AppL Radiat. 1sot. Vol. 37, No. 8, pp. 907-913, 1986 0883-2889•86 $3.00 + 0.00 Int. J. Radiat. AppL lnstrum, Part A Pergamon Journals Ltd Printed in Great Britain

Radioiodination via Isotope Exchange in Pivalic Acid

JAMEY P. WEICHERT, MARCIAN E, VAN DORT, MICHAEL P. GROZIAK, and RAYMOND E. COUNSELL*

Departments of Medicinal Chemistry and Pharmacology, The University of Michigan, Ann Arbor, MI 48109, U.S.A.

A variety of benzoic and aryl aliphatic mono and polyiodinated acids and esters (sterol, triglyceride) were radioiodinated in 55-99% radiochemical yield by isotope exchange with Naf25I in a melt of pivalic acid. In general, the reaction was complete in 1 h at 155°C with little or no substrate decompostion. High specific activity studies afforded 125I-labelediopanoic acid with a specific activity of over 700 Ci/mmol.

Introduction Although isotope exchange of aliphatic iodides introduction of radioiodine into organic molecules is usually conducted in a refluxing solvent such as ~:an be accomplished by a variety of techniques acetone, methyl ethyl ketone (MEK), water, or depending on the structure of the compound to be ethanol, °) unactivated aryl iodides generally require labeled. Excellent reviews of radioiodination methods higher reaction temperatures to effect the exchange. I~ave recentlybeen reported/~'2~ Aromatic compounds Accordingly, high boiling solvents such as propylene possessing electron donating groups (e.g. phenols glycol have been utilized with some success. °~) In ~nd anilines) are easily radiolabeled by electrophilic another method, the. substrate and radioiodide are iodination in the presence of radioiodine, iodine reacted at elevated temperatures (100--200°C) in a raonochloride, chloramine-T, iodogen, etc/~) A "melt" fashion. The molten reaction medium must x ariety of methods also exist for radioiodination possess a sufficiently high dielectric constant to solu- cf unactivated aromatic rings and, in general, the bilize both the substrate and radioiodide. Examples ~lethod of choice is dependent on the degree of of such media include benzoic acid02) (m.p. 122°C, sraecific activity sought. If high specific activity (Ci/ b.p. 249°C) and acetamide (m.p. 82°C, b.p. 221°C). nlmol) is desired, one approach involves exchange The latter has been used with some success by several of an appropriate leaving group with radioiodide groups, ester hydro- toluene and distilled under nitrogen) was added. The lysis. Subsequent studies with og-(3-amino-2,4,6-tri- vial was resealed as previously described and partially iodophenyl)-alkanoate triacylglycerols and free acids immersed in a preheated (155-160°C) oil bath. When were also successful. Based on these favorable results, ~he isotope-exchange reaction was essentially com- we continued to study the scope of pivalic acid as plete (usually 1-2 h), the reaction vial was allowed an exchange medium with a variety of different to cool, dry THF (200/zL) was added with a glass substrates as shown in Tables 1-5. syringe and the vial swirled gently. A TLC sample (1-2 #L) was removed with a 10/~L syringe and the remaining contents were transferred to the top of a silica gel-60 column (1 x 10cm) and subsequently Materials and Methods eluted with the appropriate solvent system. If neces- sary, especially when labeling polar compounds, ex- The 125I used was a no-carrier-added solution of cess pivalic acid can be removed prior to chromatog- Na~25I (5 mCi/0.1 mL) in reductant-free 0.1 N NaOH raphy by inserting a disposable syringe containing obtained from New England Nuclear. Radioactivity granulated charcoal as a trap into the reaction vial was quantified with a Searle 1185 y-counter. Tetra- while heating and allowing the pivalic acid to distill hydrofuran (THF) was distilled from LiA1H4 under into the trap. When eluting the column, a survey argon immediately prior to use. All radioiodination meter probe was placed at the outlet of the column reactions were conducted inside a plexiglass glove to serve as a radiodetector. Fractions were collected box vented with a model RIT-140 radioiodine trap and the radiochemical purity of each was monitored (Hi Q Filter Products, La Jolla, California). Thin by TLC using u.v. and radioactivity detection. The layer chromatography (TLC) analyses were per- appropriate fractions were combined and the solvents formed on Merck silica gel-60F254 polyethylene- removed with a gentle stream of nitrogen. HPLC backed plates. The plates were analyzed by u.v. analysis of the final compound confirmed both chem- and scanned for radioactivity on a Vangard 930 ical (u.v.) and radiochemical (radioactivity) purity, autoscanner immediately after development and and because u.v. and radioactivity were plotted drying. Radiochemical yield was then determined by simultaneously with a two-pen strip chart recorder, integration of the chromatogram peaks (by cut and calculation of specific activity was also possible by weigh) and calculated as the ratio of radioactivity of comparison of u.v. and radioactivity peak areas to exchanged product to the total radioactivity on the preconstructed standard calibration curves. Reaction plate. In general, two solvent systems were used times and radiochemical yields are included in (hexane/ethyl acetate, 5:2 for esters and THF/ Tables 1-5. Actual isolated yields were generally hexane, 6:4 for acids). In both systems, the Rr of 5-20% lower than shown depending on the extent of Na~ZSI was from 0 to 0.05. Column chromatography purification necessary or desired. In all cases, radio- was performed on Merck silica gel-60 (230-400 mesh) chemical purity of final compounds exceeded 95%. eluted in the same two solvent systems. The HPLC system used to confirm radiochemical purity and determine specific activity of some of the compounds, consisted of an Altex l l0A pump, Gilson 111 u.v. Results and Discussion monitor (254nm) and in-line Bicron Analyst® The pivalic acid exchange method reported here (Bicron Electronic Products, Newbury, Ohio) survey evolved during studies of the synthesis and evaluation meter equipped with a model GILE low energy y of potential organ- or tumor-specific scintigraphic scintillation probe and strip chart output. Columns and CT imaging agents. Table 1 represents a series and solvents used include a Watman Partisil 5 RAC of co-(3-amino-2,4,6-triiodophenyl)-alkanoic acids column (9.4 x 100mm) eluted with hexane/ethyl which ironically includes both the highest and also acetate (9:1) and an Excalibur 5#m (4.6 × 250 ram) some of the lowest yields listed. During radio- Ct8 reverse-phase column eluted with methanol. iodination of acid la, the reaction mixture remained Radioiodine exchange in pivalic acid 909

Table 1. Radioiodination of co-(3-amino-2,4,6-triiodophenyl) alka- Table 3. Radioiodination of cholesteryl o~-(3-amino-2,4,6-triiodo- noic acids in pivalic acid phenyl)-alkanoates in pivalic acid I o i~%.~ i X NH2 I Oo~ Compound n x Time (h) Yield (%)a la 0 H 2 59 i ..~,,.~ i x ib 1 H 1 99 Entry n x Time (h) Yield (%)a le 1 Et 1 98 ld 10 H 1 99 5a 1 H 1 94 ie 15 H 3 59b 5b 1 Et 1 98 5¢ 10 H 3 95 ~By TLC of the reaction mixture. ~Accompanied by unidentified side products (31%). =By TLC of the reaction mixture.

heterogeneous apparently due to lack of solubility of iopanoate (5b) was an improvement over the 35% the substrate. This problem was eventually overcome, yield previously reported with acetamide. (2°1 however, by substituing neodecanoic acid, a higher The results were less consistent, however, for the boiling ( > 200°C) isomeric mixture of trialkylacetic pregnenolone mono- and polyiodinated benzoate acid (methyl, ethyl, propyl mixtures), for pivalic acid. esters (6a-f) and carbamate (6g) (Table 4). As may be When conducted in neodecanoic acid at 200°C, the anticipated, the triiodinated benzoates (6a, b) were reaction mixture was homogeneous and ultimately labeled the most efficiently, and although yields were afforded radioiodinated la in 86% radiochemical lower for the remaining compounds (6c-g), there was yield. The low yield (59%) obtained for acid le was no evidence of ester hydrolysis. due to formation of a more polar side-product (31%). To further study the scope of the reaction, we Esterification of glycerol with acids la-e afforded subjected a variety of different compounds to these a series of mono-, di- and trisubstituted triacyl- radioiodination conditions as shown in Table 5. glycerols. The results obtained upon subsequent Radioiodination of o-iodohippuric acid (hippuran, 7, radioiodination of the intact triacylglycerols ranged Fig. l), a scintigraphic agent for assessing kidney from 70 to 99% radiochemical yield as shown in function, had previously been reported by at least Table 2. Formation of hydrolysis products such as five different groups. 091 Yields ranged from 2 to 57% free acids, mono- and diglycerides, was quite minimal until Wanek et al., ~2~) in 1977, and Hawkins et al.f 91 and except for 2a (11%) only ranged from 0 to in 1982, successfully obtained [mI]hippuran in 98% 6%, High radiochemical yields (94-98%) were also radiochemical yield in the presence of copper sulfate. abtained with cholesteryl-eg-(3-amino-2,4,6-triiodo- Employing pivalic acid as an exchange medium, phenyl)-alkanoates. (5a--e) as shown in Table 3. radioiodination of o-iodohippuric acid (7) was Moreover, the yield of 98% obtained for cholesteryl equally successful, and thus may be applicable for use with ~z31. ['able 2. Radioiodination of og-(3-amino-2,4,6-triiodophenyl) alka- Radioiodination of 1-(2-chlorophenyl)-l-(4-iodo- noate triacylglycerols in pivalic acida phenyl)-2,2-dichloroethane (9, Fig. 1), an analog of RIo._~O R the adrenal chemotherapeutic agent o,p'-DDD (Mi- ~---0 R Table 4. Radioiodination of pregnenolone aryl esters in pivalic acid Entry R R ~ Time (h) Yield (%)b 2a P la 1 92 2b P lb 1 80 2c P Ic I 80 2d P ld 1 71 2e P le 2 97 3a la P 2 83 3b lb P 2 96 R 3e lc P 1 94 3d ld P 1.5 91 Entry R Time (h) Yield (%)a le P 2 90 6a 2,3,5-tr/iodobenzoate 95 4a lb lb 2 92 6b 3,4,5-triiodobenzoate 94 4b lc lc 6 70 6¢ 2,5-diiodobenzoate 81 4¢ Id Id 1.5 93 6d 2-iodobenzoate 82 4d le le 2 80 6e 4-iodobenzoate 81 6f 2,6-dimethyl-3-iodobenzoate 'P = palmitate, otherwise R, R ~correspond to acyl group listed in 63 Table 1. 6g 4-(4-iodophenyl) butyryl-carbamate 68 ~By TLC of the reaction mixture. aBy TLC of the reaction mixture. 910 JAMEY P. WEICFIERTet al.

Table 5. Radioiodination of miscellaneous compounds in pivalic acid less than expected, it was the result of a single trial No. Compound Time (h) Yield(%)a and was not optimized. 7 o-iodohippuric acid 1 99 In order to test the applicability of using pivalic 8 5-iodotubercidin 1 55 acid to effect I for Br exchange, entries 10-13 were 9 1-(2-chlorophenyl)-1 -(4-iodophenyl)- 2.5 73 included. The low values obtained with 3-iodo- 2,2-dichloroethane 10 3-iodobenzoic acid } 35b benzoic acid (10) and 3-bromobenzoic acid (11) are 11 3-bromo benzoic acid d 1 15¢ in agreement with results reported by Mangner, ~6~ 12 l l -iodoundecanoic acid 1 96 13 11-bromoundecanoic acidd 1 99 who attributed the low yield of 10 to deactivation by 14 15-(4-iodophenyl) pentadecanoic acid I 74 the electron-withdrawing carboxyl group. In con- 15 1-palmitoyl-2-iopanoyl-sn-glycero- 1 60 trast, interhalogen exchange in the presence of pivalic phosphatidyl choline acid was easily effected with ll-bromoundecanoic ~By TLC of the reaction mixture. bAccompanied by an impurity (54%). acid (13). The nearly quantitative incorporation of CAccompanied by three impurities (71%). radioiodide exceeds the results typically obtained by dRadiolabeled products are [IZSll3-iodobenzoic acid for il, and I for Br exchange of aliphatic bromides conducted in [~25I]1 l-iodoundecanoic acid for 13. refluxing acetone or methyl ethyl ketone (80%). Moreover, routine HPLC separation of 17-[J23I]iodo- totane), and also of 5-iodotubercidin (8, Fig. 1), an heptadecanoic acid from its bromo precursor by adenosine kinase inhibitior, was also fairly successful. Dudczak et al., ~22~ if applicable here, offers potential In addition, 15-(4-iodophenyl)-pentadecanoic acid, for the preparation of no-carrier-added quantities a widely studied myocardial imaging agent, has been of radioiodinated fatty acids in high radiochemical radioiodinated with 1231 in a melt of benzoic acid at yields, 170~C in 95% yield by Eisenhut. "2) Although the In order to further investigate the reaction charac- 74% radiochemical yield presented in Table 5 was teristics, several additional studies were conducted.

o ~ ~ /~OIcI(cH,)CH~ O ~l NHCHzCO2H

NH 2 O II 2~C R= "OC(CHt)ICH , 7 O II ° 15 R= -O P OCHzCHzNMe3

CHCHCl 2

NO OH I

$ 9

coo.

14 Fig. 1. Miscellaneous compounds radioiodinated by exchange in pivalic acid. Radioiodine exchange in pivalic acid 911

Initial time and temperature studies were performed 100 with iopanoic acid (le) and glyceryl 1,3-dipalmitoyl- 2-iopanoate (2e) in order to ascertain the labeling efficiency of the free acid vs the ester. As can be seen • 80 in Fig. 2, radioiodination of iopanoic acid was essen- X ,'X X

tially complete by 10min at 155°C while 60min at .-'2 60 100°C were required to achieve 84% radiochemical yield. Radioiodination of iopanoate ester 2e (Fig. 3), g / on the other hand, was slower and was also accom- "E 40 z-o X tx panied by the formation and disappearance of an g apparent transient radioactive intermediate which 20 was less polar than 2c by TLC (vide supra). To investigate the acid catalysis of pivalic acid in the exchange reaction, a comparable study was con- ~0 30 60 ducted with iopanoic acid (le) in neopentyl alcohol Time ( min ) (2,2-dimethyl-l-propanol, m.p. 52°C, b.p. 114°C) at Fig. 3. Radiochemical yield of [1251]-2¢ ( × ) and transient 105°C for 1 h (Table 6). A radiochemical yield of intermediate (©) as a function of time at 155°C. 33% was obtained along with the formation of three radioiodinated side-products (49% total) as com- pared to an 84% yield obtained with pivalic acid at Throughout these studies high specific activity was essentially the same temperature. When an ester, not a goal and generally not considered achievable by cyclohexyl iopanoate, was subjected to exchange in isotope exchange under the described conditions. neopentyl alcohol, a 50% yield was obtained. Addi- However, by increasing the amount of Na~:SI tion of Amberlite, a resin containing sulfonic acid (2.5 mCi) and decreasing the amount of iopanoic acid functionality often used in catalytic applications, to (l/zg), we were able to obtain [~zsI]iopanoic acid with neopentyl alcohol resulted in an even lower yield a specific activity of over 700 Ci/mmol in roughly (39%) of [125I]cyclohexyl iopanoate. By comparison, 60% radiochemical yield (Table 8). Of interest was the pivalic acid exchange afforded [ ~25I]cyclohexyl appearance of a slightly less polar radioactive im- iopanoate in 70% yield with no accompanying side-product formation. These results imply that in Table 6. Apparent catalytic effect of pivalic acid at 105°C for 1 h addition to its solvent role, pivalic acid apparently Yield (%) also has a catalytic influence on the exchange Substrate Conditions Desired Other a reaction. Iopanoic acid Neopentyl alcohol 33 49 Because exchange occurs too rapidly with iopanoic Cyclohexyl iopanoate Neopentylalcohol 50 14 Cyclohexyl iopanoate Neopentyl alcohol 39 23 tcid in pivalic acid at 155°C, cyclohexyl iopanoate plus Amberlite ~¢as also utilized to study the relative labeling Cyclohexyl iopanoate Pivalicacid 70 0 ,'fficiencies of pivalic acid, benzoic acid, and acet- ~Total radioactive side-products excluding unreacted Na~251. tmide. As seen in Table 7, the relative efficiencies )f the three media were: pivalic acid > benzoic Table 7. Effect of various media on radioiodide exchange between Nat2Sl and cyclohexyl iopanoate ~ tcid >> acetamide. Medium Yield (%)b Acetamide 9 Benzoic acid 65 100[-- X ,X X Pivalic acid 78 'Conditions: 3 mg cyclohexyl iopanoate, 10-12 mg of media, 2juL aq. Na':SI (200/~Ci), 155°C for 1 h. bby TLC of reaction mixture

Table 8. High specific activity study with iopanoic acid' 6O Rad. Yield (%)h Spec. act. Amount (,ug) Na]25I (mCi) desired transient Ci/mmol 40 100 1.0 95 0 5.4 10 1.0 90 0 51.4 g~ 1 1.5 59 22 498c 2O 1 2.5 56 40 778c 'Conditions: substrate (1-100 #g), aq. Na~:51 (1-2.5 mCi, 2-5/zL) and pivalic acid (l mg) at 155°C for l h in a Wheaton micro V-vial 10 30 60 (0.3-mL) bby TLC of reaction mixture (silica gel, THF/hexane (6:4), Time (rain) iopanoic acid Rr = 0.50, transient R r = 0.64). Ccalculated by comparison of HPLC radioactivity trace vs a Fig. 2. Radiochemical yield of [125I]iopanoic acid (lc) as a standardized radioactivity calibration plot and quantity of starting function of time and temperature. 1 mg of le was exchanged iopanoic acid. Since we were unable to see iopanoic acid by u.v. at at 155 (x) and 100°C (O). 254nm at this scale, this represents a minimum specific activity.

A ~.1 37/8--Q 912 JAMEY P. WEICHERTet al. purity much like that formed with glyceryl iopanoate withdrawing groups such as 3-iodobenzoic acid, 2e at 105°C (vide infra). An important difference, seems to favor an electrophilic rather than a nucleo- however, was that in the previous case the substrate philic mechanism. (2c) was present in a very large excess on a molar basis relative to Na~25I. At the highest specific activity Conclusions level (1/~g of iopanoic acid, 2.5 mCi of high specific A variety of benzoic and aryl aliphatic mono and activity Na125I, 17.4 Ci/mg Na~25I) the molar ratio of polyiodinated acids and esters, as well as substrate to radioiodide was only about 1.85. Radio- 5-iodotubercidin, an analog of o,p'-DDD (Mito- active side-products have been reported (23'24) upon tane), and l l-bromoundecanoic acid were radio- co-iodination of fatty acids by halogen exchange with iodinated by isotope exchange with Nal25I in a melt Na123I and Na13q in MEK, and in one case, was of pivalic acid at 155°C. In general, the reaction was found by GC-MS analysis to be butyl iodide. (24~ complete in 1 h with little or no substrate decom- In order to gain insight into the nature of the position. Radiochemical yields were good to excellent transient intermediate, it was isolated from the (55-99%) except for 3-iodo- and 3-bromobenzoic 2.5mCi high specific activity reaction mixture by acids, which were apparently deactivated by the HPLC and portions subjected to the following electron-withdrawing effect of the carbox~¢l group. experimental conditions: In addition, I for Br exchange with l l-bromo- (1) High specific activity Na~251 and pivalic acid undecanoic acid afforded almost quantitative yields were heated in the usual manner without iopanoic of [ 125I]iodoundecanoic acid which, if separated from acid. A large radioactive peak (88%) with Rf = 0.61 the bromo substrate according to a recently reported (THF/hexane, 6:4) and possessing no u.v. absorb- method, (22) could give rise to carrier-free co-iodo- ance by TLC was formed. Further treatment of the alkanoic acids. Radioiodination of cyclohexyl reaction mixture with excess sodium thiosulfate iopanoate in pivalic and benzoic acids as well as resulted in total conversion of radioactivity with acetamide resulted in the following relative labeling Rf= 0.61 to the origin (Na12SI), thus implying the efficiencies: pivalic acid >benzoic acid>>acetamide. formation of some oxidized form of iodide which is Additional high specific activity studies afforded much less polar than Nat25I itself. Accordingly, a [125I]iopanoic acid with a specific activity of over portion of the transient intermediate fraction was 700 Ci/mmol. Because of its rapidity, the method is treated with sodium thiosulfate in a similar manner appropriate for use withmI (tt/2 = 13.3 h), currently to see if it also could be "reduced" back to Na~2SI. the iodine nuclide of choice for scintigraphic imaging. There was no change, however, indicating that the Finally, because the method is most successful with intermediate formed in the presence of iopanoic acid lipophilic substrates, it is very complimentary to the was not the same one seen previously upon heating ammonium sulfate method, "6) which is best suited Na125I and pivalic acid. toward water-soluble compounds, and as such, pro- (2) When the transient intermediate fraction was vides the radiochemist with an efficient method of reheated in the presence of additional iopanoic acid radioiodinating a wide variety of compounds. and pivalic acid, all of the radioactivity became Acknowledgements--The authors would like to thank Judy associated with the iopanoic acid. The result was Weichert for preparation of the manuscript and the inconclusive, however, because it was not known National Cancer Institute (CA-08349) for support of this whether the transient intermediate had radio- work. iodinated the new iopanoic acid or was actually an References intermediate iopanoate-iodine species which upon further heating was converted to [~:5I]iopanoic acid. 1. Seevers R. H. and Counsell R. E. Chem. Rev. 82, 575 (1982). In response to this uncertainty, a final experiment was 2. Stocklin G. Int. J. Appl. Radiat. Isot. 28, 13l (1977). conducted in which another portion of the transient 3. Kabalka G. Acc. Chem. Res. 17, 215 (1984). intermediate was simply reheated in the presence of 4. Kulkarni P. V. and Parkey R. W. J. Labeled Compd. pivalic acid. The radioactivity was indeed converted Radiopharm. 21, 1113 (1984). 5. Foster N. I., Dannals R., Burns H. D. and Heindel to [~25I]iopanoic acid. Therefore, based on these N. D. J. RadioanaL Chem. 65, 95 (1981). preliminary studies, it appears that the transient 6. Knapp F. F. and Callahan A. P. Proe. 4th Int. Syrup. "impurity" may actually be an iopanoic acid- Radiopharm. Chem., Julich, p.53. (1982). radioiodine complex which upon further heating can 7. Moerlein S. M. and Coenen H. H. J. Labeled Compd. be converted to labeled iopanoic acid. Further assess- Radiopharm. 21, 1076 (1984). 8. Machulla H. J., Marsmann H. J. and Dutschka K. ment of mechanistic detail is difficult in view of the J. Radioanal. Chem. 56, 256 (1980). non-classical behavior of iodine in very small (nmol) 9. Reske S. N., Sauer W., Machulla H. J. and Winkler C. quantities, where for instance, oxidation species like J. Nucl. Med. 25, 1335 (1984). HIO and HIO2 are known to be quite stable. ~'16'2~'24~ 10. Eckelman W. C., Adams H. R. and Paik C. H. Int. J. Nucl. Med. Biol. 11, 163 (1984); (and references cited Nevertheless, the implication of the involvement of therein). an oxidized iodine species coupled with the poor 11. Wu J. and Wieland D. M. J. Labeled Compd. Radio- results obtained with compounds containing electron pharm. 16, 172 (1979). (Abstr. AI5). Radioiodine exchange in pivalic acid 913

12. Eisenhut M. Int. J. AppL Radiat. Isot. 33, 499 (1982). 20. Seevers R. H., Groziak M. P., Weichert J. P., 13. Elias H. and Lotterhos H. F. Chem. Bet. 109, 1580 Schwendner S. W., Szabo S. M., Longino M. A. and (1976). Counsell R. E. J. Med. Chem. 25, 1500 (1982). 14. Wester G. and VanGijlswijk H. J. M. J. Labeled Compd. 21. Wanek P. M., Hupf H. B. and O Brien H. P. J. Nucl. Radiopharm. 16, 174 (1979). (Abstr. AI7). Med. 15, 638 (1977). (Abstr.). 15. Seevers R. H., Schwendner S. W., Swayze S. L. and 22. Dudczak R., Kletter K., Frischauf H., Losert U., Counsell R. E. J. Med. Chem. 25, 618 (1982). Angelberger P. and Schmoliner R. Eur. J. Nucl. Med. 9, 16. Mangner T. J., Wu J. and Wieland D. M. J. Org. Chem. 81 (1984). 47, 1484 (1982). 23. Laufer P., Machulla H. J. and Michael H. J. Labeled 17. Eisenhut M., Kimmig B., Bubeck B., Sinn H., Zum Compd. Radiopharm. lg, 1205 (1981). Winkel K., Taylor D. M. J. Labeled Compd. Radio- 24. Mertens J., Vanrycheghem W., Bossuyt A., Van den pharm. 21, 1130 (1984). Winkel P. and Vandendriessche R., J. Labeled Comp. 18. Mills S. Int. J. Appl. Radiat. Isot. 33, 467 (1982). Radiopharm. 21(9), 843-56 (1984). 19. Hawkins L., Elliot A., Shields R., Herman K., Horton 25. Kahn M. and Wahl, A. C. J. Chem. Phys. 21, 1185 P., Little W. and Umbers C. Eur. J. Nucl. Med. 7, 58 (1953). (1982); (and references cited therein). 26. Cvoric, J. J. J. Chromatogr. 44, 349 (1969)