More than 6 years ago the initial report on clinical application of bone marrow-derived mononuclear cells (BMMNCs) in patients with acute myocardial infarction (AMI) opened up a new era of regenerative cardiology and brought with it great enthusiasm and expectations.1 Since then, numerous clinical trials have been carried out with the aim of assessing the efficacy and safety of stem cell therapy. In all cases the safety of the therapy was proved, but its effectiveness still remains a matter of controversy and fierce debate.2
Since stem cell therapy trials in patients with AMI began, the primary outcome measure has been the change in left ventricular (LV) ejection fraction (EF). The LV volumes have also been analysed to assess more accurately the cause of the changes in LVEF after bone marrow cell (BMC) therapy.1,3–6 So far, the data on the effectiveness of BMC infusion in improving the global LV contractility remain equivocal.
For example, the ASTAMI trial showed no improvement of LVEF and no reduction of end-diastolic volume in patients treated with intracoronary transfer of BMCs. In this trial, a significant improvement of LVEF was observed not only in patients treated with BMCs, but also in the control group.7 The BOOST trial showed an initial improvement in LVEF 6 months after BMC infusion, but the difference between the BMC arm and the control group was no longer significant at 18 months.8
On the other hand, REPAIR-AMI, so far the largest randomized placebo-controlled study, showed a significant improvement in LVEF at 4 months follow-up by 2.5% in patients receiving BMCs. The increase in LVEF co-existed with an improvement of regional contractility including the segments located in the area of infarction.3
It has been hypothesized that these inconsistencies can be due to differences in trial designs, including time from reperfusion to BMC transfer, the type and number of infused cells, and the method of cell isolation.9 It seems that cell therapy has more favourable effects when initiated later—at least 4 days after reperfusion.3 However, in some studies which showed no significant increase of the LVEF, the bone marrow harvest and BMC transfer were performed later (4–7 days) after primary percutaneous coronary intervention (PCI).7,8 Also the BMC preparation protocol is the pivotal issue because the preservation of cell viability and migratory response seems to be related to the outcome.10
Surrogate endpoints other than the LVEF, such as the infarct size, timing of EF recovery, and improvement in regional LV systolic function within the infarct and border zones, have been considered. In the study by Janssens et al., 6 although no difference in EF change between the BMC group and the controls could be detected, patients receiving active treatment had a smaller infarct size. In the BOOST trial, the recovery of LVEF after the transfer of BMCs was significantly faster than in the control group.8 Both in BOOST5 and in REPAIR-AMI,3 an improvement in blood flow into the border zone was identified as the predominant factor leading to an increase in systolic performance.
Janssens et al.6 were the first to publish a double-blind, placebo-controlled comparison of BMC therapy in patients with AMI. In this trial, conducted in patients with ST-segment elevation AMI who underwent successful primary PCI with stent implantation, 67 subjects were enrolled. Patients were randomized to receive either placebo or intracoronary infusion of BMMNCs within 24 h after the primary PCI. The median number of MNCs harvested from the bone marrow and isolated using standard Ficol gradient centrifugation was 172 × 106. LVEF similarly increased both in the active treatment and in the placebo arm, with no significant difference between the groups. Also the assessment of myocardial perfusion and oxidative metabolism using PET imaging showed similar changes over time in both groups.
Herbots et al.11 have reported on a new analysis of the changes in LV systolic function in the same group of patients. They focus on novel parameters that might possibly serve as surrogate endpoints for the assessment of the effect of BMC therapy on LV systolic function recovery. The authors chose strain rate imaging as the most accurate method of regional LV function measurement, which serves this purpose better than the less accurate and less reproducible wall motion score index.12 Interestingly, they found that at 4 months end-systolic strain and peak systolic strain rate improved significantly more in the BMC than in the control group only in the segments with a transmural infarct. In the myocardial border zone, the recovery of end-systolic strain was equal in both groups.
Several mechanisms can be responsible for the observed fine improvement in regional LV function. First, it can be directly dependent on the grafted cells which, via neovascularization, antiapoptotic, and other paracrine effects, may reduce the extent of cardiomyocyte loss, activate and recruit resident cardiac stem cells, and improve myocardial perfusion.9 Previously published data from the same trial showed that the BMC treatment did not improve the perfusion, but might increase the oxidative metabolism.6 All other mechanisms remain largely speculative. Interestingly, the authors suggest that the reduction of afterload related to a decrease in the blood pressure in patients treated with BMCs might be involved in the beneficial effects on regional LV function.
To summarize, the study of Herbots et al.11 shows that measurement of strain and strain rate makes it possible to detect LV function recovery after BMC treatment that cannot be reflected by the LVEF changes, and that some improvement can be present in the transmural infarct zone. However, this trial6,11—similarly to many previous studies—also illustrates how difficult it is to differentiate between treatment-induced and spontaneous improvement, and therefore how difficult it is to define success in cell therapy.
Clinical outcome is the ultimate measure of success, and outcome trials with unselected mononuclear BMCs will be soon conducted. So far, however, we do not know how changes in LV systolic function, as measured by the EF or more sensitive methods, would translate into clinical benefit. The only indication that such a benefit may be present comes from the REPAIR-AMI study,3,14 where the pre-specified clinical endpoint including death, recurrence of myocardial infarction, and any revascularization procedure was significantly reduced in the BMC group as compared with the placebo group (P = 0.01). Also other combinations of clinical endpoints tended to show a benefit. This study, however, was too small to assess clinical outcome, so a new, properly powered study is needed.
While the efficacy of therapy with unselected mononuclear bone marrow-derived cells is being clinically evaluated, much effort should also be directed to understand better the biology of progenitor cells in order to identify alternative cell populations for potential use in patients (CD133+, CD34+CXCR4+, very small embryonic-like cells) and find methods which enhance cell homing and transdifferentiation into functional myocardium.15,16
The initial enthusiasm for the expected rapid and unquestionable benefit from stem cell therapy in patients with myocardial infarction is now over, and the number of unanswered questions related to this treatment method does not seem to decrease. The method still holds much promise, but definitely—success does not come easy.
Conflict of interest: none declared.