Adipose-Derived Regenerative Cells in Ischemic Cardiomyopathy
Adipose-Derived Regenerative Cells in Ischemic Cardiomyopathy
In this multicenter, randomized, placebo-controlled, double-blind study, we have shown for the first time that the harvest and transendocardial injection of ADRCs are safe and feasible in no-option patients with ischemic cardiomyopathy. Moreover, cell therapy with ADRCs in these patients may preserve functional capacity over time and could exert a modest beneficial effect on myocardial perfusion, scar size, and left ventricular contractility.
Adipose tissue comprises a mixed cell source containing a high proportion of MSCs. The stromal fraction has been shown to contain multipotent stem cells. Our approach allowed us to isolate ADRCs in high numbers in real time (<2 hours), during which time baseline mapping procedures were initiated in patients in preparation for cell injections.
Adipose-derived regenerative cells, unlike other cell types such as hematopoietic cells, have no specific cell surface markers. Thus, we used stromal-associated and stem cell–associated markers in our flow cytometric analysis and confirmed that the SVF comprises a heterogeneous cell population. The SVF cell population contained low to medium levels of cells positive for the mesenchymal marker CD105, as previously shown, as well as low to medium levels of cells with endothelial (CD34 and CD31) and hematopoietic (CD45) markers. Although the percentage of CD34 cells in the SVF varies, we found a strong positivity for CD34 cells. In contrast, CD45 cells were scarce, as previously reported. Our data on cell phenotype should be interpreted cautiously because the SVF is not a homogeneous population, and the final composition can be affected by the methods used to harvest ADRCs and perform flow cytometry. The use of automatic rather than manual isolation may have contributed to our low percentage of CD105 cells. Finally, the mesenchymal proliferative capacity (CFU-F) of ADRCs in our study supports that reported by Dromard et al.
In the present study, we provide preliminary evidence that harvesting ADRCs and delivering them by transendocardial injection are safe in patients with advanced CAD. No major safety issues were noted, specifically regarding cardiac arrhythmias and clinical adverse events. The safety profile appears to be comparable to that of delivering bone marrow–derived mononuclear cells, and the logistics are suitable for a point-of-care, same-day injection model. The rates of death were similar in our study to those reported in the largest registry of no-option patients. We reported 2 cardiac deaths at 18 months (7.4%), for a total mortality of 18.5%.
Maximal oxygen consumption levels were preserved in the ADRC-treated group at 6 and 18 months as compared with the natural progression of the disease seen in the control group (in whom MVO2 levels progressed toward eligibility levels for heart transplantation). This finding is interesting, considering that patients received only a single treatment with ADRCs over an 18-month period. Moreover, studies show that MVO2 has significant prognostic value in patients with severe heart failure and that directional changes in MVO2 over time have prognostic significance. We believe that monitoring changes in MVO2 levels may be more meaningful than evaluating LVEF in patients with chronic heart failure.
Our results are comparable to those of previous studies with bone marrow–derived cells and MSCs in no-option patients. Both Losordo et al and Haack-Sorensen et al also showed the safety of transendocardial injections and improvement in exercise tolerance as well as quality of life scores in these patients. However, in contrast to MSCs, ADRCs in our study did not reduce scar size or increase LVEF but instead appeared to result in scar stabilization. This may indicate different mechanisms of actions between the 2 cell types.
Recent studies have shown a reduction of scar size after infusion of cardiac stem cells in the subacute and chronic phases of MI. Considering that our patients were in advanced phases of ischemic heart disease, a left ventricular mass increase of 21 g and the maintenance of scar size after ADRC injection suggest a beneficial response to cell therapy in these patients. Although mechanistic study of ADRCs is beyond the scope of this study, we could hypothesize that ADRCs have cardiogenic and angiogenic paracrine effects. Accordingly, a recent report showed that ADRC therapeutic function is induced primarily by paracrine-mediated cardioprotection and angiogenesis. Improvements in perfusion and left ventricular mass may support these hypotheses, but confirmation in a larger study is warranted.
This study has several limitations. The small sample size limits the statistical rigor of our findings and our efficacy conclusions. Because baseline MRI and SPECT measurements varied between the 2 groups, imaging results should be interpreted cautiously and considered as hypothesis generating. However, the double-blind nature of the study and the use of a blind core reading of efficacy measures avoid biases in end point assessments. Furthermore, medication use, which is the most important factor that can influence clinical course in these patients, was similar between groups. Age was also significantly different; the treatment group was older, which may have biased them toward a worse outcome. Furthermore, a 3:1 randomization scheme resulted in a smaller control group, thus reducing the power of the study. Finally, MRI data were available only at baseline and 6 months, precluding longer follow-up.
Discussion
In this multicenter, randomized, placebo-controlled, double-blind study, we have shown for the first time that the harvest and transendocardial injection of ADRCs are safe and feasible in no-option patients with ischemic cardiomyopathy. Moreover, cell therapy with ADRCs in these patients may preserve functional capacity over time and could exert a modest beneficial effect on myocardial perfusion, scar size, and left ventricular contractility.
Adipose tissue comprises a mixed cell source containing a high proportion of MSCs. The stromal fraction has been shown to contain multipotent stem cells. Our approach allowed us to isolate ADRCs in high numbers in real time (<2 hours), during which time baseline mapping procedures were initiated in patients in preparation for cell injections.
Adipose-derived regenerative cells, unlike other cell types such as hematopoietic cells, have no specific cell surface markers. Thus, we used stromal-associated and stem cell–associated markers in our flow cytometric analysis and confirmed that the SVF comprises a heterogeneous cell population. The SVF cell population contained low to medium levels of cells positive for the mesenchymal marker CD105, as previously shown, as well as low to medium levels of cells with endothelial (CD34 and CD31) and hematopoietic (CD45) markers. Although the percentage of CD34 cells in the SVF varies, we found a strong positivity for CD34 cells. In contrast, CD45 cells were scarce, as previously reported. Our data on cell phenotype should be interpreted cautiously because the SVF is not a homogeneous population, and the final composition can be affected by the methods used to harvest ADRCs and perform flow cytometry. The use of automatic rather than manual isolation may have contributed to our low percentage of CD105 cells. Finally, the mesenchymal proliferative capacity (CFU-F) of ADRCs in our study supports that reported by Dromard et al.
In the present study, we provide preliminary evidence that harvesting ADRCs and delivering them by transendocardial injection are safe in patients with advanced CAD. No major safety issues were noted, specifically regarding cardiac arrhythmias and clinical adverse events. The safety profile appears to be comparable to that of delivering bone marrow–derived mononuclear cells, and the logistics are suitable for a point-of-care, same-day injection model. The rates of death were similar in our study to those reported in the largest registry of no-option patients. We reported 2 cardiac deaths at 18 months (7.4%), for a total mortality of 18.5%.
Maximal oxygen consumption levels were preserved in the ADRC-treated group at 6 and 18 months as compared with the natural progression of the disease seen in the control group (in whom MVO2 levels progressed toward eligibility levels for heart transplantation). This finding is interesting, considering that patients received only a single treatment with ADRCs over an 18-month period. Moreover, studies show that MVO2 has significant prognostic value in patients with severe heart failure and that directional changes in MVO2 over time have prognostic significance. We believe that monitoring changes in MVO2 levels may be more meaningful than evaluating LVEF in patients with chronic heart failure.
Our results are comparable to those of previous studies with bone marrow–derived cells and MSCs in no-option patients. Both Losordo et al and Haack-Sorensen et al also showed the safety of transendocardial injections and improvement in exercise tolerance as well as quality of life scores in these patients. However, in contrast to MSCs, ADRCs in our study did not reduce scar size or increase LVEF but instead appeared to result in scar stabilization. This may indicate different mechanisms of actions between the 2 cell types.
Recent studies have shown a reduction of scar size after infusion of cardiac stem cells in the subacute and chronic phases of MI. Considering that our patients were in advanced phases of ischemic heart disease, a left ventricular mass increase of 21 g and the maintenance of scar size after ADRC injection suggest a beneficial response to cell therapy in these patients. Although mechanistic study of ADRCs is beyond the scope of this study, we could hypothesize that ADRCs have cardiogenic and angiogenic paracrine effects. Accordingly, a recent report showed that ADRC therapeutic function is induced primarily by paracrine-mediated cardioprotection and angiogenesis. Improvements in perfusion and left ventricular mass may support these hypotheses, but confirmation in a larger study is warranted.
This study has several limitations. The small sample size limits the statistical rigor of our findings and our efficacy conclusions. Because baseline MRI and SPECT measurements varied between the 2 groups, imaging results should be interpreted cautiously and considered as hypothesis generating. However, the double-blind nature of the study and the use of a blind core reading of efficacy measures avoid biases in end point assessments. Furthermore, medication use, which is the most important factor that can influence clinical course in these patients, was similar between groups. Age was also significantly different; the treatment group was older, which may have biased them toward a worse outcome. Furthermore, a 3:1 randomization scheme resulted in a smaller control group, thus reducing the power of the study. Finally, MRI data were available only at baseline and 6 months, precluding longer follow-up.
