52mMn generator

Appl. Radiat. lsot. Vol. 48, No. 8, pp. 1097-1101, 1997

Pergamon PII:

S0969-8043(97)00106-1

,c~, 1997 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0969-8043/97 $17.00 + 0.00

Reinvestigation of a Physiological Eluate of the 52Fe/52mMnGenerator PETER BL,AUENSTEIN, RAIMO PELLIKKA a n d P. A U G U S T S C H U B I G E R Division of Radiopharmacy, Paul Scherrer Institute, 5232, Villigen, PSI, Switzerland (Received 25 November 1996)

We have achieved a significant step forward in the potential application of 52mMn-'+ (T~ = 0.35 h, 13+ = 97%) as a myocardial imaging agent with positron emission tomography (PET) by the introduction of a 5% (physiological) glucose solution as an eluent for the 5-'Fe/5-'mMngenerator. Our experiments have demonstrated the favourable properties of a glucose solution with minimal breakthrough ( < 0.3%) of ~-'Fe and yields of up to 90% 5-'mMn"+.Although it has been shown that lower 52Fe breakthrough is attainable using other eluents, due to the short half life of 52Fe (8.27 h) breakthrough up to 1% would not appear to significantly alter the efficacy of the 52mMneluted with this 5% glucose solution. The primary advantage of this approach lies in its convenience of application, in that a 5% glucose solution may be administered directly into patients thereby circumventing the major problem of non-injectable eluates previously associated with this generator. @ 1997 Elsevier Science Ltd

Introduction Generators are an important source of diagnostic radionuclides for application in nuclear medicine (Lambrecht, 1983; Knapp and Mirzadeh, 1994). One such radionuclide, manganese-52m (T~,2=0.35 h, 13÷ = 97%), as the cation 52mMn2* is considered a suitable candidate for myocardial imaging with positron emission tomography (PET) (Atkins et aL, 1979), based on earlier tissue distribution studies with ~Mn 2÷ (Chauncey et al., 1977). A number of procedures aimed at producing ~2mMn2÷ have been attempted. For example, Daube and Nickles (1985) described direct isolation of 52mMn2+ following irradiation of a chromium target according to the reaction 52Cr(p,n)5~mMn. This process, however, requires a complicated chemical separation procedure in close proximity to the PET scanner and is unable to provide for more than a single study at any one time. More useful approaches would therefore appear to involve elution of s2mMn2+ as a generator product enabling a number of studies to be performed in the course of a day (Atcher et al., 1980; Herscheid et al., 1983; Lambrecht, 1983). Despite the favourable physical properties of 52mMn2+ as an agent for myocardial imaging studies, it has received only limited attention when compared to the more commonly used PET tracers 23N-ammonia, ~50-water, S2Rb+ and 62Cu(PTSM) (PTSM = pyruvaldehyde-bis(N4methylthiosemicarbazone). The reasons for the apparent lack of interest in ~"mMn2÷ as a PET tracer

are unclear, but may lie in the different methodology required in the preparation of the respective radionuclides. For example, where a cyclotron and PET scanner exist within the same facility, the short-lived radionuclides ~3N and ~O can be produced and incorporated into ammonia or water, respectively, these compounds being considered the optimal tracers for determining regional cardiac blood flow (Huang et al., 1985; Hutchins et al., 1990). The other nuclides 8:Rb, 62Cu and 52mMn are obtained from a generator and can thus be used in a satellite PET facility. To produce the parent nuclides, however, a cyclotron is needed that provides a high current beam of protons with an energy of more than 45 MeV in the case of 5~Fe and 82Sr and 15 to 45 MeV for 62Zn (Pellikka et al., 1992, Zweit et al., 1992, Robinson et al., 1979). The practical use of the generators is dependent on the half lives of the mother and daughter nuclides (Table 1). Due to the long half life of 82Sr (25.5 d), prolonged and expensive irradiation is needed, although the generator can be used for several weeks. In contrast, the production of 52Fe or 6:Zn (half lives ~ 0 . 3 d) requires only a short irradiation time and use of the generator is restricted to only about 2 d. Thus, the practical application of these particular generators may ultimately depend upon differences in their production costs and useful working life. The most important aspects of a generator system are considered to be the simplicity of the elution procedure and the suitability of the eluate for

1097

1098

Peter Bl~iuenstein et al.

preparation and injection into patients. In an attempt to circumvent potential problems with earlier '-'mMn generators which required elution with 6-9 molar solutions of hydrochloric acid (Atcher et al., 1980), Herscheid et al. (1983) prepared a hydroxamate resin which they eluted with acetate buffer or physiological saline. Despite promising results, this generator did not receive wider application, perhaps due to the limited availability of the hydroxamate resin. More recently, the use of a tartaric acid solution in a Dowex 1 × 8 column has been proposed as eluent for the 5-'Fe/~-'mMngenerator (Schwarzbach et al., 1991). This approach is preferable to earlier methods because of the commercial availability of the solid phase and the superiority of the solvent to hydrochloric acid. Tartaric acid solutions contain salt concentrations lower than physiological levels, thereby enabling the addition of isotonic amounts of sodium chloride without the need for solvent evaporation. Despite this significant improvement in the preparation of the 5-'Fe/~"mMn generator, additional manipulation of the eluate is still required before application in the clinical situation. In comparison with other generator systems, the '-'Fe/'-'mMn generator does not seem to possess particular disadvantages. For example, the use of the 62Zn/62Cu generator involves a number of preparative steps prior to obtaining an injectable solution. These steps using strongly acidic, hypertonic solutions of either 0.1 M hydrochloric acid containing 100 mg/ mL sodium chloride (Robinson et al., 1979) or 2 M hydrochloric acid (Green et al., 1990) have resulted in only moderate yields following PTSM labelling. In a manner analogous to the ~-'Fe/5-'"lMn generator, simplification of this elution process is underway using an eluent of 0.3 M hydrochloric acid in 40% ethanol and permitting high product yields following direct PTSM labelling (Zweit et al., 1992). Elution of the s-'Sr/SZRbgenerator, on the other hand, appears to be a simpler procedure which can be performed using physiological saline. The short half life of the S2Rb, however, requires immediate administration of the activity after elution. Thus a special device has been developed which allows on line elution, measurement of the activity and infusion into a patient (Brihaye et al., 1987). Although directly applicable to patients, a number of concerns regarding the infusion pressure and total injected activity administered during this procedure remain. With a further simplification of the 52Fe/5-'mMn generator procedure as our principal aim, the goal of the present study was to identify both a readily available material to fill the generator column and a physiological eluate for direct administration to

patients. We investigated the potential of the commercially available anion exchange resin Dowex 1 × 8 and compared as eluent a 5% glucose solution against the previously used 2 mM tartaric acid and distilled water.

Experimental Material Jbr target and generator chemistry

The target consisted of natural high purity nickel containing 1 ppm iron (Goodfellow Metals Ltd., Cambridge, U.K.). The 1 mm thick plate was cut into squares of 18 × 18 mm each, weighing 2.8 g. Nitric acid, hydrochloric acid, cobalt(II) chloride, nickel(I1) chloride and potassium-sodium-tartrate were obtained from Merck (Germany) in the highest available purity grade, L( + )-tartaric acid was from Fluka (Switzerland). Dowex A G 1 × 8, 200-400 mesh anionic exchange resin (Cl-form) used for the separation of the iron from the impurities was obtained from Bio-Rad (Switzerland). Preparation of the parent nuclide, Fe-52

The 1 mm thick nickel target was irradiated with protons (68-61 MeV) for up to 11 h with a beam current of 70 laA producing 5-'Fe according to the nuclear reactions 5SNi(p,~p2n)52Fe and ~ONi(p,~p4n)5'Fe. Following proton bombardment the target was dissolved in 80 mL of 7 M nitric acid under strong heating and evaporated to dryness. The residue was dissolved in 6 N hydrochloric acid and loaded on a Dowex A G 1 × 8 column (0.5 × 5 cm). The column was rinsed thoroughly with 40 mL carrier solution consisting of 1 g nickel(II) chloride and 250 mg cobalt(II) chloride in 4 M hydrochloric acid. The remaining nickel and cobalt impurities were washed off with 140 mL of 4 M hydrochloric acid and the column run dry in order to remove as much of the hydrochloric acid as possible. In this manner the amount of water required to elute the iron from the resin was reduced to 8 mL. The remaining hydrochloric acid present on the resin was sufficient to prevent the precipitation of iron(IlI) hydroxide. An aliquot of 10 laL was withdrawn from the eluate and used for the determination of the yield and radionuclide purity of S:Fe, whilst the remainder was evaporated to dryness and reconstituted with 8 mL of 2 mM tartaric acid. Preparation o f the 52Fe/S:mMn generator

The anionic exchange resin Dowex A G 1 × 8 was also used for the generator column. It was converted to the tartaric form in a glass beaker with 50 mL of 0.1 M potassium--sodium-tartrate (0.1 M), decanted

Table I. Comparisonof the respectivehalf livesof the three differentradionuclidegeneratorsused for PET (cardiac)studies:~-'Sr/X~Rb.~:Zn/~:Cuand S:Fe/~2mMn(S-'Mn) Nuclide S2Sr ~-~Rb ~:Zn ~-'Cu ~2Fe S2mMn 52Mn Half life 25.5 d 1.27 mm 9.13 h 9.74 rain 8.27 h 21 min 5.6 d

Reinvestigation of a phpysiological eluate of the 52Fe/~-'~Mngenerator T

F

B

@

1099

using thioglycolate bouillon and trypticase soja bouillon from Becton Dickinson, Basle. The pyrogenicity (bacterial endotoxin) was tested using the LAL test (limulus polyphemus) from Whittaker, obtained from Serva, Heidelberg.

Results and Discussion 0

5 cm

Fig. 1. The ~-'Fe/5-'mMngenerator column (top) and the profile of the activity distribution (below) measured with PET. T: top of the column, F: glass frit, B bottom of the column. The shaded part represents the anion exchanger resin, the blank part is filled with glucose solution. The x-axis of the activity profile and the column are drawn to the same scale. The smoothed curve represents acquired counts per pixel in a linear scale in the region of interest drawn over the whole column. and rinsed twice with 50 mL distilled water to remove excess tartrate prior to loading the column with iron tartrate complex. The generator column consists of a glass tube with a glass frit at the lower end. The upper part was filled with Dowex 1 x 8 (2 mL bed volume, 8 mm diameter, 40 mm length), then both ends of the tube were closed with rubber stoppers and aluminium caps (see Fig. 1) equipped with luer needles. The column was washed with a mixture of 70% ethanol and 30% water for sterilisation purposes, then with pure water to eliminate the alcohol. The 52Fe-tartrate solution was loaded onto the column at 4 mL/min and the column washed with 100 mL 5% glucose solution. The column was eluted 1 h later for quality control. Elution

The generators were eluted for clinical trials at time intervals of around 90 min, when the build-up of 52mMnTM was close to the maximum activity (calculated build-up and decay curves published by Herscheid et al. (1983)). Pre-evacuated sterile I0 mL glass vials connected with a sterile filter were used to collect the eluate. The elution profile of the generator was recorded using a peristaltic pump dispensing 0.5 mL fractions at 4 mL/min. The 5:mMn activity of each fraction was measured in a Capintec CR-100 dose calibrator, breakthrough of 52Fe and other possible radioactive impurities being measured 2 to 3 h after elution using a high purity Ge-detector connected to a multichannel analyser. For comparison, the elution of the generator was performed using pure water, 2 mM tartaric acid solution or 5% glucose solution; however, for clinical studies only the glucose solution was used. Pharmaceutical quafity

The sterility was tested according to standard procedures described in the Pharmacopoea Helv. VII

The average yield of 52Fe achieved by proton irradiation was 6.4 _ 1.4 MBq/IaA h at the end of bombardment (EOB). We obtained approximately 10% less radioactivity than had been calculated from the results of Steyn et al. (1988) who irradiated 2 mm thick targets with a beam current of 50 rtA for a period of 1 h. In the present experiments, following correction for the radioactive decay because of the longer irradiation period and thinner target material used, the lower yield and higher variation could be partly explained by the losses which occurred during the purification process and by the precision of radioactivity measurement using a 10 p.L aliquot from an 8 mL sample. However, we cannot explain the 25% lower yield compared to the results of Smith-Jones et al. (1990) using similar experimental conditions. The reasons for this large discrepancy remain unclear. The 55Fe activity was determined after the decay of the other iron nuclides using a much simpler apparatus than Steyn et al. (1988). We obtained very low values which may be due to some absorption processes on the glass surface of the vials during storage or due to an inappropriate measuring technique. The dosimetry calculations were therefore based on the results of Steyn et al. (1988). Purification of the 52Fe by rinsing the column with acid solutions containing inactive Co 2+ and Ni 2+ salts reduced the levels of the respective contaminating radioisotopes to trace amounts (Table 2). Further reduction of these levels occurred during the preparation of the generator as Fe is strongly bound while Co, Ni and Mn are washed off the column. Thus the levels of contaminating radioactive Co and Ni on the generator were always below the limits of detection. Depending on the duration of irradiation, the generators were prepared using a 52Fe activity range of between 400 and 2200 MBq. This variation in the 52Fe activity did not appear to significantly affect the loading yield of the generator which was routinely greater than 95%. The initial elution experiments were performed using tartaric acid according to the method of Schwarzbach et al. (1991). Using this approach we had expected to observe a stabilisation of the iron-tartrate complex because of the increased Table 2. Relative radionuclidecomposition of the purified 52Fe solution calculatedat calibrationtime (8-9 h after EOB) Nuclide S2Fe 59Fe ~Co 57Co ~SCo ~Ni Relative activity 1000 0.8 0.006 0.003 0.003 0.004

1100

Peter Bl~.uensteinet al.

Table 3. Yield and breakthrough in 10 mL eluates of Fe-Mn generators. The 5-'Fepresent on the column at the time ofelution was set to 100%, the yield and breakthrough values are calculated relative to this value

Yield Breakthrough Number of eluates

Water

Glucose

Tartrate

80 + 19 0.036 + 0.022 14

90 + 17 0.144 + 0.15 13

101 + 6 0,74 + 0.q5 15

concentration of free ligand in the solution and, in accord with the Law of Mass Action, a stable binding of the iron activity on the anion exchange resin. The high and variable levels of 52Fe breakthrough (0.03-2.2%), however, precluded the routine use of tartaric acid as eluent. In particular, this approach often resulted in the generation of yields greater than 100% of eluted 52mMn,primarily due to such a high breakthrough of 5-~Fe. The reason for the high variation remained unclear. Elution with water decreased the 52Fe breakthrough to more acceptable levels (0.006-0.07%), but required the addition of sodium chloride to obtain a physiological solution and reduced the elution yield by some 20%. Elution with a 5% glucose solution was found to be the most suitable for direct patient application, 52Fe breakthrough being at acceptable levels (0.03-0.37%) despite being up to five times that observed when water was used as eluent (Table 3). The short shelf-life (2 d) of the generator usually allows 14 to 20 elutions (7-10/d) under routine conditions. After performing some 30 elutions on a single column using 10 mL of a 5% glucose solution, PET imaging was carried out in order to determine the distribution of the 52Fe activity (Fig. 1). Results showed that even after this unusually high number of elutions, the limits of the column capacity had not been reached and breakthrough of 52Fe would not be expected to be a significant problem under conditions of particularly heavy use. The first elution on the second day was found to give lower 5'mMn yields ( < 70%) than previous elutions as well as containing significantly higher quantities of 52Mn, and for these reasons was not used. The amount of ~-'Mn is proportional to the elapsed time between the last elution of the preceding day and the first elution on the second day. In addition, the reduced yield may be a result of the slow (overnight) radiolytic oxidation of Mn -'+ on the column leading to formation of insoluble MnO, which would be expected to be trapped on the column resin. If radiation dosimetry is a consideration, the

generator should be eluted twice in the morning of the second day in order to reduce the unwanted 52Mn to the normal acceptably low level. In the same respect the iron breakthrough of first day eluates does not contribute significantly to the radiation dose. However, this is no longer true the second day, that is, after about four to five half lives of ~2Fe. This reduces the activity of the mother nuclide (and hence the activity of 52"Mn) leading to the need to inject a larger portion of the eluate. The activity of 59Fe (T~, = 45 d) and ~SFe (T~2 = 2.7 d) remain virtually constant, which means that in relation to the injected dose the proportion of these nuclides increases. Based on calculations of Robertson et al. (1983) for iron isotopes and Atkins et al. (1979) for 52"Mn, we can estimate that the radiation dose to the spleen and the red bone marrow is increased by a factor of about five after applying an eluate from the late evening of the second day compared to any one eluate from the first day, while the dose to the other organs and specifically to the total body shows a rather small increase similar to the error of the dose estimation. To perform the dosimetric calculations of the second day we assumed a 55Fe content of 1.7% at EOB (Steyn et al., 1988), elution of the generator 48 h after EOB (six half lives of 5'Fe) and iron breakthrough of 0.3% (which is twice the average and the maximum ever obtained). If we assume a patient dose of 100 MBq of 5'mMn, the total dose is similar to that of 700 MBq 2-[~SF]fluoro-2-deoxy-D-glucose (['SF]FDG) (Jones et al., 1982; Jones et al., 1983), which is close to twice a usual patient dose of around 350 MBq ~8F (Table 4). Additionally it can be considered that 10 to 15 MBq S2Fe have already been used for patient studies (Roelcke et al., 1996), which is around 50 times more than might occur due to unwanted breakthrough. The elution profile, which gives an indication of the performance of a generator, showed a slightly higher peak and shorter tailing when the 5% glucose solution was used as eluent compared to pure water (Fig. 2). This profile is likely to be a reflection of the properties of glucose which is a weak ligand and acts as a weak reducing agent, and therefore it may be able to stabilise Mn 2÷ . An elution volume of about 5 - 1 0 m L was routinely used, however, under conditions where high concentrations of 52mMn2+ were required it was possible to fractionate the eluate and thereby to collect the majority of 5'mMn2+ activity in about 2-3 mL, which is also reflected by the elution profile.

Table 4. Radiation dostmetry of ~2mMnZ+ after injection of an eluate of the first day. an eluate of the evening of the second day and [~F]FDG for comparison. Patient doses of 100 MBq 52mMn"+ and 300 MBq [~F]FDG are assumed Organ Liver Spleen Kidney Red marrow Whole body

mGy/100 MBq ~:mMn"+ first day eluate 8 5 II 1 8

mGy/100 MBq 5-''Mn -'~ second day eluate I1 21 I1 5 9

mGy/300 MBq [~8F]FDG 4.7 I 1.6 5.7 3.4 3.2

Reinvestigation of a phpysiological eluate of the 5-'Fe/~:mMngenerator

40 -35 --

A

E 31)"7,

20-

0

1.0

2.0

3.0

4.0

5.0

Fraction [mL] Fig. 2. Radioactivity profile (ten fractions of 0.5 mL) of elutions with 5% glucose solution (0) and water (11). The radioactivity in the single fraction is given as a percentage of the total eluted activity. The bars represent the standard deviation (n = 3).

The sterility and pyrogenicity were never a problem, despite the fact that ion exchange resins and glucose solutions are good media for bacterial growth. All tested eluates were sterile and had endotoxin levels below the detection limit. This finding may be due to the sterile working conditions and a sterilising effect of the radiation o f the radionuclides on the column. The present study demonstrates a convenient and straightforward method to produce and elute the 52Fe/52mMn generator in a form suitable for direct administration to patients. This methodology circumvents the drawbacks of previously published approaches using concentrated hydrochloric acid solutions which require considerable manipulation prior to their clinical application.

References Atcher g. W., Friedman A. M., Huizenga J. R., gayudu G. V. S., Silverstein E. A. and Turner D. A. (1980) Manganese-52m, a new short-lived, generator-produced radionuclide: A potential tracer for positron tomography. J. Nucl. Med. 21, 565-569. Atkins H. L., Som P., Fairchild R. G., Hui J., Goldman A. and Ku T. (1979) Myocardial positron tomography with manganese-52m. Radiology 133, 769-774. Brihaye C., Guillaume M., O'Brien H. A. Jr., Raets D., de Landsheere Ch. and Rigo P. (1987) Preparation and evaluation of a hydrous tin(IV)oxide ~2Sr/~2Rb generator system for continuous elution. Int. J. Appl. Radiat. lsot. 38, 213-217. Chauncey D. M., Schelbert H. R., Halpern S. E., Delano F., McKegney L. M., Ashburn W. L. and Hagan P. L. (1977) Tissue distribution studies with radioactive manganese: A potential agent for myocardial imaging. J. Nucl. Med. 18, 933-936. Daube M. E. and Nickles R. J. (1985) Development of myocardial perfusion tracers for positron emission tomography. Int. J. Nucl. Med. Biol. 12, 303-314.

1101

Green M. A., Mathias J. C., Welch M. J., McGuire A. H., Perry D., Fernandez-Rubio F., Perlmutter J. S., Raichle M. and Bergmann S. R. (1990) Copper-62 labelled pyruvaldehyde-bis(N4-methylthiosemicarba zone)copper(ll): Synthesis and evaluation as a positron emission tomography tracer for cerebral and myocardial perfusion. J. Nucl. Med. 31, 1989-1996. Herscheid J. D. M., Vos C. M. and Hoekstra A. (1983) Manganese-52m for direct application: A new Fe-52/Mn52m generator based on a hydroxamate resin. Int. J. Appl. Radiat. Isot. 14, 883-886. Huang S. C., Schwaiger M., Carson R. E., Carson J., Hansen H.. Selin C., Hoffman E. J., MacDonald N., Schelbert H. R. and Phelps M. E. (1985) Quantitative measurement of myocardial blood flow with oxygen-15 water and positron computed tomography: An assessment of potential and problems. J. Nucl. Med. 26, 616-625. Hutchins G. D., Schwaiger M., Rosenspire K. C., Schelbert H. and Kuhl D. E. (1990) Noninvasive quantification of regional blood flow in the human heart using N-13 ammonia and dynamic positron emission tomographic imaging. J. Am. Coll. Cardiol. 15, 1032-1042. Jones S. C., Alavi A., Christman D. and Reivich M. (1982) The radiation dosimetry of 2[F-18]fluoro-2-deoxy-o-glucose in man. J. Nucl. Med. 23, 613-617. Jones S. C., Alavi A., Christman D. and Reivich M. (1983) Letters to the Editor, Reply concerning the above cited article of Jones et al. (1982). J. Nucl. Med. 24, 447-448. Knapp F. F. Jr. and Mirzadeh S. (1994) The continuing important role of radionuclide generator systems for nuclear medicine. Eur. J. Nucl. Med. 21, 1151-1165. Lambrecht R. M. (1983) Radionuclide generators. Radiochim. Acta 34, 9-24. Pellikka R., Huszar I., Schwarzbach R., Bl~iuenstein P. and Schubiger P. A. (1992) Iron-52 and strontium-82, two parent nuclides for generator systems produced at the 72 MeV cyclotron of the Paul Scherrer Institute (PSI). Radioact. Radiochem. 3, 59-60. Robertson J. S., Price R. R., Budinger T. F., Fairbanks V. F. and Pollycove M. (1983) Radiation absorbed doses from iron-52, iron-55, and iron-59 used to study ferrokinetics. J. Nucl. Med. 24, 339-348. Robinson G. D. Jr., Zielinski F. W. and Lee A. W. (1979) The zinc-62/copper-62 generator: A convenient source of copper-62 for radiopharmaceuticals. Int. J. Appl. Radiat. Isot. 31, 111-116. Roelcke U., Leenders K. L., von Ammon K., Radue E. W., Vontobel P., Guenther I. and Psylla M. (1996) Brain tumor iron uptake measured with positron emission tomography and 52Fe-citrate. J. Neuro-Oneol. 29, 157-165. Schwarzbach, R., Smith-Jones, P. M., Leenders, K. L., Weinreich, R., Mficke, M., Tschudin, P., Bl~iuenstein, P. and Schubiger, P. A. (1991) Production and use of S-'Fe. Proceedings of the 4th European Symposium on Radiopharmaceuticals, Baden, p. 106.

Smith-Jones P., Schwarzbach R. and Weinreich R. (1990) The production of S2Fe by means of a medium energy proton accelerator. Radioehim. Acta 50, 33-39. Steyn G. F., Mills S. J., Nortier B. R., Simpson B. R. S. and Meyer B. R. (1988) Production of 5-'Fevia proton-induced reactions on manganese and nickel. Appl. Radiat. lsot. 41, 315-325. Zweit J., Goodall R., Cox M., Babich J. W., Potter G. A., Sharma H. L. and Ott R. J. (1992) Development of a high performance zinc-62/copper-62 radionuclide generator for positron emission tomography. Eur. J. Nucl. Med. 19, 418-425.