Cyclotron production of ⁶⁸Ga via the ⁶⁸Zn(p,n)⁶⁸Ga reaction in aqueous solution

Abstract

The objective of the present work is to extend the applicability of the solution target approach to the production of ⁶⁸Ga using a low energy cyclotron. Since the developed method does not require solid target infrastructure, it offers a convenient alternative to ⁶⁸Ge/⁶⁸Ga generators for the routine production of ⁶⁸Ga. A new solution target with enhanced heat exchange capacity was designed and utilized with dual foils of Al (0.20 mm) and Havar (0.038 mm) separated by helium cooling to degrade the proton energy to ~14 MeV. The water-cooled solution target insert was made of Ta and its solution holding capacity (1.6 mL) was reduced to enhance heat transfer. An isotopically enriched (99.23%) 1.7 M solution of ⁶⁸Zn nitrate in 0.2 N nitric acid was utilized in a closed target system. After a 30 min irradiation at 20 μA, the target solution was unloaded to a receiving vessel and the target was rinsed with 1.6 mL water, which was combined with the target solution. An automated module was used to pass the solution through a cation-exchange column (AG-50W-X8, 200-400 mesh, hydrogen form) which efficiently trapped zinc and gallium isotopes. ⁶⁸Zn was subsequently eluted with 30 mL of 0.5 N HBr formulated in 80% acetone without any measurable loss of ⁶⁸Ga. ⁶⁸Ga was eluted with 7 mL of 3 N HCl solution with 92-96% elution efficiency. The radionuclidic purity was determined using an HPGe detector. Additionally, ICP-MS was employed to analyze for non-radioactive metal contaminants. The product yield was 192.5 ± 11.0 MBq/μ·h decay-corrected to EOB with a total processing time of 60-80 min. The radionuclidic purity of ⁶⁸Ga was found to be >99.9%, with the predominant contaminant being 67Ga. The ICP-MS analysis showed small quantities of Ga, Fe, Cu, Ni and Zn in the final product, with ⁶⁸Ga specific activity of 5.20-6.27 GBq/μg. Depending upon the user requirements, ⁶⁸Ga production yield can be further enhanced by increasing the ⁶⁸Zn concentration in the target solution and extending the irradiation time. In summary, a simple and efficient method of ⁶⁸Ga production was developed using low energy cyclotron and a solution target. The developed methodology offers a cost-effective alternative to the ⁶⁸Ge/⁶⁸Ga generators for the production of ⁶⁸Ga.

Introduction

In the recent past, a rapid increase has been noted in both clinical and preclinical studies involving ⁶⁸Ga-labeled radiopharmaceuticals [1-5]. This increase can be attributed to the favorable physical characteristics of ⁶⁸Ga (E_βmax 1.8 MeV, β⁺ 89%, T½ = 67.7 min) for imaging various rapidly changing processes (proliferation, apoptosis, angiogenesis) and targets (growth hormones, myocardial and pulmonary perfusion, inflammation and infection), and to some extent, to newer, more reliable production and labeling methods [1-5]. Gallium-68 labeled somatostatin analogs have already shown their superiority over the existing agent ¹¹¹In-DTPA-octreotide through enhanced sensitivity, specificity, accuracy and cost effectiveness for the diagnosis of patients with neuroendocrine tumors [1,6-9].

The clinical promise of ⁶⁸Ga-labeled radiopharmaceuticals clearly warrants growth of the supply of ⁶⁸Ga to meet the increasing demand in various nuclear medicine facilities. Presently, ⁶⁸Ga can be produced by two different approaches, 1) solid targetry [10,11] and 2) the ⁶⁸Ge/⁶⁸Ga generator [12]. The former requires high capital cost and expertise and specialized cyclotron facilities that accommodate solid targets, whereas, the latter is more broadly accessible in nuclear medicine facilities not equipped with an on-site cyclotron. The simplicity and lower capital cost of the ⁶⁸Ge/⁶⁸Ga generator have made it more popular among the nuclear medicine facilities with relatively lower number of requirements for ⁶⁸Ga labeled doses [1,12]. However, the breakthrough of trace quantities of the long-lived ⁶⁸Ge parent isotope (t½ = 271 d) into the eluted ⁶⁸Ga remains a concern [13]. Furthermore, with increasing applicability of ⁶⁸Ga-labeled radiopharmaceuticals, one can foresee a need for alternative production methods to meet the increasing demand especially for the relatively busy nuclear medicine centers having an on-site cyclotron. There have been previous attempts to produce ⁶⁸Ga using a cyclotron, initially employing a solid target method using ⁶⁸Zn electrodeposition on a copper substrate [10,14] and more recently using a solution target containing an enriched ⁶⁸ZnCl₂ solution [15]. The solid target methods require a lengthy separation step, which is not optimal for short-lived isotopes like ⁶⁸Ga, as well as expensive solid target infrastructure.

Our initial attempts to reproduce the solution target method described by Jensen et al. [15] showed that irradiation of aqueous ⁶⁸ZnCl₂ solutions resulted in rapid pressure increased due to radiolysis-mediated release of hydrogen and oxygen. Additionally, we noticed a pinhole in the Havar target window foil, which may be related to a reaction of the Havar foil and with the ZnCl₂ solution. We have recently performed an extensive study on the mechanistic aspects of water radiolysis during production of ⁸⁹Zr in a solution target [16,17]. This study showed that the use of nitrate salts in dilute nitric acid solutions dramatically decreased rates of water radiolysis during radiometal production [17]. Eliminating the use of ZnCl₂ could also prolong the life of the Havar foil. In the present work, we extended our solution target approach to the production of ⁶⁸Ga using a 1.7 M solution of zinc nitrate (Isotopically enriched) in 0.2 N nitric acid for the production of ⁶⁸Ga in a closed solution target system using 20 μA beam current over 30 min proton irradiation. We also report, an automated separation of ⁶⁸Ga from ⁶⁸Zn using cationexchange resin.

Materials and methods

Targetry details

In this study we designed and developed a new solution target having reduced solution capacity (1.6 mL) with a tantalum target body insert having dual foils of Al (0.20 mm) and Havar (0.038 mm) separated by helium cooling to degrade the proton energy to ~14 MeV. The new conical shape target showed enhanced heat exchange capacity, resulting in reduced gas formation during irradiation that enabled it to be run as a closed system, pressurized at ~40 psi with oxygen. The design of the target is depicted in Figure 1. A semi-automated target loading and unloading system was utilized as described in our previous study [17]. In this study, 30-min irradiations were performed with 1.7 M solutions of ⁶⁸Zn nitrate (99.23% isotopic enrichment) in 0.2 N nitric acid. The proton beam current was 20 μA.

Figure 1.

Design of solution target used for ⁶⁸Ga production.

Chemicals

Zn-68 (99.23%) enriched metal was purchased from Cambridge Isotopes Laboratory (Tewksbury, MA). Hydrochloric acid (34-37% as HCl) and nitric acid (67-70% as HNO₃) both trace metal basis were purchased from Fisher Scientific (Suwanee, GA). Hydrobromic acid (48% as HBr) and acetone were purchased from Sigma-Aldrich (St. Louis, MO). AG-50W-X8 and Chelex-100 (50-100 mesh sodium form) resins were purchased from Bio-Rad (Hercules, CA).

Instrumentation

High-purity germanium gamma spectrometer (Canberra, Meriden, CT) counters running Genie 2000 software was utilized to measure the radionuclide purity. The activity readings were measured using a CRC dose calibrator (#416 setting, CRC-55tPET, Capintec, Ramsey, NJ). A Dionex cation analysis HPLC system equipped with an IonPac CS5A analytical column (4 x 250 mm, Dionex) and in-line radioactivity detector (Carroll and Ramsey Associates, Berkeley CA) was employed to analyze for non-radioactive metal contaminants. The flow rate of mobile phase (Dionex MetPac Eluent) was 1.2 mL/min. For anion analysis, a Dionex ICS-2100 Ion Chromatography System by Thermo Scientific was used, employing an IonPac AS19 analytical column with ion suppression, and mobile phase of 70 mM KOH. A Perkin Elmer ELAN DRC II ICP mass spectrometer by Perkin Elmer was also employed to measure trace metal contaminants.

Method for ⁶⁸Ga separation, determination of specific activity and recovery of ⁶⁸Zn(NO₃)₂

Separation of ⁶⁸Ga from ⁶⁸Zn(NO₃)₂

An in-house built automated system was developed for the separation of ⁶⁸Ga radioisotope from the ⁶⁸Zn(NO₃)₂ target solution as outlined in Figure 2. After irradiation, the target dump (1.6 mL) was received in a collection vial along with a 1.6 mL water rinse of the target. The rinse and dump solutions were mixed and passed through a cation-exchange column (AG-50W-X8, 1.3 g, 200-400 mesh, hydrogen form), which was preconditioned by washing with 15 mL of water followed by air. Both ⁶⁸Ga and ⁶⁸Zn were effectively trapped on the cation exchange column. After trapping, the column was washed with 5.0 mL of water (chelexed water to avoid any metal contamination) to remove short-lived ¹¹C and ¹³N isotopes. A cation-exchange guard column (AG-50W-X8, 5.0 g, 200-400 mesh, hydrogen form) was placed before the waste bottle to trap any potential breakthrough of ⁶⁸Zn during purification. The cation-exchange column was placed inside the dose calibrator to monitor the total radioactivity trapped on the column.

Figure 2.

Schematic diagram of ⁶⁸Ga isolation using a semi-automated system.

Zinc-68 was eluted from the column using 30 mL of 0.5 N HBr in 80% acetone solution and collected in a separate recovery vial followed by a 3 mL water rinse to remove any remaining HBr-acetone. Finally, ⁶⁸Ga was eluted with 3 N HCl (7 mL) to a product vial.

Recovery of ⁶⁸Zn(NO₃)₂ after irradiation was achieved by evaporating the 0.5 N HBr in 80% acetone solution to dryness using a rotary evaporator followed by re-dissolving the residue in concentrated nitric acid (3-5 mL) and evaporating to dryness again on the rotary evaporator. This process was repeated a total of three times to ensure the complete conversion of ⁶⁸Zn to ⁶⁸Zn(NO₃)₂. After conversion, the identity of anionic species was confirmed as nitrate using an HPLC anion chromatography method (Dionex ICS-2100 Ion Chromatography System). The HPLC method did not show the presence of any other anion. The obtained salt was still a viscous material, which on freeze-drying obtained as a solid.

Specific activity and trace metal analysis

Specific activity (GBq/μg) of ⁶⁸Ga was measured by estimating the total Ga metal present in the in the final product after purification. The other metal contaminants including Zn, Fe, Cu, Ni, and Ga were also analyzed using an ICP-mass spectrometer.

Measurement of radionuclide purity

For the measurement of radionuclidic purity we used a gamma ray spectrometer (Canberra, Meriden, CT, DSA1000) equipped with a high-purity germanium (HPGe) detector.

Results and discussion

Initial experiments, separation and automation

Based on the preliminary study by Jensen et al [15], our first irradiations used 1.7 M ⁶⁸ZnCl₂ in the solution target system. At a beam current of 20 μA, the target pressure rose rapidly to >150 psi within 5 min, and the run had to be aborted. This result was not surprising, as we had observed similar results in our previous studies with yttrium chloride [16]. Based on our previous results, we switched to ⁶⁸Zn(NO₃)₂ in 0.2 N nitric acid solution, pressurized with oxygen at ~40 psi. Furthermore, due to the enhanced cooling of the new target design, the target was operated as a closed system. Under these conditions, the target was irradiated for 30 min at 20 μA. The in-target pressure was maintained below 100 psi over the run.

To accomplish the separation of ⁶⁸Ga from ⁶⁸Zn, we modified the cation-exchange method developed by Strelow [18] in the early 1980s for the separation of non-radioactive Ga salts from zinc and other metal salts. A column of 1.3 g of AG-50W-X8 resin was used. ⁶⁸Ga was trapped at more than 99% efficiency. We did not observe loss of ⁶⁸Ga during the elution of ⁶⁸Zn with 0.5 N HBr in 80% acetone or during subsequent rinsing of the column with water. Finally, ⁶⁸Ga was eluted using 3 N HCl solution (7 mL) with 92-96% elution efficiency. The elution efficiency can be increased by increasing the amount of 3 N HCl solution, but at the cost of increasing the volume of acid in the subsequent processing steps. The isotope separation process was automated using an in-house radiochemistry module with actuated valves.

Production conditions of the solution target were 1.7 M ⁶⁸Zn(NO₃)₂ in 0.2 N HNO₃ solution at 20 μA beam current over a 30 min irradiation. The production yield of ⁶⁸Ga was found to be 192.5 ± 11.0 MBq/μA·h after isotope separation having specific activity in the range of 5.20-6.27 GBq/μg (Table 1). These values are decay-corrected to end of bombardment (EOB).

Table 1.

Optimized yields of ⁶⁸Ga in solution target (n = 3)

Uncorrected Yield (GBq)	Corrected Yield (GBq)	Yield (MBq/μA·h)	Specific activity (GBq/μg)	Specific activity (mCi/μg)	ICP-MS determined quantities of Ga and other trace metals in product* (μg, n = 2)
0.96 ± 0.10	1.93 ± 0.11	192.5 ± 11.0	5.20-6.27	140-169	Ga (0.32, 0.39), Fe (17.0, 95.1), Cu (51.4, 12.6), Ni (3.5, 33.7), Zn (1.9, 1.9)

^*These amounts were estimated for the total 7 mL of the product solution.

At 2 h after EOB, a purified sample of ⁶⁸Ga was subjected to HPGe spectrometry and only two peaks were evident, at 511 keV and 1077 keV both corresponding to ⁶⁸Ga. Therefore, the radionuclidic purity of ⁶⁸Ga was >99.9% (Figure 3). However, when same sample was reanalyzed 36 h after EOB additional peaks were observed corresponding ⁶⁷Ga, which on decay correction to EOB, was estimated to be <0.07% of the total ⁶⁸Ga (Figure 4). The Ga-67 emissions were predominated by the Ga-68 emissions at 2 h after EOB. The presence of ⁶⁷Ga impurity can be attributed to two possible nuclear reactions: ⁶⁸Zn(p,2n)⁶⁷Ga and ⁶⁷Zn(p,n)⁶⁷Ga [19,20]. We would anticipate ~0.1% ⁶⁷Ga impurity in the ⁶⁸Ga product for a 60 min irradiation. Radiochemical purity, as measured by cation-exchange HPLC, was >99.9% (Figure 5).

Figure 3.

Characterization of ⁶⁸Ga by high-purity germanium (HPGe) radiation detection after production.

Figure 4.

Trace quantity of ⁶⁷Ga (0.07%) produced during ⁶⁸Ga production and was characterized by high-purity germanium (HPGe) spectroscopy 36 h after EOB.

Figure 5.

HPLC chromatogram of the purified ⁶⁸Ga (Retention Time 2.62 min).

The saturation yields for ⁶⁸Ga before and after the separation of isotopes were found to be 0.43 ± 0.01 GBq/μA and 0.36 ± 0.02 GBq/μA, respectively (Table 2). We attempted to make a comparison with the saturation yield using the solid target method as reported by Sadeghi et al [11]. These authors reported production yield as 5.032 GBq/μA·h, presumably without taking the isotope separation into consideration. The calculated saturation yield from their data is 35.3 GBq/μA. Thus, our yields were approximately 83 times less than the solid target yields. However, Sadeghi et al [11] did not specify the time of isotope separation so we cannot compare isolated product yields. We would anticipate that the comparison of isolated yields would be somewhat more favorable for the solution target method because it does not require a target dissolution step.

Table 2.

Saturation yields of ⁶⁸Ga^*

Before separation (mCi/μA)	Before separation (GBq/μA)	After separation (mCi/μA)	After separation (GBq/μA)
11.5 ± 2.8	0.43 ± 0.01	9.8 ± 0.57	0.36-0.02

^*n = 3 and all values decay corrected to EOB.

Considering that the primary end-use of the ⁶⁸Ga would be in for labeling of molecular targeted peptides, we also analyzed for the presence of trace metal contamination in the final product using ICP-MS. Although very low quantities of Ga were found (0.32-0.39 μg), relatively higher amounts of Fe, Cu and Ni were observed (Table 1). If necessary, these contaminants can be further reduced before labeling using SnCl₂/TiCl₃ and Amberchrom CG-161m resin as described by Van der Meulen et al [21]. Furthermore, in order to achieve the optimum labeling conditions (pH) for various chelators, the final ⁶⁸Ga solution (7 mL, 3 N HCl) can be concentrated using an anionexchange column, where [⁶⁸GaCl₄]- will be effectively trapped and can be further eluted with water as previously described [1,22]. The final pH can be readjusted using the buffer required to meet the labeling conditions. We estimate the cost of a single dose (370 MBq) of the ⁶⁸Ga produced from a cyclotron using the ⁶⁸Zn(NO₃)₂ in 0.2 N HNO₃ solution target method to be $20-25 assuming 85-90% recovery of ⁶⁸Zn, which we found in our initial attempts.

The unoptimized separation time was in the range of 60-80 min, but we would anticipate reduction of this time to ~45 min with further optimization. We were able to make ~25 mCi (0.92 GBq) ⁶⁸Ga at end of separation. By increasing ⁶⁸Zn nitrate concentration (2X) and beam current (2X), we anticipate that >100 mCi (3.7 GBq) can be practically achieved. This would serve the need for 2-4 patients, depending on labeling yields.

Conclusions

A solution target approach for production and automated separation of ⁶⁸Ga was successfully developed employing a solution of ⁶⁸Zn nitrate in 0.2 N nitric acid. The production yield was found to be 192.5 ± 11.0 MBq/μA·h decay-corrected to EOB with a specific activity in the range 5.20-6.27 GBq/μg. Radiochemical and radionuclidic purities were both >99.9%. Increasing the target solution concentration of ⁶⁸Zn and irradiation time may further increase the production yield. The isotope separation method employed AG 50W-X8 resin eluted with a solution of 0.5 N HBr in 80% acetone to remove the zinc isotopes, followed by elution of ⁶⁸Ga in 3 N HCl. The new target design with reduced target volume (1.6 mL) and enhanced heat transfer allowed irradiation as a closed system. Gallium-68 can, therefore, be produced on a low energy cyclotron in sufficient quantities to provide a viable alternative to the ⁶⁸Ge/⁶⁸Ga generator for those facilities that have an on-site cyclotron.

Disclosure of conflict of interest

None.