Recommendations for nomenclature and definition of cell products intended for human cardiovascular use 

Abstract

Exogenous cell-based therapy has emerged as a promising new strategy to facilitate repair of hearts damaged by acute or chronic injury. However, the field of cell-based therapy is handicapped by the lack of standardized definitions and terminology, making comparisons across studies challenging. Even the term ‘stem cell therapy’ is misleading because only a small percentage of cells derived from adult bone marrow, peripheral blood, or adipose tissue meets the accepted haematopoietic or developmental definition of stem cells. Furthermore, cells (stem or otherwise) are dynamic biological products, meaning that their surface-marker expression, phenotypic and functional characteristics, and the products they secrete in response to their microenvironment can change. It is also important to point out that most surface markers are seldom specific for a cell type. In this article, we discuss the lack of consistency in the descriptive terminology used in cell-based therapies and offer guidelines aimed at standardizing nomenclature and definitions to improve communication among investigators and the general public.

1. Introduction

Studies of cell-based repair of injured cardiac tissue have been ongoing for over 20 years¹; however, clinical translation of this therapy has been slow. The nomenclature of the various cell types identified and used in cell-w3therapy studies suffers from an inconsistent heterogeneity of phenotypic and functional definitions.² This variability makes direct comparisons of studies difficult and leads to potential misinterpretation of results and unnecessary duplication of research.

A rational and universal nomenclature is clearly needed in regenerative medicine because similar cell populations are often referred to by different names (often because of intellectual property claims), and different cell types can be referred to by the same name. For example, bone marrow (BM) stromal or mesenchymal cells have been referred to by using at least 12 different names, including mesenchymal stem cells, mesenchymal stromal cells, multipotent stromal cells, marrow stromal cells, multipotent mesenchymal stromal cells, mesodermal stem cells, medicinal signalling cells, medicinal stem cells, skeletal stem cells, stromal stem cells, mesenchymal progenitor cells, and mesenchymal precursor cells.² Similar cells derived from adipose tissue have been called adipose-derived mesenchymal progenitors, adipose tissue-derived stem cells, adipose-derived repair cells (ADRCs), adipose-tissue mesenchymal stem cells, or adipose-derived stem/stromal cells (ASCs)^3,4—again, names that often are used to differentiate cells based on intellectual property claims. Conversely, cells that are neither endothelial nor progenitor have been referred to as endothelial progenitor cells (EPCs) because they were initially thought to give rise to endothelial cells, whereas cells found within the heart were initially labelled as cardiac stem or progenitor cells but are now considered endothelial precursor cells.⁵

This confusion is further magnified by the use of the term ‘stem cell’. In haematology, haematopoietic stem cells (HSCs) are defined as cells that can self-renew and progressively differentiate into a hierarchy of committed progenitors that ultimately give rise to mature blood cells.⁶ In developmental biology, stem cells are defined more broadly according to their differentiation potential as totipotent, pluripotent, multipotent, or tissue-specific (Table 1). The cell populations used in cardiovascular studies [e.g. BM mononuclear cells (BM-MNCs) or peripheral blood CD34⁺ cells] contain only a small fraction of true stem cells but are instead relatively enriched with ‘functionally defined’ cells that do not meet the criteria to be considered stem cells. An exception is one of the most promising cell types in clinical use—mesenchymal cells—that instead meet the developmental biology definition of multipotent stem cells, as they can differentiate down multiple pathways giving rise to both mesodermal and endodermal cell types.⁷

The purpose of this article is to (i) discuss the lack of consistency in the descriptive terminology used in cardiovascular cell-based therapies; and (ii) offer guidelines that apply both to preclinical and clinical studies of cell-based therapy and are aimed at standardizing nomenclature and definitions (see Graphical Abstract). Adopting consistent, standardized definitions and nomenclature for stem and progenitor cells will improve communication among investigators and the public at large.

2. Standardizing cell nomenclature

2.1 Defining cells by surface markers

Many cells isolated from diverse sources, including peripheral blood, BM, and other tissues, are categorized according to their expression of cluster of differentiation (CD) proteins on the cell surface. The use of CD markers, each assigned a unique number, to determine lineage and differentiation state of haematologic cells is well-accepted.⁸ In part because early cardiovascular cell products were analysed in haematology settings, CD marker-based definition has been utilized for many cardiovascular cell-based products. This undeniably useful surface-marker numerical system is confounded by several issues. First, CD marker expression is dynamic both in vivo and during culture and even as cells age; thus, CD markers should be measured on cells at the time of their use and cell function as well as tissue source should be included in the cell description and discussion of any analyses, Furthermore, because CD markers represent biologically relevant receptors or adhesion molecules, multiple cell populations share the same CD markers; thus, cell markers are seldom specific for a cell type. To specify cells more precisely, in addition to reporting the conditions of cell isolation or manufacture, researchers should use more than one CD marker where possible, report all cell markers used, and specify the level of expression for each marker.

Typically, expression of CD molecules is presented as (⁺) or (⁻), although ‘high’ or ‘bright’ and ‘low’ or ‘dim’ are frequently used interchangeably as relative expression conditions. Relative marker expression is quantified either against an ‘isotype control’ antibody or by using Fluorescence Minus One (FMO). Isotype control antibodies serve to ensure that the observed staining is due to specific antibody binding rather than an artefact. They are raised against an antigen that is not found on the cell type or sample being analysed. With FMO, samples containing all antibodies minus the one of interest serve as a strong negative control and help to identify gating boundaries. The number of cells and cost required to run FMO, however, limit its use, and isotype controls are currently the more common method, although they do not reliably discriminate what is positive and negative. Table 2 displays the cell marker combinations that we propose be used for phenotypic characterization of the cells most commonly used in cardiovascular cell-based therapies. These combinations are based on recommendations of professional organizations relevant to cell therapy and on current literature.

2.2 Defining cells by cell function

In addition to surface markers, cells can be characterized based on their in vitro function as determined by cell culture assays. The colony-forming unit (CFU) assay is a haematopoietic functional assay, which is often used to measure the function or potency of haematopoietic progenitors present in the peripheral blood or BM. The colony-forming unit-fibroblast (CFU-F) assay is widely used as a functional method to quantify MSC progenitors. The CFU-endothelial cell assay (currently known as CFU-Hill assay) is used to enumerate angiogenic precursors, such as monocytes/macrophages. Endothelial colony-forming cells (ECFCs) (vide infra) are adult clonogenic EPCs defined functionally by in vitro clonogenicity over a period of weeks, capable of differentiating to regenerate endothelial cell populations.⁹ These assays are often used as surrogates for cell potency and may ultimately become functional biomarkers of in vivo efficacy, although, interlaboratory protocol standardization is needed.¹⁰

The proliferative capacity of BM cells in vitro can be associated with positive clinical outcomes or with the physiological state of enrolled participants from whom the BM was obtained.^11,12 Cells with high proliferative or regenerative capacity can be selected based on metabolic activity. For example, cells that express high aldehyde dehydrogenase activity are enriched in haematopoietic, endothelial, and mesenchymal progenitor lineages¹³ and have been evaluated in peripheral vascular disease as a tool to stimulate angiogenesis and restore blood flow in ischaemic tissues.¹⁴ Importantly, an association between patient’s age and proliferative potency of these cells has been reported.¹⁵ In addition, abnormal function of stromal precursor cells has been related to several diseases, invalidating the use of these cells for cell therapy.¹⁶

2.3 Gene expression analysis

The recent advent of genomic technologies now enables robust and comprehensive transcriptional characterization of cell types. These technologies play a critical role in deciphering molecular mechanisms and signals mediating cell mobilization, migration, and differentiation. Studies using RNA sequencing have demonstrated that MSCs from different sources present a distinct molecular signature, mostly in their secreted or membrane-bound glycosylated proteins.¹⁷ For example, Billing et al.¹⁷ identified ∼2500 significantly differentially expressed genes when comparing human BM-derived and embryonic stem cell (ESC)-derived MSCs using RNA sequencing and 40–200 proteins using mass spectrometry-based proteomics. In another report, RNA sequencing showed that 200 genes were differentially expressed in BM-MSCs when compared with placenta-derived MSCs.¹⁸ Rossini et al.¹⁹ reported that despite a remarkable similarity in growth, morphology, and immunophenotype, MSCs from cardiac and BM sources differ significantly in gene expression, microRNA content, protein expression, and growth kinetics.

Moving forward, there is a need to refine methods of cell characterization. It seems clear that current methods are woefully inadequate to capture the complexity and multifaceted nature of the phenotype of the cell products used. Perhaps, single-cell RNA sequencing, flow cytometry, ‘omics’, and functional assessments of cells can be combined to deliver extensively characterized cell population products, which should improve the reproducibility, effectiveness, and comparability of cell-therapy studies. Given the overwhelming evidence that transplanted cells act via paracrine and/or endocrine mechanisms,²⁰ an even more useful approach to characterize cells products used for cell therapy of cardiovascular disease may be one that is based upon the composition of the cells’ secretome, including characterization of exosomes and non-exosomal components.

Below, we discuss some of the most frequently used cells in cardiovascular cell-therapy studies, their naming history, and recommendations for standardizing their nomenclature.

3. Naming history and recommendations

3.1 BM mononuclear cells (BM-MNCs)

Isolated from BM aspirates by density-gradient centrifugation, BM-MNCs were one of the first cell types used in cell-based therapy for cardiovascular disease.¹ BM-MNCs comprise a heterogeneous population of white blood cells grouped into lymphocytes, monocytes, and granulocytes. However, mononuclear cells also include, at low frequencies, mature and immature HSCs, pro-angiogenic cells originally called EPCs, and mesenchymal stromal cells (often called MSCs). Most (80–98%) BM-MNCs are CD45⁺ cells; the remainder are CD45^dim or negative cells that can express a variety of progenitor cell markers. Although preclinical studies and early clinical trials suggested promising results with BM-MNCs after acute myocardial infarction (AMI), all major clinical trials in the past decade have consistently shown no benefit on scar size, cardiac function, and major adverse cardiac events, or have been inconclusive.²¹ The reason for these discrepancies is unknown. However, use of different cell isolation techniques (Ficoll vs. Sepax) and lack of standard release criteria regarding cell recovery, purity, and viability may contribute to inconsistent results. For example, contamination of isolated BM-MNCs with red blood cells (RBCs) may be a significant independent predictor of reduced recovery of left ventricular ejection fraction (LVEF) in treated patients.²² Moreover, adding RBCs to BM-MNCs impaired cell function both in vitro and in vivo.²² Therefore, reporting the percentage of RBC contamination in isolated BM-MNCs or developing stricter release criteria to limit contamination to a minimal level (<5%) could potentially increase the efficacy of BM-MNCs therapies. However, the lack of definitive data showing a clinical benefit of BM-MNCs for the treatment of cardiovascular disease has motivated researchers to pursue other cell types as clinical therapies.

Suggested nomenclature: bone marrow mononuclear cells.

3.2 Mesenchymal stromal cells

Stromal cells derived from BM and adipose tissue are often referred to as mesenchymal ‘stem’ cells (MSCs) because they can be expanded in vitro and can differentiate into osteoblasts, adipocytes, and chondroblasts in permissive cultures. The frequency of MSCs is 1 in 10–15 000 in BM and 1 in 100 000 cells in peripheral blood.²³ The International Society for Cellular Therapy (ISCT) has established minimum criteria for MSC characterization under normal culture conditions.⁷ First, MSCs should adhere to the bottom of plastic culture dishes. Second, cells should express the surface markers CD90, CD73, and CD105, and not express the CD45, CD34, CD14, CD11β, CD79α, CD19, or HLA-DR markers. Third, MSCs should have the capacity to differentiate into osteoblasts, adipocytes, and chondroblasts (mesodermal lineage) in vitro. MSCs can be isolated from different sources including BM, umbilical cord tissue, adipose tissue, and myocardium.⁷ Theoretically, cells derived from any tissue source could be considered MSCs if they meet these criteria in vitro. However, MSCs are highly sensitive to their environment and it is likely that MSCs from different sources are not identical and may have different efficacy as therapeutics.¹⁹ At present, head-to-head comparisons of therapeutic efficacy among MSCs of different sources are not available, although, as discussed above, there is mounting evidence that MSCs from different tissue sources have distinct molecular phenotypes, both in terms of surface markers and secretome components.¹⁷

Even though MSC definition guidelines have been established by the ISCT, not all trials using ‘MSCs’ have performed the characterizations described above. For example, data are inconsistent regarding the expression of CD105 in umbilical cord MSCs.⁷ Furthermore, as mentioned earlier, up to 12 different names have been given to these cells depending on their source of origin or unique definitions based on intellectual property claims, confounding the interpretation of results from clinical trials.

MSCs reduce inflammation and fibrosis, benefiting cardiac function.^20,24 These cells have been used to treat numerous cardiovascular disorders, including refractory angina, AMI, chronic ischaemic cardiomyopathy, heart failure, and ischaemic stroke.^20,24 MSC therapy has been investigated by the Cardiothoracic Surgical Network and the Cardiovascular Cell Therapy Research Network (CCTRN) in left ventricular assist device patients,²⁵ in cancer survivors with anthracycline-induced cardiomyopathy (SENECA),²⁶ in patients with ischaemic heart failure (CONCERT-HF),²⁷ and in patients with either ischaemic or non-ischaemic heart failure (DREAM-HF).²⁸ The MSCs used in the CCTRN trials (SENECA and CONCERT-HF) were analysed in every participant in terms of cell surface markers, morphology, and viability; cells that did not meet the release criteria were not administered.^26,27 Specifically, the release criteria for SENECA were: sterility, endotoxin <5.0 EU/mL, negative mycoplasma tests, viability >70%, negative bacterial and fungal cultures, negative in vitro virus tests, >95% positivity for CD105/90/73 and <2% positivity for CD45/14/19 and CD34 (flow cytometry), colony formation CFU-F (positive growth), and evidence of trilineage differentiation.^26,27 For CONCERT-HF, the release criteria for MSCs were: count of cryopreserved cells ≥75 million MSCs, endotoxin ≤5.0 EU/mL, negative mycoplasma PCR tests, viability ≥70%, negative aerobic, anaerobic, and fungal cultures, >80% positivity for CD105+ and <2% positivity for CD45+, colony formation CFU-F (positive growth).^26,27 This characterization process is not always reported in MSC studies and, therefore, variability in MSC expression markers or cell purity may account in part for differing results.

We propose that cells be called MSCs only if they express the surface markers CD90, CD73, and CD105 and are negative for CD45, CD34, CD14 or CD11b, CD79 alpha or CD19, and HLA-DR. Furthermore, the cell source should be specified; for example, adipose MSCs, BM MSCs, and umbilical cord MSCs, etc.

Suggested nomenclature: mesenchymal stromal cells.

3.3 Adipose-derived stromal cells

Adipose tissue is another rich source of progenitor cells. Adipose tissue-derived stromal cells are typically plastic-adherent and have cell surface markers similar to those of BM-MSCs. These cells have been called ASCs, adipose-derived adult stem cells, adipose-derived stem/progenitor cells (ADSCs), or ADRCs.²⁹

ADSCs contribute to angiogenesis, act as pericytes, and differentiate into osteoblasts, adipocytes, and chondrocytes; they are phenotypically positive for CD44, CD73, CD90, CD105, CD34, and negative for CD31, CD45, and CD235a.^29,30 Advantages of adipose tissue-derived cells are that they can be harvested with minimal invasiveness and expanded faster compared with BM-MSCs, and they have low immunogenicity.³⁰ The International Federation of Adipose Therapeutics and Science (IFATS) reached a consensus to adopt the term ‘adipose-derived stem cells’ to identify fat-derived, plastic-adherent, multipotent cells.²⁹

ASCs should meet the criteria established by IFATS and be fat-derived, plastic-adherent, and multipotent.

Suggested nomenclature. ASCs.

3.4 Cardiac-derived stem or progenitor cells

The identification of stem or progenitor cells in the adult heart would offer the potential of isolating and expanding autologous cells that could regenerate dead myocardium. Adult cardiac tissue is composed of cardiomyocytes and a non-cardiomyocyte fraction. Several subtypes of cardiac progenitor cells have been described, including Sca-1⁺ cardiac progenitor cells,³¹ cardiosphere-derived cells (CDCs),³² epicardium-derived cells,³³ cardiac side population cells, CD117⁺ (c-kit⁺) cardiac stem cells,³⁴ and cardiac mesenchymal cells (CMCs).³⁵ However, the expression of cell surface markers greatly overlaps among these cell populations.³⁶ It is not clear whether these multiple subsets represent transient physiological states of a single-cell lineage or unrelated cell types.

More importantly, it is doubtful that any of the aforementioned cell types is a true ‘stem’ or ‘progenitor’ cell in the sense that it has the potential to differentiate into cardiac myocytes. c-kit⁺ cardiac ‘stem’ cells were initially thought to be able to differentiate into cardiac myocytes³⁴—a finding later debunked by the Bolli lab³⁷ and others; thus, the word ‘stem’ is inappropriate in the name of these cells.²⁰ Because c-kit⁺ cardiac cells can generate endothelial cells,⁵ they are technically ‘progenitor’ cells, but given the controversy inherent in their alleged myogenic potential, we recommend the descriptive term ‘c-kit-positive cardiac cells’ (CPCs). It must be pointed out, however, that lack of stemness does not mean lack of therapeutic efficacy. Although (like MSCs) CPCs do not differentiate into myocytes and do not engraft after transplantation into an injured heart,^37,38 they are consistently effective in improving LV function in animal models of acute or chronic MI, likely via paracrine mechanisms.³⁸ In addition, in the CONCERT-HF study, CPCs reduced major adverse cardiovascular events in patients with ischaemic heart failure.²⁷

CDCs are a heterogeneous admixture of cardiac cells isolated from adult myocardium and characterized by their ability to form spheres in vitro. Human CDCs likely comprise different cell types and may express, CD105, CD29, CD34, CD31, CD117 and variable levels of CD90.³⁶ They can give rise to smooth muscle and endothelial cells and express muscle-specific genes in vitro.³⁹ Clinical trials have shown that CDCs fail to reduce scar size and improve LV function or clinical outcome in patients with ischaemic heart failure.²⁰

Recently, CMCs, a new population of reparative CD117⁻ cardiac cells, have been described.³⁵ CMCs express surface markers of mesenchymal lineage (CD90, CD29, CD73, CD105, CD106, CD146, and CD44) but lack expression of haematopoietic (CD45), endothelial (CD34 and CD31), pericyte (CD146), major histocompatibility complex (HLA-DR), and monocyte/macrophage (CD14) markers.⁴⁰ Among other mesenchymal markers, CMCs express CD106, CD9, and CD13, but not CD271 and CD166. Expression of CD117 is low (<10%). Isolated by simple plastic adherence, CMCs exhibit robust reparative properties in acute or chronic MI models and are a promising therapeutic candidate.³⁵ CMCs offer several advantages over CPCs: in vitro expansion of CMCs is easier, faster, less expensive, and more reproducible because it does not require sorting cells with CD117 antibodies. Furthermore, the starting number of CMCs in myocardial tissue is much greater than that of CPCs, which makes it possible to generate clinically relevant numbers of cells with fewer in vitro passages and thus less senescence.

In summary, many types of progenitor or stem cells have been described in the heart but, to date, there is no convincing evidence that these cells are able to differentiate into cardiac myocytes and regenerate dead myocardium. Nevertheless, at least some of these cell types, e.g. CPCs, CDCs, and CMCs, have beneficial effects in experimental models of heart disease, likely via paracrine or endocrine mechanisms, and CPCs have recently been found to improve clinical outcome in heart failure patients.²⁰ Given their therapeutic potential (irrespective of their stem/progenitor nature), proper characterization of these cell products is important.

The diversity of potential markers has hindered unambiguous identification and molecular definition of cardiac-derived cells. We propose the use of alternative methodologies, such as transcriptome analysis by sequencing and/or secretome composition, to more accurately define the various cardiac cell types. Furthermore, the designation ‘stem cells’ should be avoided unless the cell product is shown to meet the criteria for stemness enunciated above.

Suggested nomenclature: c-kit-positive cardiac cells (CPCs), CDCs, CMCs.

3.5 Endothelial progenitor cells

A population of blood-derived cells with potent angiogenic and vasculogenic actions was originally thought to give rise to mature endothelium after vascular injury; these cells were called ‘endothelial progenitor cells’.⁴¹ EPCs were isolated from peripheral blood and BM and were negative for CD45 but positive for CD34, CD133, and KDR.⁴¹ When cultured in vitro, EPCs generate a heterogeneous population, including cells with endothelial phenotypes and CD45⁺ cells expressing monocytic markers.⁴² Similar cells isolated by other groups have been characterized as pro-angiogenic cells that are CD31⁺CD34^brightCD45^dimCD133⁺.⁴³

The original assumption that EPCs give rise to mature endothelium is unlikely. Improved functional and genomic analyses have shown that EPCs are pro-angiogenic monocytes⁴⁴ that can promote vascular repair via paracrine actions on endogenous endothelial cells. Based on these findings, a consensus statement was recently published suggesting that BM- and blood-derived CD34⁺ cells be referred to as ‘myeloid angiogenic cells’ (MACs).⁴¹ The CD34⁺ cells used in most clinical trials are not cultured but are selected after mobilization from BM using G-CSF. In contrast, MACs are defined as cultured cells derived from peripheral blood mononuclear cells that are grown under endothelial cell culture conditions.⁴¹ MACs are characterized as cells positive for CD45, CD14, and CD31, and negative for CD146, CD133, and CD202b (Tie2). Furthermore, MACs do not have the capacity to become endothelial cells but promote angiogenesis and vascular repair through a paracrine mechanism.⁴¹ MACs include a category previously known as ‘early EPCs’ (cells positive for CD34, VEGFR2, CD45, and CD133).⁴²

Regardless of the name given to these cells, angiogenic precursor cells derived from blood and BM are defined differently by multiple groups and are clinically relevant in treating cardiovascular disease. Multiple cell surface markers are used as the criteria to isolate and identify blood-derived cells for cardiac and vascular therapeutic applications. These include CD45⁻CD34⁺CD133⁺ (mature EPCs), CD45⁺KDR⁺ (early EPCs), CD31⁺CD34^brightCD45^dimCD133⁺, and CD34⁺CD133⁺KDR⁺.⁴²

In addition to CD34⁺ and CD133⁺ pro-angiogenic cells in blood, normal adults are estimated to have 2.6 ± 1.6 circulating endothelial cells per millilitre of peripheral blood that give rise to colonies in vitro.⁴⁵ These blood-derived, non-monocytic cells with endothelial cell properties are defined functionally and called ‘blood outgrowth endothelial cells’ (BOECs). BOECs were identified as CD14⁻ ECFCs that appeared at 14 days after culture of peripheral blood in vitro and that expressed markers similar to mature endothelial cells.⁹ These cells participate in the repair of injured endothelium and belong to a cell type with strong angiogenic capability. Because BOECs aid in blood vessel formation and release paracrine factors, they act as trophic mediators. BOECs are now more commonly referred to as ECFCs.⁹ ECFCs include a category previously known as ‘mature EPCs’ (cells positive for CD34 and CD133 and negative for CD45).⁴² The origin of circulating BOECs and ECFCs is unknown. A major hindrance to developing these cells as a therapeutic agent for treating cardiovascular disorders is that there are no known cell surface markers.⁴¹

Due to the heterogeneous nature of the EPCs and the disparate results of different clinical studies, the identity of these cells remains elusive, and a uniform definition is not currently possible. Therefore, as was the case for cardiac-derived cells, in vitro functional characterization of EPCs and further phenotypical delineation of specific cell products (e.g. transcriptome and secretome analysis) are critical for proper nomenclature and for clinical application.

Suggested nomenclature (based on markers or function): cells derived from peripheral blood should be described based on the expression of cell surface markers (e.g. peripheral blood CD34+ cells) or on functional properties (e.g. ECFCs).

The term MACs should be used to describe cultured cells derived from peripheral blood mononuclear cells grown under endothelial cell culture conditions, which are immunophenotypically CD45⁺, CD14⁺, CD31⁺, CD146⁻, CD133⁻, and Tie2⁻.

Table 2 presents a list of cell populations including mature and immature haematopoietic cells, MSCs, EPCs, ASCs, and cardiac-derived cells with their expression markers and sources; we propose this as a guide for the nomenclature to be used by scientists and clinicians.

4. Factors limiting comparisons across cell-based therapy trials

Multiple factors can affect outcomes of cell-based clinical trials, including intrinsic variations among patients and the timing, route, and dose of treatment. However, suboptimal characterization of cell products in many studies and the lack of standards regarding the type and potency of the cells used in cell-therapy trials also limit the comparison of results between trials and thus the optimization of treatment. Discrepancies in the results of different studies and variation in patient responses within studies suggest possible heterogeneity in the cell populations being used and their potency. However, these issues are poorly understood.

The importance of such cell characterizations is illustrated by a recent study showing a direct association between the frequency of BM-derived CD31⁺ and CD34⁺ cells and infarct size reduction or LVEF improvement in patients enrolled in the CCTRN FOCUS, TIME, and Late-TIME trials.^11,46 HSCs make up ∼1–3% of BM-MNCs. Although the unfractionated BM cell populations used for cardiac repair contain 1–3% stem and 97–99% non-stem cells, the assumption has been that ‘stem cells’ are responsible for the therapeutic effects. Our studies of cell product composition and its association with outcomes suggest that the specific composition of haematopoietic and immune cells, as well as endothelial and angiogenic precursors, is associated with functional improvement.^47–49 There is also phenotypic variation in BM populations in relation to patient comorbidities and age.⁴⁸ It is plausible that, in addition to quantitative differences, there may also be qualitative differences in autologous cell products among studies and within studies, related to comorbidities, age, and concomitant medications. For example, the potency of MSCs appears to be affected by age and beta-blocker treatment of donors has been reported to affect the phenotype of CDCs.⁵⁰ Taken together, all of these considerations emphasize the importance of careful characterization of cell products when comparing the outcomes of clinical trials.

5. Exosomes: the future of cell-based therapy?

Over the past two decades, we have learned that the beneficial effects of cardiovascular cell-based therapy are likely mediated via paracrine (or even endocrine) mechanisms. As a result, it has been suggested that cell-based therapy could move towards the use of cell-free products, such as exosomes, which carry many cellular paracrine factors. However, the use of exosomes involves similar challenges as the use of cells in that their composition is affected by the isolation, culture, and other cell-source factors.⁵¹ Exosomes produced by heterogeneous cell populations are also heterogeneous and carry undefined, unpredictable groups of proteins and molecular factors. Likewise, exosome cell surface markers vary, and their therapeutic potency may vary as well, depending on the cell sources.⁵¹ The classification of stem and progenitor cells proposed herein should also assist in the analysis of cell-free products, but that discussion is beyond the scope of this review.

6. Pluripotent stem cells: unlimited source for cell-based therapy?

The ability of pluripotent stem cells—ESCs and induced pluripotent stem cells (iPSCs)—to differentiate into several cell types opened a new avenue for the treatment of several diseases. However, despite improvements in the reprogramming process⁵² and clinical-grade large-scale cell expansion,⁵³ the use of these cells is limited by safety concerns, such as genetic instability, epigenetic abnormalities, potential for teratoma formation,⁵⁴ and immaturity of the subsequent cell product.⁵⁵ Moreover, despite recent preclinical studies with cardiomyocytes derived from ESC demonstrated long-term engraftment, the beneficial effect in improve cardiac function were discouraged.^56,57 In addition to these potential complications, the clinical applicability of ESCs is impeded by the occurrence of malignant ventricular arrhythmias in non-human primate models and the need for life-long immunosuppression, which is itself a disease.⁵⁵ The problems outlined above complicate the use of ESCs and their derivatives as a clinical treatment, making it unlikely that ESCs and ESC-derived cells will find clinical application as a therapy for heart disease in the near future.⁵⁵ Clinical use may be found for iPSCs and their derivatives, but considerable work remains to be done to minimize teratoma risk and improve cost-effectiveness. Minimal criteria to translate the use of iPSCs into cell-based therapy were proposed,⁵⁸ but the lack of consensus between the societies with respect to pluripotent stem cell purity standards must be addressed to effectively and safely use these cells as a clinical product.

7. Summary and recommendations

This article highlights issues in the nomenclature and descriptions of regenerative/reparative cell products, and offers recommendations intended to improve clarity and provide a framework for moving forward. We propose a standardization of cell nomenclature and definitions to enhance communications among researchers and with the lay public. Where feasible, we have utilized recommendations made by working groups or societies with more expertise in a specific cell field.

We believe that adopting a consistent, standard nomenclature for stem and progenitor cell populations is essential. Specifically, we recommend the following:

• Strictly speaking, most cells used in studies of cardiovascular cell therapy do not meet the definition for stem cells; therefore, the term ‘stem cell therapy’ should be avoided and replaced with ‘cardiovascular cell-based therapy’ or simply ‘cell-based therapy’ or ‘cell therapy’. Adhering to this terminology is important because the phrase ‘stem cell’ often is taken to imply ‘regeneration’ or ‘transdifferentiation’, which does not occur with most cardiovascular cell therapies.

• Progenitor cells or mixed cell populations should not be referred to as ‘stem’ cells.

• Where possible, cells should be described based on their source, phenotype, and function, e.g. peripheral blood-derived CD34⁺ angiogenic cells.

• Surface CD markers are seldom specific for a cell population. Therefore, in addition to reporting the conditions of cell isolation or manufacture, more than one CD marker should be used if possible; furthermore, all cell markers used should be reported and the level of expression of each marker should be specified.

Definitions of the most commonly used cells should be agreed upon, which will enable comparisons across trials. Table 2 is a suggested source of definitions.

• Accepted consensus cell nomenclature should be used where it exists. Table 2 summarizes consensus nomenclature and its origin.

• The phenotype of cells in the final product should be assessed (e.g. by polychromatic flow cytometry) and reported, because the expression of markers may change in culture.

• The tissue source of cells should be specified because cell characteristics may differ based on their source.

• Every time a preclinical or clinical study is performed, the cell type should be characterized in vitro to include the above features (CD panel and tissue source) plus the cell age (population doublings and doubling time in hours).

• When proprietary cell names are utilized, related nomenclature should also be specified; this will improve understanding of the product, as is the case for pharmaceutical agents.

• Finally, the phrase ‘regenerative therapy’ or ‘regenerative cardiology’ should be avoided, because cell therapy does not regenerate myocardium but more accurately repairs damaged tissue, possibly via anti-inflammatory, anti-fibrotic, and/or anti-apoptotic effects.²⁰

8. Conclusions

Progress in reparative medicine requires effective communication. The simple terminology used for drugs is inadequate for cells because, unlike drug therapy, cell-based therapy carries the additional complication that the products delivered are unstable and can be highly variable among laboratories and sources of production and even within the same laboratory at different times. It is critical that investigators be cognizant of this variation and ensure their products consistently meet standard definitions.

It would be impossible to resolve all problems in terminology in one step. In this review, we have highlighted inconsistencies in the terminology currently used for cell products and suggested guidelines that would alleviate some of the resulting confusion and improve standardization. We have also discussed unresolved challenges that will require in-depth discussion and thoughtful implementation of a new and widely accepted taxonomy by the cardiovascular community. Adopting and adhering to a uniform nomenclature regarding cell-specific terminology in reparative medicine will make interexperimental comparisons more accurate and effective, both at a basic and clinical level. Moreover, such a practice may facilitate identification of cell products that will impart beneficial therapeutic outcomes in cell-based clinical trials.

Authors’ contributions

D.A.T., L.C.-A., L.C.S., M.G.d.H., E.C.P., F.C.P.M., T.D.H., J.H.T., C.J.P., J.M.H., M.P.M., P.C.Y., K.L.M., R.F.E., and R.B. contributed to the design, literature review, formulation of concepts, and conceptual content of the article and to the writing of the manuscript; R.W.V. contributed to the writing of the manuscript.

Funding

This work was supported by National Institutes of Health grants P01 HL078825 and UM1 HL113530 to RB. Funding for the Cardiovascular Cell Therapy Research Network was provided by the NHLBI under cooperative agreement UM1 HL087318.

References

Author notes