At present, only thrombolysis with tissue plasminogen activator has been approved by regulatory agencies (including the US Food and Drug Administration) in several countries for the treatment of acute stroke. Numerous neuroprotective agents have been tested in both poorly designed and well-designed clinical trials, but so far all have failed to provide convincing evidence of safety and efficacy. Many drugs, and some devices, have been shown to reduce neurological injury and histological damage to the brain in a variety of animal stroke models, but these encouraging experimental findings have not resulted in successful treatments for human stroke victims. Although it is commonly thought that the animal models do not predict the results of human clinical trials, I would contend that some animal models are actually reasonably accurate predictors of human responses to cerebral ischemia and the effects of some types of therapies. There are numerous possible reasons for the trial failures, and we have encountered obstacles at virtually every step in the process of translating theory into clinically practical therapies. I will discuss the reasons for many of the past difficulties and possible ways to achieve better results in the future.
Zivin J. Translational Challenges: From Animal to Human Experimentation. Arch Neurol. 2003;60(2):295. doi:10.1001/archneur.60.2.295-a