivalent dose was calculated as 9.6E−02 mSv/MBq for [⁵⁵Co]Co-DOTATATE. Conclusion: [⁵⁵Co]Co-DOTATATE demonstrated superior image contrast compared with [⁶⁴Cu]Cu-DOTATATE and [⁶⁸Ga]Ga-DOTATATE for PET imaging of somatostatin receptor–expressing tumors, warranting translation into clinical trials. Dosimetry calculations found that effective doses for [⁵⁵Co]Co-DOTATATE were comparable to those for both [⁶⁴Cu]Cu-DOTATATE and [⁶⁸Ga]Ga-DOTATATE.

Neuroendocrine tumors represent a broad heterogeneous group of malignancies. For diagnostic workup and treatment response monitoring, imaging by both functional and anatomic modalities has played a vital role. For functional molecular imaging, the somatostatin receptor (SSTR) has been found to be overexpressed in most neuroendocrine tumors (NETs) (1), paving the way for functional imaging by targeting SSTR. Historically, functional molecular imaging of SSTRs has been performed with SPECT, using somatostatin analogs labeled first with ¹²³I (2) and subsequently with ¹¹¹In (2,3) and ^99mTc (4). With the advent of clinical PET scanners offering higher sensitivities and higher spatial resolution, peptides labeled with positron emitters such as ⁶⁸Ga, ¹⁸F, ⁸⁶Y, and ⁶⁴Cu have been investigated (5–9). Currently, ⁶⁸Ga has attracted most of the attention because of its positron yield of 89% and half-life of 68 min, which match the pharmacokinetics of many peptides. ⁶⁸Ga is also readily available from the ⁶⁸Ge/⁶⁸Ga generator, with the parent isotope, ⁶⁸Ge, having a half-life of 270.8 d, which reduces the generator exchange to 1–2 times per year. Moreover, the relatively simple labeling procedure has further pushed the use of [⁶⁸Ga]Ga-DOTATOC, [⁶⁸Ga]Ga-DOTATATE, and [⁶⁸Ga]Ga-DOTANOC to become the standard for functional molecular imaging in NETs. Although SSTR is the general targeted receptor, the 3 somatostatin analogs have distinct affinity profiles for SSTR subtypes. Of these analogs, [⁶⁸Ga]Ga-DOTATATE exhibits the highest affinity toward SSTR subtype 2 (SSTR2) (10), which is the predominant subtype in NETs.

Besides development of new targeting molecules and analogs, large affinity differences have been observed in SSTR tracers that are identical but labeled with different isotopes (11). This has given rise to an increasing interest in radiotracer labeling with radiometals other than ⁶⁸Ga. In this respect, Reubi et al. (12) found a half-maximal inhibitory concentration of 11.7 ± 1.7 nM and 2.5 ± 0.5 nM for Y-DOTATOC and Ga-DOTATOC, respectively, for SSTR2, whereas respective values of 1.6 ± 0.4 nM and 0.2 ± 0.04 nM were reported for Y-DOTATATE and Ga-DOTATATE, demonstrating large affinity differences for identical vectors but different isotopes in cell lines transfected with SSTR subtypes. These results were extended by Heppeler et al. (13), who synthesized and tested [⁵⁷Co]Co-DOTATOC as a surrogate for the positron-emitting radioconjugate [⁵⁵Co]Co-DOTATOC and found a half-maximal inhibitory concentration of 0.44 ± 0.11 nM, demonstrating, to our knowledge, the highest affinity seen toward SSTR2 in an AR42J tumor cell line. Further, DOTATOC, DOTANOC, and DOTATATE have been labeled with different cobalt isotopes (⁵⁵Co/⁵⁷Co or ^58mCo) for theranostic applications (14), where the cobalt-labeled DOTATATE showed the highest translocation to the cell nucleus (15). Compared with [¹⁷⁷Lu]Lu-DOTATATE, [^58mCo]Co-DOTATATE was significantly more efficient in cell killing per cumulated decay, showing promise for a future theranostic pair using ⁵⁵Co/^58mCo.

Clinically, [⁶⁴Cu]Cu-DOTATATE has been introduced to overcome some of the shortcomings of [⁶⁸Ga]Ga-DOTA-X (16), such as the short half-life lowering the specific activity with time and the high β⁺ energy resulting in low-resolution images. In this respect, [⁵⁵Co]Co-DOTATATE represents a promising candidate for NET imaging because of its physical half-life, allowing late-time-point imaging, and because of the high affinity for SSTR shown for Co-DOTATOC.

The purpose of this study was to perform a head-to-head comparison of [⁵⁵Co]Co-DOTATATE, [⁶⁴Cu]Cu-DOTATATE, and [⁶⁸Ga]Ga-DOTATATE and to characterize the in vivo imaging characteristics in a SSTR-positive xenografted mouse model.

MATERIALS AND METHODS

Isotope Production

⁶⁸Ga was obtained from a ⁶⁸Ge/⁶⁸Ga generator (Eckert and Ziegler), whereas ⁵⁵Co and ⁶⁴Cu were produced in-house using a PETtrace cyclotron. The production method for ⁵⁵Co has been described previously (14,17). Briefly, highly enriched ⁵⁴Fe was irradiated with 18 μA of 8.4-MeV deuterons for 6 h the day before labeling, followed by radiochemical separation using anion-exchange resin (Dowex 1·8; DuPont) and 4 M HCl as eluent, with further purification using a Chromafix 30-PS-HCO3 cartridge. ⁶⁴Cu was produced by proton bombardment of highly enriched ⁶⁴Ni using a fully automated solid-target irradiation system (QIS; ARTMS). The irradiation was performed with 60 μA of 13-MeV protons for 4 h. After the irradiation, the target capsule was pneumatically transferred to a hot cell and mounted in a dissolution cell for dissolving the ⁶⁴Cu/⁶⁴Ni using 6 M HCl. After the dissolution process, the resulting solution was transferred to a GE Healthcare FastLab 2 Developer programmed to automatically separate the ⁶⁴Cu from the ⁶⁴Ni target material using anion-exchange resin (AG 1-X8; Bio-Rad) and HCl as eluent (6.5, 4, and 0.01 M HCl).

Radiolabeling of DOTATATE

DOTATATE was purchased in good-manufacturing-practice grade from ABX GmBH. Radiolabeling with ⁶⁸Ga was performed by the standard Eckert-Ziegler cassette synthesis on the Modular Lab Pharm Tracer synthesis unit in connection with the sterile GalliPharm ⁶⁸Ge/⁶⁸Ga generator. Briefly, in-house–prepared kits of DOTATATE (28 nmol; ABX) in 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid buffer (0.75 M, pH 4.0–4.1, 2 mL) were radiolabeled with ⁶⁸Ga at 95°C for 400 s and then purified by standard C18 solid-phase extraction.

Radiolabeling with ⁵⁵Co was performed using 4.9 nmol of DOTATATE (1 μg/μL) buffered in 50 μL of sodium acetate (pH 4.6; Sigma-Aldrich). This precursor solution was mixed with the [⁵⁵Co]CoCl₂ (86 MBq in 50 μL of 0.04 M HCl) and heated to 90°C using dynamic PETwave microwave irradiation for 2 min.

Radiolabeling with ⁶⁴Cu was performed using 2.8 nmol of DOTATATE (1 μg/μL) buffered in 30 μL of sodium acetate (pH 4.6). This precursor solution was mixed with the [⁶⁴Cu]CuCl₂ (40 MBq in 20 μL of 0.01 M HCl) and left at room temperature for 10 min, after which the solution was loaded onto an Empore cartridge (4315SD), washed with 2 mL of water for injection, and eluted dropwise with aqueous ethanol (70%).

All radioconjugates were finally diluted with phosphate-buffered saline containing 0.1% bovine serum albumin. After dilution, a concentration of 9 vol% ethanol remained in [⁶⁴Cu]Cu-DOTATATE whereas below 2 vol% remained in [⁶⁸Ga]Ga-DOTATATE. No ethanol was used in the formulation of [⁵⁵Co]Co-DOTATATE.

Radiochemical purities were assessed using analytic reverse-phase high-performance liquid chromatography (Hitachi LaChrom Elite system and Phenomenex Jupiter C18 300A column, 150 × 4.60 mm; 5 μm) as previously described (14). [⁶⁸Ga]Ga- and [⁶⁴Cu]Cu-DOTATATE were further analyzed by instant thin-layer chromatography (Biodex 150-771) and eluted in 0.1 M citric acid on which radiometals elute with the solvent front.

In Vivo Animal Studies

Male NOD-SCID mice (n = 10) 7–9 wk old were used from an in-house breeding program. The mice were inoculated subcutaneously with 10⁶ AR42J cells in a Matrigel-to-medium (Matrigel; Corning) ratio of 1:1 in the upper left flank. PET/CT scans were performed after confirmation that an approximately 5-mm tumor had developed as measured by an external caliper. Before scanning, the mice were anesthetized by 1.5%–2% isoflurane in 100% oxygen and placed prone on the scanner bed. All animal experiments were approved by the Animal Experiments Inspectorate in Denmark.

PET/CT scans were performed on a preclinical Inveon PET/CT scanner (Siemens Preclinical Solutions) in docked mode. Before the PET data acquisition, a CT scan was acquired with a full 360° rotation using 270 projections acquired with 80 kV and 500 μA exposures for subsequent attenuation and scatter correction use. List-mode data were reconstructed after scanning using a maximum a posteriori algorithm in conjunction with a 3-dimensional ordered-subset expectation maximum algorithm, giving a central resolution of approximately 1.5 × 1.5 × 1.5 mm.

The injected activity was varied in the mice to match peptide amounts across all groups. Accordingly, the mice were injected intravenously with activities between 1.7 MBq and 4.6 MBq, corresponding to 0.12 ± 0.02 nmol per animal. Four mice were scanned with [⁶⁴Ga]Ga-DOTATATE, 3 with [⁶⁴Cu]Cu-DOTATATE, and 3 with [⁵⁵Co]Co-DOTATATE.

All mice were scanned at 3 selected time points after injection, namely a dynamic scan at 0–1 h, a static scan at 4 h, and a static scan at 24 h. For [⁶⁸Ga]Ga-DOTATATE, the scans at 4 and 24 h were omitted because of the short physical half-life of ⁶⁸Ga.

All mice were awakened from anesthesia between longitudinal scans and allowed to roam freely in cages with unrestricted access to food and water.

Data Analysis

All PET/CT scans were analyzed using Inveon Research Workplace software (Siemens Preclinical Solutions). Tumors were segmented by applying a fixed 40% threshold of the maximum uptake. Muscle uptake volumes of interest were drawn freehand in the right anterior lower gastrocnemius muscle or alternatively in the upper left extensors in cases of high bladder uptake. Such high uptake in the bladder has a tendency to spill into the gastrocnemius muscle, making it poorly suited for a representative measure of background uptake. Additionally, both kidneys were segmented with a procedure akin to the tumor segmentation, whereas the liver was segmented as a freehand ellipsoidal volume of interest below the lungs (according to an anatomic atlas) because of the poor soft-tissue contrast on CT.

After segmentation, volume-of-interest means, SDs, and maximums were calculated. Furthermore, all data were decay-corrected between longitudinal scans to facilitate direct comparison between different isotopes using half-lives of 12.7 h and 17.5 h for ⁶⁴Cu and ⁵⁵Co, respectively.

Dosimetry

Because of animal ethics, dosimetry calculations for [⁵⁵Co]Co-DOTATATE were based on a known biodistribution (15) published previously by our group for [⁵⁷Co]Co-DOTATATE. No kinetic distribution differences are expected between isotopes of the same element. The biodistribution for [⁵⁷Co]Co-DOTATATE is hence equal to that for [⁵⁵Co]Co-DOTATATE. The accumulated activity in the source organs, ${\tilde{A}}_h$, was calculated from the integration of activity in the source organs with time, A_h(t)

$${\tilde{A}}_h=\underset{0}{\overset{\infty }{{\displaystyle \int }}}{A}_h\left(t\right)dt.$$

Source organ data were fitted by an exponential of the following form:

$$A_h\left(t\right)=\frac{A}{k_a-k_b}\times \left(e^{-k_b\cdot t}-e^{-k_a\cdot t}\right),$$

where k_b and k_a are uptake and elimination rates, respectively. Mice organ activity data were organ-to-mass scaled to human equivalents akin to a previous approach (18):

$${\tilde{A}}_{h,o}={\tilde{A}}_{a,o}\times \frac{O_{m,h}/m_h}{O_{m,a}/m_a},$$

where ${\tilde{A}}_{a,o}$ is the cumulated activity in the animal, $O_{m,(h,a)}$is the organ mass in a normal human (19) and mouse, respectively, and $m_{(h,a)}$ is the total mass of a normal human and animal (25 g). Absorbed doses were calculated by the MIRD formalism (20) using S factors in the adult model as implemented in OLINDA/EXM 1.1 (21).

Statistical Analysis

Statistical analysis was performed in Prism, version 6 (GraphPad Software), using 2-way ANOVA and follow-up tests with either the Tukey or the Holm–Šídák approach to correct for multiple comparisons.

RESULTS

[⁶⁸Ga]Ga-DOTATATE was obtained in a yield of 841 ± 159 MBq (n = 6) and a radiochemical purity of 97.8% ± 0.3% at the end of synthesis. [⁶⁴Cu]Cu-DOTATATE was obtained in a radiochemical and radionuclidic purity above 99% (n = 1) at the end of synthesis. [⁵⁵Co]Co-DOTATATE was obtained in 99.5% yield and a radiochemical purity above 99.5% (n = 1) at the end of synthesis. The apparent molar activities at the end of synthesis were approximately 18, 16, and 41 MBq/nmol for [⁵⁵Co]Co-DOTATATE, [⁶⁴Cu]Cu-DOTATATE, and [⁶⁸Ga]Ga-DOTATATE, respectively. Quantitative imaging by PET/CT showed a rapid increase in tumor uptake from 0 to 60 min for all radioconjugates, reaching a maximum at 1–4 h after injection. The comparison of absolute uptake concentrations normalized to injected activity at 1, 4, and 24 h are shown in Figure 1A. Absolute uptake did not significantly differ between radioconjugates at equivalent time points, although a generally lower uptake was found for [⁶⁸Ga]Ga-DOTATATE. Late-time-point imaging as made possible by the longer half-life of both ⁵⁵Co and ⁶⁴Cu than of ⁶⁸Ga resulted in a higher tumor-to-kidney ratio as illustrated by Figure 1C.

FIGURE 1.

Tumor uptake and ratios to liver, kidneys, and muscle with SD error bars. Statistical significance (P < 0.001) is indicated by stars. No significant difference in tumor uptake was observed, whereas statistically significant differences were found among the 3 radioconjugates in all tumor–to–normal-tissue ratios. [⁵⁵Co]Co-DOTATATE showed the highest ratio in liver, kidneys, and muscle and showed highest tumor-to-liver and tumor-to-muscle ratio at 4 h after injection (Supplemental Table 1; supplemental materials are available at http://jnm.snmjournals.org).

Compared with 1 h after injection, the liver concentration was reduced by a factor of 9 at 24 h for [⁵⁵Co]Co-DOTATATE. Oppositely, the liver concentration for [⁶⁴Cu]Cu-DOTATATE increased slightly, but nonsignificantly, from 1 h to 24 h after injection, resulting in a generally lower tumor-to-liver ratio (Fig. 1B). For [⁵⁵Co]Co-DOTATATE, the tumor-to-liver ratio increased to 65 ± 16 at 4 h and 50 ± 6 at 24 h, which were 15 (P < 0.0001) and 29 (P < 0.0001) times higher, respectively, than the corresponding ratios for [⁶⁴Cu]Cu-DOTATATE and 5 (P < 0.0001) times higher than that for [⁶⁸Ga]Ga-DOTATATE at 1 h. Furthermore, the tumor-to-kidney ratio at 24 h for [⁵⁵Co]Co-DOTATATE was 4 (P < 0.0001) times higher than that for [⁶⁴Cu]Cu-DOTATATE and 8 (P < 0.0001) times higher than that for [⁶⁸Ga]Ga-DOTATATE at 1 h. Similarly, as a surrogate for background uptake, the tumor-to-muscle ratio at 24 h for [⁵⁵Co]Co-DOTATATE was 11 (P < 0.0001) times higher than that for [⁶⁴Cu]-DOTATATE and 8 (P < 0.0001) times higher than that for [⁶⁸Ga]Ga-DOTATATE at 1 h.

Dosimetry calculations based on [⁵⁷Co]Co-DOTATATE as a surrogate for [⁵⁵Co]Co-DOTATATE are shown in Table 1.

TABLE 1.

[⁵⁵Co]Co-DOTATATE Dosimetry Results Scaled to Adult Human from Co-DOTATATE Biodistribution

Organ	Ãh,o (h)	Effective dose (mSv/MBq)
Adrenals	0.002	3.2E−04
Kidneys	0.12	7.2E−04
Small intestine	0.13	2.8E−04
Liver	0.013	1.2E−03
Lungs	1.77	6.5E−02
Spleen	0.006	2.0E−04
Pancreas	0.03	6.5E−04
Stomach	0.24	2.1E−02
Effective dose		9.6E−02

From the organ source data, the effective dose equivalent was calculated to be 9.6E−02 mSv/MBq, with the highest individual doses being in the lungs, stomach, and kidneys, at 6.5E−02, 2.1E−02, and 7.2E−04 mSv/MBq, respectively (Table 1).

DISCUSSION

Improvement in image contrast and hence detectability of malignancies from PET imaging depends not only on finding new targeting mechanisms with new tracers but also on the imaging properties of the radionuclide. The development of new tracers by combining existing targeting mechanisms, such as SSTR analogs, but with isotopes with other radiophysical properties, can have a significant impact on diagnostic performance (22,23). In this respect, ⁵⁵Co is a promising isotope because of its high positron yield of 75.9% (24), compared with only 17.6% for ⁶⁴Cu (25), its suitable half-life of 17.53 h, and the demonstrated superior affinity for SSTR2 when bound to DOTATOC (13), which is the predominant overexpressed subtype in NETs. Further, combining ⁵⁵Co with the therapeutic isotope ^58mCo, which is a known Auger electron emitter (17), a promising theranostic pair becomes apparent. Drawbacks of ⁵⁵Co are the high-energy γ-lines of 931.3 keV (75%), 1,316 keV (7.1%), and 1,408.4 keV (16.9%). These prompt γ-lines do not, however, influence image quality significantly on newer-generation PET scanners with modern reconstructions or time-of-flight capability (26), and ⁵⁵Co is therefore a viable option for routine clinical imaging with proper radiation protection.

Dosimetric calculations of [⁵⁵Co]Co-DOTATATE yielded an effective dose equivalent of 9.6E−02 mSv/MBq. For comparison, an effective dose of 3.2E−02 mSv/MBq has been reported for [⁶⁴Cu]Cu-DOTATATE on the basis of biodistributions in humans (16). Similarly, an effective dose of 2.1E−02 mSv/MBq has been reported for [⁶⁸Ga]Ga-DOTATATE (27). The effective dose for [⁵⁵Co]Co-DOTATATE is hence higher per injected activity than is the effective dose of [⁶⁴Cu]Cu-DOTATATE or [⁶⁸Ga]Ga-DOTATATE. However, the low positron yield of ⁶⁴Cu of only 17.6% requires higher amounts of administered activities to obtain diagnostic-quality images. Consequently, in the study by Pfeifer et al. (16), a mean injected activity of 207 MBq was used. To obtain the same equivalent number of annihilation events in identical scanners, the positron fraction must be normalized between ⁶⁴Cu and ⁵⁵Co. The ratio between positron yields from ⁶⁴Cu and ⁵⁵Co is 4.2 (0.76/0.18 = 4.2), suggesting that the injected activity of [⁵⁵Co]Co-DOTATATE should be a factor of 4.2 lower than that of [⁶⁴Cu]Cu-DOTATATE. Scaling the mean injected activity of [⁶⁴Cu]Cu-DOTATATE by 4.2 results in a projected injected activity of 49 MBq of [⁵⁵Co]Co-DOTATATE. An injected activity of 49 MBq corresponds to an effective total dose of 4.7 mSv, compared with a dose commitment of 6.5 mSv of [⁶⁴Cu]Cu-DOTATATE from an injected dose of 207 MBq. [⁵⁵Co]Co-DOTATATE dosimetry is hence comparable to [⁶⁴Cu]Cu-DOTATATE dosimetry from a patient-dosimetry point of view. [⁶⁸Ga]Ga-DOTATATE does indeed lower the effective dose if also adjusted for positron yield, but because of the relatively higher tissue uptake, [⁶⁸Ga]Ga-DOTATATE suffers from low contrast, and because of the short physical half-life, [⁶⁸Ga]Ga-DOTATATE does not allow for late-time-point imaging.

Another dosimetry aspect of ⁵⁵Co is the decay to ⁵⁵Fe, which is also unstable and subsequently decays to stable ⁵⁵Mn with a half-life of 2.7 y. ⁵⁵Fe decays by electron capture with subsequent emission of low-energy x-rays. Most iron in the body is bound to erythrocytes and is hence carried around the body with blood (28). It has previously been stated that ⁵⁵Fe could pose an additional radiation dose (29) after in vivo decay of ⁵⁵Co. This possibility has, however, been disputed, and a radiation dose contribution from ⁵⁵Fe with realistic injected activities of ⁵⁵Co is estimated to be around 1% of the ⁵⁵Co dose if all ⁵⁵Fe is retained in the body, neglecting biologic excretion. The total radiation dose from ⁵⁵Fe is hence believed to be negligible (30).

Previously, the highest known affinity for SSTR2 was found in vitro for [⁵⁵Co]Co-DOTATOC. The high affinity was not directly demonstrated in our data, but a nonsignificant increased uptake rate from 0 to 1 h after injection of [⁵⁵Co]Co-DOTATATE, compared with [⁶⁸Ga]Ga-DOTATATE, was found, supporting previous findings (13) and suggesting that [⁵⁵Co]Co-DOTATATE has a higher affinity for SSTR than does [⁶⁸Ga]Ga-DOTATATE.

Late-time-point imaging to allow clearance of activity from normal tissues with [⁵⁵Co]Co-DOTATATE and [⁶⁴Cu]Cu-DOTATATE at 24 h after injection proved to provide very little gain from imaging at 4 h, based on tumor uptake alone. For [⁶⁴Cu]Cu-DOTATATE, a nonsignificant decrease in the tumor-to-liver ratio was observed, suggesting a concentration increase of free ⁶⁴Cu due to [⁶⁴Cu]Cu-DOTATATE instability, resulting in higher liver uptake with time (Figs. 2D–2F). The ⁶⁴Cu-DOTA complex is known to undergo transchelation in vivo (31,32), leading to activity retention in the liver (33), which agrees with our findings (Figs. 2D–2F). In contrast, [⁵⁵Co]Co-DOTATATE appeared extremely stable in vivo, with statistically significant higher ratios for tumor to liver, tumor to kidney, and tumor to muscle, along with comparable tumor uptake. As such, a decrease in general background was found for [⁵⁵Co]Co-DOTATATE over time, favoring imaging at 24 h. Conversely, a tumor uptake decrease over time resulted in maximum ratios for tumor to liver and tumor to muscle at 4 h. Only with regard to the tumor-to-kidney ratio was imaging at 24 h beneficial. The very high tumor-to-kidney ratio for imaging at very late time points opens up new possibilities for imaging of SSTR, such as imaging of SSTR-positive renal cell carcinomas, which previously has been difficult because of the high kidney and bladder uptake with normal excretion (34–36).

FIGURE 2.

Representative maximum-intensity projection from PET/CT scans comparing [⁵⁵Co]Co-DOTATATE (A–C), [⁶⁴Cu]Cu-DOTATATE (D–F), and [⁶⁸Ga]Ga-DOTATATE (G). Arrows indicates tumor in each animal. Decrease in background was observed for [⁵⁵Co]Co-DOTATATE for late imaging. Higher relative liver uptake with time was observed for [⁶⁴Cu]Cu-DOTATATE, whereas high bladder uptake was observed for [⁶⁸Ga]Ga-DOTATATE. All images are scaled to maximum uptake in tumor.

CONCLUSION

The somatostatin analog [⁵⁵Co]Co-DOTATATE is a promising tracer for PET imaging of SSTR-expressing tumors. The fast clearance of [⁵⁵Co]Co-DOTATATE from normal tissues resulted in significantly higher tumor-to-background ratios than what was observed for [⁶⁴Cu]Cu-DOTATATE and [⁶⁸Ga]Ga-DOTATATE. Importantly, from 4 h after injection, uptake of [⁵⁵Co]Co-DOTATATE in the tumor was at least 50 times higher than that in the liver, which is a major site of metastasis for NETs. Thus, [⁵⁵Co]Co-DOTATATE provides excellent image contrast and warrants further development for translation into the clinic.

DISCLOSURE

The bioimaging experiments reported in this paper were performed at DaMBIC, a bioimaging research core facility at the University of Southern Denmark. DaMBIC was established by an equipment grant from the Danish Agency for Science Technology and Innovation and by internal funding from the University of Southern Denmark. No other potential conflict of interest relevant to this article was reported.

KEY POINTS

• QUESTION: Can [⁵⁵Co]Co-DOTATATE improve PET/CT image contrast by delayed imaging compared with [⁶⁴Cu]Cu-DOTATATE and [⁶⁸Ga]Ga-DOTATATE in somatostatin-positive tumors?

• PERTINENT FINDINGS: [⁵⁵Co]Co-DOTATATE provides a statistically significant improvement in image contrast in NOD-SCID mice bearing subcutaneous SSTR-expressing AR42J tumors, as exemplified by the tumor-to-liver, tumor-to-kidney, and tumor-to-muscle ratios. Delayed imaging with [⁵⁵Co]Co-DOTATATE hence offers image contrast superior to either [⁶⁸Ga]Ga-DOTATATE or [⁶⁴Cu]Cu-DOTATATE.

• IMPLICATIONS FOR PATIENT CARE: The results warrant further translation into clinical practice and hold promise for improved patient diagnostics and future applications for delayed PET/CT imaging.