Regenerative Cell-Based Therapy for the Treatment of Cardiac Disease

Unfractionated Bone Marrow Cells

The capacity for unselected or unfractionated BMCs to promote cardiac repair has been controversial, yet they remain the most thoroughly investigated cell type in clinical trial. Their popularity is driven largely by practical reasons. For example, BMCs are (1) derived from straightforward harvest protocols, (2) do not require complex or prolonged ex vivo management, and (3) contain a small niche of stem and progenitor cells despite representing a heterogeneous population. BMCs have been investigated in AMI, chronic ICM, and HF, but for the sake of brevity the following section will focus on trials of AMI. Readers interested in the results of BMC therapy in other clinical settings can consult several excellent reviews.

The safety and feasibility of BMC therapy was established from early, small trials of AMI such as the “Transplantation of Progenitor Cells and Regeneration Enhancement in Acute Myocardial Infarction” (TOPCARE-AMI). The demonstration that cell therapy may be safe, and possibly effective, prompted further inquiry into mid-sized, double-blind, randomized trials such as the “Reinfusion of Enriched Progenitor Cells and Infarct Remodelling in Acute Myocardial Infarction” (REPAIR-AMI) and the “BOne marrow transfer to enhance ST-elevation infarct regeneration” (BOOST) trials. Initial follow-up analysis from both the REPAIR-AMI and BOOST trials reported benefits associated with BMC therapy. BMCs were infused 3-7 days after successful reperfusion and were associated with superior cardiac function (5.5%-6.7% improved left ventricular ejection fraction [LVEF]) compared with placebo at 6 months postintervention. In addition, cell therapy was not associated with an increased risk for adverse clinical events. Substudies of both trials were subsequently released by trial investigators with equally promising results. For example, patients receiving BMCs in BOOST demonstrated improvements in diastolic function, whereas there was a reduction in a prespecified cumulative end-point of death, MI, or necessity for revascularization at 1-year follow-up of the REPAIR-AMI. Several other trials also support the benefit of BMC therapy in patients with AMI. Trials by Assmus and colleagues, Cao and colleagues, Yousef and colleagues, and the “FINish stem CELL study” (FINCELL) reported positive benefits associated with BMC therapy including improved global ventricular function, a reduction in morbidity and mortality, and a neutral effect on severe adverse reactions, including arrhythmias and restenosis of the stented coronary artery.

Unfortunately, despite the positive early experience with BMCs in patients, other trials have not been so promising. Moreover, follow-up from some of the earlier, positive trials were beginning to release disappointing long-term results. For example, the 2-year results of the REPAIR-AMI could not conclusively associate BMC therapy with improved morbidity and mortality ; in BOOST, whereas the difference in LVEF improvement between cell therapy and placebo were significant at 6-month follow-up, the effect was not sustained at 18 months. A subgroup of patients with pronounced transmural infarcts may have benefited from cell therapy, although this was required to be tested prospectively in a randomized trial. Regardless, the lack of beneficial effects associated with BMCs in BOOST was sustained in the 5-year results. Examples of results from recent trials include the “BONe marrow in Acute Myocardial Infarction” (BONAMI), which have been mixed, and those from the “Autologous Stem-Cell Transplantation in Acute Myocardial Infarction” (ASTAMI) trial, which were altogether negative. Aside from a small improvement in exercise time, ASTAMI did not associate BMCs with improvements in global LV function, or ventricular remodelling.

Before reaching clinical trial, many elements such as cell dose, frequency, or time of delivery had not been systematically determined for stem cell therapies. In some trials, such as the REPAIR-AMI, only until patients were stratified based on time of delivery after reperfused MI did some suggestion arise that timing of delivery influenced the efficacy of the therapy. This prespecified subgroup analysis associated later cell delivery (about 7 days after reperfused MI) with more pronounced benefits. Thus, several recent trials including the TIME, late-TIME, and SWISS-AMI sought to test this hypothesis comparing the time of delivery between cells delivered early, mid, or late postinfarction. Unfortunately, results from all three trials did not associate beneficial effects with BMC therapy for AMI, consistent with the BOOST and ASTAMI trials, regardless of when cells were delivered after AMI. Results from the TIME trial may, however, require careful scrutiny as cells were procured from a novel, convenient extraction method that had not been previously validated in bioactivity assays or animals models as opposed to the already well-supported and published methods for isolation.

Establishing a concrete position on the efficacy of BMC therapy for AMI has so far been difficult and results from recent trials such as TIME, late-TIME, or SWISS-AMI have not been able to answer this question conclusively. The outlook for BMC therapy is less than optimistic. Given the conflicting evidence and emerging alternatives, favor may lean toward products with well-demonstrated capacity to produce improvements in global LV function, mortality, and quality of life. Moreover, although some cumulative evidence from recent meta-analyses suggests that BMCs are associated with modest improvements in LV function (2%-4% LVEF), the question remains of whether these improvements are clinically meaningful. The upcoming European large, phase 3 “Effect of Intracoronary Reinfusion of Bone Marrow–derived Mononuclear Cells on All-Cause Mortality in Acute Myocardial Infarction” (BAMI) trial ( NCT01569178 ) will probably represent the final consensus of BMCs in cardiac cell therapy in AMI.

Hematopoietic Stem Cells

Hematopoietic stem cells (HSCs) are a subset of BMCs that are multipotent and self-renewing and from which all blood cell types are derived. They are possibly the first stem cell therapeutic investigated in patients dating back from early trials of bone marrow transplantation for fatal leukemias. Their use can be potentially curative for genetic blood disorders such as thalassemia and immune deficiency, or malignancies such as leukemia. The prevailing hypothesis is that HSCs give rise to two lineage restricted progenitor cell types, including myeloid progenitors and lymphoid progenitors that, in turn, supply the number of lymphocytic and granulocytic cells, macrophages, platelets, and erythrocytes found in the blood. However, specific identification and classification of HSCs has been difficult largely because of the limitations of current imaging and lineage tracing methods. Generally, HSCs are identified by a lack of hematopoietic or myeloid lineage (lin) markers and, depending on the report, the presence of a number of antigenic surface markers such as CD34, CD38, CD90, CD133, CD105, CD45, and c-kit (CD117) or other markers such as the presence of aldehyde dehydrogenase. Their use as a cardiac regenerative cell product was stimulated by early studies indicating that lin ^– , c-kit ⁺ HSCs could restore necrotic myocardium by transdifferentiating into all the major cardiac cell types (cardiomyocyte, endothelial cell, and smooth muscle cell) conferring salutary effects. Unfortunately, this notion was met with sharp criticism and conflicting data. Subsequent evidence instead indicates that HSCs are, in fact, not cardiomyogenic but adopt mature hematopoietic fates in ischemic myocardium. Whether the benefits of HSCs are derived from an ability to restore damaged myocardium or from a separate, possibly paracrinic, mechanism remains a topic of intense debate.

Consensus on what combination of surface, cytosolic, or functional markers define the HSC is lacking, thus making direct comparisons between trials difficult. Given this loose definition of HSCs, some of the trials discussed in the following section might apply equally to other BMC products, and vice versa. The MAGIC trial assessed the feasibility and efficacy of HSC infusion in patients with MI and who underwent coronary stenting. Granulocyte colony-stimulating factor (G-CSF), a well-established stem-cell mobilizer, was administered to mobilize bone marrow–derived HSCs into the peripheral blood 4 days before apheresis and cell procurement. This trial involved three separate arms including the HSC group, a control group (not placebo), and patients receiving G-CSF alone. At 6 months after infusion, patients treated with HSCs showed significant improvements in exercise capacity (450 vs. 578 seconds), myocardial perfusion (perfusion defect, 11.6% vs. 5.3%), and systolic function (LVEF, 48.7% vs. 55.1%). Although no adverse events occurred because of cell infusion combined with G-CSF, enrollment of the G-CSF cohort was eventually discontinued because of high rates of in-stent restenosis, highlighting a potential safety concern and cautioning future trials using G-CSF administration as a strategy for endogenous stem cell mobilization. Recent trials, though, such as Perin and colleagues have not reported any significant adverse cardiovascular or cerebrovascular events associated with HSC infusion in similarly high-risk patient cohorts with advanced ischemic heart failure and CAD ineligible for PCI or surgical revascularization. The difference, perhaps, is that Perin and colleagues elected to aspirate bone marrow from the iliac crest as opposed to administering G-CSF. In their trial, HSCs were then purified based on the presence of cytosolic ADH, constituting a population of primitive Lin ^– CD34 ⁺ CD38 ^– cells. When administered to patients with advanced ischemic heart failure ADH ⁺ HSCs were associated with a significant decrease in left-ventricular end-systolic volume and a trend toward maximal oxygen consumption at 6 months. It is difficult to draw meaningful conclusions given that only 20 patients were enrolled in the study. Overall, HSCs appear to be beneficial for the treatment of ischemic cardiomyopathies, but larger, randomized, controlled trials are required to confirm early results.

Endothelial Progenitor Cells

Before the characterization of a circulating EPC in 1997 by Asahara and colleagues, vasculogenesis, the process of de novo blood vessel development, was believed to be limited to the embryonic development of vasculature from endothelial cell progenitors or angioblasts. Now it is understood that a subset of bone marrow cells, possibly the putative EPC, mobilizes from the bone marrow to the peripheral circulation ultimately reaching sites of tissue injury or ischemia to participate in postnatal vasculogenesis. Circulating EPC levels have also been recognized as possibly providing clinical information on atherosclerotic burden and future cardiovascular risk. Despite the significant attention that has been devoted to studying EPC biology, consensus on an unambiguous definition of the EPC is still lacking and, likewise, the success of many studies has been hindered by the absence of such a definition. EPCs were first characterized as a population of cells that were CD34 ⁺ , VEGFR2 ⁺ , and adherent to fibronectin ( Fig. 101-3 A ). Today, they are either broadly identified as the mononuclear cells adherent to fibronectin or by cell surface expression that may or may not include CD31, FLK-1 (VEGFR2), CD34, CD133 (AC133), c-kit, and other endothelial markers including VE-cadherin. The criteria that distinguish an HSC from an EPC, and vice versa, may seem unclear given they seem to share many of the same surface antigenic markers. As a cell therapy product, however, EPCs are distinguished from HSCs based on specific ex vivo culture requirements. Separate techniques exist to produce what can be referred to as the early or late EPC . Both early and late EPCs first require isolating the mononuclear cell fraction from the peripheral circulation, and early EPCs are generated through subsequent culture on fibronectin for 3 to 7 days, whereas late EPCs are produced via culture for greater than 14 days on collagen type I. For a more detailed discussion on the characterization of EPC phenotypes or potential diagnostic use of circulating EPCs, interested readers are referred to an excellent review by Fadini and colleagues.

In preclinical animal models of ischemic cardiomyopathy EPC transplantation prevented cardiac myocyte apoptosis, resulted in increased blood vessel formation, decreased infarct size, and improved function. The capacity for human EPCs to transdifferentiate into cardiomyocytes has also been reported, but not all studies support this conclusion. Instead, recent evidence suggests that human early EPCs may be proangiogenic, enacting their salubrious properties through paracrine mechanisms. The proangiogenic effects of late EPCs, instead, is thought to be via direct incorporation into the vasculature.

Considering the attention EPCs have received in preclinical animal models of myocardial ischemia, there is comparatively little experience with their use in patients. This is likely owing to pragmatic considerations and investigator preference for unfractionated or sorted BMC products neither requiring ex vivo culture. EPC cell therapy has been explored in several settings, including AMI and chronic ischemic heart disease. As a therapy delivered shortly after reperfused AMI, the small TOPCARE-AMI trial reported that EPC transplant was safe, feasible, and potentially beneficial. EPCs were infused approximately 4 to 5 days after MI and were associated with improved LVEF (51.6 ± 9.6% to 60.1 ± 8.6%), improved regional wall motion in the infarct zone, and profound reduction in end-systolic left-ventricular volumes at 4-month follow-up, with benefits persisting in the final 1-year follow-up. Indeed, contrasting results were released by the same group in a similar trial assessing EPC cell therapy for the treatment of chronic ischemic heart disease. The randomized, controlled TOPCARE-CHD trial assessed the efficacy of BMC, EPC, or placebo treatment later after MI in patients who had a myocardial infarction at least 3 months before enrollment and a well-demarcated region of left ventricular dysfunction. Cells were infused into the vessel supplying the most dyskinetic area of the ventricle, and the results of this study contradict those of TOPCARE-AMI, as EPC delivery was significantly inferior to BMC therapy in terms of improving global left ventricular function. The absolute improvement witnessed with BMC therapy, similar to their experience in AMI, was modest. BMC therapy was associated with an increase in approximately 2.9 percentage points according to left ventricular angiography and 4.8 percentage points according to MRI. The precise mechanisms underlying the benefit of BMC therapy were not investigated in the trial, and it remains unclear as to why EPCs proved inferior in this setting. A higher number of progenitors isolated from the bone marrow aspiration may account for this discrepancy. Moreover, given that patients with CHD have been shown to have pronounced EPC dysfunction, limited cell recruitment and action at the site of injury may have also contributed.

Fractionated CD34 ⁺ Bone Marrow Cells

As an alternative to heterogeneous cell products, in recent years focus has shifted to BMCs isolated based on CD34 expression given their potent proangiogenic properties. CD34 is believed to be a cell-to-cell adhesion factor expressed in some cell types, including hematopoietic cells, endothelial cells, muscle satellite cells, hair follicle stem cells, and fibrocytes. The function of CD34 remains largely unknown, although in HSCs some evidence suggests that CD34 alters engraftment potential and other properties, such as cell adhesion and migration. Peripheral blood CD34 ⁺ cells have been shown to transdifferentiate into most of the major cardiac lineages, including endothelial cells, vascular smooth muscle cells, and cardiomyocytes, principally via cell fusion. CD34 ⁺ cells purified from peripheral blood BMCs have been shown to exhibit superior efficacy and safety for therapeutic neovascularization after MI compared with total mononuclear cells in rodent studies. They have also been shown to exhibit superior efficacy for preserving integrity and function following MI compared with unselected mononuclear cells with an equivalent number of CD34 ⁺ cells, suggesting that a purified CD34 ⁺ cell product might possess greater therapeutic potential than a mixed mononuclear cell population.

The REGENT trial investigated the therapeutic potential of fractionated CD34 ⁺ mononuclear cells for the treatment of significant systolic dysfunction (LVEF < 37%) secondary to reperfused anterior MI. Two hundred patients were randomized to intracoronary infusion of CD34 ⁺ CXCR4 ⁺ purified cells, total mononuclear cells, or a control group receiving no therapy at all (no placebo) and were followed for 6 months. Similar to REPAIR-MI, both the purified CD34 ⁺ CXCR4 ⁺ and BMC therapies were associated with a trend in modest improvement (≈3%) in systolic function. Unfortunately, the study was plagued by a high dropout rate in the control cohort, probably because of the open-label design, which may have potentially accounted for some results not reaching statistical significance. Nonetheless, for patients who received CD34 ⁺ cell therapy, a clear trend for better outcomes was evident in patients with a baseline ejection fraction (EF) below the median (<37%) compared to those whose baseline EF was above the median. In addition, a low baseline EF (<37%) was an independent predictor for pronounced benefits in terms of cardiac function. This pattern is consistent with the results reported in REPAIR-AMI. The authors also highlighted that the use of a relatively small number of CS34 ⁺ CXCR4 ⁺ cells was associated with a similar trend in functional improvement as the use of 100 times more total mononuclear cells, similar to the results published by Kawamoto and colleagues in rats. The mechanisms that could account for these benefits were not explored, but it is unlikely that it was a result of de novo cardiomyocytes. Instead, recent laboratory evidence indicates that CD34 ⁺ cells secrete exosomes that have independent angiogenic activity, potentially accounting for a significant component of the paracrinic effect of these cells.

Indeed, the hallmark of ischemic cardiomyopathies, such as myocardial infarction, is the loss of myocardium. In this scenario, cell therapy may be a futile endeavor if no meaningful restoration of myocardium can be accomplished. Moreover, the data supporting the cardiomyogenic potential of CD34 ⁺ purified cells is less than convincing. However, in situations in which instead neovascularization may be preferred, CD34 ⁺ cells could offer some benefit. Early evidence for the feasibility, safety, and benefit of purified CD34 ⁺ cells mobilized by G-CSF administration in 24 patients with refractory angina was provided by Losordo and colleagues in a pilot phase I/IIa randomized trial. Follow-up was performed in a larger double-blind, randomized phase II trial in 167 patients. CD34 ⁺ purified cells, or an equal volume of diluent (placebo), were injected with an electromechanical (NOGA) mapping injection catheter. Low-dose CD34 ⁺ therapy improved both weekly angina frequency and exercise tolerance measures. Mortality was not different between control and cell therapy cohorts at 12 months, although cell mobilization and collection procedures were associated with elevations in cardiac enzymes that will require follow-up in future studies.

CD34 ⁺ purified cells might represent a natural progression from unfractionated BMC therapies, and a more standard product compared with HSC- or EPC-based therapies based on their properties and preclinical results. However, cumulative evidence reviewed in a recent meta-analysis of BMC trials suggests that, overall, CD34 ⁺ purified cell therapy might not offer meaningful benefits in the settings of acute MI or chronic HF, possibly because of a lack of clinically relevant cardiomyogenic potential. Given their potent proangiogenic paracrinic properties, CD34 ⁺ cells could offer potential in settings where neovascularization in itself could offer substantial benefits, such as refractory angina or chronic ischemic cardiomyopathy.

Mesenchymal Stem Cells

Mesenchymal stem cells are stromal cells found within the bone marrow or adipose tissue and can be expanded via ex vivo culture (see Fig. 101-3 B ). Similar to other bone marrow cell types, they are loosely defined based on surface expression and culture characteristics which include adherence to plastic in cell culture and the surface antigen expression of CD105 ⁺ /CD90 ⁺ /CD73 ⁺ /CD34 ^– /CD45 ^– /CD11b ^– or CD14 ^– /CD19 ^– or CD79 ^– α ^– /HLA ^– DR1 ^– . MSCs have been identified as promising candidates because of their relative ease of procurement, multilineage differentiation potential, robust secretome, antiapoptotic, immunomodulatory properties. They have also been shown to be capable of activating resident cardiac stem cells. Given their immune-privileged status, they may serve as ideal allograft candidates, although the long-term benefits as allogeneic products have been challenged. Some authors contend that MSCs can adopt cardiomyocyte-like phenotypes when injected into both healthy and infarcted myocardium; however, these results have been scrutinized, and others attribute these observations as either MSC fusion with resident cardiomyocytes or nonexistent altogether.

Purified human-derived MSCs have been shown to attenuate contractile dysfunction and to prevent adverse remodeling in rodent and swine models of myocardial infarction. In contrast with the classical method of collecting cells and deploying via infusion or injection, initial clinical attempts at harnessing MSCs assessed pharmacologic mobilization of the cells from the bone marrow using G-CSF. Assessing G-CSF mobilized cells soon after AMI, results from the “G-CSF ST-segment Elevation Myocardial Infarction” (G-CSF STEMI), “STEM cells in Myocardial Infarction” (STEMMI), and “Regenerate Vital Myocardium by Vigorous Activation of Bone Marrow Stem Cells” (REVIVAL-2) trials showed that G-CSF administration did not result in improved systolic function compared with placebo. A substudy of the STEMMI trial sought to describe the cells mobilized by G-CSF administration and to determine the association, if any, between plasma concentration of these cell types and changes in left ventricular systolic function. The authors associated an inverse relationship between circulating MSCs and systolic functional improvement following G-CSF administration after MI. This relationship was left relatively unexplored, but it could be a consequence of several mechanisms. It may be that reduced circulating MSCs is a reflection of increased homing to the infarcted myocardium, or given the large size of MSCs their increased concentration in the circulation may contribute to further microembolization and infarction.

MSCs are one of the few cell types to date that may offer a true “off-the-shelf” product given their potential immune-privileged status. Hare and colleagues assessed the safety of this type of allogeneic product in a double-blind, randomized, dose-ranging trial of 53 patients with reperfused MI. The primary end-point of the study was safety, whereas efficacy measured by ejection fraction and ventricular volumes constituted exploratory, secondary end-points. Safety profiles were similar between the allogeneic MSC product (Prochymal; Osiris Therapeutics Inc., Baltimore, MD) and placebo treatments. Patients receiving MSCs experienced fewer episodes of ventricular tachycardia and improved pulmonary forced expiratory volumes. In terms of efficacy, MSCs were associated with improved global symptom score and ejection fraction. These results highlight some potential advantages of MSCs versus other BMC products. For example, the product used in this trial was a readily available allogeneic product prepared from healthy donors, circumventing any potential issues with patient clinical condition on cell potency. It also obviates the need for an invasive and highly painful bone marrow aspiration. Results from the recent POSEIDON trial reinforce some of these benefits. POSEIDON was a small, randomized, pilot trial that compared the safety and efficacy of autologous or allograft MSC transendocardial delivery in patients with chronic ischemic LV dysfunction secondary to MI. The primary end-point was safety measured by incidence of severe adverse reactions. None of the cell products reached the predefined stopping point for incident severe adverse reactions, supporting the low immunogenicity of MSCs. In terms of secondary end-points, reflecting efficacy, the combined results appear encouraging. Autologous, but not allogeneic, therapy resulted in improvements in the 6-minute walk test and Minnesota Living with Heart Failure Questionnaire (MLHFQ) scores. Moreover, both cohorts demonstrated improvements in New York Heart Association (NYHA) class. Both were associated with reduced early enhancement defect, a measure of infarct size, end-systolic volumes, and sphericity. In addition, patients having received allogeneic therapy demonstrated improved end-diastolic volumes. Changes in EF were not apparent from the aggregate results, although a trend for an inverse dose-response relationship was observed. Increases in EF and pronounced changes in ventricular geometry were evident in patients receiving a lower dose of MSCs. The results are promising but will require substantiation in several follow-up trials given the small sample size, open-label design, and lack of a placebo-controlled group.

MSCs are an exciting candidate as the field of cell therapy progresses. Given their potent paracrinic properties and immune privilege, they may truly offer the potential for an off-the-shelf product. A major knowledge gap that remains before the potential for these cells can be fully realized is the mechanisms underpinning observed benefits. Nevertheless, given their purported benefits and considerable secretome, MSCs appear to be promising as cardiac cell therapy evolves.