Stem cell therapy for patients with acute myocardial infarction: a systematic review of clinical trials

Abstract

Introduction

Stem cell therapy has emerged as a potential regenerative approach for Acute myocardial infarction (AMI). Despite decades of research and advancement in acute myocardial infarction (AMI) management, translating innovative therapies from bench to bedside remains a central challenge. Nonetheless, clinical outcomes exhibit considerable variability. This review provides a comprehensive overview of the clinical landscape of stem cell therapy for AMI, specifically focusing on how variations in cell type, delivery timing, routes, and dosages can affect cell therapy efficacy.

Methods

This study is a systematic review of randomized clinical trials. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed, and the study was conducted in accordance with the Cochrane Handbook for Systematic Reviews of Interventions.

Results

After searching the relevant databases, a total of 5276 studies were assessed, and 43 trials were considered eligible for inclusion in the present systematic review. The safety and efficacy of various types of stem cells, including bone marrow-derived mononuclear cells (BM-MNCs), mesenchymal stem cells (MSCs), cardiac progenitor cells, and, more recently, induced pluripotent stem cells, have been evaluated in numerous clinical trials and meta-analyses. Among these, BM-MNCs and MSCs have been the most extensively studied. Although results vary from trial to trial and can even be contradictory, from frank failures to monumental achievements, overall, the evidence supports modest but statistically significant improvements in surrogate endpoints, such as left ventricular ejection fraction (LVEF), ventricular remodeling, and reduced infarct size.

Conclusion

We have critically reviewed how methodological approaches—especially the definitions of endpoints and clinical outcome measures—have significantly influenced the reported efficacy and direction of the field. The interpretation of clinical trial results in cell therapy for AMI is heavily impacted by the specific metrics used to define success. A key focus is distinguishing between clinical trials on patients with acute and recent myocardial infarction (which is the main focus of this review) and those with chronic ischemic or non-ischemic cardiomyopathies, as they involve different treatment strategies. Patient selection is essential for improving responses in patients with AMI. Those with a severely reduced LVEF (LVEF < 40%) and younger age tend to benefit more. Limiting the transplantation window to the first 3–7 days after AMI may improve the intervention’s effectiveness.

Introduction

Despite the significant advances in treatment regimens, acute myocardial infarction (AMI) remains a leading cause of morbidity and mortality worldwide [1]. Current medical therapy, including thrombolytic therapy and primary percutaneous intervention (PCI), primarily focuses on reperfusion and limiting scar expansion, but falls short in restoring the adverse remodeling of the myocardium [2]. Thus, many patients continue to face risks of heart failure (HF) despite receiving the optimal medical therapy [3]. Considering the substantial impact of HF on the patient’s quality of life and health care costs, there is an urgent need to find treatments that can address the fundamental problem of cardiomyocyte loss and subsequent replacement by noncontractile scar tissue.

The limited regenerative capacity of the adult heart, coupled with the significant loss of cardiomyocytes associated with AMI, has prompted investigation into stem cell therapy as a novel approach to managing AMI. Cell therapy has advanced rapidly from experimental to clinical work. The first clinical trials began around 2002, demonstrating the feasibility and safety of infusing bone marrow-derived mononuclear cells (BMMNCs) into the infarcted area of the myocardium [4]. These pioneering studies encouraged more extensive clinical trials, investigating various types of stem cells and their ability to repair cardiac and vascular damage.

Despite decades of research and advancement in acute myocardial infarction (AMI) management, translating innovative therapies from bench to bedside remains a central challenge. The field of cardiovascular regenerative medicine, particularly the application of cell therapies for myocardial repair after AMI, has been a vibrant area of investigation, marked by significant preclinical promise and a cascade of clinical trials. Despite numerous studies over the last two decades, Critical questions regarding stem cell type, optimal timing of administration, delivery methods, cell dosage, and patient selection remain the subject of ongoing investigation. While initial trials showed promising results in improving left ventricular ejection fraction and reducing infarct size, subsequent larger randomized studies have yielded inconsistent outcomes, highlighting the complexity of translating preclinical success to clinical benefit. Here, it is essential to be careful separating trials on patients with acute and recent myocardial infarction from those with chronic ischemic or non-ischemic cardiomyopathies, as their general principles for treatment differ greatly.

While numerous reviews have summarized the heterogeneous evidence surrounding stem cell therapies in AMI, few have systematically addressed the persistent disconnect between preclinical promise and clinical outcomes, or integrated practical lessons from firsthand clinical trial experience. This gap in comprehensive, experience-driven analysis hinders progress and obscures the true potential of this transformative therapeutic modality. This systematic review aims to examine the current evidence on stem cell therapy for AMI, address the questions as mentioned earlier, and provide a comprehensive overview of this rapidly evolving field. Furthermore, by explicitly bridging insights from basic science and mechanistic studies with real-world clinical data, we aim to illuminate how and why specific regenerative strategies may fail or succeed in practice. Most importantly, we provide a critical examination of how methodological approaches—especially the definitions of endpoints and clinical outcome measures—have profoundly shaped the reported efficacy and direction of the field. The interpretation of clinical trial results in cell therapy for AMI is heavily influenced by the specific metrics used to define success. Early trials often focused on surrogate markers, such as left ventricular ejection fraction (LVEF) and infarct size reduction. While these are important, they may not fully capture the multifaceted nature of cardiac function or the long-term clinical benefits that patients derive from it. We will analyze how variations in the timing of assessments, the imaging modalities used (e.g., echocardiography, cardiac MRI), and the statistical methods employed have led to divergent conclusions regarding the efficacy of similar cell therapies.

Methods

This systematic review was designed and conducted in accordance with the methodological framework outlined in the Cochrane Handbook for Systematic Reviews [5], and its findings were reported in compliance with the PRISMA 2020 (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [6].

Search strategy

A comprehensive search strategy was developed to find studies evaluating the efficacy of cell therapy in patients with acute myocardial infarction. Electronic databases, including PubMed, Embase, and Web of Science, were searched from inception through March 2025 to identify relevant studies. No language or geographic restrictions were applied in the search. The keywords used combined condition-specific terms (acute myocardial infarction, AMI, STEMI, NSTEMI, heart attack, myocardial infarct) with intervention-related terms (cell therapy, stem cell therapy, cellular therapy, regenerative therapy, progenitor cell therapy). To enhance the comprehensiveness of the search strategy, synonymous terms were incorporated for both cell populations and delivery methods. Cell types included BM-MNCs, HSCs, MSCs, CSCs, skeletal myoblasts, iPSCs, and ESCs. Delivery approaches were similarly expanded to include intracoronary, intramyocardial, transendocardial, and intravenous routes, along with their associated terminologies.

Eligibility criteria

Included were randomized controlled trials and controlled clinical trials. Observational studies, case series, case reports, reviews, and animal studies were excluded from the analysis. Peer-reviewed, full-text studies were included if they involved adult participants (≥ 18 years) with a confirmed diagnosis of acute myocardial infarction, including both STEMI and NSTEMI. Studies were eligible if the mentioned forms of stem cell transplantation were administered, regardless of delivery route. The cell types included bone marrow mononuclear cells (BMMNCs), hematopoietic stem cells (HSCs), mesenchymal stem/stromal cells (MSCs), cardiac stem/progenitor cells (CSCs), skeletal myoblasts, induced pluripotent stem cells (iPSCs), and embryonic stem cells (ESCs). Studies were required to report at least one clinically relevant outcome, such as left ventricular ejection fraction (LVEF), infarct size, mortality, rehospitalization, and major adverse cardiac events (MACE).

Study selection and data extraction

All records retrieved from the database search were imported into EndNote, a reference management software, and duplicate entries were systematically removed. Two reviewers (A.A. and V.K.) screened the databases. This process involved an initial assessment of titles and abstracts to identify potentially relevant studies, followed by a full-text review of articles that met the preliminary inclusion criteria. The two authors independently reviewed the full-text papers and selected the studies for inclusion. Disagreements were resolved by consensus among the two researchers.

A custom-designed spreadsheet was developed in advance to facilitate consistent data extraction. For each study, general information was recorded, including the first author’s name, year of publication, country of origin, trial or registry name, and study design (e.g., randomized, single-blind, double-blind). Key study variables, including the stem cell type used for transplantation, the method of administration, the dosage, and the sample size, were documented on the extraction sheet. Data extraction was independently performed by two authors (A.A. and H.M.) to ensure methodological rigor and redundancy.

Results

After searching relevant databases, 5276 studies were assessed, and 43 trials were deemed eligible for inclusion in the present systematic review. Table 1 summarizes the main clinical trials in stem cell therapy for AMI. Figure 1 shows the PRISMA flowchart for the study.

Table 1.

Clinical trials of cell-based therapy after acute myocardial infarction

Study name/First author, year	Design/Phase	Number of participants	Cell type/Dose	Route of delivery	Timing from AMI to cell injection	Imaging modality	Follow-up/results
TOPCARE Assmus [4] (2002)	clinical trial Phase I/II	Treated: 20 Control: 11	BMSCs/245 ± 72 × 106 or CPCs /10 ± 7 × 106	IC	4.3 ± 1.5 days	Echo PET scan	4 mos: Improvement in LVEF and RWM, ESV reduction.
BOOST Wollert [12] (2004) Meyer [75] (2009)	RCT Phase I/II	Treated: 30 Control: 30	Nucleated BM cells 24.6 ± 0.94 × 10	IC	4–6 days	Echo CMR	6 mos: Improvement in LVEF 18 mos: No improvement in LVEF, LV volumes, and RWM 5 yrs: No improvement in LVEF, LV volumes, infarct size, and RWM
Chen [80] (2004)	RCT Phase I/II	Treated: 34 Control: 35	Autologous BM MSCs/48 to 60 × 109 cells	IC	18 days	Echo PET	3 mos: LVEF and wall motion velocity improvements, reduction in LV volumes 6 mos: The LVEF improvement observed at three months was maintained at six months
Döbert [81] (2004)	clinical trial	Treated: 26	BMCs/245 ± 72 × 106 or EPCs/10 ± 7 × 106	IC	4 ± 2 days	PET scan SPECT	4 mos: LVEF improvement
Katritsis (2005)	Clinical trial Phase I	Treated: 11 Control: 11	MSCs and EPCs/2–4 × 106	IC	242.4 ± 464.0 days (range, 8–1560)	SPECT Echo RNV	4 mos: No significant changes in LVEF and LV volumes, improvement of myocardial contractility, reduction in WMSI and myocardial scar segments
Janssens [58] (2006)	RDBC, Phase II	Treated: 33 Control: 34	BMSC/3 ± 1.3 × 108	IC	24 h after PCI	CMR	4 mos: No improvement in LVEF, reduction in infarct size
ASTAMI Beitnes [82] (2009)	RCT	Treated: 47 Control: 50	BMMNCs/ 0.7 × 108	IC	4–8 days	SPECT Echo CMR	6 mos: No improvement in EF, infarct size, and LV volumes 12 mos: No improvement in EF, LV volumes, and RWM 3 yrs: No improvement in EF, LV volumes, LV mass, infarct size, and RWM
REPAIR-AMI Schächinger [59] (2006) Assmus [83, 84] (2010, 2014)	RDBCT	Treated: 101 Control: 98	BMMNC/ 2.36 ± 1.74 × 108	IC	3–6 days	LV angiography	4 mos: Improvement in EF, ESV and RWM 1 year: ↓MACE, ↑RWM, 2 yrs: ↓MACE, ↑RWM, ↓infarct size, no improvement in EF and LV volumes 5 yrs: ↓MACE
TACT-PB-AMI Tatsumi [85] (2007)	clinical trial, phase I/II	Treated: 18 Control: 36	PBMNC/4.9 × 10⁹	IC	2.5 ± 0.5 days	Echo SPECT	6 mos: improvements in global/regional LVEF, improved ΔEF
Meluzín [86] (2008)	RCT, phase II	Treated: 40 Control: 20	BMMNC/low dose: 107, high dose: 108	IC	5–9 days	Echo SPECT	3, 6 and 12 mos: LVEF improvement in the high dose group
Hare [69] (2009)	RDBC, phase I	Treated: 34 Control: 19	BM MSC/0.5, 1.6, and 5.0 × 106/kg	IV	1 to 10 days	Echo MRI	3, 6 and 12 mos: LVEF improvement
Cao [87] (2009)	RCT, phase II	Treated: 41 Control: 45	BMMNC/5 + 1.2 × 107	IC	7 days	Echo SPECT	6 mos, 1 year and 4 years: Improved ΔLVEF and decreased ESV, No significant difference in infarct size
Grajek [88] (2010)	RCT, phase II	Treated: 31 Control: 14	BMMNC/2.34 + 1.2 × 109	IC	4–6 days after PCI	Echo SPECT	6 mos and 12 mos: improved perfusion index, no difference in ΔEF between groups
FINCELL Huikuri [89] (2009)	RDBCT	Treated: 39 Control: 38	BMMNCs/402 ± 196 × 106	IC	2–6 days	Echo LV angiography	6 mos: ΔEF improvement
MYSTAR Gyöngyösi [90] (2009,2016)	RCT, phase II	Treated: 60 (Early treatment : 30, late treatment: 30)	BMMNCs/1.56 ± 0.40 × 10⁹ in early and 1.55 ± 0.44 × 10⁹ in the late group	Combined IM and IC	Early treatment : 21–42 days, late treatment: 3–4 mos	SPECT MRI	3 mos: reduction in infarct size and LVEF improvement in both groups, but no difference between groups 1 year: improvement in LVEF 5 year: preserved the 1 year results
REGENT Tendera [91] (2009)	RCT, phase II	Treated: 160 Controls: 40	BMMNCs/ 1.78 × 108 CD34+/CXCR4+ 1.9 × 106	IC	3–12 days	CMR LV angiography	6 mos: No improvement in EF and LV volumes
Piepoli [92] (2009)	RCT, phase II	Treated: 19 Controls: 19	BMMNCs/418 ± 186 × 10⁶	IC	4–7 days after PCI	Echo SPECT	12 mos: Improvement in LVEF, Enhanced perfusion and global/regional function
Wöhrle [93] (2010)	RDBCT	Treated: 29 Controls: 13	BMMNC/381 × 106	IC	5–7 days	CMR	3, 6 mos: No improvement in LVEF, infarct size, or LV volumes
HEBE Hirsch [94] (2011)	RCT, phase II	Treated: 66 (PBMNC), 69 (BMMNC) Controls: 65	BMMNC/296 ± 164 × 106 vs. PBMNC/287 ± 137 × 106	IC	3–8 days	CMR	4 mos: No improvement in global or regional EF, no difference in infarct size between groups
COMPARE Mansour [28] (2010,2011)	RDBCT, phase II	Treated: 17 Control: 20	autologous CD133 + BMSCs	IC	6.4 ± 2.2 days after PCI	Echo MRI	4 mos: LVEF improvement 1 year: LVEF improvement 10 yrs: safety reported, no improvement in LVEF
LateTIME, Traverse [14] (2011)	RDBCT, phase II	Treated: 58 Placebo: 29	BMMNCs/ 1.5 × 108	IC	15–20 days	CMR	6 mos: No improvement in EF, LV volumes, RWM, and infarct size
TIME Traverse [13] (2012)	RDBCT, phase II	Treated: 80 Placebo: 40	BMMNCs/ 1.5 × 108	IC	3–7 days	CMR	6 mos: No improvement in EF, LV volumes, RWM, and infarct size
Penn [95] (2011)	Open-label clinical trial, phase I	Treated: 19 Control:6	Allogeneic multipotent adult progenitor cells (MultiStem)/20, 50, or 100 million	Adventitial (perivascular)	2–5 days	Echo	4 mos: improvement in EF and stroke volume in Combining both the 50 million and 100 million dose groups
Turan [96] (2012)	RCT, Phase II	Treated: 42 Control:20	BMSC/1.9 × 10⁹	IC	7 days	LV ventriculography	3, 12 mos: improvement in LVEF and reduction of infarct size
CADUCEUSMakkar [97] (2012), Malliaras [98] (2014)	RCT, phase I	Treated: 17 Control: 8	CDCs/12.5, 17.3 or 25 milion	IC	65 ± 14 days	MRI	6 mos: reductions in scar mass, increase in regional contractility, no improvement in LVEF, LV volumes 1 year: same as the previous F/U
Gao [99] (2013)	RCT, Phase I/II	Treated: 21 Control: 22	Autologous BM MSCs/3.08 ± 0.52 × 106	IC	17.1 ± 0.6 days after reperfusion	SPECT Echo	6, 12, 24 mos: no significant difference in LVEF absolute changes, LV volumes, and WMSI between the two groups
Rodrigo [100] (2013)	Clinical trial	Treated: 9 Control: 45	autologous BM MSCs/31 × 10⁶	IM	21 ± 3 days	Echo SPECT	3 mos: reduction in perfusion defect size 12 mos: Improvement in LVEF 4–5 yrs: Sustained LVEF improvement, safety established
SWISS-AMI Surder [61] (2013)	RCT	Treated early: 60 Treated late: 58 Controls: 49	BMMNCs/ 1.4–1.6 × 108	IC	5–7 days 3–4 wks	CMR	4 mos: No improvement in EF, LV volumes, and scar mass
Wang [101] (2014)	RCT/Phase I/II	Treated: 28 Control: 30	BM MSC/2 × 10⁸	IC	Day 15 ± 1 after PCI	Echo	1, 3, and 6: no significant differencence in LVEF, LV diameters and infarct size between groups
SEED-MSC Lee [102] (2014)	RCT, Phase I/II	Treated: 30 Control: 28	Autologous BM MSCs/7.2 ± 0.9 × 107	IC	~ 1-month	SPECT Echo Angiography	6 mos: Improved ΔLVEF, No significant differences in LV volumes and WMSI between groups
Benedek [103] (2014)	RCT, Phase I/II	Treated: 9 Control: 9	BMMNCs/1.66 ± 0.32 × 109	IC	3 weeks − 3 mos	Echo CCTA	4 years: Reduction in plaque burden, no difference in MACE between groups
TECAM San Román [104] (2015)	RCT, Phase II	Treated: 30 (BMMNC), 30 (G-CSF), 29 (BMMNC + G-CSF) Control: 31	BMMNC/83 × 10⁶ if no G-CSF, 560 × 10⁶ if G-CSF pre-treated)/G-CSF 10 µg/kg/day × 5 days	IC	3–5 days after PCI	CMR, LV angiographyEcho	12 mos: No improvement in LVEF in the 3 treatment groups; only a modest reduction in infarct size observed.
Gao [105] (2015)	RDBCT, Phase II	Treated: 58 Control: 58	Wharton’s jelly-derived MSCs/6 × 106	IC	5–7 days after reperfusion	PET, SPECT and echo	4, 12 mos: Improved ΔLVEF, WMSI reduction 18 mos: Improved ΔLVEF, WMSI reduction, greater reduction in LV volumes
Nicolau [108] (2017)	RDBCT, Phase II	Treated: 66 Control: 55	BMMNCs/108	IC	6–9 days	CMR	6 mos: No improvement in LVEF and LV volumes
REGENERATE-AMI Choudry [107] (2016) Mathur [108] (2022)	RDBCT, Phase II	Treated: 54 Control:44	Autologous BMCs/≈ 60 × 106	IC	First 24 h	CMR or CT LV angiography	1 year: non-significant improvement in LVEF, significant myocardial salvage improvement and, reduction in infarct size 5 years: no difference in MACE
PreSERVE-AMI Quyyumi [27] (2016)	RCT, Phase II	Treated: 78 Control: 81	Autologous BM CD34 + cells/14.9 ± 8 × 10⁶	IC	11 days after PCI	SPECT, CMR	6 mos: Improved ΔLVEF in patients receiving > 20 million cells and trend toward reduced infarct size 12 mos: decreased mortality
BOOST-2 Wollert [109] (2017)	RDBCT, Phase II	Treated: 127 Control:26	autologous BMCs/ 20.6 ± 7.7 × 10⁸ or 7.0 ± 2.9 × 10⁸	IC	8.1 ± 2.6 days after PCI	CMR	6 mos: No improvement in LVEF, no change in infarct size or LV volumes
Kim [63] (2018)	RCT, Phase I/II	Treated: 14 Control: 12	BM MSCs/7.2 ± 0.9 × 107	IC	30 ± 1.3 days after PCI	SPECT and echo	4 mos: LVEF improvement, no significant differences in LV volumes 12 mos: LVEF improvement, no significant differences in LV volumes
CAREMI, Avilés [110] (2018)	RDBCT, Phase I/II	Treated: 33 Placebo: 16	Allogenic Human CSCs/35 × 106	IC	5–7 days after PCI	MRI	1, 6 mos: no significant differences in infarct size between the 2 groups 12 mos: No significant differences between groups in terms of ventricular volumes, LVEF, or RWM
ALLSTAR Makkar [35] (2020), Ostovaneh [16] (2021)	RDBCT, phase II	Treated: 90 Control: 44	CDCs/25 × 106	IC	4.6 ± 3.1 mos	MRI	6 mos: no reduction in scar size, improvement in LV volumes and segmental Ecc, no improvement in LVEF
BAMI Mathur [77] (2020)	Open label trial	Treated: 185 Control:190	BMMNC/25–500 × 106	IC	2–8 days after PCI	-	2 years: all-cause mortality was low and similar between groups (3.26% vs. 3.82%)
Zhang [111] (2021)	RCT, Phase II	Treated: 21 Control: 22	Autologous BM MSCs/2–5 × 10⁶	IC	< 1 month	Echo SPECT PET scan	6 mos: No significant improvement in myocardial perfusion or metabolic defect index 12 mos: LVEF improved in both groups, but no between-group difference
BOOSTER-TAHA7 Attar [112] (2023)	RCT, Phase II	Treated: 20 (single dose), 20 (double dose) Control: 25	Wharton’s jelly MSCs/	IC	3–7 days	Echo CMR	6 mos: greater improvement in ΔLVEF in repeated infusion group compared to single MSC transplantation and control groups

RCT, randomized clinical trial; RDBCT, randomized double blind; BM, bone marrow; BMMNCs, bone marrow mononuclear cells; MSCs, mesenchymal stromal/stem cells; BMSC, Bone Marrow–Derived Stem Cell; CPCs, circulating blood–derived progenitor cells; EPCs, endothelial progenitor cells; PBMNC, peripheral blood–derived mononuclear cells; CDCs, Cardiosphere-derived cells; CSCs, cardiac stem cells; IC, intracoronary; IV, intravenous; IM, intramyocardial; PCI, Percutaneous Coronary Intervention; CMR, cardiac magnetic resonance imaging; Echo, echocardiography; SPECT, single-photon emission computed tomography, PET, positron emission tomography; CCTA, Coronary Computed Tomography Angiography; RN, Radionuclide Ventriculography; EF, ejection fraction; ESV, end-systolic volume; LV, left ventricular; EDV, end-diastolic volume; WMSI, Wall Motion Score Index, MACE, major adverse cardiovascular events, RWM, regional wall motion; mos, months

Fig. 1.

PRISMA flow chart diagram of the study

MI and stem cell therapy

Cell-based therapy in clinical medicine primarily began with trials focused on AMI. The aim was to administer an intracoronary infusion of stem cells as an adjunct to percutaneous coronary intervention to mitigate cardiomyocyte necrosis and prevent the progression of heart failure (HF), an approach commonly referred to as cardioprotection [7]. The mechanism through which stem cell therapy enhances cardiac function remains a topic of debate. While the hypothesis that transplanted cells could differentiate into cardiomyocytes has proven to be a rare event, growing evidence suggests that paracrine signaling is a key mode of action in stem cell therapy [8, 9]. Transplanted cells can produce growth hormones, cytokines, metalloproteinases, and exosomes, which could increase vascularity and collateral growth, promote cardiomyocyte proliferation, limit or reduce fibrosis, and/or activate endogenous resident stem cells [8, 9]. These mechanisms target the adverse remodeling of the damaged myocardium and ultimately reduce the scar size [2].

Experimental and clinical work has advanced in parallel, rapidly translating fundamental discoveries into clinical trials. The first clinical study by Strauer et al. (2002) [10] marked a turning point, reporting improved LVEF in AMI patients treated with intracoronary BM-MNC transplantation. Early trials such as TOPCARE-AMI [4, 11] and BOOST [12] further supported the safety and feasibility of BM-MNC therapy, suggesting modest improvements in LVEF during a short follow-up period (4–6 months) and enhancements related directly to cardiac remodeling. However, results were inconsistent throughout the studies. For instance, the TIME [13], LateTIME [14], and SWISS AM [15] trials found no improvement in LVEF between the placebo and BMMNC groups, highlighting the need for further research into optimal cell types and delivery methods.

The field progressed with trials exploring various cell types, including cardiosphere-derived cells and MSCs. For example, the ALLSTAR trial [16] investigated the therapeutic effects of allogeneic cardiosphere-derived cells, demonstrating significant enhancements in segmental myocardial function. A recent meta-analysis by Attar et al. demonstrated a mean improvement of LVEF by 3.67% with MSC therapy in patients following AMI [17]. However, due to the limited number of studies and major adverse cardiac events (MACE) occurrences in MSC therapy research, there is no consistent reduction in mortality or MACE [18]. This discrepancy underscores the need for ongoing research to optimize multiple parameters, including cell type, isolation protocols, timing, delivery route, and patient selection, before stem cell therapy can be routinely implemented in cardiovascular care.

Types of stem cells used in cardiac repair post-MI

A wide range of cells with various characteristics are classified as stem cells (Fig. 2). Properties such as lineage-differentiation potential, self-renewal capacity, degree of purification, and numerous other factors can affect the efficacy of cell therapies. Therefore, it is anticipated that various cells contribute to myocardial regeneration. Here, we provide a comprehensive overview of various clinical and some preclinical studies investigating the application of different stem cell types for cardiac repair after myocardial infarction (MI).

Fig. 2.

Various cell types have been used in clinical trials for patients with acute myocardial infarction

Bone marrow mononuclear cells (BM-MNCs)

BM-MNCs are a heterogeneous population of cells that can be easily achieved via bone marrow aspiration. Then, a simple Ficoll-gradient centrifugation can yield an achievable autologous cell mixture (Fig. 3). This mixture contains both highly potent stem cells and highly differentiated terminal cells. As the cell-obtaining procedure is not costly and fast, this source of cells has been most widely studied in regenerative cardiology as an autologous stem cell source.

Fig. 3.

A schematic illustration of how bone marrow mononuclear cells (BM-MNCs) are prepared. To obtain these cells, bone marrow aspiration is typically performed at the iliac crest. The cells are then processed through either machine-based apheresis or Ficoll gradient centrifugation

BM-MNCs have demonstrated potential for cardiac repair in both experimental models and clinical applications. In vitro researches indicate that BM-MNCs are capable of engrafting into injured myocardial tissue and differentiating into cells with cardiomyocyte-like characteristics [19]. In addition, over the past twenty years, numerous preclinical and clinical investigations have explored the use of autologous BM-MNCs as a therapeutic approach for treating cardiovascular disorders [20]Around 45 trials, enrolling more than 3,300 patients, have evaluated the efficacy of BM-MNCs in AMI, and the results are highly variable [21].

A meta-analysis combining findings from 22 RCTs with 1360 patients aimed to evaluate the efficacy of intracoronary BM-MNCs therapy on left ventricular (LV) function, remodeling, and both LV diastolic and systolic output in ST-elevation myocardial infarction (STEMI) patients. The pooled statistics demonstrated a significant enhancement in LVEF [2.58, p < 0.001], LVESV [4.67, p < 0.001], and a significant reduction in LVEDV [−3.73, p = 0.02] in the BM-MNCs group compared to the control group. In the sensitivity tests, a notable decrease in LVEDV was eliminated; however, the results for LVEF and LVESV remained constant [22]. Additionally, according to the meta-analysis on 22 RCTs conducted by de Jong et al. [23], LVEF increased by 2.10% (P = 0.004) in the BM-MNCs group in comparison with controls, attributed to a preservation of LVESV (−4.05 mL, P = 0.006) and a decrease in infarct size (−2.69%, P = 0.01). However, an analysis of only RCTs (n = 9) utilizing MRI-derived endpoints reveals no impact on infarct size, heart function, and volumes. Furthermore, no beneficial effect on the incidence of major adverse cardiac events was observed following BM-MNC infusion, with a median follow-up of 6 months.

A collaborative meta-analysis involving 16 studies enrolling 1,641 patients clarified that patient selection is a key factor in achieving better success. In that study, the absolute enhancement in LVEF was more pronounced in patients treated with BM-MNCs compared to the control group [2.55% increase, P < 0.001]. Therapies with cells significantly diminished LVEDV and LVESV (−3.17 mL/m², P < 0.001; −2.60 mL/m², P < 0.001, respectively). The therapeutic advantage of LVEF enhancement was notably greater in younger patients than in older patients (age < 55 years: 3.38%; age ≥ 55 years: 1.77%; P = 0.03). This variability in treatment effect was similarly visible in the decrease in LVEDV and LVESV. In addition, patients with baseline LVEF < 40% obtained higher outcomes from intracoronary BM-MNCs therapy (LVEF < 40%, 5.30%, LVEF ≥ 40%, 1.45%, P < 0.001) [24].

Attar and colleagues performed a meta-analysis, collecting data from 23 randomized controlled trials (RCTs) involving 2286 participants, and considered clinical rather than surrogate endpoints, such as LVEF. They found that the transplantation of BM-MNCs may decrease the combined endpoint of hospitalization for congestive heart failure (CHF), reinfarction, and cardiovascular mortality (91/1191 vs. 111/812, p = 0.002), primarily by lowering rates of re-infarction (23/1159 vs. 30/775, p = 0.046) and being hospitalized for heart failure (47/1220 vs. 62/841, p = 0.005), but not cardiovascular mortality (28/1290 vs. 31/871, p = 0.207) [20].

Hematopoietic stem cells (HSCs)

Hematopoietic stem cells (HSCs) are a group of multipotent stem cells that primarily reside in the bone marrow and are responsible for maintaining the hematopoietic system. These cells are capable of secreting paracrine factors and growth mediators that may contribute to cardiomyocyte protection, promoting new blood vessel formation, immunomodulation, antioxidant activity, inhibition of apoptosis and inflammation, reduction of fibrosis, and enhancement of cardiac contractile function [25]. Several studies have investigated the effects of HSC injection in patients following myocardial infarction (MI). In the REVIVAL-2 trial, granulocyte colony-stimulating factor (G-CSF) was used to mobilize stem cells from the bone marrow. Between baseline and follow-up, the left ventricular infarct size measured by scintigraphy decreased by 6.2% ± 9.1% in the G-CSF group compared to 4.9% ± 8.9% in the placebo group (P = 0.56). LVEF showed a slight improvement of 0.5% ± 3.8% in the G-CSF group versus 2.0% ± 4.9% in the placebo group (P = 0.14). Angiographic restenosis was observed in 35.2% of patients in the G-CSF group and 30.9% in the placebo group (P = 0.79). Although G-CSF increased circulating CD34 + cells, it did not significantly reduce infarct size or improve LV function. Furthermore, there was no significant difference in the incidence of restenosis between the treatment and placebo groups [26]. The PreSERVE-AMI Phase 2 trial focused on intracoronary infusion of autologous CD34 + hematopoietic stem cells and showed more promising results. The results indicated a significant improvement in myocardial perfusion in the treatment group after 6 months (p < 0.001); however, no significant difference in LVEF improvement was detected between the groups. Secondary analyses demonstrated a dose-dependent relationship between CD34 + cell treatment and enhancements in LVEF, a decrease in infarct size, and an increased duration of patient survival outside the hospital, after adjusting for ischemia time (p = 0.05) [27]. CD133 + cells have also been investigated in previous studies. Patients with acute MI and left ventricular dysfunction received intracoronary injections of CD133 + bone marrow-derived cells in the COMPARE-AMI trial. The preliminary results demonstrated a significant enhancement in LVEF, rising from 41.3%±5.3 at baseline to 52.1%±7.1 at four months post-treatment (P < 0.001), with no major adverse cardiac events (MACE) or accelerated atherosclerosis observed. The results indicate a potential regenerative effect supported by the angiogenic and anti-apoptotic characteristics of CD133 + cells [28]. Similarly, in a PET-based imaging study assessing myocardial perfusion and metabolism, patients treated with intracoronary injection of CD133 + bone marrow-derived stem cells (Group A) showed a significant reduction (P < 0.05) in the quantity of scarred segments and infarct size, followed with an elevation in myocardial blood flow (MBF) (P < 0.05) and a slight enhancement in cardiac function. In contrast, those who received CD133 + peripheral blood-derived stem cells (Group B) exhibited no reduction in infarct size, which correlated with a significant decline in MBF (P = 0.01) and cardiac dysfunction. At the same time, patients under standard therapy (Group C) showed inconsistent or minimal improvements. In addition, no major adverse events happened in Group A, confirming the procedure’s safety. These findings emphasize that the origin and selection of stem cells significantly affect therapeutic results, with CD133 + bone marrow cells demonstrating greater effectiveness in maintaining and improving cardiac function after infarction [29].

Mesenchymal stem cells (MSCs)

Mesenchymal stem cells (MSCs), which should better be called as multipotent stromal cells, are known for their regenerative capacity and anti-inflammatory and immunomodulatory effects [30]. MSCs offer ease of isolation, ex vivo growth, in vitro proliferation, and immune-privileged properties, so their use in clinical trials is expanding rapidly [31]. According to the POSEIDON clinical trial on MSC transplantation, allogeneic MSCs are safe and as effective as autologous MSCs [32]. Thus, they can be easily prepared as off-the-shelf products from allogenic sources, making them an exciting resource of cells.

MSCs can originate from diverse sources, including bone marrow (BM-MSCs), umbilical cord (UC-MSCs), and adipose tissue (AD-MSCs). Human umbilical cord MSCs (hUC-MSCs), also known as Wharton’s jelly-derived MSCs (WJ-MSCs), exhibit strong multipotency and self-renewal capabilities, making them promising for replacing damaged heart cells. WJ-MSCs may be more beneficial in myocardial infarction, as they exhibit more potent immunomodulatory effects, lower immunogenicity, faster replication, and smaller size than MSCs derived from adipose tissue or bone marrow, according to prior clinical and animal research [33].

Gao et al. [34] conducted the largest MSC trial in AMI, a multicenter study involving 116 patients with acute ST-elevation MI. Participants were randomly assigned to receive either an intracoronary infusion of WJMSCs or a placebo into the infarct artery five to seven days after successful reperfusion. After 18 months, the group treated with WJMSCs showed significantly greater improvements in myocardial viability (PET) and infarcted territory perfusion (SPECT), with increases of 6.9 ± 0.6% (95% CI, 5.7 to 8.2) and 7.1 ± 0.8% (95% CI, 5.4 to 8.8), respectively, compared to the placebo group’s 3.3 ± 0.7% and 3.9 ± 0.6% at four months (P < 0.0001 and P = 0.002). Additionally, at 18 months, the increase in LVEF was significantly higher in the WJMSC group (7.8 ± 0.9) than in the placebo group (2.8 ± 1.2), with a P value of 0.001. Attar et al. [17] performed a meta-analysis of clinical trials, revealing a significant increase in left ventricular ejection fraction (LVEF) in comparison with baseline among those receiving MSCs, including these three types of cells (WMD = 3.78%). The effect was significantly greater when the transplantation occurred throughout the first week post-AMI (MD = 5.74%). Subgroup analysis also revealed that the efficacy of this novel medication becomes significantly greater over a follow-up period reaching 12 months (WMD = 4.21%). According to another subgroup analysis done for stem cell source, it was demonstrated that umbilical cord-derived stem cells had a more beneficial impact on EF (WMD = 5.1%) compared to bone marrow-derived and adipose-derived cells (WMD = 1.98% & WMD = 5.1%, respectively). The efficacy of intracoronary infusion was similar to that of transendocardial injection (WMD = 3.565% vs. 4%, respectively). MSC dosages of lower and more than 10⁷ cells could not increase LVEF differentially (WMD = 5.24% vs. 3.19%, respectively).

Since MSCs are purer stem cells than BM-MNCs, they might be expected to perform better [8]. However, this cannot be confirmed without appropriate, well-designed clinical studies or meta-analyses. The only trial directly assessing this is the TAC-HFT trial, which was conducted in patients with chronic ischemia rather than in those with AMI. While this trial might support the hypothesis, it has notable limitations. First, the sample size was small, with 19 participants in the MSC group and 19 in the BM-MNCs group. A meta-analysis compared the efficacy of BM-MNCs and MSCs in patients with AMI, including 36 trials (26 on BM-MNCs and 10 on MSCs) with 2489 patients (1466 transplanted, including 1241 with BM-MNCs and 225 with MSCs, and 1023 as controls). Both cell types showed significant short-term improvements in ejection fraction (BM-MNCs: WMD = 2.13%, p < 0.001; MSCs: WMD = 3.71%, p < 0.001), with MSCs being more effective according to ICEMAN criteria (Fig. 4).

Fig. 4.

Forest plot comparing the effect of bone marrow mononuclear cells (BM-MNCs) with mesenchymal stem cells (MSCs) transplantation on improving left ventricular ejection fraction (LVEF) after myocardial infarction

Cardiac stem cells (CSCs)

Despite a previous idea that the heart is a terminally differentiated organ with no regenerative capacity, we now know resident stem cells are within the myocardium (Fig. 5) [34]. Intracoronary injection of cardiac stem cells, particularly cardiosphere-derived cells (CDCs), has emerged as a promising approach to support myocardial repair following MI [35]. Cardiosphere-derived cells (CDCs) have been widely investigated as a potential treatment for cardiac injury but growing preclinical and clinical evidence suggest that their effects on heart tissue repair and functional myocardial recovery are still limited [36, 37]. For instance, Zhao et al. [36] conducted a preclinical study to investigate the effects of CDCs isolated from neonatal mice on cardiac function in aging mice. Cardiac function was assessed in aging mice that received either PBS (n = 15) or CDCs injections (n = 19). Echocardiographic analysis revealed similar LVEF (57.46% ± 3.57% vs. 57.86% ± 2.44%) and LV fractional shortening (30.67% ± 2.41% vs. 30.51% ± 1.78%) in both groups. These findings suggest that CDC administration does not enhance cardiac function or overall physiological performance in aging mice. Moreover, in another study conducted in Wistar rats, MI was induced, and the animals were treated with human CDC (hCDC), rat CDC (rCDC), and a placebo. However, no significant variations in cardiac functional parameters were detected by echocardiography between these three treatments [38]. Similarly, in the ALLSTAR trial, 142 eligible patients were randomly assigned, and 134 ultimately received treatment (90 in the CDCs group and 44 in the placebo group). At the 6-month follow-up, there was no significant difference between the CDCs and placebo groups in the percentage change in scar size (P = 0.51). Intracoronary administration of allogeneic CDCs in patients with left ventricular dysfunction following MI was well tolerated. Still, it did not lead to a reduction in scar size compared with placebo after 6 months [35]. However, CDC-treated patients showed significant decreases in LV end-diastolic volume (LVEDV), LV end-systolic volume (LVESV), and NT-proBNP (NT-proBNP) levels (all P = 0.02) compared with placebo. In a separate trial, comparable to the previous study, CDCs were delivered via intracoronary infusion to patients following MI. Administration of CDCs in these patients led to an enhanced segmental myocardial circumferential strain (Ecc) compared with placebo (p = 0.05) [16]. These results underscore the importance of assessing segmental myocardial function as a key endpoint in future clinical studies involving post-myocardial infarction (MI) patients [16]. Another study that administered Allogeneic human cardiac stem cells (AlloCSC-01) in Patients with STEMI and left ventricular dysfunction showed that AlloCSC-01 can be safely injected early after revascularization. At the same time, reductions in infarct size and ventricular remodeling at 12 months were limited and did not reach statistical significance [39].

Fig. 5.

Overview of cardiac stem and progenitor cell derivation with potential to be used following myocardial infarction. This schematic illustrates the origin and classification of cardiac-derived stem and progenitor cells used in post–myocardial infarction (MI) therapy. Human myocardial biopsy specimens are processed to isolate different cell populations. Single-cell suspensions derived from cardiospheres contain various cell types expressing markers such as Cx43, c-kit, and CD105. Stem cells isolated using side population flow cytometry can be identified by markers including c-kit, Sca-1, Abcg2, CD34, and CD45. Cardiosphere cells from this source may also express Musashi-1 and Nestin. Multiple progenitor subsets are expanded and characterized from myocardial biopsies, including CFUFs (Cardiac Fibroblast–Derived Uncommitted Cells), Isl1⁺ cells, Sca-1⁺ cells, c-kit⁺ cells, EPDCs (Epicardial-Derived Progenitor Cells), and CDCs (Cardiosphere-Derived Cells). Their molecular signatures involve combinations of surface and lineage markers—such as CD31, CD34, CD90, CD105, and FLK1—and regulatory factors like GATA4, NKX2-5, MEF2C, and Isl1. After in vitro characterization, selected cell types can be delivered to the myocardium after MI to support myocardial regeneration and restore function

Although MRI findings in some studies may suggest scar reduction, these observations should also be confirmed histopathologically, since MRI may mistake increased tissue perfusion for a decrease in scar size, even when the cells themselves have not truly regenerated but only received improved blood supply [40].

Skeletal myoblasts

Autologous skeletal myoblasts were considered a promising cellular source for myocardial repair due to their unique ability to form muscle and undergo contraction, as well as their biochemical and functional analogies to cardiac cells, significant proliferation potential in vitro, and resilience to ischemia [41]. Furthermore, engrafted myoblasts have the potential to repair post-infarction scars and enhance cardiac function [41]. These advantages were demonstrated in a 2004 study, which showed that transcatheter injection of autologous skeletal myoblasts in postinfarction patients with significant left ventricular dysfunction is feasible, safe, and an effective therapeutic approach. This trial demonstrated the technical success of skeletal myoblast transplantation in all six patients, with no complications; an average of 19 ± 10 injections was administered per patient, resulting in the implantation of 210 × 10⁶ ± 150 × 10⁶ cells per patient. LVEF increased from 24.3% ± 6.7% at baseline to 32.2% ± 10.2% after 12 months post-myoblast implantation (p = 0.02 compared to baseline and p < 0.05 compared to controls); in the control group, LVEF declined from 24.7% ± 4.6% to 21.0% ± 4.0% (p = NS). At 1 year, NYHA functional class and walking distance showed considerable improvement (p = 0.001 and p = 0.02 compared to baseline), while matched controls showed no change [42]. In addition, the POZNAN trial aimed to evaluate the safety and feasibility of autologous skeletal myoblast injection conducted by a percutaneous trans-coronary-venous method in post-infarction patients with left ventricular dysfunction. At a 6-month follow-up, NYHA classification improved in each patient, and EF increased by 3–8% in 6 of the nine patients evaluated [43]. Despite these promising preliminary results, trials investigating the effects of skeletal myoblast transplantation in patients with heart failure and ischemic cardiomyopathy showed the development of ventricular arrhythmias. Consequently, further studies in patients with MI were stopped [44–47].

Endothelial progenitor cells

The dominant paradigm, indicating that in adults the formation of new blood vessels occurs only by migration and proliferation of mature endothelial cells in a process termed angiogenesis, was overturned in recent years by the discovery of endothelial progenitor cells (EPC), which differentiate into endothelial cells in a process referred to as vasculogenesis. There are at least three subtypes of EPCs. Their role as markers for the diagnosis or screening of cardiovascular disorders, such as coronary artery disease [45], has been demonstrated. However, no clinical study has been done regarding their effect in patients with AMI. Some preclinical studies have been performed [48]. In a clinical trial by Katritsis et al., a combination of MSCs and EPCs was transplanted via coronary arteries, and the wall motion score index significantly decreased at follow-up in the transplantation group (P = 0.04), but not in the control group. During stress echocardiography, five of 11 patients (all with recent infarctions) showed improvement in myocardial contractility in one or more previously nonviable segments, while none of the controls did (P = 0.01). Additionally, six of 11 transplant patients exhibited restored Tc(99 m) sestamibi uptake in previously nonviable myocardial scars, compared to none of the controls (P = 0.02).

Induced pluripotent stem cells (iPSCs)

A more recently introduced cellular source for heart regeneration is induced pluripotent stem cells (iPSCs), which result from a significant advancement in converting somatic cells into pluripotent stem cells. The iPSCs can grow indefinitely in vitro and develop into various cardiac lineages, including cardiomyocytes, cardiac progenitors, smooth muscle cells, and endothelial cells. The iPSCs can be sourced from patients for autologous therapy, or repositories of human leukocyte antigen (HLA) diverse iPSCs may be utilized for allogeneic therapy [49]. However, Concerns about iPSC technology remain, particularly about the characteristics of iPSC-derived cardiomyocytes: these cells generally display immature structural and functional characteristics analogous to fetal cardiomyocytes, which may facilitate the modeling of early-onset diseases but complicate drug testing and clinical uses, and potentially masking critical processes of adult-onset cardiac diseases; differentiation outcomes can vary between batches; and the reprogramming process may result in the keeping of epigenetic marks of the original somatic cells, mutations, chromosomal anomalies and genomic instability [50]. As with other experimental approaches, stem cell therapy must meet various criteria before entering the clinical trial stage, and animal studies are essential for verifying the safety of these innovative treatments. Preclinical investigations in large-animal models of MI suggest that iPSC therapy may be a promising strategy for treating cardiovascular diseases (CVD) [51].

Embryonic stem cells (ESCs)

Embryonic stem cells (ESCs) originate from the inner cell mass of an early-stage embryo known as the blastocyst. These cells are undifferentiated and have pluripotent capabilities [52]. ESC-derived cardiomyocytes (hESC-CMs) exhibit the morphology of adult heart cells, including well-organized sarcomeric proteins and autonomous beating, with atrial, ventricular, and nodal cell types [53]. However, the use of these cells in clinical trials raises particular concerns. The allogeneic origin of ESCs requires continuous immunosuppressive treatment, which carries significant risks and side effects. Moreover, the discovery that hESC-CMs may be arrhythmogenic introduces another considerable concern [54]. Therefore, Caspi et al. [55] conducted a preclinical trial in which they transplanted hESC-CMs into infarcted rat hearts. The results demonstrated that implantation of hESC-CMs promotes the formation of sustained cardiac grafts, reduces pathological remodeling, and provides functional improvements in rat models of severe MI. Further studies are needed to evaluate the safety and efficacy of this cell type under these conditions.

When should the transplantation be done?

The timing of stem cell transplantation is critical, particularly in AMI, as the myocardial microenvironment evolves rapidly, significantly impacting cell survival, engraftment, and overall therapeutic outcomes. During the first 1 to 2 days following AMI, the infarcted tissue is characterized by a pronounced inflammatory response and elevated oxidative stress, conditions that can increase apoptosis of transplanted stem cells and thereby diminish their therapeutic efficacy. In contrast, between days 6 and 9 post-infarction, the cardiac debris begins to undergo liquefactive necrosis and is subsequently absorbed [56]. Delaying stem cell transplantation beyond this window may adversely affect cell survival and differentiation, especially as fibrosis is established in the affected myocardium [17].

Clinical trials have investigated a range of transplantation windows, yielding mixed results that reflect the complexity of timing. Some studies have administered stem cells within 12 to 24 h after PCI [2]. Theoretically, this approach could be advantageous as it capitalizes on the peak release of chemotactic stimuli [57]. However, immediate injection (within hours) has often demonstrated modest or inconsistent improvement in LVEF or infarct size reduction, or even no significant improvement compared to the control group [2]. For instance, Janssens et al. delivered stem cells within 24 h following PCI but found no improvement in LVEF [58]. While the early administration is safe and feasible, the acute inflammatory phase may adversely affect cell survival or function.

Evidence from several trials suggests that the subacute phase is a promising window for cell delivery. In studies such as REPAIR-AMI, stem cells were administered an average of 4 days after AMI, with significant improvements in LVEF observed in the treatment group compared to controls [59]. The TIME [60] trial specifically examined the effect of the timing of cell delivery (three versus seven days after reperfusion) on LVEF improvement. TIME was developed in parallel with LateTIME [14] trials that investigated whether delayed delivery of cell therapy (2–3 weeks) in a similar STEMI population would also enhance recovery of LV function. However, regardless of the delivery timing, neither demonstrated a significant reduction in infarct size or improvement in LVEF compared to the placebo group. Similarly, the SWISS-AMI trial compared cell delivery on days 5 to 7 versus 3 to 4 weeks and found no treatment effect of BM-MNCs on LV functional recovery at 4 months or 1-year follow-up [61].

In contrast, some trials have suggested that intracoronary stem cell injections are better performed between 7 and 14 days after PCI [62]. This approach can be beneficial as it provides sufficient time for cell culture and for recovering the damaged myocardium and coronary arteries [18]. Additionally, a study by Kim et al. delivered BM-MSCs one month after PCI in patients with anterior wall ST-segment elevation MI and reported significant improvement in LVEF at 4 and 12 months post-treatment [63]. However, this timeframe may coincide with potentially irreversible local damage and the onset of ventricular remodeling, contradicting the findings of the previously mentioned meta-analysis.

A meta-analysis underscored that the optimal window for BMMNC transplantation is 3 to 7 days post-AMI. It was superior to transfer within 24 h or more than 7 days after AMI in improving LVEF and decreasing LV end-systolic and end-diastolic dimensions [64]. Similarly, a more recent meta-analysis demonstrated that administering MSCs within the first week of AMI correlates with notable improvements in LVEF, with an observed increase of up to 5.74% [17]. This early, though not immediate, window offers a balanced approach, avoiding the acute inflammatory burst while capitalizing on the myocardial healing process.

Despite advances in trial design, the optimal timing of stem cell transplantation is still unknown due to several factors. One significant challenge is heterogeneity in trial methodology. Variability in cell type, dosage, and delivery techniques contributes to inconsistent outcomes. It is also important to note that patient-specific factors, including baseline cardiac function, infarct size, and the extent of myocardial damage, further influence both the timing and efficacy of treatment. These multiple variables highlight the complexity of transplantation timing and warrant further standardized protocols to shed light on this matter.

How should the cells be transplanted?

The route of stem cell transplantation is one of the key factors that can influence the efficacy of stem cell therapy in myocardial infarction (MI). Although a range of delivery routes has been introduced for stem cell delivery to the myocardium, only intracoronary infusion, subendocardial injection, and intravenous delivery are applicable in the setting of acute myocardial infarctions. All other ways have been investigated in the setting of chronic ischemic or non-ischemic cardiomyopathies.

Intracoronary infusion

This is the most common approach in clinical trials. It entails infusing stem cells directly into the coronary arteries, allowing them to home in on the infarcted area. It is minimally invasive and uses the natural blood supply to deliver cells to the damaged myocardium. However, the hostile conditions of the ischemic environment can limit the efficacy of intracoronary infusion by reducing cell survival and engraftment. Intracoronary injection of leukapheresis-derived bone marrow-derived mononuclear cells (BM-MNCs) versus placebo aspiration in the REPAIR-AMI trial demonstrated an increase in left ventricular ejection fraction (LVEF) by 5.5% at 4 months and a reduction of adverse cardiac events at 12 months [59]. Similarly, the BOOST study demonstrated a 7% improvement in LVEF at 6 months after intracoronary BM-MNC infusion [12]. Conversely, ASTAMI did not observe a significant improvement in left ventricular ejection fraction (LVEF) or infarct size following intracoronary BM-MNC therapy, highlighting the discrepancy in the results [65].

Transendocardial administration

This procedure involves electrical mapping and transcatheter transendocardial direct injection. It has potential benefits, as it can detect hibernating myocardium and directly transfer cells to the needed area. Also, direct intramyocardial injection may enhance cell retention and is less invasive than CABG. In the POSEIDON trial of patients with ischemic cardiomyopathy, the intramyocardial injection of allogeneic MSCs improved LVEF and reduced scar size; however, the magnitude of the change was modest [32]. In contrast, the intramyocardial injection of autologous MSCs did not improve LVEF in patients with chronic ischemic cardiomyopathy, as reported in the TAC-HFT trial; however, it reduced scar size and improved quality of life [66]. In the meta-analyses comparing this route with intracoronary transplantations, no differences in final efficacy were observed [17].

Intravenous administration

Intravenous administration has shown promising results recently using Muse cells (multilineage-differentiating stress-enduring cells). A study by Yamada et al. showed that intravenous injection of Muse cells significantly decreased the area of damage and improved the LVEF at 2 weeks post-MI in a MI rabbit model (44% reduction and 28% increase, respectively). Additionally, these cells are selectively homed to the injured myocardium and differentiate into cardiomyocytes and vascular cells. They persisted for 6 months without inducing the need for immunosuppression, probably owing to their expression of the immunotolerant molecule HLA-G [67]. Similarly, a follow-up study by Tanaka et al. [68] demonstrated that Muse cells, administered by intravenous injection in a rabbit MI model, improved cardiac function and attenuated left ventricular remodeling, with benefits maintained for 6 months. These results indicate that Muse cells exhibit unique characteristics that enhance their ability to be delivered intravenously.

Earlier clinical studies have also evaluated the intravenous administration of allogeneic mesenchymal stem cells in post-myocardial infarction (MI) patients. In a phase I/II randomized trial, MSC was shown to be safe and well-tolerated following intravenous infusion in ST-elevation MI patients. However, improvements in cardiac function were not statistically significant [69]. Similarly, a dose-ranging trial of intravenous administration of Prochymal (allogeneic hMSCs) demonstrated a favorable safety profile and reported reduced ventricular arrhythmias and improved ejection fraction in a subset of anterior MI patients, suggesting potential therapeutic benefit [70]. However, a significant proportion of cells are sequestered in the pulmonary circulation, resulting in limited delivery efficiency and reduced myocardial retention, which diminishes therapeutic efficacy. Additionally, this approach is associated with less improvement in left ventricular ejection fraction compared to direct myocardial administration [71, 72].

Cell dose

The number of transplanted cells is a central determinant of cell-based cardiac repair efficacy. Increasing the dose can theoretically enhance myocardial retention, paracrine signaling, and neovascularization; consequently, early trials often assumed a “more is better” paradigm. Indeed, several studies have reported a dose–dependent relationship between cell number and functional recovery. For instance, an escalation in cell dose (≥ 10⁸ cells) has been associated with greater improvement in left ventricular ejection fraction (LVEF) and a reduction in infarct size, suggesting that higher engraftment may amplify myocardial repair processes by increasing the secretion of pro-angiogenic and anti-apoptotic cytokines.

The REPAIR-AMI trial, which infused 200 million bone marrow–derived mononuclear cells (BM-MNCs), demonstrated a statistically significant improvement in LVEF at four months and a concomitant reduction in major adverse cardiac events [59]. In contrast, the ASTAMI trial, using a comparatively lower dose of 68 million BM-MNCs, did not detect significant differences in either LVEF or infarct volume versus placebo [65]. These discrepancies highlight the importance of cell dose as a driver of reproducible outcomes.

However, supraphysiologic doses may yield paradoxical effects. Extremely high cell concentrations can increase blood viscosity and microvascular plugging, precipitating transient ischemia in the microcirculation. Moreover, excessive intracoronary delivery may trigger arrhythmic events or inflammatory responses that offset potential benefits [73]. The TRIDENT trial, designed to interrogate the dose–response curve of mesenchymal stem cell (MSC) therapy, provided key insights into this non-linear relation. Surprisingly, patients receiving 20 million MSCs exhibited larger gains in LVEF than those treated with 100 million cells, implying that saturation or negative feedback mechanisms may emerge at higher doses [74].

Collectively, evidence suggests a U-shaped dose–response curve in cardiac cell therapy: too few cells may be insufficient to produce meaningful paracrine signaling, while excessively high doses risk microvascular obstruction and pro-arrhythmic complications. Ongoing studies employing dose-escalation designs and advanced delivery methods (e.g., intramyocardial versus transendocardial injection) aim to identify an optimal biological dose that maximizes therapeutic efficacy without compromising safety.

Safety issues

Across the 42 randomized controlled trials identified in this review, cell delivery after AMI was consistently reported to be procedurally safe, with no study reporting an excess of periprocedural death, coronary dissection, or sustained ventricular arrhythmia attributable to the cell product itself. The early BM-MNC trials that opened the field (Strauer et al. [10], TOPCARE-AMI [4], and BOOST [12, 75]) explicitly stated that adjunctive cell infusion in conjunction with primary PCI was feasible and did not worsen infarct-related artery patency or trigger acute inflammatory reactions. These safety impressions were reproduced in later, methodologically stricter trials. The second wave of studies that tested timing and dose (TIME [13], LateTIME [14], SWISS-AMI [61]) likewise failed to show procedure-related safety penalties, even though they deliberately infused cells outside the “very acute” window, suggesting that the absence of excess harm is robust to moderate shifts in timing and to small–moderate differences in cell number. In line with our findings, a meta-analysis by Moeswir et al. [76] showed that the pooled adverse event rate was even lower in the cell groups (OR 0.66, 95% CI 0.44–0.99). No cardiac-related malignancies were reported, but the authors stressed that follow-up was short-term and mid-term, and longer surveillance is needed. Although most studies in this meta-analysis used BM-MNCs, another meta-analysis focusing on MSC therapy in diverse type of disorder, not only cardiovascular, found no excess in major events, including death 0.99 (95% CI 0.66–1.49), infection 1.03 (95% CI 0.70–1.53), vascular events 1.17 (95% CI 0.52–2.62), CNS disorders 1.13 (95% CI 0.61–2.12), arrhythmia 0.62 (95% CI 0.36–1.07; trend to lower), dermatitis/urticaria 0.93 (95% CI 0.38–2.26). Notably, MSCs were clearly associated with transient fever 3.65 (95% CI 2.05–6.49) and administration-site reactions 1.98 (95% CI 1.01–3.87), plus smaller increases in constipation 2.45 (95% CI 1.01–5.97), fatigue 2.99 (95% CI 1.06–8.44) and sleeplessness 5.90 (95% CI 1.04–33.47) [76]. Almost all of these manifestations were mild and occurred within 48 h, and were slightly more common in younger/cardiac populations and in early-phase studies, which seems to be mostly related to the administration procedure rather than the stem cells themselves. The authors still recommend long-term surveillance because MSCs can theoretically aggregate or be immunogenic. Additionally, no new systematic safety signal emerged from cardiac-progenitor sources and the allogeneic cardiosphere-derived products, such as those exemplified by ALLSTAR [16, 35]. Because the included trials mixed several cell sources (BM-MNCs, MSCs, CSCs/progenitors, and a few exploratory iPSC-related approaches) and several delivery routes, the convergence of their findings is essential: cell therapy in the context of reperfused AMI is consistently non-inferior to standard care in terms of safety. It can therefore be considered clinically deployable, with the main limitations being the short follow-up in the published trials and the heterogeneity in reporting of late events.

Discussion

How choosing the trial outcomes affects the results

When designing a clinical trial, the most critical issue is determining the primary endpoint and how it will be measured. Almost all trials in regenerative cardiology that have focused on patients with AMI have used surrogate endpoints such as LVEF or scar size, except for the BAMI trial [77]. In clinical trials, surrogate endpoints are variables that serve as substitutes for direct clinical outcomes. These are often laboratory measures, imaging results, or biomarkers expected to predict the effect of a therapy on meaningful clinical outcomes. While using surrogate endpoints can make trials faster and simpler, there are several significant reasons why they cannot fully replace clinical endpoints. Consequently, only interventions proven in a trial with a clinical endpoint are considered for treatment planning in clinical practice. First of all, there is an Indirect Relationship with True Clinical Outcomes, and Surrogate endpoints do not always accurately reflect the actual clinical effect of an intervention. A treatment may improve a surrogate marker but have no meaningful impact on clinical outcomes (such as mortality or quality of life). In addition, relying on surrogate endpoints may lead to the approval of treatments that are actually ineffective or even harmful. A classic example is class I antiarrhythmic drugs, which reduced ventricular arrhythmias (the surrogate end point) but increased mortality (the clinical end point).

That is what has happened in the field of regenerative cardiology. Although many trials using surrogate endpoints, such as LVEF or scar size, reached statistical significance, none have directly shown reductions in clinical endpoints, including major cardiovascular events (MACE). In that case, the BAMI is an exception.

The BAMI trial was the first phase III study to assess whether intracoronary transplantation of BM-MNCs after MI could reduce all-cause mortality. Initially designed for 3,000 patients, it was stopped early after enrolling only 375. Of these, 185 received BM-MNCs via intracoronary infusion 2–8 days post-primary PCI, while 190 received standard medical therapy as controls. After two years, all-cause mortality was 3.26% (6 deaths; 95% CI: 1.48–7.12%) in the BM-MNC group, compared to 3.82% (7 deaths; 95% CI: 1.84–7.84%) among controls. The unexpectedly low mortality was a key factor in these results. When the trial began in 2011, literature indicated a 12% two-year mortality rate post-AMI for patients with LVEF ≤ 45% after reperfusion [77]. However, during the study, a 3.85% mortality rate was observed, likely due to improvements in primary angioplasty. Notably, only five patients (2.7%, 95% CI: 1.0–5.9%) who received BM-MNCs were hospitalized for heart failure during follow-up, versus 15 patients (8.1%, 95% CI: 4.7–12.5%) in the medical therapy group (HR: 0.33, 95% CI: 0.12–0.88). This was the only significant clinical benefit identified. The BAMI trial suggested that relying on mortality as a primary endpoint in stem cell therapy studies can be challenging with moderate sample sizes. Heart failure incidence might be a more suitable primary endpoint. A recent meta-analysis confirmed that BM-MNC injection lowers the risk of combined outcomes, such as hospitalization for CHF, reinfarction, and cardiac death (91/1191 vs. 111/812; RR = 0.643, 95% CI: 0.489–0.845; p = 0.002). These benefits primarily resulted from a reduction in CHF episodes (47/1220 vs. 62/841; RR = 0.568, 95% CI: 0.382–0.844; p = 0.005) and lower reinfarction rates (23/1159 vs. 30/775; RR = 0.583, 95% CI: 0.343–0.991; p = 0.046). Cardiac-related mortality was not significantly affected (28/1290 vs. 31/871; RR = 0.722; 95% CI: 0.436–1.197; p = 0.207) (Fig. 6).

Fig. 6.

Forest Plots demonstrating the efficacy of transplantation of bone marrow-derived myonuclear cells (BM-MNCs) on various clinical outcomes

How choosing outcomes measurement tools affects results

Left ventricular ejection fraction (LVEF) is the percentage of blood volume that is pumped from the left ventricle to the aorta. This parameter is the most widely used in cardiology and serves as an endpoint for several clinical trials. Based on LVEF values, patients are categorized into three subgroups of normal (Preserved; LVEF ≥ 50%), mildly reduced (40%≤LVEF < 50%), and severely reduced cardiac function (LVEF < 40%). LVEF can be measured using several techniques, with echocardiography being the most widely used method. Although cardiac MRI (CMR) is the gold standard, it is rarely used in clinical practice due to its high cost, limited availability, and time-consuming nature. Other techniques, such as Single-photon emission computed tomography (SPECT) and LV angiography, are not routinely used nowadays.

Another critical issue is how the primary endpoint has been assessed in the clinical trial. For example, when measuring LVEF, different modalities would yield different results. Fisher et al., in a meta-analysis, proved that BM-MNCs augment the left ventricular ejection fraction (LVEF) following AMI by roughly 2.72% when measured by echocardiography, 5.63% if measured by SPECT, 6.43% if measured by angiography, and no change if measured by CMR [2].

Limitations of stem cell therapy

Once first introduced, Stem cell therapy and regenerative medicine were believed to become a promising new tool in cardiology. The primary hope was that the transplanted stem cells could migrate and differentiate into myocytes, replacing the lost and scarred myocardial tissue. However, as the clinical trials continued, this hope gave way to disappointment. After raising disappointments, through indirect evidence, some new hopes again emerged, and the story went on with rises and falls. Overall, the field is neither promising nor disappointing. We now know that the stem cells currently under study do not home to or transdifferentiate into myocardial tissue. Instead, it exerts indirect effects through paracrine signaling and modulates myocardial inflammation, leading to reduced scar size and prompting endogenous cardiac stem cells to replace lost cells. In light of that, all interpretations and future trial designations should be made with these facts in mind.

Another essential aspect is shifting trials from small-scale, surrogate-endpoint studies to large-scale phase III studies with clinical endpoints that can help change daily practices. In this context, all direct attempts have failed, and only subgroup or secondary analyses have shown some effects. For example, in the BAMI trial, a reduction in heart failure incidence, a secondary endpoint, was observed [77], and this effect was further confirmed by meta-analyses [20].

Finally, it should be noted that not all patients will benefit from stem cell therapy. For example, patients whose baseline LVEF is above 40%, especially those with more than 50% LVEF, are already nearing full recovery, and there is no further gain that stem cell therapy can offer them. This may be one of the fundamental reasons for failures in some trials.

Conclusion and future research direction

Stem cell therapy has raised some hopes, adding to the conventional treatment of AMI. However, its clinical application is currently impeded by a lack of standardization in study designs and variability in factors such as cell sources, cell dosage, timing, and delivery route. Therefore, to fully exploit the capabilities of stem cell therapy in AMI, it is essential to standardize protocols for identifying the most effective cell populations and optimizing delivery techniques. Patient selection is key to achieving better response, and those with a severely reduced LVEF (LVEF < 40%) and younger age tend to gain greater benefit. Limiting the timing of transplantation to the first 3–7 days post-AMI may increase the efficacy of this intervention. Current evidence suggests that WJ-MSCs are the best available and studied source for AMI stem cell therapy. Innovative delivery strategies, such as biomaterial scaffolds and cell patches, may enhance cell retention and survival in the endangered myocardium. Table 2 demonstrates the ongoing trials in the field.

Table 2.

Ongoing clinical trials of cell therapy in patients with acute myocardial infarction

NCT Number	Trial name	Phase	Cell therapy type	Sample size	Primary outcome	Route of administration
NCT00275977	–	Phase 1	Autologous bone marrow derived stem cells	10	Safety	Coronary catheterization and stem cell infusion
NCT05669144	–	Phase 1/Phase 2	MSC-derived exosomes + autologous mitochondria	20	Ejection fraction, allergic reactions	Intracoronary and intra-myocardial injection
NCT00984178	TECAM2	Phase 2	Bone marrow mononuclear cells ± G-CSF mobilization	120	Change in left ventricular ejection fraction and end-systolic volume	Intracoronary transplantation; G-CSF subcutaneous injection
NCT01394432	ESTIMATION	Phase 3	Autologous mesenchymal stem cells (endocardial implantation after PCI)	50	Reduction in left ventricle systolic volume ≥ 15% (MRI)	Endocardial injection via NOGA system
NCT06258447	–	Not Applicable	Autologous bone marrow mononuclear cells (ABM MNC) delivered via CardiAMP cell therapy system	250	Composite efficacy endpoint (death, MACCE events, quality of life by MLHFQ	Intramyocardial injection via trans endocardial catheter
NCT01625949	COAT	Not Applicable	Autologous bone marrow stem cells injected post-PCI	40	Change in left ventricular function and myocardial viability (ECHO, PET)	Intracoronary injection
NCT00437710	CARDIAC	Phase 1/2	Autologous bone marrow–derived mononuclear cells	50	Safety and efficacy of intracoronary bone marrow cell transplantation after MI	Intracoronary injection
NCT06147986	–	Phase 2	Allogeneic umbilical cord mesenchymal stem cells (UMSC01)	41	Incidence of AEs, SAEs, and SUSARs; changes in VO₂ (CPET)	Intracoronary + Intravenous
NCT00725738	TRACIA	Phase 2/3	Autologous bone marrow–derived CD34⁺ stem cells	80	Change in LVEF at 6 months; Major adverse cardiovascular events	Intracoronary
NCT06364150	MAGICcell6	Phase 3	Autologous angiopoietin-primed peripheral blood stem cells	30	LVEF at 12 months	Intracoronary injection
NCT00501917	MAGIC Cell-5	Phase 2/Phase 3	G-CSF ± darbepoetin–mobilized peripheral blood stem cells	116	Change in left LVEF by MRI	Intracoronary infusion
NCT01536106	AMIRST	Phase 1 & 2	Concentrated autologous bone marrow mononuclear cells	30	Number of adverse events (safety outcome)	Intracoronary infusion of cells through intraoperative point of care
NCT02666391	SEESUPIHD	Phase 1 & 2	Umbilical cord mesenchymal stem cells	64	Change in global LVEF by echocardiography and infarct size and myocardial viability measured by ECT	Intracoronary infusion during percutaneous coronary intervention
NCT01652209	–	Phase 3	Autologous bone marrow-derived mesenchymal stem cells	90	LVEF at 13 months after cell treatment	Injected into the infarct coronary artery using balloon tipped catheter
NCT04551443	WAIAMI	Phase 2	Wharton’s jelly-derived MSCs	200	Composite of major adverse cardiovascular events, LVEF, infarct size, and perfusion defect	Intravenous infusion
NCT02504437	TPAABPIHD	Phase 1 & 2	autologous bone marrow mesenchymal stem cells with or without hypoxia pre-condition and endothelial pre-induction	200	LVEF at 12 months	Unknown
NCT00936819	ENACT-AMI	Phase 2	Autologous early endothelial progenitor cells (EPCs)	47	Global LVEF at 6 months	Stem cells administered into the infarct related artery
NCT07134712	SEMIAMI	Phase 2	Umbilical cord blood derived nucleated cells	100	All adverse events, development of GVHD, and cardiovascular composite events	Peripheral intravenous injection
NCT00555828	–	Phase 1 & 2	Allogeneic mesenchymal precursor cells	25	Safety and feasibility at 30 days	Transendocardial delivery using the Cordis Biosense NogaStarTM mapping catheter
NCT05554484	AMI-DC	Phase 1 & 2	Autologous peripheral blood-derived tolerogenic dendritic cells	30	Adverse cardiovascular events and cumulated incidence of MACE at 6 months	Subcutaneous administration at 1–4 sited in the left axillary regions
NCT03902067	–	Phase 1	Umbilical cord mesenchymal stem cells	40	MACE	Catheter transplantation
NCT03047772	TEAM-AMI	Phase 2	Autologous bone marrow mesenchymal stem cells	124	Change in LVEF after 12 months	Intracoronary infusion
NCT02323620	RACE-STEMI	Phase 3	Autologous bone marrow-derived mononuclear cells	200	LVEF 12-month change measured by computed tomography	Intracoronary infusion

Another important avenue is combining stem cell therapy with other regenerative therapies, such as gene therapy or pharmacological agents, to enhance therapeutic effects. For instance, the Akt-MSC study showed that preconditioning MSCs with the Akt gene enhances their survival and paracrine activity, resulting in improved cardiac function [78]. With advances in genomic and proteomic technologies, tailored stem cell treatments can be designed based on a patient’s individual genetic and molecular profile. Finally, emerging cell types such as Muse cells show promise, but their potential must be confirmed through larger trials involving Muse and other novel stem cells [79].

Abbreviations

AMI

Acute myocardial infarction

HF

Heart failure

CHF

Congestive heart failure

BM-MNC

Bone marrow mononuclear cell

MSC

Mesenchymal stem cell

LVEF

Left ventricular ejection fraction

LVEDV

Left ventricular end-diastolic volume

LVESV

Left ventricular end-systolic volume

STEMI

ST elevation myocardial infarction

SD

Standard deviation

RR

Risk ratio

CI

Confidence interval

WMD

Weighted mean difference

ICEMAN

Instrument for assessing the Credibility of Effect Modification of Analyses

CMR

Cardiac magnetic resonance

SPECT

Single-photon emission computed tomography

PCI

Percutaneous coronary intervention

Author contributions

Ar.At designed the review. H.M, V.K, Al.Ar, SA.M., A.KJ contributed to performing the necessary searches for the study. A.H performed statistical analyses. All contributed to preparing the first draft of the manuscript, and AA finalized the version. All the authors accept responsibility for what is submitted.

Funding

No funding has been received for this study.

Data availability

This is a systematic review and therefore does not contain any outcome data to be shared.

Declarations

Ethics approval and consent to participate

Not applicable.

AI use

We utilized an AI tool (Grammarly) to enhance the quality of the English language in the manuscript.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.