Cell therapies for the central nervous system: how do we identify the best candidates?

Central to the obstacles to be overcome in moving promising cell-based therapies from the laboratory to the clinic is that of determining which of the many cell types being examined are optimal for repairing particular lesions.

Our studies on astrocyte replacement therapies demonstrate clearly that some cells are far better than others at promoting recovery in spinal cord injury and that, at least in some cases, transplanting undifferentiated precursor cells is far less useful than transplanting specific astrocytes derived from those precursor cells. But further comparison between different approaches is hindered by the difficulties in replicating results between laboratories, even for well defined pharmacological agents and bioactive proteins. These difficulties in replication appear most likely to be due to unrecognized nuances in lesion characteristics and in the details of delivery of therapies.

We propose that the challenge of reproducibility provides a critical opportunity for refining cell-based therapies. If the utility of a particular approach is so restricted that even small changes in lesions or treatment protocols eliminate benefit, then the variability inherent in clinical injuries will frustrate translation. In contrast, rising to this challenge may enable discovery of refinements needed to confer the robustness needed for successful clinical trials.