Abstract
Satellite cells are committed precursor cells residing in the skeletal muscle. These cells provide an almost unlimited regeneration potential to the muscle, contrary to the heart, which, although proved to contain cardiac stem cells, possesses a very limited ability for self-renewal. The idea that myoblasts (satellite cell progenies) may repopulate postinfarction scar occurred around the mid-1990s. Encouraging results of preclinical studies triggered extensive research, which led to the onset of clinical trials. These trials have shown that autologous skeletal myoblast transplantation to cure heart failure is feasible and relatively safe (observed incidences of arrhythmia). Because most of the initial studies on myoblast application into postischemic heart have been carried out as an adjunct to routine surgical procedures, the true clinical outcome of such therapy in regard to cell implantation is blurred and requires to be elucidated. The mechanism by which implantation of skeletal myoblast may improve heart function is not clear, especially in the light of inability of these cells to couple electromechanically with a host myocardium. Successful myoblast therapy depends on a number of factors, including: delivery to the target tissue, long-term survival, efficacious engraftment, differentiation into cardiomyocytes, and integration into the new, unique microenvironment. All these steps constitute a potential goal for cell manipulation aiming to improve the overall outcome of such therapy. Precise understanding of the mechanism by which cells improve cardiac function is essential in giving the sensible direction of further research.
Introduction
Satellite cells were first described in 1961 by Mauro as the committed precursor cells residing in the skeletal muscle (29). The basic role of these cells is restoration of skeletal muscle function in case of injury. Satellite cells serve as a reserve of quiescent mononucleated muscle precursor cells. After activation they may proliferate, differentiate, and fuse into the new muscle fibers (16).
Myoblasts are progeny of satellite cells obtained upon biopsy of skeletal muscle followed by in vitro culture. In the beginning, the most prominent use of myoblasts was associated with the treatment for muscular dystrophy, first as a source of correct acting genes (allogenic transplantation) and later as vehicles for ex vivo therapy (autologous cells) (9). The idea that myoblasts may substitute for cardiomyocytes lost upon ischemic insult occurred around the mid-1990s (12). Skeletal myoblast transplantation for cardiac repair was first evaluated by Marelli et al. in a dog heart model of cryoinjury (27). The encouraging data obtained in the study triggered extensive research resulting in the onset of preclinical and subsequently clinical trials. Results of these clinical trials have shown safety, efficacy, and feasibility of skeletal myoblast transplantation into damaged myocardium (68). Although one might argue that myoblast implantation carries the risk of arrhythmia, that these cells are not committed precursor cells of myocardium as cardiac progenitor cells (36) and fail to couple with neighboring cardiomyocytes (68), still skeletal myoblasts, due to their unique properties (contractility and stress resistance), are considered some of the most suitable cells for myocardial repair.
Treatment for Ischemic Cardiomyopathy
The incidence of heart failure (HF), with myocardial infarction being the primary cause, has reached epidemic proportions. Contemporary medical therapies include new drugs, organ transplantation, and cardiac devices that reshape the dilated left ventricle. Although patients' survival rate is dramatically improving, these therapies are not satisfactory; heart transplantation is limited to a small number of patients (32). In this context, transplantation of myoblasts has emerged as a potential new approach of reducing cardiac failure incidence. The limited regenerative ability of an adult myocardium is related to the fact that mature cardiac cells are postmitotic and incapable of replication, thus unable to repair tissue after myocardial infarction (16). Although, in the past several years, some groups demonstrated the existence of cardiac stem cells within adult myocardium (60), it seems that their regenerative capacity is insufficient to replace dead cells after ischemic insult. Cardiomyocyte death caused by reduced blood flow triggers a cascade of events, including inflammation, formation of noncontracting fibrous scar, alteration to the workload of the surrounding viable myocardium (compensation and remodeling), and, if the damaged area is large enough, this may lead to congestive heart failure (CHF) (60). The overall objective of cell transplantation is to repopulate postinfarction scar with contractile cells, thus improving systolic function and preventing or regressing the remodeling process. Approximately 15 years of preclinical data have shown that autologous skeletal myoblasts are not only capable of fusion and differentiation into striated muscle cells within damaged myocardium (15) but also of augmenting systolic and diastolic performance in animal models of acute myocardial infarction and heart failure (20, 22, 40, 60). These encouraging results led to the initiation of clinical trials.
Clinical Trials
Due to the myogenic potential and myoblast contractility, several clinical trials involving these cells have been conducted to estimate myoblast efficacy to treat heart failure in its late stages, and, to much lesser extent, to treat acute myocardial infarction (AMI) (9). The first autologous skeletal myoblast transplantation was performed by Menasche et al. in June 2000 and has set the stage for phase I clinical trial (35). The cell therapy, intramyocardial injection (IM) of autologous myoblasts into nonviable scar, was combined with coronary artery bypass grafting (CABG), and included patients diagnosed with severe left ventricular dysfunction with no controls. After 10-month follow-up there was an improvement in NYHA class, an increase in left ventricular ejection fraction (LVEF), and improvement in systolic shortening in cell-implanted scars (35). As a continuation, the MAGIC (Myoblast Autologous Graft in Ischemic Cardiomyopathy) phase II clinical trial was under-taken as a multicentric, randomized, placebo-controlled, dose-ranging design. The inclusion criteria for patients eligible for the trial were LVEF ≥ 15% and ≤35% and akinesia in at least three wall segments. Two doses of myoblasts were injected intramyocardially: high dose (HD), low dose (LD), and placebo suspension medium as a control. All cases were concomitant with CABG and all patients received internal cardioverter/defibrillator (ICD) prior to discharge from the hospital. After a 6-month follow-up there was an improvement in LVEF in the HD group and recovery of movement in at least one wall segment (33).
The trial performed by Pagani et al. involved patients suffering from ischemic cardiomyopathy and refractory HF. Autologous myoblasts were injected intramyocardially during left ventricular assist device implantation (LVAD) procedure. The survival of implanted cells and their development into myotubes occurred in parallel to the host's myocardial fibers and was accompanied by an increased blood vessel count (37). Zhang et al. reported phase I study in patients with history of coronary heart disease. Autologous satellite cell transplantation adjunct to CABG resulted in an uneventful recovery. Four-month follow-up examination showed increase in LVEF, decreased left ventricular diastolic diameter, as well as improved ventricular wall thickness and perfusion at the satellite cell implantation site (70). Smits et al. performed autologous myoblast transplantations on patients with NYHA class ≥ II, with previous anterior wall myocardial infarction and depressed LV function. Percutaneous IM injection was guided by electromechanical mapping via NOGA-guided catheter system. The improved LVEF was detectable at 3-month follow-up only via angiography but after 6 months both angiographic and nuclear LVEF assessments showed a trend toward increased left ventricular ejection fraction. The MRI analysis also revealed regional increase in wall thickening at the target areas and less wall thickening in remote areas, which hypothetically can be a consequence of LV remodeling (53).
Herreros et al. conducted phase I clinical study on patients with old myocardial infarction and ischemic coronary artery disease. They also underwent bypass surgery (CABG) and IM injection of autologous myoblasts cultured with autologous serum (19). Also in the trial performed by Chaques et al. autologous myoblasts were cultivated using patients' own serum in order to prevent malignant arrhythmias and sudden deaths. This study gathered patients diagnosed with impaired left ventricular (LV) function, LV wall postischemic scars, and surgical indication for CABG (8). The results from both trials after 6- and 14-month follow-up, respectively, are comparable and presented in Table 1.
Clinical Trials Involving Myoblast Transplantation
CABG, coronary artery bypass grafting; ICD, internal cardioverter/defibrillator; LVAD, left ventricular assist device; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; HD, high dose; LD, low dose; LV, left ventricle.
Next Ince et al. evaluated transcatheter transplantation of skeletal myoblasts as a stand-alone procedure in patients with ischemic heart failure and compared them to control patients. The procedure involved multiple injections via MyoCath™ delivery catheter system. The 6-month follow-up revealed the improvement in both LVEF and NYHA functional class whereas in matched controls LVEF decreased and NYHA class remained unchanged (21). Siminiak et al. conducted two distinct phase I clinical trials, one involving patients subjected to IM cell transplantation during CABG (51) and the second involving patients who underwent percutaneous trans-coronary-venous cell transplantation. The percutaneous procedure was performed using TransAccess™ catheter system under fluoroscopic and intravascular ultrasound guidance (50). A global LVEF increase was observed in all cases in regard to cell transplantation adjunct to CABG, but only in six out of nine when cell implantation was a sole therapy. In the latter case, after 6-month follow-up, there was an improvement in NYHA class in all patients (50).
Dib et al. conducted nonrandomized, multicenter pilot study of autologous skeletal myoblast transplantation concurrent with CABG or LVAD implantation. Twenty-four-month follow-up showed overall improvement in LVEF, new areas of viability within the infarction scar and reduction in left ventricular systolic and diastolic volumes in CABG patients. Also improvement in NYHA class was evaluated. Histological analysis carried out in four of six patients who underwent heart transplantation further documented survival and engraftment of implanted myoblasts (11).
Over the past 2 years an additional four trials involving catheter-based cell delivery were conducted. Biagini et al. performed autologous skleletal myoblasts transplantation on patients diagnosed with dilated ischemic cardiomyopathy as assessed by two-dimensional echocardiography and tissue Doppler imaging (TDI) during dobutamine infusion. Percutaneous cell transplantation was made using a 7F NOGASTAR catheter for a mapping and Myostar catheter for an injection. Sequential follow-ups performed at 1, 3, and 6 months and especially 1 year revealed decreased NYHA functional class. There were no significant changes in wall motion score index (WMSI), ejection fraction, and left ventricular volumes at rest, while contractile reserve improved during follow-up. Whereas at low-dose dobutamine infusion such parameters as LVEF and end-systolic volumes improved, the peak systolic velocity in the regions of myoblasts injection significantly increased at 1 year (6). Next Sherman et al. conducted the Myogenesis Heart Efficiency and Regeneration Trial (MYOHEART) as an open label, dose escalation study of intramyocardial autologous skeletal myoblast (ASM) administration. Patients enrolled in this phase I study had CHF from post-MI systolic left ventricular dysfunction (SLVD) (49), and were divided into four dose groups. They underwent ASM administration using an endoventricular needle-based catheter delivery system (MyoCath™). Positive effects were defined as a increase in 6-min walk, quality of life score, and LVEF. Obtained data, although preliminary, couple with feasibility and safety assessments. They have led to formulation of phase II–III study, similar to MYOHEART trial, called MARVEL-Multicenter Study of the Safety and Cardiovascular Effects of Myoblasts in Congestive Heart Failure, which was anticipated to begin late in 2007 (48). Dib et al. conducted the Catheter-based delivery of Autologous Skeletal Myoblasts for Ischemic Cardiomyopathy: Feasibility, safety and improvement in cardiac performance (CAuSMIC) trial. Patients suffered from previous MI and congestive heart failure were divided into four dose groups plus control and subjected to ASM transplantation using Noga® System and MyoStar Catheter. At 6-month follow-up the myoblasts improved heart failure, as measured by average NYHA class. Also, quality of life scores and left ventricular systolic and diastolic parameters improved while the control group had no change (10). In the SEISMIC study performed by Serruys et al. patients with congestive heart failure were subjected to injections of muscle cells into scarred areas of the heart using a needle-tipped catheter. Unfortunately, 6-month follow-up did not reveal improvement in LVEF and the left ventricle did not shrink as hoped. However, cell therapy was associated with an increase in 6-min walk and a couple of patients experienced some relief of heart failure symptoms as estimated by one-step improvement in NYHA functional class. These researchers concluded that injection of myoblasts is feasible and may improve symptoms, but had no detectable effect on heart function or size (47).
To summarize, several points must be here emphasized. It is worthy of notice that in a majority of trials routes of delivery were a direct cell implantation, most optimal for cell homing in myocardium (23). At least half of the trials, however, included myoblast implantation as an additional manipulation to surgical procedures; therefore, the sole myoblast effect is still difficult to estimate. A negative view can be further supported by lack of the proper control groups. A glimpse of hope, however, can be brought by some correlation between ejection fraction improvement and number of implanted cells (33, 50), while 13 out of 15 trials performed indicated LVEF improvement, although not always statistically significantly. Summary of the discussed clinical trials is shown in Table 1.
The results obtained in these clinical trials are encouraging but still preliminary. A true concern is associated with the incidence of cardiac arrhythmia occurring in patients after myoblast transplantation. Menasche et al. reported sustained monomorphic ventricular tachycardia (VT), which occurred early after the operation. Its resistance to pharmacological treatment necessitated the implantation of an automatic internal cardioverter/defibrillator (35).
In the MAGIC trial there was not one single death due to the procedure or due to arrhythmic events (33). Patients participating in the Pagani trial suffered from cardiac arrhythmias (i.e., atrial fibrillation) and from VT (37). In the Smits et al. pilot study there was one episode where the patient needed to be admitted to hospital at 6 weeks after the procedure due to long asymptomatic runs of nonsustained ventricular tachycardia (53). In the Herreros et al. trial, 40 days after surgery one patient developed nonsustained ventricular tachycardia (19). Patients from the Ince trial experienced VT episodes: at 30 days postmyoblast implantation in one patient and at 27 and 41 days in the other patient (21). During the early postoperative period two episodes of sustained VT were also reported in the Siminiak trial (70). In the Dib et al. trial serious adverse events were noted in patients undergoing CABG as well as LVAD implantation. Patients suffered from atrial fibrillation, ventricular tachycardia (sustained and nonsustained), and also ventricular fibrillation (11). However, most incidences of arrhythmia were clinically well tolerated.
Because many of the patients included in these trials represented a group at high risk for negative electrical events (MADIT II criteria, Multicenter Automatic Defibrillator Implantation Trial II) (9), the observed arrhythmia might not be directly related to cell implantation. The result of the Ince trial, where three patients of the control group experienced a single episode of VT, may support this hypothesis (21). Some adverse events, however, also occurred in the Biagini et al. trial where during the 29-month follow-up one patient was readmitted to the hospital due to heart failure symptoms, and two of the six with an ICD had appropriate shocks due to major arrhythmias (6). The follow up (1–12 months) in the MYOHEART trial revealed one single death, brief runs of nonsustained VT, and atrial fibrillation (49). In the CAuSMIC trial one subject in each dose group was hospitalized for congestive heart failure (10). The most common side effects noted are shown in Table 2.
Clinical Trials Involving Serious Adverse Events
P, placebo.
To overcome the potential risk of these events, a lot of patients enrolled in more recent trials received automatic implantable cardio-defibrillators (AICD) or low doses of antiarrhythmic agents prophylactically. This approach resulted in a substantial reduction of the arrhythmia incidence observed in patients participating in the MAGIC trial (33). It seems possible that the local tissue injury caused by intramyocardial injection can be in part responsible for the observed arrhythmia because most cases of such negative electrical events in the early clinical trials have been associated with intramyocardial rather than intracoronary injection (26). Although bone marrow stem cells (BMSC) show much different properties than myoblasts, their intracoronary administration could be an explanation for the extremely rare cases of malignant arrhythmia (4). Furthermore, experiments with HeLa cells suggest that any electrically isolated cell introduced into the myocardium may increase the arrhythmia risk (12). Nevertheless, regardless of arrhythmia origin and in spite of rather transient characteristics of these negative electrical events, the long-term follow-up studies of myoblast recipients are essential to minimize any risk associated with the cell therapy.
There is a tremendous need for the randomized placebo controlled trials designed to answer the most urgent questions. The development of strict inclusion and exclusion criteria, better establishment of the target population, cell delivery routes, precise cell numbers with defined phenotype, and good methods for end-point measurements should make unambiguous data interpretation and the creation of central data registries. Again, keeping in mind that most of the initial studies on myoblast transplantation have been carried out as an adjunct to routine surgical procedures such as CABG and LVAD implantation, such approach blurs the true clinical outcome in regard to cell transplantation per se and makes it impossible to directly relate the cell implantations to the beneficial effects observed.
How Do Myoblasts Improve Cardiac Function?
The mechanism by which implantation of skeletal myoblasts improves heart function still remains unclear. There are at least three hypotheses in regard to the cells' actions within the cardiac muscle. The first hypothesis claims that implanted cells might act as a strengthening scaffold within the ventricular wall and thus limit infarct expansion and subsequent ventricular remodeling (34). The improved structural support might facilitate the survival of grafted cells and residual cardiomyocytes (32). This assumption is strongly supported by improved regional diastolic function observed not only after myoblast transplantation but also after noncontractile cell delivery, including fibroblasts (20) and bone marrow-derived CD133+ cells (2). What remains to be determined is whether or not the cell engraftment affects extracellular matrix organization (through the inhibition of matrix metalloproteinases and/or an increase in collagen network). It should be noted, however, that this hypothesis excludes the possibility of myoblast influence on the completed remodeling process. Improvement in cardiac function observed in patients having undergone myoblast transplantation in old infarcts (32) implies that the proposed mechanism, if it is a real one, is surely not the only one to be involved.
In this context, the second hypothesis postulates that myoblasts, due to their contractile properties, contribute directly to cardiac systolic function (32). This hypothesis was first challenged by the observation that skeletal myoblasts neither transdifferentiate into cardiomyocytes after cardiac grafting nor express cardiac-specific genes [connexin 43 (Cx43), N-cadherin] required for electromechanical coupling with cardiomyocytes (25, 43, 45). Although gap junction communication and synchronous contraction was demonstrated in coculture of myoblasts with cardiomyocytes (14, 43), there are no data confirming that this process takes place in vivo as well. There are, however, a considerable number of indirect arguments suggesting that the lack of positive staining for gap junction proteins does not preclude a direct role of grafted cells in the enhancement of cardiac muscle contractility (32). Firstly, the comparison between fetal cardiomyocyte implantation and myoblasts failed to show any difference in functional outcome. Secondly, the cells lacking contractile properties (e.g., fibroblasts that have been shown to improve diastolic parameters) failed to improve systolic function or even worsened it (20), and thirdly, intramyocardially generated skeletal myotubes not only aligned in parallel to host myocardial fibers but also displayed well-defined striations (32).
The fact that a contractile apparatus of skeletal muscle fibers tends to disappear in noninnervated, noncontractile muscles (e.g., during neurological diseases with nerve degradation) (17) suggests that the contractile apparatus of grafted cells might be somehow active. An increase in the expression of slow myosin heavy chain isoform (11), which renders the myotubes fatigue resistant, further supports the hypothesis of the myoblast's direct involvement in cardiac muscle contraction. Because no gap junction presence within grafted myotubes has been documented, the mechanism is difficult to understand and remains to be elucidated. It is conceivable that myotubes could contract in response to the mechanical stress exerted by the surrounding cardiomyocytes (34). This, however, implies the existence of some physical link between the cellular graft and extracellular matrix to which native cardiomyocytes are connected (e.g., to integrins that are known to participate in cell adhesion and signal transduction) (32). Another possible explanation involves a field effect generated by cardiomyocytes that is directly channeled through the cell membranes to evoke action potentials within myotubes (32).
It could be also that transplanted cells constitute a potent source of growth and/or angiogenic factors. As for the angiogenic effect of grafted cells, most of the studies failed to prove increased angiogenesis beyond the one seen in control hearts receiving a culture medium alone (32). The paracrine effect exerted by grafted myoblasts on putative resident cardiac myocytes or cardiac stem cells, however, cannot be excluded.
The proposed hypotheses do not exclude each other. It seems that the observed functional outcome is multifactorial in its origin and might engage different mechanisms—not all of them directly connected to the grafted cell.
Successful Myoblast Therapy and Emerging Opportunities for Improvement
There are a number of steps necessary for successful cell therapy, such as: delivery to the target tissue, long-term survival within regenerated organ, efficacious engraftment, differentiation into contractile cells, and integration into the new, unique microenvironment. All of these steps constitute a potential goal for cell manipulation aiming to improve the overall outcome of such therapy.
Routes of Cell Delivery and its Early Survival
Cell delivery is an important issue not only in terms of safety and efficacy of the procedure but also in regard to the feasibility of data comparison between different research groups. Given that the majority of early clinical trials with myoblasts were conducted adjunct to standard surgical procedures such as CABG and LVAD implantation, cell injections were usually accomplished under a direct control through multiple epicardial punctures. To reduce the procedural invasiveness, percutaneous approaches have been initiated. This was facilitated by the considerable development in the area of catheter design and the navigation system that currently makes intracoronary, intravenous, and endoventricular injection of cells possible (32). In spite of the successful intracoronary (57) and intravenous (50) administration of autologous myoblasts, more emphasis is being put on the endoventricular route (32). But one of the constraints is an evident lack of robust animal data (as it is in the case of the epicardial injection during surgical approach) proving functional efficacy of this procedure.
Regardless of the cell delivery method, the major limitation of cell therapy is the high degree of cell death. As for the cell administration approach, this excessive attrition rate is attributed to physical strain exerted on myoblasts during injections (31) and the inflammatory process triggered within a graft destination site by multiple punctures. On the other hand, mechanical stress (in this case epicardial puncture alone) has been shown to increase angiogenesis (34), the process that might positively impact a cell engraftment. In order to enhance skeletal myoblast survival in the first phase of implantation, heat shock and cryopreservation treatment prior to injection have been evaluated (28, 58). These processes are known to induce expression of self-protecting proteins, which might attenuate the deleterious effects of inflammation and free radicals generated upon mechanical stress, and cell implantation during the initial phase of myoblast delivery. The early survival rate of pretreated myoblasts was shown to be significantly improved. However, heat shock treatment and cryopreservation blunted the subsequent proliferation abilities of engrafted cells (28). Therefore, in total, these procedures failed to improve the ultimate magnitude of myoblast engraftment.
Recently, investigation on a porcine model revealed that limited success in cell engraftment is caused by inadequate microvascular environment in infarcted tissue. The researchers proposed transmyocardial laser revascularization (TMR) as a pretreatment before cell implantation in order to provide early survival. Results of this investigation are encouraging because the use of TMR as a pretreatment for cell injection increases early survival in infarcted myocardium without increased adverse events (38). Suzuki et al. showed that IL-1β-mediated inflammatory response caused by mechanical damage associated with cell injection is involved in a graft loss (55). This group demonstrated that coadministration of myoblasts and anti IL-1β antibodies improved graft survival and resulted in a twofold increase in the total number of implanted cells (55). Such an approach might considerably enhance the therapeutic effects of myoblast transplantation, especially in the treatment of acute myocardial infarction in which IL-1β is upregulated before cell administration. The other promising method of suppression of the inflammation associated with cell implantation and/or infarction involves ex vivo gene therapy, where myoblasts would be transfected with genes known to restrict apoptosis and inflammatory reactions, and those that limit synthesis and release of free radicals (e.g., catalase, insulin-like growth factor, TGF-β1, Hsp70, cytokine signaling inhibitors) (18).
Other dilemmas in the area of cell delivery include: the number of injected cells and the timing of injection (early vs. late postinfarct injections, single vs. repetitive implantation). The positive correlation between the extent of improvement in cardiac function and the number of cells injected is widely accepted (41, 59). There is, however, a serious risk of tissue overgrowth and massive cell death in cases where a large number of cells is introduced at one time. An additional problem lies in the practical limitation concerning multiple cell passages required to obtain myoblasts in vast quantities. Firstly, the duration of in vitro cultures should remain within the clinically acceptable time frame of 4–6 weeks; secondly, the prolonged in vitro culture might lead to the development of differentiation-defective population of cells (5). The efficacy and feasibility of the repetitive implantation of skeletal myoblasts have been evaluated in rats and pigs, and the preliminary results are encouraging (59). Premaratne et al. showed that repeated myoblast implantation significantly improved LV function and resulted in significantly larger engrafted volume and LV contractility compared with single transplantation in rats (42). This strategy allows for the implantation of the greater number of cells with the simultaneous circumvention of the risk associated with massive cell death after single injection of high cell dosage. The optimal time frame for cell transplantation still needs to be defined as both too early and too late postinfarct injections may be equally ineffective—the former due to ongoing infarct-induced inflammation and the latter due to the completed remodeling process (31).
The most innovative approach to cell delivery involves tissue-engineered constructs (30). The main advantage of this technology over standard cell implantation lies in the preservation of myoblast microcellular communication and matrix, which is lost upon trypsin treatment in the typical procedure of cell preparation (30). The use of a scaffold for tissue-engineering technology, although attractive, is subject to biocompatibility-, biodegradability-, and cytotoxicity-associated problems. In order to eliminate these problems, new cell sheets devoid of scaffolds have been established (30, 46). Sawa et al. reported that myoblast sheet implantation improved global cardiac function to a greater extent than the injection of cell suspension (46). In regard to these results, it seems that myoblast sheets might become a universal cell carrier for heart repair in the future.
Long-Term Survival
The long-term survival of engrafted myoblasts constitutes a serious and unsolved problem in cell therapy. Although skeletal myoblasts are considerably tolerant to poor graft environment, the majority of cells transplanted into myocardium do not survive due to ischemia and apoptosis (64). Depending on the time course of cell injection, there are distinct approaches to increase myoblast survival. This is due to completely different environmental conditions and the desired outcome in early or late postinfarction therapy. In cases where chronic ischemia prevails, the primary objective is to eliminate the extremely hypoxic conditions by increasing angiogenesis and to ameliorate myoblast survival. The early postinfarction scenario requires that transplanted myoblasts are capable to overcome the ongoing acute inflammation process. At the moment, it seems that the most promising approach to increase graft survival is genetic manipulation. There are multiple advantages of combining gene therapy with cell therapy: firstly, overexpression of relevant genes in the myoblast serves to improve their viability and engraftment; secondly, myoblasts might be treated as vehicles to introduce desired genes and their products into a defined area of damaged tissue. Ex vivo gene therapy opens a new perspective for effective cell transfection methods with the exclusion of possible immune responses observed after in vivo use of adenoviral vectors.
Because skeletal myoblasts have been mainly used to treat heart failure in its late stages, the fundamental problem of myoblast-based cell therapy is an inadequate blood supply. Without an integrated vascular system the grafted cells will not survive in sufficient quantities. It seems that myoblasts do not initiate effective vasculogenesis (new vessel formation in the infarct bed) and/or angiogenesis (proliferation of preexisting vasculature) themselves as most of the trials failed to prove a substantial increase in vessel density within the grafted area. In order to enhance blood supply, a great number of preclinical trials, involving a wide range of proangiogenic growth factors, have been conducted. So far, vascular endothelial growth factor (VEGF) has been the most extensively investigated growth factor in regard to skeletal myoblast therapy. In 2003, Chachques et al. failed to demonstrate that the concomitant myoblast implantation and the injection of VEGF exerted any angiogenic effect in the sheep model (7). Profound results suggesting the prior role of combined therapy involving myoblasts and gene introduction were obtained by Askari et al., at least in ischemic rat myocardium (3). This group demonstrated that VEGF transduced myoblasts improved both cardiac function and vascular density while direct adenoviral delivery of VEGF resulted only in neovascularization. Myoblast implantation alone improved cardiac function, but failed to induce neovascularization. The additional finding of this study was that cell-based delivery of VEGF attenuated the inflammatory response and reduced apoptosis ongoing in cardiomyocytes and skeletal myoblasts within infarct zone (3). An increase in vascular density after VEGF expressing myoblasts was also confirmed by Xia et al. (63). Ye and coworkers showed that human skeletal myoblasts transfected with VEGF increased vascular density and heart function upon their implantation into a rodent cryoinjured heart (67). In order to eliminate the risk of excessive angiogenesis and angioma formation, these preclinical trials evaluated only short-term gene expression. Yau et al. demonstrated that transient VEGF expression is confined to transplanted myoblasts, sustained up to 4 weeks and limited only to a scar and its border zone (65). The same group reported that VEGF transfection of skeletal myoblasts induced angiogenesis and limited the apoptosis of transplanted cells (64). Law et al. demonstrated that human myoblasts transduced with the VEGF gene produced six times more capillaries in porcine myocardium than in controls (24).
The protective effect of VEGF may involve more than one mechanism. In addition to its proangiogenic activity, the vasodilatory effect of VEGF may also play a role by improving local perfusion and thus supporting the transplanted myoblasts (64). The protective effect of VEGF vasodilatory action was evaluated by Suzuki et al. in the rat model of acute myocardial infarction (56). This group demonstrated that the grafting of skeletal myoblasts expressing VEGF provided an advanced benefit in reducing infarct size and preserving cardiac function; these effects correlated with the enhanced angiogenesis observed in comparison to the hearts transplanted with control myoblasts (56).
The other growth factors examined in combination with myoblast transplantation included fibroblast growth factor (FGF) and hepatocyte growth factor (HGF). Tambara et al. investigated the efficacy of these factors in the controlled release system of gelatin hydrogel sheets as an adjunct to skeletal myoblast therapy in the rat model (59). FGF coadministration greatly increased survival of grafted cells and enhanced vessel density (59). Investigation conducted by Unzek et al. showed that mesenchymal stem cells engineered to overexpress stromal cell-derived growth factor 1 (SDF-1) led to significant decrease in cardiac myocyte apoptosis and increases in vascular density and cardiac function compared to control (62).
In spite of these encouraging results, it is now recognized that the administration of a single angiogenic factor has a limited therapeutic efficacy. Ye at al. functionally assessed human skeletal myoblasts transduced with bicistronic vector carrying VEGF and agiopoietin-1 (66). The rationale for this study was to evaluate whether or not the combination of VEGF with angiopoietin-1, which is responsible for blood vessel maturation, constitutes an efficient strategy to develop functional and mature vasculature. The conditioned medium from transduced skeletal myoblasts stimulated HUVEC to proliferate much faster. The coculture of these cells resulted in enhanced formation of capillary-like structure (66). Another approach to improving angiogenic gene therapy is the use of a master gene that can control the expression of the wider array of downstream effectors. Azarnoush et al. evaluated the functional outcome of therapy involving concomitant administration of skeletal myoblasts and adenovirus-encoded hypoxia-inducible factor 1α (HIF-1α), a master gene that controls the expression of angiogenic growth factors genes, in the rat model (5). The most salient finding of this study was that the simultaneous intramyocardial administration of cells and HIF-1α improved LV systolic function to a greater extent than myoblast grafting alone. This improvement correlated with an increase in angiogenesis, cell survival, and graft area (5). The lack of functional benefits in hearts injected with HIF-1α alone implies that increasing angiogenesis in scar tissues devoid of cardiomyocytes is unlikely to affect the contractile properties of that fibrous area. These results further support the necessity to apply combined therapy for heart failure (cells with contractile capacity and angiogenic factors or cells transfected with angiogenic genes).
Functional Integration Into Myocardium
The great number of preclinical and clinical studies provided compelling evidence that skeletal myoblasts differentiate into myotubes with well-defined striations within a recipient's myocardium (32). Unfortunately, there are no data confirming the mechanical and electrical coupling between myotubes and host cardiomyocytes. The lack of gap junctions is partly blamed for the incidence of cardiac arrhythmia observed in patients after skeletal myoblast transplantation (52). In order to eliminate these side effects and to further improve systolic parameters in the failing heart, different groups conducted experiments aiming to enhance the electrical and mechanical coupling between myotubes and cardiomyocytes. It has been shown that the major adhesion and the gap junction proteins (N-cadherin and connexin 43, respectively), although expressed in undifferentiated skeletal myoblasts, are markedly downregulated after differentiation into myotubes and are virtually absent in skeletal muscle grafts in injured hearts (43). The drastic decline in the connexin 43 gene expression, at least in cardiomyocytes, was shown to be a consequence of inflammation (13), which might also be a case in regard to the engrafted myoblasts that do not express these proteins at all (43). The in vitro coculture of myoblasts and cardiomyocytes was shown to enhance the connexin 43 expression in undifferentiated cells and resulted in functional gap junction formation. This effect required direct cell-to-cell contact between the two cell types (the conditioned medium containing soluble factors released from cardiomyocytes failed to substantially increase connexin 43 expression) and was reinforced by treatment with relaxin, a cardiotropic hormone (14).
Although encouraging, these results indicate that spontaneous connexin 43 upregulation in grafted myoblasts is rather implausible because there is hardly any direct contact between transplanted cells and resident cardiomyocytes. Suzuki et al. generated the connexin 43-overexpressing rat skeletal myoblast cell line with enhanced fusion and differentiation capacity compared to the control-transfected myoblasts (54). The transgene expression was sustained both in myoblasts and myotubes and caused the significant enhancement in intercellular dye transfer, implying the existence of functionally active gap junctions (54). Tolmachov et al. demonstrated that the coculture of connexin 43-transduced skeletal myoblasts with cardiac myocytes exhibited the extracellular activation rates of field action potential similar to those observed in pure cultures of cardiac myocytes (61). Contrary to the results obtained by the Suzuki group, Reinecke et al. reported that constitutive expression of connexin 43, although successful in cultured myoblasts, caused a significant death rate upon cell differentiation (44). In contrast, transfection of already differentiated myotubes was nontoxic, implying a window of vulnerability during the differentiation process. To delay Cx43 overexpression until after myoblast differentiation, myoblasts were transfected with vector containing Cx43 under the muscle creatine kinase promoter, which is active only in mature myotubes. This approach resulted in high levels of Cx43 expression in differentiated myotubes, did not cause cell death, and enabled viable cell graft formation upon percutaneous or IM myoblast transplantation (44).
However, the existence of functional gap junctions between the host and grafted cells is still under investigation. If such coupling takes place in vivo, the next step will be to estimate whether or not this promotes contractile function and prevents arrhythmia. The preliminary results obtained by Abraham et al. suggest that connexin 43 overexpression might eliminate arrhythmia (1). This group observed a decreased arrhythmogenicity in cocultures of cardiomyocytes and Cx43-overexpressing myoblasts (1).
Treatment for Nonischemic Cardiomyopathy
Most of the experiments and clinical trials involving skeletal myoblasts were focused on the efficacy of cell transplantation in heart failure of ischemic origin. At this stage, very little is known about the effect of cell transplantation on globally dysfunctional hearts, such as those in dilated cardiomyopathy (DCM). DCM is characterized by a substantial decrease in cardiac muscle contractility, as well as ventricular dilation and ventricular wall thinning, all of which lead to progressive heart failure. Cardiac transplantation, which is the most effective life-saving therapy in advanced stages of DCM, has limited applicability due to the shortage of donor organs and immunorejection (69). In this context, transplantation of contractile cells constitutes an attractive alternative to improve systolic function of globally dysfunctional heart.
Pouly et al. evaluated the feasibility and efficacy of multiple direct intramyocardial injections of autologous skeletal myoblasts into a globally dilated myocardium (39). In order to reflect the clinical setting of the disease as closely as possible, the experiment was conducted on a Syrian hamster strain exhibiting a mutation in the δ-sarcoglycan gene, which leads to a global DCM evolving towards heart failure. The major finding of this study was that multiple IM injections of skeletal myoblasts did result in successful cell engraftment, seemed to preserve LV function, and did not increase disease-associated interstitial fibrosis. The myoblasts' inability to improve diastolic function in this model of nonischemic cardiomyopathy suggests that the scaffolding effect of the grafted cells does not play a role. The plausible mechanism may involve changes in the extracellular matrix composition, direct contraction of engrafted myotubes in response to gap junction-independent electronic currents fired by the neighboring cardiomyocytes and/or secretion of growth factors, which might trigger activation of cardiac stem cells. Although encouraging, this study has a major limitation (a potential proarrhythmic risk of cell transplantation was not addressed in this experiment and only 1-month follow-up was not sufficient to confirm the long-term survival of grafted cells) (39).
In this regard, there is a tremendous need for new trials designed to address the diffuse nature of the disease with a special emphasis put on the optimal cell delivery route and a clinically relevant animal model of DCM. In our first attempt (unpublished data) to implement autologous myoblast to the patient with nonischemic cardiomyopathy brought expected arrhythmia attacks, specifically in the first month of follow-up (although not in the first 2 weeks after administration) as well as unexpectedly high improvement in ejection fraction (EF >10%).
Conclusions
Despite the ongoing extensive research involving cell-based therapy for heart failure, our current knowledge remains quite poor. We have not yet recognized the mechanisms by which cells (myoblasts or bone marrow stem cells) improve cardiac function. It is known that cell implantation increases cardiac contractility and/or limits infarct expansion and heart remodeling. But in order to improve these properties appropriate gene therapy application turned out to be relevant. Selected genes should promote myogenesis, revascularization, and electromechanical coupling between transplanted cells and cardiomyocytes in targeted regions. The precise understanding of these “healing processes” is absolutely essential to guide the sensible direction of further research.
In regard to skeletal myoblast transplantation, there is still unsolved safety issues concerning the risk of arrhythmia. However, improved cell replacement therapy may recover cardiac function to such an extent that any induced arrhythmic risk will prove clinically unimportant.
There is an ongoing debate about which cell type provides the best option for clinical application: skeletal myoblast or BMSC. Each cell type enjoys superiority over the other in certain aspects, and the choice between those two depends on the required outcome of the procedure (angiogenesis vs. myogenesis). Nevertheless, one new aspect that should be taken into consideration in order to eliminate arhythmic risk is an option to find appropriate cell fraction from heterogenous skeletal muscle tissue populations. Among at least three fractions of skeletal muscle stem cells, muscle-derived stem cells (MDSC)-like cells seems to bring a promising future for tissue regeneration. Better understanding of all these approaches could shed light on cell-based therapy and point out a new possible direction of tissue engineering.
Footnotes
Acknowledgments
This work was supported by Polpharma III/II/2004, PBZ-KBN-099, and N403 065 31/3011 grants.