Found in translation: preclinical stroke research predicts human pathophysiology, clinical phenotypes, and therapeutic outcomes

Stroke presents a major global burden to patients, their relatives, and whole economies.1 Worldwide intense efforts are being undertaken to understand the pathobiology of stroke and to develop novel, effective treatments. This has led to the discovery of a plethora of preclinically promising treatment strategies,2 but intravenous thrombolysis remains the only specific pharmacological treatment with proven efficacy in acute ischemic stroke.3 Apparently, of several hundred controlled clinical trials4 aiming to protect the brain against ischemia (neuroprotection) with compounds previously found effective in animal experiments, not a single one has led to the regulatory approval of a drug for this indication. Many pharmaceutical companies have, therefore, stopped developing treatments for this devastating disease, and nihilism is spreading in the field. A roadblock seems to exist between bench and bedside, preventing the translation of highly effective preclinical neuroprotectants into the clinical realm.5 Despite the prevailing mantra of translational medicine,6 other fields report similar problems when making the preclinical-to-clinical transition. Much has been speculated about the reasons underlying the translational roadblock, and the phrase lost in translation has entered the titles of numerous biomedical publications. In this article, we propose that contrary to current opinion, preclinical stroke research can indeed predict clinical outcomes and phenotypes. Numerous neuroprotectants were reported as lost in translation (Figure 1A), but there is reason to argue that the clinical trials were based on rather weak preclinical evidence, and that many pathophysiological principles and therapies were actually discovered in translation (Figure 1B). We propose that there is little reason for nihilism. We present a concise list of potential improvements that may help to reduce the current attrition rate of bench-to-bedside translation, even in neuroprotection. Stroke may even serve as a model disease because our analysis may apply to other fields where development of …