Here we show that transcatheter, intramyocardial injection of autologous bone marrow preparations in patients with chronic ischemic cardiomyopathy is well tolerated and produces functional recovery in scarred myocardium and reverse remodeling of the LV chamber. Importantly, the functional recovery is evident at 3 months following injection, and precedes reverse remodeling. As all patients had documented MI and were previously revascularized, these findings support myocardial regeneration and mirror the phenotype observed in porcine models of chronic ischemic cardiomyopathy.1,5
The present findings have several major implications. First, we have established that bone marrow derived cell therapy clearly can enable reverse remodeling of dilated hearts, a phenotypic change that would be expected to have positive clinical implications. Second, the reverse remodeling caused substantial parallel declines in systolic and diastolic volumes, so that an EF increase was obscured; accordingly, this strongly suggests that EF is not a good primary endpoint for assessing responses to cell therapy in dilated hearts and that chamber size, MI size, or regional function are more likely to reflect a favorable outcome. Finally, these data strongly indicate that a high-resolution imaging tool such as MRI, capable of detecting the latter outcomes, is necessary for trials of cardiac regeneration.
While there is enormous enthusiasm for the possibility of repair of ischemic cardiomyopathy with cell-based therapy, mechanisms and manifestations of this strategy remain highly controversial. To address this, we phenotyped patients using cardiac MRI, which is recognized as the “reference standard” for quantitative assessment of LV function and has been shown to be the most accurate and reproducible imaging modality to assess global and regional LV function.6,7
Regional functional improvements (peak Ecc) of the infarct territory were calculated by harmonic phase MRI, arguably the most accurate non-invasive analytical imaging modality to measure regional strain8,9
The early improvements in regional contractility of a treated infarct predicted later reverse remodeling. Regional strain of the infarct territory improved as early as 3 months following cell injection, which persisted at 6 months and 1 year (). The reduction in chamber volumes was not evident until 6 months and the correlation with improved regional strain was not apparent until 6 months and continued at 1 year (). This suggests that the first manifestation of bone marrow cell therapy on the heart is improved regional function that later contributes to reverse remodeling. As a corollary to this finding, infarct size was reduced by 3 months after treatment.
Preclinical studies of bone marrow cell therapy for ischemic heart disease have demonstrated improved EF, but this has not translated to early phase studies in humans. We show a strong parallel decrease in both EDV and ESV (r2
=0.87, p=0.002) 12 months after treatment. Indeed this finding strongly suggests that EF may not be a good outcome measure for studies of cell therapy for remodeled ventricles. Importantly, both LV size and infarct size are important clinical indicators of outcome in patients with ischemic cardiomyopathy,10
and the results here highlight a highly adaptive response to cell therapy, even in the absence of a net increase in EF.
This study lacks the power to determine superiority between MSCs or whole bone marrow, and does not have a placebo comparison group. Although, only 4 patients had longitudinal assessment of scar size (due to artifact distortion from delayed enhancement imaging from ICD leads), MI size was reduced in each of these subjects. Each of these limitations will be definitively addressed in the ongoing TAC-HFT and POSEIDON trials, which together will treat 90 patients in placebo-controlled, blinded fashion.
In conclusion, MRI provides a unique opportunity to image the in-vivo structural and functional changes after bone marrow stem cell therapy for the heart. Our data suggest human autologous bone marrow progenitor cells increase regional contractility of injected myocardial scar tissue within 3 months of treatment, and these functional changes are associated with later reverse remodeling. These findings strongly support ongoing clinical investigation of cell-based therapy for cardiomyopathy, and support the use of regional function and chamber size as important endpoints.