Implantation site and lesion topology determine efficacy of a human neural stem cell line in a rat model of chronic stroke.

Stroke remains one of the most promising targets for cell therapy. Thorough preclinical efficacy testing of human neural stem cell (hNSC) lines in a rat model of stroke (transient middle cerebral artery occlusion) is, however, required for translation into a clinical setting. Magnetic resonance imaging (MRI) here confirmed stroke damage and allowed the targeted injection of 450,000 hNSCs (CTX0E03) into peri-infarct tissue, rather than the lesion cyst. Intraparenchymal cell implants improved sensorimotor dysfunctions (bilateral asymmetry test) and motor deficits (footfault test and rotameter). Importantly, analyses based on lesion topology (striatal vs. striatal + cortical damage) revealed a more significant improvement in animals with a stroke confined to the striatum. However, no improvement in learning and memory (water maze) was evident. An intracerebroventricular injection of cells did not result in any improvement. MRI-based lesion, striatal and cortical volumes were unchanged in treated animals compared to those with stroke that received an intraparenchymal injection of suspension vehicle. Grafted cells only survived after intraparenchymal injection with a striatal + cortical topology resulting in better graft survival (16,026 cells) than in animals with smaller striatal lesions (2,374 cells). Almost 20% of cells differentiated into glial fibrillary acidic protein+ astrocytes, but <2% turned into FOX3+ neurons. These results indicate that CTX0E03 implants robustly recover behavioral dysfunction over a 3-month time frame and that this effect is specific to their site of implantation. Lesion topology is potentially an important factor in the recovery, with a stroke confined to the striatum showing a better outcome compared to a larger area of damage.