Assessing post-stroke cognition in pre-clinical models: Lessons and recommendations from a multi-center study.

Cognitive decline is a significant long-term consequence of stroke with no available treatments. To aid in therapy development, we sought to achieve robust detection of cognitive performance after stroke in a multi-site design. Ischemic stroke was induced in adult and middle-aged male C57BL/6 mice utilizing permanent distal middle cerebral artery occlusion (dMCAO), dMCAO with hypoxia or transient MCAO by filament insertion. Cognitive outcomes were assessed by Novel Object Recognition (NOR) and Barnes Maze (BM) tests prior to surgery, and during sub-acute (1-2 weeks) and chronic (8 weeks) phases post-stroke. Histology and immunostaining were used to assess infarct size, tissue damage and neuronal loss, and plasma neurofilament light chain levels were quantified. We did not detect a cognitive deficit using NOR, but detected significantly worse performance in dMCAO mice with the BM. Multi-site BM validation will be a key next step. Overall, our study highlights common challenges in detecting post-stroke cognitive impairment within the pre-clinical stroke community, and underscores the complexities of designing and executing these studies, particularly as applied to a multi-site structure. We provide recommendations and suggest important considerations for future stroke cognition studies, whether operating as an individual lab or a multi-site group.