- Email: [email protected]
Neutral technetium(II)-99m complexes as potential brain perfusion imaging agents
Nucl. Med. Biol. Vol. 14, No. 5, pp. 503-510, 1987 hf. J. Radial. Appl. Instrum. Part B Printed in Great Britain. All rights reserved
Copyright 0
0883-2897/87 $3.00 + 0.00 1987 Pergamon Journals Ltd
Neutral Technetium(II)-99m Complexes as Potential Brain Perfusion Imaging Agents MARIA NEVES,* KAREN LIBSON and EDWARD
DEUTSCHT
Departments of Chemistry and Radiology, University of Cincinnati, Cincinnati, OH 45221, U.S.A. (Received 2 October 1986)
A series of eight neutral technetium(II)-Wm complexes of general formula rr-[99”r~*‘D,X,]~, where D represents a chelating ditertary phosphine or arsine ligand and X represents a halide or pseudohalide ligand, has been prepared and characterized by HPLC comparisons to the known %z analogs. Several members of this series exhibit significant brain uptake in rats, with maximum uptake (1.2% dose/brain at 1 min after injection) being exhibited by fr-[99mTc”(DIARS)2C12]0 where DIARS represents ophenylenebis(dimethylarsine). Surprisingly, several cationic Tc(III) analogs, tr-[99”rcD,X,]+, also exhibit significant brain uptake, presumably via a mechanism which involves in oivo reduction to the neutral Tc(II) form. The complicated interrelationships among the chemical and biological properties of the rr[~c’u~“D,X,]+~ couples, and the dependencies of these properties on the nature of D and X, provide possible means by which this class of complexes might be manipulated to yield an effective 99mTcbrain
perfusion imaging agent.
Introduction Numerous authors and groups have noted the great potential of SPECT (single-photon emission computed tomography) imaging techniques in the diagnosis and management of cerebrovascular disease.(‘4) In the context of brain perfusion imaging, 133Xehas been used for some time to determine regional cerebral blood flow,“) but tomographic imaging with this isotope requires specially dedicated SPECT instrumentation which is not optimal for other tomographic applications. Much more useful results have been obtained in the past few years with the lz31 labelled amines 1Z31-IMPand ‘231-HIPDM(C6)(IMP = N-isopropyl-iodoamphetamine; HIPDM = N,N,N’trimethyl-N-(2-hydroxy-3-methyl-5-iodobenzyl)-1,3propanediamine). These amines exhibit high brain uptake and long cerebral retention times,“) and clinical studies with these agents have clearly demonstrated the ultimate utility of single-photon emitting radiopharmaceuticals that monitor regional cerebral blood flow.” However, since iz31must be produced in a cyclotron and possesses a physical half-life of 13.3 h, it is expensive and not universally available for daily use. Clearly, the ideal SPECT brain perfusion *IAEA Fellow on leave from LNETI, Instituto de Energie, Portugal. t Author to whom all correspondence should be addressed. 503
agent should be based on the readily available, inexpensive *Tc isotope.(a9) This has been widely recognized ever since Oldendorfs 1978 calI for a new class of %Tc radiopharmaceuticals that would penetrate the intact blood-brain barrier (BBB) and have prolonged retention in the brain. While %Tc brain perfusion imaging agents have been systematically sought since about 1978,(i”~i1)it has only been in the last three years that significant progress towards such agents has been achieved. Neutral Tc(V) complexes, containing derivatives of the tetradenat*N, S2 dithiol ligand HSCH,CH,NHCH,CH,NHCH,CH,SH, have been shown to cross the BBB, but the underivatized complexes are not sufficiently retained in the brain. Attachment of pendant amine functionalities to the core Tc(V) complexes leads to significant brain retention in primates,“3~‘4Jsimilar to the brain retention observed for the 1231labelled amines.@-‘) A totally different class of neutral Tc(V) complexes, containing derivatives of the tetradentate_N, bis(oxime) ligand HON= CHCH2 NHCH,CH, CH, NHCH,CH=NOH, has also been shown to cross the BBBuS) and several of these complexes are retained in the brain long enough to allow clinically useful SPECT images to be obtained in humans.““‘*) Interestingly, these complexes do not contain pendant amine functionalities. In fact, derivatization of the core Tc(V) complex with pendant amine functionalities is detrimental to brain
504
MIWA NIMS et al.
uptake and retention.(19*M) Among the bis(oxime)Tc(V) complexes, those that are retained in the brain differ from those that are not, solely by the distribution of methyl groups about the periphery of the complex. (“Jo) The biological mechanisms of action of these agents have yet to be elucidated.(2’) Thus, all of the %Tc brain pe rf u sion imaging agents evaluated to date are neutral, lipophilic complexes of technetium(V) which contain Tc==O linkages. It is known that a molecule must be neutral and lipophilic in order to passively diffuse across the BBB.(22*23) But since almost nothing is known about the biological mechanisms of action of the %Tc agents, it is not all clear whether either the Tc(V) oxidation state or the Tc=O linkage, is necessary in order for a *Tc complex to cross the BBB. In order to probe this issue we have evaluated a series of neutral, lipophilic complexes of Tc(I1) (which do not contain T& linkages) for their ability to cross the BBB in rats. These complexes have the general formula fr-[ *Tc” D, X, 1’ wher e D represents a chelating ditertiary phosphine or arsine ligand and X represents a halide or pseudohalide ligand.(%“) Moreover, in order to obtain some information about the role of in viuo redox reactions in determining the biodistributions of these complexes,(2*) we have similarly evaluated some of the analogous Tc(II1) cations tr-PgmT~‘i’&X~] +. The results of this comparative study are reported herein.
(X = Cl, Br) were generated in siru by chemical reduction of the corresponding Tc(II1) complex, and then were characterized by the well established visible-u.v. spectrophotometric parameters of these Tc(II) complexes. (24,25)No ?Ic standards were available for I~-[CA”TC”(DMPE),I,~+ and tr[99”Tc”(DMPE), I,]‘; these complexes were indirectly characterized by their syntheses, which are analogous to those of the bromo complexes, and by comparing their HPLC retention times to those of the corresponding chloro and bromo analogs. AH *Tc complexes were prepared from “no carrier added” wmTcO; obtained by elution of a commercial 99Mo/99”T~ generator with normal saline. If the presence of chloride would interfere with preparation of the 99mTccomplex, the *TcO; was freed from chloride by chromatographic separation as (nC4Hs)4pT~04 on a C, or Cl* reversed phase cartridge, followed by elution with absolute ethanol; this procedure has been described for the purification of 186ReOq.(2*)The preparations of the 9h”rc(III) complexes were generally performed as previously detailed for tr-[99rnT~‘*i(DMPE)~X,]+ (X = Cl, Br);‘26) all syntheses were conducted anaerobically using a dry heat bath to maintain the appropriate reagents at the desired temperature in borosilicate vials (Wheaton) sealed with Teflon lined caps. After being heated for the desired length of time these vials are slowly cooled to room temperature. Cautiowrugid’ cooling the vial can cause it to break with explosive
Materials and Methods
force; borosilicate serum vials are used to minimize this hazard. Most of the 99mTc(II) complexes were pre-
Unless otherwise noted all chemicals were of reagent grade. Solvents used in HPLC analyses were of HPLC grade. The o-phenylenebis(dimethylarsine) (DIARS), 1,2-bis(dimethylphosphino)ethane (DMPE), and 1,2-bis(diphenylphosphino)ethane (DPPE) ligands were obtained from Strem Chemical Co. and were used without further purification. The air sensian inert atmosphere, and solution of these ligands were stabilized with strong acids (HCl, HBr or HClO,) as previously described.(26) HPLC (high performance liquid chromatographic) analyses were conducted using a previously detailed(26) apparatus equipped with both U.V. (254nm) and radiometric (centered at 140 keV) detectors and an integrating recorder. The use of these two detection systems allows the positive identification of “unknown” *Tc complexes by simultaneous injection of the “unknown” 99m Tc complex and a “known” 99Tc complex which has been characterized by classical chemical techniqueso6) Observation of simultaneously eluting radiometric and U.V. HPLC peaks confirms that the %Tc “unknown” and the well characterized qc complex are the same chemical species. Well characterized samples of tr-[99Tc’u(DIARS),XJ+ (X = Cl, Br), ~~-[~~Tc”‘(DPPE)~X,]+ (X = Cl, Br), ~~-~c~(DPPE)~(NCS)J~ and tt-~c’u(DMPE),X,] + (X = Cl, Br) were available from previous studies 9(U,27,29*M) while samples of ~~-~c”(DMPE),X,]~
pared by chemical reduction of the corresponding %Tc(III) complex. Details of the various *Tc(III) and *Tc(II) sy nthetic procedures are specified below. The radiochemical purities of the final preparations were monitored by HPLC using a silica based cation exchange column (Vydac, 10 micron: 150 x 4.6 mm) and a 90/10 methanol/water mobile phase that was 0.04 M in (n-C,H,),NH,PO,. In all cases, except one, the radiochemical purity was greater than 95%; in the case of the single exception, tr -pTc” (DPPE), (NCS), Jo, the radiochemlcal purity was only 80% and the primary radiochemical Before being used in animal impurity was *TcO;. studies, all *Tc preparations were adjusted to a pH value between 6 and 8 using either H3 PO4 or Na, PO,, filtered through either a 0.45 or a 0.20 micron sterile filter (Gelman), and then diluted with saline so that the final concentration of ethanol was less than 10%. The various %Tc agents were screened for their ability to cross the BBB in anesthetized (Metofane), female Sprague-Dawley rats of ca 200 g weight. At each of various times after intrajugular injection of the agent, three rats were sacrificed by cervical dislocation, blood samples were collected, and the brains were excised. Tissue samples were weighed (average brain weight = 1.5 g) and assayed for *Tc by standard techniques. The brain and blood uptakes are expressed in Table 1 as the average (n = 3)
_
0.060 (I)
0.32 (2) 0.050(l) 0.23 (2) 0.19(l) 0.15(l) 0.08 (1) 0.10(l) 0.040(l)
0.62 (3) -
0.66 (3) 0.07(l) 0.33 (2) 0.19 (2) 0.15(l) 0.14(l) -
0.79 (3) 0.09 (I) 0.38 (3) 0.31 (2) 0.26 (2) 0.14(l) 0.24 (I) 0.10(l) 0.12(l)
0.64 (3) 0.060(I) 0.20 (2)
-
0.040(I)
0.19 (2) 0.18(l) 0.14(l) O.IO(l)
0.23 (I) 0.06(l) 0.10(l) 0.17(l) 0.18(l)
0.28 (1) 0.07 (I) 0.09 (I) 0.16(l) 0.17(l)
0.08(l) -
0.30 (2) 0.09(l) 0.07 (I) 0.15 (I) 0.16 (I)
0.32 (2) 0.11 (I) 0.09(l) 0.17(l) 0.19(l)
0.040 (I)
0.040 (I)
0.050 (I)
20
agents
0.060(l)
6
of Tc
2
4
uptakes
t=l
Brain
I. Brian and blood
0.12(2) 0.040(l) 0.26 (2) 0.22(l) 0.18 (I) 0.07 (I) 0.09 (1) 0.040(l)
0.15(l) 0.04(l) 0.11 (I) 0.18(l) 0.20 (I)
0.040(l)
6Omin
2
(3) (3) (2) (3) (3) (5)
(2) (2) (1) (I) (2)
0.24 (2) -
-
0.48 (2) 0.28 (2) 0.18(l) 0.24 (3)
0.19 (2) -
4
Me,PCH,CH,PMe,;
6
20
0.24 (2) 0.15(l) 0.06 (I) 0.15(l) 0.26 (2) 3.22 (4) 0.28 (2) 0.75 (3)
0.34 (2) -
6Omin
0.08 (I) 0.040(l) 0.14(l) 0.22 (2) 0.30 (2) 3.39 (5) 0.70 (3) 0.36(2)
0.67 (I) 0.040(l) 0.08 (I) 0.08 (I) 0.14(i)
0.81 (2)
DPPE
represents
1,2-bis(diphenyl-
At least three rats were sacrificed
-
0.12(l) 0.20 (2) 0.24 (2) 3.31 (5)
0.79 (I) 0.10 (I) 0.05 (I) 0.06 (I) 0.12(l)
I .02 (3) I .Ol (2) 0.16(l) 0.09 (I) 0.05 (I) 0.18 (2)
I .38 (3)
Blood
digit is given in parentheses.
1.70(4)
0.64 0.35 0.25 0.40 0.40 3.50
1.21 0.38 0.30 0.50 0.42 4.20 0.34 2.61
(5) (3) (2) (4) (3) (6) (3) (5)
I .22 0.24 0.12 0.07 0.30
I .78 (4)
of time
I .45 (4) 0.34 (3) 0.21 (2) 0.12(l) 0.38 (3)
2.20 (4)
1=I
in rats as a function
‘Expressed as percent injected dose per gram of tissue at the time of assay. The standard deviation of the last significant at each time point. Data are corrected for the physical decay of 99”rc. 1,2-bis(dimethylphosphino)ethane, bDIARS represents o-phenylenebis(dimethylarsine), o-C,H,(AsMe&; DMPE represents phosphino)ethane, Ph,PCH,CH2PPh,. (Me = CH,; Ph = C,H,).
Aneatb
Table
506
MAIU Naves et at.
percentage of injected dose per gram of tissue at any time of assay. Synthesis of 99mTccomplexes (i) tr-[99”Tc”‘(DIARS), Cl,]+. An HCl stabilized 1% solution of DIARS is prepared by combining 30 ~1 of DIARS ligand with 0.5 mL 1.0 M HCl and 2.5 mL ethanol. Fifty microliters of the DIARS solution, 30 PL of 6 M HCl, and 0.40 mL of 95% ethanol are added to the desired amount of 99”Tc0; in 1.0 mL of saline. This reaction mixture is heated at 140°C for 20 min. (ii) tr-[99mTc”(DIARS)2 Br,]+. To the desired amount of (n-C,H,),p”TcO; in 1.OmL absolute ethanol are added 0.30 mL of 9 M HBr and 40 microliters of a 1% solution of DIARS in absolute ethanol. This reaction mixture is heated at 100°C for 20 min. (iii) tr-[99mTc’1’(DMPE)2 Cl,]+. An HCl stabilized 1% solution of DMPE is prepared by combining 30 ~1 of DMPE ligand with 0.5 mL 1.0 M HCl and 2.5 mL ethanol. Fifty microliters of this DMPE solution and 0.80mL of 95% ethanol are added to the desired amount of %TcO; in 1.OmL of saline. This reaction mixture is heated at 140°C for 30 min. (iv) tr-[99mTc”(DMPE)2Br2]+. An HBr stabilized 1% solution of DMPE is prepared by substituting HBr for HCl in the above formulation. Fifty microliters of this DMPE solution, 0.8 mL of 3 M NaBr, and 50 ~1 of 4.5 M HBr are added to the desired amount of (n - CqH9)q~mT~04 in 0.70 mL ethanol. This reaction mixture is heated at 140°C for 30 min. (u) tr-[W”Tc”(DMPE),Z,)+. An HC104 stabilized 1% solution of DMPE is prepared by combining 30 ~1 of DMPE with 0.5 mL of 0.25 M HC104 and 2.5mL ethanol. One hundred microliters of this solution and 100 PL of 3 M NaI are added to the desired amount of (n-C,H,),N99mTc0, in 1.0 mL ethanol. This reaction mixture is heated at 145°C for 15min. (oi) tr-[ w”‘Tc”(DZARS)2 Cl,]“. (A) The tr ~Tc”‘(DIARS)&~~] + preparation is brought to pH 5 with NaOH, 3 mg of NaBH, are added, and the reaction mixture is allowed to stand at room temperature for 30 min. (B) Alternatively, the %Tc(III) preparation is brought to pH 4-5 with NaOH, either 100 PL of 2-mercaptoethanol or 100 PL of 1.OM ascorbic acid are added, and the reaction mixture is allowed to stand at room temperature for 2-3 min. (vii) tr-[99mTc’f(DIARS)2Br2]o. (A) The tr~Tc”‘(DIARS), Br,]+ preparation is brought to pH 6 with NaOH, 100 PL of 2-mercaptoethanol are added, and the reaction mixture is heated at 90°C for 20 min. (B) Alternatively, the *Tc(III) preparation is brought to pH4 with NaOH, 1OOpL of 1.0 M ascorbic acid are added, and the reaction mixture is allowed to stand at room temperature for 30min. (viii) tr-[99”‘Tc”(DMPE)2 CIJ”. The corresponding *Tc(III) preparation is treated with NaBH, as described above for the DIARS analog.
(ix) tr-~‘“Tc’(DMPE)zBrz]o. The corresponding *Tc(III) preparation is treated with 2-mercaptoethanol as described above for the DIARS analog. (x) tr-[wmTcr’(DMPE)212]o. The corresponding “mTc(III) preparation is treated with 2-mercaptoethanol as described above for tr~9mTc*‘(DIARS),Br,]o except that the reaction mixture is heated for 30 min. (xi) tr-[99mTc’(DPPE),C&]o. One hundred microliters of a 0.01 M solution of DPPE in absolute ethanol and 0.2 mL 6 M HCl are added to the desired amount of (n-C,H,),~mTcO, in l.OmL ethanol. This reaction mixture is hea@d at 140°C for 45 min. (xii) tr-[99mTcr’(DPPE)2Br,]o. One hundred microliters of a 0.04 M solution of DPPE in absolute ethanol and 0.2 mL 4.5 M HBr are added to the desired amount of (n-C, H,),N*TcO, in 1.0 mL ethanol. This reaction is heated at 140°C for 45 min. (xiii) tr-[99”‘Tc”(DPPE),(NCS)2]o. One hundred PL of a 0.04 M solution of DPPE in absolute ethanol, 200 PL of 1.OM ascorbic acid, 1.0 mL water, and 100 p L of 1.OM NaSCN are added to the desired amount of (n-C,H,),N*TcO, in l.OmL ethanol. This reaction mixture is heated at 100°C for 20 min.
Results and Discussion Chemistry
The cationic tr-[99mTd”D2X2]+ complexes with D = DIARS are prepared by procedures similar to those detailed for preparation of the complexes with D = DMPE.(26) These syntheses involve the use of the D ligand to reduce %TcOi, and to coordinate to the resulting technetium center, in the presence of HX. Careful control of reaction parameters (time, temperature, concentration of HX and D, solvent etc.) is required in order to obtain the desired 99”Tc(III) complexes in greater than 95% radiochemical purity.(26) The neutral [9A”Tc11D2X2]o complexes with D = DIARS or DMPE are prepared by chemical reduction of the analogous %Tc(III) agent, and, again, careful control of reaction parameters is required in order to obtain radiochemical purities in excess of 95%. In addition to the time and temperature of reaction, pH, solvent, etc. the successful preparation of the %Tc(II) complexes is dependent on the choice and concentration of the reductant. This in turn is dependent on the E”’ value of the Tc(III)/Tc(II) redox couple, which is controlled by the identity of the D and X ligands of the trp9”TcD2X2]+Io core.(24*25) For a given D ligand, Tc(II1) complexes with X = Br (or I) are easier to reduce than those with X = Cl. Thus, preparation of the DMPE Tc(I1) complexes with X = Br (or I) utilizes the relatively mild reductants 2-mercaptoethanol or ascorbic acid, while preparation of the chloro analog requires a use of the potent reductant NaBH,. Borohydride is too strong a reductant for the preparation of the DIARS and DMPE complexes with X = Br (or I), and its use in
Neutral *Tc(II) complexes these cases leads to large amounts of unidentified reaction products. Even in the case of the two chloro analogs, under many reaction conditions NaBH, can lead to several unidentified reaction products in addition to the desired Tc(I1) complex. For a given X ligand, complexes with D = DPPE are easier to reduce than those with D = DIARS or DMPE. Therefore, the original conversion of %TcO; to the tr-Pg”TcD,X2]+‘0 core leads to Tc(II1) complexes for D = DIARS or DMPE (which then must be reduced to Tc(I1) in a subsequent step), but leads directly to Tc(I1) complexes for D = DPPE. In fact, generation of Tc(III)-DPPE complexes from the %TcO; reaction mixture require a subsequent oxidation of the initial Tc(I1) product. Thus, the neutral Tc(II) tr-ph”Tc”(DPPE), X,]” (X = Cl, Br) complexes are directly prepared by the DPPE reduction of %TcO; using procedures similar to those detailed for preparation of the cationic Tc(II1) tr[99mTc”‘(DMPE),X,]+ complexes.(26’ Preparation of the tr-~mTc1’(DPPE)2(NCS)2]o complex is a special case because of the inherent instability of the NCS ligand under the conditions used to reduce *TcO; . To keep the reaction conditions as mild as possible, ascorbic acid is added to the original %TcO; reaction mixture. But even with this expedient, only about an 80% yield of the desired agent is obtained. This agent is of interest in the context of radiopharmaceutical development because the N-bonded NCS ligand so greatly stabilizes the Tc(I1) oxidation state that tr-[99mTc”(DPPE)2(NCS),l” can be isolated and manipulated in the presence of air.(27) Chromatography
When macroscopic amounts of qc complexes are used, standard reversed phase chromatography on CB or C,, columnso6’ successfully separates the cationic Tc(II1) species from the neutral Tc(I1) species. However, in our hands, this type of chromatography is not successful when used with 99mTccomplexes prepared at the “no carrier added” concentration level, presumably because of adventitious oxidation of the Tc(I1) species on the surface of the column support. Thus, when either of tr-[99mTc1”D2X2]f or tr~Tc” D2X2]o is subjected to reversed phase chromatography, only a single peak with the retention time of tr-~Tc”’ D2X2]+ is observed (after the known effect of injected mass on retention time is taken into account).(26) This problem is avoided by the use of cation exchange chromatography wherein the neutral 99mTc(II) complexes exhibit retention times of 4-5 min and the cationic 99mT~(III) complexes exhibit retention of 10-13 min. This type of chromatography therefore allows the %Tc(III) and %Tc(II) species to be easily distinguished even at the very low “no carrier added” concentrations of technetium encountered in 99Mo/99”Tc generator eluents. HPLC provides the crucial link between Yc and %Tc chemistry that is necessary for the preparation
so7
and biological evaluation of *Tc agents of known composition and structure. This link is effected by the simultaneous use of two detectors; U.V. detection to monitor macroscopic amounts of 99Tc complexes, and radiometric detection to monitor 99”Tc complexes.(26’ Characterized samples of all the trp9TcD2X2]+‘” cores used in this work, except trwere available from previous ~Tc(DMPE)~I~]+‘~, studies.‘24~25.27*“~‘r’) Therefore, for these complexes, coelution of simultaneously injected tr-f’%D2X2]+” “knowns” and tr-pTcD,X,]+” “unknowns”, establishes that the two are indeed the same chemical species. The tr-Pg”Tc(DMPE)21z]+‘o complexes are identified by (a) their preparation via procedures analogous to those used to synthesize the known bromo complexes, but employing iodide rather than bromide in the reaction mixture, and (b) their HPLC retention times (RT) which are consistent with those observed for the chloro and bromo analogs (for tr-PgmTc”‘(DMPE),X,]+, RT = 8.8, 11.4 and 12.6 min X = Cl, Br and I respectively; for tr~Tc~~(DMPE),X,]O, RT =4.1, 4.6 and 5.2min for X = Cl, Br and I respectively). Biodistribution
Table 1 summarizes the brain and blood uptake values observed in rats as a function of time for various 99mT~complexes. There are several aspects of these data that are relevant to the development of %Tc brain perfusion imaging agents. (i) The biodistribution of *TcO; is included as a point of reference since this species is well known not to cross the BBB. At all times the brain uptake of %TcO; is indeed very low, even though the amount of *TcO,- in the blood is relatively high. *Tc(II) complex tr(ii) The neutral ~Tc”(DIARS),CI~]~ clearly exhibits significant brain uptake. At 1 min after injection approximately 1.2% of the injected dose is in the brain (0.79% dose/g X 1.5 g/brain). This uptake is undoubtedly due to passive diffusion of the neutral Tc(I1) complex across the BBB. As the Tc(I1) agent clears from the blood, passive diffusion also carries it out of the brain leading to a relatively short retention time of the brain for tr -P)“Tc”(DIARS)* Cl,]O. Note that from 20 to 60min after injection the brain/blood ratio is relatively constant even though the amount of activity in the blood decreases by a factor of three. This implies that at long times after injection the amounts of tr-[99mTc’1(DIARS),C12]0 in the brain and blood are related by a simple equilibrium, presumably involving passive diffusion of this agent across the BBB. (iii) The neutral fr-~Tc”(DMPE)rClJ’ complex also exhibits brain uptake, but considerably less than does the DIARS analog. This lower brain uptake presumably results from the DMPE complex being less lipophilic than the DIARS analog, and thus less capable of diffusing across the BBB. However, the
508
MARU NEW et al.
tr-[ ggmTc”lDIARS)2
8
0.6 0.4 0.3 0.2 0.1 0
0
1
6
12
16
24
30
36
.?
46
6.
60
Time (Hid
Fig. 1. Time dependent brain uptake in rats for three related rr-ph”TcnD,Cl,P complexes. very lipophilic complex tr-~mTcll(DPPE)zCIJo exhibits essentially no brain uptake because its excessive lipophilicity causes it to bind tightly to blood components. Note the very high levels of this agent in the blood even 1 h after injection (Table 1). Thus it appears that in order to exhibit reasonable brain uptake, a neutral (Tc(I1) complex must be sufficiently lipophilic to cross the BBB, but not so lipophilic so as to bind tightly to blood components. Figure 1 compares the time dependent brain uptakes observed for the three related tr-PgmTcilD,CIJO complexes with D = DIARS, DMPE and DPPE, graphically illustrating the intermediate uptake of the DMPE complex. (iv) Figure 1 also shown that while the DIARS complex continually washes out of the brain, the DMPE complex first washes out and then appears to slowly reaccumulate. While this slow increase in brain uptake of tr-~Tcl’(DMPE),CIJo at long times is barely of statistical significance, the phenomenon is reproducible and also occurs for the bromo and iodo DMPE analogs (Table 1). Thus, the slow reaccumulation of the tr-~Tc”(DMPE),XJ” (X = Cl, Br, I) complexes in the brain appears to be a real effect which warrants further investigation. (v) As expected, the cationic *Tc(III) complex tr-~Tc(DMPE)2C12]+ exhibits no detectable brain uptake. Since this agent bears a formal charge, as does *TcG;, it is unable to traverse the BBB. Surprisingly, injection of the cationic analog tr~Tc(DIARS)~C~J+ does lead to a significant amount of brain uptake. This uptake, graphically illustrated in Fig. 2, presumably occurs by in uiuo conversion of the cation to a neutral form. This conversion could occur via in vivo reduction of Tc(II1) to Tc(II), a phenomenon which is now well established,@) or by ion pair formation with a lipophilic anion. In either case, brain uptake of Tc(II1) cations is a complicated process that is dependent on
in vivo reaction mechanisms that are not clearly understood. (vi) Given the possible importance of in vivo redox reactions, plus the known requirement that a complex must have intermediate lipophilicity in order to cross the BBB, it is not at all surprising that few systematic trends in brain uptake are observed as the X groups of tr -phllT~‘*rD~X,]+ and tr -pAmTcrrD,X,]O complexes are varied. When X is varied from Cl to Br to I for a given D ligand, several properties of the complexes are altered simultaneously; the Tc(II1) cations become easier to reduce’“*“) to the more lipophilic Tc(I1) forms, and both the Tc(II1) and Tc(I1) complexes become inherently more lipophilic. For the three Tc(II1) DMPE cations tr-Pg”Tc’u(DMPE)J,]+ (X = Cl, Br, I), increasing the lipophilicity and reducibility by varying Cl to Br to I results in the expected increased brain uptake. However, for the pair of Tc(II1) DIARS cations tr-~Tc”‘(DIARS),_YJ+ (X = Cl, Br) increasing the lipophilicity and reducibility by varying Cl to Br leads to reduced brain uptake, perhaps because both the tr~Tc”‘~“(DIARS)~ Br,]+” complexes are excessively lipophilic. Neither of the tr-~Tcn’~*‘(DIARS)2Br,]+” complexes exhibit significant brain uptake. For the three Tc(I1) DMPE species tr~Tc**(DMPE)J,]~ (X = Cl, Br, I), increasing the lipophilicity increases brain uptake as expected. However, for the three Tc(I1) DPPE species trP~“Tc~~(DPPE),X,~” (X = Cl, Br, SCN) brain uptake appears to be controlled by blood binding; only the Br complex exhibits any significant brain uptake since it is the only one of the three agents that is not tightly bonded to blood components (Table 1).
Conchsions The data of Table 1 establish that *Tc complexes other than those containing Tc(V) centers and Tc=O
509
Neutral 99mTc(II)complexes
01 0
t 6
1P
19
24
30
3s
42
49
94
4
so
Time (Mln)
Fig. 2. Time dependent brain uptake in rats for three %Tc species.
linkages can indeed cross the BBB. If a *Tc complex is neutral and lipophilic, but not so lipophilic that it is tightly bonded to blood proteins, then it can cross the BBB and thus provide the chemical foundation for development of a *Tc brain perfusion imaging agent. The class of *Tc(II) complexes trph”T~~‘&&]~, where D represents a chelating ditertiary phosphine or arsine ligand and X represents a halide or pseudohalide ligand, contains members which meet these conditions and which correspondingly exhibit brain uptake in rats. Surprisingly, several of the cationic *Tc(III) analogs, trphll’Tc1*iD2X2]+, also exhibit brain uptake presumably because in vivo they are converted to a neutral form. The most likely mechanism for this conversion is in vivo reduction of the *Tc(III) cation to the *Tc(II) neutral species. The properties of the tr$“‘Tct”‘t’D2 X2]+” couples that are relevant to their potential application in brain imaging (e.g. lipophilicity, redox potential, blood binding) depend on the nature of D and X in a complicated, interrelated fashion. These complicated interrelationships among biological and chemical properties constitute both a difficulty and an opportunity in developing the class of tr-p”rc”D,X$ complexes as potential *Tc brain perfusion imaging agents. Acknowledgements-Financial support from the National Institutes of Health, Grant No. CA-42179, Mallinckrodt Inc., and the IAEA (for a fellowship to M. Neves), is gratefully acknowledged. The authors also thank Mr J. Bugaj for assistance with the animal studies and Dr M. ElEtri for contirming the results obtained with the DIARS complexes.
References I. Coleman R. E., Drayer B. P. and Jasxcxak R. J. J. Nucl.
Med. 23, 266 (1982). 2. Alaaraki N. P., Alderson P. O., Eckelman W. et al. Inoest. Radiol. 19, 570 (1984).
3. Magistretti P. L. (Ed.) FunctionalRadionuclideImaging of the Bruin (Raven Press, New York, 1983). 4. Holman B. L., Hill T. C., Polak J. F. ef. al. In RadionuclideImaging of the Brain (Ed. Hohnan B. L.) p. 163 (Churchill Livingstone, New York, 1985). 5. von Schulthess G. K., Ketz E., Schubiger P. A. and Bekier A. J. Nucl. Med. 26, 9 (1985). 6. Kung H. F., Tramposch K. M. and Blau M. J. Nucl. Med. 24, 66 (1983). 7. Holman B. L. Lee R. G. L., Hill T. C. et al. J. Nucl. Med. 25, (1984). 8. Loberg M: D. j. Nucl. Med. 21, 183 (1980). 9. Oldendorf W. H. Ann. Neural. 10. 207 (1981). 10. Oldendorf W. H. J. Nucl. Med. 19, 1182 (1978). 11. Loberg M. D., Corder E. H., Fields A. T. and Caller P. S. J. i&cl. Med. 20, 1181 (1979). 12. Kunn H. F.. Molnar K. M.. Billinas J. et al. J. Nucl. MedyU, 326 (1984). ’ 13. Lever S. Z., Bums H. D., Kervitsky T. M. et al. J. Nucl. Med. 26, P18 (1985) (Abstr.). 14. Kung H. F., Efange S., Yu C. C. er al. J. Nucl. Med. 26, P18 (1985) (Abstr.). 15. Volkert W. A., Hoffman T. J., Seger R. M. et al. Eur. J. Nucl. Med. 9, 511 (1984). 16. El1 P. J., Jarrit P. H., Hocknell J. M. L. et al. J. Nucl. Med. 26, 828 (1985) (Abstr.). 17. Anderson A., Holn S., Vorstrup S. et al. J. Nucl. Med. 26, 827 (1985) (Abstr.). 18. Nowotnik D. P., Canning L. R., Cumming S. A. et al. J. Nucl. Med. All. Sci. 29, 209 (1985) (Abstr.). 19. Chaplin S. B., Oberle P. O., Hoffman R. J. et al. J. Nucl. Med. 26, P18 (1985) (Abstr.). 20. Nechvatal G., Canning L. R., Cumming S. A. et al. J. Nucl. Med. All. Sci. 29, 208 (1985) (At&r.). 21. Jurisson S., Schlemper E. O., Troutner D. E. et al. J. Nucl. Med. All. Sci. 29, 201 (1985) (Abstr.). 22. Oldendorf W. H. In Handbook of Neurochemistry(Ed. Lattha A.) 2nd edn, p. 485. (Plenum Press, New York, 1982). 23. Oldendorf W. H. In FunctionalRadionuclideImaging of the Brain o. 1. (Raven Press. New York, 1983). 24. Libson K.: Bamett B. L. and Deutsch E.. Inor;. Chem. 22, 1695 (1983). 25. Hurst R. W., Heineman W. R. and Deutsch E. Inorg. Chem. 20, 3298 (1981). 26. Vanderheyden J-L., Libson K. Nosco D. L. et al. Int. J. Appl. R&at. Isot. 34, 1611 (1983).
510
Mm
NEWS et al.
27. Bandoli G., Mazzi U., Ichimura A. et al. Inorg. Chem. 23, 2898 (1984). 28. Vanderheyden J-L., Heeg M. J. and Deutsch E. Inorg. Chem. 24, 1666 (1985).
29. Glavan K. A., Whittle R. Johnson J. F. et al J. Am. Chem. Sot. 102, 2103 (1980). 30. Glavan K. A., Ph.D. Thesis, University of Cincinnati (1982).
Copyright 0
0883-2897/87 $3.00 + 0.00 1987 Pergamon Journals Ltd
Neutral Technetium(II)-99m Complexes as Potential Brain Perfusion Imaging Agents MARIA NEVES,* KAREN LIBSON and EDWARD
DEUTSCHT
Departments of Chemistry and Radiology, University of Cincinnati, Cincinnati, OH 45221, U.S.A. (Received 2 October 1986)
A series of eight neutral technetium(II)-Wm complexes of general formula rr-[99”r~*‘D,X,]~, where D represents a chelating ditertary phosphine or arsine ligand and X represents a halide or pseudohalide ligand, has been prepared and characterized by HPLC comparisons to the known %z analogs. Several members of this series exhibit significant brain uptake in rats, with maximum uptake (1.2% dose/brain at 1 min after injection) being exhibited by fr-[99mTc”(DIARS)2C12]0 where DIARS represents ophenylenebis(dimethylarsine). Surprisingly, several cationic Tc(III) analogs, tr-[99”rcD,X,]+, also exhibit significant brain uptake, presumably via a mechanism which involves in oivo reduction to the neutral Tc(II) form. The complicated interrelationships among the chemical and biological properties of the rr[~c’u~“D,X,]+~ couples, and the dependencies of these properties on the nature of D and X, provide possible means by which this class of complexes might be manipulated to yield an effective 99mTcbrain
perfusion imaging agent.
Introduction Numerous authors and groups have noted the great potential of SPECT (single-photon emission computed tomography) imaging techniques in the diagnosis and management of cerebrovascular disease.(‘4) In the context of brain perfusion imaging, 133Xehas been used for some time to determine regional cerebral blood flow,“) but tomographic imaging with this isotope requires specially dedicated SPECT instrumentation which is not optimal for other tomographic applications. Much more useful results have been obtained in the past few years with the lz31 labelled amines 1Z31-IMPand ‘231-HIPDM(C6)(IMP = N-isopropyl-iodoamphetamine; HIPDM = N,N,N’trimethyl-N-(2-hydroxy-3-methyl-5-iodobenzyl)-1,3propanediamine). These amines exhibit high brain uptake and long cerebral retention times,“) and clinical studies with these agents have clearly demonstrated the ultimate utility of single-photon emitting radiopharmaceuticals that monitor regional cerebral blood flow.” However, since iz31must be produced in a cyclotron and possesses a physical half-life of 13.3 h, it is expensive and not universally available for daily use. Clearly, the ideal SPECT brain perfusion *IAEA Fellow on leave from LNETI, Instituto de Energie, Portugal. t Author to whom all correspondence should be addressed. 503
agent should be based on the readily available, inexpensive *Tc isotope.(a9) This has been widely recognized ever since Oldendorfs 1978 calI for a new class of %Tc radiopharmaceuticals that would penetrate the intact blood-brain barrier (BBB) and have prolonged retention in the brain. While %Tc brain perfusion imaging agents have been systematically sought since about 1978,(i”~i1)it has only been in the last three years that significant progress towards such agents has been achieved. Neutral Tc(V) complexes, containing derivatives of the tetradenat*N, S2 dithiol ligand HSCH,CH,NHCH,CH,NHCH,CH,SH, have been shown to cross the BBB, but the underivatized complexes are not sufficiently retained in the brain. Attachment of pendant amine functionalities to the core Tc(V) complexes leads to significant brain retention in primates,“3~‘4Jsimilar to the brain retention observed for the 1231labelled amines.@-‘) A totally different class of neutral Tc(V) complexes, containing derivatives of the tetradentate_N, bis(oxime) ligand HON= CHCH2 NHCH,CH, CH, NHCH,CH=NOH, has also been shown to cross the BBBuS) and several of these complexes are retained in the brain long enough to allow clinically useful SPECT images to be obtained in humans.““‘*) Interestingly, these complexes do not contain pendant amine functionalities. In fact, derivatization of the core Tc(V) complex with pendant amine functionalities is detrimental to brain
504
MIWA NIMS et al.
uptake and retention.(19*M) Among the bis(oxime)Tc(V) complexes, those that are retained in the brain differ from those that are not, solely by the distribution of methyl groups about the periphery of the complex. (“Jo) The biological mechanisms of action of these agents have yet to be elucidated.(2’) Thus, all of the %Tc brain pe rf u sion imaging agents evaluated to date are neutral, lipophilic complexes of technetium(V) which contain Tc==O linkages. It is known that a molecule must be neutral and lipophilic in order to passively diffuse across the BBB.(22*23) But since almost nothing is known about the biological mechanisms of action of the %Tc agents, it is not all clear whether either the Tc(V) oxidation state or the Tc=O linkage, is necessary in order for a *Tc complex to cross the BBB. In order to probe this issue we have evaluated a series of neutral, lipophilic complexes of Tc(I1) (which do not contain T& linkages) for their ability to cross the BBB in rats. These complexes have the general formula fr-[ *Tc” D, X, 1’ wher e D represents a chelating ditertiary phosphine or arsine ligand and X represents a halide or pseudohalide ligand.(%“) Moreover, in order to obtain some information about the role of in viuo redox reactions in determining the biodistributions of these complexes,(2*) we have similarly evaluated some of the analogous Tc(II1) cations tr-PgmT~‘i’&X~] +. The results of this comparative study are reported herein.
(X = Cl, Br) were generated in siru by chemical reduction of the corresponding Tc(II1) complex, and then were characterized by the well established visible-u.v. spectrophotometric parameters of these Tc(II) complexes. (24,25)No ?Ic standards were available for I~-[CA”TC”(DMPE),I,~+ and tr[99”Tc”(DMPE), I,]‘; these complexes were indirectly characterized by their syntheses, which are analogous to those of the bromo complexes, and by comparing their HPLC retention times to those of the corresponding chloro and bromo analogs. AH *Tc complexes were prepared from “no carrier added” wmTcO; obtained by elution of a commercial 99Mo/99”T~ generator with normal saline. If the presence of chloride would interfere with preparation of the 99mTccomplex, the *TcO; was freed from chloride by chromatographic separation as (nC4Hs)4pT~04 on a C, or Cl* reversed phase cartridge, followed by elution with absolute ethanol; this procedure has been described for the purification of 186ReOq.(2*)The preparations of the 9h”rc(III) complexes were generally performed as previously detailed for tr-[99rnT~‘*i(DMPE)~X,]+ (X = Cl, Br);‘26) all syntheses were conducted anaerobically using a dry heat bath to maintain the appropriate reagents at the desired temperature in borosilicate vials (Wheaton) sealed with Teflon lined caps. After being heated for the desired length of time these vials are slowly cooled to room temperature. Cautiowrugid’ cooling the vial can cause it to break with explosive
Materials and Methods
force; borosilicate serum vials are used to minimize this hazard. Most of the 99mTc(II) complexes were pre-
Unless otherwise noted all chemicals were of reagent grade. Solvents used in HPLC analyses were of HPLC grade. The o-phenylenebis(dimethylarsine) (DIARS), 1,2-bis(dimethylphosphino)ethane (DMPE), and 1,2-bis(diphenylphosphino)ethane (DPPE) ligands were obtained from Strem Chemical Co. and were used without further purification. The air sensian inert atmosphere, and solution of these ligands were stabilized with strong acids (HCl, HBr or HClO,) as previously described.(26) HPLC (high performance liquid chromatographic) analyses were conducted using a previously detailed(26) apparatus equipped with both U.V. (254nm) and radiometric (centered at 140 keV) detectors and an integrating recorder. The use of these two detection systems allows the positive identification of “unknown” *Tc complexes by simultaneous injection of the “unknown” 99m Tc complex and a “known” 99Tc complex which has been characterized by classical chemical techniqueso6) Observation of simultaneously eluting radiometric and U.V. HPLC peaks confirms that the %Tc “unknown” and the well characterized qc complex are the same chemical species. Well characterized samples of tr-[99Tc’u(DIARS),XJ+ (X = Cl, Br), ~~-[~~Tc”‘(DPPE)~X,]+ (X = Cl, Br), ~~-~c~(DPPE)~(NCS)J~ and tt-~c’u(DMPE),X,] + (X = Cl, Br) were available from previous studies 9(U,27,29*M) while samples of ~~-~c”(DMPE),X,]~
pared by chemical reduction of the corresponding %Tc(III) complex. Details of the various *Tc(III) and *Tc(II) sy nthetic procedures are specified below. The radiochemical purities of the final preparations were monitored by HPLC using a silica based cation exchange column (Vydac, 10 micron: 150 x 4.6 mm) and a 90/10 methanol/water mobile phase that was 0.04 M in (n-C,H,),NH,PO,. In all cases, except one, the radiochemical purity was greater than 95%; in the case of the single exception, tr -pTc” (DPPE), (NCS), Jo, the radiochemlcal purity was only 80% and the primary radiochemical Before being used in animal impurity was *TcO;. studies, all *Tc preparations were adjusted to a pH value between 6 and 8 using either H3 PO4 or Na, PO,, filtered through either a 0.45 or a 0.20 micron sterile filter (Gelman), and then diluted with saline so that the final concentration of ethanol was less than 10%. The various %Tc agents were screened for their ability to cross the BBB in anesthetized (Metofane), female Sprague-Dawley rats of ca 200 g weight. At each of various times after intrajugular injection of the agent, three rats were sacrificed by cervical dislocation, blood samples were collected, and the brains were excised. Tissue samples were weighed (average brain weight = 1.5 g) and assayed for *Tc by standard techniques. The brain and blood uptakes are expressed in Table 1 as the average (n = 3)
_
0.060 (I)
0.32 (2) 0.050(l) 0.23 (2) 0.19(l) 0.15(l) 0.08 (1) 0.10(l) 0.040(l)
0.62 (3) -
0.66 (3) 0.07(l) 0.33 (2) 0.19 (2) 0.15(l) 0.14(l) -
0.79 (3) 0.09 (I) 0.38 (3) 0.31 (2) 0.26 (2) 0.14(l) 0.24 (I) 0.10(l) 0.12(l)
0.64 (3) 0.060(I) 0.20 (2)
-
0.040(I)
0.19 (2) 0.18(l) 0.14(l) O.IO(l)
0.23 (I) 0.06(l) 0.10(l) 0.17(l) 0.18(l)
0.28 (1) 0.07 (I) 0.09 (I) 0.16(l) 0.17(l)
0.08(l) -
0.30 (2) 0.09(l) 0.07 (I) 0.15 (I) 0.16 (I)
0.32 (2) 0.11 (I) 0.09(l) 0.17(l) 0.19(l)
0.040 (I)
0.040 (I)
0.050 (I)
20
agents
0.060(l)
6
of Tc
2
4
uptakes
t=l
Brain
I. Brian and blood
0.12(2) 0.040(l) 0.26 (2) 0.22(l) 0.18 (I) 0.07 (I) 0.09 (1) 0.040(l)
0.15(l) 0.04(l) 0.11 (I) 0.18(l) 0.20 (I)
0.040(l)
6Omin
2
(3) (3) (2) (3) (3) (5)
(2) (2) (1) (I) (2)
0.24 (2) -
-
0.48 (2) 0.28 (2) 0.18(l) 0.24 (3)
0.19 (2) -
4
Me,PCH,CH,PMe,;
6
20
0.24 (2) 0.15(l) 0.06 (I) 0.15(l) 0.26 (2) 3.22 (4) 0.28 (2) 0.75 (3)
0.34 (2) -
6Omin
0.08 (I) 0.040(l) 0.14(l) 0.22 (2) 0.30 (2) 3.39 (5) 0.70 (3) 0.36(2)
0.67 (I) 0.040(l) 0.08 (I) 0.08 (I) 0.14(i)
0.81 (2)
DPPE
represents
1,2-bis(diphenyl-
At least three rats were sacrificed
-
0.12(l) 0.20 (2) 0.24 (2) 3.31 (5)
0.79 (I) 0.10 (I) 0.05 (I) 0.06 (I) 0.12(l)
I .02 (3) I .Ol (2) 0.16(l) 0.09 (I) 0.05 (I) 0.18 (2)
I .38 (3)
Blood
digit is given in parentheses.
1.70(4)
0.64 0.35 0.25 0.40 0.40 3.50
1.21 0.38 0.30 0.50 0.42 4.20 0.34 2.61
(5) (3) (2) (4) (3) (6) (3) (5)
I .22 0.24 0.12 0.07 0.30
I .78 (4)
of time
I .45 (4) 0.34 (3) 0.21 (2) 0.12(l) 0.38 (3)
2.20 (4)
1=I
in rats as a function
‘Expressed as percent injected dose per gram of tissue at the time of assay. The standard deviation of the last significant at each time point. Data are corrected for the physical decay of 99”rc. 1,2-bis(dimethylphosphino)ethane, bDIARS represents o-phenylenebis(dimethylarsine), o-C,H,(AsMe&; DMPE represents phosphino)ethane, Ph,PCH,CH2PPh,. (Me = CH,; Ph = C,H,).
Aneatb
Table
506
MAIU Naves et at.
percentage of injected dose per gram of tissue at any time of assay. Synthesis of 99mTccomplexes (i) tr-[99”Tc”‘(DIARS), Cl,]+. An HCl stabilized 1% solution of DIARS is prepared by combining 30 ~1 of DIARS ligand with 0.5 mL 1.0 M HCl and 2.5 mL ethanol. Fifty microliters of the DIARS solution, 30 PL of 6 M HCl, and 0.40 mL of 95% ethanol are added to the desired amount of 99”Tc0; in 1.0 mL of saline. This reaction mixture is heated at 140°C for 20 min. (ii) tr-[99mTc”(DIARS)2 Br,]+. To the desired amount of (n-C,H,),p”TcO; in 1.OmL absolute ethanol are added 0.30 mL of 9 M HBr and 40 microliters of a 1% solution of DIARS in absolute ethanol. This reaction mixture is heated at 100°C for 20 min. (iii) tr-[99mTc’1’(DMPE)2 Cl,]+. An HCl stabilized 1% solution of DMPE is prepared by combining 30 ~1 of DMPE ligand with 0.5 mL 1.0 M HCl and 2.5 mL ethanol. Fifty microliters of this DMPE solution and 0.80mL of 95% ethanol are added to the desired amount of %TcO; in 1.OmL of saline. This reaction mixture is heated at 140°C for 30 min. (iv) tr-[99mTc”(DMPE)2Br2]+. An HBr stabilized 1% solution of DMPE is prepared by substituting HBr for HCl in the above formulation. Fifty microliters of this DMPE solution, 0.8 mL of 3 M NaBr, and 50 ~1 of 4.5 M HBr are added to the desired amount of (n - CqH9)q~mT~04 in 0.70 mL ethanol. This reaction mixture is heated at 140°C for 30 min. (u) tr-[W”Tc”(DMPE),Z,)+. An HC104 stabilized 1% solution of DMPE is prepared by combining 30 ~1 of DMPE with 0.5 mL of 0.25 M HC104 and 2.5mL ethanol. One hundred microliters of this solution and 100 PL of 3 M NaI are added to the desired amount of (n-C,H,),N99mTc0, in 1.0 mL ethanol. This reaction mixture is heated at 145°C for 15min. (oi) tr-[ w”‘Tc”(DZARS)2 Cl,]“. (A) The tr ~Tc”‘(DIARS)&~~] + preparation is brought to pH 5 with NaOH, 3 mg of NaBH, are added, and the reaction mixture is allowed to stand at room temperature for 30 min. (B) Alternatively, the %Tc(III) preparation is brought to pH 4-5 with NaOH, either 100 PL of 2-mercaptoethanol or 100 PL of 1.OM ascorbic acid are added, and the reaction mixture is allowed to stand at room temperature for 2-3 min. (vii) tr-[99mTc’f(DIARS)2Br2]o. (A) The tr~Tc”‘(DIARS), Br,]+ preparation is brought to pH 6 with NaOH, 100 PL of 2-mercaptoethanol are added, and the reaction mixture is heated at 90°C for 20 min. (B) Alternatively, the *Tc(III) preparation is brought to pH4 with NaOH, 1OOpL of 1.0 M ascorbic acid are added, and the reaction mixture is allowed to stand at room temperature for 30min. (viii) tr-[99”‘Tc”(DMPE)2 CIJ”. The corresponding *Tc(III) preparation is treated with NaBH, as described above for the DIARS analog.
(ix) tr-~‘“Tc’(DMPE)zBrz]o. The corresponding *Tc(III) preparation is treated with 2-mercaptoethanol as described above for the DIARS analog. (x) tr-[wmTcr’(DMPE)212]o. The corresponding “mTc(III) preparation is treated with 2-mercaptoethanol as described above for tr~9mTc*‘(DIARS),Br,]o except that the reaction mixture is heated for 30 min. (xi) tr-[99mTc’(DPPE),C&]o. One hundred microliters of a 0.01 M solution of DPPE in absolute ethanol and 0.2 mL 6 M HCl are added to the desired amount of (n-C,H,),~mTcO, in l.OmL ethanol. This reaction mixture is hea@d at 140°C for 45 min. (xii) tr-[99mTcr’(DPPE)2Br,]o. One hundred microliters of a 0.04 M solution of DPPE in absolute ethanol and 0.2 mL 4.5 M HBr are added to the desired amount of (n-C, H,),N*TcO, in 1.0 mL ethanol. This reaction is heated at 140°C for 45 min. (xiii) tr-[99”‘Tc”(DPPE),(NCS)2]o. One hundred PL of a 0.04 M solution of DPPE in absolute ethanol, 200 PL of 1.OM ascorbic acid, 1.0 mL water, and 100 p L of 1.OM NaSCN are added to the desired amount of (n-C,H,),N*TcO, in l.OmL ethanol. This reaction mixture is heated at 100°C for 20 min.
Results and Discussion Chemistry
The cationic tr-[99mTd”D2X2]+ complexes with D = DIARS are prepared by procedures similar to those detailed for preparation of the complexes with D = DMPE.(26) These syntheses involve the use of the D ligand to reduce %TcOi, and to coordinate to the resulting technetium center, in the presence of HX. Careful control of reaction parameters (time, temperature, concentration of HX and D, solvent etc.) is required in order to obtain the desired 99”Tc(III) complexes in greater than 95% radiochemical purity.(26) The neutral [9A”Tc11D2X2]o complexes with D = DIARS or DMPE are prepared by chemical reduction of the analogous %Tc(III) agent, and, again, careful control of reaction parameters is required in order to obtain radiochemical purities in excess of 95%. In addition to the time and temperature of reaction, pH, solvent, etc. the successful preparation of the %Tc(II) complexes is dependent on the choice and concentration of the reductant. This in turn is dependent on the E”’ value of the Tc(III)/Tc(II) redox couple, which is controlled by the identity of the D and X ligands of the trp9”TcD2X2]+Io core.(24*25) For a given D ligand, Tc(II1) complexes with X = Br (or I) are easier to reduce than those with X = Cl. Thus, preparation of the DMPE Tc(I1) complexes with X = Br (or I) utilizes the relatively mild reductants 2-mercaptoethanol or ascorbic acid, while preparation of the chloro analog requires a use of the potent reductant NaBH,. Borohydride is too strong a reductant for the preparation of the DIARS and DMPE complexes with X = Br (or I), and its use in
Neutral *Tc(II) complexes these cases leads to large amounts of unidentified reaction products. Even in the case of the two chloro analogs, under many reaction conditions NaBH, can lead to several unidentified reaction products in addition to the desired Tc(I1) complex. For a given X ligand, complexes with D = DPPE are easier to reduce than those with D = DIARS or DMPE. Therefore, the original conversion of %TcO; to the tr-Pg”TcD,X2]+‘0 core leads to Tc(II1) complexes for D = DIARS or DMPE (which then must be reduced to Tc(I1) in a subsequent step), but leads directly to Tc(I1) complexes for D = DPPE. In fact, generation of Tc(III)-DPPE complexes from the %TcO; reaction mixture require a subsequent oxidation of the initial Tc(I1) product. Thus, the neutral Tc(II) tr-ph”Tc”(DPPE), X,]” (X = Cl, Br) complexes are directly prepared by the DPPE reduction of %TcO; using procedures similar to those detailed for preparation of the cationic Tc(II1) tr[99mTc”‘(DMPE),X,]+ complexes.(26’ Preparation of the tr-~mTc1’(DPPE)2(NCS)2]o complex is a special case because of the inherent instability of the NCS ligand under the conditions used to reduce *TcO; . To keep the reaction conditions as mild as possible, ascorbic acid is added to the original %TcO; reaction mixture. But even with this expedient, only about an 80% yield of the desired agent is obtained. This agent is of interest in the context of radiopharmaceutical development because the N-bonded NCS ligand so greatly stabilizes the Tc(I1) oxidation state that tr-[99mTc”(DPPE)2(NCS),l” can be isolated and manipulated in the presence of air.(27) Chromatography
When macroscopic amounts of qc complexes are used, standard reversed phase chromatography on CB or C,, columnso6’ successfully separates the cationic Tc(II1) species from the neutral Tc(I1) species. However, in our hands, this type of chromatography is not successful when used with 99mTccomplexes prepared at the “no carrier added” concentration level, presumably because of adventitious oxidation of the Tc(I1) species on the surface of the column support. Thus, when either of tr-[99mTc1”D2X2]f or tr~Tc” D2X2]o is subjected to reversed phase chromatography, only a single peak with the retention time of tr-~Tc”’ D2X2]+ is observed (after the known effect of injected mass on retention time is taken into account).(26) This problem is avoided by the use of cation exchange chromatography wherein the neutral 99mTc(II) complexes exhibit retention times of 4-5 min and the cationic 99mT~(III) complexes exhibit retention of 10-13 min. This type of chromatography therefore allows the %Tc(III) and %Tc(II) species to be easily distinguished even at the very low “no carrier added” concentrations of technetium encountered in 99Mo/99”Tc generator eluents. HPLC provides the crucial link between Yc and %Tc chemistry that is necessary for the preparation
so7
and biological evaluation of *Tc agents of known composition and structure. This link is effected by the simultaneous use of two detectors; U.V. detection to monitor macroscopic amounts of 99Tc complexes, and radiometric detection to monitor 99”Tc complexes.(26’ Characterized samples of all the trp9TcD2X2]+‘” cores used in this work, except trwere available from previous ~Tc(DMPE)~I~]+‘~, studies.‘24~25.27*“~‘r’) Therefore, for these complexes, coelution of simultaneously injected tr-f’%D2X2]+” “knowns” and tr-pTcD,X,]+” “unknowns”, establishes that the two are indeed the same chemical species. The tr-Pg”Tc(DMPE)21z]+‘o complexes are identified by (a) their preparation via procedures analogous to those used to synthesize the known bromo complexes, but employing iodide rather than bromide in the reaction mixture, and (b) their HPLC retention times (RT) which are consistent with those observed for the chloro and bromo analogs (for tr-PgmTc”‘(DMPE),X,]+, RT = 8.8, 11.4 and 12.6 min X = Cl, Br and I respectively; for tr~Tc~~(DMPE),X,]O, RT =4.1, 4.6 and 5.2min for X = Cl, Br and I respectively). Biodistribution
Table 1 summarizes the brain and blood uptake values observed in rats as a function of time for various 99mT~complexes. There are several aspects of these data that are relevant to the development of %Tc brain perfusion imaging agents. (i) The biodistribution of *TcO; is included as a point of reference since this species is well known not to cross the BBB. At all times the brain uptake of %TcO; is indeed very low, even though the amount of *TcO,- in the blood is relatively high. *Tc(II) complex tr(ii) The neutral ~Tc”(DIARS),CI~]~ clearly exhibits significant brain uptake. At 1 min after injection approximately 1.2% of the injected dose is in the brain (0.79% dose/g X 1.5 g/brain). This uptake is undoubtedly due to passive diffusion of the neutral Tc(I1) complex across the BBB. As the Tc(I1) agent clears from the blood, passive diffusion also carries it out of the brain leading to a relatively short retention time of the brain for tr -P)“Tc”(DIARS)* Cl,]O. Note that from 20 to 60min after injection the brain/blood ratio is relatively constant even though the amount of activity in the blood decreases by a factor of three. This implies that at long times after injection the amounts of tr-[99mTc’1(DIARS),C12]0 in the brain and blood are related by a simple equilibrium, presumably involving passive diffusion of this agent across the BBB. (iii) The neutral fr-~Tc”(DMPE)rClJ’ complex also exhibits brain uptake, but considerably less than does the DIARS analog. This lower brain uptake presumably results from the DMPE complex being less lipophilic than the DIARS analog, and thus less capable of diffusing across the BBB. However, the
508
MARU NEW et al.
tr-[ ggmTc”lDIARS)2
8
0.6 0.4 0.3 0.2 0.1 0
0
1
6
12
16
24
30
36
.?
46
6.
60
Time (Hid
Fig. 1. Time dependent brain uptake in rats for three related rr-ph”TcnD,Cl,P complexes. very lipophilic complex tr-~mTcll(DPPE)zCIJo exhibits essentially no brain uptake because its excessive lipophilicity causes it to bind tightly to blood components. Note the very high levels of this agent in the blood even 1 h after injection (Table 1). Thus it appears that in order to exhibit reasonable brain uptake, a neutral (Tc(I1) complex must be sufficiently lipophilic to cross the BBB, but not so lipophilic so as to bind tightly to blood components. Figure 1 compares the time dependent brain uptakes observed for the three related tr-PgmTcilD,CIJO complexes with D = DIARS, DMPE and DPPE, graphically illustrating the intermediate uptake of the DMPE complex. (iv) Figure 1 also shown that while the DIARS complex continually washes out of the brain, the DMPE complex first washes out and then appears to slowly reaccumulate. While this slow increase in brain uptake of tr-~Tcl’(DMPE),CIJo at long times is barely of statistical significance, the phenomenon is reproducible and also occurs for the bromo and iodo DMPE analogs (Table 1). Thus, the slow reaccumulation of the tr-~Tc”(DMPE),XJ” (X = Cl, Br, I) complexes in the brain appears to be a real effect which warrants further investigation. (v) As expected, the cationic *Tc(III) complex tr-~Tc(DMPE)2C12]+ exhibits no detectable brain uptake. Since this agent bears a formal charge, as does *TcG;, it is unable to traverse the BBB. Surprisingly, injection of the cationic analog tr~Tc(DIARS)~C~J+ does lead to a significant amount of brain uptake. This uptake, graphically illustrated in Fig. 2, presumably occurs by in uiuo conversion of the cation to a neutral form. This conversion could occur via in vivo reduction of Tc(II1) to Tc(II), a phenomenon which is now well established,@) or by ion pair formation with a lipophilic anion. In either case, brain uptake of Tc(II1) cations is a complicated process that is dependent on
in vivo reaction mechanisms that are not clearly understood. (vi) Given the possible importance of in vivo redox reactions, plus the known requirement that a complex must have intermediate lipophilicity in order to cross the BBB, it is not at all surprising that few systematic trends in brain uptake are observed as the X groups of tr -phllT~‘*rD~X,]+ and tr -pAmTcrrD,X,]O complexes are varied. When X is varied from Cl to Br to I for a given D ligand, several properties of the complexes are altered simultaneously; the Tc(II1) cations become easier to reduce’“*“) to the more lipophilic Tc(I1) forms, and both the Tc(II1) and Tc(I1) complexes become inherently more lipophilic. For the three Tc(II1) DMPE cations tr-Pg”Tc’u(DMPE)J,]+ (X = Cl, Br, I), increasing the lipophilicity and reducibility by varying Cl to Br to I results in the expected increased brain uptake. However, for the pair of Tc(II1) DIARS cations tr-~Tc”‘(DIARS),_YJ+ (X = Cl, Br) increasing the lipophilicity and reducibility by varying Cl to Br leads to reduced brain uptake, perhaps because both the tr~Tc”‘~“(DIARS)~ Br,]+” complexes are excessively lipophilic. Neither of the tr-~Tcn’~*‘(DIARS)2Br,]+” complexes exhibit significant brain uptake. For the three Tc(I1) DMPE species tr~Tc**(DMPE)J,]~ (X = Cl, Br, I), increasing the lipophilicity increases brain uptake as expected. However, for the three Tc(I1) DPPE species trP~“Tc~~(DPPE),X,~” (X = Cl, Br, SCN) brain uptake appears to be controlled by blood binding; only the Br complex exhibits any significant brain uptake since it is the only one of the three agents that is not tightly bonded to blood components (Table 1).
Conchsions The data of Table 1 establish that *Tc complexes other than those containing Tc(V) centers and Tc=O
509
Neutral 99mTc(II)complexes
01 0
t 6
1P
19
24
30
3s
42
49
94
4
so
Time (Mln)
Fig. 2. Time dependent brain uptake in rats for three %Tc species.
linkages can indeed cross the BBB. If a *Tc complex is neutral and lipophilic, but not so lipophilic that it is tightly bonded to blood proteins, then it can cross the BBB and thus provide the chemical foundation for development of a *Tc brain perfusion imaging agent. The class of *Tc(II) complexes trph”T~~‘&&]~, where D represents a chelating ditertiary phosphine or arsine ligand and X represents a halide or pseudohalide ligand, contains members which meet these conditions and which correspondingly exhibit brain uptake in rats. Surprisingly, several of the cationic *Tc(III) analogs, trphll’Tc1*iD2X2]+, also exhibit brain uptake presumably because in vivo they are converted to a neutral form. The most likely mechanism for this conversion is in vivo reduction of the *Tc(III) cation to the *Tc(II) neutral species. The properties of the tr$“‘Tct”‘t’D2 X2]+” couples that are relevant to their potential application in brain imaging (e.g. lipophilicity, redox potential, blood binding) depend on the nature of D and X in a complicated, interrelated fashion. These complicated interrelationships among biological and chemical properties constitute both a difficulty and an opportunity in developing the class of tr-p”rc”D,X$ complexes as potential *Tc brain perfusion imaging agents. Acknowledgements-Financial support from the National Institutes of Health, Grant No. CA-42179, Mallinckrodt Inc., and the IAEA (for a fellowship to M. Neves), is gratefully acknowledged. The authors also thank Mr J. Bugaj for assistance with the animal studies and Dr M. ElEtri for contirming the results obtained with the DIARS complexes.
References I. Coleman R. E., Drayer B. P. and Jasxcxak R. J. J. Nucl.
Med. 23, 266 (1982). 2. Alaaraki N. P., Alderson P. O., Eckelman W. et al. Inoest. Radiol. 19, 570 (1984).
3. Magistretti P. L. (Ed.) FunctionalRadionuclideImaging of the Bruin (Raven Press, New York, 1983). 4. Holman B. L., Hill T. C., Polak J. F. ef. al. In RadionuclideImaging of the Brain (Ed. Hohnan B. L.) p. 163 (Churchill Livingstone, New York, 1985). 5. von Schulthess G. K., Ketz E., Schubiger P. A. and Bekier A. J. Nucl. Med. 26, 9 (1985). 6. Kung H. F., Tramposch K. M. and Blau M. J. Nucl. Med. 24, 66 (1983). 7. Holman B. L. Lee R. G. L., Hill T. C. et al. J. Nucl. Med. 25, (1984). 8. Loberg M: D. j. Nucl. Med. 21, 183 (1980). 9. Oldendorf W. H. Ann. Neural. 10. 207 (1981). 10. Oldendorf W. H. J. Nucl. Med. 19, 1182 (1978). 11. Loberg M. D., Corder E. H., Fields A. T. and Caller P. S. J. i&cl. Med. 20, 1181 (1979). 12. Kunn H. F.. Molnar K. M.. Billinas J. et al. J. Nucl. MedyU, 326 (1984). ’ 13. Lever S. Z., Bums H. D., Kervitsky T. M. et al. J. Nucl. Med. 26, P18 (1985) (Abstr.). 14. Kung H. F., Efange S., Yu C. C. er al. J. Nucl. Med. 26, P18 (1985) (Abstr.). 15. Volkert W. A., Hoffman T. J., Seger R. M. et al. Eur. J. Nucl. Med. 9, 511 (1984). 16. El1 P. J., Jarrit P. H., Hocknell J. M. L. et al. J. Nucl. Med. 26, 828 (1985) (Abstr.). 17. Anderson A., Holn S., Vorstrup S. et al. J. Nucl. Med. 26, 827 (1985) (Abstr.). 18. Nowotnik D. P., Canning L. R., Cumming S. A. et al. J. Nucl. Med. All. Sci. 29, 209 (1985) (Abstr.). 19. Chaplin S. B., Oberle P. O., Hoffman R. J. et al. J. Nucl. Med. 26, P18 (1985) (Abstr.). 20. Nechvatal G., Canning L. R., Cumming S. A. et al. J. Nucl. Med. All. Sci. 29, 208 (1985) (At&r.). 21. Jurisson S., Schlemper E. O., Troutner D. E. et al. J. Nucl. Med. All. Sci. 29, 201 (1985) (Abstr.). 22. Oldendorf W. H. In Handbook of Neurochemistry(Ed. Lattha A.) 2nd edn, p. 485. (Plenum Press, New York, 1982). 23. Oldendorf W. H. In FunctionalRadionuclideImaging of the Brain o. 1. (Raven Press. New York, 1983). 24. Libson K.: Bamett B. L. and Deutsch E.. Inor;. Chem. 22, 1695 (1983). 25. Hurst R. W., Heineman W. R. and Deutsch E. Inorg. Chem. 20, 3298 (1981). 26. Vanderheyden J-L., Libson K. Nosco D. L. et al. Int. J. Appl. R&at. Isot. 34, 1611 (1983).
510
Mm
NEWS et al.
27. Bandoli G., Mazzi U., Ichimura A. et al. Inorg. Chem. 23, 2898 (1984). 28. Vanderheyden J-L., Heeg M. J. and Deutsch E. Inorg. Chem. 24, 1666 (1985).
29. Glavan K. A., Whittle R. Johnson J. F. et al J. Am. Chem. Sot. 102, 2103 (1980). 30. Glavan K. A., Ph.D. Thesis, University of Cincinnati (1982).