Stem cell tracking in human trials: a meta-regression

The potential effectiveness of cell therapies is dependent upon homing of transplanted cells to relevant target organs. In this study we firstly characterise the range of methods employed in all human therapeutic-cell studies published to date investigated with cell-tracking. Secondly, we determine factors that predict target-organ cell uptake efficiency by meta-regression. Following a comprehensive literature search, we identified 19 relevant trials, representing 145 patients over the following 7 diseases: myocardial infarction; Chagasic cardiomyopathy; ischemic stroke; traumatic injury of brain or spinal cord; diabetes and cirrhosis. Cell-labelling strategies employed were: 18-fluorodeoxyglucose-PET, 111-indium-SPECT; 99-technetium-SPECT, and iron oxide-MRI. The following methodological parameters were extracted: label type; label dose; labelling efficiency; stability; cell dose; percentage labelled cells; disease type and chronicity; cell purity; cell type; and cell uptake efficiency. Meta-regression techniques were used to identify predictors of cell-labelling efficiency; viability and cell uptake efficiency. These analyses found that labelling efficiency is proportionate to cell dose, while cell viability is lowest with indium and long label incubation times. Uptake efficiency of cells is predicted by stem cell purity (positive association) and cell infusion number (negative association), although these two variables are themselves strongly negatively correlated between studies. In summary the methodological factors associated with enhanced therapeutic-cell homing from both our own analysis, and within-trial comparisons, are: acute (versus chronic) disease, selective stem cells (versus unselected cells), and intra-arterial (versus intravenous) delivery. However, future trials need to keep cell doses and imaging times constant so as to enable meaningful comparisons in uptake efficiency.