Efficacy and safety of bone marrow mononuclear cell therapy for myocardial infarction: a dose–response meta-analysis

Background Bone marrow mononuclear cells (BMMNCs) have been investigated as a potential regenerative therapy for myocardial infarction (MI). However, whether their clinical benefits are sustained over time and what constitutes an optimal therapeutic dose remain uncertain. This meta-analysis aimed to evaluate the efficacy, safety, and dose–response relationship of BMMNC therapy in patients with MI. Methods Randomized controlled trials (RCTs) comparing BMMNC therapy with standard therapy or placebo in patients with MI were systematically searched in relevant databases. Statistical analyses were performed using R. Continuous outcomes were summarized as mean differences (MDs) with 95% confidence intervals (CIs), and dichotomous outcomes were summarized as risk ratios (RRs) with 95%CIs. Random-effects models were used for pooled analyses. Subgroup analyses, sensitivity analyses, meta-regression, and restricted cubic spline models were performed to explore the effects of follow-up duration and cell dose on treatment outcomes. Results A total of 43 RCTs involving 2,213 patients receiving BMMNC therapy and 1,686 control patients were included. BMMNC therapy significantly improved left ventricular ejection fraction (LVEF) at 6 months (MD = 2.57%, 95%CI: 0.79 to 4.35; p = 0.007), but this effect was attenuated at 12 months and did not reach statistical significance (MD = 1.59%, 95%CI: −0.06 to 3.23; p = 0.057). At the final follow-up, BMMNC therapy was still associated with a significant improvement in LVEF (MD = 2.24%, 95%CI: 0.77 to 3.71; p = 0.004). Meta-regression showed that the improvement in LVEF decreased significantly with longer follow-up duration (β = −0.296 per month, 95%CI: −0.549 to −0.043; p = 0.024). No significant linear or nonlinear dose–response relationship was observed between cell dose and LVEF improvement. Safety analysis showed that BMMNC therapy significantly reduced the risk of major adverse cardiac events (MACE) overall (RR = 0.62, 95%CI: 0.45 to 0.86; p = 0.0036). However, this benefit was mainly observed in low- and intermediate-dose studies (<10 9 cells), whereas high-dose studies (>10 9 cells) did not show a clear benefit (RR = 1.53, 95%CI: 0.40–5.82), suggesting a potential threshold in the dose–safety relationship. Further meta-regression indicated that increasing cell dose was significantly associated with a higher risk of MACE (β = 0.341, 95%CI: 0.104 to 0.578; p = 0.0062). Subgroup analyses suggested that a dose range of 10 8 to 10 9 cells may provide a favorable balance between LVEF improvement and safety. Conclusion BMMNC therapy may improve LVEF in the short term and reduce the risk of MACE in patients with MI; however, the LVEF benefit appears to diminish over time. Current evidence does not support a significant linear or nonlinear dose–response relationship between cell dose and LVEF improvement, although higher cell doses may be associated with increased MACE risk. Taken together, 10 8 to 10 9 cells may represent a potentially appropriate therapeutic dose range, although this finding requires confirmation in further high-quality RCTs with longer follow-up and standardized cell-processing protocols.