Investigating cerebral ischemia requires animal models relevant to human stroke. A precise knowledge of the strengths and shortcomings of available models is mandatory for effective research in neuroprotection 1 2 . Initially described in the rat 3 and subsequently adapted in mice 4 , permanent middle cerebral artery (MCA) occlusion (pMCAO) by electrocoagulation is a widely used model of focal ischemia. Invasive surgical procedures are required: temporal muscle dissection, in some cases by electrical cauterization 5 6 , subtemporal craniotomy and MCA electrocoagulation. This model, by interrupting blood flow at the level of the parietal cerebral artery branch of the MCA (distal occlusion), has the advantage of producing smaller, cortical-restricted, more reproducible and better-tolerated infarcts compared to suture MCAO, which endoluminal occluder is situated in the internal artery, at the birth of the MCA (proximal occlusion) and gives rise to extensive cortico-striatal infarcts 7 8 . It may however induce traumatic brain damage.

To our knowledge, sham-operated animals have only been studied using immunohistology at the subacute stage of cerebral ischemia. Early monitoring with magnetic resonance imaging (MRI) may facilitate in vivo identification of traumatic brain injury during pMCAO.

Temporal muscle cauterization in the absence of MCA electrocoagulation (group A) consistently produced an extensive lesion across the frontoparietal cortex (n = 6/6). These lesions appeared hyperintense on T₂WI within the first hour after surgery, with an early mass effect and reduced ADC (Figure 1A). Temporal muscle incision and craniotomy in the absence of MCA electrocoagulation (group B and C, respectively) caused no visible brain injury on day 0 or day 1 MRI (n = 4/4 in groups B and C, Figure 1B and 1C). MCA electrocoagulation (group D) induced an ischemic lesion in the frontoparietal cortex with reduced ADC on day 0 (n = 6/6). These lesions were not visible on T₂WI within 1 h post-surgery (Figure 1D). No bleeding was noted during the surgery, but MCA electrocoagulation occasionally produced a small superficial traumatic lesion (n = 2/6 in group D, figure 1D). When pMCAO was performed using temporal muscle cauterization (group E), the lesion appeared as a superimposition of the almond-shaped cauterization lesion and of the MCA-territory bounded ischemic lesion (n = 2/2, Figure 2). Infarcts from group A, D and E were clearly delineated on day 1 T₂WI.

Our results showed that temporal muscle cauterization in the absence of ischemic stimulus produced a hyperintense area on T₂WI highly suggestive of traumatic injury. Indeed, T₂ increase is not expected in ischemic lesions within the first hours of occlusion in adults, although early T₂ increase has been reported in neonates after MCA electrocoagulation 10 or hypoxia-ischaemia 11 . Early mass effect is uncommon after ischaemia in the adult rat 12 or mouse 13 . Acute ADC decrease is not specific of ischemic insult and has been described in both experimental and human traumatic brain injury 14 15 . Signal suggestive of intra- or extra-cranial surgery-related bleeding (T2 hypointensity) would have been ascertained using T2*WI, but were not noted. No traumatic injury was observed after temporal muscle incision. Accordingly, incision should be preferred over cauterization for temporal muscle dissection.

The MRI appearance of lesions resulting from both temporal muscle cauterization (traumatic damage) and occlusion (ischemic damage) may mimic an infarction, especially if imaging is done at later time points. Histopathological analyses are usually performed after a delay of 24-72 h, when traumatic and ischemic damage may not be discernable, while early histological examination of intracerebral coagulation necrosis would be required to ascertain the thermal origin of traumatic damage induced by muscle cauterization.

In the last decade, high resolution MRI has become a valuable tool for monitoring tissue damage in rodent models of cerebral ischemia. Early MRI monitoring may help to identify non-specific brain injury that could hamper neuroprotective drugs assessment.

The authors declare that they have no competing interests.

FC carried out the MRI experiments, participated in the design of the study and helped to draft the manuscript. SM carried out surgery. MW designed and optimized the MRI acquisition protocol, and helped to draft the manuscript. JCB participated in the MRI experiments. YB and NN conceived the study, and participated in its design and coordination. THC conceived the study, participated in its design, performed image analysis, and drafted the manuscript. All authors read and approved the final manuscript.