The biomedical chemistry of technetium and rhenium Jonathan R. Dilwortha and Suzanne J. Parrottb a Inorganic Chemistry Laboratory University of Oxford South Parks Road Oxford UK OX1 3QR b Department of Biological and Chemical Sciences University of Essex Wivenhoe Park Colchester UK CO4 3SQ This review describes recent developments in the chemistry of both first and second generation 99m-technetium-based imaging agents. The material is presented according to the biological target for the agent and where possible actual images are presented to indicate the type of information available to the clinician. Beta emitting isotopes of rhenium offer a possible method for the in situ treatment of cancerous tissue using analogous targeting strategies to those for technetium.Recent developments in the relevant coordination chemistry of rhenium and their extension to in vitro and in vivo studies are presented. 1 Introduction Modern medicine demands progressively more sophisticated methods for the accurate diagnosis of disease states and there is a massive worldwide research effort into developing and improving imaging techniques. Images can be produced either by measuring the absorption of externally applied radiation (e.g. X-ray ultrasound MRI imaging) or by administering a small amount of a radioactive compound and detecting the radiation escaping from the body. All of these techniques enhanced by computerised tomographic methods can produce remarkably high quality images of locations deep inside the body.To an extent these techniques are complementary and the method selected will depend not only on the type of image required but also on other factors such as the availability of equipment. Nuclear medicine has traditionally been favoured for imaging biological function and while some of this area is being taken over by developments in MRI SPECT (single photon emission computerised tomography) and PET (positron emission tomography) remain the methods of choice for imaging low capacity high density receptors. The use of external radiation for treatment of cancer is extremely well developed but now the greater ability to target radiopharmaceuticals has led to the Jon Dilworth graduated from the University of Oxford in 1967 and then studied for a DPhil degree at the Unit of Nitrogen Fixation University of Sussex with Professor Joseph Chatt.After a period on the permanent staff of the Unit he took the Chair of Chemistry at the University of Essex in 1985 and has now taken a post in the Inorganic Chemistry Laboratory University of Oxford (September 1997). His research interests involve the applications of coordination chemistry in biology medicine and catalysis. Outside he hopes one day to master a topspin tennis backhand and a reliable method to get out of bunkers. Suzanne Parrott Jon Dilworth O CH3 N B N O O N Tc O H N H N N O O CI possibility of using b-emitting compounds to deliver radiation in situ to cancer sites. In this review we attempt to present a brief account of the major developments in the use of g-emitting technetium complexes for imaging and recent work directed towards producing b-emitting rhenium compounds for therapeutic use.This is given very much from a chemical perspective although we have wherever possible given examples of the type of diagnostic image that can be produced. The restriction of space means that it cannot be comprehensive and the material has been selected to provide what are hopefully interesting illustrations of the potential scope of the radioactive isotopes of Tc and Re for diagnosis or therapy. Technetium-99m although heavily used (over 90% of all diagnostic nuclear medicine imaging studies carried out worldwide use this isotope) is only one of a range of metallic radionucleides used for medical imaging or therapy but space restricts our coverage here to Tc and Re.There are several reviews available on Tc and Re chemistry1–4 and specialised texts5 and published conference proceedings6 provide more comprehensive surveys of the medical applications of these elements. 2 Technetium The element was first predicted by Mendeleev (ekamanganese number 43) and first isolated by Segr�e and Perrier in 1938. It was separated from a molybdenum target plate that had been bombarded with deuterons in the Berkeley cyclotron. Currently there are no fewer than 20 known isotopes (91Tc–110Tc) and seven nuclear isotopes.7 The 99mTc nuclear isotope is used for medical imaging due to its near ideal nuclear characteristics of a 6 h halflife and g-ray emission energy of 141 keV.The diagram in Fig. 1 shows the radioactive decay series involving the medically important Tc isotopes. The practical use of 99mTc for regular imaging depends totally on the ready availability of Suzanne Parrott received a PhD in Chemistry from the University of Essex in 1993 in the area of rhenium coordination chemistry. After working as a Post-doctoral Fellow at The duPont Merck Pharmaceutical Company in Massachusetts USA she returned to the University of Essex as a Senior Research Officer and is currently a Lecturer in Chemistry at the University. Her research interests include rhenium and technetium coordination chemistry with applications to nuclear medicine.43 Chemical Society Reviews 1998 volume 27 99mTc beta 66 h 88.75% 99Ru gamma 6 h 142 keV 99Mo beta 66 h 12.5% beta 2.14 x 105 y 0.292 meV 99Tc Fig. 1 the isotope using the 99Mo/99mTc generator developed in Brookhaven in the early 1960s. This consists of [99MoO4]2 absorbed at the top of an alumina ion exchange column. The 99Mo decays continuously to 99mTc which can be preferentially eluted with physiological saline (0.15 m NaCl) over a period of 7–10 days. A typical eluate used to prepare an imaging agent will be about 1027 to 1028 m in [TcO4]2. The chemical consequences are that the synthesis of radiopharmaceuticals has to proceed in very dilute aqueous solution directly from [TcO4]2. The very dilute nature of the [99mTcO4]2 solutions means that characterisation of complexes by routine spectroscopic and analytical methods is not possible and HPLC or other chromatographic methods with g detection are virtually the only way to monitor the chemistry.The very long lived b-emitting 99Tc isotope is used (typically on a 10–20 mg of technetium scale) to isolate technetium complexes and characterise them fully using the full range of spectroscopic techniques including X-ray crystallography. HPLC of the 99Tc complexes (UV and b-detection) is then used to infer the structures of the 99mTc analogues. The weak b-emitting properties of the 99Tc isotope means that complexes can be handled safely in conventional glassware with appropriate precautions. With the widespread use of nuclear reactors technetium is no longer a rare element and it has been estimated that 160 000 kg of technetium will in principle be available by the year 2000.It is a remarkable statistic that there is already more of this entirely artificial element available in the world than its stable naturally occurring congener rhenium! Very recently there have been reports of the preparation of the 94Tc isotope by irradiation of 94Mo in a cyclotron. Both the 94mTc and 94gTc isotopes are positron emitters. The ground state isomer also emits g radiation and offers the interesting possibility of also deploying SPECT. The 94mTc isotope has been used for PET imaging of the heart using an isocyanide complex (see section 2.3.2 on heart imaging below).2.1 Imaging techniques for technetium The eluate from the 99mTc generator described above is introduced by syringe via a septum into a vial containing the reagents necessary to produce the imaging agent. After a suitable incubation period the radiopharmaceutical is injected into the patient and after time for biodistribution to occur the image data is collected by a gamma camera equipped with a NaI scintillation detector and photomultiplier system (Fig. 2). The camera is rotated around the patient or a multidetector array is usedo create a tomographic image by use of a sophisticated computerised program which reconstruct the image from a series of projections. A successful imaging agent will generally direct in the order of 1–5% of the injected dose of activity to the target organ the bulk of the remainder generally being excreted via the kidneys.The total radiation dose from a technetium scan is comparable with that from a conventional X-ray. 2.2 Types of technetium imaging agents The first use of technetium for medical imaging was in 1961 and involved the use of [99mTcO4]2 for diagnosis of thyroid disease based on the principle that the pertechnetate anion would behave similarly to iodide known to be taken up by the thyroid. The biodistribution and targeting ability thus depended solely on the size and charge of the complex. This was the first of a Chemical Society Reviews 1998 volume 27 44 Fig. 2 A patient undergoing a technetium scan using a gamma camera. Reproduced with permission from Amersham International.series of the so-called ‘technetium essential’ or first generation agents. These are represented diagrammatically in Fig. 3(A) and such agents have been deployed with great success to image organs such as the heart the brain the kidney and the liver and are discussed in more detail below. However the growing demand for ever more specific agents has prompted the development of second generation agents [Fig. 3(B)]. Here the targeting capability resides in a biologically active molecule (BAM) covalently linked to an appropriate technetium complex. The BAM is typically a small peptide molecule which acts as an agonist or antagonist for a specific receptor site or a monoclonal antibody. The targeting ability of the BAM can be adversely affected by the presence of the technetium complex and the site of attachment to the BAM the size charge and lipophilicity of the conjugate and the length of the covalent linker all need to be optimised for maximum receptor binding.Tc Tc Tc BAM (C) (B) (A) Fig. 3 The ideal situation would be where the outer surfaces of the complex itself contain the groups necessary for receptor binding [Fig. 3(C)]. This approach is far more challenging in terms of the chemistry involved and developments are currently in the early stages. Examples of all three strategies are presented in the discussions of individual imaging agents below. 2.3 First generation technetium imaging agents 2.3.1 Brain imaging The dominant requirement for an agent that will accumulate in the brain is that it is capable of traversing the blood–brain barrier (BBB).Viable complexes must therefore be moderately lipophilic and have an overall neutral charge. Research at the University of Missouri in the 1980s demonstrated that a series of neutral amine–oxime complexes could readily be prepared directly from [TcO4]2 in the presence of SnCl2 as reducing agent. Further development at Amersham International led to the commercially successful Ceretec agent utilising the hexamethylpropyleneamineoxime proligand (HMPAO hexametazime) which loses three protons and forms a neutral square pyramidal TcV mono-oxo complex [Fig. 4(A)].8 The HMPAO O O N N HN N EtO2C CO2Et Tc Tc N N S S O O H (B) (A) Fig.4 derivative was selected from more than 100 structural variants for its optimal biodistribution characteristics. The proligand has two chiral centres and both the d–l and meso-HMPAO have been investigated. The greater effectiveness of the d–l complex is dependent on the formation of a more hydrophilic species once the complex has traversed the BBB which prevents diffusion back out of the brain. The mechanism of this reaction is not clear but glutathione appears to be involved and the complex from d–l proligand is less stable than that from the meso. The Ceretec agent generally has limited stability in solution and considerable effort has been expanded in increasing the lifetime by addition of agents such as CoII. The CoII is rapidly converted into CoIII which is believed to be the active stabilising agent although the exact details of the mechanism are uncertain.In principle the complex from the meso proligand could form two isomeric complexes differing in the orientation of the oxo group relative to the two methyl groups. In practice only the complex with the TcNO group syn to the methyls has been isolated. The TcV complexes of a wide range of bisamidedithiol proligands have been investigated as potential agents for imaging the brain. The ethylenecysteine diester (ECD) complex is commercially available from Dupont as Neurolite. The proligand loses three protons on reaction with [TcO4]2 to give the neutral square pyramidal complex [Fig. 4(B)],9 which readily crosses the BBB.It provides a striking example of the importance of stereochemistry in determining biochemical function. The l–l form of the complex is trapped once across the BBB due to enzymatic hydrolysis of one ester group by an esterase enzyme generating a more hydrophilic complex. The corresponding d–d complex is inert to enzymatic hydrolysis and diffuses back across the BBB. Such enzymatic conversion reactions also occur in the blood during biodistribution and although these impair brain uptake they facilitate clearance from the blood and non-target tissue via the kidneys. These agents actually provide images of regional cerebral blood flow (rCBF) and these are conventionally presented as computer-generated colour pictures such as that in Fig.5 obtained using the Ceretec agent. This represents a tomographic slice through a normal brain (front to back transaxial) with blue–green colours indicating low and orange high Tc concentrations and therefore high rCBF. The advent of advanced computer techniques has permitted the alternative representation of rCBF as three-dimensional surface rendered images where blood flow deficits appear as holes in the surface. Fig. 6 Fig. 5 A transaxial scan of a normal brain using Ceretec. Reproduced with permission from Amersham International. Fig. 6 A three-dimensional surface-rendered SPECT rCBF image for a patient with a stroke in the left upper (parietal) area of the brain. Reproduced with permission from the publishers from M. D. Devous in Clinical SPECT Imaging ed.E. L. Kramer and J. J. Sanger ch. 6 Raven Press Ltd New York 1995. provides a dramatic surface rendered 99Tc-SPECT image for a stroke patient with large zone of depleted blood flow in the left parietal area of the brain. The rCBF is dependent on a wide number of factors other than disease such as anxiety time of day and cognitive involvement and these have to be taken into account in interpreting the images. During epileptic fits there is enhanced rCBF (hyperperfusion) in the site of the EEG abnormality. If EEG is used to monitor the outset of the fit (ictus) then SPECT imaging can be used to image the focus of the abnormality within the brain. The three dimensional surface rendered image in Fig. 7 shows the outline of the area of hyper-perfusion due to the seizure as a 45 Chemical Society Reviews 1998 volume 27 Fig.7 Three dimensional surface-rendered SPECT images for a patient suffering from seizures. The pinkish area shows the area of hyperfusion during the seizure and the upper white area indicated secondary activation of the motor region of the brain. Reproduced with permission from the publishers from M. D. Devous in Clinical SPECT Imaging ed. E. L. Kramer and J. J. Sanger ch. 6 Raven Press Ltd New York 1995. pink solid body within the brain which is delineated in blue mesh. The white areas in the upper portions of the brain are due to the activation of the sensory motor area that accompanies an epileptic seizure. Such images permit precise location of the site of seizure origin.Subsequent surgical partial temporal lobectomy coupled with drug therapy is the best treatment for seizures which are not responsive to drugs alone.10 It has been estimated that in the United States alone there are over 50 000 patients suffering from this type of seizure. Only 1% of these are able to have surgical treatment due to the difficulties of locating the focus of the seizure by techniques such as depth EEG. There are also a number of psychiatric conditions which give rise to characteristic rCBF patterns and SPECT shows promise to be able to assist in precise diagnosis of such disorders. In schizophrenia there is frequently frontal lobe dysfunction which is particularly evident when the patient is carrying out a task requiring cognitive skills.11 The left-hand surface-rendered SPECT HMPAO-99mTc image in Fig.8 is for a schizophrenic Fig. 8 Three dimensional surface-rendered images for a patient suffering from schizophrenia. The left image was taken while the patient was performing a simple number matching exercise; the right shows rCBF during the Wisconsin card sort exercise which normally would enhance perfusion in the frontal areas of the brain. The decreased flow seen for this patient is typical for schizophrenia. Reproduced with permission from the publishers from M. D. Devous in Clinical SPECT Imaging ed. E. L. Kramer and J. J. Sanger ch. 6 Raven Press Ltd New York 1995. patient carrying out a simple task requiring little brain activation. The right-hand image was taken during a Wisconsin Card Sort task which requires intellectual input and would normally result in enhanced perfusion in the frontal lobes.It is characteristic of schizophrenia to observe the decreased perfu- Chemical Society Reviews 1998 volume 27 46 sion in the frontal lobe of the brain. There have also been reports of altered rCBF patterns in cases of depression Alzheimer’s disease and obsessive-compulsive disorder. The use of 99mTc labelled neurotransmitter molecules for the possible diagnosis of psychiatric conditions is reviewed below (section 2.4.2). 2.3.2 Heart imaging Initially it was postulated that lipophilic unipositively charged complexes would accumulate in heart tissue via the Na/K ATPase mechanism as K+ ion mimics.This concept prompted the synthesis of the cationic 99mTc complex [99mTc(dmpe)2Cl2]+ where dmpe = 1,2-bis(dimethylphosphino) ethane (Fig. 9) as a potential myocardial perfusion agent.12 + Me Me Me Me Cl P P Tc P P Cl Me Me Me Me Fig. 9 It was later found that this complex undergoes in vivo reduction to the neutral TcII complex [99mTc(dmpe)2Cl2] which having lost the positive charge has unacceptably fast washout from the heart and accumulates in the liver. Approaches are currently being pursued to lower the susceptibility of the metal ion to reduction which includes evaluation of complexes such as [99mTc(diars)2(SR)2]+ where diars = o-phenylenebis(dimethylarsine) and SR2 = thiolate. The thiolate ligands have been shown to increase the reduction potentials of the TcIII complexes relative to the chlorides.Further development of cationic complexes as myocardial perfusion agents led to the approval and availability of [99mTc(MIBI)6]+ where MIBI is 2-methoxy-2-methylpropylisocyanide Cardiolite which is shown in Fig. 10.13 The O O N N N HN EtO2C CO2Et Tc Tc N N S S O O H (B) (A) Fig. 10 X-ray structure of the tert-butyl isocyanide 99Tc analogue shows an octahedral arrangement for the isocyanide ligands around the central metal core An investigation into the mechanism of uptake has led to the belief that cations such as [Tc(MIBI)6]+ accumulate via a diffusion mechanism and electrostatic binding due to a high mitochondrial membrane potential. The lipophilicity of the complex is known to be important for uptake into the heart.The TcI oxidation state is surprisingly easily accessible directly from pertechnetate as the complex is synthesized by the reaction of 99mTcO42 with [Cu(MIBI)4][BF4] and SnCl2 as reducing agent. The uptake in the human heart is observed to be about 1.5% of the injected dose which slowly decreases to 1% after 4 h. A good organ to background ratio is achieved due to low uptake in the blood lungs liver and spleen. The presence of the alkoxy groups on the monodentate isonitrile ligands is believed to reduce this background activity. 99m The first approved neutral myocardial perfusion agent is Tc-teboroxime (Cardiotec) Fig. 11 which is a member of the BATO class of complexes (BATO—boronic acid adducts of technetium dioximes).The complex has the formula [TcCl(CDO)(CDOH)2BMe] where CDOH2 = cyclohexane-O CH3 N B N O O N Tc N N O H N H O O CI Fig. 11 2Cl2]+ previously discussed. dione dioxime and is prepared by the reaction of 99mTcO42 with a mixture of cyclohexane-1,2-dione dioxime and methyl boronic acid with SnCl2 as a reducing agent. 5 Min after injection 2.2% of the injected dose of the TcIII complex is found to accumulate in the heart via a mechanism which is unknown at this time however the complex exhibits rapid myocardial clearance in normal myocardium. The complex attains a seven coordinate geometry which consists of the three dioxime ligands bound to the TcIII centre via all six nitrogens with one end of the complex capped by a boronic acid derivative.The seventh coordination site is occupied by a chloride ligand.14 Two protons are believed to be shared between the three uncapped oxime ligands. The chloride ligand has been shown to be labile and susceptible to Cl/OH exchange which may be responsible for the initiation of the mechanism for the fast washout. One mechanism for this washout has been suggested which involves in vivo equilibrium between [Tc(OH)(CDO)(CDOH)2BMe] and the cationic complex [Tc(OH2)(CDO)(CDOH)BMe]+. It has been postulated that the neutral complexes may be washed out of the heart and it is the cationic complex which is subsequently retained. This is consistent with the results found with the dmpe complex [Tc(dmpe) A new class of technetium imaging agents containing the 99mTcN2+ core 99mTcN–NOET has been evaluated for use as myocardial imaging agent.In addition to having a new core it also differs from some other heart imaging agents in not carrying a positive charge which confirms that the cationic charge for myocardial perfusion imaging agents is not essential. The exact mechanism by which this neutral complex is accumulated in the heart remains to be determined. An essential feature of the viability of 99mTcN–NOET where NOET is N-ethoxy-N-ethyldithiocarbamate for pharmaceutical use has been the development of a high yield synthesis from methyl-N2-methyldithiocarbazate and 99mTcO42 in the presence of a tertiary phosphine as a reducing agent.15 This generates a nitride intermediate of uncertain structure in high yield and subsequent addition of the dithiocarbamate ligand gives the required complex.The lower charge on the Tc·N2+ core as compared with TcNO3+ means that in complexes with comparable ligands the nitrides will generally be more negatively charged. Some recent comparisons of the biological behaviour of oxo and nitrido complexes of DADS and MAG3 (for structures of these ligands see section 2.3.4) support this view. In human volunteers 99mTcN–NOET showed an uptake in the heart of 4.8–5.2% of the injected dose and slow clearance from normal myocardium. The X-ray crystal structure of the analogous dithiocarbamate complex [99TcN(S2CNEt2)2] Fig. 12 shows the technetium to have a five-coordinate square pyramidal geometry with the nitride ligand in an axial site.S N Tc S S S C C N N Et Et Et Et Fig. 12 Two more recent cationic imaging agents which are now in clinical trials are 99mTc-P53 {a trans-dioxobis[bis(2-ethoxyethyl) phosphino]ethane TcV cation} also known as tetrofosmine or Myoview Fig. 13 and 99mTc-Q12 TechneCard a mixed N2O2-donor Schiff base/phosphine TcIII cation Fig. 14. Both complexes carry a single positive charge and contain coordinated phosphine ligands. + OEt EtO O P P OEt EtO Tc OEt EtO O P P OEt EtO Fig. 13 + PR3 N N Me Me Tc O O Me Me O O Me Me PR3 Me Me R = CH2CH2OMe Fig. 14 Due to the ease of reduction of [Tc(dmpe) The complex [99mTcO2(P53)2]+ Myoview is synthesized via SnCl2 reduction of 99mTcO42 in the presence of the diphosphine ligand P53.In contrast to [TcCl2(dmpe)2]+ Myoview contains the dioxo TcV core which does not undergo in vivo reduction. The complex contains eight alkoxy groups on the bidentate phosphine ligands which help to reduce the background activity in the blood and liver. Uptake of the complex is 1.2% of the injected dose which has slow clearance and reduces to 1% 2 h post injection.16 Myoview rapidly enters the myocardial cells due to its lipophilic properties and the mechanism of uptake is believed to be similar to that of the 99mTc-MIBI complex. The dioxobisdiphosphine complex also exhibits rapid lung and liver clearance. The structure of the 99Tc analogue has been determined to be close to octahedral with the two bidentate ligands in the equatorial plane.A representation of the structure is shown in Fig. 13. Fig. 15 and 16 were produced using Myoview and show a healthy (Fig. 15) and defective heart (Fig. 16). The zone of non-functional heart muscle at the apex of the horseshoe is evident as a dark area. The horseshoe-shaped image is a consequence of the particular cross section of the heart due to the orientation of the scan. 2Cl2]+ 99mTc-Q12 was designed to incorporate fewer phosphine ligands and so reduce the susceptibility of the metal ion to undergo detrimental reduction. The TcIII complex attains an octahedral geometry with the tetradentate ligand occupying the equatorial plane and the two monodentate tertiary phosphines occupying the axial sites.The Schiff base ligand is 1,2-bis-{[(dihydro-2,2,5,5-tetramethylfuran-3(2H)-onato)methylene]amino}ethane which contains a furanone group to aid clearance of the complex from the blood lung and liver to achieve a good background.17 Due to the reducing ability of the phosphine ligands the addition of a separate reducing agent such as SnCl2·2H2O was found not to be necessary. 2.3.3 Liver imaging Technetium(iii) complexes of HIDA [2,6-dimethylphenylcarbamoylmethyl) iminodiacetic acid] derivatives have been shown to be suitable for imaging the hepatobiliary system.21 Currently there are three 99mTc-HIDA analogues which have been approved for this purpose; 99mTc-Lidofenin (TechneScan HIDA) 99mTc-Mebrofenin (Choletec) and 99mTc-Disofenin (Hepatolite).The exact nature of the complexes is uncertain but 47 Chemical Society Reviews 1998 volume 27 Fig. 15 SPECT image of a normal heart taken using Myoview. Reproduced with permission from Amersham International. Fig. 16 SPECT image of a diseased heart taken using Myoview. The diseased regions show as gaps in the horseshoe shape seen for the healthy heart. Reproduced with permission from Amersham International. Fig. 17 shows the proposed structure. The complex is believed to contain two ligands coordinated in an octahedral configuration and bear a single negative charge. _ O O C R1 R2 O O Tc C N CH R 2 R = HN R3 O H H2C R N 2C CH2 C R4 R5 C O O O Lidoferin R1 = CH3 Disoferin R1 = isopropyl Mebroferin R1 = R3 = CH3 R2 = Br Fig.17 Tc-sulfur colloid is also used for liver imaging and is believed to be made up of 99mTc2S7 and colloidal sulfur. The Tc-sulfur colloid is produced by the sodium dithionite reduction of TcO42 in an acidic solution. 80–85% of the colloid is accumulated in the liver via uptake in Kupffer cells by phagocytosis. A normal liver scan taken using the Tc-colloid is shown in Fig. 18. The two images are taken from the front (upper) and side and show uptake of the tracer in the liver and spleen (to the right in upper image). Some uptake in the bone marrow of the spine can also be seen (pale purple in the upper image). Chemical Society Reviews 1998 volume 27 48 Fig.18 Liver SPECT images using Tc-sulfur colloid. The upper scan is taken from the front the lower from the side. The spleen appears as a fainter spot to the right in the upper image and uptake by the bone marrow is evident with a pale purple outline of the spine. Reproduced with permission from Dr S. J. Mather Department of Nuclear Medicine St Bartholomews Hospital London. 2.3.4 Kidney imaging [99mTcO(glucoheptonate)2]2 Glucoscan also known as TechneScan or Glucoheptate is an early kidney imaging agent the precise structure of which is unknown although it is believed to have the five coordinate structure shown in Fig. 19.18 The _ O O O Tc O O O O (CHOH)4 (CHOH)4 CH2OH CH2OH Fig. 19 99m complex is not currently used widely as an imaging agent due to the availability of better alternatives such as ultrasound and CT X-ray imaging.However the complex is regularly used as a precursor for the synthesis of other TcV species via ligand exchange. The complex is synthesized by the reaction of 42 with calcium glucoheptonate in the presence of the TcO reducing agent SnCl2·2H2O. A 99mTc-DMSA complex (DMSA is dimercaptosuccinic acid) has been used to image the kidney for a number of years. The TcIII complex (of unknown structure) is prepared from the reaction of 99mTcO42 with DMSA in the presence of the reducing agent SnCl2·2H2O. Three hours post injection 50% of the injected dose has accumulated in the kidneys and specifically localizes in the proximal convoluted tubule.The TcV complex [TcO(DMSA)2]2 is also known and the complex has three possible conformations syn-endo anti or syn-exo of the carboxylic acid groups with respect to the TcNO core. Fig. 20 shows the syn-endo orientation. The crystal structure of the analogous rhenium complex [ReO(DMSA)2]2 has been determined and displays a square pyramidal geometry of the ligands around the central rhenium atom.20 (see section 3.2.1.2). 99mTc-DTPA DTPA = diethylenetriaminepentaacetic acid has approval for use as a kidney imaging agent. The structure of the 99Tc analogue has not yet been determined and it is unclear at present as to whether the complex contains technetium in the +IV or +V oxidation state. If the complex contains technetium _ O R R S Tc R R S S S syn endo R = COOH Fig.20 in the +IV oxidation state [Fig. 21(A)] the DPTA is proposed to coordinate as a hexadentate ligand or if the correct oxidation state is +V [Fig. 21(B)] the complex is proposed to contain a pentadentate DPTA ligand and the TcNO core. In both cases the complex is likely to have octahedral geometry. The complex is prepared by the reaction of 99mTcO42 with DPTA with SnCl2 acting as a reducing agent. + + O O O N O O N HOOC Tc N COOH N N Tc O O N O O COOH HOOC O O (B) (A) Fig. 21 Fritzberg10 developed the most recent and widely used anionic kidney imaging agent [99mTcO(MAG3)]2 99mTcOmercaptoacetyltriglycine which is shown in Fig. 22.17 [99mTcO(MAG3)]2 contains a free carboxylic acid group which is believed to be necessary for efficient renal excretion.The TcV – O H2C C O N O N C CH2 Tc H2C C N S O CH2 HO C O Fig. 22 complex attains a square pyramidal geometry with the oxo group in the apical position. The structure of the rhenium analogue has been determined. In contrast to the previous kidney imaging agents there is no chiral centre and therefore no problems arise from the existence of isomers.The complex is prepared by the reaction of 99mTcO42 with benzoyl mercaptoacetyltriglycine and the reducing agent SnCl2 when loss of the benzoyl protecting group occurs. The benzoyl protecting group prevents ligand oxidation and therefore increases kit stability and reliability. A few minutes post injection about 1–2% of the injected dose is found in the kidneys.It is the passage into and through the kidneys which provides a measure of renal function. The presence of the thiol group provides additional reducing power to convert TcVII to TcV and assists in the stabilisation of the complex. Current research is directed at variations in the MAG3 ligand by substitution of glycine by l-alanine thereby modifying the renal excretion characteristics. 2.3.5 Bone imaging A series of 99mTc complexes of phosphonate ligands have been developed as bone-imaging agents. The initial developments in this area used pyrophosphate but it was later shown that diphosphonates such as methylenediphosphonate [MDP Fig. 23(A)] gave much improved performance.Typically the agent is prepared by reaction of the [99mTcO4]2 generator eluate with MDP in the presence of SnCl2·2H2O as reductant. The coordination chemistry involved is not simple and the number of species formed is dependent on pH concentration and reductant used. The concentration dependence complicates attempts to characterise the 99mTc complexes by extrapolation from the 99Tc level as the concentrations are hugely different (1028 m for 99mTc 1023–104 m for 99Tc). There is a consensus that the dominant oxidant state for 99mTc/MDP is TcIV and that a mixture of oligomers is formed. At the 99Tc level reaction of [99TcBr6]22 with H4MDP led to the isolation and structural characterisation of a polymeric complex [Fig. 23(B)].22 A hexameric complex has also been isolated and an X-ray structure determination carried out.OH OH O O HO Tc O CH2 O O P O O Tc O O P P HO OH HO O OH P CH2 (B) (A) Fig. 23 The mechanism of absorption on bone is believed to be via co-ordination of the free phosphoryl oxygens to calcium ions on the hydroxyapatite bone surface. Since stressed bone has higher concentrations of calcium ions such areas appear as ‘hot spots’ on the scan. The 99mTc-MDP bone scan in Fig. 24 (front and rear view) gives a clear picture of the skeletal structure with an intense (red) area corresponding to the bladder. The area of increased tracer uptake in the right ankle is caused by arthritis. One of the main uses for 99mTc bone-scans is for cancer patients to identify if there has been metastasis into the bone the metasteses appearing as bright spots on the scan.SPECT 99mTc bone images can in general provide information on lesions which may not be visible by conventional X-ray methods. 99mTc-based scans can also be valuable for diagnosis of problems with joints such as the elbow or knee as it can show up bone damage not immediately visible from MRI imaging. Such images then enable the surgeon to screen for those patients who will benefit most from expensive keyhole type exploratory surgery. 2.4 ‘Second generation’ technetium imaging agents These are classified according to the receptor site or biological function that is targeted. 2.4.1 Steroid receptors About 60–70% of breast tumours are estrogen receptor positive and endocrine therapy with drugs such as tamoxifen is effective in about half of cases with such estrogen receptor positive cancers.If a molecule which binds to such sites could be radiolabelled it would provide a method of monitoring the progress of therapies with agents such as tamoxifen. Most prostate cancers are androgen and progesterone receptor positive and could be imaged with an appropriate labelled receptor hormone. The structures of the three relevant hormones are shown in Fig. 25. The 99mTc labelling of the progesterone receptor has been studied utilising conjugation to N2S2 ligands via a phenyl spacer (Fig. 26).23 The key to success is to find a site of attachment to the steroid which does not impair receptor binding and the 11b site proved to be optimal.The conjugates contain stereoisomers a syn pair and two diastereoisomers and remarkably the syn pair had an affinity for the progesterone receptor 161% of progesterone itself. Although the conjugates showed high binding in vivo studies also showed a high level of non-specific binding. 49 Chemical Society Reviews 1998 volume 27 Fig. 24 Bone SPECT images taken using MDP diphosphonate agent. The skeletal structure shows clearly and the bright red area in the centre is due to accumulation in the bladder. The red area on the right ankle is due to arthritis in the joint. Reproduced with permission from Dr S. J. Mather Department of Nuclear Medicine St Bartholomews Hospital London. OH HO Estradiol O H Dihydrotestosterone Fig.25 Nevertheless the approach is clearly promising although further fine-tuning of biodistribution characteristics is required. An alternative to the pendant receptor ligand approach which was discussed in the introduction above is to integrate the receptor binding sites directly onto the outer periphery of the Tc ligands. A proposed structure for such a complex is shown in Fig. 27(A) and the overall similarity to progesterone is apparent. Chemical Society Reviews 1998 volume 27 50 O O Progesterone OH N S O Tc S OH C CCH3 O Fig. 26 Some initial steps towards producing an analogue of estradiol (Fig. 25) have been made with the synthesis of the complex shown in Fig. 27(B).24 Reaction of a TcV precursor with a 1 1 mixture of the two bidentate N–S ligands leads to a good yield of the mixed complex shown rather than a statistical mixture.The receptor-binding affinity was found to be low as perhaps expected for this initial model but the approach offers interesting possibilities for the future. OH S S O O N Tc Tc N N N HO S S O (B) (A) Fig. 27 2.4.2 Central nervous system (CNS) receptors There are a number of important diseases and psychiatric conditions that are associated with changes in the densities of neurotransmitter receptor sites in the brain benzodiazepene (epilepsy) muscarinic and nicotinic (Alzheimer’s disease) dopaminergic (Parkinson’s disease psychiatric conditions) serotonergic (psychiatric conditions).Most of the initial studies in imaging have used PET but this imaging modality is of limited availability due to the necessity of being close to a cyclotron. The g-emitting iodine-123 has been used for SPECT brain receptor imaging but again this isotope is expensive and unlike 99mTc not widely available. There is currently a worldwide effort directed to producing 99mTc CNS receptor imaging agents via the pendant bioconjugate approach. We give two examples of different approaches to the construction of the conjugate. The molecule ketanserin [Fig. 28(A)] is a potent antagonist for serotonin (5-HT) receptor sites. Detailed biochemical studies have established where modifications can be made without impairing receptor binding and appropriate fragments have been bound via thiolate or isocyanide groups to a TcV oxocore with a tridentate NS2 22 or SS2 22 ligand.25 The structure of an Re analogue [Fig.28(B)] shows the square pyramidal MNO core and the flexible side chain containing the receptor binding sites. Binding studies have been made using rat brain homogenate rich in 5-HT receptor sites and have shown high affinity for the derivative with an OC6H5 group in place of the phenyl group and a isocyanide group to provide ligation to the Tc. The higher affinity of this particular derivative appears to be associated with better ability to traverse the BBB. Cocaine [Fig. 29(A)] and analogues block dopamine transporter sites and iodine-123 substituted derivatives have been explored for the diagnosis of Parkinson’s disease.Linking of the cocaine derivative via the seven-membered ring to a TcV oxo-O H N F N N O O ketanserin (A) N O S S Re S S O S S Tc CH3 CH3 N N N N CO2Me O O C CI Tc-TRODAT (A) (B) Fig. 28 Fig. 29 (B) core via an N2S2 ligand produces a conjugate (‘Tc-TRODAT’) shown in Fig. 29(B). This has produced the first in vivo images of D2 transporter sites in man (Fig. 30) using technetium-99m. Uptake (coloured yellow) in the areas of the brain rich in D2 Fig. 30 A series of SPECT scans taken with 99mTc-TRODAT at the time intervals shown. The initial image accords with normal rCBF and the 60–80 and 120–140 min scans show significant uptake in the regions of the brain rich in dopamine transporter sites as two bright yellow areas in the centre.Reprinted with permission from H. F. Kung H.-J Kim M.-P Kung S. K. Meegala K. Plossl and H.-K Lee Eur. J. Nucl. Med. 1996 23 1527. receptors is evident in the centre of the image taken after 120 min. This exciting advance confirms the viability of the conjugate approach to the 99mTc imaging of CNS receptor sites. 2.4.3 Monoclonal antibodies or antibody fragments Monoclonal antibodies or their fragments the so-called ‘magic bullets’ are potentially ideal vehicles to target radioisotopes to specific sites providing of course the radiolabel can be introduced without interfering with binding to the receptor site. The relatively large size of whole antibodies generally confers undesirably slow biodistribution kinetics for imaging purposes and attention is now directed to antibody fragments [F(ab’)2 Fv Fab’ or Fab] which retain the specific binding characteristics.The use of the fragments also reduces the possibility of immunogenicity and adverse allergic reactions. The crucial aspect of the development of 99mTc labelled antibodies and their fragments is the mode of attachment of the metal and the link must be sufficiently stable to prevent premature release of the radioisotope. The first approach to 99mTc labelling of antibodies involved the reduction of the disulfide groups holding the F(ab’)2 fragments together and binding of the Tc to the resulting SH groups.27 Although attractive in its simplicity the conjugates do not always have high in vivo stability.However Fig. 31 shows a series of images produced by a 99mTc direct labelled antibody PR1A3 for colorectal tumours. The image after 5 min shows uptake mainly in the heart blood pool and the 7 h image additionally shows liver uptake (below the heart from this angle) and slight uptake in the tumour at 4 o’clock. This tumour uptake has increased significantly after 20 h indicating the relatively slow targeting of the monoclonal antibody. Fig. 31 Scans taken with 99mTc labelled PR1A3 monoclonal antibody against colorectal tumours at time intervals shown. The 20 h image shows clear uptake in the tumour as a small orange area at 4 o’clock. Reproduced with permission from Dr S.J. Mather Department of Nuclear Medicine St Bartholomews Hospital London. This has prompted a search for more stable conjugates and variants of the bifunctional chelate approach appear to offer the most promise. Two principal strategies have been adopted. The first (post-formed chelation) involves initial attachment of metal binding groups to the antibody or fragment followed by insertion of the 99mTc. Two of the many examples are shown in Fig. 32. A hydrazine nicotinamide derivative can be bound to lysine groups on an antibody or fragment as shown in Fig. 32(A). The hydrazine group then reacts with [99mTcO4]2 to give an uncharacterised but stable conjugate which may contain TcNN–NH– groups. An alternative elegant route to binding the antibody is via the thiolactone in Fig.32(B) which generates an N2S2 diaminedithiol ligand system which forms a stable neutral square pyramidal TcNO species with [99mTcO4]2.28 51 Chemical Society Reviews 1998 volume 27 O O NH NH2•HCI N N O O AbNH2 O NH NH2•HCI N AbNH (A) Fig. 32 Antibody engineering techniques have also enabled the incorporation of (gly)4cys peptide into single chain antibody proteins. The peptide binds the 99mTc in an analogous fashion to mercaptoacetyltriglycine (MAG3) (see kidney imaging) and provides a stable conjugate. Studies have been made of the protein-coupled fragments of antibodies specific for human ovarian cancer using tumours grafted into mice and significant uptake of 99mTc into the tumour was observed.29 The alternative approach (preformed chelation) requires the initial synthesis of a technetium chelate with an activated ester group and subsequent attachment of the antibody or fragment.This is illustrated in Fig. 33 for a diamidedithiolate ligand with a pendant tetrafluorophenyl (TFP) activated ester. Kits based on this procedure for attaching antibody fragments for targeting melanoma and lung cancer have undergone chemical trials and many other related promising systems are currently in development. F F CO2 NH NH F F S S Tc-gluconate O O Fig. 33 2.4.4 Imaging hypoxia Cells become hypoxic in several disease states. A significant fraction of certain types of tumour are hypoxic (80% for head and neck squamous cell carcinoma) and imaging of such tumours would be of great advantage in devising suitable treatments.Impaired blood flow in the heart gives rise to transient or persistent tissue hypoxia and accurate diagrams of such areas show where medical intervention to restore blood flow would be beneficial. 2-Nitroimidazoles have been shown to be trapped in hypoxic cells due to their reduction to a series of products which either cannot diffuse out or become bound inside the cell. Several groups have investigated the possibility of linking 2-nitroimidazoles to 99mTc chelates for hypoxic site imaging.30 Conjugate A (Fig. 34) shows some promise for hypoxic imaging in vivo but is too lipophilic and clears slowly from background Chemical Society Reviews 1998 volume 27 52 N N O SH S AbNH2 O N N NHAb O SH HS (B) CO2TFP O NH N O O Tc S S AbNH2 CONHAb O NH N O O Tc S S tissue.A more hydrophilic version [Fig. 29(B)] with an oxygen in the backbone is more promising with more rapid liver clearance. N N NO2 O N O N N O N Tc Tc N N N N N N O O O O NO2 H H (B) (A) Fig. 34 A variant on this theme involved conjugation of the 2-nitroimidazole to a amineoxime ligand but with a four- rather than three-carbon backbone (Fig. 35). However a control experiment for hypoxic imaging using the 99mTc complex without the imidazole group showed this to be more effective HN NH N N N N O O NO2 H H Fig. 35 than with the conjugated imidazole. The co-ordination chemistry of the Tc in this type of amine–oxime complex is strongly dependent on the backbone length.With three carbons in the backbone a monooxo complex is formed whereas with five carbons a trans dioxo system is favoured. The four-carbon system apparently undergoes biochemical reduction and is trapped inside the hypoxic cell. This redox behaviour may be associated with labile protic equilibria involving protonation and/or aquation of the oxo-core. This last complex is showing promise as a hypoxic imaging agent in human clinical trials. 2.4.5 Thrombus imaging Current research is being directed at the area of diagnostic agents for imaging thrombi. One group has used the approach of conjugating platelet glycoprotein IIb/IIIa antagonists onto a chelate complex of technetium.The glycoprotein IIb/IIIa complex is expressed on the membrane surface of activated platelets and plays an integral role in platelet aggregation and thrombus formation. Cyclic peptides which incorporate the sequence Arg–Gly–Asp (RGD) have been shown to be high affinity antagonists for the glycoprotein receptor. The glycoprotein IIb/IIIa receptor is expressed only on activated platelets so therefore radiolabelled cyclic IIa/IIIa receptor antagonists were anticipated to bind to only the platelets involved in the thromboembolic event. Fig. 36 shows one example of a technetium complex conjugated to the cyclic glycoprotein IIa/IIIa receptor antagonist. In this case the coordination environment for technetium is based around the N2S2 diamide dithiol chelate ligand with the cyclic peptide conjugated onto the backbone via an active ester.This complex has also been synthesized from the direct reaction of 99mTcO42 with the preformed conjugate ligand which eliminates the synthesis of the active ester complex and subsequent step of peptide conjugation.31 Fig. 37 shows selected images derived from studies of this complex in a dog with implanted deep vein thrombi. The complex was actively incorporated into the growing thrombi with images being clearly detectable within 15 min post-injection. NH O H2N O N H N H N O OH HN O O NH HN O NH N H O O N N O O Tc– O Fig. 36 S 3 Rhenium S An alternative mode of incorporation of the cyclic peptide is via N-hydroxysuccinimidyl hydrazinonicotinate (S-Hynic).32 Synthesis of the conjugate shown in Fig.38 was achieved by the reaction of 99mTcO42 with the Hynic conjugated cyclic peptide in the presence of EDDA (ethylenediaminediacetic acid) and SnCl2·2H2O. The monoanionic ligand X is believed to be chloride and the hydrazine unit is postulated to be coordinated as an isodiazene ligand (NNNHR) as opposed to a diazenide ligand (–NNR). There is a possibility of a number of isomers for this type of complex. Only investigations at the macroscopic level using technetium-99 or rhenium will confirm the structure of the complex and the mode of coordination of Hynic ligand. Another group has also investigated the synthesis of technetium complexes containing receptor binding peptides which bind to the glycoprotein IIb/IIIa receptor.They focused on high affinity peptides containing the receptor binding sequence –Apc–Gly–Asp– where Apc is S-(3-aminopropyl)cysteine. Again a bisamide bisthiol chelator is used to coordinate technetium which was incorporated into the dimeric peptide. The 99mTc complex of this peptide P357 was synthesized at room temperature. This complex has given excellent clinical images of deep vein thrombosis and of pulmonary embolism. 33 The element rhenium (Z = 75) was discovered in 1925 by the Noddacks and is one of the rarest elements occurring naturally as a mixture of two non-radioactive isotopes 185Re (37.4%) and Fig. 37 Images produced by a technetium complex conjugated to a cyclic glycoprotien IIa/IIIa receptor antagonist in a dog with implanted deep vein thrombi.Reprinted with permission from S. Liu D. S. Edwards R. J. Looby M. J. Poirier M. Rajopadhye J. P. Bourque and T. R. Carroll Bioconj. Chem. 1996 7 83. Copyright 1996 American Chemical Society. CON- P P = peptide NH N O O Tc O O MeV g-Energy/ keV 186Re NH Range in tissue/ mm 5 11 N H are summarised in Table 1. 1.07 (71%) 2.1 (100%) 137 (9%) 155 (15%) 90 17 188Re X Fig. 38 187Re (62.6%). The radioactive isotopes of interest in nuclear medicine are 186Re and 188Re the nuclear properties of which Table 1 Radioactive isotopes of rhenium Halflife/ Max. b energy/ h Both isotopes are suitable for therapeutic use by means of b-irradiation.The 5 mm range for 186Re means that it is suitable for small tumours whereas the greater 11 mm range for 188Re is more appropriate for large masses. The selection of isotope is also governed by factors such as halflife and technical aspects of their production. 186Re is generated by neutron radiation of 185Re and there is inevitable contamination with non-radioactive 185Re. On the other hand 188Re is available by radioactive decay of 188W and separable by analogous ion-exchange methods to those used for 99mTc and commercial generators are available. Such a generator with 0.5 Ci of 188W has the potential to provide therapeutic treatments to several hundred patients over its 2–6 month lifetime.The major disadvantage of 188Re for therapeutic applications is the relatively short halflife of 17 h. 3.1 Comparison of the chemistry of technetium and rhenium The ‘lanthanide contraction’ ensures that the complexes of the two elements are very similar in terms of their physical characteristics (size lipophilicity etc). Significantly however rhenium complexes are easier to oxidise (harder to reduce) and more kinetically inert than their technetium analogues. The Chemical Society Reviews 1998 volume 27 53 relative ease of oxidation of rhenium means that in vivo oxidation to [ReO4]2 is common. This can be an advantage in that it provides an ultimate elimination route for the radioactive isotope via the kidneys. 3.2 Rhenium radiopharmaceuticals 3.2.1 ‘Rhenium essential’ agents As discussed in section 2.2 above this is class of therapeutic agents where the biodistribution is determined by the size charge and lipophilicity of the complex.Technetium complexes of this type are used to study major organs and there are few examples of rhenium complexes which have the required specificity to be used for treatment of cancers and other therapies. 3.2.1.1 Agents for the palliation of bone pain Externally applied radiation (sealed source) is widely used for pain relief. However this is difficult to deploy when metastases are widely distributed through the skeletal structure. This has prompted the search for b-emitting radiopharmaceuticals that can be targeted to bone lesions and rhenium is one of several radioactive elements under investigation.Among the more developed agents for palliation of bone pain are rhenium radiopharmaceuticals and are based on diphosphonate ligands. 34 The underpinning co-ordination chemistry is directly analogous to that discussed for 99mTc bone imaging agents above. Typically [186ReO4]2 is incubated at 100 °C for 10 min with diphosphonate (HEDP) in the presence of SnCl2 as reductant giving a 90% labelling yield. As with 99mTc HPLC measurements indicate that a mixture of polymeric complexes are formed but bind preferentially to sites of skeletal damage. The exact mechanism by which bone pain is reduced is still unclear but these agents have proved to be of real benefit in clinical trials.3.2.1.2 Medullary thyroid carcinoma A TcV complex of dimercaptosuccinic acid (DMSA) is used widely as an agent for the imaging of a relatively rare medullary thyroid carcinoma. The 99mTc kidney imaging agent using the same ligand is believed to contain TcIII rather than TcV. The ReV complex has the expected square pyramidal structure with an apical oxo-group (Fig. 39) and exists as a mixture of isomers – O CO2H HO2C S S Re HO CO2H 2C S S Fig. 39 in solution depending on the orientation of the carboxylate groups with respect to the S4 plane. This Re complex displays selective uptake in tumour tissue analogous to that of the Tc species and offers a possibility of therapeutic treatment of this disease.35 3.1.2.3 Monoclonal antibodies and fragments Monoclonal antibodies and their fragments are potentially powerful targeting agents for the therapeutic uses of radionucleides.In principle the methods described in section 2.4.3 above for the attachment of 99mTc can also be used for the radioactive isotopes of Re. The mercaptoacetyltriglycine (MAG3) technetium system is used widely as an imaging agent for renal function (see section 2.3.4. above) forming the familiar square pyramidal oxo complex. This system has been adapted by the attachment of an activated ester group via an amide group (Fig. 40) to permit conjugation to antibodies or fragments as described above. Conjugation of 186Re to a murine (mouse derived) antibody for adenocarcinomas using this technology gave promising results in animals.36 The slower kinetics of complex formation for Re means that the preformed Chemical Society Reviews 1998 volume 27 54 O NH HN O NH O F SH F O O F F Fig.40 chelate approach needs to be used as the conditions for complexation of the Re can denature the antibody. Subsequent developments have included using chimaeric human–mouse antibodies which reduced immunogenicity and biodegradable linker groups which accelerate the excretion of non-targeted radioactivity elsewhere in the body as the Re chelate is released and then eliminated by the kidneys. There have also been studies of the direct labelling of antibodies using the prereduction step to generate free SH groups and subsequent binding to rhenium after stannous chloride reduction of either [186ReO4]2 or [188ReO4]2.As with this approach to labelling antibodies with 99mTc there is a problem with the stability of the conjugates. This is more acute for the therapeutic Re isotopes as it can lead to undesirable radiation doses at locations other than the tumour. 3.2.1.4 Steroids and bioactive peptides Certain tumours (pituitary malignant breast pancreatic) have a large number of receptors for the tetradecapeptide somatostatin and its cyclic analogue octreotide. The latter has disulfide bridges and reduction of this and reaction with 188ReO42/ stannous/citrate provides a direct high yield labelling route. Studies in mice have shown that this is retained on injection in tumours and can induce necrosis of the tumour tissue.Bioconjugates of octreotide using the N2S2 ligand approach described in Section 2.4 have been made for technetium-99m and have been investigated as a potential method for the imaging of tumours. Steroids can be attached to rhenium oxo-complexes of N2S2 ligand systems in a directly analogous fashion to that described for technetium in Section 2.4.1 and show take up in target tissue rich in steroid receptors. This therefore is a promising approach to the delivery of therapeutic radiation to appropriate tumours. An alternative approach to the labelling of steroids involves the attachment of a rhenium cyclopentadienyltricarbonyl unit to the 17 position of estradiol derivatives as shown in Fig. 41.37 OH CICH2 Re(CO)3 HO Fig.41 Surprisingly the derivative (shown) with an 11-chloromethyl group shows higher binding to receptors than the parent steroid. This was attributed to the ability of the acetylene linked Re unit to bend out of the way behind the steroid molecule and an interaction of the CH2Cl group with a Lewis acid group at the receptor site. It has yet to be shown if this interesting approach can be extended to the radioactive isotopes with the requirement to use aqueous [ReO4]2 as the starting point for the chemistry. 4 References Due to the limit on the number of references allowed the authors have quoted some pieces of work without proper acknowledgement. The choice of work referenced is of necessity somewhat arbitrary and will we trust not cause offence to those we appear to have ignored.1 E. Deutsch K. Libson and S. Jurisson Prog. Inorg Chem. 1983 30 75. 2 M. Clarke and L. Podbielski Coord. Chem. Rev. 1987 78 253. 3 F. Tisato F. Refosco and G. Bandoli Coord. Chem. Rev. 1994 135 235. 4 J. R. Dilworth and S. Parrott Directions in Radiopharmaceutical Research and Development ed. S. Mather Kluwer Netherlands 1996 1 and other articles therein. 5 Technetium and Rhenium in Nuclear Medicine ed. M. Nicolini G. Bandoli and U. Mazzi S. G. Editorali Padova 1995 vol. 4 and references therein. 6 Clinical SPECT Imaging ed. E. L. Kramer and J. J. Sanger Raven Press New York 1995. 7 K. Yoshihava Top. Curr. Chem. 1996 176 2. 8 J. P. Leonard D. P. Novotnik and R. D. Neirinckx J. Nucl.Med. 1986 27 1819. 9 E. H. Cheesman M. A. Blanchette M. V. Ganey L. J. Maheu S. J. Miller and A. D. Watson J. Nucl. Med. 1986 29 288. 10 B. L. Holman and M. D. Devous J. Nucl. Med. 1992 33 1888. 11 M. D. Devous Brain Imaging Applications in Psychiatry ed. N. Andreason American Psychiatric Press Washington 1988 p. 147 and references therein. 12 M. C. Gerson E. A. Deutsch H. Nishyama K. G. Libson R. J. Adolph L. W. Grossman V. J. Sodd J. L. Fortman C. C. Vanderhagden C. C. Williams and E. L. Salinger Eur. J. Nucl. Med. 1983 8 371. 13 B. L. Holman A. G. Jones J. Lister-James A. Davison M. J. Abrams J. M. Kirschenbaum S. S. Tubeh and R. J. English J. Nucl. Med. 1984 14 K. E. Linder M. F. Malley J. Z. Gongoutas S. E. Unger and 25 1350. A. D.Nunn Inorg. Chem. 1990 29 2428. 15 R. Pasqualini and A. Duatti J. Chem. Soc. Chem. Commun. 1992 1354. 16 J. D. Kelly A. M. Forster B. Higley C. M. Archer F. S. Booker L. R. Canning K. W. Chiu B. Edwards H. K. Gill M. McPartlin K. R. Noyle I. A. Lathan R. D. Pickett A. E. Storey and P. M. Webbon J. Nucl. Med. 1992 24 353. 17 M. A. DeRosch J. W. Brodack G. D. Grummon M. E. Marmian D. L. Nosco K. F. Deutsch and E. A. Deutsch J. Nucl. Med. 1992 22 850. 18 C. E. Costello J. W. Brodack A. G. Jones A. Davison D. L. Johnson S. Kasina and A. R. Fritzberg J. Nucl. Med. 1983 24 353. 19 J. A. Ponto H. M. Chilton and N. E. Watson Pharmaceuticals in Medical Imaging ed. D. P. Swanson H. M. Chilton and J. H. Thrall MacMillan Publishing Co. New York 1990 501.20 J. Singh A. K. Powell S. E. M. Clarke and P. J. Blower J. Chem. Soc. Chem. Commun. 1991 15. 21 A. R. Fritzberg S. Kasina D. Eshima and D. L. Johnson J. Nucl. Med. 1986 27 111. 22 K. Libson E. A. Deutsch and B. Barnett J. Am. Chem. Soc. 1980 102 2476. 23 M. J. Welch J. B. Downer and J. A. Katzenellenbogen Current Directions in Radiopharmaceutical Research and Development ed. S. J. Mather Kluwer Academic Press Netherlands 1996 p.137 and references therein. 24 D. Y. Chi J. P. O’Neil C. J. Anderson M. J. Welch and J. A. Katzenellenbogen J. Med. Chem. 1994 37 928. 25 H. Spies T. Fietz M. Glacer H.-J. Pietsch and B. Johanssen in Technetium and Rhenium Chemistry and Nuclear Medicine ed. M. Nicolini G. Bandoli and U. Mazzi S. G. Editorali Padova Italy 1995 vol. 4 p. 243 and references therein. 26 M. E. Kung H.-J. Kim M.-P. Kung S. K. Meegallu K. Plossl and H.-K. Lee Eur. J. Nucl. Med. 1996 23 1527. 27 A. R. Fritzberg and D. S. Wilbur in Handbook of Targeted Delivery of Imaging Agents ed. V. P. Torchilin CRC Press Inc. New York 1995 p. 83 and references therein. 28 D. J. Hnatowich G. Mardirossan M. Ruscowski M. Fargarasi F. Firzi and P. Winnard J. Nucl. Med. Chem. 1993 34 172. 29 A. J. T. George F. Jamar M.-S. Tai B. T. Heelan G. P. Adams J. E. McCartney L. L. Houston L. M. Weiner H. Opperman A. M. Peters and J. S. Huston Proc. Natl. Acad. Sci. 1995 92 8538. 30 C. M. Archer B. Edwards and N. A. Powell in Current Directions in Radiopharmaceutical Research and Development ed. S. J. Mather Kluwer Academic Press Netherlands 1996 p. 81 and references therein. 31 S. Liu D. S. Edwards R. J. Looby M. J. Poirier M. Rajopadhye J. P. Bourque and T. R. Carroll Bioconj. Chem. 1996 7 203. 32 S. Liu D. S. Edwards R. J. Looby M. J. Poirier M. Rajopadhye J. P. Bourque and T. R. Carroll Bioconj. Chem. 1996 7 83. 33 J. Lister-James W. J. McBride S. Buttram E. R. Civetello L. J. Martel D. A. Pearson D. M. Wilson and R. T. Dean in Technetium and Rhenium in Chemistry and Nuclear Medicine ed. M. Nicolini G. Bandoli and U. Mazzi S. G. Editorali Padova 1994 vol. 3 269. 34 W. A. Volkert and E. A. Deutsch in Advances in Metals in Medicine ed. M. J. Abrams and B. A. Murrer JAI Press USA 1993 p. 115 and references therein. 35 P. J. Blower J. Singh S. E. M. Clarke M. M. Bisundan and M. J. Went J. Nucl. Med. 1990 31 768. 36 A. R. Fritzberg L. M. Gustavson M. D. Hylandes and J. M. Reno in Chemical and Structural Approaches to Rational Drug Design ed. D. B. Weiner and W. B. Williams CRC Press Inc. Boca Raton USA 1994 p. 125 and references therein. 37 S. Top M. Elhafa A. Vessieres J. Quivy J. Vaissermann D. W. Hughes M. J. McGlinchey J. P. Mornon E. Thoreau and G. Jaouen J. Am. Chem. Soc. 1995 117 8372. Received 13th June 1997 Accepted 6th August 1997 55 Chemical Society Reviews 1998 volume 27