Predicting experimental success: a retrospective case-control study using the rat intraluminal thread model of stroke.

The poor translational success rate of preclinical stroke research may partly be due to inaccurate modelling of the disease. We provide data on transient middle cerebral artery occlusion (tMCAO) experiments, including detailed intraoperative monitoring to elaborate predictors indicating experimental success (ischemia without occurrence of confounding pathologies). The tMCAO monitoring data (bilateral cerebral blood flow, CBF; heart rate, HR; and mean arterial pressure, MAP) of 16 animals with an 'ideal' outcome (MCA-ischemia), and 48 animals with additional or other pathologies (subdural haematoma or subarachnoid haemorrhage), were checked for their prognostic performance (receiver operating characteristic curve and area under the curve, AUC). Animals showing a decrease in the contralateral CBF at the time of MCA occlusion suffered from unintended pathologies. Implementation of baseline MAP, in addition to baseline HR (AUC, 0.83, 95% c.i. 0.68 to 0.97), increased prognostic relevance (AUC, 0.89, 95% c.i. 0.79 to 0.98). Prediction performance improved when two additional predictors referring to differences in left and right CBF were considered (AUC, 1.00, 95% c.i. 1.0 to 1.0). Our data underline the importance of peri-interventional monitoring to verify a successful experimental performance in order to ensure a disease model as homogeneous as possible.