Impact of mesenchymal stem cell therapy on cardiac function and outcomes in acute myocardial infarction: A meta-analysis of clinical studies

Abstract

This meta-analysis evaluated the efficacy of mesenchymal stem cell (MSC) treatment on cardiovascular function and major adverse cardiac events (MACE) in patients with acute myocardial infarction (AMI) at various follow-up intervals. Clinical studies comparing MSC therapy with control treatments for AMI were identified from databases including Cochrane, Web of Science, PubMed, Embase, CNKI, and Wanfang, covering publications up to August 2024. Data analysis was conducted using Review Manager 5.4 software. MSC treatment significantly improved left ventricular ejection fraction (LVEF) compared to controls at follow-up intervals <6 months (MD = 3.42; P < 0.0001), 6 months (MD = 4.15; P = 0.006), and 12 months (MD = 2.77; P = 0.006). However, no significant effect on LVEF was observed after 12 months (MD = 3.50; P = 0.17). MSC therapy did not significantly affect left ventricular end-diastolic volume (LVEDV) at any interval. Left ventricular end-systolic volume (LVESV) significantly decreased only within the first 6 months (MD = −11.35; P = 0.11) but not at subsequent follow-ups. Wall motion score index (WMSI) significantly improved at <6 months (MD = −0.06; P < 0.0001), 6 months (MD = −0.04; P = 0.006), and >12 months (MD = −0.03; P = 0.02). However, the improvement at 12 months was borderline significant (MD = −0.06; P = 0.06). MSC therapy showed no significant reduction in MACE (odds ratio [OR] = 1.61; P = 0.10). Subgroup analyses indicated intracoronary MSC administration notably improved LVEF (MD = 4.27; P < 0.0001), while intravenous MSC administration showed no significant effect. Neither administration route significantly affected MACE outcomes. No publication bias was detected. In conclusion, MSC therapy significantly enhances LVEF and WMSI within the first 12 months post-AMI, with intracoronary administration showing greater efficacy than intravenous delivery. However, MSC treatment did not significantly reduce MACE incidence. Further rigorous clinical trials are needed to confirm these findings.

Graphical Abstract

Keywords

mesenchymal stem cell acute myocardial infarction left ventricular ejection fraction left ventricular end-diastolic volume left ventricular end-systolic volume major adverse cardiac eventsIntroduction

Coronary artery disease is a major cause of mortality, with acute myocardial infarction (AMI) being a major contributor to the increasing prevalence of heart failure and mortality¹. AMI, a frequent cardiovascular emergency, is characterized by prolonged myocardial ischemia, hypoxia, and necrosis due to the narrowing or blockage of coronary arteries^2,3. Clinically, it manifests as severe, persistent chest pain behind the sternum lasting over 30 minutes, often radiating to the back, jaw, and upper limb^4,5. Certain people might encounter lightheadedness, queasiness, vomiting, and difficulty breathing. Patients typically show dynamic electrocardiogram (ECG) changes and elevated cardiac enzymes, with complications like arrhythmias, heart failure, and shock posing life-threatening risks^4,5. Risk factors for AMI include genetic predisposition, smoking, high cholesterol, hypertension, diabetes, older age, and male gender^6–8. Diagnosis is generally based on clinical symptoms, ECG results, cardiac biomarkers, and coronary angiography^6–8. Advances in drug therapy and interventional techniques have significantly reduced the acute mortality rate for myocardial infarction patients^6–8. However, the incidence of heart failure postmyocardial infarction remains high. Therefore, improving cardiac function after percutaneous coronary intervention (PCI) and enhancing long-term survival rates for AMI patients is a pressing issue⁹.

Mesenchymal stem cells (MSCs) are a variety of stromal cells that can be isolated from various adult tissues, including bone marrow, umbilical cord, and adipose tissue¹⁰. Recently, they have attracted attention in various medical fields due to their potential therapeutic benefits. MSCs are mature stem cells capable of differentiating into various cell types and demonstrating considerable plasticity^11–13. Under specific conditions, they can transform into mesodermal and ectodermal tissue cells, including adipocytes, osteocytes, chondrocytes, skeletal myocytes, cardiomyocytes, smooth muscle cells, hepatocytes, and neuron-like cells^11–13. MSCs can be conveniently and safely harvested with minimal harm and are easy to expand in vitro, making them suitable for autologous transplantation^11–13. They have the capacity to repair and regenerate various functional cells, making them important “seed cells” for tissue engineering^11–13. Several prior studies have demonstrated that after MSC transplantation in patients with cardiovascular diseases, these stem cells can colonize the infarcted myocardial tissue, directly proliferating, and differentiating into cardiomyocyte-like cells and vascular endothelial cells^14–16. They can also secrete various vascular endothelial growth factors through paracrine mechanisms to promote blood vessels formation, enhance myocardial contractility, and reduce myocardial fibrosis while preventing left ventricular remodeling^14–16. Currently, MSCs are used in treating various heart conditions, including heart failure and myocardial infarction^17,18.

MSC transplantation in patients with AMI is typically performed using three methods: intracoronary injection, intramyocardial injection, and intravenous injection^19,20. Intravenous infusion often results in low cell survival rates within the myocardium, which can greatly impact the effectiveness of the treatment^19,20. Research indicates that direct intramyocardial injection of stem cells is less effective than intracoronary injection and may cause uneven distribution of stem cells, leading to dys-synchrony in ventricular wall motion²¹. In addition, intramyocardial injection carries a higher risk of triggering malignant arrhythmias. Consequently, intracoronary injection of stem cells is considered the most optimal transplantation method²¹. To our best knowledge, most studies reported small sample sizes or conflicting results^22–35, making it difficult to draw a reliable and authoritative conclusion for clinical guidance. Hereby, we carried out a meta-analysis to summarize efficacy and safety of MSC therapy in patients with AMI and to investigate the immediate and prolonged impacts of MSC on patients with AMI.

Methods

Literature retrieval strategy

According to the principles and relevant guidelines proposed by the Cochrane Handbook, we conducted a computer search in databases including “PubMed,” “The Cochrane Library,” “Embase,” “Web of Science,” “CNKI,” and Wanfang. The search aimed to identify clinical studies on MSCs therapy for treating AMI patients, covering period from inception of databases up to August 2024. Search terms or keywords in English were: “mesenchymal stem cells,” “mesenchymal stromal cells,” “stem cell(s),” “cell transplantation” “acute myocardial infarction,” “AMI,” “myocardial infarction,” “MI,” and “coronary artery disease.”

Inclusion and exclusion criteria

Inclusion criteria were (1) clinical studies with a minimum follow-up of 2 months. (2) Both experimental and controls received standard treatment, with the control group not receiving stem cell transplantation. (3) Patients clinically diagnosed with AMI who underwent PCI. (4) Studies in which the experimental group received cell therapy within 3 months post-PCI. (5) MSCs sourced from human bone marrow, umbilical cord, or adipose tissue. (6) Primary endpoints included left ventricular ejection fraction (LVEF), with secondary endpoints being left ventricular end-diastolic volume (LVEDV) and left ventricular end-systolic volume (LVESV). (7) Only MSCs were used for transplantation, with no restrictions on cell dose or administration timing.

The exclusion criteria were: (1) Studies involving cytokine interventions. (2) Trials lacking a control group. (3) Research involving specific populations, such as pregnant women. (4) Studies that did not assess any of the primary or secondary endpoints. (5) Studies with incomplete data or duplicate publications.

Literature selection and data extraction

Two reviewers independently performed the literature search based on the defined criteria. All selected literature was evaluated according to the inclusion and exclusion criteria, removing irrelevant trials, reviews, or nonclinical studies. Any disagreements or differing viewpoints were addressed through discussion and consultation with a third reviewer.

Two researchers independently extracted data from the included studies, ensuring accuracy through cross-checking. Discrepancies were resolved via discussion or by consulting a third reviewer. The extracted data included: general study details (author, publication year, number of cases, demographic information, follow-up duration, and baseline LVEF), the exact number of MSCs transplanted, and the time elapsed from PCI to transplantation. In addition, measurement methods including echocardiography, magnetic resonance imaging (MRI), SPECT, positive emission tomography (PET) and angiography, and outcome measures, such as, LVEF, LVEDV, LVESV, wall motion score index (WMSI), and major adverse cardiac events (MACE) were also recorded.

Quality assessment

Two researchers assessed the quality of the included studies across seven dimensions. Any disagreements or differing views were resolved through discussion and consultation with a third reviewer. The evaluation criteria were based on the quality assessment standards for RCTs outlined in the Cochrane Handbook. This assessment covered aspects such as random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, completeness of outcome data, selective reporting, and other potential biases. The results were classified as low risk of bias, high risk of bias, or unclear risk of bias.

Statistical analysis

Data from the included studies were organized and analyzed using Review Manager 5.3. For continuous variables, the mean difference (MD) was used as the effect size in studies employing the same statistical methods, helping to adjust for variations in absolute values across studies. For categorical variables, the odds ratio (OR) was used. Statistical significance was defined as P < 0.05. Heterogeneity among studies was evaluated using chi-square tests to compute P and I² values. If P > 0.1 and I² < 50%, a fixed-effect model was applied for the meta-analysis; otherwise, a random-effect model was used after investigating sources of heterogeneity. Funnel plots were employed to evaluate potential publication bias.

Results

Study flow

Meta-analysis flow is illustrated in Figure 1. Initially, 1,475 articles were retrieved, with an extra 27 obtained from other sources. Of these, 983 records were excluded. The remaining 62 studies were reviewed in detail, resulting in the inclusion of 13 studies involving 727 patients (360 in the experimental group and 367 in the control group). Table 1 provides a summary of the characteristics of these included studies, which were published between 2004 and 2023.

Figure 1.

Flow diagram of the meta-analysis.

Table 1.

Characteristics of the included studies for meta-analysis.

Author	Year of publication	Sample size		Cell types	Cell origin	Route of cell administration	Method of measurement	Follow-up period
Author	Year of publication	Control	Experimental	Cell types	Cell origin	Route of cell administration	Method of measurement	Follow-up period
Attar et al²²	2023	25	20	Wharton’s jelly mesenchymal stromal cells	Allogeneic	Intracoronary	Echocardiography, CMR	6 months
Chen et al²³	2004	35	34	Autologous bone marrow mesenchymal stem cell	Autologous	Intracoronary	Echocardiography, PET	6 months
Chullikana et al²⁴	2015	10	10	Bone marrowe derived allogeneic mesenchymal stromal cells	Allogeneic	Intravenous	Echocardiography, MRI, SPECT	24 months
Gao et al²⁵	2013	20	19	Bone marrow mesenchymal stem cells	Autologous	Intracoronary	Echocardiography, SPECT	24 months
Gao et al²⁶	2015	55	57	Wharton’s jelly–derived mesenchymal stem cells	Allogeneic	Intracoronary	Echocardiography, SPECT	18 months
Grajek et al²⁷	2010	14	31	Bone marrow mesenchymal stem cells	Autologous	Intracoronary	Radionuclide ventriculography, SPECT	12 months
Hare et al²⁸	2009	21	39	Allogeneic human mesenchymal stem cells	Allogeneic	Intracoronary	Echocardiography, MRI	12 months
Houtgraaf et al²⁹	2012	4	9	Adipose tissue derived regenerative cells	Autologous	Intracoronary	CMR, SPECT	6 months
Kim et al³⁰	2018	12	14	Bone marrow mesenchymal stem cells	Autologous	Intracoronary	Echocardiography, SPECT	12 months
Lee et al³¹	2014	28	30	Bone marrow mesenchymal stem cells	Autologous	Intracoronary	Echocardiography, SPECT	6 months
Mathur et al³²	2022	46	39	Autologous bone marrow mesenchymal stem cells	Autologous	Intracoronary	CMR, angiography	60 months
Rodrigo et al³³	2013	45	9	Bone marrow mesenchymal stem cells	Autologous	Intracoronary	Echocardiography, SPECT	60 months
Wang et al³⁴	2014	30	28	Bone marrow mesenchymal stem cells	Autologous	Intracoronary	Echocardiography, angiography	6 months
Zhang et al³⁵	2021	22	21	Bone marrow mesenchymal stem cells	Autologous	Intracoronary	Echocardiography	12 months

Quality assessment of included literature

Risk bias assessment indicated varying degrees of uncertainty and potential bias among the eligible studies. Specifically: Four studies were categorized as having an “unclear risk” of selection bias. One study was classified as having a “high risk” of selection bias due to insufficient descriptions of random sequence generation. Five studies had an “unclear risk” of selection bias due to a lack of clarity regarding allocation concealment. Three studies were labeled as having an “unclear risk” and one as having a “high risk” due to insufficient details concerning the blinding of participants and personnel. Three studies were deemed to have an “unclear risk” of detection bias because of incomplete information on the blinding of outcomes assessment. Seven studies were categorized as having an “unclear risk” of other forms of bias. A visual representation of the quality assessment of the eligible clinical studies is shown in Figure 2.

Figure 2.

Quality assessment analysis: (a) risk of bias graph and (b) risk of bias summary.

LVEF assessment at different time points

For LVEF assessed at <6 months, eight studies compared LVEF between experimental (234 cases) and control groups (229 cases). The results indicated heterogeneity in studies (I² = 81%, P < 0.00001); thus, a random effect was used. Meta-analysis showed a notable difference in LVEF < 6 months between these groups [MD = 3.42, P < 0.0001, 95% CI (1.86, 4.98)] (Figure 3a). This indicates a substantial elevation in LVEF in experimental group within 6 months following treatment compared to controls.

Figure 3.

Forest plot of meta-analysis of LVEF: (a) Forest plot of LEVF <6 months; (b) Forest plot of LVEF at 6 months; (c) Forest plot of LVEF at 12 months; (d) Forest plot of LVEF >12 months.

For LVEF assessed at 6 months, 10 studies compared LVEF between experimental (227 cases) and control groups (223 cases). The results showed heterogeneity among studies (I² = 94%, P < 0.00001); thus, a random-effects model was used. The meta-analysis revealed a notable difference in LVEF at 6 months between these groups [MD = 4.15, P = 0.006, 95% CI (1.18, 7.11)] (Figure 3b), indicating a substantial increase in LVEF in the experimental group at 6 months following treatment.

For LVEF assessed at 12 months, seven studies compared LVEF between experimental (227 cases) and control groups (223 cases). The results indicated heterogeneity for included studies (I² = 94%, P < 0.00001); thus, we used a random effect. Meta-analysis showed a notable difference in LVEF at 12 months between these groups [MD = 2.77, P = 0.006, 95% CI (0.78, 4.75)] (Figure 3c). This indicates a substantial elevation in LVEF in experimental group at 12 months following treatment compared with controls.

For LVEF assessed >12 months, three studies compared LVEF between experimental (86 cases) and control groups (85 cases). The results showed heterogeneity for included studies (I² = 98%, P < 0.00001); thus, we used a random-effects model. There was no significant difference in LVEF >12 months between these groups [MD = 3.50, P = 0.17, 95% CI (−1.51, 8.50)] (Figure 3d), indicating no changes in LVEF in experimental group comparing with controls at >12 months following treatment.

Funnel plots of the included studies did not show discernible asymmetry, indicating minimal evidence of publication bias (Supplemental Figures S1A–S1D); however, the small number of heterogeneous trials and variable reporting standards limit its sensitivity to detect publication bias.

Subgroup analysis based on administration route revealed that intracoronary administration of MSCs led to a significant enhancement in LVEF at 6 months [MD = 4.27, P < 0.0001, 95% CI (1.17, 7.37)] compared to the control group (Supplemental Figure S2A). Conversely, intravenous infusion did not lead to a substantial improvement in LVEF comparing with controls [MD = 2.55, P = 0.49, 95% CI (−4.76, 9.86)] (Supplemental Figure S2A). For the subgroup analysis of LVEF >12 months, intracoronary infusion of MSCs failed to significantly increase LVEF comparing with controls [MD = 3.12, P = 0.28, 95% CI (−2.56, 8.81)] (Supplemental Figure S2B). Consistent findings were observed in patients receiving intravenous infusion of MSCs [MD = 3.50, P = 0.17, 95% CI (−1.51, 8.50)] (Supplemental Figure S2B).

When stratified by cell origin, allogeneic MSCs, but not autologous MSCs, significantly improved LVEF at <6 months [allogeneic: MD = 4.40; P < 0.0001; 95% CI (3.96, 4.85); autologous: MD = 2.96, P = 0.05, 95% CI (–0.02, 5.94)] versus controls (Supplemental Figure S3A). In subgroup analysis, allogeneic MSCs continued to confer a significant benefit [MD = 5.04, P < 0.0001, 95% CI (4.10, 5.99)], whereas autologous MSCs did not reach significance (MD = 4.04, P = 0.08, 95% CI (–0.43, 8.51)) compared with controls (Supplemental Figure S3B). However, neither allogeneic [MD = 3.43; P = 0.11; 95% CI (–0.78, 7.64)] nor autologous MSCs [MD = 2.12; P = 0.05; 95% CI (0.04, 4.21)] achieved statistical significance versus controls at 12 months (Supplemental Figure S3C).

In a further analysis, combining all outcomes regardless of follow-up duration showed heterogeneity for included studies (I² = 97%, P < 0.00001); thus, we used a random-effects model. Meta-analysis demonstrated a notable difference in LVEF between these groups [MD = 3.52, P < 0.0001, 95% CI (2.39, 4.65)] (Table 2), indicating a substantial increase in LVEF in experimental group following MSC treatment comparing with controls.

Table 2.

Meta-analysis of LVEF, LVEDV, LVESV, and WMSI.

Outcomes	Sample size		Mean difference	Analysis method	Tau²	Chi²	df	I ²	Overall effect
Outcomes	Experimental	Control	IV (95% CI)	Analysis method	Tau²	Chi²	df	I ²	Overall effect
LVEF	729	714	3.52 (2.39, 4.65)	Random effect	6.4	860.91	27 (P < 0.0001)	97%	Z = 6.13 (P < 0.0001)
LVEDV	588	573	–13.19 (–19.05, –7.33)	Random effect	143.31	1088.17	20 (P < 0.0001)	98%	Z = 4.41 (P < 0.0001)
LVESV	588	573	–12.40 (–17.00, –7.79)	Random effect	94.29	1851.7	20 (P < 0.0001)	99%	Z = 5.28 (P < 0.00001)
WMSI	401	435	–0.06 (–0.06, –0.05)	Fixed effect	-	18.59	13 (P = 0.14)	30%	Z = 25.96 (P < 0.00001)

LVEDV assessment at different time points

For LVEDV assessed at <6 months, six studies compared LVEDV between experimental (183 cases) and control groups (175 cases). The results indicated heterogeneity for included studies (I² = 94%, P < 0.00001); thus, we used a random-effects model. Meta-analysis showed no difference in LVEDV <6 months between these groups [MD = −11.56, P = 0.11, 95% CI (−25.70, 2.59)] (Figure 4a), indicating no changes in LVEDV in the experimental group within 6 months following treatment compared to controls.

Figure 4.

Forest plot of meta-analysis of LVEDV: (a) Forest plot of LVEDV <6 months; (b) Forest plot of LVEDV at 6 months; (c) Forest plot of LVEDV at 12 months; and (d) Forest plot of LVEDV >12 months.

For LVEDV assessed at 6 months, six studies compared LVEDV between experimental (147 cases) and control groups (146 cases). The results indicated heterogeneity for included studies (I² = 99%, P < 0.00001); thus, we carried out a random-effects model. Meta-analysis exhibited no difference in LVEDV at 6 months between these groups [MD = −21.10, P = 0.17, 95% CI (−51.54, 9.33)] (Figure 4b), indicating no changes in LVEDV in the experimental group at 6 months following treatment compared to controls.

For LVEDV assessed at 12 months, seven studies compared LVEDV between experimental (182 cases) and control groups (177 cases). The results indicated heterogeneity for included studies (I² = 94%, P < 0.00001); thus, we used a random-effects model. Meta-analysis showed no significant difference in LVEDV at 12 months between these groups [MD = −5.69, P = 0.23, 95% CI (−15.08, 3.69)] (Figure 4c), indicating no changes in LVEDV in the experimental group at 12 months following treatment compared to the control group.

For LVEDV assessed >12 months, two studies compared LVEDV between experimental (76 cases) and control groups (75 cases). The results indicated heterogeneity among studies (I² = 96%, P < 0.00001); thus, we used a random-effects model. No difference was detected in LVEDV >12 months between these groups [MD = −8.47, P = 0.12, 95% CI (−19.05, 2.11)] (Figure 4d), indicating no changes in LVEDV in experimental group compared to controls at >12 months following treatment.

The funnel plots for studies likewise appeared symmetrical, suggesting little publication bias (Supplemental Figures S4A–S4D). However, the limited number and heterogeneity of trials reduce the analysis’s power to detect it.

Subgroup analysis by cell origin confirmed that neither allogeneic [MD = −14.96, P = 0.11, 95% CI (–39.06, 9.15)] nor autologous MSCs [MD = −9.94, P = 0.20, 95% CI (–25.17, 5.30)] significantly reduced LVEDV versus controls (Supplemental Figure S5A). Subgroup analyses indicated that allogeneic and autologous MSCs trended toward a reduction [allogeneic: MD = −58.11, P = 0.07, 95% CI (–120.92, 4.17); autologous: MD = −3.85; P = 0.05; 95% CI (–7.70, 0.01)], but neither reached significance (Supplemental Figure S5B). Neither allogeneic [MD = −15.59; P = 0.23; 95% CI (–41.26, 10.09)] nor autologous MSCs [MD = −2.98; P = 0.08; 95% CI (–6.37, 0.41)] significantly altered LVEDV (Supplemental Figure S5C).

We also combined all outcomes regardless of follow-up duration. The results revealed heterogeneity for included studies (I² = 98%, P < 0.00001); hence, we employed a random-effects model. Meta-analysis indicated a difference in LVEDV between these groups [MD = −13.19, P < 0.0001, 95 CI (−19.05, −7.33)] (Table 2). This suggests that, following MSC treatment, experimental group experienced a substantial reduction in LVEDV compared to controls (Table 2).

LVESV assessment at different time points

For LVESV assessed at <6 months, three studies compared LVESV between experimental (96 cases) and control groups (107 cases). Results showed heterogeneity for included studies (I² = 93%, P < 0.00001); thus, we used a random-effects model. Meta-analysis indicated that experimental group exhibited a decrease in LVESV <6 months compared to controls [MD = −11.35, P = 0.11, 95% CI (−20.55, −2.16)] (Figure 5a).

Figure 5.

Forest plot of meta-analysis of LVESV. (a) Forest plot of LVESV <6 months; (b) Forest plot of LVESV at 6 months; (c) Forest plot of LVESV at 12 months; (d) Forest plot of LVESV >12 months.

For LVESV assessed at 6 months, six studies compared LVESV between experimental (147 cases) and control groups (146 cases). The results revealed heterogeneity for included studies (I² =100%, P < 0.00001); thus, we used a random-effects model. The meta-analysis showed no significant difference in LVESV at 6 months between these groups [MD = −18.23, P = 0.17, 95% CI (−48.24, 11.78)] (Figure 5b). This indicates that, at 6 months following treatment, experimental group showed no change in LVESV compared to controls (Figure 5b).

For LVESV assessed at 12 months, seven studies compared LVESV between experimental (182 cases) and control groups (177 cases). The results showed heterogeneity for included studies (I² = 96%, P < 0.00001); thus, we used a random-effects model. Meta-analysis indicated no difference in LVESV at 12 months between these groups [MD = −7.62, P = 0.06, 95% CI (−15.71, 0.47)] (Figure 5c). This indicates that, at 12 months following treatment, experimental group showed no change in LVESV comparing with controls (Figure 5c).

For LVESV assessed >12 months, two studies compared LVESV between experimental (76 cases) and control groups (75 cases). The results revealed heterogeneity for included studies (I² = 98%, P < 0.00001); thus, we used a random-effects model. Results showed no difference in LVESV >12 months between these groups [MD = −8.25, P = 0.13, 95% CI (−18.93, 2.43)] (Figure 5d). This indicates that, at >12 months following treatment, experimental group showed no change in LVESV comparing with controls (Figure 5d).

No obvious asymmetry was observed in the funnel plots shown in Supplemental Figures S6A–S6D, indicating minimal publication bias; however, the small, diverse sample of trials and uneven reporting standards constrain the sensitivity of this assessment.

Subgroup analysis based on cell origin revealed no significant LVESV reduction at <6 months for allogeneic [MD = −15.26, P = 0.05, 95% CI (–30.74, –0.22)] or autologous MSCs [MD = −8.36; P = 0.28; 95% CI (–23.65, 6.93)] versus controls (Supplemental Figure S7A). At 6 months, only allogeneic MSCs significantly reduced LVESV [MD = −49.08, P = 0.04, 95% CI (–96.84, 1.31)], whereas autologous MSCs did not [MD = −0.46; P = 0.74, 95% CI (–3.18, 2.26)] (Supplemental Figure S7B). Subgroup analysis demonstrated that neither allogeneic [MD = −17.73; P = 0.09; 95% CI (–38.11, 2.65)] nor autologous MSCs [MD −2.51, P = 0.07; 95% CI (–5.19, 0.17)] reached significance in reduced LVESV at 12 months (Supplemental Figure S7C).

In another analysis, combining all outcomes regardless of follow-up duration, results revealed heterogeneity for included studies (I² = 97%, P < 0.00001); thus, we used a random-effects model. Results exhibited a difference in LVEF between these groups [MD = −12.40, P < 0.0001, 95% CI (−17.00, −7.79)] (Table 2). This suggests that, following MSC treatment, experimental group exhibited a substantial decrease in LVESV compared to controls (Table 2).

WMSI score assessment at different time points

For the WMSI assessed at <6 months, three studies compared WMSI between experimental (96 cases) and control groups (107 cases). The results showed no heterogeneity for included studies (I² = 0%, P = 0.84); thus, we used a fixed-effects model. Results showed a notable difference in WMSI <6 months between these groups [MD = −0.06, P < 0.0001, 95% CI (−0.07, −0.05)] (Figure 6a). This suggests that, within 6 months following treatment, experimental group exhibited a substantial decrease in WMSI score compared to controls (Figure 6a).

Figure 6.

Forest plot of meta-analysis of WMSI: (a) Forest plot of WMSI < 6 months; (b) Forest plot of WMSI at 6 months; (c) Forest plot of WMSI at 12 months; and (d) Forest plot of WMSI >12 months.

For the WMSI score assessed at 6 months, five studies compared WMSI between experimental (118 cases) and control groups (128 cases). The results revealed no heterogeneity for included studies (I² = 0%, P = 0.60); thus, we used a fixed-effects model. Results indicated a notable difference in WMSI at 6 months between these groups [MD = −0.04, P = 0.006, 95% CI (−0.06, −0.01)] (Figure 6b). This suggests that, at 6 months following treatment, experimental group exhibited a substantial increase in LVEF compared to controls (Figure 6b).

For the WMSI assessed at 12 months, four studies compared WMSI score between experimental (111 cases) and control groups (125 cases). Results exhibited heterogeneity for included studies (I² = 62%, P = 0.05); thus, we used a random-effects model. Meta-analysis indicated a notable difference in WMSI at 12 months between these groups [MD = 2.77, P = 0.06, 95% CI (−0.09, −0.03)] (Figure 6c). This suggests that, at 12 months following treatment, experimental group exhibited a substantial decrease in WMSI score compared to controls (Figure 6c).

For the WMSI assessed >12 months, two studies compared WMSI between experimental (76 cases) and control groups (75 cases). The results showed no heterogeneity for included studies (I² = 0%, P = 0.62); thus, we used a fixed-effects model. Meta-analysis indicated a notable difference in WMSI at >12 months between these groups [MD = −0.03, P = 0.02, 95% CI (−1.51, 8.50)] (Figure 6d). This suggests that, at >12 months following treatment, experimental group exhibited a substantial decrease in WMSI score comparing with controls (Figure 6d).

Funnel plots of the included studies did not show discernible asymmetry, indicating minimal evidence of publication bias (Supplemental Figures S8A–S8D); however, the small number of heterogeneous trials and variable reporting standards limit its sensitivity to detect publication bias.

In a further analysis, combining all outcomes regardless of follow-up duration, results showed no heterogeneity for included studies (I² = 30%, P = 0.14); thus, we used a fixed-effects model. Meta-analysis indicated a notable difference in LVEF between these groups [MD = −0.06, P < 0.0001, 95% CI (−0.06, −0.05)] (Table 2). This suggests that, following MSC treatment, experimental group exhibited a substantial decrease in WMSI comparing with controls (Table 2).

MACE assessment

For MACE, 10 studies compared MACE between experimental (268 cases) and control groups (247 cases). The results showed no heterogeneity for included (I² = 0%, P = 0.92); thus, we used a fixed-effects model. Meta-analysis indicated no difference in MACE between these groups [OR = 1.61, P = 0.10, 95% CI (0.92, 2.83)] (Figure 7). This suggests that, at 12 months following treatment, experimental group exhibited no change in MACE comparing with controls (Figure 5c). Funnel plots of the included studies did not show discernible asymmetry, pointing to minimal publication bias (Supplemental Figure S9); yet the small scale and heterogeneity of included trials, together with variable reporting, may obscure subtle biases.

Figure 7.

Forest plot of meta-analysis of MACE.

Subgroup analysis by administration route, intracoronary [OR = 1.62, P = 0.11, 95% CI (0.90, 2.92)] and intravenous [OR = 1.56, P = 0.64, 95% CI (0.24, 9.91)], also showed no effect on MACE (Supplemental Figure S10). Finally, cell-origin stratification confirmed that neither allogeneic [OR = 0.91, P = 0.05, 95% CI (0.30, 2.76)] nor autologous MSCs [OR = 1.98, P = 0.05, 95% CI (1.02, 3.85)] significantly affected MACE compared with controls (Supplemental Figure S11).

Discussion

Stem cell transplantation therapy has been a hot topic in recent years. Historically, stem cell transplantation was primarily used for hematological diseases, with over 40 years of clinical experience. Recently, numerous clinical trials both domestically and internationally have evaluated safety and efficacy of stem cell transplantation in patients with AMI post-PCI^18,36. These studies have demonstrated that stem cell transplantation is effective to some extent in AMI patients and that intracoronary injection is a safe and feasible method. Certain animal experiments and clinical studies have suggested that the therapeutic efficacy of MSC transplantation for treating myocardial infarction might surpass that of bone marrow-derived mononuclear cells³⁷. However, regarding long-term efficacy, especially benefits observed after 6 months, there are no significant differences compared to control groups³⁸. Therefore, the clinical efficacy of MSCs for AMI requires extensive clinical trials to be substantiated. Our research shows that AMI patients who received MSC transplantation had an improvement in LVEF across different follow-up periods. For LVEF assessed at <6 months, meta-analysis of eight studies revealed a notable difference in LVEF between these groups, indicating a substantial increase in LVEF in experimental group within 6 months following treatment comparing with controls. Similarly, for LVEF assessed at 6 months, meta-analysis of 10 studies showed a significant difference, and for LVEF assessed at 12 months, meta-analysis of seven studies demonstrated a significant improvement. However, for LVEF assessed at >12 months, no difference was detected between these groups.

The mechanisms by which MSCs exert their myocardial repair effects have been widely confirmed, including promoting neovascularization, reducing myocardial fibrosis, restoring vascular contractile function through direct differentiation, and stimulating endogenous cardiac stem cells³⁹. Paracrine mechanisms play a crucial role, whereby MSCs secrete various bioactive molecules that can inhibit fibroblast proliferation, promote the secretion of metalloproteinases, stimulate neovascularization, and mobilize endogenous cardiac stem cells to reduce fibrosis. Moreover, MSCs have the ability to differentiate into endothelial cells in vitro and cardiomyocytes in vivo. Furthermore, the transplanted MSCs influence the proliferation and differentiation of native cardiac stem cells and stimulate cell cycle of cardiomyocytes through cell-cell interactions, thereby repairing damaged myocardium, promoting angiogenesis, and achieving myocardial repair^40,41.

Regarding left ventricular remodeling, our study found that intracoronary MSC transplantation does not effectively improve LVESV and LVEDV. For LVEDV assessed at <6 months, meta-analysis showed no difference between these groups. Similarly, no differences were detected at 6, 12, and >12 months. For LVESV, meta-analysis also showed no significant differences at <6, 6, 12, and >12 months. In terms of follow-up time for stem cell transplantation, previous studies have shown that MSC transplantation in AMI patients significantly increases LVEF in the short term (follow-up time < 6 months) compared to controls, with statistically significant differences. However, in the long term (follow-up time >6 months), the results did not show statistical differences.

Regarding the baseline levels of patients undergoing MSC therapy, previous studies by Tendera et al indicated that stem cell transplantation for AMI is more effective in patients with lower baseline LVEF levels²³. However, recent large-scale RCT studies found that stem cell transplantation does not significantly improve outcomes in AMI patients with larger infarct areas and lower LVEF compared to control groups^25,42. Since only a portion of MSCs injected via coronary arteries are retained in the myocardium, the dosage of MSCs used in the injection impacts the efficacy. Previous meta-analysis results by Meluzin et al⁴³ regarding bone marrow mononuclear cells for AMI patients indicated that only those receiving a high dose of MSCs showed sustained LVEF improvement, suggesting a dose-response relationship. Another meta-analysis concluded that only when the number of transplanted MSCs exceeded 10⁷ could the patient’s LVEF improve⁴⁴.

Furthermore, the WMSI assessment revealed significant improvements in the experimental group at different follow-up points. For WMSI assessed at <6 months, meta-analysis revealed a notable decrease. At 6 months, the WMSI also significantly improved, and similar positive results were observed at 12 months and >12 months. In contrast, MACE analysis indicated no difference between these groups, suggesting no change in MACE following MSC treatment. Subgroup analysis based on administration route also revealed no significant impact of intracoronary administration or intravenous infusion on MACE.

Although our meta-analysis involving a large number of subjects demonstrated that intracoronary MSC transplantation improves LVEF in AMI patients, this study also has its limitations. In the subgroup analysis, allogeneic MSCs demonstrated clear early benefits, significantly improving LVEF at both <6 months and 6 months, and reducing LVESV at 6 months. In contrast, autologous MSCs failed to reach statistical significance at these timepoints. By 12 months, gains in LVEF for both allogeneic and autologous cells had diminished to nonsignificant levels, and neither cell type produced meaningful changes in LVEDV or LVESV beyond this period. Importantly, there was no impact on MACE rates for either group. However, the allogeneic subgroup is based on relatively few studies, which may introduce bias and limit generalizability; larger, well-powered trials are needed to validate these early advantages. Across the trials included in our meta-analysis, cell doses varied widely, and emerging evidence suggests a clear dose–response relationship. Mechanistically, higher cell numbers may improve myocardial retention, enhance paracrine cytokine release, and more effectively attenuate adverse ventricular remodeling. Nonetheless, the small number of studies employing high-dose regimens introduces potential bias, and optimal dosing remains undefined. Future randomized trials should therefore incorporate multiple dose arms and standardized reporting of MSC viability and delivery techniques to establish the minimum effective dose and maximize therapeutic benefit. We also recognize that AMI subtype influences MSC efficacy: STEMI’s larger, transmural infarcts and earlier cell delivery generally yield greater LVEF improvements, whereas non-STEMI patients with smaller, subendocardial injuries, and preserved baseline function often show minimal gains. Differences in infusion timing and infarct characteristics likely contribute to these varied responses. Future trials should stratify by AMI subtype, standardize delivery windows, and report infarct size and transmurality to identify which patient groups benefit most from MSC therapy. Variability in imaging modalities introduces significant bias. Echocardiography can underestimate early LVEF gains due to operator dependence and poor acoustic windows; Cardiac magnetic resonance (CMR) delivers gold-standard volumetric accuracy but is limited by cost, scanner availability, and scan time; and SPECT offers high reproducibility at the expense of temporal resolution, potentially masking subtle functional improvements. These intermodality differences amplify between-study heterogeneity and can skew meta-analytic outcomes. To mitigate this, future MSC trials should predefine a single-core modality, which will enhance data comparability and more accurately reflect MSC-mediated cardiac repair. All included trials carry inherent bias. Many early studies were small, single-center with inconsistent randomization, introducing selection bias, and lack of blinding in cell delivery and assessment created performance and detection bias. Differential follow-up rates also raise attrition bias concerns. These factors may have inflated treatment effects relationships. Future multicenter, double-blind RCTs with standardized cell-processing and imaging protocols are needed to address these limitations.

Conclusions

MSC treatment significantly improves LVEF and enhances WMSI in patients with AMI, particularly within the first 12 months posttreatment. Intracoronary administration appears to be more effective than intravenous infusion. However, no significant effect on MACE was observed. Additional rigorous studies are required to validate these results.

Supplemental Material

sj-docx-1-cll-10.1177_09636897251359773 – Supplemental material for Impact of mesenchymal stem cell therapy on cardiac function and outcomes in acute myocardial infarction: A meta-analysis of clinical studies

Supplemental material, sj-docx-1-cll-10.1177_09636897251359773 for Impact of mesenchymal stem cell therapy on cardiac function and outcomes in acute myocardial infarction: A meta-analysis of clinical studies by Mei Zhao, Yanpeng Xue, Qingqing Tian, Yan Deng and Tao Tang in Cell Transplantation

Footnotes

Acknowledgements

The authors acknowledge using ChatGPT-4 by OpenAI for English language editing and confirm the accuracy and originality of the AI-assisted content.

ORCID iDs

Mei Zhao

Tao Tang

Ethical Considerations

Not applicable.

Statement of human and animal rights

There are no human subjects in this article and informed consent is not applicable.

Statement of informed consent

Not applicable.

Authors’ contributions

MZ, YD, and TT designed the whole study; MZ, YX, QT, YD, and TT analyzed the data; YD and TT prepared the figures; MZ, YX, QT, YD, and TT wrote the manuscript; all the authors approved the manuscript for submission.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availabilitystatement

The data are available upon request.

Supplemental material

Supplemental material for this article is available online.

References

Duggan

Peters

Trachiotis

Antevil

. Epidemiology of coronary artery disease. Surg Clin North Am. 2022;102(3):499–516.

Damluji

van Diepen

Katz

Menon

Tamis-Holland

Bakitas

Cohen

Balsam

Chikwe

. Mechanical complications of acute myocardial infarction: a scientific statement from the American Heart Association. Circulation. 2021;144(2):e16–35.

Anderson

Morrow

. Acute myocardial infarction. N Engl J Med. 2017;376(21):2053–64.

Krittanawong

Khawaja

Tamis-Holland

Girotra

Rao

. Acute myocardial infarction: etiologies and mimickers in young patients. J Am Heart Assoc. 2023;12(18):e029971.

Kapur

Thayer

Zweck

. Cardiogenic shock in the setting of acute myocardial infarction. Methodist Debakey Cardiovasc J. 2020;16(1):16–21.

Murphy

Goldberg

. Mechanical complications of myocardial infarction. Am J Med. 2022;135(12):1401–409.

Bajaj

Sethi

Rathor

Suppogu

Sethi

. Acute complications of myocardial infarction in the current era: diagnosis and management. J Investig Med. 2015;63(7):844–55.

Zimetbaum

Josephson

. Use of the electrocardiogram in acute myocardial infarction. N Engl J Med. 2003;348(10):933–40.

Berezin

. Adverse cardiac remodelling after acute myocardial infarction: old and new biomarkers. Dis Markers. 2020;2020:1215802.

10.

Uccelli

Moretta

Pistoia

. Mesenchymal stem cells in health and disease. Nat Rev Immunol. 2008;8(9):726–36.

11.

Mishra

Shih

Parveen

Lenzen

Ito

Chan

. Identifying the therapeutic significance of mesenchymal stem cells. Cells. 2020;9(5):1145.

12.

Samsonraj

Raghunath

Nurcombe

Hui

van Wijnen

Cool

. Concise review: multifaceted characterization of human mesenchymal stem cells for use in regenerative medicine. Stem Cells Transl Med. 2017;6(12):2173–85.

13.

Ding

Chang

Shyu

Lin

. Human umbilical cord mesenchymal stem cells: a new era for stem cell therapy. Cell Transplant. 2015;24(3):339–47.

14.

Pan

Jiang

Liu

. Mesenchymal stem cell-derived exosomes in cardiovascular and cerebrovascular diseases: from mechanisms to therapy. Biomed Pharmacother. 2023;163:114817.

15.

Weng

Wang

Ouchi

Liu

Qiao

Zhao

. Mesenchymal stem/stromal cell senescence: hallmarks, mechanisms, and combating strategies. Stem Cells Transl Med. 2022;11(4):356–71.

16.

Hoang

Pham

Bach

Ngo

ATL

Nguyen

Phan

TTK

Nguyen

PTT

Hoang

Forsyth

Heke

, et al. Stem cell-based therapy for human diseases. Signal Transduct Target Ther. 2022;7(1):272.

17.

Matsushita

. Heart failure and adipose mesenchymal stem cells. Trends Mol Med. 2020;26(4):369–79.

18.

Clavellina

Balkan

Hare

. Stem cell therapy for acute myocardial infarction: mesenchymal stem cells and induced pluripotent stem cells. Expert Opin Biol Ther. 2023;23(10):951–67.

19.

Yamada

Minatoguchi

Kanamori

Mikami

Okura

Dezawa

Minatoguchi

. Stem cell therapy for acute myocardial infarction–focusing on the comparison between Muse cells and mesenchymal stem cells. J Cardiol. 2022;80(1):80–87.

20.

Henning

. Current status of stem cells in cardiac repair. Future Cardiol. 2018;14(2):181–92.

21.

Yang

Bai

. Research progress on cardiac tissue construction of mesenchymal stem cells for myocardial infarction. Curr Stem Cell Res Ther. 2024;19(7):942–58.

22.

Attar

Farjoud Kouhanjani

Hessami

Vosough

Kojuri

Ramzi

Hosseini

Faghih

Monabati

. Effect of once versus twice intracoronary injection of allogeneic-derived mesenchymal stromal cells after acute myocardial infarction: BOOSTER-TAHA7 randomized clinical trial. Stem Cell Research & Therapy. 2023;14(1):264.

23.

Chen

Fang

Liu

Qian

Shan

Zhang

Chunhua

Liao

Lin

Sun

J-P

. Effect on left ventricular function of intracoronary transplantation of autologous bone marrow mesenchymal stem cell in patients with acute myocardial infarction. Am J Cardiol. 2004;94(1):92–95.

24.

Chullikana

Majumdar

Gottipamula

Krishnamurthy

Kumar

Prakash

Gupta

. Randomized, double-blind, phase I/II study of intravenous allogeneic mesenchymal stromal cells in acute myocardial infarction. Cytotherapy. 2015;17(3):250–61.

25.

Gao

Pei

Ding

Chen

Zhang

Chen

Wang

Zhu

Liu

, et al. A critical challenge: dosage-related efficacy and acute complication intracoronary injection of autologous bone marrow mesenchymal stem cells in acute myocardial infarction. Int J Cardiol. 2013;168(4):3191–99.

26.

Gao

Chen

Zhang

Yang

Liu

Wang

Yan

Wang

Zhu

Wang

, et al. Intracoronary infusion of Wharton’s jelly-derived mesenchymal stem cells in acute myocardial infarction: double-blind, randomized controlled trial. BMC Med. 2015;13:162.

27.

Grajek

Popiel

Gil

Bręborowicz

Lesiak

Czepczyński

Sawiński

Straburzyńska-Migaj

Araszkiewicz

Czyż

Kozłowska-Skrzypczak

, et al. Influence of bone marrow stem cells on left ventricle perfusion and ejection fraction in patients with acute myocardial infarction of anterior wall: randomized clinical trial: impact of bone marrow stem cell intracoronary infusion on improvement of microcirculation. Eur Heart J. 2009;31(6):691–702.

28.

Hare

Traverse

Henry

Dib

Strumpf

Schulman

Gerstenblith

DeMaria

Denktas

Gammon

Hermiller

, et al. A randomized, double-blind, placebo-controlled, dose-escalation study of intravenous adult human mesenchymal stem cells (prochymal) after acute myocardial infarction. J Am Coll Cardiol. 2009;54(24):2277–86.

29.

Houtgraaf

den Dekker

van Dalen

Springeling

de Jong

van Geuns

Geleijnse

Fernandez-Aviles

Zijlsta

Serruys

Ducker

. First experience in humans using adipose tissue-derived regenerative cells in the treatment of patients with ST-segment elevation myocardial infarction. J Am Coll Cardiol. 2012;59(5):539–40.

30.

Kim

Cho

Lee

Kim

Lee

Kang

Lee

Ahn

Jeong

, et al. Improvement in left ventricular function with intracoronary mesenchymal stem cell therapy in a patient with anterior wall ST-segment elevation myocardial infarction. Cardiovasc Drugs Ther. 2018;32(4):329–38.

31.

Lee

Youn

Ahn

Kim

Yoo

Yoon

Kwon

Hong

Lee

Kwan

, et al. A randomized, open-label, multicenter trial for the safety and efficacy of adult mesenchymal stem cells after acute myocardial infarction. J Korean Med Sci. 2014;29(1):23–31.

32.

Mathur

Sim

Choudry

Veerapen

Colicchia

Turlejski

Hussain

Hamshere

Locca

Rakhit

Crake

, et al. Five-year follow-up of intracoronary autologous cell therapy in acute myocardial infarction: the REGENERATE-AMI trial. ESC Heart Fail. 2022;9(2):1152–59.

33.

Rodrigo

van Ramshorst

Hoogslag

Boden

Velders

Cannegieter

Roelofs

Al Younis

Dibbets-Schneider

Fibbe

Zwaginga

, et al. Intramyocardial injection of autologous bone marrow-derived ex vivo expanded mesenchymal stem cells in acute myocardial infarction patients is feasible and safe up to 5 years of follow-up. J Cardiovasc Transl Res. 2013;6(5):816–25.

34.

Wang

. The beneficial effects of intracoronary autologous bone marrow stem cell transfer as an adjunct to percutaneous coronary intervention in patients with acute myocardial infarction. Biotechnol Lett. 2014;36(11):2163–68.

35.

Zhang

Wang

Cai

Chen

Yang

Liu

. Bone marrow mesenchymal stem cells transfer in patients with ST-segment elevation myocardial infarction: single-blind, multicenter, randomized controlled trial. Stem Cell Res Ther. 2021;12(1):33.

36.

George

. Stem cell therapy in acute myocardial infarction: a review of clinical trials. Transl Res. 2010;155(1):10–19.

37.

Zhang

Sun

Qian

Lin

Zhao

Wang

Zou

. Comparison of various kinds of bone marrow stem cells for the repair of infarcted myocardium: single clonally purified non-hematopoietic mesenchymal stem cells serve as a superior source. J Cell Biochem. 2006;99(4):1132–47.

38.

Minguell

Erices

. Mesenchymal stem cells and the treatment of cardiac disease. Exp Biol Med (Maywood). 2006;231(1):39–49.

39.

Bronckaers

Hilkens

Martens

Gervois

Ratajczak

Struys

Lambrichts

. Mesenchymal stem/stromal cells as a pharmacological and therapeutic approach to accelerate angiogenesis. Pharmacol Ther. 2014;143(2):181–96.

40.

Kinkaid

Huang

Weisel

. What’s new in cardiac cell therapy? allogeneic bone marrow stromal cells as “universal donor cells.” J Card Surg. 2010;25(3):359–66.

41.

Ubil

Duan

Pillai

ICL

Rosa-Garrido

Bargiacchi

Stanbouly

Huang

Rojas

Vondriska

, et al. Mesenchymal–endothelial transition contributes to cardiac neovascularization. Nature. 2014;514(7524):585–90.

42.

Fan

Chen

Liu

Jiang

Tian

Sun

Tian

. The effect of age on the efficacy of human mesenchymal stem cell transplantation after a myocardial infarction. Rejuvenation Res. 2010;13(4):429–38.

43.

Meluzín

Janousek

Mayer

Groch

Hornácek

Hlinomaz

Kala

Panovský

Prásek

Kamínek

Stanícek

, et al. Three-, 6-, and 12-month results of autologous transplantation of mononuclear bone marrow cells in patients with acute myocardial infarction. Int J Cardiol. 2008;128(2):185–92.

44.

Martin-Rendon

Brunskill

Hyde

Stanworth

Mathur

Watt

. Autologous bone marrow stem cells to treat acute myocardial infarction: a systematic review. Eur Heart J. 2008;29(15):1807–18.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.99 MB