In the beginning there is the stem cell; it is the origin of an organism’s life. It is a single cell that can give rise to progeny that differentiate into any of the specialized cells of embryonic or adult tissues. —Stewart Sell
Cardiovascular disease is the leading cause of mortality and disability, responsible for one third of deaths worldwide. Although there have been significant reductions in morbidity and mortality associated with coronary artery disease, many patients with advanced disease have few therapeutic options. Recent studies suggest that the adult heart is capable of limited regeneration after injury, and improvement in cardiac function after a myocardial infarction has been demonstrated after transplantation of various bone marrow–derived stem cell populations that are capable of generating cardiomyocytes.
For nearly half a decade, the delivery of allogeneic or autologous cellular populations from bone marrow (i.e., bone marrow transplantation) has successfully treated chronic diseases such as solid tumors, hematologic malignancies, and storage diseases. However, challenges arising from the complexities of the adult heart (i.e., its three-dimensional geometry and highly structured cellular architecture and its ability to adapt and remodel in response to injury and changing hemodynamics and loads), and the heterogeneity of study designs, have yielded variable results in preclinical trials of stem cell therapy for cardiac disease. Nevertheless, studies that reported encouraging results motivated the design of clinical trials of cell therapy for ischemic cardiomyopathy. A meta-analysis of these relatively small, short-term clinical trials confirmed the feasibility and safety of stem cell transplantation, although the effects on left ventricular (LV) remodeling and LV ejection fraction (LVEF) were modest at best. However, the long-term (3-year) follow-up of the randomized controlled Autologous Stem Cell Transplantation in Acute Myocardial Infarction study, which followed 100 patients with ST-segment elevation myocardial infarction who received intracoronary autologous bone marrow cells, failed to show any significant effects on LV global systolic function, remodeling, regional systolic function (using deformation analysis), or diastolic function compared with the control group. Thus, before cell-based therapy of cardiovascular disease becomes a clinical reality, several gaps in our understanding of the process need to be confronted.
The critical unanswered issues in stem cell transplantation include (1) defining the patient population likely to benefit most from cell therapy (e.g., ischemic vs nonischemic cardiomyopathy), (2) identifying the optimal cell population (e.g., bone marrow mononuclear cells, mesenchymal stem cells, umbilical stem cells, cardiospheres), and (3) understanding how to prepare (e.g., fresh vs cultured cells, expanded, reprogrammed), dose (e.g., number of cells delivered, number of treatments), and deliver (e.g., infarct vs remote zone using intracoronary, intramyocardial, or intravascular route) cells most effectively. Arguably, a deeper understanding of the processes responsible for repair after stem cell therapy is paramount. In this regard, either differentiation of transplanted stem cells into functioning cardiomyocytes or the elaboration of paracrine factors (cytokines and growth factors) may be operative. A number of studies, including the study by Bonios et al. in this month’s JASE , support the concept that release of paracrine factors by transplanted cells are responsible for the beneficial effects of cell therapy.
Bonios et al. examined the effect of cardiosphere-derived stem cells (CDCs) injected into rat myocardium on LVEF, circumferential strain, and dyssynchrony 1 and 4 weeks after an experimental anterior myocardial infarction. CDCs (a heterogeneous cell population with a high proliferative capacity capable of forming contractile cardiomyocytes cultured from a myocardial biopsy) from male rats were transplanted into infarcted syngenic females, permitting quantification (by subsequent polymerase chain reaction of a region of Y chromosome) of the engraftment. CDC localization (infarct and border zone) and perfusion (mean infarct size, 15.4 ± 3.6% of the left ventricle) were determined using 18 F-fluorodeoxyglucose and 13 NH 3 positron emission tomography, respectively. Despite trivial engraftment (2.4 ± 3.3%), LVEFs improved 4 weeks after infarction in animals treated with CDCs (by about 7%) and decreased (by about 8%) in the controls. This was associated with improved mean apical circumferential strain (an averaged strain of all segments in the short-axis view that included the infarct) and regional strain (lowest strain identified in the apical short-axis view) of the infarcted region and dyssynchrony (the highest measured delay of time to peak strain between two opposing segments) in CDC-treated animals and corresponding decreases of these indices in the control animals. Interestingly, there were no differences at 1 week, suggesting that the beneficial effects were not due to a direct paracrine effect of the CDCs on viable infarct and border zone myocyte contractility, because the maximal secretion of growth factors by CDCs occurs during the first week. In addition, there were no changes in the remote (basal short-axis) zone at 1 and 4 weeks, indicating that improvement in function was due to salutary changes in strain and dyssynchrony in the infarct region alone.
This study is notable insofar as it exploits a clever study design and multimodality imaging to quantify engraftment and cellular localization, which allows for an elegant exploration of the mechanisms responsible for an increase in global LV function after cellular transplantation. Although strain and dyssynchrony (both previously shown to improve with umbilical and bone marrow stem cells ) improve via paracrine mechanisms (the minimal engraftment of CDCs is unlikely to be directly responsible), it remains uncertain whether the improvement is due to paracrine-induced regeneration (by recruiting endogenous cardiac stem cells), enhancement of angiogenesis, or limitation of apoptosis or the fibroproliferative response to injury. The study also serves to highlight the important role of imaging in general and echocardiography in particular, in addressing the unresolved issues perplexing stem cell therapy.
Imaging may be used to profile and assess the severity and course of the disease phenotype, to guide the delivery of stem cells, to assess the efficacy of cellular therapy on cardiac function, to evaluate the fate and function of engrafted cells, and to help define (as in the current study) the mechanisms responsible for the beneficial effects of cellular transplantation. Echocardiography has a potentially important role and is likely to have a significant impact in each of these areas in the future.
Disease Profile and Assessment of Severity, Course, and Efficacy
Ultrasound evaluation of LV volumes and LVEF (typically using the biplane method of disks in clinical studies and M-mode echocardiographic dimensions in preclinical studies) is the method most often used to profile the nature of the disease and its course and is the standard surrogate outcome measure in stem cell trials. However, in view of the variability, limited reproducibility, and load sensitivity of the ejection fraction, attention has been drawn to other techniques. For example, real-time three-dimensional echocardiographic imaging enables the quantification of global systolic function with accuracy and reproducibility similar to magnetic resonance imaging and computed tomography, and objective determinations of regional ejection fractions and synchronous contraction are readily obtained. Regional function may also be assessed quantitatively and objectively using a variety of strain rate imaging algorithms and platforms (e.g., tissue Doppler, two-dimensional speckle tracking, and vector velocity imaging).
Doppler and two-dimensional echocardiographically derived myocardial strain imaging represent exciting advances in the field of noninvasive cardiac imaging. Strain and strain rate are sensitive (i.e., they detect early improvement), provide objective regional measurements, correlate well with other measures of cardiac function, and detect changes in myocardial contractility. They provide novel insights into the complex deformations of the left ventricle and appear likely to enhance the detection of viable myocardium and predict the sequelae of myocardial infarction, the response to therapy, and prognosis. Although the accuracy, precision, and reproducibility of measurements of myocardial deformation and their variability across different platforms and vendors have raised concerns, further study, technological refinements, automation, and establishment of normative values will add greatly to echocardiographic strain imaging the needed objectivity for the evaluation of cell-based therapies.
Bonios et al. confirm that LV dyssynchrony may help explain mechanistically the effects of CDCs on global LV function. Chang et al. studied 40 patients with acute myocardial infarction who were randomized to the infusion of either stem cells or control. Tissue synchronization imaging was used to determine dyssynchrony, measured as the standard deviation of the difference in time to peak systolic velocity using a 12-segment model. Dyssynchrony was similar at baseline, but at 6-month follow-up, the standard deviation of the difference in time to peak systolic velocity improved in the cell therapy group but not in controls. The improvement correlated with an improvement of magnetic resonance imaging–determined ejection fraction and symptom-limited treadmill testing. Thus, stem cell therapy improves LV dyssynchrony in rodents and humans and helps explain an improvement in global systolic function but is another surrogate that needs to be validated. The improvement in synchrony, by means of less energy wastage, and salutary changes in regional blood flow and regional metabolism may be responsible. However, precisely how transplantation causes LV resynchronization is unknown.
The ability to recognize the spatial distribution of myocardial perfusion and quantify blood volume noninvasively without ionizing radiation is a major advantage of myocardial contrast echocardiography (MCE). MCE may be useful in selecting patients suitable for cell therapy and in following the effects of cell therapy on myocardial perfusion. Unfortunately, the overinflated concerns of the safety of ultrasound contrast agents have discouraged clinical trials with MCE in the United States. Low-dose dobutamine determination of viability may also be useful in patient and myocardial site selection for cell therapy and to assess its effects. Karatasakis et al. infused bone marrow stem cells in the left anterior descending coronary artery in 10 patients with remote anterior myocardial infarction and nonviable myocardium documented by 201 Tl scintigraphy and conventional low-dose dobutamine. Peak systolic strain rate and to a lesser extent strain (but not postsystolic strain) improved in the scarred anteroapical region s6 months after infusion. Low-dose dobutamine increased longitudinal deformation, but there was no change in the global ejection fraction (which may not be suitable to detect small changes in function). Although intriguing, the clinical significance of this finding is unclear, as there are no validation studies of these surrogate end points. In any event, taken together, these studies demonstrate that advanced echocardiographic techniques allow a comprehensive, noninvasive measurement of myocardial performance and contractility with acceptable interpretative variability that can be used to profile and assess the severity and course of ischemic cardiomyopathy and determine the efficacy of stem cell therapy.
Microbubbles have been used clinically for improvement of diagnostic ultrasound images for several years, but their therapeutic use has only recently gained interest. Ultrasound-mediated microbubble destruction (USMBD) in the capillary lumen has been shown to elicit arteriogenesis and hyperemic blood flow in ischemic rat skeletal muscle by creating pores (sonoporation) in the capillary walls, which is dependent on ultrasound power. This strategy has since been used for targeted cell and gene delivery to injured myocardium and has been shown to increase attachment of the delivered therapy to the endothelium after exposure to USMBD. Targeted delivery of bone marrow–derived mononuclear cells combined (not complexed) with microbubbles and ultrasound increased LV fractional shortening, capillary density, and regional blood flow in cardiomyopathic hamsters. Physical binding of microbubbles to stem cells as a platform for targeted delivery has not been evaluated, although studies have shown the capability of biotinylated microbubbles to attach to mature vascular endothelial cells.
Stem cell transplants themselves may be targeted with microbubbles linked to small molecules, genetic material, and other biological agents to enhance the function of the transplant. Recently, Fujii et al. delivered plasmids containing vascular endothelial growth factor and stem cell factor genes, empty (control) plasmids, and green fluorescent dye (used to visualize gene delivery) using USMBD (8 MHz, mechanical index 1.6, for 20 min in intermittent mode) in mice 1 week after myocardial infarction. Vascular endothelial growth factor and stem cell factor (and green fluorescent dye) protein expression, capillary and arteriolar density by immunohistochemistry, qualitatively determined MCE perfusion, and LVEF were all improved. Advantages of this technique include low toxicity, limited immunogenicity, and the ability to noninvasively and repetitively target myocardium.
A therapeutic high-intensity focused ultrasound–based imaging probe system was recently designed and shown to accurately target a region of interest and deliver ultrasound energy; this wideband annular array coupled with a motorized positioning system successfully and precisely delivered fluorescently labeled plasmid deoxyribonucleic acid into murine myocardium. Although these studies are encouraging, there is a need to define USMBD requirements for the agent, the delivery vehicle, the therapeutic ultrasound probes for drug delivery and for subsequent translation to humans.
Ghanem et al. validated percutaneous, echocardiography-guided closed chest cell delivery of mononuclear and mesenchymal stem cells and assessed contractile reserve with reconstructive three-dimensional echocardiography 3 weeks after transplantation in rats with reperfused myocardial infarction. Closed chest injections were as effective as the open chest approach. The technique facilitated injections into the functionally impaired peri-infarct zone; cell transplants were identified as ultrasound opaque deposits.
Tracking the Fate of Cells
Imaging transplanted cells in the myocardium and confirming their survival and integration at the site of transplantation is a critical step in evaluating the efficacy and mechanistic hypotheses underpinning cellular transplantation. Reporter genes and fluorescent dyes are usually used in preclinical trials and are essential to understanding the mechanisms of engraftment, but these approaches (in vivo positron emission tomography used by Bonios et al. notwithstanding) are invasive and not suitable for human studies. Bara et al. established the feasibility of in vivo detection of donor cells that were labeled with ferromagnetic nanoparticles. In their study, hyperdense areas on transesophageal echocardiography identified the myocardial distribution of the graft at the time of injection and spreading of the transplant 8 weeks later in a pig model of myocardial infarction; that visualization is qualitative and that the signal from the nanoparticles does not equate with cell survival and integration are important limitations of this technique.