Response to Letter Regarding Article, “Adult Bone Marrow Cell Therapy Improves Survival and Induces Long-Term Improvement in Cardiac Parameters: A Systematic Review and Meta-Analysis”

Abstract

We thank Dr Delewi and colleagues for a careful review of our report 1 and thoughtful comments. The authors point out that our meta-analysis of 50 trials did not include 3 large trials (REGENT [Myocardial Regeneration by Intracoronary Infusion of Selected Population of Stem Cells in Acute Myocardial Infarction], BONAMI [Bone Marrow in Acute Myocardial Infarction], and HEBE).2-4 As mentioned under "Study Selection," our meta-analysis included trials that reported data as mean and standard deviation. The REGENT trial 2 reported data with median values and therefore could not be included. As a preliminary screening step, we reviewed the titles of all reports retrieved through a stepwise search strategy, and the title of the HEBE trial stated that mononuclear cells were derived from bone marrow or peripheral blood. Because it is a standard practice to mobilize bone marrow cells prior to harvest from peripheral blood, we assumed mobilization with cytokine. After reviewing the full article, it appears that Hirsch et al did not mobilize bone marrow cells and reported results from bone marrow group and peripheral blood group separately.3 In the report by Herbots et al, 4 we were unable to definitively ascertain whether magnetic resonance imaging data were from the same patient population as in the earlier report from this group. Finally, our search strategy failed to identify the BONAMI trial. 5In light of the above, we analyzed data for all primary end points de novo after including data from BONAMI and HEBE and excluding data from Herbots et al. 4 In this new analysis (51 trials, 2787 patients), compared with control subjects, bone marrow cell-treated patients exhibited greater left ventricular ejection fraction (3.87%; 95% confidence interval, 2.81%-4.93%; P<0.00001) and smaller infarct size (−3.42%; confidence interval, −4.74% to −2.10%; P<0.00001), left ventricular end-systolic volume (−8.34 mL; confidence interval, −10.92 to −5.76 mL; P<0.00001), and left ventricular end-diastolic volume (−4.57 mL; confidence interval, −6.77 to −2.37 mL; P<0.00001). Importantly, these results are very similar to those reported in our original paper and uphold our original major conclusions. 1As for the subgroup analysis, we do agree that cell therapy mechanisms are likely different in acute versus chronic ischemic heart disease, but we also recognize that the difference in myocardial pathology is not the only difference among these studies. Since subgroup analysis is limited by the number of trials (and patients) in each subgroup, we chose to include all eligible studies when assessing the impact of one specific factor (for example, timing) despite the presence of other differences (pathology, cell number, ejection fraction, and such) among included studies. This inherent limitation with meta-analysis at present can be circumvented only with data from a much larger number of trials and patients in the future. DisclosuresNone. Vinodh

Emerging Therapies for Congestive Heart Failure

“…The close relationship between the surrogates of infarct size and LV function and subsequent mortality, as determined in very large clinical outcomes studies, serves as a clear basis for the use of such assessments for clinical effectiveness . Encouragingly, a series of meta‐analyses ( Table ) suggest that bone marrow cell administration in the days post‐MI results in a highly statistically significant improvement in ejection fraction (EF), a finding that has been replicated multiple times …”