Molecular cardiovascular imaging using scintigraphic methods

Eur Radiol (2007) 17: 1422–1432DOI 10.1007/s00330-006-0541-6 MOLECULAR IMAGING

Lars SteggerKlaus SchäfersKlaus KopkaStefan WagnerSven HermannPeter KiesMarilyn LawOtmar SchoberMichael Schäfers

Received: 5 September 2006Revised: 2 November 2006Accepted: 24 November 2006Published online: 6 January 2007# Springer-Verlag 2007

Molecular cardiovascular imaging usingscintigraphic methods

Abstract Molecular cardiovascularimaging plays an increasingly impor-tant role both in basic research and inclinical diagnosis. Scintigraphicmethods have long been used to studypathophysiological changes on a cel-lular and molecular level, and they arelikely to remain important molecularimaging modalities in the foreseeablefuture. This article provides an over-view over current developments incardiovascular molecular imagingusing scintigraphic methods. Thefocus lies on imaging of cardiacinnervation, plaque instability, hyp-oxia and angiogenesis, gene expres-sion and stem and progenitor cellmigration and proliferation.

Keywords Molecular imaging .Scintigraphic . Cardiovascular .SPECT . PET

Introduction

Owing to the developments in molecular and cellularbiology in recent years, many of the genetic aberrations andmolecular changes associated with cardiovascular diseaseshave been identified. Furthermore, the advent of molecularand cellular therapies has fundamentally changed theprospect of treatment. Therefore, imaging of processes onthe molecular and cellular level becomes an increasinglyimportant field in both preclinical and clinical imaging.

Radiotracer techniques have been a cornerstone offunctional and molecular cardiovascular imaging sincethe early 20th century. In 1927 Blumgart, Yens and Weisspublished results of a study where they used the radiumdecay product “radium C” (historical name for Bismuth-

214) to investigate circulation times from the right arm tothe left arm in humans [1]. Since then a large number ofdifferent radiotracers labelled with gamma- or positron-emitting nuclides have been used for cardiovascularautoradiography, planar scintigraphy, single-photon emis-sion computerized tomography (SPECT) and positronemission tomography (PET). Recently, devices dedicatedto small-animal imaging have been developed andevaluated (Fig. 1) (e.g., [2]).

In the following chapters an overview of cardiovascularmolecular imaging with radiotracer techniques is given.The focus lies on new developments in imaging of cardiacinnervation, plaque instability, hypoxia and angiogenesis,gene expression and stem and progenitor cell migration andproliferation. These new methods are being used in

L. Stegger (*) . K. Schäfers .K. Kopka . S. Wagner . S. Hermann .P. Kies . M. Law . O. Schober .M. SchäfersDepartment of Nuclear Medicine,University Hospital of Münster,Albert-Schweitzer-Str. 33,48149 Münster, Germanye-mail: [email protected].: +49-251-8347362Fax: +49-251-8347363

preclinical and clinical research. Some of them may find aplace in cardiovascular routine imaging in the future.

Myocardial perfusion and metabolism

Perfusion imaging with radiotracer techniques under bothstress and resting conditions is clinically used to identifyhemodynamically relevant coronary stenoses in patients witha low to intermediate a priori cardiovascular risk as assessedby the Framingham or PROCAM (Prospective Cardiovas-cular Münster Heart Study) [3] scores. The negativepredictive value with regard to major cardiovascular eventsof this methodology is excellent for this patient group [4, 5].Perfusion imaging is also used in patients with knowncoronary artery disease to assess the hemodynamic relevanceof known stenoses and the need for therapeutic intervention.A large number of radiotracers such as 201Tl-TlCl (thallouschloride), 99mTc-methoxyisobutylisonitrile (99mTc-MIBI)and 99mTc-tetrofosmin for SPECT and 13N-NH3 (ammonia),15O-H

2O (water) and 82Rb-RbCl (rubidium chloride) for PETimaging have been developed and extensively evaluated,each having specific advantages and disadvantages (e.g.,[6]). PET allows for biomathematical quantitation of manyphysiological parameters making it well suited as a researchtool. Gated acquisition with additional assessment of globaland regional left-ventricular function has become routine inrecent years (Fig. 2) [7, 8] and provides independentprognostic information [9].

The assessment of myocardial glucose metabolism withthe radiolabelled glucose derivative 18F-2-fluoro-2-deoxy-

D-glucose (18F-FDG) is clinically used to distinguish viable,but non-functional (hibernating) myocardium from non-viable tissue with PET (e.g., [10]). It can be used to establishthe potential benefit of interventional or surgical revascular-isation on an individual basis. 18F-FDG is also widely usedin preclinical research and can be used, for example, to studyphenotypic consequences of genetic aberrations in mice[11]. Other PET and SPECT tracers have been developed toassess fatty acid metabolism [12].

Innervation

The autonomous nervous system plays an important role inregulating cardiac function and is implicated in manycardiac diseases ranging from coronary heart disease,congestive heart failure and cardiomyopathies to life-threatening arrhythmias. Although less established inclinical practice than perfusion and metabolic imaging,examination of the cardiac autonomous nervous system ispotentially equally valuable.

Both the pre- and postsynaptic functions of thesympathetic and parasympathetic nervous system areassessable by radionuclide techniques. At present, thesympathetic arm has received most attention. Presynapticfunction of sympathetic innervation has been examinedwith the SPECT radiotracer 123I-meta-iodobenzylguanidin(123I-MIBG) and the PET radiotracer 11C-meta-hydro-xyephedrine (11C-MHED). Both behave similarly tonorepinephrine, the physiological neurotransmitter. Thedensity of postsynaptic β-adrenoceptors has been quanti-

Fig. 1 PET scan of a mouse.(a) Photograph of the dedicatedsmall animal PET scannerquadHIDAC (Oxford-Positron-Systems, Weston-on-the-Green,UK) based on a multiwire gaschamber detector design;(b) mouse preparation withisoflurane/oxygen anaesthesia,ECG electrodes (on paws) and apressure pad for measurement ofrespiration (under belly); (c)result of an ECG-gated acquisi-tion of left-ventricular 18F-FDGuptake. Short axis (SA), hori-zontal long axis (HLA) andvertical long axis (VLA) imagesat the midventricular level areshown for all uneven gates of a12-gate image acquisition [2].Picture of a normal-sized matchprovided for size comparison

1423

fied with radiolabelled β-adrenoceptor antagonists andPET. 11C-CGP 12177 is a non-selective β-adrenoceptorantagonist that has seen considerable application inclinical research [13–16]. More recently, 11C-CGP12388 has been synthesised and used to assessβ-adrenoceptor downregulation in idiopathic dilatedcardiomyopathy [17]. Labelled β-adrenoceptor antago-nists selective for the β1-adrenoceptor-subtype are underdevelopment and await application in imaging research[18]. The PET tracer 11C-GB67 has been developed forimaging of α1-adrenoceptors [19], but has not yet beenused clinically.

It was shown that after non-Q-wave myocardial infarc-tion (MI), denervation occurs at the site of the infarct. AfterQ-wave MI denervation extends also “downstream” withrespect to sympathetic innervation [20]. Clinically relevantarrhythmias after MI often arise in denervated but viabletissue. Sympathetic function is impaired for a prolongedtime after an ischaemic event as Inobe et al. have shown inpatients with coronary spastic angina using 123I-MIBG[21]. They suggested a potential role in the emergencydepartment to answer the question of previous ischaemia.

Studies have shown that shortly after cardiac transplanta-tion a complete denervation exists and that later onre-innervation occurs in the transplanted organ [22]. Ithas been shown with 123I-MIBG that sympathetic functionhas a prognostic value. In patients with congestive heartfailure, for example, prognosis is worse for patients withimpaired presynaptic function [23]. Recently, it was shownthat in patients with idiopathic ventricular fibrillation 123I-MIBG was able to predict future arrhythmic events [24].123I-MIBG can be used to predict functional improvementof left-ventricular function after initiation of β-blocker andangiotensin converting enzyme (ACE) inhibitor therapies[25]. Assessment of sympathetic innervation may revealadverse effects of anthracycline therapy on cardiac functionmuch earlier than the established parameter of left-ventricular ejection fraction as a study from Lekakis andcolleagues has shown [26].

Cardiomyopathies that give rise to potentially lethalarrhythmias are associated with typical aberrations insympathetic function. It has been shown with PET usingthe radiotracers 11C-MHED and 11C-CGP 12177 for pre-and postsynaptic imaging that in patients with hypertrophic

Fig. 2 Evaluation of left ventricular perfusion, end-diastolic andend-systolic cavity volumes, global and regional wall motion andwall thickening using ECG-gated single photon emission tomo-graphy (SPECT) and the radiotracer 99mTc-tetrofosmin in apatient. Endo- and epicardial contours superimposed on verticallong (a, b) and short (c, d) axis slices at end-diastole (a, c) andend-systole (b, d). Values for tracer uptake are normalised with

respect to the myocardial maximum found in any of the ECG-defined gates. Three-dimensional display of myocardial perfusion(e), derived from the ungated data, and wall motion (f), viewedinferiorly. The solid grey surface in subimage F represents themidmyocardial contour (halfway between the endocardial andepicardial contours) for the end-systolic gate, the wire meshsurface that for the end-diastolic gate

1424

cardiomyopathy (HCM), arrhythmogenic right ventriculardysplasia (ARVD) and right-ventricular outflow tractarrhythmia (RVOT) presynaptic neurotransmitter recyclingand postsynaptic receptor density are downregulated(Fig. 3). In contrast, in patients with the Brugada syndromethe presynaptic neurotransmitter recycling is upregulatedwhile postsynaptic receptor densities were within thenormal range (Fig. 3) [13–16]. This change in innervationmirrors a difference in clinical appearance: whereas theformer cardiomyopathies lead to arrhythmias under stressconditions, the Brugada syndrome is associated withpotentially lethal arrhythmias occurring at rest. A reliablenon-invasive test to identify patients at risk for suddencardiac death who need an implanted cardioverter-defib-rillator (ICD) device is one of the potential future roles ofinnervation imaging.

The parasympathetic nervous system uses acetylcholineas neurotransmitter with muscarinic receptors postsynapti-cally. 11C-methylquinuclidinylbenzilat (11C-MQNB) is aradiotracer that allows quantitation of muscarinic receptordensities [27]. A study from Le Guludec et al. exemplifiesthe application of this radiotracer [28]: they foundupregulation of muscarinic receptor density in patientswith idiopathic dilated cardiomyopathy.

Plaque imaging

The imaging of the vessel lumen by angiography is notadequate to assess the risk for individual plaques to rupture.Many myocardial infarctions are caused by plaques thatrupture before they cause critical stenoses. These plaquescan escape detection by imaging modalities that visualisethe vessel lumen instead of the plaque itself [29]. Theassessment of calcification within the arterial wall bymeans of electron beam or multislice CT may yield anindependent risk factor for coronary heart disease [30] butit, like angiography, also fails to identify unstable plaquesthat are likely to rupture and require intervention. Theassessment of plaque instability may also play an importantclinical role in assessing the efficacy of systemic therapiesaimed at plaque stabilisation such as oral intake of3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA)reductase inhibitors (“statins”) on an individual basis.

Several different radiotracers for imaging plaqueinstability (“plaque imaging”) have begun to emerge inrecent years and several of them show great promise.

18F-FDG 18F-FDG is accumulated in areas with highconcentrations of activated macrophages and leucocytes

Fig. 3 Quantitation of pre-synaptic and post-synaptic sympatheticfunction in groups of patients with different cardiomyopathiesknown to be associated with ventricular arrhythmias measured withPET and the radiotracers 11C-MHED and 11C-CGP 12177 (see textfor details). The graph shows post-synaptic beta receptor density onthe abscissa and pre-synaptic catecholamine recycling on theordinate. Given are the mean and the standard deviation for each

type of cardiomyopathy. Data from patients with differentcardiomyopathies [arrhythmogenic right ventricular cardiomyopathy(ARVCM, number of patients: n=8), right outflow tract tachycardia(RVOT, n=8), hypertrophic cardiomyopathy (HCM, n=9) andBrugada syndrome (n=9)] and from normal controls (n=19 for 11C-CGP and n=10 for 11C-MHED) are included

1425

and therefore serves as a surrogate marker of inflamma-tion. Studies in the carotid artery have shown thatsymptomatic plaques show higher 18F-FDG uptake thannon-symptomatic plaques [31]. Our own studies in thecarotid and great arteries of the body with combined PET/CT have shown that 18F-FDG uptake and calcification arediscordant in many cases, underscoring the assumptionthat measurement of calcification cannot be used toidentify unstable plaques. Application of this surrogatemarker for coronary artery imaging is hampered by theoften significant background uptake in the left-ventricularmyocardium and by heart movement during imageacquisition. Although controlled metabolic conditionsmay limit myocardial glucose utilisation, 18F-FDG isstill not optimal for effective coronary plaque imaging.

Matrixmetalloproteinase inhibitors The integrity of thefibrous cap of a plaque, which separates the thrombogeniccore from the vessel lumen, is an important determinant ofplaque stability. Proteolysis of the extracellular matrix byactivated matrixmetalloproteinases (MMPs) can serve as apotent indicator for plaque instability. Radiotracers basedon MMP inhibitors that only bind to activated MMPs havebeen developed and successfully tested in animal experi-ments [32, 33]. Radiotracer development can takeadvantage of the fact that MMPs are well characterised

and that a number of synthetic MMP inhibitors have beendeveloped for cancer treatment. As an example, our grouphas modified and labelled with the gamma emitter Iodine-123 (123I) the broad-spectrum nonpeptidyl sulfonamide-based hydroxamate-derivative CGS 27023A to yield theradiotracer 123I-HO-I-CGS 27023A [32]. Enzymaticassays confirmed that the capacity to inhibit MMP-2 andMMP-9 subtypes of the MMPs was retained afterlabelling. Specific uptake of this radiotracer into plaqueswas shown in an animal experiment using apolipoproteinE-deficient (Apo E−/−) mice undergoing ligature of thecarotid artery and being put on a high cholesterol diet withsubsequent development of MMP-rich vascular lesions atthe site of the ligature (Fig. 4) [33]. PET imaging ispossible when using 124I or 18F as the radiolabel. Severalother MMP inhibitors have been successfully labelled with11C or 18F for PET imaging [34]. Recently precursors forSPECT- and PET-imaging based on barbiturate derivativeswith high MMP binding potential have been developedthat promise improved MMP selectivity [35].

Annexin V and caspase inhibitors The presence of apo-ptosis of macrophages and other cells can also serve as amarker for unstable plaques. Annexin V is known to bindto phosphatidylserine, a negatively charged membranephospholipid externalised to the cell surface during the

Fig. 4 In vivo imaging of MMPactivity within a MMP-richvascular lesion. Apo E-deficientas well as wildtype mice withprevious temporary ligation ofthe left common carotid arteryand Apo E-deficient sham-op-erated mice (all on a high cho-lesterol diet) were examinedwith SPECT and the radiotracer123I-HO-I-CGS 27023A. MMP-rich vascular lesions are knownto develop only in Apo E-deficient mice at the site ofligature. Considerable traceruptake in the neck (white arrow)was seen in the operated trans-genic mice (a), uptake in thesame mice predosed with coldcomponent (b), in transgenicsham-operated mice (c) and inoperated wildtype mice (d) wasconsiderably lower. Activityvalues are expressed as percentof total injected dose (%ID) [32]

1426

early apoptotic signalling cascade. Annexin V labelledwith 99mTc can be used to visualise vulnerable plaques inanimals and in humans using SPECT. Kolodgie et al. haveobserved in rabbits that plaques have a high uptake of thisradiotracer and that this uptake reflects the amount ofapoptosis [36]. Kietselaer and coworkers have success-fully applied 99mTc-Annexin V in humans [37]. Hartungand colleagues have shown in rabbits that cholesterollowering therapy by dietary modification and statin intakeleads to a decrease in 99mTc-Annexin V accumulation inatherosclerotic lesions, implying plaque stabilisation [38].However, phosphatidylserine is not an exclusive in-vivotarget for apoptosis. It is also elevated in other forms ofcell stress not necessarily leading to apoptosis. Activationof caspases inside the cells signals the initiation of anirreversible cascade towards apoptosis. Caspases, there-fore, are likely to be better targets for molecular imagingof apoptosis. Targeting an intracellular enzyme with aradiotracer, e.g., a radiolabelled caspase inhibitor, isconceptually more challenging than targeting a cell-surface receptor such as phosphatidylserine. The futureapplicability of this approach therefore remains to beseen.

Apoptosis also plays an important role in diseasesaffecting cardiomyocytes such as myocardial infarction,myocarditis and non-ischaemic cardiomyopathies. It hasbeen shown, for example, that imaging of Annexin Vaccumulation depicts tissue at risk after myocardialinfarction with reversible damage [39]. Elevated AnnexinV uptake can be found after heart transplantation in case ofacute transplant rejection [40] that correlates with caspase-3 staining in histology.

The radiotracer 99mTc-glucarate, the 99mTc complex of asix-carbon dicarboxylic acid, depicts necrosis early aftercell death and may complement Annexin V for theassessment of cardiac tissue [41].

Endothelin Endothelin (ET) is a strong vasoconstrictor. Ithas three isoforms ET-1, ET-2 and ET-3 [42] that mainlybind to the two ET receptors, ETAR and ETBR. ETAR canbe found in smooth muscle cells, whereas ETBR is locatedprimarily in vascular endothelial cells. ET-1 is generatedfrom its precursor big ET-1 by help of the endothelin-converting enzyme (ECE)-1 and plays an important role inthe regulation of the vasotonus. Elevated levels of ET-1lead to endothelial dysfunction. Additionally, ET-1 isimplicated in angiogenesis, restenosis after balloon angio-plasty and atherosclerosis. ET-1, ECE-1 and ET-receptorupregulation has been found in atherosclerotic lesionsespecially in the inflammatory stages and may therefore besuitable targets for plaque imaging [43]. Johnström andcolleagues have successfully labelled ET-1 with thepositron emitter 18F. Biodistribution and organ uptakewere visualised in rats with a dedicated small animal PETscanner [44]. Myocardial binding of 18F-ET-1 with goodET-receptor affinity was observed in vitro. Relevant

myocardial uptake in vivo was only observed afterblockage of the ETBR receptors, with a high density inlung and kidneys, with the antagonist BQ788 to preventrapid clearance from the blood in the lung and kidneysbefore reaching the myocardium. Dinkelborg, Meding andcoworkers have shown an elevated uptake of a 99mTc-labelled endothelin derivative in atherosclerotic plaques ina rabbit model both in vivo and ex vivo [45]. Severalantagonists of ET-receptors, either selective or unselectivefor ETAR or ETBR, have been radiolabelled and success-fully tested in animal studies, most of them for use withPET (e.g., [46]). Johnström et al. have seen an uptake ofthe ETBR selective antagonist 18F-BQ3020 in athero-sclerotic plaques co-localised to macrophages, so that thistracer may have a potential to detect unstable macrophage-rich plaques [46]. In addition to plaque imaging, ET-related radiotracers can also be used for investigations ofendothelial dysfunction, angiogenesis and restenoses.

Imaging technologies for plaque imaging Molecularplaque imaging is easier for the large arteries of thebody and the carotid arteries than for the small andconstantly moving coronary arteries. Before imaging ofcoronary arteries with radiotracer techniques becomesclinically feasible, a considerable amount of developmentin the fields of physics and instrumentation, mathematicsand computer science will be required. Cardiac andrespiratory gating techniques, where data acquisition islimited to a well-defined position within the heart orrespiratory cycles, have been implemented, however witha negative impact on count statistics. More elaboratemethods for motion compensation are currently beingdeveloped and evaluated [47, 48]. They rely on list modedata acquisition techniques already implemented in mostPET scanners where every single event during the scan isstored individually together with biomedical data such asthe electrocardiogram (ECG) and measurements of respi-ratory movement. In the future, combined molecular-anatomical imaging with SPECT or PET combined withCT, MRI or echocardiography is likely to play anincreasingly important role, especially for imaging withhighly specific radiotracers that do not reveal a lot ofanatomical background information.

Hypoxia and angiogenesis

Several radiolabelled PET and SPECT tracers for imagingof hypoxic tissue have been developed in recent years,mainly for applications in oncology, but also for applica-tions in cardiovascular diseases. These include the PETradiotracers 18F-fluoromisonidazole (18F-FMISO) and *Cu(II)-diacetyl-bis(N4-methylthiosemicarbazone) (*Cu-ATSM) where * can stand for any of the positron-emittingcopper nuclides 60Cu, 61Cu, 62Cu or 64Cu. The applicabilityof both radiotracers for myocardial imaging has been

1427

established in a number of animal experiments [49, 50].Imaging of hypoxia may potentially be used in patients todetermine jeopardised, but salvageable myocardium in acutecoronary syndromes.

Interest in the imaging of angiogenesis has been fosteredmainly by cancer research in order to evaluate the effect oftherapeutic interventions aimed at preventing tumor-in-duced angiogenesis. In cardiovascular medicine there istremendous interest not only in therapeutic approaches thatstrive for preventing angiogenesis, e.g., in the context ofrestenoses after angioplasty with or without stent implan-tation, but also in approaches that promote angiogenesiseither by administration of angiogenic factors, such as thevascular endothelial growth factor (VEGF) or the basicfibroblast growth factor (bFGF) [51, 52], or by genetherapeutic approaches [53] in patients with ischaemicheart disease. The non-invasive molecular imaging ofangiogenesis is a formidable means to improve therapymonitoring.

VEGF is a key mediator of angiogenesis. Lu et al. haveshown in an in vivo animal experiment in rabbits thatuptake of VEGF121 labelled with the SPECT radionuclide111In is considerably higher in an ischaemic hind legcompared to the contralateral non-ischaemic hind leg [54],reflecting the higher expression of the VEGF receptors 1(FLT-1) and 2 (KDR). Another important mediator is theαvβ3-integrin receptor, a heterodimeric transmembraneglycoprotein with α and β subunits that mediates not onlyadhesion, but also proliferation and differentiation ofendothelial cells [55]. Since the expression of αvβ3-integrin is upregulated with angiogenesis, it should be asuitable target for molecular imaging [55]. Peptidescontaining the amino acid sequence arginine-glycine-aspartic acid (RGD) can interact with the αvβ3-integrinreceptor. Consequently, a large number of radiotracerscontaining the RGD sequence have been developed andevaluated. The iodinated compound *I-gluco-RGD, where* can stand for any of 123I, 124I, 125I and 131I, and thefluorinated PET radiotracer 18F-galacto-RGD have beenapplied in several preclinical and clinical imaging studies[55, 56]. The structurally different radiotracer quinolinone111In-RP748 has also been successfully tested in animalmodels of myocardial infarction, vascular injury andtransplant arteriopathy (e.g., [57–59]). Ongoing develop-ment strives for optimisation of binding potency.

Reporter gene imaging

Therapeutic strategies based on the introduction of genesinto the genome of living organisms have been developedfor several cardiac diseases, e.g., atherosclerosis, myocar-dial infarction, heart failure and arrhythmias [52, 60]. Sofar, the results from the first studies have been equivocalwith respect to the therapeutic efficacy of gene therapeutic

approaches. Additionally, gene therapy in its present formhas been implicated in serious side effects such as theinduction of malignant diseases. Therefore, considerableresearch efforts are required to develop gene therapy intoan effective and safe procedure. Molecular imagingtechniques offer the unique possibility to closely monitorgene distribution and gene expression in vivo non-invasively as a direct way to assess the efficacy of genetransfer. Radiotracer methods also allow for correlativeimaging of gene expression with a wide variety offunctional parameters such as myocardial perfusion, me-tabolism, innervation and global and regional wall motionto reveal the effects of therapeutic gene expression oncardiac function.

A very elegant method to monitor gene expression is theindirect method of reporter gene imaging. The therapeuticgene is coupled with a reporter gene so that both genes areintroduced into the DNA by viral or non-viral vectors.Expression of the therapeutic gene is directly reflected inexpression of the reporter gene. The reporter gene productmust be chosen so that it interacts with a radiotracer(reporter probe) to produce a signal. Several reporter gene/probe systems have been developed for molecular radio-tracer imaging, which utilise different kinds of interaction(Fig. 5):

– Enzyme-based reporter gene imaging: the reportergene product is an enzyme that converts the reporterprobe to metabolites trapped in the cell.

– The product of the herpes simplex virus type 1thymidine kinase reporter gene (HSV1-tk) or itsmutant derivative HSV1-sr39tk is an enzyme thatphosphorylates the nucleoside reporter probes 5-iodo-1-(2-deoxy-2-fluoro-β-D-arabinofuranosyl)uracil (FIAU) and 9-(4-fluoro-3-hydroxymethylbu-tyl)guanine (FHBG). After phosphorylation, themetabolites are trapped in the cells and can beimaged by PET or SPECT with adequate tracers.This reporter gene/probe system has seen applica-tion in autoradiography, in small-animal PETimaging [61], in large animals on a clinical PETscanner [62] and in humans ([63], brain tumormodel). This exemplifies the quick translation ofmolecular imaging techniques from ex vivo (auto-radiography) over animal to human imaging offeredby radiotracer techniques.

– Receptor-based reporter gene imaging: the reportergene product is a receptor that serves as the target for areporter probe.

– The product of the dopamine D2 receptor (D2R)reporter gene is a surface receptor that serves as thetarget for 3-(2-fluoroethyl)-spiperone (FESP) suit-able for PET imaging when labelled with 18F [64].

1428

– The product of the somatostatin type 2 receptor(SSTR2) reporter gene is a surface receptor thatserves as the target for the 111In-labelled SPECTtracer pentreotide (Octreoscan, Mallinckrodt, St.Louis, MO).

– Transporter-based reporter gene imaging: the reportergene product is a transport protein that leads tointracellular accumulation of a reporter probe.

– The product of the human sodium/iodine symporter(hNIS) reporter gene is a transport protein that

facilitates iodine uptake and is suitable for SPECTand PET imaging with radioiodine nuclides [65].

Successful co-expression of the two reporter gene/probesystems HSV1-sr39tk and D2R [66] as well as co-expression of a therapeutic gene for VEGF expressionand the reporter gene HSV1-sr39tk [67] have establishedthe feasibility of monitoring “angiogenic gene expressionwith a PET reporter gene and probe noninvasively,quantitatively, and repetitively”. [67].

Cell imaging

Stem and progenitor cells have a recognised potential fortissue regeneration. Myocardial tissue damaged by coronaryartery disease and heart failure is one of the obvious targetsfor cell-based therapies. Although the first studies haveshown successful application of these therapies (e.g., [68]),many details, such as optimal cell selection and deliverystrategy as well as the mechanisms of action, are still unclear.

Non-invasive molecular imaging of the fate of stem orprogenitor cells in vivo can contribute to the futuredevelopment of this field. Again, molecular imaging withradiotracer techniques offers the additional advantage thatfunctional imaging of perfusion, metabolism, innervation,apoptosis and other parameters can be performed at thesame time to elucidate the effects of stem or progenitor celltherapies.

Direct and indirect methods for stem and progenitor cellimaging are feasible. Ex vivo labelling of the cells beforeinjection have been used to monitor cell migration [69];however, the radioactive signal is not linked to cellviability, and it is not maintained during cell division.Radioactive decay attenuates the signal, and radiation-induced cell damage may impair the validity of thisapproach.

Indirect imaging of stem cell migration and proliferationis possible when applying the reporter-gene technique tostem or progenitor cells. Wu and colleagues have labelledmurine embryonic stem cells with a triple fusion reportergene construct consisting of fluorescence, bioluminescenceand PET markers [HSV1-truncated thymidine kinase (tTK)reporter gene] [70]. They were able to show that thetransfected stem cells survived, proliferated and differ-entiated and that the reporter genes were passed on duringcell division. Transcriptional profiling of the transfectedcells revealed that expression of many genes differed fromnative cells. PET imaging using the reporter probe 18F-FHBG proved successful. Additional metabolic imagingwith 18F-FDG in the same imaging session demonstratedcorrelative cell/functional imaging and provided anatomiclandmarking not possible with 18F-FHBG imaging alone.

Fig. 5 Principles of reporter gene/probe imaging systems. (a) Thegene product of the reporter gene is an enzyme. After introduction ofthe reporter gene into the genome inside the cell nucleus by meansof a viral or non-viral vector (step 1), the reporter gene istranscripted into messanger ribonucleine acid (m-RNA) (step 2),which is then translated into the reporter enzyme at the ribosomes(step 3). The enzyme enables transformation of the labelled reporterprobe into a metabolite that is trapped inside the cell and can be usedfor imaging (step 4); (b) The gene product of the reporter gene is areceptor. After gene transfection, transcription and translation (steps1–3) the generated receptor is expressed on the cell surface andserves as the target for the the labelled reporter probe. (c) The geneproduct is a transport protein. After gene transfection, transcriptionand translation (steps 1–3) the generated transmembraneous trans-port protein enables the reporter probe to enter the cell andaccumulate

1429

Conclusions and perspectives

While the assessment of myocardial perfusion andmetabolism with radiotracer techniques has been longestablished, newer methodologies, especially in the fieldsof plaque, gene and cell imaging, have started to emergeand promise an exciting future for both basic research andclinical routine. Radiotracer techniques allow for a uniquetranslational application from animals to humans. Theyare very sensitive so that molecular targets with a lowmolar density can be assessed without disturbing thebiochemical system, provided that specific radiotraceractivity is high enough. The combination of SPECT orPET, providing true molecular information, with sonog-

raphy, CT or MRI, providing anatomic, morphologicaland functional information, in one device is likely tobecome increasingly popular. Imaging of the constantlymoving heart and especially the coronary arteries withSPECT and PET is a challenge. The development ofadequate imaging strategies requires an interdisciplinaryapproach with specialists from the fields of engineering,physics, mathematics and computer science, chemistryand medicine working together.

Acknowledgements This work was supported by the DeutscheForschungsgemeinschaft (DFG), Sonderforschungsbereich SFB 656MoBIL, Münster, Germany (Projects A1-A5, B3, C1, C2, PM3).

References

1. Blumgart HL, Yens OC (1927) Studieson the velocity of blood flow: I. Themethod utilized. J Clin Invest 4:1–13

2. Schäfers KP, Stegger L, Barnard C et al(2005) ECG-triggered high-resolutionpositron emission tomography: abreakthrough in cardiac molecular im-aging of mice. Eur J Nucl Med MolImaging 32:383

3. Assmann G, Cullen P, Schulte H (2002)Simple scoring scheme for calculatingthe risk of acute coronary events basedon the 10-year follow-up of the pro-spective cardiovascular Münster(PROCAM) study. Circulation105:310–315

4. Vanzetto G, Ormezzano O, Fagret D,Comet M, Denis B, Machecourt J(1999) Long-term additive prognosticvalue of thallium-201 myocardial per-fusion imaging over clinical and exer-cise stress test in low to intermediaterisk patients: study in 1137 patientswith 6-year follow-up. Circulation100:1521–1527

5. Hachamovitch R, Berman DS, Kiat Het al (1996) Exercise myocardial per-fusion SPECT in patients withoutknown coronary artery disease: incre-mental prognostic value and use in riskstratification. Circulation 93:905–914

6. Beller GA, Bergmann SR (2004) Myo-cardial perfusion imaging agents:SPECT and PET. J Nucl Cardiol 11:71–86

7. Germano G, Kiat H, Kavanagh PB et al(1995) Automatic quantitation of ejec-tion fraction from gated myocardialperfusion SPECT. J Nucl Med36:2138–2147

8. Stegger L, Biedenstein S, Schäfers KP,Schober O, Schäfers MA (2001) Elasticsurface contour detection for the mea-surement of ejection fraction in myo-cardial perfusion SPET. Eur J NuclMed 28:48–55

9. Sharir T, Germano G, Kavanagh PB etal (1999) Incremental prognostic valueof post-stress left ventricular ejectionfraction and volume by gated myocar-dial perfusion single photon emissioncomputed tomography. Circulation100:1035–1042

10. Schelbert HR (2002) 18F-deoxyglucoseand the assessment of myocardial via-bility. Semin Nucl Med 32:60–69

11. Stegger L, Schäfers KP, Flögel U et al(2005) Monitoring left ventricular di-lation in mice with PET. J Nucl Med46:1516–1521

12. Tamaki N, Morita K, Kuge Y,Tsukamoto E (2000) The role of fattyacids in cardiac imaging. J Nucl Med41:1525–1534

13. Schäfers M, Dutka D, Rhodes CG et al(1998) Myocardial presynaptic andpostsynaptic autonomic dysfunction inhypertrophic cardiomyopathy. Circ Res82:57–62

14. Wichter T, Schäfers M, Rhodes CG etal (2000) Abnormalities of cardiacsympathetic innervation in arrhythmo-genic right ventricular cardiomyopathy:quantitative assessment of presynapticnorepinephrine reuptake and postsyn-aptic beta-adrenergic receptor densitywith positron emission tomography.Circulation 101:1552–1558

15. Schäfers M, Lerch H, Wichter T et al(1998) Cardiac sympathetic innervationin patients with idiopathic right ven-tricular outflow tract tachycardia. J AmColl Cardiol 32:181–186

16. Kies P, Wichter T, Schäfers M et al(2004) Abnormal myocardial presyn-aptic norepinephrine recycling in pa-tients with Brugada syndrome.Circulation 110:3017–3022

17. de Jong RM, Willemsen AT, Slart RHet al (2005) Myocardial beta-adreno-ceptor downregulation in idiopathicdilated cardiomyopathy measured invivo with PET using the new radioli-gand (S)-[11C]CGP12388. Eur J NuclMed Mol Imaging 32:443–447

18. Wagner S, Law MP, Riemann B et al(2006) Synthesis of a F-18-labelledhigh affinity beta1-adrenoceptor PETradioligand based on ICI 89, 406.J Label Compd Radiopharm 49:177–195

19. Law MP, Osman S, Pike VW et al(2000) Evaluation of [11C]GB67, anovel radioligand for imaging myocar-dial alpha 1-adrenoceptors with posi-tron emission tomography. Eur J NuclMed 27:7–17

20. Inoue H, Zipes DP (1987) Results ofsympathetic denervation in the canineheart: supersensitivity that may be ar-rhythmogenic. Circulation 75:877–887

21. Inobe Y, Kugiyama K, Miyagi H et al(1997) Longlasting abnormalities incardiac sympathetic nervous system inpatients with coronary spastic angina:quantitative analysis with iodine 123metaiodobenzylguanidine myocardialscintigraphy. Am Heart J 134:112–118

22. Wilson RF, Laxson DD, ChristensenBV, McGinn AL, Kubo SH (1993)Regional differences in sympatheticreinnervation after human orthotopiccardiac transplantation. Circulation88:165–171

1430

23. Yamada T, Shimonagata T, FukunamiM et al (2003) Comparison of theprognostic value of cardiac iodine-123metaiodobenzylguanidine imaging andheart rate variability in patients withchronic heart failure: a prospectivestudy. J Am Coll Cardiol 41:231–238

24. Paul M, Schäfers M, Kies P et al (2006)Impact of sympathetic innervation onrecurrent life-threatening arrhythmiasin the follow-up of patients with idio-pathic ventricular fibrillation. Eur JNucl Med Mol Imaging 33:866–870

25. Nakata T, Wakabayashi T, Kyuma M,Takahashi T, Tsuchihashi K,Shimamoto K (2004) Cardiac metaio-dobenzylguanidine activity can predictthe long-term efficacy of angiotensin-converting enzyme inhibitors and/orbeta-adrenoceptor blockers in patientswith heart failure. Eur J Nucl Med MolImaging 32:186–194

26. Lekakis J, Prassopoulos V,Athanassiadis P, Kostamis P,Moulopoulos S (1996) Doxorubicin-induced cardiac neurotoxicity: studywith iodine 123-labeled metaiodoben-zylguanidine scintigraphy. J NuclCardiol 3:37–41

27. Delforge J, Le Guludec D, Syrota A etal (1993) Quantification of myocardialmuscarinic receptors with PET inhumans. J Nucl Med 34:981–991

28. Le Guludec D, Cohen-Solal A,Delforge J, Delahaye N, Syrota A,Merlet P (1997) Increased myocardialmuscarinic receptor density in idio-pathic dilated cardiomyopathy: an invivo PET study. Circulation 96:3416–3422

29. Falk E, Shah PK, Fuster V (1995)Coronary plaque disruption.Circulation 92:657–671

30. Greenland P, LaBree L, Azen SP,Doherty TM, Detrano RC (2004) Cor-onary artery calcium score combinedwith Framingham score for risk pre-diction in asymptomatic individuals.JAMA 29:210–215

31. Rudd JH,Warbourton EA, Fryer TD et al(2002) Imaging atherosclerotic plaqueinflammation with [18F]-fluorodeoxy-glucose positron emission tomography.Circulation 105:2708–2711

32. Kopka K, Breyholz HJ, Wagner S et al(2004) Synthesis and preliminary bio-logical evaluation of new radioiodinatedMMP inhibitors for imaging MMP ac-tivity in vivo. Nucl Med Biol 31:257–267

33. Schäfers M, Riemann B, Kopka K et al(2004) Scintigraphic imaging of matrixmetalloproteinase activity in the arterialwall in vivo. Circulation 109:2554–2559

34. Furumoto S, Takashima K, Kubuta K,Ido T, Iwata R, Fukuda H (2003)Tumor detection using 18F-labeled ma-trix metalloproteinase-2 inhibitor. NuclMed Biol 30:119–125

35. Breyholz H-J, Schäfers M, Wagner S etal (2005) C-5-disubstituted barbituratesas potential molecular probes for non-invasive matrix metalloproteinaseimaging. J Med Chem 48:3400–3409

36. Kolodgie FD, Petrov A, Virmani R et al(2003) Targeting of apoptotic macro-phages and experimental atheroma withradiolabeled annexin V: a techniquewith potential for noninvasive imagingof vulnerable plaque. Circulation108:3134–3139

37. Kietselaer BL, Reutelingsperger CP,Heidendal GA et al (2004) Noninvasivedetection of plaque instability with useof radiolabeled annexin A5 in patientswith carotid-artery atherosclerosis. NEngl J Med 350:1472–1473

38. Hartung D, Sarai M, Petrov A et al(2005) Resolution of apoptosis inatherosclerotic plaque by dietary mod-ification and statin therapy. J Nucl Med46:2051–2056

39. Hofstra L, Liem IH, Dumont EA et al(2000) Visualisation of cell death invivo in patients with acute myocardialinfarction. Lancet 356:209–212

40. Narula J, Acio ER, Narula N et al(2001) Annexin-V imaging for non-invasive detection of cardiac allograftrejection. Nat Med 7:1347–1352

41. Mariani G, Villa G, Rossettin PF et al(1999) Detection of acute myocardialinfarction by 99mTc labeled D-glu-cararic acid imaging in patients withacute chest pain. J Nucl Med 54:1832–1839

42. Inoue A, Yanagisawa M, Kimura S et al(1989) The human endothelin family:three structurally and pharmacologi-cally distinct isopeptides predicted bythree separate genes. Proc Natl AcadSci USA 86:2863–2867

43. Iwasa S, Fan J, Shimokama T, Nagata M,Watanabe T (1999) Increased immuno-reactivity of endothelin-1 and endothelinB receptor in human atheroscleroticlesions. A possible role in atherogenesis.Atherosclerosis 146:93–100

44. Johnström P, Harris NG, Fryer TD et al(2002) 18F-Endothelin-I, a positronemission tomography (PET) radioli-gand for the endothelin receptorsystem: radiosynthesis and in vivoimaging using microPET. Clin Sci103:4S–8S

45. Dinkelborg LM, Duda SH, Hanke H,Tepe G, Hilger CS, Semmler W (1998)Molecular imaging of atherosclerosisusing a technetium-99m-labeled endo-thelin derivative. J Nucl Med 39:1819–1822

46. Johnström P, Rudd JH, Richards HK etal (2006) Imaging endothelin ET(B)receptors using [18F]-BQ3020: in vitrocharacterization and positron emissiontomography (microPET). Exp BiolMed 231:736–740

47. Livieratos L, Stegger L, BloomfieldPM, Schafers K, Bailey DL, Camici PG(2005) Rigid-body transformation oflist-mode projection data for respiratorymotion correction in cardiac PET. PhysMed Biol 50:3313–3322

48. Dawood M, Lang N, Jiang X, SchafersKP (2006) Lung motion correction onrespiratory gated 3-D PET/CT images.IEEE Trans Med Imaging 25:476–485

49. Shelton ME, Dence CS, Hwang DR,Herrero P, Welch MJ, Bergmann SR(1990) In vivo delineation of myocar-dial hypoxia during coronary occlusionusing fluorine-18 fluoromisonidazoleand positron emission tomography: apotential approach for identification ofjeopardized myocardium. J Am CollCardiol 16:477–485

50. Fujibayashi Y, Cutler CS, Anderson CJet al (1999) Comparative studies of Cu-64-ATSM and C-11-acetate in an acutemyocardial infarction model: ex vivoimaging of hypoxia in rats. Nucl MedBiol 26:117–121

51. Udelson JE, Dilsizian V, Laham RJ etal (2000) Therapeutic angiogenesiswith recombinant fibroblast growthfactor-2 improves stress and rest myo-cardial perfusion abnormalities in pa-tients with severe symptomatic chroniccoronary artery disease. Circulation102:1605–1610

52. Henry TD, Rocha-Singh K, Isner JM etal (2001) Intracoronary administrationof recombinant human vascular endo-thelial growth factor to patients withcoronary aftery disease. Am Heart J142:872–880

53. Giordano FJ, Ping P, McKirnan MD etal (1996) Intracoronary gene transfer offibroblast growth factor-5 increasesblood flow and contractile function inan ischemic region of the heart. NatMed 2:534–539

54. Lu E, Wagner WR, Schellenberger U etal (2003) Targeted in vivo labeling ofreceptors for vascular endothelialgrowth factor: approach to identifica-tion of ischemic tissue. Circulation108:97–103

55. Haubner R, Wester H-J, Reuning U etal (1999) Radiolabeled αvβ3 integrinantagonists: a new class of tracers fortumor targeting. J Nucl Med 40:1061–1071

1431

56. Beer AJ, Haubner R, Sarbia M et al(2006) Positron emission tomographyusing [18F]Galacto-RGD identifies thelevel of integrin alpha(v)beta3 expres-sion in man. Clin Cancer Res 12:3942–3949

57. Meoli DF, Sadeghi MM, KrassilnikovaS et al (2004) Noninvasive imaging ofmyocardial angiogenesis following ex-perimental myocardial infarction. J ClinInvest 113:1684–1691

58. Sadeghi MM, Krassilnikova S, Zhang Jet al (2004) Detection of injury-inducedvascular remodeling by targeting acti-vated alpha V beta 3 integrin in vivo.Circulation 110:84–90

59. Zhang J, Krassilnikova S, Gharaei AAet al (2005) Alphavbeta3-targeted de-tection of arteriopathy in transplantedhuman coronary arteries: an autoradio-graphic study. FASEB J 19:1857–1859

60. Donahue JK, Kikuchi K, Sasano T(2005) Gene therapy for cardiac ar-rhythmias. Trends Cardiovasc Med15:219–224

61. Wu JC, Inubushi M, Sundaresan G,Schelbert HR, Gambhir SS (2002)Positron emission tomography imagingof cardiac reporter gene expression inliving rats. Circulation 106:180–183

62. Bengel FM, Anton M, Richter T et al(2003) Noninvasive imaging of trans-gene expression by use of positronemission tomography in a pig model ofmyocardial gene transfer. Circulation108:2127–2133

63. Jacobs A, Voges J, Reszka R et al(2001) Positron-emission tomographyof vector-mediated gene expression ingene therapy for gliomas. Lancet358:727–729

64. MacLaren DC, Gambhir SS,Satyamurthy N et al (1999) Repetitive,non-invasive imaging of the dopamineD2 receptor as a reporter gene in livinganimals. Gene Ther 6:785–791

65. Miyagawa M, Beyer M, Wagner B et al(2005) Cardiac reporter gene imagingusing the human sodium/iodide sympor-ter gene. Cardiovasc Res 65:195–202

66. Chen IY, Wu JC, Min JJ et al (2004)Micro-positron emission tomographyimaging of cardiac gene expression inrats using bicistronic adenoviral vector-mediated gene delivery. Circulation109:1415–1420

67. Wu JC, Chen IY, Wang Y et al (2004)Molecular imaging of the kinetics ofvascular endothelial growth factor geneexpression in ischemic myocardium.Circulation 110:685–691

68. Orlic D, Kajstura J, Chementi S et al(2004) Bone marrow cells regenerateinfarcted myocardium. Nature428:664–668

69. Aicher A, Brenner W, Zuhayra M et al(2003) Assessment of the tissue distri-bution of transplanted human endothe-lial progenitor cells by radioactivelabeling. Circulation 107:2134–2139

70. Wu JC, Spin JM, Cao F et al (2006)Transcriptional profiling of reportergenes used for molecular imaging ofembryonic stem cell transplantation.Physiol Genomics 25:29–38

1432