Context Numerous randomized controlled studies assessing intracoronary bone marrow cell therapy (BMC) after acute myocardial infarction (AMI) have been performed.
Objective To systematically review the effect of autologous BMC therapy on left ventricular function by performing an up to date meta-analysis of randomized controlled trials (RCTs) including long-term follow-up.
Data sources Trials were indentified through a literature search from 1980 to June 2012 of the Pubmed, Embase, Cochrane database, and the Current Controlled Trials Register.
Study selection Randomized clinical trials comparing intracoronary BMC infusion to control as treatment for AMI.
Data extraction The primary endpoint was the change in left ventricular ejection fraction (LVEF) from baseline to follow-up. Secondary endpoints were changes in left ventricular end diastolic volume (LVEDV), left ventricular end systolic volume (LVESV), infarct size and clinical outcomes.
Results Improvement of LVEF in patients receiving intracoronary BMC was significantly better within 6 months (23 studies, 2.23% (95% confidence interval (CI) 1.00 to 3.47); p<0.001). At 12 months of follow-up, this effect sustained with 3.91% more LVEF improvement (11 studies, (95% CI 2.56 to 5.27), p<0.001). At long-term follow-up, we found a trend for better LVEF improvement in favor of cell therapy (7 studies, 1.90% (95% CI −0.43 to 4.23); p=0.11). There was no clear effect in infarct size or LVEDV. However, we found a significant reduction in LVESV at 6 months (−4.81 ml (95% CI −7.86 to −1.76); p<0.001 and at 12 months (−9.41 ml (95% CI −13.64 to −5.17); p<0.001). Moreover, there was a statistically significant decrease in recurrent AMI (Relative Risk (RR) 0.44 (95% CI 0.24 to 0.79); p=0.007), and readmission for heart failure, unstable angina or chest pain (RR 0.59 (95% CI 0.35 to 0.98); p=0.04) in favour of cell therapy.
Conclusion Intracoronary BMC treatment leads to a moderate improvement of LVEF and reduction of LVESV at 6 months that sustained at 12 months follow-up, without a clear significant effect on LVEDV, or infarct size. Furthermore, we found that intracoronary cell therapy is significantly associated with a reduction in recurrent AMI and readmission for heart failure, unstable angina or chest pain.
- Cell therapy
- acute myocardial infarction
- ventricular function
Statistics from Altmetric.com
Despite optimal state-of-the-art medical and revascularisation therapies, the progression and development of congestive heart failure is still considered a large burden.1 Acute myocardial infarction (AMI) is a major cause of congestive heart failure.2 Therefore, alternative therapies to complement primary percutaneous coronary intervention (PCI) or thrombolytic therapy in the prevention of congestive heart failure after AMI are being investigated.
One of the suggested therapies for myocardial dysfunction is intracoronary bone marrow cell (BMC) treatment, which was tested positive in several animal studies. From then on, clinical trials rapidly followed to translate these exciting preclinical results to humans.3 Numerous relatively small randomised controlled studies with different types of cells, different dosages and follow-up durations were performed.4–36
The effect of cell therapy in AMI patients has been reviewed in the past,37–40 showing a small significant effect after myocardial infarction of left ventricular ejection fraction (LVEF) and other endpoints, such as left ventricular end diastolic volume (LVEDV), left ventricular end systolic volume (LVESV) and infarct size. However, these reviews did not include latest studies and follow-up. Our aim was to systematically review the effect of intracoronary injection of autologous BMC after AMI by performing an up-to-date meta-analysis of randomised controlled trials (RCTs).
Literature search and study selection
We identified all published trials comparing intracoronary infusion of BMC as treatment for AMI with control, by searching PubMed, EMBASE, Cochrane database and the Current Controlled Trials Register from 1980 to 1 June 2012. The following key words were used: ‘bone marrow cell’, ‘stem cell’, ‘bone marrow transplantation’, ‘precursor cell’, ‘progenitor cell’, ‘haemopoietic marrow’, ‘myocardial ischaemia’, ‘ischaemia’, ‘ischaemic heart disease’, ‘coronary heart disease’, ‘heart failure’, ‘cardiac’, ‘cardiomyo’, ‘ischemia’ (see online appendix 1). Additionally, we manually searched the conference abstracts of the American Heart Association, American College of Cardiology, European Society of Cardiology and Transcatheter Cardiovascular Therapeutics, KoreaMed, IndMed and LILACS. We also searched the references of identified studies and relevant review articles to identity additional randomised studies.
To be included, trials had to: be randomised; include patients with a clinical diagnosis of acute AMI treated with PCI and compare a single intracoronary infusion of autologous BMC (irrespective of the type and number of isolated cells) within one month of AMI to a comparator arm not receiving BMC (eg, control media or plasma or standard treatment). We did not include trials where follow-up was <3 months, or granulocyte (macrophage) colony-stimulating factor (G(M)-CSF) was administrated as co-intervention. Study selection was done by two independent reviewers (RD, AA), and disagreement was resolved by a third reviewer (AH).
Data extraction and data analysis
Data from the original publications, including markers of validity, study design, nature of the intervention and clinical and imaging outcomes were obtained from the original publications.
The primary endpoint was the change in LVEF from baseline to follow-up. Secondary endpoints were changes in LVESV, LVEDV, infarct size and clinical outcomes (all-cause mortality, recurrent AMI, target vessel revascularisation and readmission for heart failure, unstable angina or chest pain). For each endpoint we involved only those studies in the analysis for which information about the endpoint had been published. The quantitative information about endpoint per treatment group was obtained by extracting the mean change±SD from the publication.
When several methods were used for outcome assessment, MRI data were preferentially included in the analysis, followed by single-photon emission CT, echocardiography, and left ventricular (LV) angiography. LVESV and LVEDV were collected in ml/m2, or ml when ml/m2 was not available. Infarct size was preferably collected in per cent of LV or grams when not available. When reported or published data were incomplete, we requested additional details by correspondence, generally with the principal investigator, searched previously published Cochrane reviews, or calculated or estimated these data using the method of the Cochrane handbook.41 If trials consisted of multiple intervention comparator arms, we only used data from the comparator arm meeting the inclusion- and exclusion criteria. If the multiple intervention arms were all eligible, we used the combined data, if available, and if not available, we separated them into different studies.15 ,27
Outcomes were analysed with random-effects models. Summary results were presented as weighted mean difference with 95% CI per imaging outcome. We presented imaging outcomes split for different follow-up duration (3–6 months, 12 months and >12 months) and different units used (ml/m2 vs ml and %/LV vs grams).
We examined heterogeneity across studies by calculating an I2-value for every outcome. A standard fixed-effects model (Mantel–Haenszel method) was used in the absence of heterogeneity among studies (I2 value less than 50%). In the presence of heterogeneity, the DerSimonian and Laird random effects model was used. Funnel plots were plotted to investigate possible publication bias. Analyses were performed using RevMan 4.2 (The Cochrane Collaboration, Copenhagen, Denmark), with statistical significance for hypothesis testing set at the 0.05, two-tailed level.
From the initial 2124 citations (figure 1), 2025 of them were initially excluded at the title/abstract level. Fifty-four of the remaining 99 citations were RCTs, of which 24 studies were eligible for inclusion. These trials allocated 1624 patients to intracoronary cell therapy or standard therapy, ranging from 10 to 204 participants per trial. The intervention and comparator characteristics of the included studies are shown in table 1. Ten studies included patients with anterior myocardial infarction only. Only seven studies were double blinded, and 12 studies used MRI as diagnostic modality. Follow-up ranged from 3 to 60 months. Twenty studies used mononuclear BMCs for infusion (table 2). Most trials fulfilled our markers of validity (table 3).
Change in LVEF
The change in LVEF within 6 months was available in 23 studies (figure 2). Improvement of LVEF within 6 months in patients receiving intracoronary cell therapy was significantly augmented compared with standard therapy (2.23% (95% CI 1.00 to 3.47); p<0.001). At 12 months of follow-up, this effect sustained with 3.91% more LVEF improvement (11 studies, (95% CI 2.56 to 5.27), p <0.001) in patients receiving intracoronary cell therapy. At long-term follow-up, we found a trend for better LVEF improvement in favour of cell therapy (seven studies, 1.90% (95% CI −0.43% to 4.23%); p=0.11), although the number of studies was small. Assessment of publication bias using visual examination of the funnel plot of the primary publications indicated no significant publication bias (figure 3). Moreover, exclusion of the smallest studies (cell therapy arm <20) had little effect on the overall estimate. Improvement of LVEF in patients receiving intracoronary BMC within 6 months in 17 studies was 2.12% (95% CI 0.71% to 3.52%); p<0.001).
LV volumes and infarct size
There was no clear difference in LVEDV between cell therapy and controls at 6 months (−0.96 ml (95% CI −5.21 to 3.29); p=0.66 and −1.94 ml/m2 (95% CI −6.84 to 2.96); p=0.44), at 12 months (−4.84 ml (95% CI −10.52 to 0.83); p=0.09) and at long-term follow-up (−1.50 ml (95% CI −10.84 to 7.83); p=0.75) (figure 4). However, we found a significant reduction in LVESV at 6 months (−4.81 ml (95% CI −7.86 to −1.76); p<0.001 and −0.55 ml/m2 (95% CI −2.57 to 1.47); p=0.59), at 12 months (−9.41 ml (95% CI −13.64 to −5.17); p<0.001), and at long-term follow-up (−5.53 ml (95% CI −10.97 to −0.08); p=0.05) in patients receiving intracoronary cell therapy (online appendix 2). There were only two studies reporting on LVEDV and LVESV in ml/m2 on longer term showing no benefit of cell therapy. There was a trend towards infarct size reduction in patients receiving intracoronary cell therapy at 6 months (−2.16 g (95% CI −4.45 to 0.14), p=0.07) (online appendix 3). However, in the studies where infarct size was reported in per cent of LV, the difference was 0.01% ((95% CI −1.38 to 1.40), p=0.99). Infarct size in grams or per cent of LV was reported in nine of the 24 studies.
We found that cell therapy was associated with a statistically significant decrease in recurrent AMI (Relative Risk (RR) 0.44 (95% CI 0.24 to 0.79); p=0.007) and readmission for heart failure, unstable angina or chest pain (RR 0.59 (95% CI 0.35 to 0.98); p=0.04). There was a trend for difference between intracoronary cell therapy and control in all cause mortality (RR 0.60 (95% CI 0.34 to 1.08); p=0.09) and target vessel revascularization (RR 0.82 (95% CI 0.63 to 1.07); p=0.15). The number of events are summarized in Table 4.
In the present meta-analysis, we found that BMC treatment leads to a slightly better improvement of LVEF at 6 months and reduction of LVESV at 6 months, which was sustained at 12 months follow-up, without a clear significant effect on LVEDV, or infarct size. Furthermore, we found that intracoronary cell therapy is significantly associated with a reduction in recurrent AMI and readmission for heart failure or unstable angina or chest pain.
In the past years, earlier reports have presented a meta-analysis regarding BMC cell therapy after MI.37–40 Our meta-analysis differs from previously published meta-analyses. First, other meta-analyses also included studies that administered G(M)-CSF or bone marrow-derived cells from the peripheral blood group, and/or included both acute and chronic myocardial infarction patients making the results more heterogeneous. Second, all meta-analyses pooled different outcome measures for LVEDV, LVESV and infarct size used in the trials. We have chosen to present these outcomes separately, because this is more precise. Third, our meta-analysis is more up-to-date containing a larger number of studies and the latest follow-up. However, our results are globally in agreement with the results of the previous meta-analyses, namely that cell therapy results in a slightly better significant improvement of LVEF compared with controls, without a clear significant effect on LVEDV or infarct size.
There has been a lot of controversy regarding high-risk subgroups benefitting more from cell therapy, appropriate cell storage, cell dosage and cell type. Unfortunately, this meta-analysis does not answer these relevant questions, but bundles all trials with a high degree of heterogeneity.
Finally, the question remains whether a small increase of LVEF is clinically meaningful. However, as Reffelman et al stated when putting the moderate increases of cardiac volumes in perspective with existing therapy, we appear to be in the range of effects observed with reperfusion therapy, pharmacotherapeutic interventions influencing the renin-angiotensin-aldosterone pathway, and beta-blockers after AMI.42 More importantly, we found that intracoronary cell therapy is associated with a significant reduction in recurrent AMI, and readmission for heart failure, unstable angina or chest pain. The mechanism behind these significant reductions remains unclear, and could be biased due to the low number of clinical events or placebo effect (in case of readmission for heart failure, unstable angina or chest pain). It has been suggested that BMC therapy alters the process of restenosis development and/or atherosclerotic disease progression via enhanced reendothelialisation and potential vascular repair. Also, it has been hypothesised that BMC leads to greater recovery of coronary blood flow reserve which is inversely associated with atherosclerotic disease progression, which may, in turn, explain the lower rate of recurrent MI.24 However, these hypotheses are highly speculative, and should be investigated further.
We feel that intracoronary BMC treatment still holds promise, and should be investigated in a large clinical trial with clinical outcomes as the primary endpoint. The large BAMI trial, funded by the European Union, will investigate BMC therapy in a RCT with a primary clinical endpoint.
This meta-analysis shows that there is little space for more small RCTs assessing intracoronary BMCs in AMI with surrogate primary endpoints.
Intracoronary BMC treatment leads to a moderate improvement of LVEF and reduction of LVESV at 6 months that sustained at 12 months follow-up, without a clear significant effect on LVEDV, or infarct size. Furthermore, we found that intracoronary cell therapy is significantly associated with a reduction in recurrent AMI, readmission for heart failure, and unstable angina or chest pain.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.