Table 1.

Issues in translation from rodent experiments to human clinical trials.

Desirable characteristic of rodent experiment controlled by investigator	Limitations in translation to human subjects	Methods to mitigate human subjects limitations
EXPERIMENTAL SUBJECT POPULATION
Genetic background Rearing conditions Subject availability Age, gender medical comorbidities Inexpensive to obtain Relatively distant oversight by institutional review board and animal care committee	Most human populations genetically heterogeneous Wide variety of socioeconomic and activity backgrounds Enrollment dependent on flow of stroke patients, and lengthy periods are usually required to identify and enroll sufficient numbers of participants. Close review and oversight by Institutional Review Boards; heavy documentation required	Measure important personal characteristics, analyze as covariates Use randomization method to balance for important subject-specific covariates Adjust inclusion/exclusion criteria to minimize heterogeneity without unduly affecting enrollment Adaptive trial designs to ensure participants are randomized only to promising study arms Do studies in large centers with high patient throughput, or multicenter trials
BRAIN LESIONS
Defined time of onset Reproducible mechanism of injury Reproducible lesion location and size Injuries occur in otherwise pristine brain Muted immunological response to injury Quick and relatively complete motor recovery; greater recruitment of brainstem and extrapyramidal structures	Time of onset can be ambiguous Multiple stroke mechanisms Cause of stroke often undetermined Wide variety of stroke lesions Prior stroke and white matter changes present Greater and more variable immunological response to injury Slower and variable motor recovery	Use study designs that do not require precise time of onset (e.g., wide enrollment windows) Use randomization method to balance for important subject-specific covariates Use of stratification in design Use study designs that do not require specific stroke mechanisms, or ignore lesions and recruit based on clinical impairments Require specific lesions or lack of background brain changes
TRAINING CONDITIONS
Timing Amount Consistency of training paradigm Motivation: High with food rewards	Little control over timing of patient presentation acutely, during inpatient or outpatient rehab, or chronic care setting Need to adapt training to clinical environment, which cannot be controlled by research team Need to recruit over months or years can lead to drift in participant training and assessment methods Amount of training can be dictated by unrelated factors (insurance, transportation, etc) Training is an interaction between unique therapist and unique participant with specific impairments, thus difficult to standardize Motivation to participate in training program limited by lack of knowledge about stroke recovery, cognitive impairment, depression	Focus study activities at specific clinical milestones (rehabilitation admission, initiation of outpatient therapies, etc.) Study provides pragmatic support to overcome insurance payment and transportation barriers Treatment protocols that are flexible for participant needs, but reproducible and well quantified Scheduled audits of training protocol execution and outcome assessment Access increased motivation through patient-centered training activities
BACKGROUND CONDITIONS
Post-injury environment Follow up Diet and activity Motor training outside of study	Differences in care across multiple institutions Variations in home environment and social support Loss to follow up due to subject withdrawal, moving away, medical events Little control over diet Little control over therapies prescribed outside of study	Recruit from one large institution or standardize practices across multiple sites Measure home environment and social support, treat as covariate Select participants who are socially stable and unlikely to withdraw or move Discourage or prohibit outside therapies as a condition of study participation; or measure and treat as a covariate
ABILITY TO STUDY MECHANISM OF RECOVERY
Sacrifice of subjects for anatomic and metabolic studies Homogenous genetic background	Brain tissue rarely available Cerebrospinal fluid difficult to obtain Only non-invasive or minimally invasive assessments available	Use of inferential evaluations: multi-omics of peripheral blood, functional MRI, electrophysiology