Eur Radiol (2008) 18: 548–569
DOI 10.1007/s00330-007-0777-9 CARD IAC
Yuxiang Ye
Jan Bogaert
A key characteristic of myocardial tissue is the low from pre-clinical experiments using various proge...
be answered: how many stem cells reside after administra- Cell labeling with iron oxide
tion in the infarct ...
Table 1 Superparamagnetic iron oxide agents available for MRI iron oxide, SPIO superparamagnetic iron oxide,...
following magnetic labeling with SPIO-TA combinations, the most robust to detect SPIO tagged cells, being ...
Fig. 1a–d In vivo and ex vivo MRI of allograft hearts in rats, 1 day MPIO-particle labeling. MR microscopy at 11...
Fig. 3a–c Ferumoxide-labeled cells in the interventricular septum hypoenhancement corresponding to the zo...
the iron oxide labeling strategies may ca...
gene expression and cell targeting. Metabolic biotinylation positive findings should be regarded and interpret...
Table 2 Comprehensive use of cardiac MRI-MRS Myocardial contraction, consisting in w...
Table 3 Use of MRI in randomized clinical cell therapy trials in infarct transmurality, MPI myocardial perfusio...
Table 4 Impact of cell therapy on LV EF in acute and chronic progenitor cells, LVA left ventricular angiograp...
Table 4 (continued)
Study Baseline ...
Fig. 7 Impact of anatomical structures (i.e., papillary muscles and excluding the papillary muscles and trabec...
Fig. 9 Acute reperfused apical
infarction in a 51-year-old man.
CE-IR MRI in the first week
(1W), 4 months (4M) and...
Fig. 11 Short-axis contrast-enhanced inversion-recovery (CE-IR) myocardium. In case of true myocardial regen...
how to correctly interpret them. Simple administration of the three germ layers) or multipotent (i.e., t...
13. Kocher AA, Schuster MD, Szabolcs 19. Britten MB, Abolmaali ND, Assmus B, 25. Kang H-J, Lee ...
of 22


Published on: Mar 3, 2016
Published in: Health & Medicine      Technology      Business      

Transcripts - Nanomedicina5

  • 1. Eur Radiol (2008) 18: 548–569 DOI 10.1007/s00330-007-0777-9 CARD IAC Yuxiang Ye Jan Bogaert Cell therapy in myocardial infarction: emphasis on the role of MRI Received: 23 July 2007 Abstract Despite tremendous pro- itive study findings. Before new, Revised: 31 August 2007 gress in myocardial infarct (MI) larger clinical trials can be initiated, a Accepted: 4 September 2007 treatment, mortality rates remain sub- number of critical questions and issues Published online: 9 October 2007 stantial. Permanent loss of cardio- need to be considered starting with a # European Society of Radiology 2007 myocytes after ischemic injury, results scrutinized analysis of currently in irreversible loss of myocardial available data to extending our contractility, reduction in ventricular knowledge of the mechanism of scar- performance, and may initiate the less myocardial regeneration. Cardiac development of dilated heart failure. cell therapy necessitates a multidisci- The discovery that pluripotent pro- plinary approach, whereby imaging, genitor cells bear the capacity to in particular MRI, and the input of the differentiate to mature cardiac cells imaging specialist is crucial to the raised the hope of cell-based regen- success of cardiac cell regenerative erative medicine. Engraftment of stem medicine. MRI is an appealing tech- cells in the damaged myocardium, nique for cell trafficking depicting repair and functional improvement engraftment, differentiation and sur- appeared suddenly a nearby reality. vival. Endomyocardial cell adminis- Promising results in animal models, tration can be achieved safely with and preliminary studies reporting the MR fluoroscopy and MRI is without Y. Ye . J. Bogaert (*) feasibility and safety of adult stem cell any doubt the most accurate and Department of Radiology, therapy in MI patients led to the first reproducible technique to measure University Hospital K.U.L., double-blinded randomized, placebo- study end-points. Herestraat 49, controlled trials. The initial great 3000 Leuven, Belgium enthusiasm for this paradigm shift in Keywords MRI . Heart . e-mail: Tel.: +32-16-343681 MI treatment has been tempered by Myocardial . Infarction Fax: +32-16-343765 the mainly negative or modestly pos- Introduction mias, preventing thrombo-embolic events, and reducing LV remodeling (e.g., angiotensin-converting enzyme inhi- In the last two decades, a substantial decrease in early and bition); and thirdly, cardiac revalidation and secondary late mortality in acute myocardial infarction (MI) patients prevention programs [2]. Nevertheless, large multicenter has been achieved mainly due to wide acceptance of the studies have shown despite early and successful reperfu- open-vessel theory [1], emphasizing the need of early sion and optimized medication, the actual hospital mortal- reperfusion of the infarct-related artery by means of ity remains approximately 7% [3], not mentioning the percutaneous coronary intervention (PCI), intravenous evolution towards ischemic dilated cardiomyopathy (i.e., thrombolytic therapy, or coronary artery bypass graft chronic MI) and heart failure, and the permanently reduced (CABG) surgery; secondly, availability of supportive quality of life in a considerable number of surviving medication adequately controlling life-threatening arrhyth- patients.
  • 2. 549 A key characteristic of myocardial tissue is the low from pre-clinical experiments using various progenitor cell regenerative capacity, i.e., most myocardial cells are types (skeletal muscle myoblasts, cultured peripheral blood terminally differentiated soon after birth. In contrast to cells, bone-marrow derived hematopoietic and mesenchy- some lower species, like zebra fish, humans, and mammals mal stem cells, embryonic stem cells) as well as the indirect in general, do not bear the capacity to renew damaged stimulation and mobilization of BMCs, using cytokine myocardium [4]. Instead, ischemically damaged myocar- stimulation [granulocyte colony stimulating factor (G- dium is replaced in the weeks and months following the CSF)], in both small and large animal models have been a acute event by an afunctional, fibrotic scar. Although major stimulus toward small scale clinical trials evaluating discoveries in the past few years have demonstrated that the feasibility, safety and potential effectiveness of use of there is a pool of resident progenitor cells within the heart, adults stem cells in acute and chronic MI treatment [14– which can give rise to cardiomyocytes, endothelial cells 19]. Several larger, randomized, double-blind, place-con- and smooth muscle cells in vitro and in vivo, they appear trolled trials on the use of bone marrow cells in acute and insufficient to replace the acute loss of a large number of chronic infarct treatment reported mixed, i.e., modestly myocardial cells but rather seem to participate in the positive or negative, effects [20–34]. Although temporary continual replacement of apoptotic cardiomyocytes at a or sustained stem-cell-related effects have been reported low basal level [5, 6]. As a consequence, the amount of with improved myocardial perfusion, improved regional contractile cells lost, or the infarct size, is a crucial and/or global ventricular function, decreased infarct size, determinant of adverse LV remodeling. Loss of 1g of reduction in infarct-related mortality, after 6 years of myocardium represents approximately 20 million cardio- intensive research, the initial enthusiasm that began with a myocytes. To cause heart failure 25% of the ventricle great vision in 2001 has somewhat faded away. It is should be lost, while infarction of 40% results in cardio- currently not clear whether the above effects are truly genic shock [5]. related to cellular regeneration or need to be explained by paracrine effects. The way towards a “scarless heart” response in humans is still long. Unraveling the genetic Cardiac stem cell transplantation—hype or reality mechanism of scarless healing, creating attractive cues to stem cells while overcoming repulsive cues thus creating a In the past decade, stem cell transplantation has been hospitable environment for stem cells will be major steps widely investigated as a potential therapy for myocardial forward for myocardial regeneration [35, 36]. infarction and heart failure [1]. For the interested reader, we refer to some excellent review papers on this issue [5–9]. Cell therapy attempts to repopulate cardiac cell populations MRI and cardiac stem cell research and improve cardiac physiology by delivering progenitor cells, which have regeneration potentials of myocardium or Medical imaging techniques play a critical role in stem cell neovascularization, to the heart locally or systematically. In therapy research, by allowing in vivo tracking of cells, cardiac transplant patients, it was found that primitive assisting in stem cell delivery, and better understanding and circulating cells from the recipient have the ability to evaluation of the efficacy of the therapy. Among the migrate to the grafted heart, a process called cardiac different imaging modalities currently used [e.g. positron chimerism. This phenomenon is mainly responsible for emission tomography (PET), single-photon emission renewing endothelial cells (averaging 24%), but only rarely computerized tomography (SPECT), magnetic resonance for cardiomyocytes (averaging 0.04%) [10, 11]. Injury to a imaging (MRI), optical imaging, contrast ventriculography, target organ, such as the myocardium, is sensed by distant echocardiography], MRI has emerged to one of the stem cells, which migrate to the site of damage and undergo preferential players, enabling excellent assessment of alternate stem cell differentiation and promote structural response to myocardial regeneration therapy and can be and function repair. Anversa’s group in New York reported exploited to yield valuable insights into the mechanism of in 2001 in a mice model with an acute anterior MI, action of myocardial regeneration therapies [37–39]. formation of new myocytes, endothelial cells and smooth muscle cells generating de novo myocardium, significantly reducing infarct size and improving regional and global Cardiac MRI for cell trafficking ventricular function 9 days after injection of bone marrow stem cells (BMCs) having hematopoietic markers in the Ideally, in infarct patients, administered stem cells should peri-infarct area [12] (see Addendum). In another landmark migrate toward, home into and repopulate the irreversibly paper, Kocher et al. [13] reported in the same year damaged myocardium, improve local and global ventricular neovascularization of ischemic myocardium by bone- function, reduce life-treatening arrhythmias and overall lead narrow derived angioblasts preventing cardiomyocyte to a reduction of infarct-related mortality. One can under- apoptosis, reducing ventricular remodeling and improving stand that an essential pillar in stem cell research is labeling ventricular function. These and other encouraging results or trafficking of stem cells. Several crucial questions need to
  • 3. 550 be answered: how many stem cells reside after administra- Cell labeling with iron oxide tion in the infarct or peri-infarct area; what is the best administration route; how long and how many cells do Family of iron oxide MR contrast agents for cellular MRI survive after administration; do the stem cells transdifferen- Iron oxide nanoparticles are a family of superparamagnetic tiate and dedifferentiate into myocardial cells (i.e., cardio- contrast agents, consisting of a ferrite (maghemite or myocytes, smooth muscle cells, endothelial cells); do new magnetite) core and a polymeric coating. Based on their cardiomyocytes form electromechanical junctions with the size, they are categorized as monocrystalline iron oxide adjacent normal myocardium and do they function in nanoparticles (CLIO, MION, 10–30 nm in diameter) [54], synchronicity with the rest of the myocardium; what are ultra-small superparamagnetic iron oxide (USPIO, 10– the underlying mechanisms of improvement in myocardial 40 nm in diameter) [54, 55], superparamagnetic iron oxide perfusion and function, i.e., true cellular regeneration or (SPIO, 60–150 nm in diameter) [56, 57] and Micro- paracrine effect(s). These questions emphasize the need for metersized SPIO (MPIO, around one to a few micrometers both short-term (i.e., days to weeks) and long-term (i.e., in diameter) [58] (Table 1). Amongst them Ferucarbotran weeks to months) stem cell labeling. (Resovist), Ferumoxides (Endorem and Feridex) are FDA approved clinical grade reagents for enhancing MRI detection of liver tumors [59] and metastatic lymph General strategies to create image contrast nodes [60]. Recent attempts have demonstrated the novel applications of USPIO on inflammatory diseases [61], Medical imaging is largely based on differences in contrast, microcirculation permeability [62], blood volume [63] and enabling to discriminate different anatomical structures or MR angiography [55]. It is the strong superparamagnetic to differentiate between normal and pathological tissues or nature of these iron oxide nanoparticles that generates disease processes. As such, therapeutic cells delivered to local field gradients which enhances dephasing of the heart can not be distinguished from surrounding surrounding spins and therefore produces significant myocardium. In 1H MRI, a significant difference of signal voids on T2*-weighted sequences. By incorporating relaxation (T1, T2 or T2*) is needed to stand the iron oxide agents, cells of interest are visible as transplanted cells out of the background tissue. A variety hypointense areas on MRI. of approaches have been proposed to generate MR contrast for transplanted cells in the myocardium. Both paramag- Strategies of increasing intra-cellular iron loading For netic and superparamagnetic contrast agents [40–45] have cellular MRI, at least 5-10 pg/cell is a favorable iron been linked or incorporated to target cells, in order to loading, which provides and sustains sufficient suscep- change T1 or T2/T2* values. Fluorine-19 (19F)-labeling of tibility artifacts as a source of MR contrast even after a stem cells is an appealing alternative approach [46]. few cell divisions. By strong endocytosis, phagocytes Besides direct labeling, reporter genes have been applied such as macrophages and mononuclear cells are able to to preclinical research, in which they were engineered into incorporate SPIO particles spontaneously to an MRI- cells to serve as markers for MR cell tracking [47–50]. detectable level, which varies from 0.3 to 1 pg/cell, thereby creating only weak T2* effects. The intracellular iron may easily decline to a nondetectable level as a Direct labeling result of cell divisions or cellular excretions of inter- nalized SPIO, which is unfavorable for sensitive MRI Use of paramagnetic MR contrast agents, such as gadolin- detection and long-term cell tracing. Given the same ium (Gd) chelates [40–44] and manganese chloride coating material, it has been described that the efficiency (MnCl2) [45], is hampered by the low sensitivity on T1- of iron loading increases with the size of iron oxide weighted images, the requisite of a very high intracellular nanoparticles [64], probably due to the increase of concentration to shorten T1 times of surrounding water phagocytosis and endocytosis. Moreover, the coating molecules, and the concern of toxicity in free ion status. materials may also influence the cellular internalization Ferumoxides, a superparamagnetic agent, is dominant in of these agents [65]. Higher iron loading may be MR cell tracing, because of its robust effect on MR achieved by increasing the concentration of SPIO and relaxivity, biocompatibility and easy accessibility. These prolonging incubation time. However, impaired cell iron contents shorten T1 and T2 values at high concentra- viability and proliferation were observed in some of tion, while having significant T2* effects, which create a these settings, probably resulting from an elevated free magnetic field inhomogeneity in the surrounding tissue, radical level. In addition, the majority of cell types used thus creating a signal void [51]. By tissue digestion and for cell stem research are non-phagocytes, which are not magnetic cell sorting following transplantation, cells ready to uptake SPIOs or only internalize these nano- labeled with iron oxide may be isolated, which gives a particles at a very low rate through pinocytosis. There- means to quantify cell retention and future characterization fore, efficient methods to facilitate cellular uptake of iron of engrafted cells [52]. oxide nanoparticles are needed.
  • 4. 551 Table 1 Superparamagnetic iron oxide agents available for MRI iron oxide, SPIO superparamagnetic iron oxide, USPIO ultrasmall cell trafficking (CLIO crosslinked iron oxide, MION monocrystal- superparamagnetic iron oxide) line iron oxide nanoparticles, MPIO micrometer-sized paramagnetic Category Trade name Company Mean Coating material Other applications diameter (nm) MION VSOP-C184 Ferropharm 7 Citrate Blood pool contrast agent [53] Crosslinked aminated dextran MRI molecular imaging CLIO 45 [141] USPIO Sinerem/ Guerbet 20–40 Dextran T10T1 Metastatic lymph node imaging, Macrophage Combidex imaging, Blood pool agent [54] Advanced Magnetics CODE 7228 Advanced 18–20 Carboxyl-methyl-dextran Macrophage imaging, Blood pool agent Magnetics [55] SPIO Feridex/En- Guerbet 80–150 Dextran T10 Liver imaging dorem [56] Advanced Magnetics Resovist Schering 62 Carboxy-dextran Liver imaging, [57] MPIO Bangs Parti- Bangs La- 760–1630 Styrene/ Divinyl benzene with dragon Magnetic cell sorting cles [58] boratories green fluorescent dye In the current stage, a combination of SPIO particles complicated surface modification. Arbab and coworkers and polycationic transfection agents (TAs) is the most optimized this method in 2004, realizing a more universal widely applied strategy to enhance the efficiency of iron cell labeling strategy by using two commercially available, oxide labeling [66]. SPIO particles are coated with FDA-approved reagents [69]. Mixing Ferumoxides (FE) negative charged polymeric, which does not attach to the and protamin sulfate (Pro) at a ratio of 50:3 μg/ml, negatively charged cellular membrane. Transfection agents intracellular iron contents ranging from 1.47 pg/cell to are highly positive charged macromolecules, developed to 17.31 pg/cell, depending on the used cell type, were found facilitate macromolecule uptake into cells. With a positive after one night of incubation. No impairment was observed charge, the complex of SPIO and TA spontaneously in cell survival, proliferation, differentiation and other vital attaches to cellular membranes through electrostatic inter- cellular functions. Moreover, FE-Pro avoided aggregation action and effectively shuttles SPIO particles into cells by of SPIO particles, and showed to be less toxic to cells. endocytosis [66]. In 2002, Frank and coworkers proposed this nonspecific labeling approach, inspired by the success Bio-effects of cell labeling with iron oxide nanoparticles of magnetodendrimers [67], which is a specifically in conjunction of TA Cell-labeling with iron oxide holds synthesized SPIO coated with dendrimers allowing promise for rapid translation from bench to bedside; significantly improved iron oxide incorporation into however, intensive toxicity tests and pre-clinical examina- most mammalian cells [68]. They mixed ferumoxides tions are required for every protocol and cell type before (Feridex) or USPIO (MION-46 L) with various poly- any clinical application. The FDA-approved SPIO agents cationic transfection agents before incubation with cells. (Feridex, Resovist and Endorem) are cleaned mainly by With 25 μg Fe/ml TA-(U)SPIO complex, dramatic the reticuloendothelial system, which is capable of decrease of T2-values were observed in a variety of target metabolizing excess iron. However, non-phagocytic cells cells [e.g., mouse T-lymphocytes, human mesenchymal may not tolerate intracellular iron overloading. Moreover, stem cells (MSC)] following incubations from 4 to 48 h since the FE-Pro complex is a combination of two agents, [68]. As nonspecific magnetic labeling method, this further regulatory approval is required. Although, a variety approach was found to be as efficient as magnetoden- of cell types exhibited no significant changes in viability, drimer labeling and possesses the advantage of avoiding proliferation, differentiation and other vital functions
  • 5. 552 following magnetic labeling with SPIO-TA combinations, the most robust to detect SPIO tagged cells, being able to other studies pointed out the necessity of careful visualize single SPIO labeled cells even on a 1.5-T clinical evaluation of any labeling method on any cell type before MR scanner [80]. However, T2* sequences are prone to use in clinical trials [70–72]. any magnetic field inhomogeneity created by, not only magnetically labeled cells, but also any interface between Other methodologies to enhance labeling efficacy Besides different tissues, which might impede identification of the TAs, other efforts have been taken to improve SPIO hypointense signal of labeled cells, especially in high field internalization, mainly by modifying particle surface strengths. Conventional fast 3D gradient-echo (GE) coating. Already in 1993, Bulte and coworkers used sequences seems superior, in terms of a compromise liposomes for increased dextran-magnetite particle uptake between sensitivity of T2* effects and spatial resolution/ by human peripheral blood mononuclear cells [73]. Iron imaging time. Recently, the group of Fayad proposed an oxide nanoparticle incorporation can be significantly appealing MR sequence, named ‘Gradient echo Acquisi- enhanced by conjugating the SPIO coating with monoclo- tion for Superparamagnetic particles with Positive con- nal antibodies (mAbs), which targeted specific cell-surface trast’ (GRASP), which enables to create positive rather receptors and promoted receptor mediated endocytosis of than negative contrast of SPIO [81]. This method over- SPIOs [74]. Using HIV1 Tat-peptide, a membrane translo- comes the interference of other sources of T2* effects, and cation signal, intracellular iron concentrations in the range the hyperintense signal may increase the sensitivity and of 10–30 pg/cell can be achieved in human hematopoietic specificity of cell tracking. Exciting and refocusing the CD34+ cells, mouse neural progenitor cells, human CD4+ off-resonance water surrounding the labeled cells by lymphocytes and mouse splenocytes [75]. Moreover, the applying spectral selective radiofrequency pulses is species specific nature of mAbs requires preparation of another approach to create positive contrast [82]. specific mAbs for each animal species and the presence of Theoretically, detection of a small number of labeled receptors on targeted cellular membranes. To label cells cells should improve using a small imaging voxel size in a non-specifically, Strable and Wilhelm formulated magne- high-field environment. Detection sensitivity, however, todendrimers/MD-100, a novel SPIO particle coated with might be significantly reduced in vivo, especially when an anionic material showing a high affinity to cellular imaging the beating heart and in the presence of a membranes [67]. Despite robust magnetic labeling of a heterogeneous tissue background (e.g., hemorrhagic myo- variety of cells, these SPIO particles are not yet commercial cardial infarction). At 17.6-T, as few as 100 transplanted nor FDA-approved. cells in the rat brain can be clearly demonstrated using a Recently, Walczak et al. [76, 77] reported another 3D GE sequence, a 98-μm isotropic resolution and an appealing approach. They attempted to create intracellular intracellular iron concentration of approximately 7.1 pg endosmosis containing SPIO particles by using magne- Fe/cell [83]. At 4.7-T, single MPIO-labeled macrophages toelectroporation devices, an approach which has been in a rejected allograft rat heart could be detected using an used since the 1980s to help DNA and chemotherapeutic in-plane resolution of 156 μm (Fig. 1) [84]. Although a drugs penetrating cellular membranes. Depending on the few studies showed that at 1.5-T single-labeled cells in applied pulses, intracellular iron loadings of 1–5 pg has static organs, such as the brain [85], can be visualized in been achieved in MSCs, neural stem cells and leukocytes vivo in small animal models, it seems almost impossible to after less than 15 min, without any notable toxic effects on achieve this goal in the beating heart because of motion labeled cells. This approach is attractive because it avoided artifacts, relatively low signal-to-noise ratios and a too prolonged incubation and cells of interest can be labeled large voxel size. In an acute infarct pig model, Hill et al. with a single FDA-approved MR contrast agent. Further [86] found that a minimum of 1×105 intramyocardially research, however, is warranted to assess the bio-effects of injected MSCs could be detected by a 1.5-T clinical MR these label methods on cells in vitro and in vivo [78]. scanner for at least 12 weeks after administration (Fig. 2). Every cell contained 200 pg iron on average, using 0.9 μm Detection sensitivity of in vivo imaging with MRI The limit sized iron fluorescent particle. The feasibility of visualiz- to visualize magnetically labeled cells with MRI is ing SPIO-labeled cells has also been investigated after determined by the intra-cellular iron loading, MR se- intra-vessel delivery. Hypointense signals could be quence type, spatial resolution, magnetic field strength, detected in the infarct zone after intracoronary infusion and heterogeneity of the tissue background. The higher the but not after intravenous administration of Feridex-labeled intracellular iron content, the fewer cells are needed to cells [87, 88] (Fig. 3). shorten the relaxation times, no matter T1, T2 or T2*. T2* serves as the most sensitive MR imaging parameter for Limitations of direct iron oxide labeling Direct labeling iron oxide labeling, being 3,000- and 60-times more with SPIO particles remains the most commonly used sensitive than T1 and T2, respectively [79]. A variety of strategy for cell trafficking in the heart. Although it is T2* MR sequences have been tested and compared. simple, allowing fast and sensitive imaging, it bears Steady-state free precession (SSFP) sequence was found several limitations and shortcomings. As mentioned
  • 6. 553 Fig. 1a–d In vivo and ex vivo MRI of allograft hearts in rats, 1 day MPIO-particle labeling. MR microscopy at 11.7 Tesla using a after intravenous injection of MPIO particles. a Allograft heart on Bruker AVANCE-DBX MRI instrument with in-plane resolution of postoperational day (POD) 5. b, c Allograft heart on POD 6. Shown 40 μm of an allograft heart with MPIO-particle labeling. (The heart with 156-μm in-plane resolution at 4.7 Tesla by using a Bruker in d is not the same one with that in a, b and c) (From Wu et al. Proc Biospec AVANCE-DBX MRI instrument. d Ex vivo MRI of Natl Acad Sci USA 103:1852–1857 with permission) macrophage accumulation on POD 6 of an allograft heart with above, any signal loss arising from other origin, e.g., signal of endogenous fluorine, with other words no noise, hemorrhagic myocardial infarcts and the heart-lung background signal is created. Thus, 19F-labeled cells are interface (Fig. 4), may interfere with the detection of shown as ‘hot spots’ against a dark background [46, 97]. SPIO-labeled cells [86]. Presence of microvascular ob- Moreover, it bears the potential of quantifying the number of struction (MVO) and intramyocardial hemorrhage in particular, a frequent finding in MI patients despite successful coronary reperfusion [89–92] may interfere with the visualization of iron-oxide-labeled cells [93]. This can be explained by the superparamagnetic properties of hemoglobin breakdown products (i.e., deoxyhemoglobin) creating areas of signal void on T2*-weighted images (Fig. 5). Although several other studies [94–96] did not mention this issue, it may interfere with an accurate tracking of administered stem cells. It should be emphasized that iron oxide does not provide direct information concerning cell survival, proliferation, differ- entiation and quantification, and it is not possible to differentiate two or more groups of cells which are all labeled with iron oxide agents. Since a fixed amount of iron oxide is administered to the cells before administra- tion, every subsequent cell division will halve the intracellular iron content and thus leading to gradual reduction of the cell detection. Therefore, direct cellular labeling with iron oxide is most appealing for short-term cell tracking (i.e., to locate stem cells within the heart). Since the labeling is nonspecific, iron contents could be passed on to surrounding cells (e.g., macrophages) or simply distributed extra-cellular, which may produce false positive signals. False negative results may be due to partial volume effects or when the amount of cells is too low in one imaging voxel. Fig. 2a–d In vivo visualization immediately before and after IFP- 19 MSC injection. Long-axis SSFP MRI view of left ventricle before Fluorine 19 ( F) (a) and after (b) transcatheter injection of 4×106 IFP-labeled MSCs into infarct at apex (arrows) and into adjoining normal myocardium Use of 19F MRI rather than 1H MRI may be highly appealing (arrowhead). Delayed hyperenhancement inversion recovery FGRE for cell trafficking. Owing to a high gyromagnetic ratio of MRI highlights areas of nonviable infarcted myocardium using same view as above, before (c) and after (d) injection of IFP-labeled 40.05 MHz/T and a spin 1/2 nucleus, 19F has 0.83 NMR MSCs. MSCs appear dark against hyperenhanced infarct. (FGRE: sensitivity relative to proton and 100% naturally abundant. fast gradient echo). (From Hill et al. Circulation 108:1009–1014 The advantage of using 19F over 1H MRI is the lack of NMR with permission)
  • 7. 554 Fig. 3a–c Ferumoxide-labeled cells in the interventricular septum hypoenhancement corresponding to the zone of cell accumulation of porcine hearts with AMI (a) before and (b) after systemic under dual contrast (arrow), whereas c shows only thinning of the administration of Gd-DTPA as second contrast, compared with (c) antroseptal wall consistent with untreated infarction with absence of control animal after systemic administration of Gd-DTPA. Notice a hypoenhancement. (From Baklanov et al. Magn Reson Med shows vague darkening corresponding to cell accumulation in the 52:1438–1442 with permission) septum (arrow), b shows much better conspicuity of the zone of labeled cells in vivo based on the measured signal intensity and to image their distribution and migration into lymph [46, 98]. Superposition of the 19F MR images on 1H MR nodes. Superposition of 19F MRI and 1H MRI provided images can be used to anatomically locate the labeled cells, definite localization of the labeled cells [46]. Dr. Wickline’s and in addition, more than one type of cells can be labeled group recently demonstrated the capability of 19F MRI to with different 19F contrast agents and differentiated with MR distinguish cells labeled with two different PFCs by spectral spectroscopy [46, 98]. 19F MRI might provide interesting discrimination at 11.7 T [98]. Currently, no apparent bio- biological information. For instance, 19F MRI using effects were observed on viability and function of dendritic perfluorocarbon (PFC) has been applied to report the pO2 cells and mononuclear cells after PFC incorporation both in in a tumor micro-environment [99]. This may be extended to vitro and in vivo [46, 98]. Recently, it has been found provide information about the reaction and behavior of feasible to image myocardial infiltration of PFC labeled engrafted progenitor cells in infarcts. PFC emulsions, macrophages with 19F MRI at 9.4 T, in an acute MI mice originally developed as a red blood cell substitute, should model. Furthermore, the PFC cell labeling was shown to be well tolerated (a maximal safe dose of 3 ml/kg allow quantification of engrafted cells by using 19F MRS. administrated intravenously in human beings) [100]. These encouraging initial results suggest that 19F MRI and The first results of in-vivo tracing progenitor cells with MRS might be a valuable alternative to iron oxide labeling 19 F MRI and MRS appear very promising, in both high for cell trafficking. magnetic field (11.7 T) and clinical (1.5 T) settings. Ahrens As a nonspecific labeling, 19F agents share similar and coworkers used this technique to label dendritic cells, shortcomings as iron oxide nanoparticles, such as dilution Fig. 4 In vivo relaxometry of injected IFP-labeled MSCs. a Typical The posterobasal epicardial surface also shows typical susceptibility in vivo injection of 105 IFP-labeled MSCs is shown in this still SSFP artifact commonly attributed to adjacent diaphragmatic frame from a cinematic SSFP magnetic resonance image immedi- surfaces. The color map corresponds to T2* values indicated on ately after injection. This cell concentration represents the minimally the scale immediately to the right. c A profile of T2* values along detected dose in vivo and was used for all subsequent relaxometry. b the dotted white line from the middle panel. The center of the T2* parametric map of the same slice shows the lowest T2* in the injection has a T2* value close to the values measured in vitro. region of the injection, corresponding to the location of labeled cells. (From Hill et al. Circulation 108:1009–1014 with permission)
  • 8. 555 the iron oxide labeling strategies may can be applied. TA has shown to achieve a 26-fold increase of intracellular 19F loading [46] and the addition of Gd-DTPA administered intravenously shortens 19F spins’ T1, which may enhance 19 F signal intensity and reduce imaging time. The ever- increasing magnetic field strength of clinical MR scanners should allow a faster and more sensitive cell tracing with 19 F MR imaging and spectroscopy. Transgenic labeling strategies Although direct labeling definitely bears several advantages, such as ease of manipulation and potential to use in the clinical setting, it does not fulfill the demands of providing in-depth insight in cell survival, proliferation and differen- tiation. In this perspective, reporter gene imaging in combination with MRI or magnetic resonance spectroscopy (MRS) seems more appealing for long-term cell labeling. Using a transgene approach (e.g., replication-defective adenovirus), genes are incorporated into the cellular DNA. The products derived by the reporter genes (e.g., enzymes, surface markers, membrane receptors) are expressed only in viable labeled cells and may potentially reflect the functional and differentiation status in a quantitative fashion. In addition, in case the reporter gene is incorporated into chromosomes, even offspring cells can be traced as long as Fig. 5 a In vivo scans of an animal model at 1 week (subacute) and the engineered gene is not silent or blunt. The above 5 weeks (chronic) after infarction. Depicted are the end-diastolic advantages make transgenic labeling strategies highly frames of cine, T2*W, FPP, and DE scans at the same level. The valuable for answering questions concerning the long term infarct area is indicated by white arrows. In the T2*W image, endo- and epicardial rims of signal voids are present at 1 week. The artifact fate of labeled cells, although currently limited to preclinical caused by the heart–air interface is indicated by an asterisk. The research due to the restrictions of sufficient bio-safety profile. rims of signal voids correspond with a zone of hypoperfusion in the Earlier attempts to transfect enzyme (creatine kinase, FPP scan at 1 week, indicating MO. Furthermore, a zone of arginine kinase) genes to host tissues showed that products persistent hypoenhancement can be appreciated in the DE scan within the hyperenhanced infarct area. At 5 weeks, the area of signal of incorporated genes were detectable in vivo by 31P MRS. void in the T2*W image takes up a larger part of the infarct area. Interference of background tissue signal, however, re- Neither microvascular obstruction nor an area of persistent stricted their application, and MRS does not provide spatial hypoenhancement is observed in the FPP or DE scans. Bar indicates distribution of the marker, making enzyme reporter gene 2 cm. b The left panel shows the GE/PD-T2*W/TE20 scan, before MRS less attractive for cell trafficking. In hunting reporter injection with iron oxide-labeled cells. The middle panel shows the same slice after injection with 0.1, 1, or 4×106 iron oxide-labeled genes for MRI, a recent breakthrough is the ferritin gene cells. The right panel shows a similar series of injections in remote, [48, 49]. Ferritin is a metalloprotein consisting of 24 noninfarcted myocardium. Although the cell injections create larger subunits, both light and heavy chains, that can coat up to areas of signal voids in the middle panel, their precise location 4,500 Fe3+ ions. Unlike SPIOs, ferritin is predominately cannot be determined because of the signal voids induced by the presence of hemoglobin degradation products. Bar indicates 0.5 cm. anti-ferromagnetic, resulting in a T2/T2* shortening (From van den Bos et al. Eur Heart J 27:1620–1626 with several orders smaller than that of SPIOs. Over-expression permission) of ferritin heavy chain not only led to redistribution of intracellular iron but also to an elevation of the intracellular of intracellular labeling following cell division and passing iron content [48]. Genove et al. [49] reported after of contrast agent to surrounding cells. In addition to the stereotactical injection of the adenovirus containing the availability of 19F coils, software modifications are needed ferritin reporter gene in the striatum of mice, robust contrast to adapt conventional 1H MRI systems working for 19F on T2- and T2*-weighted images, from 5 days to at least imaging. Moreover, 19F labeling is not as sensitive as iron 39 days after injection. Currently several groups are trying oxide labeling. Despite the above, application and devel- to apply this technique in stem cell research in both small opment of 19F MRI and MRS for cell trafficking have been and large animal models of ischemic cardiomyopathy. carried out in a variety of institutes and manufactures. To Tannous et al. [47] generated a novel reporter gene increase sensitivity of 19F labeled cell detection, some of expressing a cell surface receptor allowing to monitor
  • 9. 556 gene expression and cell targeting. Metabolic biotinylation positive findings should be regarded and interpreted of cell surface receptors can be used for in vivo imaging. carefully, but findings in the placebo-controlled group Linking multimodality contrast agents to avidin, which has should match with literature [20, 30, 111]. MRI is an affinity 103–106 higher than the antigen-antibody definitely one of the preferred techniques to evaluate interaction, provides signal enhancement in fluorescent surrogate end-points in stem cell therapy. Regarding the imaging, MRI, PET and SPECT [47]. Although initially heart, in particular infarct imaging, MRI has evolved designed for monitoring gene expression in gene therapy, toward a preferential technique enabling to provide reporter genes are potentially translational for cell labeling accurate and reproducible information on ventricular and appealing for long-term observation of cell distribu- volumes, global and regional ventricular performance, tion, proliferation and differentiation. myocardial perfusion, infarct size and geometry, infarct and ventricular remodeling, MVO, and area at risk within an acceptable imaging time (Table 2). MRI guided stem cell administration Ultrafast MRI technology has opened the way toward Global and regional ventricular function MRI-guided catheter interventions. Compared with con- ventional X-ray angiography, MR fluoroscopy has several Coronary artery occlusion initiates an immediate loss of advantages such as interactive 3D steering of the imaging contractility in the ischemic myocardium. In infarct patients, plane, excellent soft tissue, and elimination or reduction of global ventricular performance is determined by extent of iodinated contrast agents and ionizing radiation [101, 102]. myocardial necrosis, amount of stunned myocardium (i.e., Additionally, high-resolution anatomic and functional viable but dysfunctional myocardium that will progressively imaging can be performed in the same study. This recover in function over time), and impact on contractility in technique has been used with success for a variety of the remote, noninfarcted myocardium. Postinfarct ventricu- cardiac applications in animal models and patients [101– lar remodeling describes the functional, morphological, 106]. MRI-guided intramyocardial delivery of stem cell geometrical changes over time. It is a complex phenomenon therapy offers (1) direct delivery of therapy to ischemic or determined by different parameters (e.g., infarct size, infarct regions, (2) high local concentrations, (3) reduction location and transmurality, MVO, area at risk, etc.). It is of intracoronary administered dose of cells [107–109]. imperative that these parameters should be taken into Although its intrinsic advantages, currently there is only account before any statements on potential treatment effects limited interest from the manufacturers to start producing can be made. For instance, presence of MVO (so-called ‘no- FDA-approved MR compatible catheters and guide wires reflow zones’) is considered an important determinant of for these kind of procedures. Moreover, catheter heating in postinfarction ventricular remodeling [90, 112, 113], but an MRI environment yields the risk of decreasing the stem most studies so far characterized infarct severity only in cell viability during delivery. For an extensive review on terms on infarct volume, ignoring other additionally MRI in guiding and assessing endomyocardial therapy, we important parameters [24–26, 29, 114] (Table 3). refer to a recent publication by Saeed et al. [109]. As mentioned above, true scarless myocardial healing is still far away. Instead, the magnitude of potential functional improvement should be estimated in the order of a few MRI surrogate measures for stem cell research percentages (Table 4). In order to depict these subtle changes over time, the cardiac imaging technique with the A crucial issue in cardiac stem cell research is assessment highest precision and accuracy of the technique should be of response to myocardial regeneration therapy, thereby selected [115, 116]. Cine MRI, using the steady-state free- obtaining insight into the mechanism of action of myocar- precession (SSFP) cine MRI is, because of its 3D dial regeneration therapies. The most objective way to volumetric approach, and its superior image quality, at achieve this goal are randomized, double-blinded, placebo- present the preferred technique. In a single individual the controlled studies. Surrogate measures [e.g., LV ejection 95% range for change (reflecting the precision) in EF, end- fraction (EF)] rather than hard end-points (e.g., patient diastolic volume and LV mass was 4.8%, 12.4 ml and mortality) are often preferred because smaller study 11.4 g, respectively (Fig. 6), representing a considerable populations can be used, thus reducing total study duration improvement compared with the spoiled gradient-echo and study cost [37, 110]. To be an effective substitute, technique with respective values of 10%, 32.2 ml, and surrogate end-points must reliability predict the overall 20.3 g (117). This enables to use significantly smaller effect on the clinical outcome. Surrogate measures should samples to evaluate ventricular remodeling with the same be evaluated and quantified with the most accurate imaging statistical power [118, 119]. To ensure the best possible techniques available using standardized protocols for data reproducibility, any potential bias during image acquisition acquisition and data analysis [111]. Data material should be and postprocessing should be avoided. Use of similar analyzed with scrutiny, by experts in the field. Not only image sequence parameters, careful positioning of image
  • 10. 557 Table 2 Comprehensive use of cardiac MRI-MRS Myocardial contraction, consisting in wall thickening, circumferential and longitudinal myocardial shortening Cine MRI using SSFP technique [112–123], is reduced and/or abolished in case of myocar- A/ global function dial necrosis. These three deformation parameters are often - end-diastolic volume used to describe the regional functional response to - end-systolic volume regeneration therapy, using the standardized 17 (or 16) - ejection fraction segment or alternatively a three-compartment approach for - stroke volume/cardiac ouput reporting [20, 24, 33, 124]. The latter is definitely - myocardial mass appealing because of its simplicity, describing morpholog- MRI myocardial tagging technique ical and functional changes in three well-defined areas (i.e., infarcted, adjacent and remote myocardium), allowing to - wall deformation in terms of 2D/3D strains and shear strains (e.g., readily depict any stem-cell related effects [24, 90] (Fig. 8). torsion) An alternative approach to assess regional ventricular - fiber/cross-fiber shortening performance, is the calculation of a regional EF parameter, - regional ejection fraction allowing to describe the contribution of the infarcted Velocity-encoded MRI – Phase-contrast MRI myocardium to the ejection of blood. This can be achieved - diastolic function by fusing information on infarct extension (see below) with - myocardial wall velocities functional information. All the above data on regional - associated valvular pathology (e.g., mitral valve regurgitation) function can be extracted from the same cine MRI data used for quantification of global volumes and function. To T2-weighted short-tau-inversion-recovery spin-echo MRI obtain reliable data on wall thickness, wall thickening and - myocardial edema (area at risk) wall motion, it is obvious that this necessitates a scrutinized - intramyocardial bleeding delineation of endocardial and epicardial borders (Fig. 7). First-pass perfusion MRI (stress-rest) Finally, the complex mechanisms of myocardial deforma- - myocardial perfusion reserve (MPR) tion and contraction can be unraveled with MRI myocar- - quantitative myocardial blood flow calculation (dual-bolus dial tagging [122, 125]. This technique has been used with approach) success to explain the mechanism of functional recovery in Contrast-enhanced inversion-recovery (CE-IR) MRI reperfused MIs [126], and because of its sensitivity might be of interest in depicting treatment effects on myocardial - microvascular obstruction (MVO) deformation [126, 127]. - infarct location - infarct volume - infarct transmurality Characteristics determining infarct severity - infarct surface/circumferential – longitudinal infarct length - (endocavitary thrombus formation) Within 15–30 min after onset of coronary artery occlusion, - (associated inflammatory pericarditis) myocardial necrosis occurs at the inner myocardial border 31 P MRS (i.e., subendocardium) and necrosis will spread in a transmural direction (“transmural wave front”) to involve -phosphocreatine/ATP ratio the entire wall within 3–6 h. Early, sustained reperfusion B/ regional function stops transmural spread and salvages the jeopardized - wall thickening myocardium. In the initial hours post infarction, the - wall motion myocardium distal to the culprit lesion, consists in a - regional ejection fraction mixture of irreversibly damaged myocardium, and a C/ regional morphology variable amount of dysfunctional but viable myocardium - wall thickness that will progressively recover in function. Infarct severity is determined by infarct volume and infarct transmurality, - wall curvatures and presence of MVO [112, 113]. Despite successful recanalization of the occlusion with restoration of a normal flow (i.e., TIMI 3 flow), a significant number of patients will have a MVO [90]. It was recently suggested that slices during acquisition, correction for through-plane presence of MVO might impair stem cell engraftment in motion and in/exclusion of papillary muscles and endo- the infarct core, thus altering the therapy response [24, 114, cardial trabeculations during image delineation are all 128]. Other parameters that might influence therapy critical issues that may significantly impact the results outcome are postreperfusion intramyocardial hemorrhage [120, 121] (Fig. 7). In many stem cell papers it remains and the relationship of infarct size to area at risk. The first unclear how these issues were addressed. condition is associated with more extensive LV remodel-
  • 11. 558 Table 3 Use of MRI in randomized clinical cell therapy trials in infarct transmurality, MPI myocardial perfusion imaging, MVO acute and chronic MI patients (AAR area at risk, EDV end-diastolic microvascular obstruction, WM wall motion, WTh (systolic) wall volume, EF ejection fraction, CE-IR MRI contrast-enhanced inver- thickening, Y data reported in the paper, (Y) data available but not sion-recovery MRI, H hemorrhagic component, IS infarct size, IT reported in the paper, NA non-applicable) Study EDV EF Mass WTh WM AAR MPI CE-IR MRI IS IT MVO H BOOST Y Y Y – Y – – Y – (Y) – REPAIR-AMI Y Y Y Y Y – – Y – – – Janssens et al. Y Y Y Y Y Y (Y) Y Y Y (Y) ASTAMI Y Y (Y) – – – – SPECT – – – STEMMI Y Y Y Y – – – Y – – – MAGIC-3 Y Y (Y) – – – – Y – – – REVIVAL-2 Y Y – – – – – SPECT – – – G-CSF-STEMI Y Y Y Y – – Y Y – Y – Assmus et al. (Y) Y – – Y NA – Y – NA NA Hendrickx et al. Y Y – Y – NA – Y – NA NA ing, while the infarct to area-at-risk ratio provides an shrinkage, with a volume loss of 30–60% [137] (Fig. 11). It estimate of the amount of stunned myocardium [129]. To is currently not clear whether myocardial regeneration summarize, several parameters determine infarct severity, therapy has any effect on infarct remodeling, though some postinfarct functional recovery and postinfarct ventricular papers mention a more extensive shrinkage in treated remodeling and need to be taken into account when patients [24]. Valuable information on this issue, e.g., assessing response to generation therapy in MI patients. infarct volume, infarct transmurality, infarct surface, Contrast-enhanced inversion-recovery (CE-IR) gradient- circumferential and longitudinal infarct length, infarct echo technique has become the preferred technique to radii of curvature in longitudinal/circumferential direction, diagnose, quantify, and follow-up, MVO, myocardial can be achieved using the CE-IR MRI in combination with infarction and myocardial scarring [90, 130–133] cine MRI in infarct patients over time. The above (Fig. 9). Myocardial edema (“area-at-risk”) can be parameters moreover allow to estimate local wall stress, estimated adding a T2-weighted short-tau inversion- and enable to get a better idea about strain-stress relation- recovery (STIR) dark-blood spin-echo MRI sequence, ships in different parts of the ventricle. while this sequence is also helpful to depict concomitant intramyocardial hemorrhage [129, 134, 135] (Fig. 10). Experience in performing and analyzing of these studies is Myocardial perfusion essential. Semiautomatic delineation programs may be of help in increasing study reproducibility [136]. To be effective, regeneration of myocytes should be matched with new blood vessel formation in the infarct- bed (vasculogenesis) and proliferation of preexisting Infarct expansion, shrinkage and remodeling vasculature (angiogenesis) improving myocardial perfu- sion and thus assuring adequate oxygen supply. Kocher et In the first days after the ischemic event, the infarcted al. in 2001, and others have shown the capability of bone- myocardium undergoes an expansion due to cellular marrow derived endothelial progenitor cells (or angio- swelling and necrosis, extracellular edema, and infiltration blasts) for revascularization of infarcted myocardium [13], of inflammatory cells. Necrotic cells and extracellular thereby protecting against apoptosis and inducing of matrix (mainly collagen bundles) are progressively re- proliferation/regeneration of endogenous cardiomyocytes moved, and replaced by granulation tissue containing a [138]. In a recently published substudy of the REPAIR- high number of myofibroblasts producing new collagen AMI trial, improved coronary flow reserve was reported in fibers. In the weeks and months post infarction the infarct BMSC treated patients [139]. Attempts to quantify thera- remodels and matures into an anisotropic scar consisting of peutic effects on myocardial perfusion have been less high density of type I collagen bundles. With the onset of successful. Janssens and coworkers’ study used PET with myocardial necrosis, myocardial tissue properties suddenly 11C-acetate as perfusion tracer, but they found no differ- shift from an active contracting tissue toward a passive ences in absolute myocardial blood flow were found visco-elastic material that progressively will decrease in between the treated and placebo group [24]. Engelmann et compliance over time. Scar maturation results in infarct al. [21] reported a temporary improvement in G-CSF
  • 12. 559 Table 4 Impact of cell therapy on LV EF in acute and chronic progenitor cells, LVA left ventricular angiography, LV EF left infarct patients. Only results of randomized, placebo controlled ventricular ejection fraction, NS not significant) studies are shown. (BMSC bone-marrow stem cells, CPC circulating Study Baseline Early FU Late FU LV EF (3–6M) (12–18M) BOOST - control (n=30) 51.3 +0.7 +3.1 - treated (n=30) 50.0 +6.7 +5.9 (P=0.0026) (P=0.27) REPAIR-AMI (LVA)a - control (n=103) 46.7 +3.0 - treated (n=101) 47.5 +5.5 (P=0.01) REPAIR-AMI (EF < 48.9%) - control (n=30) 40.5 -0.8 - treated (n=30) 38.7 +5.0 (P=0.01) Janssens et al. - control (n=30) 46.9 +2.2 +2.5 - treated (n=30) 48.5 +3.4 +2.0 (P=0.36) (P=0.84) ASTAMI - control (n=50) 53.6 +4.3 - treated (n=47) 54.8 +1.2 (P=0.054) REVIVAL-2 - control (n=47) 49.2 +2.0 - treated (n=49) 51.3 +0.5 (P=0.14) STEMMI - control (n=29) 55.7 +8 - treated (n=28) 51.2 +8.5 (P=0.9) G-CSF-STEMI - control (n=21) 44 +5.3 - treated (n=23) 41 +6.2 (P=NS) MAGIC - control (n=25) 53.2 −0.2 - treated (n=25) 52.0 +5.1 (P=0.05) MAGICb - control (n=16) 45.1 +0.2 - treated (n=16) 48.5 0 (P=NS) Assmus et al. (LVA)a, b - control (n=23) 43 −1.2 - CPC group (n=24) 39 −0.4 - BMSC group (n=28) 41 +2.9
  • 13. 560 Table 4 (continued) Study Baseline Early FU Late FU LV EF (3–6M) (12–18M) (P=0.001) Hendrickx et al.b - CABG (n=10) 39.5 3.6 - CABG + BMC (n=10) 42.9 6.1 (P=0.41) a Studies that have used LVA to estimate LV EF b Chronic MI patients treated patients of the myocardial perfusion reserve using affecting the entire ventricle (“ventricular remodeling”). first-pass adenosine stress perfusion MRI. True quantifica- Though definition of primary and secondary study end- tion of absolute myocardial blood flow with MRI is points remains essential to evaluate therapy effects on required in these studies but the step from bench to bedside ventricular remodeling, they may not suffice to clarify the has appeared to be more difficult than initially expected. mechanism of regenerative therapy. For example, a larger Promising results have been reported using dual-bolus loss in infarct volume was demonstrated in stem cell treated first-pass perfusion MRI [140]. patients [24]. Should this finding be interpreted as positive or negative? Is this a sign of true myocardial regeneration? Without additional information on infarct remodeling and Towards use of comprehensive MRI in cardiac infarct transmurality, the changes in infarct volume can not stem-cell studies be correctly interpreted. Extensive wall thinning and shrinkage of infarct surface will both result in a loss in As pointed out in previous paragraphs, MI initiates a infarct size. While shrinkage in infarct surface should be complex cascade of morphological, geometrical and func- regarded as positive, extensive infarct thinning will result tional changes involving not only the infarct area, but in increased wall stress and loading of the adjacent Variable = EF Variable = EF 6 6 5 5 4 4 3 3 2 2 1 0 1 -1 0 -2 -1 -3 -2 -4 -5 -3 -6 -4 58 59 60 61 62 63 64 65 66 67 68 69 57 58 59 60 61 62 63 64 65 66 67 68 69 70 % % Average of Reader 1 and 2 Average of Reading 1 and 2 Subject, Reading Subject 1, Reading 1 Subject 1, Reading 2 Subject, Reader Subject 1, Reader 1 Subject 1, Reader 2 Subject 2, Reading 1 Subject 2, Reading 2 Subject 2, Reader 1 Subject 2, Reader 2 Interobserver Variation Intraobserver Variation Fig. 6 Bland-Altman analyses of inteobserver and intra-observer using strict guidelines for image delineation. Despite scanning and variation of LV ejection fraction (EF) using short-axis SSFP cine analysis in optimized circumstances, LV EF still differs considerably MRI. Two healthy volunteers were studied ten times over a period of between studies, between readers and readings. This variability in 5 days (twice a day, with an interstudy delay of at least 1 h). Studies measurements, expressed in terms of “within subject standard were delineated twice separately by two experienced cardiac MRI deviation” and “95% range for change” should be taken into account readers, observing a 1-week interval between repeated readings when evaluating the effects of regenerative therapy
  • 14. 561 Fig. 7 Impact of anatomical structures (i.e., papillary muscles and excluding the papillary muscles and trabeculations to the myocar- endocardial trabeculations) on calculations of wall thickness, LV dium yield LV end-diastolic volume, EF and LV mass of 134 ml, volumes, regional and global LV function. Epicardial contours 71%, 138 g and 161 ml, 63%, and 118 g, respectively. End-diastolic (striped lines), endocardial contours including papillary muscles and wall thickness and systolic wall thickening of the lateral LV wall endocardial trabeculations to the myocardium (full lines), endocar- measured 9.5 mm and 84%, and 8.3 mm and 57%, respectively. dial contours excluding papillary muscles and endocardial trabecu- Adapted from Clinical Cardiac MRI (2005), by Bogaert J, lations to the myocardium (fine dashed lines). Including or Dymarkowski S, Taylor A. Springer Berlin-Heidelberg, Germany myocardium negatively influencing the ventricular perfor- paracrine effects improving myocardial perfusion (and thus mance. In case of true myocardial regeneration, one would function) in the peri-infarct area. These two examples expect to find thickening of the nonenhanced myocardium, emphasize the need to focus on as much parameters as reflecting viable tissue, at the expense of enhanced possible, a task that can be achieved using the versatility of myocardium, reflecting necrotic or scarred myocardium cardiac MRI. In a recently published study by Zeng et al. (Fig. 11). Or, observation of improved global function can [127], the addition of 31P MRS to the comprehensive MRI be the result of myocardial regeneration but also due to protocol enabled them to demonstrate beneficial effects of Fig. 8a, b Acute reperfused inferoseptal myocardial infarc- tion (MI) (arrowheads) with subendocardial microvascular obstruction (MVO) (black arrow) studied in the first week (1W) and at 4 months (4M) after the acute event. At 4M the necrotic tissue and MVO have been replaced by a thinner fibrotic scar. Using late en- hancement information about presence and extent of myocar- dium necrosis-scar formation, the left ventricle can be divided into three compartments, an in- farct region, a peri-infarct or adjacent region (e.g., using a 30° sector on both sides), and a remote region, allowing to de- monstrate changes over time in regional morphology and func- tion. (Partially adapted from Janssens et al. 2006 Lancet 367:113–121) (C control group, T treated group)
  • 15. 562 Fig. 9 Acute reperfused apical infarction in a 51-year-old man. CE-IR MRI in the first week (1W), 4 months (4M) and one year (1Y) after the event. Upper row: horizontal long-axis image; lower row: vertical long-axis image. Measured infarct size (normalized to LV mass) is 4.2 ml at 1W, 2.6 ml at 4M, and 2 ml at 1Y. LV EF obtained using short-axis SSFP cine MRI (not shown) is 52.8% at 1W, 60.0% at 4M, and 60.1% at 1Y. Note the presence of a small microvascular obstruction in the infarct core at 1W Fig. 10a–c Use of comprehensive T2-short-tau inversion-recovery of edema in the entire inferior LV wall (arrows, above). The edema (STIR) fast spin-echo MRI (upper row) and contrast-enhanced extends in inferior wall of right ventricle (arrowhead, above). CE- inversion-recovery (CE-IR) MRI (lower row) to better characterize IR-MRI shows complete transmural necrosis (white arrows, below) myocardial infarctions. a Acute reperfused anteroseptal myocardial and large MVO zone with 50% transmural extent (black arrows, infarction (MI). Region of increased signal intensity, representing above). c Acute, hemorrhagic MI in inferior LV wall. T2-STIR-MRI myocardial edema and reflecting the area at risk is found in the shows central dark core (due to the breakdown of hemoglobin into entire anteroseptal LV wall (arrows, above). On CE-IR MRI the the paramagnetic deoxyhemoglobin), surrounded by hyperintense transmural spread (50% transmural) can be well appreciated rim (arrows, above). On CE-IR-MRI, visualization of complete (arrows, below). b Extensive, acute reperfused MI with concomitant transmural infarction with extensive zone of MVO microvascular obstruction (MVO). T2-STIR-MRI shows large area
  • 16. 563 Fig. 11 Short-axis contrast-enhanced inversion-recovery (CE-IR) myocardium. In case of true myocardial regeneration, it can be MRI (a) enables to measure the thickness of enhanced (i.e., hypothesized to find an increase in nonenhanced myocardial infarcted/scarred) myocardium as well as nonenhanced (i.e., viable) thickness which was not found in the study by Janssens et al. (b) BMCs transplantation on phosphocreatine/ATP ratio in the completely shifted our look on treatment on cardiac infarct border zone. diseases, in particular those with irreversible myocardial damage, such as acute and chronic MI patients. True curative medicine suddenly became nearby reality. The Conclusion hype and initial enthusiasm, however, have been tempered since the recent publication of several clinical studies The discovery of pluripotent progenitor cells bearing the yielding less optimistic results in MI patients. Treatment capacity to differentiate into mature cardiac cells, has effects appear to be minimal, and questions have arisen
  • 17. 564 how to correctly interpret them. Simple administration of the three germ layers) or multipotent (i.e., they can stem cells to the damaged myocardium is likely not differentiate into several cell types). An important location sufficient to create a repopulation. Though the way towards of adult stem cells is the bone marrow, containing true scarless cardiac healing remains long, it is definitely an hematopoietic, endothelial and mesenchymal lineages. achievable target. In the search for solutions, imaging will Aspiration of bone marrow typically contains a mixed or have an important position. New MRI labeling techniques “contaminated” cell population, with less than 0.1% stem will allow better short- and long-term techniques allowing cells. The cells can be further characterized the presence or to track the fate of administered stem cells. Interventional absence of cell surface factor (or markers), whether or not MR procedures will enable improved delivery of stem cells they express differentiation factors, and the CD surface of the target of interest. Finally, the versatility of cardiac antigen (see also: MRI should be exploited to better understand the Stem cell markers are specialized receptors that have the mechanisms of action. The input of the cardiac MRI expert capability of selectively binding or adhering to other in the design and analysis of future myocardial regenerative “signaling” molecules. These receptors and the molecules studies is definitely another key toward success. that bind to them are not only a way of communicating with other cells and to carry out their proper functions in the body but researchers use these receptors to identify (and Addendum: Stem cells—types and classification separate) stem cells by attaching to the signaling molecule another molecule (or tag) that has the ability to fluoresce or Stem cells or progenitor cells are defined by their location emit energy when activated by an energy source such as an where they are found, the cell surface markers, transcrip- UV light or laser beam. Important surface markers for tion factors and proteins. Embryonic stem cells are derived cardiac progenitor cells are stem cell antigen (Sca-1) from the blastocyst, and are totipotent (i.e., they can indicative of hematopoietic and mesenchymal stem cell; differentiate into the three germ layers and extraembryonic lineage surface antigen (lin), lin− signifies stem cells that tissue such as the placenta). Adult (or somatic) stem cells do not express differentiation factors; C-kit+ are cells are undifferentiated cells found among differentiated cells carrying the receptor for stem cell factor. CD34+ represent in a tissue or organ, can renew itself, and can differentiate the hematopoietic cell line, CD133+ are endothelial to yield the major specialized cell types of the tissue or progenitors, while CD34+ Sca1+Lin− profile represents organ. They are pluripotent (i.e., they can differentiate into mesenchymal stem cells. References 1. Braunwald E (1989) Myocardial reper- 4. Bartunek J, Dimmeler S, Drexler H, 9. Dimmeler S, Zeiher AM, Schneider fusion, limitation of infarct size, re- Fernandez-Aviles F, Galinanes M, MD (2005) Unchain my heart: the duction of left ventricular dysfunction, Janssens S, Martin J, Mathur A, scientific foundations of cardiac repair. and improved survival. Should the Menasche P, Priori S (2006) The J Clin Invest 115:572–583 paradigm be expanded? Circulation consensus of the task force of the 10. Quaini F, Urbanek K, Beltrami AP, 79:441–444 European Society of Cardiology con- Finato N, Beltrami CA, Nadal-Ginard 2. Cohn JN, Ferrari R, Sharpe N (2000) cerning the clinical investigation of the B, Kajstura J, Leri A, Anversa P (2002) Cardiac remodeling—concepts and use of autologous adult stem cells for Chimerism of the transplanted heart. N clinical implications: a consensus paper repair of the heart. Eur Heart J Engl J Med 346:5–15 from an international forum on cardiac 27:1338–1340 11. Minami E, Laflamme MA, Saffitz JE, remodeling. Behalf of an International 5. Murry CE, Reinecke H, Pabon LM Murry CE (2005) Extracardiac progen- Forum on Cardiac Remodeling. J Am (2006) Regeneration gaps: observations itor cells repopulate most major cell Coll Cardiol 35:569–582 on stem cells and cardiac repair. J Am types in the transplanted human heart. 3. Steg PG, Lopez-Sendon J, Lopez de Sa Coll Cardiol 47:1777–1785 Circulation 112:2951–2958 E, Goodman SG, Gore JM, Anderson 6. Laflamme MA, Murry CE (2005) 12. Orlic D, Kajstura J, Chimenti S, FA Jr, Himbert D, Allegrone J, Van de Regenerating the heart. Nat Biotechnol Jakoniuk I, Anderson SM, Li B, Pickel Werf F (2007) External validity of 23:845–856 J, McKay R, Nadal-Ginard B, Bodine clinical trials in acute myocardial in- 7. Leri A, Kajstura J, Anversa P (2005) DM, Leri A, Anversa P (2001) Bone farction. Arch Intern Med 167:68–73 Cardiac stem cells and mechanisms of marrow cells regenerate infarcted myocardial regeneration. Physiol Rev myocardium. Nature 410:701–705 85:1373–1416 8. Marthur A, Martin JF (2004) Stem cells and repair of the heart. Lancet 364:183–192
  • 18. 565 13. Kocher AA, Schuster MD, Szabolcs 19. Britten MB, Abolmaali ND, Assmus B, 25. Kang H-J, Lee H-Y, Na S-H, Chang MJ, Takuma S, Burkhoff D, Wang J, Lehmann R, Honold J, Schmitt J, Vogl S-A, Park K-W, Kim H-K, Kim S-Y, Homma S, Edwards NM, Itescu S TJ, Martin H, Schachinger V, Dimmeler Chang H-J, Lee W, Kang WJ, Koo (2001) Neovascularization of ischemic S, Zeiher AM (2003) Infarct remodel- B-K, Kim Y-J, Lee DS, Sohn D-W, Han myocardium by human bone-marrow- ing after intracoronary progenitor cell K-S, Oh B-H, Park Y- B, Kim H-S derived angioblasts prevents cardio- treatment in patients with acute myo- (2006) Differential effect of intracor- myocyte apoptosis, reduces remodeling cardial infarction (TOPCARE-AMI): onary infusion of mobilized peripheral and improves cardiac function. Nat mechanistic insights from serial con- blood stem cells by granulocyte colo- Med 7:430–436 trast-enhanced magnetic resonance im- ny-stimulating factor on left ventricular 14. Assmus B, Schachinger V, Teupe C, aging. Circulation 108:2212–2218 function and remodeling in patients Britten M, Lehmann R, Dobert N, 20. Wollert KC, Meyer GP, Lotz J, Ringes- with acute myocardial infarction versus Grunwald F, Aicher A, Urbich C, Lichtenberg S, Lippolt P, Breidenbach old myocardial infarction. The MAGIC Martin H, Dimmeler S, Zeiher AM C, Fichtner S, Korte T, Hornig B, Cell-3-DES randomized, controlled (2002) Transplantation of Progenitor Messinger D, Arseniev L, Hertenstein trial. Circulation 114:I-145–I-151 Cells and Regeneration Enhancement B, Ganser A, Drexler H (2004) In- 26. Schächinger V, Erbs S, Elsässer A, in Acute Myocardial Infarction (TOP- tracoronary autologous bone-marrow Haberbosch W, Hambrecht R, CARE-AMI). Circulation 106:3009– cell transfer after myocardial infarction: Hölschermann H, Yu J, Corti R, 3017 the BOOST randomised controlled Mathey DG, Hamm CW, Süselbeck T, 15. Strauer BE, Brehm M, Zeus T, Bartsch clinical trial. Lancet 364:141–148 Assmus B, Tonn T, Dimmeler S, T, Schannwell C, Antke C, Sorg RV, 21. Engelmann MG, Theiss HD, Hennig- Zeiher, for the REPAIR-AMI investi- Kogler G, Wernet P, Muller HW, Theiss C, Huber A, Wintersperger BJ, gators (2006) Intracoronary bone mar- Köstering M (2005) Regeneration of Werle-Ruedinger AE, Schoenberg SO, row-derived progenitor cells in acute human infarcted heart muscle by in- Steinbeck G, Franz WM (2006) myocardial infarction. N Engl J Med tracoronary autologous bone marrow Autologous bone marrow stem cell 355:1210–1221 cell transplantation in chronic coronary mobilization induced by granulocyte 27. Schachinger V, Assmus B, Britten MB, artery disease: the IACT Study. J Am colony-stimulating factor after subacute Honold J, Lehmann R, Teupe C, Coll Cardiol 46:1651–1658 ST-segment elevation myocardial infarc- Abolmaali ND, Vogl TJ, Hofmann WK, 16. Strauer BE, Brehm M, Zeus T, tion undergoing late revascularization: Martin H, Dimmeler S, Zeiher AM Kostering M, Hernandez A, Sorg RV, final results from the G-CSF-STEMI (2004) Transplantation of progenitor Kogler G, Wernet P (2002) Repair of (Granulocyte Colony-Stimulating Factor cells and regeneration enhancement in infarcted myocardium by autologous ST-Segment Elevation Myocardial In- acute myocardial infarction: final one- intracoronary mononuclear bone mar- farction) trial. J Am Coll Cardiol year results of the TOPCARE-AMI row cell transplantation in humans. 48:1712–1721 Trial. J Am Coll Cardiol 44:1690–1699 Circulation 106:1913–1918 22. Zohnhöfer D, Ott I, Mehilli J, Schömig 28. Schachinger V, Erbs S, Elsasser A, 17. Assmus B, Walter DH, Lehmann R, K, Michalk F, Ibrahim T, Meisetschläger Haberbosch W, Hambrecht R, Honold J, Martin H, Dimmeler S, G, von Wedel J, Bollwein H, Seyfarth M, Holschermann H, Yu J, Corti R, Zeiher AM, Schachinger V (2006) Dirschinger J, Schmitt C, Schwaiger M, Mathey DG, Hamm CW, Dimmeler S, Intracoronary infusion of progenitor Kastrati A, Schömig A, for the REVI- Zeiher AM (2006) Improved clinical cells is not associated with aggravated VAL-2 investigators (2006) Stem cell outcome after intracoronary adminis- restenosis development or athero- mobilization by granulocyte colony-sti- tration of bone-marrow-derived sclerotic disease progression in patients mulating factor in patients with acute progenitor cells in acute myocardial with acute myocardial infarction. Eur myocardial infarction: a randomized infarction: final 1-year results of the Heart J 27:2989–2995 controlled trial. JAMA 295:1003–1010 REPAIR-AMI trial. Eur Heart J 18. Bartunek J, Vanderheyden M, 23. Lunde K, Solheim S, Aakhus S, 27:2775–2783 Vandekerckhove B, Mansour S, De Arnesen H, Abdelnoor M, Egeland T, 29. Ripa RS, Jorgensen E, Wang Y, Thune Bruyne B, De Bondt P, Van Haute I, Endresen K, Ilebekk A, Mangschau A, JJ, Nilsson JC, Sondergaard L, Johnsen Lootens N, Heyndrickx G, Wijns W Fjeld JG, Smith HJ, Taraldsrud E, HE, Kober L, Grande P, Kastrup J (2005) Intracoronary injection of Grogaard HK, Bjornerheim R, Brekke (2006) Stem cell mobilization induced CD133-positive enriched bone marrow M, Müller C, Hopp E, Ragnarsson A, by subcutaneous granulocyte-colony progenitor cells promotes cardiac re- Brinchmann JE, Forfang K (2006) stimulating factor to improve cardiac covery after recent myocardial infarc- Intracoronary injection of mononuclear regeneration after acute ST- elevation tion: feasibility and safety. Circulation bone marrow cells in acute myocardial myocardial infarction. Result of the 112:I178–I183 infarction. N Engl J Med 355:1199– double-blind, randomized, placebo- 1209 controlled stem cells in myocardial 24. Janssens S, Dubois C, Bogaert J, infarction (STEMMI) trial. Circulation Theunissen K, Deroose C, Desmet W, 113:1983–1992 Kalantzi M, Herbots L, Sinnaeve P, Dens J, Maertens J, Rademakers F, Dymarkowski S, Gheysens O, Van Cleemput J, Bormans G, Nuyts J, Belmans A, Mortelmans L, Boogaerts M, Van de Werf F (2006) Autologous bone marrow-derived stem-cell transfer in patients with ST-segment elevation myocardial infarction: double-blind, randomised controlled trial. Lancet 367:113–121

Related Documents