Bone marrow for cardiac repair: the importance of characterizing the phenotype and function of injected cells.

Administration of adult stem cells for cardiac repair after myocardial infarction (MI) is a revolutionary new strategy that could become clinically feasible in the near term, appears to be safe and relatively cost-effective, and offers tremendous potential to improve prognosis. Using diverse cell populations and strategies, several relatively small clinical studies have reported encouraging outcomes, with improvement in various measures of myocardial perfusion and left ventricular (LV) function. However, negative data have also emerged, generating controversy regarding the overall efficacy of cell therapy. This controversy persists even after the recent publication of the three largest randomized studies to date of cell therapy for cardiac repair. In the largest of these trials, the multicentre REPAIR-AMI study, 204 patients received an intracoronary infusion of autologous mononuclear bone marrow cells (BMCs) or placebo (medium) at 3–7 days after a reperfused acute MI. After 4 months, the absolute improvement in global LV ejection fraction (EF) was greater in patients treated with BMCs compared with control patients. After 1 year, BMC treatment was associated with a significant reduction in the combined endpoints of death, recurrent MI, and revascularization and death, recurrent MI, and re-hospitalization for heart failure. An improvement in LV function and clinical outcome after BMC treatment was also noted in the TOPCARE-CHD trial among patients with an old MI. However, no benefit was reported in the other trial of acute MI (the ASTAMI trial), in which 100 patients with reperfused acute anterior MI received either intracoronary infusion of BMCs or no intervention at a median of 6 days after MI. After 6 months, there was no difference in LVEF, infarct size, or LV enddiastolic volume between the two groups. At first glance, these conflicting results might appear to portend a rather uncertain future for adult stem cell-based approaches for cardiac repair. Importantly, the ostensibly incongruent results may be construed by the opponents of adult stem cell therapy as evidence of lack of efficacy and inappropriately used to summarily dismiss this approach. It is therefore important to understand the reason(s) for the discrepancy between the REPAIR-AMI and ASTAMI studies. Although REPAIR-AMI was larger (204 vs. 100 patients), utilized a higher dose of BMCs (median, 198 10 vs. 68 10), and had a more rigorous control group (unlike ASTAMI, control patients received bone marrow aspiration and intracoronary infusion of vehicle), it is unclear whether these differences can account for the different outcomes. A careful study by Seeger et al. provides important new insights into this issue. These investigators compared the methods for isolating BMCs in REPAIR-AMI and ASTAMI. Bone marrow was collected from healthy donors and patients with angiographically proven coronary artery disease (CAD) and tested for various parameters of phenotype and function. To obtain BMCs, aliquots from the same bone marrow aspirate were subjected to similar gradient centrifugation protocols using either Ficoll (Cambrex), as in REPAIR-AMI, or Lymphoprep (Axis-Shield), as in ASTAMI. Cells were stored overnight in either X-vivo 10 mediumþ20% autologous serum at room temperature (REPAIR-AMI protocol) or 0.9% NaClþ20% heparin-plasma at 48C (ASTAMI protocol). The results of the two protocols were quite different. Compared with Lymphoprep, the use of Ficoll yielded not only greater number of BMCs but also more CD45þ/CD34þ and CD45þ/CD133þ cells. Cells isolated with Ficoll yielded greater numbers of colony forming units, an in vitro indicator of the proliferative capacity and a surrogate measure of stemness. The number of mesenchymal stem cells (MSCs) was also greater after Ficoll separation. The authors also examined whether the overnight storage conditions might have affected the phenotype or functionality of the cells. In an in vitro assay, BMCs isolated with Lymphoprep (particularly from patients with CAD) showed reduced ability to migrate to the chemoattractant SDF-1, and this reduction was more pronounced after overnight storage. Consistent with these observations, the expression of CXCR4, the receptor for SDF-1, was significantly reduced after overnight storage in BMCs isolated with Lymphoprep. Following transplantation in a mouse model of hind-limb ischaemia, BMCs The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.