Use of Mesenchymal Stem Cells for Therapy of Cardiac Disease

Abstract

Despite substantial clinical advances over the past 65 years, cardiovascular disease remains the leading cause of death in America. The past 15 years has witnessed major basic and translational interest in the use of stem and/or precursor cells as a therapeutic agent for chronically injured organs. Among the cell types under investigation, adult mesenchymal stem cells (MSCs) are widely studied and in early stage clinical studies show promise for repair and regeneration of cardiac tissues. The ability of MSCs to differentiate into mesoderm and non-mesoderm derived tissues, their immunomodulatory effects, their availability and their key role in maintaining and replenishing endogenous stem cell niches have rendered them one of the most heavily investigated and clinically tested type of stem cell. Accumulating data from preclinical and early phase clinical trials document their safety when delivered as either autologous or allogeneic forms in a range of cardiovascular diseases, but also importantly define parameters of clinical efficacy that justify further investigation in larger clinical trials. Here, we review the biology of MSCs, their interaction with endogenous molecular and cellular pathways, and their modulation of immune responses. Additionally, we discuss factors that enhance their proliferative and regenerative ability and factors that may hinder their effectiveness in the clinical setting.

Introduction

Cardiovascular disease is the leading cause of mortality and morbidity in the US and worldwide¹. Attributed mainly to myocardial infarction (MI) and its fatal sequelae, heart failure and sudden cardiac death, cardiovascular diseases carry an enormous psychological and financial burden to patients, their families and society. Over the past half a century conventional medicine and surgery have offered many breakthroughs, resulting in a dramatic decline in CV mortality¹. Despite the major advances, medical or surgical treatment of chronic heart disease yields only temporary delay in a progressive disease process², with the only definite cure remaining heart transplantation.

The idea of using stem or precursor cells has emerged in the last decade as a leading approach for a regenerative strategy to address cardiac disease³. In this context, mesenchymal stem cells are lead candidates for cellular therapy not only for heart disease, but multiple diseases characterized by fibrosis⁴. Bone marrow (BM) and adipose derived MSCs are easily isolated, expanded and immunologically tolerated allowing for allogeneic, “off the shelf” transplantation⁵. The ability to use pre-prepared allogeneic cells for cell-based therapy allows for a level of quality control and scalability that far exceeds autologous strategies⁵. In this review we describe the biology, the mechanisms of action, the emerging preclinical and clinical trial results, and discuss potential strategies that will refine the development of MSCs as a therapeutic tool.

What constitutes a Mesenchymal Stem Cell?

In 1970 Friedenstein⁶ discovered a rare (0.0001%-0.01% of nucleated cells in human BM) population of plastic adherent stromal cells residing in the BM. These cells, which readily expand in culture, are now commonly called mesenchymal stem or stromal cells, and are recognized to play an integral role in the hematopoietic niche⁷. In 2006, the International Society for Cellular Therapy established minimal requirements⁸ for MSC definition: 1) adherence to plastic in standard culture conditions; 2) expression of the surface molecules CD73, CD90, and CD105 in the absence of CD34, CD45, HLA-DR, CD14 or CD11b, CD79a, or CD19 surface molecules and 3) a capacity for differentiation into osteoblasts, adipocytes, and chondroblasts in vitro (Figure 1). The rationale behind the selection of these criteria was to facilitate the comparison of different studies, to foster a more uniform characterization of MSC and render the exchange of data among investigators easier. However, these markers represent a range of MSC differentiation potential. Furthermore, these criteria apply to human MSCs, but do not necessarily extend to other species⁹, and also following culture these markers may be lost or new markers may arise. Some reports fail to meet these criteria, thus making an across the board comparison difficult. The convention of referring to human BMMSCs as stem cells results from their proven self-renewal capability and capacity for multilineage differentiation¹⁰ (Figure 1).

Figure 1. Schematic overview of signaling molecules and transcription factors involved in the regulation of differentiation of MSCs.

VEGF, vascular endothelial growth factor; BMP-2, bone morphogenetic protein-2; EGF, epidermal growth factor; FGF-2, Fibroblast growth factor-2; LRP5/6, low-density lipoprotein receptor-related protein-5/6; Osx, Osterix; PDGF, platelet-derived growth factor; RUNX2, runt-related transcription factor-2; TGF-b, transforming growth factor-b; MRF, myogenic regulatory factors; MEF2, myocyte enhancer factor-2. Figure modified and reporduced with permission from Augello et al. Hum Gene Ther 2010⁴⁵

Sources and Types of MSCs

MSCs are broadly distributed throughout the body¹¹ outside BM, and reside in adipose tissue, gut, lung, liver, placenta, amniotic fluid, dental pulp, periodontal ligament and recently in the heart^{12, 13}. The cells most commonly used in clinical trials to date originate from BM (MSCs and MPCs), adipose tissue, and umbilical cord³ (Table 1). Umbilical cord blood-derived MSCs have been extensively studied in models of heart disease, but their isolation can be difficult¹⁴. Wharton's jelly of the umbilical cord is also a rich source of MSCs, but mainly studied in the context of heart valve tissue engineering¹⁵. Cells that share some of the characteristics of MSCs can be identified in peripheral blood¹⁶.

Table 1.

Comparison of characteristics of MPC, BMMSC, ATMSC, and UCMSC

Feature	MPC	BMMSC	ATMSC	UCMSC
Yield	1-3% of BM mononucleated cells	0.001-0.01% of total marrow nucleated cells	0.5-5% of adipose tissue
Immunophenotype
CD90	-	+	+	+
CD105	+	+	+	+
CD73	-	+	+	+
CD31	+	-	-	-
CD146	-	±	-	+
CD271	+	+	+	+
CD34	-	-	±	-
CD49d	+	-	-	-
CD106	+	+	-	±
CD54	+	-	+	+
MHC I	+	+	+	+
MHC II	-	-	-	-
CD40	-	-	-	-
CD80	-	-	-	-
CD86	-	-	-	-
STRO-1/STRO-3	+	-	-	-
Differentiation potential*
Adipogenic	+	-	+	+
Osteogenic	+	+	-	+
Chondrogenic	±	+	+	+
Smooth muscle cells	+	+	+	+
Endothelial	+	+	+	+
Transcriptome and proteome**	IFN-γ	WNT11, WNT7B, SOX6	CCL3, FGF9, IL1R2, KDR, PAX3, SPI1, ZNF45	Vimentin, nph3, gelsolin, prohibitin,

MPC – Mesenchymal precursor cell, BMMSC – BM derived mesenchymal stem cell, ATMSC – Adipose tissue derived mesenchymal stem cell, UCMSC – Umbilical Cord derived mesenchymal stem cell.

^*There are conflicting data that imply that MSCs from different sources can respond differently to different stimuli.

^**uniquely expressed or in significantly higher levels in each population

Amniotic fluid derived MSCs possess some of the characteristics of embryonic stem cells. In't Anker et al¹⁷ described amniotic fluid derived MSCs (AFMSC), with full in vitro multipotential differentiation capacity¹⁸. In addition, AFMSCs have a ckit+ subpopulation¹⁹ that also expresses embryonic stem cell markers (Oct-4, Nanog, and SSEA-4). AFMSCs are phenotypically similar to BMMSCs²⁰, sharing similar immunologic profiles. Like BMMSCs, AFMSC express proteins and genes of HLA-ABC (MHC class I) but not those of HLA-DR (MHC class II)²¹.

Another subpopulation of BM derived mononuclear cells, called Mesenchymal Precursor Cells (MPCs), can be isolated from human BM aspirates²². A widely used methodology to identify MPCs is by selecting cells bearing the Stro-1 or Stro-3 receptor from bone marrow and then culture expanding that cell (Table 1). In humans, several surface proteins, including Stro-1, CD271, and CD146 may be used as markers for MPCs²³. Recently, heart derived MSC-like cells have been identified^{12, 13} and have been used both in animal models of heart disease and in the CADUCEUS study²⁴. It has also been shown that cardiac stromal cells treated with a cocktail of epigenetic drugs differentiate into functional cardiovascular precursors²⁵. Cardiac stromal cells exhibit many phenotypical similarities²⁶ to their bone marrow counterpart, however they demonstrate an ability to acquire a cardiomyocyte phenotype more efficiently²⁷. In 2002, Jiang et al²⁸ isolated adult cells from rodent bone marrow and suggested that under some conditions may have broad differentiation potency. They named those cells multipotent adult progenitor cells (MAPCs). MAPCs have been shown under appropriate circumstances to be able to differentiated into hematopoietic cells, osteoblasts, hepatocytes, neurons and also possess immunomodulatory capabilities²⁹.

Tissue sources of MSCs

Adipose and BMMSCs are the most heavily investigated and tested³. The abundance of MSCs derived from adipose tissue and the ease of acquiring them with a liposuction, makes this source attractive and feasible for clinical trials (Table 1, supplementary table V). However, there has been a degree of uncertainty on whether adipose derived MSCs are truly MSCs, hence they are often termed as “adipose tissue stem cells”³⁰. Adipose derived and BMMSCs share many biological characteristics; however, there are some differences in their immunophenotype, differentiation potential, transcriptome, proteome, and immunomodulatory activity (Table 1). Some of these differences may represent specific features, while others are suggestive of the inherent heterogeneity of both populations³¹. To date there is no conclusive head to head in vivo comparison of those two sources and BMMSCs are more widely studied.

Isolation and Expansion of MSCs

It is possible to obtain (after expansion) 50-400 million or even more cells from a BM aspirate of 10ml¹⁰. The 3 critical steps that allow MSCs to be isolated from other BM cells are: 1) use of density gradient centrifugation (i.e Ficoll or Percoll) to separate non nucleated red blood cells from nucleated cells or cell mobilization and isolation; 2) the ability of MSCs to adhere to plastic; and 3) the ability of monocytes to be separated from MSCs by trypsinization³². To isolate MSC from a BM aspirate, cord blood or peripheral blood, the samples are fractionated by density gradient for mononuclear cell isolation, resuspended in appropriate culture medium containing selected batches of fetal bovine serum (FBS) and allowed to adhere to plastic dishes for 2 days; then, non-adherent cells are removed and the remaining cells allowed to grow for 2–3 weeks. Cells initially generate a heterogeneous adherent cell layer including fibroblast-like and small round-shaped cells, while they appear uniformly spindle shaped after several passages in culture. Confluent cells are trypsinized and allowed to expand for as many as 40 generations without loss of multipotentiality. A panel of monoclonal antibodies against epitopes expressed on their surface is used for phenotypic characterization¹⁰.

Biochemical cross-talk and other factors regulating MSC differentiation

Colony Forming Units (CFU), once isolated, can commit to particular cell lineages through treatment with bioactive factors in vitro or in vivo (Figure 1). When vascular endothelial growth factor (VEGF)³³ is present, CFUs preferentially take up vascular endothelial fate. Similarly, 5-azacytidine treatment induces their cardiac differentiation in vitro³⁴. However, other factors shown to induce CFU differentiation into cardiac cells recently, including dexamethasone and ascorbic acid³⁵, bone morphogenetic protein-2 (BMP-2) and fibroblast growth factor-4 (FGF-4)³⁶, represent alternative avenues of CFU pretreatment. Treating CFU with dexamethasone, 1 3 -glycerol phosphate, and ascorbic acid, prompts differentiation into osteogenic cells that forms a mineralized extracellular matrix³⁷. Chondrogenic differentiation is achieved by using dexamethasone and TGF-β3³⁸, adipogenic through dexamethasone, insulin, indomethacin and 1-methyl-3-isobutylxanthine¹⁰. Exposed to basic fibroblast factor, dimethylsulfoxide, β-mercaptoethanol, and butylated hydroxyanisole, MSC differentiate into cells expressing neuronal phenotype³⁹. MSCs also differentiate into other mesoderm-derived tissue, notably myocytes, as shown in some but not all studies⁴⁰.

MSC themselves also secrete several cytokines. Many cytokines pertinent to hematopoietic cells proliferation and differentiation are among them (interleukin-6, Flt-3 ligand, G-CSF, GM-CSF). MSCs express several adhesion-related antigens (CD166, CD54, CD31, CD106) and integrins (CD49, CD29, CD11, b4-integrin)¹⁰. Mechanical loading, as shown by Pijnappels et al⁴¹, may explain the increased propensity for MSC differentiation to occur in vivo or in co-culture situations that add mechanical forces.

When mouse MSC and rat ventricular myocytes were co-cultured, MSCs became actin positive and formed gap junctions with the native myocytes⁴². On the contrary, when a semipermeable membrane divided the two cell types there were no such findings, suggesting that differentiation also requires direct cell-cell contact⁴². Although this is not always true⁴⁰, in a similar experiment the MSCs started to contract, express SERCA2 and ryanodine receptor and were positive for troponin, sarcomeric actinin and desmin⁴⁰. All these suggest that cell-cell contact, cardiac microenvironment and factors secreted by myocytes play intertwined roles in MSC transdifferentiation to cardiomyocytes.

Two main molecular pathways govern MSC differentiation: the Wnt and the TGF-β superfamily pathway⁴³ (Figure 1). The Wnt pathway is critical in skeletogenesis by promoting osteoblast proliferation⁴⁴ and by suppressing chondrocyte formation. Activation of Wnt receptors on MSCs leads to downstream signal transduction that regulates cell proliferation and differentiation⁴⁴. The TGF-β pathway is involved in skeletal tissue growth and regulation of MSC differentiation into chondrocytes⁴⁵. This is achieved by up-regulation of gene expression in MSCs via several intracellular cascades, including extracellular-signal regulated kinase (ERK1/2), SMAD proteins, mitogen-activated protein (MAP) kinases, p38, and JNK⁴⁵.

Immunomodulatory properties of MSCs

MSCs are potent modulators of the immune system by suppressing white blood cells and triggering anti-inflammatory subsets (Figure 2). MSCs were used to treat therapy resistant severe graft-versus-host disease⁴⁶ based on the fact that MSCs inhibit T-cell proliferation in vitro⁴⁷. The success of MSCs has sparked vigorous investigation of their immunomodulatory capabilities. Moreover, the immunomodulatory properties in addition to the immunosuppressive properties further underlie the ability of MSCs to be used as an allograft⁵.

Figure 2. Immunomodulatory capabilities of MSCs.

Schematic overview of the interactions between MSC and the immune system. Via multiple pathways MSC suppress proliferation of both T helper (TH) and cytotoxic T cells (Tc). In addition, differentiation to TH2 and regulatory T-cells (Treg) is triggered, resulting in an anti-inflammatory environment. Maturation of dendritic cells (DC) is inhibited via IL-6, blocking upregulation of CD40, CD80, and CD86, which in turn reduce T-cell activation. Monocytes are induced by MSC to differentiate preferentially towards the M2 phenotype. IL-10 produced by the M2 macrophages can boost the formation of Treg while simultaneously reducing tissue migration of neutrophils. Neutrophils (polymorphonuclear granulocytes; PMN) are allowed longer life span but reactive oxygen species production is decreased. Natural Killer (NK) cell proliferation is suppressed as well as their cytotoxic activity. B-cell proliferation is inhibited and the production of antibodies is reduced. Modified from van den Akker et al¹⁷⁵.

MSCs when co-cultured with T-cells up-regulate indoleamine-pyrrole-2-3-dioxygenase (IDO) leading to tryptophan depletion and accumulation of metabolites such as kynurenine, both of which reduce T-cell proliferation⁴⁸. MSCs also express PD-L1 and PD-L2 ligands that activate PD-1 receptor on a T-cell resulting in decreased production of the pro-inflammatory cytokines IFN-γ, TNF-α and IL-2. In addition, MSCs have been shown to secrete TSG-6, a powerful anit-inflammatory factor⁴⁹. Moreover MSCs also down-regulate the activating receptors of natural killer cells NKp30, NKp44 and NKG2D⁵⁰. MSCs are able to arrest B-cell maturation in G0/G1 phase and simultaneously reduce the chemotactic activity of these cells and block maturation of dendritic cells resulting in reduced expression of antigens and costimulatory molecules necessary to activate T-cells⁵¹.

After an MI two major types of macrophages can be found in the heart: 1) M1 (inducible nitric oxidase synthase, MHC class II, CD80, CD86) that is clearing the debris and produces proinflammatory IL-1β, TNF-α and IFN-γ⁵² and after five days 2)M2 (arginase, macrophage mannose receptor CD206) which has an anti-inflammatory phenotype reducing the release of proinflammatory cytokines and in the same time stimulating scar formation and angiogenesis⁵³. In the presence of MSC, differentiation into the M2 subtype was boosted⁵⁴ while the debris-cleaning function remained intact⁵⁵.

The effects described above were observed in controlled in vitro studies. In vivo, transplanted cells first have to survive in order to then suppress or modulate the immunologic milieu. Their survival is aided by the fact that MSCs express moderate levels of HLA class I, lack HLA class II, B7 and CD40 ligand⁵⁶ expression. In some⁵⁷ but not all⁵⁸ in vivo preclinical studies allogeneic MSCs may lose immunoprivilege. The most important assessment of MSC immunology in humans is The Percutaneous Stem Cell Injection Delivery Effects on Neomyogenesis (POSEIDON; NCT01087996) clinical trial where allogeneic MSCs did not induce an immunologic reaction⁵ twelve months after transplantation. That was also confirmed in the clinical trial by Ascheim et al⁵⁹ in which MPCs were administered to LVAD patients.

In Vivo Mechanism of Action

There is substantial accumulating data from in vitro^{60, 61}, pre-clinical^62-65 and clinical^66-68 settings supporting a multi-factorial mechanism of action for the cardioreparative effects of MSCs. Three main mechanisms of action underlie the favorable actions of MSCs in disease: 1) reduction of fibrosis⁶⁹, 2) Stimulation of angiogenesis⁷⁰ and 3) restoration of contractile function^{66, 67} through engraftment⁷¹, differentiation and stimulation of endogenous cardiac stem cells to proliferate and differentiate⁶³ (Figure 3). These effects occur in concert and together lead to the replacement of scarred or dysfunctional myocardial tissue with contractile and perfused tissue^{65, 66}.

Figure 3. Mechanisms of action of MSCs.

The proposed mechanism of action of MSCs form an intertwined cycle of paracrine, autocrine and direct effects that include vascular regeneration, myocardial protection, cardiomyocyte regeneration that ultimately lead to cardiac repair.

Most studies that examine engraftment of cell therapy reveal low retention rates^{72, 73}. Toma and colleagues⁷¹, reported that after 4 days, 0.44% of transplanted MSCs resided in the myocardium. This low retention rate prompted the exploration of other mechanisms of action driving the recovery in cardiac structure and function.

Cardioprotection

Many early studies revealed that the border zone of an experimentally induced myocardial infarction was characterized by a reduction in apoptotic myocytes and an augmentation of vascularity^{62, 74}. Scar tissue reduction and cardioprotection after MSC transplantation is well described in both preclinical models and clinical trials. Amado et al^{64, 65} using serial CT imaging showed in-vivo reappearance of myocardial tissue and restoration of contractility after MSC implantation in swine. Gnecchi et al⁶⁰ showed that cell culture medium conditioned by hypoxic MSC reduces apoptosis and necrosis of isolated rat cardiomyocytes exposed to low oxygen tension. These investigators interpreted the findings as consistent with paracrine signaling, given the use of cell culture medium as opposed to cells themselves. This cardioprotective effect could be enhanced when MSCs were engineered to overexpress Akt-1⁶⁰. This was later confirmed in vivo⁷⁵ by a study showing enhanced survival of Akt-MSCs post transplantation. Additionally, Akt-MSCs secrete a protein (Sfrp2) that exerts a prosurvival effect through modulation of Wnt signaling⁷⁶. Rehman et al⁷⁷ showed that adipose derived stem cells secrete angiogenic and antiapoptotic factors providing further evidence of cardioprotection. The responsible pathways may include activation of inflammation and stress response associated signaling pathways mediated by IGF-1 and inhibition of transcription factor NF-kB⁷⁸. The cardioprotective effect was further enhanced by preconditioning the MSCs with TGF-β or by activating the TNF receptor-2⁶¹. In vivo, MSC cardioprotection leads to reduced number of apoptotic cells around the MSC injections⁷⁹. Thus, together these findings support the idea that in addition to other key cardioreparative effects, MSCs also create a milieu that protects or reduces stimuli driving ongoing loss of myocytes.

Neoangiogenesis

The formation of new vessels is the cornerstone of any meaningful cardiac repair. There are three mechanisms of postnatal neovascularization: 1) angiogenesis, 2) arteriogenesis, and 3) postnatal vasculogenesis⁸⁰ where endothelial precursors originating from the BM assemble to create new blood vessels. It is still under debate whether the observed increase in capillary density and tissue perfusion is due to differentiation of MSCs to endothelial cells and vascular smooth muscle cells or due to secretion of paracrine mediators and generation of new pericytes⁸¹. There is evidence that MSCs act as pericytes, perivascular cells that are essential to vascularization by stimulating the endothelial cells to form tube-like structures and subsequently vascular networks⁸². Expression of MSC markers was also detected at the surface of native, non-cultured perivascular cells. Thus, blood vessel walls harbor a reserve of progenitor cells that may be integral to the origin of the MSCs and other related adult stem cells⁸³.

In vitro, MSCs express α-smooth muscle actin and β-actin filaments⁸⁴ while in in vivo studies it is shown that MSCs expresses an endothelial phenotype that enhanced microvascular density⁷⁰. Despite this evidence, several groups suggest that only a small number of vessels contain donor cells. Therefore it is proposed that it is the release of proangiogenic and proarteriogenic factors from the MSCs that play the most important role in neovasculogenesis⁸⁵. In their experiments, there was a significant increase in the levels of VEGF and bFGF in the MSC treated animals, and that was also documented in a gene expression profiling of MSCs under hypoxia. Further supporting this theory, Markel and colleagues⁸⁶ showed that MSCs underexpressing VEGF have significantly less cardioreparative capabilities. Therefore, MSCs either directly through exosomes, or indirectly through soluble factors are strong proangiogenic factors.

Antifibrosis triggers neomyogenesis

The dominant lesion in the remodeling heart following infarction is the replacement of necrotic myocardium with scar tissue. It is the scar burden that leads to both infarct expansion and extension, and the former sets the stage for the overall remodeling and transformation of the shape of the ventricle from ellipsoid to spherical. In a fibrotic environment, type 1 collagen accumulates resulting in decreased expression of a wide array of genes, growth factors and cytokines that inhibits endogenous reparative potential of muscle⁸⁷. In addition degradation of extracellular matrix components play a key role in regulating muscular tissue regeneration⁸⁸. Thus, reducing the fibrotic tissue directly enhances endogenous myogenesis⁸⁹. Willems et al⁹⁰ showed that a 1,4-dihydropyridine inducer of type 2 TGF-β receptor (TGFBR2) degradation-1 (ITD-1) selectively enhanced the differentiation of uncommitted mesodermal cells to cardiomyocytes. Extensive studies in the field of bone and cartilage regeneration have shown us that mechanical forces and extracellular matrix components significantly influence MSCs⁹¹ but also MSCs modulate the matrix by secreting metalloproteinases and their tissue inhibitors by shifting the balance towards domination of matrix-degrading effects⁶⁹. This may require the secretion of hepatocyte growth factor⁹² and heme oxygenase-1⁹³ by MSCs. However, MSCs do not only interact with the matrix directly. It is reported that MSCs suppress the proliferation of fibroblasts and promote their metalloproteinase secretion⁹⁴.

Direct MSC stimulation of endogenous repair

MSC transplantation has now been shown by multiple groups to stimulate proliferation and differentiation of endogenous cardiac stem cells^{63, 95, 96} (Figure 4). This discovery provides a highly plausible explanation for the replacement of scarred tissue with new contractile myocardium. Neomyogenesis occurs by two related mechanisms: stimulation of endogenous cardiac stem cells (c-kit+ and other lineages) and enhancement of myocyte cell cycling⁶³. In the first demonstration of this phenomenon, GFP+ allogeneic MSCs were injected in infarcted swine hearts and led to the formation of chimeric clusters containing immature MSCs and endogenous ckit+ cardiac stem cells. These clusters exhibited cell-cell interactions mediated by connexin-43-mediated gap junctions and N-cadherin mechanical connections. Importantly, there was a 20-fold increase in the endogenous ckit+ population in MSC treated animals relative to controls, and the c-kit+ cells had much greater capacity for myocyte lineage commitment⁶³. Beltrami and colleagues co-cultured MSCs and resident cardiac stem cells and they reported maturation and increased viability of the latter⁹⁷. Loffredo et al⁹⁵ used a lineage tracing mouse to examine the capacity of cell therapy to stimulate endogenous myogenesis, and compared the ability of BM derived c-kit+ stem cells and BM-MSCs to induce proliferation of endogenous CSCs after MI. Mice were genetically modified to express GFP in cardiomyocytes. At 8 weeks after transplantation, BM c-kit+ stem cells led to a significant reduction in the GFP+ cardiomyocyte pool and parallel increases in β-galactosidase+ cardiomyocytes compared to control, suggesting increased progenitor activity induced by BM c-kit+ stem cells. However, BM-MSCs did not exhibit similar findings, suggesting that MSCs may not stimulate endogenous progenitors. This murine finding contrasts with several studies in pigs and may reflect species differences. Suzuki et al⁹⁶ found that in pigs with hibernating myocardium induced by left anterior descending artery stenosis, treatment with intracoronary injections of autologous GFP+-MSCs led to an improvement in regional wall thickening at both 2 and 6 weeks after injection compared with the control. They also noted a 4-fold increase in c-kit+ and CD133+ populations that also co-expressed GATA-4 and NKx2.5 at 3 days through 2 weeks in animals receiving MSCs. Markers of proliferation (Ki67) were significantly increased in hibernating myocardium in MSC treated animals. More rarely, fusion of transplanted MSC with resident cells may take place as proposed by certain groups⁹⁸. In a preclinical study by our group⁹⁹, the combination of human MSCs and ckit+ CSCs led to enhanced cardioreparative than that seen in either cell type alone. Together these findings exhibit important biological interactions between c-kit+ CSCs and MSCs.

Figure 4. Direct MSC stimulation of endogenous repair.

(A) Graph depicting the contribution of cardiomyocyte precursors following exogenous administration of MSCs (green line) and endogenous CSCs (orange line), during cardiac repair after MI. MSC differentiation occurs rapidly after delivery. At 2-weeks, MSCs activate endogenous expansion c-kit+ CSCs (orange line). (B) Two weeks following TEI, the number of C-kit+ cells co-expressing GATA-4 is greater in MSCs vs. non-MSCs treated hearts. The cardiac precursors are preferentially located in the IZ and BZ of the MI, indicating an active process of endogenous regeneration (‡p=0.019 and †p<0.0001) (C,D) The 2-week old chimeric myocardium contains mature cardiomyocytes (open arrow), immature MSCs (arrowheads, inset) and cardiac precursors of MSCs origin (arrow), coupled to host myocardium by connexin-43 gap junctions; Interestingly, endogenous c-kit+ CSCs are found in close proximity to MSCs (D). (E) Cluster of c-kit+ CSCs in an MSCs-treated heart; numerous CSCs are committed to cardiac lineage documented by GATA-4 and MDR-1 co-expression (arrows). (F) Few, isolated c-kit+ cells were found in non-MSC treated animals. Figure reproduced with permission from Hatzistergos et al. Circulation Research 2010⁶³

Cellular Effect of MSCs

Although the paracrine theory seems to play a role, it is the direct cellular mechanisms (exosomes, mitochondrial transfer, connexin43, etc.) that convincingly explain the effects observed in preclinical and clinical studies (Figure 3).

Extracellular Vesicles and Exosomes

MSCs exert a host of direct cellular effects by transmitting exosomes and mitochondria into recipient cells (Figure 3). Cells¹⁰⁰ continuously secrete a large number of extracellular vesicles (EV), microvesicles, macromolecular complexes, and small molecules called exosomes. Exosomes have been reported to contain significant amounts of miRNA, other non-coding RNAs, as well as mRNA¹⁰¹. Valadi et al¹⁰² reported that some full-length molecules are present, and translated extracted RNA to identifiable full-length. Several papers^{100, 103} indicate that the RNA content of exosomes differs from that of the parental cell. They facilitate immune responses and participate in antigen presentation¹⁰⁴, play roles in programmed cell death, angiogenesis, inflammation, and coagulation¹⁰⁵. Exosomes' most unique function might be specific interaction with a target recipient cell, enabling cell–cell communication, putatively between widely separated locations in the body. More recent studies have demonstrated that exosomes are not only specifically targeted to recipient cells to exchange proteins and lipids or to trigger downstream signaling events, but also deliver specific nucleic acid cargo¹⁰⁶. Embryonic stem cells secrete EVs containing Wnt-3¹⁰⁷, the protein Oct-4 but also mRNA. Those EVs derived from endothelial progenitor cells triggered angiogenesis in endothelial cells by horizontal transfer of mRNA¹⁰⁸. In addition to mRNAs those vesicles can also transfer miRNAs, further supporting the pivotal role of EVs in signaling within the stem cell niche¹⁰⁹. Both mRNAs and miRNAs contained in EVs have been suggested to mediate a bidirectional exchange of genetic information between the stem cells and the injured cells¹¹⁰. EVs of endothelial origin that were released by ischemic muscle, were able to induce the differentiation of BMMNCs into endothelial cells and thus promoting post natal vasculogenesis¹¹¹.

MSC derived exosomes were first investigated in 2010 in a mouse model of ischemia/reperfusion injury¹¹². MSCs produce increased quantities of exosomes relative to myoblasts and human embryonic kidney cell line¹¹³. MSC derived exosomes express both the common surface markers of exosomes and markers expressed on MSCs (CD29, CD44, and CD73). Pretreating MSCs with RNase completely abolishes their renal protective effect¹¹⁴. Treatment of neurons and astrocytes with MSC-derived exosomes leads to an increase of miR-133b in these cells, which promotes functional recovery in Parkinson's disease and spinal cord injury. This finding suggests that MSCs regulate neurite outgrowth at least partly by transferring miR-133b to neurons and astrocytes via the release of exosomes¹¹⁵. Purified exosomes administered to a mouse ischemia-reperfusion injury model revealed that MSCs mediate their cardioprotective paracrine effect by exosome secretion¹¹². Thus, MSC derived exosomes play a key role in cell-cell interactions, regulating everything from immune responses to neoangiogenesis in their surrounding tissue.

Mitochondrial transfer

In addition to exosomes, MSCs may transfer mitochondria through tunneling nanotubes to injured cells. Spees et al¹¹⁶ co-cultured somatic cells depleted of their mitochondrial DNA with MSC, and demonstrated rescue of respiration via the transfer of mitochondria from the healthy cells to the respiratory deficient cells. Similar results can be reproduced using cardiomyocytes¹¹⁷. While full details of the molecular mechanism remain to be elucidated, it seems to be mediated by actin-based extensions named tunneling nanotubes (TNT) and by gap junctions containing connexin 43¹¹⁸. Ahmad et al¹¹⁹ showed that Miro1 (mitochondrial Rho-GTPase) plays a key role mediating the mitochondrial transfer between cells. Interestingly, an association between Miro1 levels and mitochondrial transfer was also reported when comparing different cell types with different mitochondrial transfer capacity, with MSC expressing higher levels of Miro1 compared to lung epithelial cells and fibroblasts. Thus, MSCs through TNTs and gap junctions are able to directly rescue impaired myocytes.

Improvement of cardiac function by reconstituting the cardiac stem-cell niche

Stem cell niches are described in the bone marrow¹²⁰, hair follicles¹²¹, intestinal epithelium¹²², and in the heart¹²³. In the heart, the niche is comprised of supporting cells and cell-cell interactions that have crucial regulatory roles. MSC transplantation triggers chemokine and cytokine cascades that initiate and boost an endogenous repair mechanism through the restoration of a cellular and molecular collective with the properties of a stem cell niche¹²⁴. One key axis that modulates the niches is the SDF-1α/CXCR4 axis that also regulates the homing of hematopoietic stem cells to the injured myocardium. Shi and colleagues¹²⁵ reported high levels of CXCR4 expression both on the surface and intracellularly in BMMSCs even after several passages. When the MSCs were exposed to Flt-3 ligand, stem cell factor, IL-6, hepatocyte growth factor, and IL-3 they up-regulated the expression of CXCR4. Cardiac myocytes can be induced to reenter the cell cycle after treatment with periostin, FGF-1/p38 MAP kinase inhibitor¹²⁶, neuropeptidase and TGF-β. The fact that some of these factors are secreted by MSCs underlines their role in activating the niche's proliferative potential and in fact modulating the cardiovascular microenvironment towards a reparative mode.

Harnessing and enhancing the therapeutic potential of MSCs

Numerous efforts to further enhance the therapeutic potential of MSCs by genetically modifying or pretreating MSCs with various drugs and cytokines are under way. While combining MSCs and simvastatin¹²⁷ did not impact the reduction in perfusion defect, scar size or EDV, it did potentiate increases in EF and systolic wall thickening as well as improved MSC retention and survival. Trimetazidine pre-conditioned MSCs decreased LDH levels and enhanced production of survival proteins including HIF-1α, survivin, phosphorylated Akt, and Bcl-2 protein levels and Bcl-2 gene expression¹²⁸.

Guo and colleagues¹²⁹ pretreated MSCs with IGF-1 that led to overexpression of CXCR4 resulting in improved wall thickness and left ventricular function as compared to naïve MSCs alone. Similarly, preconditioning MSCs with VEGF resulted in increased activation of the Akt pathway, lower expression of cell aging markers and cell cycle inhibitors (p16 and p21)¹³⁰. In another study, heme oxygenase-1 was used, that improved MSC survival in hypoxic conditions and up-regulated the secretion of various cytoprotective cytokines by MSCs¹³¹. The combination of various factors was also quickly put to test by various research groups. One such combination¹³² was a cocktail of FGF-2, IGF-1, and BMP-2 to pretreat MSCs before transplantation that led to increased cardiac specific markers and also interestingly enhanced expression of phosphorylated Akt and phosphorylated cAMP response by native cardiomyocytes. Surprising Akt activation not only enhanced the survival of the transplanted MSCs but also decreased the apoptosis of the surrounding myocytes¹³².

Another intriguing and more difficult approach is that of genetic modification of MSCs. MSCs that were modified to overexpress Bcl-2, had reduced apoptosis and increased secretion of VEGF that led to longer survival even at 6 weeks post transplantation in vivo⁷⁹. Transfected MSCs to overexpress SDF-1 receptor, a chemotactic factor for lymphocytes, resulted in both increased MSC retention and expression of SDF-1 by the ischemic myocardium that triggered increased intracellular activation of Akt¹³³. Genetic modification of MSCs, although tremendously interesting, will have greater challenges in reaching the clinic due to ethical and safety concerns.

Preclinical Trials of MSC Therapy

In a pioneering study by Toma et al⁷¹, human MSCs were injected in murine hearts and were shown to adopt cardiac fate. The immunoprivileged capabilities of MSCs were shown by Amado et al⁶⁴, where allogeneic MSCs were transplanted into 3 day old immunocompetent porcine infarcted hearts that resulted in long term engraftment and a large decrease of scar tissue without any inflammatory response.

Subsequent to those early groundbreaking studies, MSCs have been tested in numerous cardiovascular settings. First, in acute myocardial infarction where the inflammatory microenvironment and the necrotic/apoptotic signals are the dominant opposing forces to the therapeutic activities of MSCs. However, the presence of homing signals and the antifibrotic milieu may be of benefit. Another study¹³⁴ where MSCs were intracoronarily infused into porcine heart 5 days after infarction, showed improvement in EF and scar reduction in MSC treated animals. In another study¹³⁵, where different doses of MSCs were used three days after MI, there was no dose dependent effect on scar size reduction or EF improvement 12 weeks post transplantation. In contrast, in a study¹³⁶ where four different doses of BM derived STRO-3+ MSCs were directly injected into sheep hearts 1 hour post MI, there were improvements in EDV only in the two lower doses, although EF increased universally. These seemingly contradicting reports suggest that there might be a therapeutic threshold in the total number of cells that can be injected, although other factors certainly play a role. A preclinical study¹³⁷ where Stro-3+ MPCs were intracoronarily infused in sheep's hearts immediately after MI showed a 40% decrease of infarct size, attenuation of remodeling, and a 50% increase of blood vessel density in border and remote zones.

In the chronic setting the repair processes have been completed, the scar has stabilized and the newly formed network of blood vessels is disorganized and inadequate to halt any disease deterioration. Silva and colleagues⁷⁴ injected MSCs into canine hearts that resulted in increase in EF, vascular density and decrease in scar tissue. Schuleri and colleagues⁶² in a similar setting reported that autologous MSCs produce reverse remodeling and functional recovery. Our group⁷³, has transendocardially injected allogeneic MSCs in swine hearts 3 months post infarction. Although adverse remodeling is nearly complete in that setting, we have observed significant decrease of scar tissue and improved contractility in MSC treated animals, three months post transplantation. Left ventricular global function was also improved thus indicating that MSCs may be able to reverse chronic adverse remodeling.

The intermediate setting between chronic and acute, the subacute model also presents with its own challenges. Although, there are fewer preclinical studies^{64, 138} in this setting, the accumulating data appear consistent with a benefit of MSCs. Our group⁶⁴ administered allogeneic MSCs to the border and infarct zones of infarcted porcine myocardium 3 days after MI via intramyocardial injections. Eight weeks after transplantation there was a 50% reduction in scar size that was coupled with improvements in EF, LV end diastolic pressure, relaxation time and systolic compliance in the treated animals. Another group¹³⁸ directly injected autologous MSCs into 2 week old swine infarcts. Four weeks after transplantation, a trend towards improved wall thickness and systolic thickening was observed in the MSC treated animals. Supplementary table I summarizes the preclinical trials described here.

Non-ischemic heart disease models

Ohnishi and colleagues¹³⁹ in a rat model of non-ischemic dilated cardiomyopathy reported that MSCs attenuate myocardial injury and ventricular dysfunction. In canines, Plotnikov and colleagues¹⁴⁰ suggest that MSCs may provide a platform for sustained biological pacemaker function. That experiment was rather provocative because whether or not cell therapy exerts pro- or anti-arrhythmic effects is crucial to address. These early concerns though have been largely entertained especially after the encouraging results from clinical trials supported the anti-arrhythmic effects⁶⁸.

Autologous MSCs were even used to recellularize tissue-engineered heart porcine valves that were then transplanted to lambs¹⁴¹. Four months later, the MSC valves exhibited better transvalvular and distal gradients as well as less inflammatory reaction and structural deterioration than the BMMNC counterparts.

Safety of MSC Therapy

Although the vast amount of preclinical and clinical data suggest that MSC possess a great potential in treating cardiac diseases and in spite of all the evidence for the safety of MSC transplantation¹⁴², it is important to point out that very long term safety requires additional study. There are two main theoretic concerns regarding MSC therapy, arrhythmogenicity and tumorogenicity. The former, as discussed earlier has been proven not to be an issue in all the clinical trials that have been completed. Miura and colleagues¹⁴³ showed that there is accumulating chromosomal instability in murine BMMSCs that may lead to malignant transformation. Although seen in rodent models, this observation has yet to be described in numerous preclinical and clinical studies^{5, 62-64, 66, 67, 74, 136}.

Clinical Trials of MSC Therapy for Cardiac Repair

Acute Myocardial Infarction

Several studies have examined both autologous and allogeneic MSCs for acute MI. In a phase I randomized study⁶⁸, 53 patients were randomized to receive either allogeneic MSCs or placebo 7 to 10 days after MI and in different doses. The intravenous infusion of allogeneic MSCs in this study resulted in improvement in overall clinical status 6 months after infusion, fewer arrhythmic events, and improved EF. The success of this pilot study led to a phase II trial where as it was preliminary reported by Osiris, the trial sponsor, that intravenous infusion of allogeneic MSCs within 7 days of an acute MI resulted in significantly reduced cardiac hypertrophy, stress-induced ventricular arrhythmia, heart failure, and rehospitalizations for cardiac complications¹⁴⁴. Chen and colleagues¹⁴⁵, administered autologous MSCs intracoronarily in patients with subacute MI and observed decreased perfusion defect, improved LVEF and LV remodeling 3 months after therapy. In CArdiosphere-Derived aUtologous stem CElls to reverse ventricUlar dySfunction (CADUCEUS) trial²⁴ where autologous cardiospheres (that included MSC-like cardiac cells) were injected 2-4 weeks after MI, there were scar reduction, increase in viable heart mass, and regional contractility. However, changes in end-diastolic volume, end-systolic volume, and LVEF did not differ between groups by 6 months.

In addition to BMMSCs, adipose derived MSCs have also been tested for acute MI in the APOLLO trial¹⁴⁶, a trial of 14 patients, which exhibited the safety of intracoronary infusion of freshly isolated adipose derived MSCs in the acute setting of an STEMI. This study also demonstrated a trend toward improved cardiac function, accompanied by a significant improvement of the perfusion defect and a 50% reduction of myocardial scar formation (Figure 5). Thus, these studies suggest a potential role for MSCs in the setting of acute MI.

Figure 5. Adipose Derived MSCs (PRECISE and APOLLO trials) and autologous CSC (C-CURE) in clinical trials.

(A) In the PRECISE trial, Adipose Derived Regenerative Cells (ADRCs) failed to reduce scar size in patients suffering from HF, 6 months after implantation. (B) In the APOLLO trial, where similar cells were employed in the setting of an acute myocardial infarction, there was a significant scar size reduction. (C and D) In the C-CURE trial, where autologous cardiopoietic stem cells (CSCs) were employed, LVEF increased significantly compared to baseline and to control group 6-months following treatment. That was coupled with decreases in left ventricular end-diastolic (ED) and end-systolic (ES) volumes. Panel A is reproduced with permission from Perin et al. American Heart Journal, 2014¹⁴⁷, panel B from Houtgraaf et al, Journal of American College of Cardiology, 2012¹⁴⁶ and panels C and D from Bartunek et al, JACC, 2013¹⁴⁹.

Chronic Ischemic Cardiomyopathy

Currently, chronic ischemic cardiomyopathy is the setting in which there is the highest level of consensus for MSC efficacy. There are 6 published or ongoing clinical trials that together show that MSCs have anti-fibrotic effects, induce neoangiogenesis, enhance contractility and improve the quality of life of the recipient patients. Thus, the mechanisms of actions identified preclinically appear to be operative in humans with chronic ischemic cardiomyopathy. Perhaps the most consistent finding is that of a 30-50% decrease in the size of the MI scar in BMMSC trials^{5, 66, 67} (Figure 6). Importantly, a decrease in MI size is not evident in the one published trial of adipose MSCs for ischemic cardiomyopathy (Figure 5). Currently, no clinical study has directly compared head-to-head BMMSCs with ADMSCs.

Figure 6. Bone marrow derived MSCs in clinical trials (POSEIDON, TAC-HFT, PROMETHEUS).

(A) In the POSEIDON trial, autologous and allogeneic MSCs both decreased the scar size (as measured by MDCT) and (B) the end-systolic volume 13 months after transplantation. (C) In the TAC-HFT trial, twelve months after injection, scar mass reduced as the percentage of left ventricular mass for patients treated with mesenchymal stem cells (MSCs) and those in the placebo group. (D) In the PROMETHEUS trial, at 18 months after injection of autologous MSCs in patients undergoing CABG, scar tissue, perfusion, wall thickness and thickening,andsystolic strain improved preferentially in the MSC injected segments. There was a progressive drop-off in regional phenotypic improvement as distance from injection site increased. Panels A and B reproduced with permission from Telukuntla et al, Journal of American Heart Association, 2013¹⁴⁴. Panel C reproduced with permission from Heldman et al. Journal of American Medical Association, 2014⁶⁷. Panel D reproduced with permission from Karantalis et al, Circulation Research, 2014⁶⁶

Importantly scar reduction is shown to be accompanied by reperfusion and restoration of contractile performance. Perfusion was also preserved or increased in the PRECISE¹⁴⁷ (A Randomized Clinical Trial of Adipose Derived Stem and Regenerative Cells In the Treatment of Patients With Non Revascularizable Ischemic Myocardium) and in the PROMETHEUS trials⁶⁶. These effects also led to a restoration of contractile function as measured by echocardiography¹⁴⁷ or Eulerian circumferential strain using MRI at the site of injection^{66, 67}. The last is also an interesting point and highlights the importance of imaging studies in the assessment of stem cell efficacy. In two studies from our group, a substudy of POSEIDON¹⁴⁸ and the PROMETHEUS trial⁶⁶ it was shown that MSC therapy exerts its strongest impact at the site of injection and there is a drop off effect in the adjacent myocardial segments that drive the global functional improvement (Figure 6). Maybe one of the most interesting findings that the clinical trials have offered is the improvement in functional capacity and quality of life these patients experience after cell therapy^{5, 67, 149}

There is also an ongoing clinical trial employing adipose derived MSCs, the MesenchYmal STROMAL CELLTherapy in Patients With Chronic Myocardial Ischemia (MyStromalCell)¹⁵⁰, that is using culture-expanded adipose tissue-derived MSCs, and is designed to investigate the safety and efficacy of intramyocardial delivery of VEGF-A165 stimulated autologous adipose tissue-derived MSCs to improve myocardial perfusion and exercise capacity and reduce symptoms in patients with chronic ischemic cardiomyopathy.

In a recently published clinical trial⁵⁹ where allogeneic MPCs were injected in the hearts of patients that were undergoing left ventricular assist device (LVAD) implantation, 50% of the patients were successfully temporarily weaned from LVAD at 90 days post transplantation. Although this trial was small and exploratory it demonstrated the safety of MPC implantation. The efficacy end points though were derived only from a comparison of 10 patients in the MPC group and only two in the control group. In addition, an ongoing clinical trial, the Allogeneic Mesenchymal precursor cell Infusion in myoCardial Infarction (AMICI) trial, aims to prove safety, feasibility, and efficacy of MPC therapy in the acute ST-elevation MI (NCT01781390). Supplementary tables II-V summarize important completed and ongoing clinical trials.

Phase III Clinical trials

There are currently two phase III clinical trials employing MSCs. the C-CURE trial¹⁴⁹ where MSCs were treated ex-vivo with cytokines in order to enhance their commitment to cardiopoietic lineage (Figure 5). Bartunek and colleagues reported significant improvements in EF, end systolic volume and walked distance compared to controls. Another phase III study is currently underway aiming to enroll 1730 patients with CHF, in which MPCs will be injected intramyocardially (NCT02032004).

Non Ischemic Cardiomypoathy

The PercutaneOus StEm Cell Injection Delivery Effects On Neomyogenesis in Dilated CardioMyopathy (The POSEIDON-DCM; NCT01392625) ongoing clinical trial¹⁵¹ is testing transendocardial injections of autologous and allogeneic MSCs in a non-ischemic setting of dilated cardiomyopathy.

Demographic and Biologic Factors affecting MSC therapy efficacy

When the results of preclinical studies are compared to those of clinical trials there is a relative decrease of the impact seen in the former. Simply the fact that the animals used in preclinical trials are younger, healthier and followed up in a closed and controlled environment compared to humans enrolled in clinical trials, could account for the best part of this discrepancy in translation. But there are many lessons to be drawn from this phenomenon that will lead to further optimizing the cell therapy itself but also detecting the patient population that will benefit the most from it.

The cell products produced in different laboratories as well as the treated population are very heterogeneous, heavily influencing the clinical outcome. Timing, delivery method, and optimal dose are still open to debate although already there are hints for the existence of time windows and maximal and minimal effective doses.

Sethe and colleagues¹⁵² reported that the proportion of MSCs in the BM decreases with age and the MSC yield is lower in direct association with the age of the donor. Although the inherent ability to differentiate in different cell lineages seems to be preserved in older MSCs there are quantitative differences between older and younger cells¹⁵³. Zhang and colleagues¹⁵⁴ in a rat model found that increasing donor age adversely impacts the beneficial effect of MSCs. Khan et al¹⁵⁵ suggested that only MSCs derived from young and healthy donors can effectively regenerate senescent rat hearts. Understanding the effect of aging on MSCs is crucial for autologous therapy for older patients, who are typically afflicted by cardiovascular diseases. In a recently published sub-study¹⁵⁶ from the TAC-HFT and POSEIDON trials, it was reported that older individuals did not have an impaired response to MSC therapy.

Gender may influence the biology of MSCs. Some studies suggest that females have a lower reservoir of MSCs but compared to male MSCs exhibit greater resistance to injury¹⁵⁷. But different conditions in the patients themselves may play a role. Less injury and inflammation is found in female patients compared to males after acute cardiac injuries. Estradiol contributes by decreasing the expression of the Bcl-2 family of genes and JNK pathways¹⁵⁸.

It is also important to understand whether and what differences exist between MSCs obtained from normal and diseased individuals. Zhao and colleagues¹⁵⁹ described the characteristics of MSCs derived from various malignant hematopoietic diseases. The morphology and phenotype of MSCs remained unchanged even after 20 passages. Patients with rheumatoid arthritis on the other hand, although having normal reserves of MSCs they seem to display decreased proliferative and clonogenic potential compared to normal¹⁶⁰. Some studies suggest that in diabetics, MSCs become exhausted and lose their differentiation potential. These changes appear to be due to induction of apoptosis and senescence by advanced glycation end-products¹⁶¹. Interestingly the changes in MSCs caused by the diabetic environment are probably permanent, because culturing MSCs in medium without glucose does not reverse the effect. Another risk factor for cardiovascular disease, cigarette smoking, seems that adversely affects MSCs. A study demonstrated that nicotine decreased the proliferation of MSCs in a dose dependent manner¹⁶².

The Future of MSC therapy

Indications, timing, delivery of MSCs

MSCs have been used to treat pediatric cardiomyopathy¹⁶³, congenital heart diseases (hypoplastic left heart syndrome)¹⁶⁴, refractory angina¹⁶⁵, acute MI^{68, 145, 146, 166}, and chronic ischemic cardiomyopathy^{5, 66, 67, 147, 150}, are being investigated in non-ischemic dilated cardiomyopathy¹⁵¹, and have been proposed for doxorubicin induced cardiomyopathy¹⁶⁷. In the setting of refractory angina and in acute MI, the chemokines, cytokines, and growth factors released by the injured myocardium provide migratory cues for endogenous resident stem cells as well as bone marrow derived stem cells¹⁶⁸; hence, the rationale behind locally (intracoronary) and systemically (intravenous) administered stem cells, respectively, in these settings. However, both delivery methods have limitations. After intravenous infusion, there is wide distribution of MSCs throughout the body, with entrapment of these cells in the lungs, spleen, and liver^169-171, while the intracoronary technique requires a transient ischemic period through the inflation of a balloon to give the cells the chance to be distributed and not washed out. In addition, there is the potential for microvascular obstruction by the infused cells, which can result in myocardial necrosis. The intracoronary delivery technique is also limited by the inaccessibility of some myocardial distributions in patients with advanced coronary artery disease. In acute MI, intramyocardial injections have not been a preferred mode of delivery due to concerns about perforation in the setting of acute ischemia and necrosis. However, Perin et al¹⁷² have shown that transendocardial delivery was more efficacious than intracoronary delivery in a canine model of acute MI.

On the other hand, in the chronic setting of heart failure due to chronic ischemic and nonischemic heart disease, the migratory cues for stem cells are thought to be absent or minimal¹⁶⁸ and a more targeted approach, namely intramyocardial injection, has been investigated. In an analysis from the POSEIDON¹⁴⁸ clinical trial, transendocardial injection of MSCs reduced scar size in both injected and non-injected myocardial segments but segmental contractility improved only in the injected scar segments and was greatest in those territories with severe baseline dysfunction. Moreover, analysis of the effects of direct surgical myocardial injection in the PROMETHEUS⁶⁶ clinical trial found a concordant reduction in scar size and improvement in perfusion and contractility in MSC-injected myocardial segments when compared to “revascularization only” and non-MSC treated segments in patients who underwent coronary artery bypass graft (CABG) surgery. In light of the PROMETHEUS trial⁶⁶ and POSEIDON substudy¹⁴⁸ findings, special attention should be invested on designing the strategy of the precise placement of the intramyocardial injections in order to cover as much of the impaired myocardium as possible. When there is no ischemia, as in the setting of non-ischemic dilated cardiomyopathy, this strategy makes even more sense, and effort should be made to inject as much area of the left ventricle as possible.

There is evidence⁵ that autologous and allogeneic sources of MSCs produce similar beneficial effects in patients with ischemic cardiomyopathy and this is currently being investigated in patients with non-ischemic cardiomyopathy as well¹⁵¹. Given these findings, the allogeneic source provides an “off-the-shelf” therapeutic option that is advantageous over the autologous source, especially in the urgent setting of an acute MI. Although there are clinical trials (TIME¹⁷³, LateTIME¹⁷⁴) that were designed to detect the optimal window of delivery of bone marrow mononuclear stem cells after an acute MI, the effect of timing of MSC delivery on therapeutic efficacy has not been specifically investigated. However, MSCs have been employed in the chronic setting of ischemic cardiomyopathy in the PROMETHEUS, POSEIDON, and TAC-HFT studies⁶⁷ and have been shown to produce favorable outcomes.

Future directions

There is now a large body of evidence from preclinical^{62, 63} and early phase clinical trials^{5, 67} that elucidates the mechanism of action and detects potential pitfalls of MSC therapy. It is also largely accepted that one of the main limitations to the long term efficacy of MSC therapy is the low retention and survival of the transplanted cells⁷². Genetic engineering and cell pretreatment may in the near future provide the appropriate lifespan to the transplanted MSCs and thus render them the new nearly permanent regenerative niche in the impaired myocardium⁷⁹ (Figure 7).

Figure 7. The evolution of MSC therapy for cardiac disease.

MSCs will play the central role in combination therapies supporting and orchestrating different cell types. Cytokines may also be given with or pretreat the target organ before MSC transplantation. A more challenging but promising way is the genetic modification of MSCs.

One of the primary advantages of the MSC is their ability to home to sites of injury, respond dynamically to the extent of injury and secrete a broad range of factors, many of whom not yet discovered⁸¹. Many research groups around the world are now moving towards enhancing the effectiveness of MSC therapy via various pretreatment cocktails of factors, although this has yet to be translated clinically¹³². The employment of sophisticated imaging like CT¹⁴⁸ and MRI⁶⁶ and the advanced understanding those two modalities offer us on the regional effect of MSC therapy, will also play a big role on accurately detecting target areas for therapy. Additionally and in conjunction to the latter, there is significant potential for improvement in design of cell delivery methods.

It is also becoming increasingly evident, that although MSC remains one of the promising cell types for a successful cell therapy, complete myocardial regeneration cannot be meaningfully achieved by just one cell type⁹⁹. The future of MSC therapy may lie in being the main supporting, trophic and orchestrating cell type in a cell therapy that will combine different cell types taking advantage of their unique characteristics and benefits.

We now have enough evidence for the safety and encouraging signs of efficacy of MSC therapy. Currently, there are ongoing phase 3 clinical trials testing whether various MSC-based strategies offer clinical benefits to patients. These trials based upon substantial basic, preclinical, and phase I/II data will take the next step in addressing the appropriate use of cell-based therapy in the armamentarium of approaches for chronic heart disease.

Non-standard Abbreviations and Acronyms

MSC

Mesenchymal Stem Cell

MNC

Mononuclear Cell

AFMSC

Amniotic Fluid Derived Mesenchymal Stem Cell

MPC

Mesenchymal Precursor Cell

CFU

Colony Forming Unit

VEGF

Vascular Endothelial Growth Factor

Miro-1

Mitochondrial Rho-GTPase

TNT

Tunneling Nanotubes

EV

Extracellular Vesicles

miRNA

microRNA