System and Process for Production of Isotopes and Isotope Compositions

Methods for purifying 89Zr are provided, 89Zr compositions are provided, isotope compositions are provided that can include: a radio isotope and a nanoparticle, and methods for radio labeling monoclonal antibodies are provided.

Skip to: Description  ·  Claims  · Patent History  ·  Patent History
Description
CROSS REFERENCE TO RELATED APPLICATION

This application is a divisional of U.S. patent application Ser. No. 15/788,724 filed Oct. 19, 2017, entitled “System and Process for Production of Isotopes and Isotope Compositions”, which claims priority to U.S. Patent Application Ser. No. 62/410,303 filed Oct. 19, 2016, entitled “System and Process for Production of Labeling-Grade Zirconium-89”, the entirety of each of which is incorporated by reference herein.

STATEMENT AS TO RIGHTS TO INVENTIONS MADE UNDER FEDERALLY-SPONSORED RESEARCH AND DEVELOPMENT

This invention was made with Government support under Contract DE-AC0576RL01830 awarded by the U.S. Department of Energy. The Government has certain rights in the invention.

TECHNICAL FIELD

The present disclosure relates to materials, methods, and compositions for use in isotope purification and/or PET isotope production. Particular embodiments provide materials, methods, and compositions for use in 89Zr purification and PET production.

BACKGROUND

Availability of longer-lived positron emitters has made PET-based imaging of tumors possible by radiolabeling monoclonal antibodies (mAbs), mAb fragments and aptamers, a process referred to as immuno-PET. Immuno-PET combines the high sensitivity and spatial resolution of PET imaging with the antigen specificity of mAbs. Use of 89Zr is gaining tremendous interest in Europe and the U.S. as an immunoPET diagnostic isotope due to its ease of production using monoisotopic (natural) yttrium targets and moderate-to-low energy medical cyclotrons. The long half-life of 89Zr (T1/2=78.4 hr) also enables the potential for off-site isotope production and distribution and for opportunities in new and emerging medical modalities.

Given the high cost of specialty proteins used in immuno-PET diagnostic imaging, it is desirable that binding yields onto these proteins be as high as possible. Another drawback of existing 89Zr purification processing is the high dose rates to personnel that can occur from process scale-up. Accordingly, new processes are needed that address the residual metal contamination for medical applications and personnel dose rates. Embodiments described in the present disclosure address these needs.

Additional advantages and novel features of the present invention will be set forth as follows and will be readily apparent from the descriptions and demonstrations set forth herein. Accordingly, the following descriptions of the present invention should be seen as illustrative of the invention and not as limiting in any way.

SUMMARY OF THE DISCLOSURE

Methods for purifying 89Zr are provided with the methods including: loading a hydroxamate resin with a loading solution comprising HCl and 89Zr; and eluting the 89Zr from the resin using an oxalic acid solution having molarity of less than 1. Useful eluting solutions are provided that can include 89Zr and less than 1 M oxalic acid.

Methods for purifying 89Zr can also include: preparing a loading solution comprising 89Zr and HCl; exposing the loading solution to an ion exchange resin; and eluting the 89Zr from resin using an HCl solution having a molarity greater than 0.3.

Methods for purifying 89Zr can also include: loading a first resin with a first loading solution comprising 89Zr; eluting the 89Zr from the first resin using a first eluting solution; loading a second resin with a second loading solution comprising 89Zr; and eluting the 89Zr from the second resin using a second eluting solution.

89Zr compositions are provided that may have bindable metal concentration ([MB]) less than 169 nmole·g−1 or effective specific activity (ESA) of greater than 800 Ci·mmole−1.

Isotope compositions are provided that can include: a radio isotope and a nanoparticle.

Other isotope compositions can include: a radio isotope; a nanoparticle; and a macromolecule.

Methods for radio labeling monoclonal antibodies are provided that can include: preparing a binding conjugate by exposing a ligand to a solution of 89Zr and less than 1 M oxalic acid.

Methods for radio labeling monoclonal antibodies are also provided that can include: exposing monoclonal antibodies to a solution of 89Zr having a have bindable metal concentration ([MB]) less than 169 nmole·g−1 or effective specific activity (ESA) of greater than 800 Ci·mmole−1.

DRAWINGS

Embodiments of the disclosure are described below with reference to the following accompanying drawings.

FIG. 1 is a depiction of resin loading and elution according to embodiments of the disclosure.

FIG. 2 is a fluidic system configured for the hydroxamate column separation of 89Zr from irradiated Y metal targets. Labels: Syringe Pump w/8-port distribution valve, 2-position Valve, Column, Detection coil, Fraction Collector, Reaction Flask, Peristaltic Pump, and in-line Filter. Blocked port is indicated by “x”; outward arrows on SP and V indicate lines to waste. V is shown in position “1”.

FIG. 3 depicts distribution coefficient (Kd) values for 88Zr tracer on hydroxamate resin as a function of H2C2O4 concentration. 88Zr tracer data is approximated by the curve y=1.87x−1.34 (R2=0.972).

FIG. 4 depicts on-line traces of column effluents during a hydroxamate column separation. (A) Loading 89Zr from dissolved irradiated Y target onto column. (B) Column washes followed by 89Zr elution. Except for the 89Zr elution peak at (5), detected activity is from the sorbed 89Zr's continuous production of unretained 89mY daughter.

FIG. 5 is a graph of four replicate elution profiles for 89Zr from a hydroxamate column (100 mg) using H2C2O4. Peak heights are normalized to each peak maxima. Time zero is set to the delivery of the eluent solution to the evacuated column delivery line. Vertical line represents the reference point (1.2±0.03 mL) for evaluation of peak parameters.

FIG. 6 depicts schematics of fluidic system configurations for the separation of 89Zr from yttrium target. (Left) System configured with a sample loop. (Right) The column receives the dissolved Y target from the flask via a peristaltic pump. Labels: Syringe Pump w/8-port distribution valve, 2-position Valve, Sample Loop, Column, Detection coil, Fraction Collector, Sample Syringe, Peristaltic Pump, and in-line Filter. Blocked port is indicated by “x”; outward arrows on SP and V1 indicate lines to waste. V1 and V2 are shown in position “1”.

FIG. 7 depicts distribution coefficient (Kd) values for 88Zr tracer as a function of HCl concentration for two strongly basic anion exchange resins.

FIG. 8 is an overlay of radiochromatograms showing several radio-Zr elutions (7.5 mL) with varying concentrations of HCl, followed by column strip, from an MP-1M strongly basic anion exchange resin column. The time between arrows represents fluid delivery delay during syringe pump reloading operations. Legend indicates HCl eluent concentrations, in moles·liter−1.

FIG. 9 is a contour plot of the radio-Zr recovery fractions observed during a 7.5 mL eluent delivery to the MP-1M column following load/wash of radio-Zr in concentrated (conc.) HCl. Optimal 88Zr recovery fraction was observed between 6 M HCl+(0.33−1.0) mM NaF. (Contour lines calculated by DeltaGraph v. 5.6.4).

FIG. 10 is a load of simulated dissolved Y foil target (20 mL of 25 mg/mL Y) that was spiked with either 88Zr or 88Y radiotracers and delivered by peristaltic pump to the column (0.5 mL/min). Next, 5 mL column wash (conc. HCl) was delivered via syringe pump (0.75 mL/min). Arrow indicates a small peak resulting from removal of 88Y from droplets in the empty tubing and interstitial spaces of the column.

FIG. 11 depicts distribution coefficient (Kd) values for radio-Zr on MP-1M (∘) and QAE-550C (□) as a function of dissolved Y metal concentration in conc. HCl.

FIG. 12 depicts radiochromatograms; (Top Row) Radiochromatogram pairs showing the elution (5 mL) and strip traces of 88Zr tracer following load/wash from 0.5 g dissolved Y metal (25 mg/mL). Elute and strip reagents are delivered in forward flow (- - -) and reverse flow (-) directions; eluent solutions are (A) 6 M HCl, (B) 6 M HCl+0.33 mM NaF, and (C) 6 M HCl+1.0 mM NaF. (Bottom Row) Corresponding cumulative recovery of 88Zr tracer across the 5 mL elution volume.

FIG. 13 depicts on-line absorptivity traces of ferric chloride across the MP-1M column elute and strip cycles (360 nm). (A) FF (- - -) and RF (-) observed during a 6 M HCl elution. (C) FF (- - -) and RF (-) observed during a 6 M HCl+1 mM NaF elution. The fluid delivery methods are identical to those utilized for 88Zr tracer experiments in FIG. 12, conditions A and C.

FIG. 14 is a depiction of a dual column purification system according to an embodiment of the disclosure.

FIG. 15 is a depiction of another dual column purification system according to an embodiment of the disclosure.

FIG. 16 depicts purification systems according to an embodiment of the disclosure: (Left) Fluidic system to evaluate the performance of the secondary column in isolation of the primary column. (Right) Fluidic system configured for the tandem column separation of 89Zr from irradiated Y metal targets. The column receives the dissolved Y metal from the flask via a peristaltic pump. Labels: Syringe Pumps w/8-port distribution valves, 2-position Valves, Columns, Sample injection Syringe, Detection coil, Fraction Collector, 89Zr product Collection Vessel, Peristaltic Pump, and in-line Filter. Blocked ports are indicated by an “x”; arrows indicate lines to waste. Valves are shown in position “1”.

FIG. 17 depicts distribution coefficient (Kd) values for 88Zr tracer as a function of NaF concentration in 6 M HCl on hydroxamate resin.

FIG. 18 depicts graphs of on-line detector traces of column effluents during a tandem column separation. (A) Loading 89Zr from dissolved irradiated Y target onto primary column. (B) Tandem column separation followed by 89Zr elution. Except for the 89Zr elution peak at (7), detected activity is from the sorbed 89Zr's continuous production of unretained 89mY daughter.

FIG. 19 depicts quadruplicate detector traces showing elution profiles for 89Zr from a hydroxamate microcolumn in H2C2O4. Peak heights are normalized to the peak maxima for comparison. Vertical line is set at the elution start reference point of 0.536 mL.

FIG. 20 depicts 89Zr binding fraction results on aminopropyl anion exchange cartridge plotted as a function of deferoxamine mesylate (Dfo-m) present in the transchelation solution in a Dfo-m titration experiment.

FIG. 21 depicts distribution coefficients (Kd) of 88Zr tracer onto Fe3O4 and Mn-doped Fe3O4 magnetic nanoparticles in varying concentrations of HCl (left) and pH-adjusted phosphate buffered saline (right).

FIG. 22 depicts distribution coefficients (Kd) of 88Y tracer onto Fe3O4 and Mn-doped Fe3O4 magnetic nanoparticles in varying concentrations of HCl (left) and pH-adjusted phosphate buffered saline (right).

FIG. 23 depicts 88Zr chloride elution profiles from QAE-550C mini-column. 88Zr oxalate is loaded onto the column; column is converted to the formate form; then low concentrations of HCl are used to elute the 88Zr tracer as ZrCl4.

 DESCRIPTION

This disclosure is submitted in furtherance of the constitutional purposes of the U.S. Patent Laws “to promote the progress of science and useful arts” (Article 1, Section 8).

The materials, methods, and compositions of the present disclosure will be described with reference to accompanying FIGS. 1-23 as well as U.S. Provisional Patent Application Ser. No. 62/410,303 filed Oct. 19, 2016, entitled “System and Process for Production of Labeling-Grade Zirconium-89”, the entirety of which is incorporated by reference herein.

Referring first to FIG. 1, an example purification method is shown using a single resin. In the first instance, system 10 is loaded with material to be purified. Material 12 can include a mixture of materials, such as an isotope M1 (14), contaminant metals Mz (16) and other materials*(18). In accordance with example implementations, material 12 can be exposed to resin material 19.

In accordance with example implementations, methods for purifying the isotope 89Zr, can include loading a hydroxamate resin material, in a column, with a loading solution comprising HCl and 89Zr. This solution can be directly obtained with the dissolution of yttrium. Example irradiated yttrium can include, but is not limited to, cyclotron bombarded Y foil, for example. Some metals of the solution, M1 and Mz, may be bound to the resin, and much of the remainder of the loaded solution removed from the resin.

In accordance with example implementations, the resin can be exposed to an eluting solution, in this case a solution of oxalic acid having a molarity less than 1. The oxalic acid solution (H2C2O4) can be between 0.3 and less than 1 M, or 0.8 and less than 1 M. In accordance with example implementations, a solution can be obtained during elution in system 20 that includes 89Zr and less than 1 M oxalic acid. For example, this results in the need for less buffering agents to be added to the 89Zr product fraction prior to mAb labeling. In accordance with example implementations, eluted 89Zr product fractions demonstrated chemical recoveries from irradiated Y foil targets, with 1 mL product volumes yielding 89±2% of the elution peak activity and 84±2% of the 89Zr from the irradiated target.

Referring to FIG. 2, delivery of dissolved yttrium target solution to a hydroxamate column, column washing, and 89Zr elution; in its simplest configuration, the process is accomplished using a syringe pump, a peristaltic pump, and a 2-position valve. The system can incorporate on-line monitoring of column effluents with the use of a NaI(Tl) scintillation detector for example. Beyond the detector, a fraction collector captures column effluents for example.

The fluidic system has been used to demonstrate the performance of the 89Zr purification method commonly used in the field. However, the method has been modified to utilize a significantly less concentrated oxalic acid eluent solution (0.8 M), which primarily serves to minimize the quantity of buffering agents required to be added prior to labeling the 89Zr product. The concentration of oxalic acid used in the field (1 M) can be close to saturation at laboratory temperatures; the solubility limit of H2C2O4 in water at 20° C. is 95.2 g·kg−1 (1.06 mol·kg−1), or a 8.69% (w/w) mass fraction. The methods and systems may reduce the amount of metal contaminants introduced into the product. With reference to FIG. 3, the lower concentration of oxalic acid was shown to be feasible by determining the distribution coefficient (Kd) of Zr (as 88Zr) on hydroxamate resin across a range of oxalic acid concentrations, values which have not previously been reported.

Optima™ grade hydrochloric acid (Fisher Scientific, Waltham, Mass.) was used for Y target dissolution and column washes, and hydrogen peroxide (29-32%, Sigma-Aldrich) was used in the Y target dissolution. TraceSELECT® oxalic acid dihydrate (>99.9999%, Sigma-Aldrich, St. Louis, Mo.) was used for 89Zr column elutions. All solutions were prepared from deionized water (18.3 MΩ·cm) using a Barnstead Nanopure Diamond water purification system (Dubuque, Iowa). Hydroxamate ligand was bound onto Accell Plus CM weak cation-exchanger (37-55 μm, Waters Corp., Milford, Mass.) by the method of Verel et al. The hydroxamate column was prepared using 100 mg resin packed into a disposable 1 cc SPE column using 2 mm thick polyethylene (PE) frits (Supelco, Bellefonte, Pa.). With an ID of 5.6 mm, the resin bed had a height of ˜1.0 cm.

Zirconium-88 (88Zr) and yttrium-88 (88Y) radiotracers (tracers), with half-lives of 83.4 and 106.6 days, respectively, were purchased through the Department of Energy Office of Science National Isotope Development Center (NIDC). The 88Zr was periodically purified of in-grown 88Y using an anion exchange separation method, whereby 88Zr was loaded onto a ˜0.8 cc AG MP-1M column (Bio-Rad Laboratories, Inc., Hercules, Calif.) in concentrated (conc.) HCl (88Y is unretained). Then, after a conc. HCl column wash, the 88Zr was eluted in 6 M HCl. 89Zr was obtained from the University of Washington (UW) PET radiochemistry center as an irradiated Y foil of 0.25 mm thickness (Alfa Aesar, Ward Hill, Mass.). Following an initial decay period of ˜2 h, the 89Zr target activity was determined by a dose calibrator and corrected to activity at end of bombardment (EOB). Upon delivery to PNNL, this activity was verified using a high purity germanium (HPGe) detector (Ortec, Oak Ridge, Tenn.) calibrated with NIST-traceable standards, with energy and efficiency checks performed daily.

Referring to FIG. 3, distribution coefficients (Kd) for each radiotracer (88Zr, 88Y) were determined on synthesized hydroxamate resin across a range of H2C2O4 concentrations. Resin contact volumes were typically 5 mL in 2-dram borosilicate glass vials. Sorbent to solution ratios were between ˜4 and 20 mg/mL. The resin was initially weighed into tared vials, and the vials were then transferred into a radiological fume hood. Additional vials were included that did not contain resins; these vessels served as controls for determination of the initial activities of 88Zr and 88Y for each spiked solution. Reservoirs of H2C2O4 solutions were spiked with either 88Zr or 88Y and were mixed thoroughly. Next, 5 mL aliquots of the spiked H2C2O4 solutions were pipetted into each resin contact and control vessel. Vessels were shaken for ˜4 h at ˜150 rpm on an orbital shaker. Following the contact period, the vials were removed from the shaker. A second set of clean 2-dram vials were brought into the radiological fume hood. A disposable plastic syringe was used to extract the contacted solution; a 17 mm diameter, 0.45 μm polyethersulfone (PES) syringe filter (Thermo Scientific, Waltham, Mass.) was then attached to the syringe and the resulting filtrate was delivered to the clean vessel. Finally, 2.0 mL of filtrate was transferred to a 12×74 mm test tube for counting on a Wizard 1470 (PerkinElmer, Meriden, Conn.) automatic gamma counter containing a well-type NaI(Tl) scintillation detector. The detector was configured with counting protocols specific to 88Zr and 88Y gamma emission regions of interest (corresponding to 393 keV for 88Zr (97.3% intensity) and 898 keV for 88Y (93.7% intensity)).

As shown in FIG. 2, the system includes a 48,000 step digital syringe pump (SP) coupled to an 8 position distribution valve (Norgren, Las Vegas, Nev.). The SP was configured with a 10 mL displacement syringe (Flex Fluidics, Las Vegas, Nev.). A 6-port 2-position Cheminert valve (V), driven by a microelectric actuator, was manufactured by Valco, Inc. (Houston, Tex.). The pump and valve were assembled into two separate boxes provided by J-Kem (St. Louis, Mo.), and were operated by a customized version of Kem-Pump software. The pump box was positioned outside the radiological contamination area (RCA) in order to minimize the amount of equipment this zone. A check valve placed in-line with the reagent delivery tubing assured that only flow direction into the RCA was possible. The system was plumbed using 0.5 mm and 0.75 mm ID× 1/16″ OD Teflon® FEP tubing for reagent transport lines. Lines were connected to the hardware using PEEK ¼-28 flangeless nuts with Tefzel® ferrules (Upchurch Scientific, Oak Harbor, Wash.) and PEEK 10-32 nuts with PEEK conical ferrules (Valco). The hydroxamate column (C) was inserted into the influent and effluent lines with PEEK luer to ¼-28 nut couplers. The column effluents were routed through an on-line detector (D) to a fraction collector (FC).

The Y foil was dissolved in a 50 mL 2-neck round-bottom reaction flask (RF) that contained a magnetic stir bar. A 4-channel Sci-Q 400 peristaltic pump (PP) (Watson-Marlow, Wilmington, Mass.) was configured to deliver the Y foil dissolving solution via a tube inserted through the side neck of the RF using pump channel 1 (not shown). Channel 2 of the PP was used to withdraw the dissolved Y foil solution via a polyetherimide (ULTEM™) sipper tube inserted through the central neck. Downstream of the PP, an in-line filter (17 mm diameter, 0.45 μm PES syringe filter (Thermo Scientific) was connected to the delivery line using PEEK luer to ¼-28 couplers (Upchurch). The dissolved target solution delivery line was then routed to the 6-port 2-position V. In position 1, the V directed the PP flow to a waste line, and in position 2, the PP connected to the column. Inversely, the SP access to the valve's waste and column delivery ports were positions 2 and 1, respectively.

The fluidic system reagent delivery protocol for purifying 89Zr on a hydroxamate column is summarized in Table 1.

TABLE 1 Automated hydroxamate resin column separation protocol for the load/ wash-elute sequence of cyclotron-produced 89Zr from a dissolved Y metal target. Flow Volume, Rate, Valve Step Purpose Reagent mL mL/min Position 1 Condition 2M HCl 5 1.0 1 2 Load Y foil in 2M HCl/ 6 0.5-0.6 2 0.1M H2O2 a 3 Wash 1 2M HCl 10 1.0 1 4 b Wash 2 H2O 10 5 Elute 0.8M H2C2O4  5 c a Foil masses ranged from 0.15-0.25 g b At the completion of the step, air was pushed through the column delivery line to purge reagent c Only the first ~1-1.5 mL was collected as the 89Zr product fraction

The column is first conditioned with 2 M HCl by the syringe pump (SP) via the valve (V), which is initially set to position 1. The irradiated Y foil (typically 150-250 mg) is placed in the 2-neck round-bottom flask. The peristaltic pump (PP) is engaged to deliver 6 mL 2 M HCl/0.1 M H2O2 to the side neck of the flask at a rate of ˜6 mL/min. The solution is stirred during dissolution with a magnetic stir bar to minimize frothing and spatter inside the flask. The solution is allowed to cool with stirring for 10-20 min, followed by 5-10 min of settling prior to initiating column load. Next, the dissolved foil solution is delivered to the column at 0.5-0.6 mL/min by the PP through an in-line filter (F) to remove residual solids, and then through the V (toggled to position 2). Once the solution is transferred onto the column, the V is returned to position 1, where the SP is once again connected to the column. The SP sequentially delivers 10 mL of 2 M HCl and 10 mL of H2O washes to the column at a 1 mL/min flow rate. At the completion of the H2O wash, and prior to the 89Zr elution step, the SP pushes air through the column delivery line until all fluid is purged from the column and detector/fraction collector delivery lines. Zirconium-89 can be eluted using 5 mL of H2C2O4, for example. The fraction collector is used to separate the elution volumes so only the first ˜1-1.5 mL of the elution volume can be used for subsequent evaluation of product labeling performance.

The 89Zr purification process is 1) fluidically automated, 2) provides near-real time monitoring of column effluents, and 3) uses a significantly lower concentration of H2C2O4 (0.8 M) eluent than that reported previously. A schematic of the fluidic device is shown in FIG. 2, while Table 1 describes the computer-controlled 89Zr separation protocol employed by the device.

On-line detector traces for a hydroxamate resin column separation of 89Zr from an irradiated Y foil target are shown in FIG. 4. For this run, the irradiated Y foil (165 mg) contained 3.46 mCi 89Zr (EOB). Note that the NaI(Tl) scintillation detector was raised above the detection loop prior to the start of the run to assure that the 89Zr elution peak would not saturate the detector. The detector trace for the column load step (via the PP) is shown in FIG. 4A (step 2 in Table 1). The trace shows a continuous increase in detector response over time. As the 89Zr concentrates on the column, an ever-increasing amount of the short-lived and un-retained 89mY daughter is being delivered to the detector. The rapid drop in detector response at ˜23 min corresponds to the completion of loading the dissolved target onto the column, as air has been pushed through the column and detector loop.

FIG. 4B shows the on-line detector trace for steps 3-5 in Table 1. During the 2 M HCl (3) and H2O (4) deliveries, the detector signal rises and plateaus. The signal corresponds to the delivery of unretained 89mY that is being continuously generated by the 89Zr parent that is sorbed on the column. For a given activity of sorbed 89Zr, the 89mY detector response is a function of the effluent flow rate and the internal tube volume between the column and the detector. Dips in the detector signal are caused by the stopping/reloading of the SP between reagent deliveries, as the 89mY quickly decays within the detector cell during these disruptions in flow. Activity spikes at each resumption in flow are 89mY that has in-grown on the column during the brief pauses of the syringe. Finally, the elution of 89Zr is shown at the end of the trace (5).

Five Y foils were irradiated and then processed as per the described hydroxamate resin column method. Table 2 provides a summary of the 89Zr activities (at EOB) and masses for each foil, as well as the chemical separation performance of 89Zr from the dissolved foils.

TABLE 2 Activity distribution of 89Zr loaded onto a hydroxamate resin column across five irradiated Y foil targets. 89Zr Foil 89Zr Distribution from Foil Activity, Mass, Hydroxamate Column, % b ID mCi a mg Load Wash 1 Wash 2 Elute d 1 1.69 244 c 1.9 0.3 0.0 97.6 2 0.16 252 c 1.2 0.1 0.0 98.6 3 0.66 249 c 0.2 0.1 0.0 99.7 4 4.51 153 0.2 0.0 0.5 99.3 5 3.46 165 0.0 0.0 0.0 99.9 Average ± 1σ e 0.7 ± 0.8 0.1 ± 0.1 0.1 ± 0.2 99.0 ± 0.9 a At end of bombardment (EOB) b Total activity recovered from the column separation c Foil mass supplemented with unirradiated Y foil d 89Zr recovery in 5 mL elution volume e Sample standard deviation

Irradiated foils 1-3 were originally ˜150-180 mg, but the mass was augmented with the addition of unirradiated Y foil to ˜250 mg in order to assure the robustness of the method. Foils 4 and 5 received no additional Y foil. For each 89Zr purification, all column effluents were collected in order to determine the fraction of 89Zr activity at each step. The total elute fraction contained 99.0±0.9% of the column-loaded activity. However, some 89Zr activity was lost to the system, either in the form of residues remaining from the foil dissolution (activity remaining in the reaction flask and in-line membrane filter) or as activity trapped in the column. Residual 89Zr activity was calculated to be 4±1% of the original foil activity. Considering the 89Zr activity lost to the system, the total 89Zr yield in the elution fractions was 95±2% of the 89Zr present in the irradiated Y foil.

The elution profiles for four replicate 89Zr purification runs are shown in FIG. 5. The vertical scale is normalized to the peak maximum of each elution profile for ease of comparison. Zero volume is set at the point at which the SP begins delivering eluent solution to the column delivery line, which was purged with air prior to this step. The elution peaks began entering the detector after 1.21±0.03 mL was delivered to the system (vertical dashed line). Thereafter, the peak maximum was obtained after 0.33±0.04 mL; 90% and 95% of the elution activity was obtained after 1.02±0.09 mL and 1.71±0.15 mL, respectively. One milliliter fractions collected from the elution profiles in FIG. 5 represent 89±2% of the 89Zr eluted from the columns. When considering the fraction of 89Zr lost to the system, the 1 mL elution fraction represents 84±2% of the total 89Zr present in the irradiated target.

With reference to FIG. 1 again, a different resin can be utilized, for example an ion exchange resin. The resin of this embodiment may be strongly basic anion exchange resin, gel-type or macroporous resin, that may include styrene divinylbenzene polymer or acrylic polymer. Accordingly, methods are provided for purifying 89Zr that can include: preparing a loading solution comprising 89Zr and HCl; exposing the loading solution to an anion exchange resin; and eluting the 89Zr from resin using an HCl solution having a molarity greater than 0.3. In accordance with example implementations, the HCl solution may have a molarity between 0.3 and 9, or between 2 and 6. The HCl solution may have a molarity of at least 0.3. In accordance with additional implementations, the HCl solution may also include fluoride, in ion form, for example NaF.

In accordance with example implementations, three strongly basic anion exchange resins were evaluated. Each had quaternary amine functional groups with different polymer supports. The supports ranged from macroporous and 10% cross-linked styrene-divinylbenzene copolymer to a hydroxylated methacrylic polymer. The resins were evaluated for their ability to load and elute Zr in neat solutions and solutions containing high concentrations of dissolved Y. Additionally, the method had to ensure removal of Fe (a metal that competes for antibody labeling sites for example) from the Zr product. In accordance with example implementations, eluent solutions are provided to remove 89Zr selectively to Fe(III). Accordingly, after a dissolved irradiated Y foil was loaded and subsequently washed, both 89Zr and Fe may remain on the resin. Eluting solutions are provided that elute 89Zr while leaving Fe(III) on the resin.

The 89Zr purification method involves cyclotron bombarded Y foil dissolution, loading of 89Zr from the dissolved target onto an anion exchange column, column wash, and 89Zr elution. This sequence of steps was accomplished with a prototype automated fluidic platform (FIG. 6) configured with an on-line gamma detector and in-line UV-Vis absorbance detector to monitor 89Zr and Fe(III) column effluents respectively in near-real time.

TraceMetal and Optima grade hydrochloric acid (34-37% HCl, Fisher Scientific) working stock solutions were prepared as-is or as dilutions into deionized water (18.3 MΩ·cm) using a Barnstead E-Pure water purification system (Dubuque, Iowa). Method development work and cyclotron bombarded Y foil processing used TraceMetal and Optima grade acids, respectively. Sodium fluoride (99.99+%) and hydrogen peroxide (30%) were purchased from Sigma-Aldrich (Milwaukee, Wis.). Several ion exchange resins were evaluated: AG MP-1M (Cr form, 200-400 mesh) and AG 1-X10 (Cl− form, 100-200 mesh) strongly basic anion exchangers, each on styrene-divinylbenzene co-polymer (Bio-Rad Life Science, Hercules, Calif.), and Toyopearl QAE-550C strongly basic anion exchanger (Cl− form, 50-150 μm) on hydroxylated methacrylic polymer (TOSOH Bioscience, King of Prussia, Pa.).

As depicted in FIG. 6, a fluidic system was configured with a 48,000 step digital syringe pump (SP) with an 8-position distribution valve (Kloehn, Las Vegas, Nev.) and a 10 mL zero-dead-volume syringe (FlexFluidics, Las Vegas, Nev.). Two Cheminert valves coupled to microelectric actuators (Valco Instruments Co, Inc., Houston, Tex.) allowed fluids to be directed into various flow paths. (FIG. 6, left). Valve 1 (V1) was a 6-port 2-position valve and Valve 2 (V2) was a 4-port 2-position valve. The hardware was packaged into separate pump and valve boxes by J-KEM (St. Louis, Mo.), and was controlled by a modified version of Kem-Pump software (J-KEM). All transport and reagent lines were made from 0.5 mm and 0.75 mm i.d. fluorinated ethylene propylene (FEP) tubing (Upchurch Scientific, Oak Harbor, Wash.), respectively. Fittings were made from polyether ether ketone (PEEK), of either ¼-28 (Upchurch) or 10-32 (Valco) thread size. Ferrules were ethylene tetrafluoroethylene (ETFE) or FEP. The column was 4.6×50 mm (0.83 cc internal volume), made by OmegaChrom (Upchurch), with frits from IsoLab, Inc. (Akron, Ohio). The column delivery and exit lines were plumbed into the 4-port 2-position V2 such that flow to the column could be directed to either the forward flow (FF) or reverse flow (RF) direction, depending on valve position.

Two automated fluid delivery protocols were prepared for optimization of 88Zr tracer separation methods on ion exchange columns. For initial column performance studies, 88Zr tracer was introduced to the column using a 0.29 mL sample loop configured within a 6-port 2-position valve (V1). A description of the protocol is shown in Table 3, and the system is illustrated in FIG. 6, left.

TABLE 3 Chemical separation protocol for load/wash-elute-strip sequence on columns for determination of optimal eluent for 88Zr. Volume, Step Purpose Reagent mL a VI Position b 1 Condition Conc. HCl 5 1 2 Load c 88Zr in Conc. HCl 5 2 3 Wash Conc. HCl 4 Elute Variable HCl 7.5 1 concentrations 5 Strip 3M HCl + 0.1M NaF 4 a Flow rate 0.75 mL/min; all flow in forward direction through column b V2 is in position 1 for duration of run c Delivered via 0.290 mL sample loop through a 6-port 2-position valve

The SP was programmed to deliver reagents at 0.75 mL/min. With V1 in position 1, the sample loop (SL) was isolated, and the SP delivered solutions directly to the column to be conditioned with conc. HCl. Next, the 88Zr was loaded onto the column by toggling V1 to position 2, whereby a 5 mL aliquot of load solution from the SP pushed the 88Zr-bearing solution to the column and subsequently washed the column. Next, 7.5 mL of 88Zr eluent solution (varying between 0.1 and 9 M HCl) was delivered to the column, followed by 4 mL of strip solution to ensure all 88Zr was removed from the column and the tubing.

Once the optimal separation media and 88Zr tracer eluent solution was selected, a second fluid delivery protocol was prepared, and is summarized in Table 4.

TABLE 4 Chemical separation protocol for the load/wash-elute-strip sequence of 88Zr and 88Y from dissolved Y metal (or 89Zr from a dissolved irradiated foil target) loaded onto an MP-1M column. Flow rate 0.75 mL/min, unless otherwise specified. Volume, V1 Flow Step Purpose Reagent mL Position Direction a 1 Condition Conc. HCl 5 1 FF 2 Load b 88Zr in Y/ 20 2 Conc. HCl 3 Vessel Rinse c Conc. HCl 0.5 (×2) 4 Wash Conc. HCl 5 1 5 Elute 6M HCl + 5 FF/RF e NaF d 6 Strip 3M HCl + 7.5 0.1M NaF a V2 configured for FF (forward column flow) or RF (reverse column flow) b Delivery of dissolved Y target solution to column via peristaltic pump (0.5-0.6 mL/min) c 2 × 0.5 mL conc. HCl wash of dissolution vessel walls after dissolved target solution was completely aspirated from vessel d Eluents evaluated: 6M HCl, 6M HCl + 0.33 mM NaF, and 6M HCl + 1.0 mM NaF e The eluent solutions in (d) were evaluated under FF and RF directions

The hardware was reconfigured to that shown in FIG. 6, right. This protocol enabled the delivery of dissolved Y metal solution to the column via a PP with V1 in position 2, immediately following a 5 mL delivery of conditioning solution via the SP with V1 set to position 1. After delivery of the dissolved Y solution, V1 was reset to position 1 for the remainder of the protocol. This allowed the SP to deliver reagents directly to the column.

Batch contact experiments were performed in order to determine the distribution coefficients (Kd) of Zr and Y (as 88Zr and 88Y) on two anion exchange resins across a range of HCl concentrations (0.1 to ˜12 M). AG MP-1M (MP-1M) and Toyopearl QAE-550C (QAE-550C) were evaluated. A Kd map for Zr on each resin as a function of HCl concentration is shown in FIG. 7. The MP-1M exhibited superior Kd values at high HCl concentrations (˜7× higher) when compared to QAE-550C. In no case was 88Y observed to sorb onto the resins.

The Kd map shows low Kd at HCl concentrations below ˜9 M, and steeply rising Kd as the HCl concentration approaches ˜12 M. As HCl concentration increases, Zr chloride complexes transition from ZrCl4→ZrCl5−1→ZrCl6−2, the latter of which is most strongly retained on the quaternary amine group of the resins.

Both MP-1M and QAE-550C were evaluated for their performance in loading and eluting 88Zr tracer from a 4.6×50 mm column. The fluidic platform was configured as shown in FIG. 6, left, and the fluid delivery protocol shown in Table 3 was loaded into the instrument controller. Approximately 10 kBq of 88Zr tracer in conc. HCl was loaded into the 0.29 mL sample loop. The fluid delivery protocol was activated, and it delivered conditioning, sample load/wash, 88Zr elution, and strip reagents to the column. The on-line detector monitored the activity of the column effluents over time, and the fraction collector, in tandem with the detector, collected each reagent into a separate vial for off-line counting.

FIG. 8 shows a zoomed-in view of the radiochromatograms collected using an MP-1M column, with time zero corresponding to the beginning of the syringe pump delivery of eluent to the 88Zr tracer-loaded and washed column. The eluent solutions (7.5 mL) were varied between 0.1 and 9 M HCl, with only 1, 6, 7.5, and 9 M HCl eluent radiochromatograms included in the figure. Subsequent to the delivery of the eluent, the syringe pump loaded 4 mL of strip solution (3 M HCl+0.1 M NaF) and delivered it to the column to remove any residual Zr. The concentration of HCl in the eluent solution had a dramatic effect on the elution profile of the 88Zr tracer, with peaks broadening as the HCl concentration increased. At 9 M HCl, no 88Zr was removed from the column during the elution cycle. The other observation of note is the incomplete elution of 88Zr during the elution cycle across all concentrations of HCl. In each case, a discernable fraction of 88Zr tracer remained uneluted within the 7.5 mL elution window, only to be expeditiously removed from the column during the strip cycle. The fraction of 88Zr found in the elute and strip solutions for 0.1 to 9.0 M HCl are listed in Table 5.

TABLE 5 Performance of anion exchange resins on the recovery of 88Zr following the load/wash-elute-strip routine described in Table 3. Values represent the fraction of 88Zr tracer measured between the elute and strip solution deliveries vs. total loaded activity. 4.6 x 50 mm Column a, b MP-1M QAE-550C [HCl], M Elute Strip Elute Strip 0.1 0.43 0.56 0.3 0.53 0.47 0.92 0.08 1.0 0.65 0.35 0.93± 0.07± 0.04 0.04 2.0 0.72 0.28 0.93± 0.03± 0.01 0.01 3.0 0.78± 0.22± 0.96± 0.04± 0.01 0.00 0.02 0.02 4.5 0.83 0.17 0.97± 0.03± 0.01 0.01 6.0 0.86± 0.14± 0.97± 0.03± 0.04 0.05 0.01 0.01 7.5 0.82± 0.18± 0.98± 0.02± 0.03 0.02 0.02 c 0.02 9.0 <0.01 1.00 0.97± 0.02± 0.02 c 0.01 a 88Zr activity distribution determined by off-line gamma counting of collected fractions b 88Zr load/wash fraction not shown (typically <1%), but is included in calculations c Despite high apparent 88Zr recovery, elution peaks were very broad

The activity in the eluent fraction from the MP-1M resin is poor at low concentrations and increases to a maximum at 6 M HCl (0.86±0.04 elute/0.14±0.05 strip), above which it decreases slightly, and ultimately falls to near zero at 9 M HCl.

The incomplete elution of 88Zr tracer on the macroporous MP-1M prompted the evaluation of the more traditional gel-type anion exchange resin AG 1-X10 (Bio-Rad) under the same conditions. However, the elution profile at 6 M HCl was virtually identical to MP-1M at the same acid strength (0.81 elute/0.13 strip).

The experiments were repeated with the column packed with QAE-550C across a range of HCl concentrations QAE-550C exhibited significantly improved chromatographic performance, with a substantially higher elute fraction across all HCl concentrations tested. 88Zr tracer elution recoveries were observed between HCl molarities of 0.3 (˜92%) to 9 (˜97%). Eluent condition ranges can also be between 2 and 6 M HCl (˜97%) (Table 5).

Once the optimal eluent solutions were obtained for MP-1M (6 M HCl) and QAE-550C (2 M to 6 M HCl), the automated protocol was repeated with 15.9 kBq 88Y tracer in conc. HCl loaded into the 0.29 mL sample loop. Yttrium-88 tracer was observed to be thoroughly removed from the column during the 88Zr during the load/wash step. Using a 6 M HCl elute solution, the MP-1M and QAE-550C 88Y distributions for the load/wash—elute—strip fractions were 0.999/0.001/0.000 and 0.998/0.002/0.000, respectively.

Despite MP-1M resin's ability to better retain Zr in the presence of high concentrations of Y, its chromatographic performance was poor, with incomplete elution (vide supra). An approach was therefore sought to improve the column's Zr elution performance. Zr produces chlorofluoro complexes in the general form (ZrCliFj)(4-i-j). The presence of low concentrations of HF across a range of HCl concentrations can have a profound effect on the anion exchange behavior of Zr when compared to the behavior of Zr in HCl alone. Specifically, the presence of HF prevents the steep rise in Zr affinity for anion exchange resin at high HCl concentrations. This is likely caused by the formation of chlorofluoro complexes with high negative charges.

An experimental matrix was performed to determine how effective the fluoride ion (F−) was in enhancing the Zr elution performance. Solution combinations of HCl and fluoride ion (as NaF) were prepared with concentrations ranging between 1 and 6 M (1, 2, 3, 4, 6 M) and 0 and 1 mM (0, 0.1, 0.33, 1.0 mM), respectively. Each solution in this matrix was evaluated for its ability to efficiently elute 88Zr tracer from a pre-loaded column. FIG. 9 shows the results of the column elution performance matrix, with the z-axis representing the fraction of 88Zr tracer collected within the 7.5 mL elution volume. The poorest elution performance was exhibited at the lower left corner of the matrix (1 M HCl with no NaF), while the optimal elution performance was at the diagonal opposite corner of the matrix: 6 M HCl with NaF ranging between ˜0.33 to 1 mM NaF. At 6 M HCl, this optimal NaF concentration range had a pronounced effect on the 88Zr tracer elution recovery fraction, assuring>99% recovery from the column within the 7.5 mL elution volume, as compared to 86±4% recovery when using 6 M HCl alone (Table 5).

Simulated dissolved irradiated Y targets were prepared using non-irradiated Y metal spiked with 88Zr or 88Y tracers following metal dissolution and solution filtration. One-half gram of Y (Alfa Aesar pieces) was utilized in order to robustly verify method performance, as an actual Y target mass would be expected to be significantly lower than this. Two allotments of 0.5 g Y metal were dissolved in 20 mL conc. HCl each using a 50 mL round-bottom flask (with magnetic stirring). After allowing the solution to cool, the solution was filtered (0.45 μm PES, Thermo Scientific) and spiked to ˜1 kBq/mL 88Zr (solution 1) and 6.5 kBq/mL 88Y (solution 2) tracers, respectively. The fluidic platform was configured as shown in FIG. 6, right, and the fluid delivery protocol shown in Table 4 was loaded into the instrument controller. FIG. 10 shows the resulting detector traces for the dissolved Y metal load and column wash portions of the 88Zr and 88Y tracer-spiked runs through the MP-1M column. During the 20 mL load volume, no 88Zr tracer breakthrough was observed.

The peristaltic pump was operated until the entire sample volume, plus rinses (0.5 mL×2), had been delivered to the column (the rapid drop in 88Y tracer signal at the end of the load is air pushed through the detection coil). The small 88Y peak shown in the wash segment (arrow) represents residual 88Y droplets removed from the fluid delivery lines plus removal of residual sample load solution trapped in the interstitial spaces of the column. The 88Y tracer, used as an indicator of the dissolved Y metal in the simulated dissolved target solution, shows virtually complete removal of Y after just a couple of milliliters of wash solution. Of 130 kBq 88Y tracer in 0.5 g dissolved Y metal delivered to the column, <6 Bq were observed in the subsequent Zr elute fraction (signal was below the gamma detector detection limit). This activity is equivalent to <23 μg Y, and represents a Y decontamination factor (DF) of >22,000.

An additional set of experiments was performed in order to evaluate the Zr loading performance of the MP-1M column across a range of Y concentrations and load volumes in conc. HCl. Table 6 summarizes the load conditions and the observed Zr breakthrough levels.

TABLE 6 88Zr loading performance on MP-1M column at various load volumes and Y concentrations. Load Flow Y Conc., Vol., Rate, 88, 89Zr Breakthrough, % mg/mL mL mL/min Load Wash 11.9 ± 0.8 a 20 0.5-0.6 1.1 ± 0.8 0.0 19.2 10 1.0 0.1   25.0 b 20 0.0 0.0 39.3  5 0.3 0.1 a Performed on irradiated Y targets (89Zr, n = 3) b 88Zr spike in 0.5 g dissolved Y, as shown in FIG. 10

Despite the Zr Kd dependence on Y concentration (FIG. 11), no significant level of Zr breakthrough was observed during the load steps, nor was it observed in the subsequent column washes for the Y concentrations and load volumes indicated.

Six separation protocol variations, each representing only a change in the eluent solution delivery, were run using the Table 4 protocol. The load solution contained 0.5 g dissolved Y metal and 88Zr tracer in 20 mL conc. HCl. The elute (reduced to 5 mL) and strip portion of the detector traces for each run are shown in FIG. 12 (top row). The time gap between the arrows represents the time required for the syringe pump to reload itself with the strip solution (between ˜6.6 and 8 min). FIG. 12 (bottom row) shows the cumulative recovery of eluted 88Zr tracer obtained during the 5 mL eluent solution delivery. The dashed traces represent forward flow (FF) of the elute and strip solution deliveries to the column, while the solid lines show the same solutions delivered under reverse flow (RF) conditions.

Set A shows FF and RF elute and strip traces for 88Zr with an eluent solution of 6 M HCl. By the end of the FF elution volume, only 91.4% of the retained 88Zr was removed from the column. For the 6 M HCl RF trace, 98.5% of the 88Zr was recovered, with 97% being removed in 2.9 mL. Set B shows FF and RF traces when the eluent solution was 6 M HCl+0.33 mM NaF. The presence of fluoride ion causes a discernible sharpening of the elution peak for the FF condition, with the RF condition showing only a slightly enhanced elution profile. The FF condition resulted in 98.9% of the retained 88Zr tracer being removed in 5 mL, with 97% removal within 2.9 mL, while the RF condition yielded 99.4% recovery in 5 mL and 97% removal within 2.4 mL of the eluent solution. Set C shows the same for an eluent delivery of 6 M HCl+1.0 mM NaF. Here, the increased complexing strength of the higher fluoride ion concentration results in almost indistinguishable 88Zr tracer elution profiles between FF and RF. The 5 mL elution recoveries are 99.5% and 99.3%, respectively, while 97% is recovered in 2.4 mL for both flow direction conditions. Based on these observed elution profiles, it was concluded that a solution of 6 M HCl+0.33 mM NaF provided essentially equivalent elution profiles of Zr from the MP-1M column, regardless of eluent solution flow direction.

The test Y foils (Alfa Aesar, 0.25 mm thickness) were cyclotron irradiated to produce 89Zr activities ranging between ˜0.2 and 1.6 mCi. The foils were dissolved in conc. HCl, and the resulting solutions delivered to the column as described above; the solutions had Y concentrations between ˜9 and ˜19 mg/mL.

The fluid delivery protocol outlined in Table 4 was implemented to perform the 89Zr/Y separation. The 89Zr was eluted with 6 M HCl+0.33 mM NaF prior to delivery of the column strip solution. Table 7 provides a summary of the performance from eight 89Zr purification runs performed under three conditions: 1) RF elution direction with no H2O2 added to the dissolving solution; 2) RF direction with H2O2 added to the dissolving solution; and 3) FF elution direction with H2O2 added to the dissolving solution. The table lists the column effluent distributions for 89Zr determined by off-line measurement of all collected fractions.

In order to account for the distribution of all 89Zr in the irradiated foil (not only 89Zr delivered to the column), post-run fluidic system washes were performed, and the solutions were counted off-line. These washes included the round bottom flask and the dissolved target delivery line and filter. Additionally, the washed in-line filter and column were sent to the counting lab to determine 89Zr activity in the trapped solids. The last two columns of Table 7 present 1) the fraction of 89Zr lost in the fluidic system as droplets and as insoluble particulate matter (the primary source of 89Zr activity loss), and 2) the 89Zr column elute fraction compared to the total 89Zr present in the irradiated foil.

TABLE 7 Results from the optimized MP-1M column separation of 89Zr from cyclotron irradiated foils. Results shown with and without H2O2 treatment during Y target dissolution. Flow direction of 89Zr elute and strip are indicated. Fraction of Total 89Zr 89Zr Distribution on Produced 89Zr, % Foil Activity, [Y], Flow H2O2 Column, % c Insoluble Elute ID mCi a mg/mL Dir. Added Load Wash Elute Strip Residues d Fraction e 1 0.23  8.7 RF No 8.1 0.3 85.8 5.8 10.2  77.0 2 0.53  8.8 10.2  0.5 83.1 6.1 9.2 75.5 Average 9.2 0.4 84.5 5.9 9.7 76.2 (±difference) (2.1) (0.2)  (2.7) (0.3) (1.0)  (1.6) 3 0.79 18.6 RF Yes b 0.4 0.0 98.4 1.2 2.1 96.4 4 0.18 12.1 0.1 0.0 99.6 0.3 7.9 91.7 5 0.97 13.1 0.4 0.1 98.8 0.7 1.6 97.3 Average 0.3 0.0 98.9 0.7 3.9 95.1 (±1σ) (0.2) (0.0)  (0.6) (0.4) (3.5)  (3.0) 6 0.19 11.6 FF Yes b 1.7 0.0 96.8 1.5 1.5 98.0 7 0.17 11.3 1.5 0.0 96.4 2.1 2.3 94.2 8 1.56 12.8 0.2 0.0 98.6 1.3 2.3 96.3 Average 1.1 0.0 97.3 1.6 2.0 96.2 (±1σ) (0.8) (0.0)  (1.1) (0.4) (0.4)  (1.9) a Activity determined at EOB b 0.1M H2O2 added to conc. HCl immediately prior to target dissolution c Activity fraction for all 89Zr-bearing solutions delivered to the column d Combined activity of 89Zr trapped in insoluble residues, remaining in dissolution vessel, trapped in in-line filter and column, etc. e Fraction of 89Zr in elute fraction, considering all activity in irradiated foil

The conc. HCl used to dissolve the first irradiated targets did not contain H2O2 (foils 1 and 2, RF direction). These were small targets, each ˜200 mg. Surprisingly, a significant fraction (8-10%) of 89Zr was lost to the column during the sample load, and ˜6% was recovered in the strip fraction. Approximately 10% of the 89Zr activity was found as insoluble residue. Combined, this resulted in a ˜76% 89Zr eluent recovery from the overall foil activity. The poor 89Zr yield in the elute solution had not been observed during simulated dissolved Y target studies with 88Zr spike, and is perhaps evidence of a persistent non-anionic Zr complex or colloidal solid formed during cyclotron irradiation or during target dissolution (e.g., zirconium oxide or zirconyl species).

The effect of H2O2 in the HCl dissolver solution on 89Zr recovery was therefore evaluated. Foils 3-5 (RF direction) were processed as above, but the conc. HCl reagent was spiked to 0.1 M H2O2 immediately prior to target dissolution. Despite the increased masses of these foils (between ˜240 and ˜370 mg Y), the amount of 89Zr lost in the load dropped significantly to <1%. Additionally, the amount of 89Zr in the strip fraction dropped from ˜6% to 0.7±0.4%. The 89Zr elution fraction recovery now reached 98.9±0.6% of column-delivered effluents, and 95±3% of the overall foil activity. Foils 6-8, performed with H2O2 in the dissolver solution and the preferred FF column elute direction, performed equally as well, with 97±1% of the column delivered activity, and 96±2% of the overall foil activity, found in the 89Zr elute fraction.

The 89Zr product fractions from Foils 6-8 were analyzed for Y, Fe, and Zr by inductively coupled plasma-mass spectrometry. It was determined that the three product fractions contained 0.74±0.38 μg Y from Y target masses that averaged 0.23±0.02 g. This represents a Y decontamination factor of ˜3×105. Iron in the 89Zr product fractions was below the detection limit of the ICP-MS. Therefore, the mass of Fe present could only be reported as <0.16 μg. Based on the certificates of analysis (COA) for the Y metal foils used in this study, the irradiated foils would be expected to contain 66±38 μg Fe; therefore, a conservative Fe decontamination factor of >410 is reported. However, given the excellent separation of Zr from Fe using this method (FIG. 13), the Fe decontamination factor is likely much higher than this. The COA did not provide a value for elemental Zr in the foil, but the ICP-MS results of the analyzed 89Zr product fractions indicated that elemental Zr is present in the foil at 0.4±0.1 μg/g, after correction for the 96±2% recovery in the product fraction. Because the concentration of natural Zr in the chosen Y foil will have a pronounced effect on the quality of 89Zr labeling onto proteins for example, it is imperative that each Y foil source be analyzed. Some Y foils screened over the course of this study contained Zr levels as high as 240 μg/g; Zr concentrations at these levels would not provide good 89Zr labeling performance, regardless of the non-Zr elemental purity.

Referring next to FIG. 14 an example depiction of a dual resin system 120 that includes a first resin column 122 having a first resin 124 and a second resin column 126 having a second resin 128. In accordance with example implementations the volume of resin 128 can be a fraction of the volume of resin 124. For example, the first resin is at least twice the volume of the second resin. Further, the solution used to elute metals of importance from column 122 can be a loading solution for column 126, making these solutions, eluting and loading, the same in accordance with example implementations.

In accordance with additional implementations and with reference to FIG. 15, system 130 can include the columns and resins of FIG. 14, however valving/pumping configurations can be included in the form of valving/pumping system 132 and 134 respectively. These systems can be operatively coupled to reservoirs that may receive or provide solutions as dictated by the configuration to allow for system 130 to be operated automatically.

Accordingly, methods for purifying 89Zr are provided with the methods including loading a first resin with a first loading solution. This first loading solution can include dissolved Y foil that includes 89Zr, for example. The method can then include washing the first resin and then eluting the 89Zr from the first resin using a first eluting solution. In the case where the first resin is an ion-exchange resin, the elution can be performed with an HCl solution or an HCl and fluoride solution. The method can then include loading the second resin with a second loading solution comprising 89Zr. In some embodiments, the loading solution for the second resin can be 89Zr in HCl solution. In some embodiments, the loading solution for the second resin can be 89Zr in HCl obtained from the eluting solution from the first resin. The 89Zr on the second resin can be eluted with a second eluting solution such as oxalic acid. In accordance with example implementations, the first resin can be an ion exchange resin and the second resin is a hydroxamate resin or HDEHP resin.

In accordance with example implementations, TraceMetal and Optima grade hydrochloric acid (34-37% HCl, Fisher Scientific) working stock solutions were prepared as-is or as dilutions into deionized water (18.3 MΩ·cm) using a Barnstead E-Pure water purification system (Dubuque, Iowa). Method development work and irradiated Y foil processing utilized TraceMetal and Optima grade acids, respectively. Sodium fluoride (99.99+%), hydrogen peroxide (30%), and TraceSELECT® oxalic acid dihydrate (>99.9999%) were purchased from Sigma-Aldrich (St. Louis, Mo.). AG MP-1M strongly basic anion exchange resin (CI− form, 200-400 mesh), with quaternary amine functional groups on a macroporous styrene divinylbenzene co-polymer (Bio-Rad Life Sciences, Hercules, Calif.) was used for the primary column. Hydroxamate ligand was synthesized onto Accell Plus CM weak cation-exchanger (37-55 μm, Waters Corp., Milford, Mass.) and used for the secondary column. The MP-1M column had internal dimensions of 4.6×50 mm (0.83 cc internal volume) (OmegaChrom column, Upchurch Scientific, Oak Harbor, Wash.), and used frits from IsoLab, Inc. (Akron, Ohio). The hydroxamate column was constructed from a 2 cm long, ˜27 μL internal volume, internally tapered mini-column with PTFE frits (GlobalFIA, Fox Island, Wash.).

Two fluidic system configurations were utilized for this work. Initially, a simple fluidic system was assembled to evaluate the performance of a second resin, in this case a single hydroxamate microcolumn, as shown in FIG. 16 (Left). The system included a 48,000 step digital syringe pump (SP) coupled to an 8 position distribution valve (Norgren, Las Vegas, Nev.). The SP was configured with a 2.5 mL displacement syringe (Flex Fluidics, Las Vegas, Nev.). A 6-port 2-position Cheminert valve (V1), driven by a microelectric actuator, was manufactured by Valco, Inc. (Houston, Tex.). The SP and V1 were assembled into separate boxes, and were remotely by computer. The pump box was positioned outside the radiological zone to minimize the amount of hardware therein. A back-flow preventer was installed between the SP and V1 to ensure no radiological solution could backflow out of the radiological zone. A sample loop (SL) of 0.032 mL inner volume was plumbed within V1 as shown. With V1 in position 1, the loop could be filled with a radiotracer spiked solution using a hand-held sample injection syringe (SS), while the SP could deliver reagents directly to the column (C). With V1 in position 2, the SP could drive the contents of the SL to the column. All column effluents were routed through an on-line detector (D), and then to a fraction collector (FC). In accordance with example implementations, the internal volume of the microcolumn housing the hydroxamate was about ˜27 microliters; while the primary AnIX column was about 0.83 mL.

The second system was configured for an automated tandem column purification of 89Zr from irradiated Y foils. It was comprised of two digital syringe pumps (SP1 & SP2), each with 8-position distribution valves. The SP1 and SP2 syringe displacement volumes were 10 mL and 2.5 mL (Flex Fluidics), respectively. A 6-port 2-position valve (V1) and two 4-port 2-position valves (V2 & V3) were Valco Cheminert valves with microelectric actuators. The pumps and valves were assembled into separate. Once again, back-flow preventers were installed in the SP1 and SP2 reagent delivery lines where they crossed the radiological boundary. In lieu of a SL in V1, the second system incorporated the ability to push a dissolved irradiated target solution from a 50 mL round-bottom flask via a peristaltic pump (PP) to the primary column when V1 was in position 2; SP1 could deliver reagent to the C1 when V1 was in position 1. The target dissolution chamber and dissolved target delivery system have been described herein.

Two automated fluid delivery protocols were prepared in order evaluate the load/wash-elute performance of 88Zr tracer using column-based separations. The first protocol was designed to evaluate the performance of a hydroxamate microcolumn in isolation from the primary anion exchange column. The steps of the automated protocol are presented in Table 8.

TABLE 8 Automated chemical separation protocol for load/wash-elute sequence on hydroxamate microcolumn for determination of column load/elute properties of 88Zr and 88Y. Volume, Step Purpose Reagent mL a V1 Position 1 Condition 6M HCl + NaF 2   1 2 Load b 88Zr in 6M HCl ± NaF c 2.5 2 3 Wash 6M HCl ± NaF c 4 Elute 0.8M H2C2O4 2.5 1 a Flow rate is 0.2 mL/min; all flow is in forward direction through column b Sample delivered in a 0.032 mL sample loop via a 6-port 2-position valve c [NaF] was varied between 0 and 10 mM

Solutions were aspirated into the 2.5 mL syringe and delivered at 0.2 mL/min. With V1 in position 1, the SP could deliver the conditioning solution to the column, while 88Zr or 88Y tracer spiked samples remained isolated in the loop. The sample was loaded onto the column, along with 2.5 mL wash solution, after toggling V1 to position 2. Finally, 2.5 mL of 88Zr tracer eluent solution (H2C2O4) was delivered to the column after V1 had returned to position 1.

Once ideal load/wash-elute conditions were obtained for the hydroxamate microcolumn, a tandem column purification method for 89Zr from dissolved Y foil was developed. The separation protocol is presented in Table 9, and the fluidic system shown in FIG. 16 (right).

TABLE 9 Automated MP-1M → hydroxamate resin tandem column separation protocol for the purification of cyclotron-produced 89Zr from dissolved Y metal targets. Column a Volume, Flow Rate, Valve Position Step Purpose (Pump) Reagent mL mL/min V1 V2 V3 1   Condition C1 Conc. HCl 5   1.5 1 1 1 (SP1) C2 2M HCl 2.5 0.4 1 1 1 (SP2) 2   Load C1 Y foil in 20   0.5-0.6 2 1 1 (PP) Conc. HCl b 3 c Wash 1 C1 Conc. HCl 5   1.5 1 1 1 (SP1) 4   Col. Transfer C1→C2 6M HCl + 5   0.2 1 2 1 (SP1) 0.33 mM NaF 5 c Wash 2 C2 2M HCl 2.5 0.4 1 1 1 6 c Wash 3 (SP2) H2O 2.5 1 1 1 7 c Elute 0.8M   2.5 d 0.2 1 1 2 H2C2O4 a For C1/C2 and SP1/SP2/PP designations see FIG. 16 (Right) b Target dissolver solution consisted of conc. HCl + 0.1M H2O2 c At the completion of the step, air was pushed through the column delivery line to purge reagent d Only first ~0.25-0.30 mL was collected as the 89Zr product fraction

Initially, the primary (C1) and secondary (C2) columns were conditioned with conc. HCl and 2 M HCl using SP1 and SP2, respectively. Next, the primary column was loaded with the 89Zr from the dissolved Y metal target using the PP. Following sample load, the primary column was washed with additional conc. HCl using SP1 to remove all traces of dissolved Y. The 89Zr on the primary column was then transferred to the secondary column by toggling V2 to position 2 and delivering 5 mL of 6 M HCl+0.33 mM NaF with SP1 to both columns. With all 89Zr activity now on the secondary column, SP2 commenced to deliver 2 M HCl and H2O wash solutions, respectively. Finally, the 89Zr was eluted from the secondary column, using H2C2O4, to 89Zr product collection vessels (CV) with V3 positioned to position 2. While in route from the microcolumn, the 89Zr elution profile was monitored by the in-line gamma detector.

As herein described, a primary column (MP-1M) eluent solution of 6 M HCl+0.33 mM NaF could be utilized as a transfer agent to a secondary microcolumn packed with hydroxamate resin. Additionally, it demonstrated how a solution of H2C2O4 could be utilized to efficiently elute Zr from the hydroxamate microcolumn in a volume (˜0.29 mL) that is significantly less than that specified for the traditional method (typically reported as ˜1 mL or greater). The next step was to develop a fully automated tandem column Zr separation that would combine the optimized method for an MP-1M column with the optimized hydroxamate microcolumn method as primary and secondary purification stages, respectively.

The hydroxamate column can be traditionally conditioned in 2 M HCl, after which the dissolved irradiated Y target (in ˜2 M HCl±H2O2) is loaded to adsorb the 89Zr. The column is subsequently washed with 2 M HCl to remove residual Y ions, and is then washed with H2O to remove residual HCl. The 89Zr is traditionally eluted with 1 M H2C2O4, and the first 1 mL eluent fraction is typically collected for subsequent radiolabeling.

Given the familiarity and reliability of this separation method, it was replicated to the extent possible in the latter steps of the MP-1M→hydroxamate resin tandem column separation scheme. Table 9 summarizes the tandem column separation protocol that was employed using the fluidic system shown in FIG. 16 (Right). Steps 1-4 largely cover the primary column (C1) separation of 89Zr from the dissolved Y target, as described previously. Step 4 (column transfer) coincides with the initiation of the secondary column separation, as the 89Zr eluent from the primary MP-1M column is synonymous with the load of the 89Zr onto the secondary microcolumn. The column transfer step is made possible via the fluidic coupling of the primary column effluent with the secondary column influent by setting V2 to position 2 and delivering the transfer solution with SP1.

With reference to Table 10 and FIG. 17, the hydroxamate resin microcolumn (C2), which was already conditioned using 2 M HCl during step 1, is capable of receiving the 89Zr in 6 M HCl (Kd˜11,600 mL/g, Table 10) or in 6 M HCl+0.33 mM NaF solution (Kd˜9000 mL/g, FIG. 17). Following the column transfer step, the C2 is washed with 2 M HCl (using SP2) to remove any trace contaminants that may have bled from the primary column. The column is then washed with H2O to remove HCl from the column and its connecting lines. After purging the column influent/effluent lines with air, the 89Zr is eluted from the column with 6 M HCl+0.22 mM NaF solution, through an on-line detector, and finally to 89Zr product collection vials using 2.5 mL of H2C2O4. This is accomplished by toggling V3 to position 2. The first 0.25-0.3 mL of column effluent was collected as the 89Zr product fraction; the remaining elution volume was collected in a separate vessel in order to determine total 89Zr elution recovery.

TABLE 10 Distribution coefficient (Kd) values determined using 88Zr tracer on hydroxamate resin across a range of HCl concentrations. [HCl], Kd, M mL/g n 1   5800 1 2 12,300 1 3 11,300 1 6 11,600 ± 2200 4 7.5 11,000 ± 100  2 9 14,700 1 12   3000 1

On-line detector traces for a tandem column separation of 89Zr from an irradiated Y foil target are shown in FIG. 18. Numbers on the detector traces correspond to the steps listed in Table 9. For this run, the column effluent lines were reconfigured so that the entire separation process could be monitored by the on-line detector. Because of the vastly higher activity levels in the irradiated foil vs. the 88Y and 88Zr tracer studies, the detector was raised ˜15 cm above the detection flow cell. The detector trace for the dissolved target load on the primary column is shown in FIG. 18A. The irradiated Y foil contained 2.11 mCi 89Zr (EOB). The foil was dissolved in 20 mL conc. HCl+0.1 M H2O2 and was loaded onto the primary column at a flow rate of ˜0.6 mL/min. FIG. 18B shows the on-line detector trace for steps 3-7 in Table 9. The positive detector responses across steps 2-6 are not due to 89Zr. Rather, these signals are due to the unretained and short-lived (15.66) 89mY that is generated by the 89Zr parent. An explanation of 89mY behavior during the 89Zr purification process on a 100 mg hydroxamate column has been given previously. Finally, the 89Zr elution is shown at the end of the trace (7).

A total of seven cyclotron bombarded yttrium metal foils were cyclotron irradiated and processed as per the described tandem column method. Table 11 provides a summary of the 89Zr foil activities (at EOB) and masses, as well as the chemical separation performance of 89Zr from the dissolved foils.

TABLE 11 89Zr separation performance for MP-1M → Hydroxamate tandem column method across seven irradiated Y foil targets. 89Zr Foil 89Zr Distribution from Tandem Columns, % a Activity, Mass, C1 C1 C1→C2 C2 C2 Foil ID mCi mg Load Wash Transfer Wash b Elute c Strip d 1 2.11 251 0.2 0.0 0.0 0.0 97.5 2.3 2 4.33 240 0.4 0.0 3.7 0.6 94.5 0.7 3 1.43 252 0.9 0.0 1.6 0.3 96.2 1.0 4 1.80 224 0.1 0.0 0.1 0.1 98.3 1.4 5 0.72 250 0.3 0.0 0.1 0.0 96.8 2.0 6 5.61 175 0.1 0.0 0.8 0.2 98.3 0.6 7 5.48 179 0.2 0.0 1.4 1.2 96.0 1.2 Average ± 1σ 224 ± 34 0.3 ± 0.3 0.0 ± 0.0 1.1 ± 1.3 0.3 ± 0.4 96.8 ± 1.4 1.3 ± 0.6 a Total activity recovered from the column separation b Cumulative activity fraction across 2M HCl and H2O washes c 89Zr recovery from entire 2.5 mL elution volume d Strip solution was 3M HCl + 0.1M NaF, which assured complete removal of any 89Zr remaining on the column

Irradiated foils 1-5 were originally between 78 and 180 mg, but the mass was augmented with the addition of non-irradiated Y foil to ˜250 mg in order to assure that the hydroxamate microcolumn performance was robust. Foils 6 and 7 received no additional Y foil. Irradiated Y foils were dissolved with a solution of 20 mL conc. HCl+0.1 M H2O2. For each of the seven 89Zr purifications, all column effluents were collected in order to determine the fraction of 89Zr activity at each step. The hydroxamate microcolumn elute fraction contained 96.8±1.4% of the column-recovered activity. However, some 89Zr activity was lost to the system, either in the form of residues remaining from the foil dissolution (activity remaining in the reaction flask and in-line membrane filter) or as activity trapped in the two columns. Residual 89Zr activity was calculated to be 1.9±0.4% of the original foil activity. Considering the 89Zr activity lost to the system, the total 89Zr yield in the elution fractions was 95.1±1.3% of the total 89Zr foil activity.

The elution profiles for 89Zr from the hydroxamate microcolumn are shown for four replicate runs in FIG. 19. The vertical scale is normalized to the peak maximum of each elution profile. Zero volume is set at the point at which the syringe pump begins delivering eluent solution to the reagent delivery line, which is purged with air prior to this step. The elution peak begins entering the detector after 0.536±0.002 mL has been delivered to the system. The peak maximum is obtained after 0.045±0.005 mL; full peak width at half the peak maximum (FWHM) is 0.12±0.03 mL. Ninety percent and 95% of the elution activity is obtained after 0.29±0.06 mL and 0.50±0.10 mL, respectively. Given the above result of 95.1±1.3% of the 89Zr found in the total 2.5 mL elution volume with respect to the activity in the irradiated foil, a 0.29±0.06 mL product volume (90% of elution peak) would be expected to yield 86±5% of the activity in the foil. Accordingly, 89Zr compositions are provided having 86±5% of the activity in the irradiated foil.

In order to establish a purity measurement specific for 89Zr compositions analysis methods were developed. For example, Deferoxamine is a bacterial siderophore that is the principal conjugate used to radiolabel proteins with 89Zr in immunoPET applications. However, metal contaminants that may persist in the 89Zr product fraction following hydroxamate column purification from the irradiated foil may decrease 89Zr binding yields. These metals include Zr, Fe, Ga, Al, etc. In general, the concentration of these potentially interfering metals in the 89Zr product fractions are dependent on the impurity levels in the Y target metal, the target mass required for irradiation, and the decontamination factors made possible by the chemical purification method employed. Commercially available deferoxamine mesylate (Dfo-m) was used to assess the quality of the 89Zr product fractions that were collected from the automated hydroxamate resin column purification method described above.

The following reagents were used for 89Zr labeling of deferoxamine mesylate (Dfo-m): Deferoxamine mesylate salt was obtained from Santa Cruz Biochemicals (≥98%) and Sigma-Aldrich (99.0%). Dimethyl sulfoxide (DMSO, ACS Reagent ≥99.9%), sodium carbonate (TraceSELECT®), sodium chloride (TraceSELECT®), and 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES, 99.5%) were supplied by Sigma-Aldrich. 2,5-Dihydroxybenzoic acid (gentisic acid, 99%) was supplied by Acros Organics (New Jersey).

Parameters that were evaluated in this assessment included bindable metal concentration ([MB]) and effective specific activity (ESA). Two 89Zr/Dfo-m binding solution matrices were used to determine these parameters for 89Zr product fractions from each of three irradiated Y targets purified by the method described above.

TABLE 12 Chemical compositions used in 89Zr/Dfo-m binding studies. Preparation Volume, Method Tube Reagent μL 1 1 89Zr Product Mix a 40-80  0.2-24 μM Dfo-m 3-60 Total Vol. 43-140 2 1 89Zr Product b 10   Buffer mix 1 c 19.5 2 Buffer mix 2 d 0.6-70.5 Buffer mix 3 e 0-69.9  Total Vol. 100 a 89Zr product in 0.8M H2C2O4 diluted with 0.2M Na2CO3 and H2O b 89Zr product in 0.8M H2C2O4 c Buffer mix 1 = 0.37M Na2CO3; 0.36M HEPES d Buffer mix 2 = 0.15M NaCl; 0.25M HEPES; 0.014M gentisic acid; 8.37 μM Dfo-m; 0.77M DMSO e Buffer mix 3 = 0.15M NaCl; 0.25M HEPES; 0.014M gentisic acid

Method 1 employed a simple solution preparation that consisted of only H2C2O4, Na2CO3 and H2O in an attempt to minimize the addition of competitive metals from impurities in reagents (Table 12); Method 2 used a more complex solution mixture that incorporated saline, buffers (Na2CO3 and HEPES), and gentisic acid (an anti-oxidant), similar to that reported by Verel et al. Both methods resulted in solutions with a pH of ˜7.

Following solution mixing and incubation, the 89Zr binding fraction was determined with the use of an aminopropyl-based anion exchange cartridge. Zirconium-89 oxalate (likely 89Zr (C2O4)4−4) that was successfully transchelated to Dfo-m (89Zr/Dfo-m) was not retained on the resin, while non-transchelated 89Zr(C2O4)4−4 was retained. The 89Zr/Dfo-m binding fraction was plotted as a function of added Dfo-m concentration, and a sigmoidal curve was fitted (FIG. 20). The curve was then used to identify the Dfo-m concentration that corresponded to 50% 89Zr binding, and the bindable metal (MB) content and effective specific activity (ESA) were subsequently calculated. In order to more easily compare the MB content between Y foil targets of different masses (and, hence, different levels of metal impurities introduced into the purification process), the MB concentration ([MB]) was calculated (where [MB]=MB/g Y).

Table 13 provides details of the Dfo-m transchelation experiments for Methods 1 and 2 using 89Zr product fractions obtained from Foils 3-5 using the hydroxamate resin column-based 89Zr purification method.

TABLE 13 Labeling performance of 89Zr products evaluated by DFOM transchelation using two solution preparation methods. 89Zr product Final Foil vol. fraction, [H2C2O4], [MB], ESA ID Method v/vtot, % mM nmole · g−1 Ci · mmole−1 a TBq · mmole−1 a 3 1 3.7 29 b 75.9   1200 44.24 2 10 80   36.5   2494 92.26 4 1 1.3 11 c 110     1358 50.23 3.5 83 b 130     1144 42.31 2 10 80   41.1   3625 134.1 5 1 1.3 11 c 95.0   1447 53.55 1.9 45 b 141 d   973 35.98 3.5 83 b 140     982 36.32 2 10 80   508 d   2705 100.1 Method 1 mean (±1s) e 115 ± 27 1180 ± 190 44 ± 7 Method 2 mean (±1s) e 43 ± 7 2940 ± 600 109 ± 22 a Based on 22.7 mCi (841 MBq) 89Zr produced in a 1 h target irradiation. b Na2CO3 and H2C2O4 concentrations were equimolar. c Na2CO3 molarity twice the H2C2O4 molarity. d 89Zr/DFOM binding fraction plots shown in FIG. 20. e Where “s” is the sample standard deviation.

Method 1 employed a range of 89Zr product volume fractions ranging from 1.3 to 3.7%; H2C2O4 and Na2CO3 additions varied, but their concentrations were equimolar so that complete solution neutralization was achieved, and solutions were diluted with H2O. Whereas a single 89Zr/Dfo-m transchelation curve was generated for Foil 3, duplicate and triplicate curves of different transchelation solution mixtures were generated for Foils 4 and 5, respectively. The mean [MB] value was determined to be 115±27 nmole·g−1, with an ESA of 1180±190 Ci·mmole−1.

In contrast, Method 2 employed a consistent 10% volume fraction between 89Zr product and the buffered Dfo-m transchelation mixture. The only H2C2O4 added to the experiment was from the 89Zr product, which resulted in a final H2C2O4 concentration of 80 mM. Across the three foils evaluated, Method 2 provided consistent [MB] values averaging 43±7 nmole·g−1. ESA was calculated as 2940±600 Ci·mmole−1. The [MB] and ESA results for Methods 1 and 2 were within a factor of ˜2.5 of each other, thus providing good agreement.

While the ESA term is used ubiquitously in reporting medical isotope quality, it cannot be used to compare isotope products of different activity and/or originating target mass; it is simply the ratio of isotope activity to bindable metals content (MB). The term [MB] expresses the concentration of metals present in the isotope product fraction that compete for binding to the labeling chelate (MB), normalized to the mass of target material from which the isotope is produced. Isotope activity is not considered, because the radioactive isotope produced within the target (89Zr=25 nmol/Ci) constitutes an insignificant portion of the competing metal presence (e.g., elemental Zr, Fe, etc.) within the purified product. Consideration of target mass is important, as the target metal (in this case, Y foil) is likely the greatest source of metal impurities introduced into the separation process. Since any given chemical purification process is capable of a finite decontamination of non-Zr metal impurities from the target metal, the use of a lower-mass target should yield a lower MB. Additionally, a lower-mass target will contain less elemental Zr contamination in the 89Zr product, which should likewise assure a lower MB. Normalizing MB to the mass of target from which the isotope was produced therefore negates target mass differences and allows for a direct comparison of a purification method's efficacy. This assumes, of course, that the metal contaminant concentrations in high purity Y targets are approximately consistent.

Performance of the tandem column separation method was demonstrated by processing seven cyclotron irradiated Y foil targets. The 89Zr elution from the secondary column represented 96.8±1.4% of the total activity collected from the dual column process, and 95.1±1.3% of the total activity present in the irradiated foil. The evaluation of replicate 89Zr elution traces lead to the determination that 90% of the elution activity is found in the first 0.29±0.06 mL of H2C2O4 eluent solution; this elution fraction represents 86±5% of the activity in the irradiated Y foil. This represents an 89Zr product that is times more concentrated than that reported by this team and others when performing the traditional single hydroxamate column method. The combination of an increased 89Zr product concentration (reduced product volume) and reduced H2C2O4 concentration should result in improvements in the labeling performance of 89Zr onto conjugated mAbs.

In accordance with the materials and methods described herein, 89Zr compositions are provided that can have bindable metal concentration ([MB]) and effective specific activity (ESA), respectively of less than 169 nmole·g−1 and/or greater than 800 Ci·mmole−1.

In accordance with example implementations, these compositions can be used as starting materials for methods for radio labeling monoclonal antibodies, other proteins, aptamers, etc.

Methods are also provided for binding 89Zr to a macromolecule. The methods can include binding the 89Zr to a nanoparticle and then subsequently binding the 89Zr-nanoparticle to a macromolecule such as protein including immunoglobulin(s). Accordingly, isotope compositions are provided that can include a radio isotope and a nanoparticle and/or a macromolecule. In accordance with example implementations, the isotope composition of claim 25 wherein the radio isotope is one or more of 89Zr, 86Y, and/or 90Y.

In accordance with example implementations, batch contact studies were performed to evaluate the uptake of both 88Zr and 88Y (both in the chloride form) onto magnetic iron oxide nanoparticles. Chemically unmodified, commercially available Fe3O4 (“FeO”, ˜27nm) and in-house synthesized manganese-doped Fe3O4 (“MnFeO”), were used. The distribution coefficient values are shown in FIG. 21 (88Zr tracer) and FIG. 22 (88Y tracer) across a range of dilute HCl conditions and in pH-adjusted physiological saline. The results show that the MnFeO has improved 88Zr tracer uptake in HCl when compared to FeO. However, FeO is equally effective at 88Zr tracer retention in the higher ionic strength physiological saline matrix. For yttrium (as 88Y tracer), the overall uptake is lower for FeO and MnFeO in HCl matrix when compared to uptake measured for 88Zr at HCl concentrations above 0.001M. A strong pH dependence is shown for 88Y tracer in physiological saline, with Kd values at near-neutral saline being very high.

The nanoparticle uptake of highly purified 89Zr was obtained directly from the 0.8 M oxalic acid elution matrix. Each of these methods result in 89Zr product fractions in the oxalate form (likely as Zr(C2O4)4−4). Initial experiments to determine the Kd of 88Zr tracer onto the magnetic nanoparticles from oxalic acid resulted in low Kd's (on the order of ˜100 mL/g). This distribution coefficient is insufficient for effective isotope loading. Next, the Kd's of 88Zr oxalate on the magnetic nanoparticles were determined after the solution was buffered using physiological saline and HEPES buffer. An increase in 88Zr tracer binding was observed, with Kd values up to 4700 and 46,000 mL/g for FeO and MnFeO, respectively. This represented a substantial improvement in uptake. However, the Kd values were still substantially lower than those observed in the ZrCl4 system. Therefore, the team evaluated a method to convert the 89Zr oxalate product to the chloride form.

In order to obtain the very high Kd values for Zr on the magnetic nanoparticles, as shown in FIG. 21, a method to convert Zr oxalate isotope product to Zr chloride can be utilized. Zirconium oxalate is known to bind strongly to anion exchange (An IX) resins due to its -4 charge. One approach was to load Zr(C2O4)4−4 onto a column, wash excess oxalate from the column, and then elute with HCl to obtain ZrCl4. Both AG MP-1M and QAE-550C resins demonstrated high uptake of Zr(C2O4)4−4. Unfortunately, HCl concentrations of >1 M were required to elute the Zr. The magnetic nanoparticles are incompatible with high acid concentrations, as they will dissolve. One option was to evaporate the acid from the ZrCl4 eluent solution and then dissolve the ZrCl4 in a more benign solution. However, there was a risk of hydrolyzing part or all the Zr to the zirconyl (ZrO−2) ion during this process.

A modified method was devised to elute Zr from the AnIX column using HCl concentrations as low as 0.15 M HCl. This was accomplished by inserting an intermediate step in the AnIX separation process. Zr(C2O4)4−4 was loaded onto a micro-column (27 μL column) packed with a strongly basic anion exchange resin (QAE-550C) that had been previously converted to the oxalate form). Following the load, the column was washed with clean 0.8 M H2C2O4, and then the column was converted to the formate form via the addition of formic acid. Zirconium has a high level of retention on the AnIX column in formate form, so it remains fixed on the column during this step.

The formate anion (HCO2−) has a low “relative selectivity coefficient” on the AnIX resin. For MP-1M resin, formate has a selectivity coefficient of 4.6. This compares to a value of 22 for the chloride anion. An oxalate value is not specified directly in the literature, but it is estimated to be very high—likely somewhere below citrate (220) and above iodide (175). By converting the column from oxalate form (˜200) to formate form (4.6), the Zr can be eluted from the AnIX column easily using moderate concentrations of Cl− ion.

FIG. 23 demonstrates the elution of 88Zr tracer using HCl between 0.15 and 0.5 M HCl after performing the formate column conversion on QAE-550C strongly basic AnIX resin. Note that when not using the formate conversion step, Zr would not elute from the column unless HCl concentrations of 1 M HCl or greater were utilized. Additionally, even with high HCl eluent concentrations, the Zr elution volumes were significantly greater than those shown in FIG. 23. The formate conversion step resulted in the ability to effectively elute Zr (chloride form) in low volumes (˜0.3 to 0.7 mL) of relatively dilute HCl. The addition of a moderate amount of buffering agent to the Zr chloride eluent would assure that dissolution of the iron oxide-based magnetic nanoparticles (MNPs) would not occur, or that damage to labeling proteins would not occur, for example.

Given the new column method to produce ZrCl4 from Zr(C2O4)4−4, efficient labeling of the magnetic nanoparticles was now possible. The above chemical method was used to convert purified 88Zr tracer from the oxalate form to the chloride form as a necessary step in order to successfully label the isotope onto the magnetic nanoparticles. 88Zr tracer was eluted with HCl concentrations of 0.15 M and 0.3 M. Additionally, the 88Zr tracer was eluted with 0.15 M and 0.3 M HCl that was diluted 1:1 with phosphate buffered saline (PBS). Prior to conducting batch contact experiments to determine the distribution coefficients across the various matrices, each of the four 88Zr tracer eluent solutions was diluted again into PBS. The final PBS dilution, resulting pH, and chloride concentration is indicated in Table 14.

TABLE 14 Measured distribution coefficients (Kds) for Zr (as 88ZrCl4) loaded onto Fe3O4 (FeO) and Mn-doped Fe3O4 (MnFeO) magnetic nanoparticles from a range of HCl concentrations diluted into phosphate buffered saline (PBS). Zr Eluent Log Kd, mL/g Dilution [Cl−], (% Loading) Trial ID Zr Eluent into PBS, v/v pH M FeO MnFeO 1 0.3M HCl 1:2 1.03 0.194 4.2 5.2 (88) (99) 2 0.15M HCl 1:2 1.35 0.144 4.9 5.2 (95) (98) 3  0.3M HCl + PBS a 1:2 1.37 0.194 4.2 5.2 (90) (99) 4 0.15M HCl + PBS a 1:2 1.75 0.144 5.2 5.3 (99) (99) 5 0.15M HCl + PBS a 1:3 1.91 0.143 5.2 5.3 (99) (99) 6 0.15M HCl + PBS a 1:4 1.96 0.143 5.1 5.5 (99) (99) a HCl solution diluted 1:1 into PBS prior to column elution.

Measured distribution coefficient values are presented as Log Kd; percent 88Zr tracer uptake values are listed in parentheses.

Within the solutions, which ranged between pH 1 and pH 2, the FeO and MnFeO nanoparticles were observed to be relatively stable over the contact period. For MnFeO MNPs, a 1 h contact time resulted in <1% Fe mass solubilization and ˜17% Mn mass solubilization fractions. Despite the loss of Mn from the MnFeO MNP media, however, the material was still capable of analyte sorption.

Distribution coefficient values for both nanoparticle types were very high, with Kd's typically measured at levels exceeding 10,000 mL/g for FeO and exceeding 100,000 mL/g for MnFeO. The Kd values are in good agreement with those observed in a matrix of pH-controlled physiological saline (FIG. 21, Right). This study indicates that the Zr oxalate-to-chloride conversion method is effective at removing oxalate from the Zr eluent, which results in dramatic increases in Zr retention on the MNPs.

The superparamagnetic metal oxides Fe3O4 and Mn-doped Fe3O4 were shown to exhibit high affinities for Zr from dilute HCl solutions and solutions of physiological saline across a pH range. Measured affinities in oxalic acid were low, but were improved with the addition of buffering agents.

In accordance with example implementations, wherein 89Zr is eluted in an oxalate form, for example from a hydroxamate resin alone or from a hydroxamate resin as part of the dual column system described herein, it may be converted to a chloride form. A column method was developed in order to convert the 89Zr oxalate product fraction to 89ZrCl4, using formic acid to enable efficient 89Zr elution from the column using a low concentration of HCl. This resulting eluent solution, diluted into PBS, demonstrated that 89Zr Kd values were high for FeO (10,000 to >100,000 mL/g) and very high for MnFeO (>100,000 mL/g).

In compliance with the statute, embodiments of the invention have been described in language more or less specific as to structural and methodical features. It is to be understood, however, that the entire invention is not limited to the specific features and/or embodiments shown and/or described, since the disclosed embodiments comprise forms of putting the invention into effect. The invention is, therefore, claimed in any of its forms or modifications within the proper scope of the appended claims appropriately interpreted in accordance with the doctrine of equivalents.

Claims

1. A method for radio labeling monoclonal antibodies, the method comprising preparing a binding conjugate by exposing a ligand to a solution of 89Zr and less than 1 M oxalic acid.

2. The method of claim 1 wherein the solution comprises from 0.3 M to less than 1 M oxalic acid.

3. The method of claim 1 wherein the solution further comprises fluoride.

4. The method of claim 3 wherein the solution further comprises NaF.

5. The method of claim 1 wherein the conjugate comprises mAbs.

6. The method of claim 5 wherein the conjugate comprises deferoxamine.

Patent History
Publication number: 20210386878
Type: Application
Filed: Aug 30, 2021
Publication Date: Dec 16, 2021
Applicant: Battelle Memorial Institute (Richland, WA)
Inventor: Matthew J. O'Hara (Richland, WA)
Application Number: 17/460,986
Classifications
International Classification: A61K 51/10 (20060101); B01D 59/30 (20060101); G21G 1/00 (20060101); C22B 3/20 (20060101); G21G 4/08 (20060101); C01G 25/00 (20060101);