The brain is highly sensitive to disturbance of its blood supply. Stroke is a devastating disease and is the third most common cause of death, and the most common cause of motor and mental disability in adults, in developing countries 1. Complex pathophysiological events occur in brain during ischemic processes, and these are considered responsible for cell damage leading to neuronal death (for review see 23). However, it is now generally accepted that the mammalian brain may be more resistant to ischemia than previously thought. This raises the possibility of therapeutic intervention before brain damage has become irreversible.

A number of interacting and sequentially evoked events tend to reinforce the initial ischemic insult. A key role in these processes is played by post-ischemic inflammation. The Ca²⁺-related activation of intracellular second messenger systems, the increase in reactive oxygen species formation, as well as hypoxia itself triggers the expression of a large number of pro-inflammatory genes following cerebral ischemia. Thus, mediators of inflammation such as platelet-activating factor (PAF), tumor necrosis factor α (TNFα), interleukin 1β (IL-1β), chemokines (IL-8, monocyte chemoattractant protein-1) and other pro-inflammatory factors are produced by the ischemic brain tissue 3. In addition, the expression of adhesion molecules with the subsequent infiltration of polymorphonuclear leukocytes occurs following ischemic stroke. Results from several studies also suggest that the marked and sustained expression of inflammation-related enzymes such as inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) plays an important role in the secondary events that amplify cerebral injury after ischemia 456789101112.

Nimesulide (N-(4-nitro-2-phenoxyphenyl)-methanesulfonamide) is a non-steroidal anti-inflammatory drug with potent effects. It shows a high affinity and selectivity for COX-2 with a COX-2/COX-1 IC₅₀ selectivity ratio of 0.06 (whole blood assay) 13. Nimesulide readily crosses the intact blood-brain barrier in both humans and rodents 1314. Several recent studies have demonstrated a marked neuroprotective effect of nimesulide on chronic cerebral hypoperfusion 15, kainate-induced excitotoxicity 16, quisqualic acid-induced neurodegeneration 17, diffuse traumatic brain injury 1819, glutamate-mediated apoptotic damage 20 and induction of the expression of the B subunit of endogenous complement component C1q (C1qB) in transgenic mice with neuronal overexpression of human COX-2 21.

Recently, we have found a significant neuroprotective effect of nimesulide both in global cerebral ischemia 1022, a type of injury that mimics the clinical situation of cardio-respiratory arrest, and in a rat model of ischemic stroke induced by the transient (1 h) occlusion of the middle cerebral artery 12.

Since most cases of human ischemic stroke are caused by permanent occlusion of cerebral arteries 23242526, the present study was conducted to assess whether nimesulide would also show neuroprotective efficacy on the cerebral infarction induced by permanent middle cerebral artery occlusion (pMCAO) in the rat, a clinically relevant model of ischemic stroke. The effects of the COX-2 inhibitor nimesulide had not been previously investigated in a model of permanent ischemic stroke.

The present study was prompted by our previous encouraging results with nimesulide in a model of transient focal cerebral ischemia, which show that this COX-2 inhibitor is able to potently reduce infarct volume and improve functional recovery 12. These neuroprotective effects are also observed when treatment is delayed until even 24 h after the onset of ischemia 12. Since we believe that it is very important to perform thorough, multifactorial and well-designed pre-clinical studies before assuming definitive conclusions on the neuroprotective effect of any compound, and considering that in stroke patients a very early spontaneous recanalization of an obstructed brain vessel is, unfortunately, only rarely found, we conducted the present investigation to shed more light into the effects of nimesulide on ischemic damage using a permanent stroke model in the rat considering that this model might be more relevant to the clinical situation of stroke, as suggested previously 23242526.

The core findings of this study are: (i) administration of clinically relevant doses of nimesulide confers protection against the damage induced by permanent focal cerebral ischemia in two modalities (reduction of infarct size, and improvement of functional outcome) and (ii) nimesulide's neuroprotection is still evident when the first administration is delayed until 2 h after the onset of stroke.

Depending on the experimental conditions, the temporal evolution of ischemic damage may vary considerably 304041. Thus, it is very important to characterize the time course of brain damage, especially if one wants to interpret correctly the effects of a given compound using delayed treatment schedules. Our results showed that in the permanent model of stroke induced by the occlusion of the middle cerebral artery using an intraluminal suture, the infarct size progresses very fast in the subcortical areas (mainly striatum) and much slower in cortical areas, but in general the evolution of damage is relatively quick, reaching maximal values by 24 h after the insertion of the filament (Fig. 1A). These findings are in line with those published previously in this model of stroke 4243. Although infarct size continues to increase between 12 and 24–48 h of ischemia (Fig. 1A), the neurological deficits and motor impairment reached their maximum by 12 h, and the animals did not showed any further deterioration of their neurological functions (Fig. 1B and 1C). This might reflect the fact that unlike ischemic injury to many other tissues, the severity of disability is not predicted well by the amount of brain tissue lost. For example, damage to a small area in the medial temporal lobe may lead to severe disability, while damage to a greater volume elsewhere has little effect on function 2. There is not always a direct correlation between the lesion size and the severity of neurological deficits as demonstrated before in animal models 2944 and in stroke patients 45. For that reason, it is essential to evaluate the neuroprotective effects of agents by combining both histological and functional measures. The present study offers a good example of this: even when the lowest doses of nimesulide did not reduce infarct volume in pMCAO (Table 1), one can not minimize the beneficial effects of these doses since a significant reduction in neurological deficits and an improvement of rotarod performance were observed (Table 2). Thus, further studies would be required to better characterize the effects of the lowest doses of nimesulide (3 and 6 mg/kg) in models of cerebral ischemia.

Repeated treatments with nimesulide afforded a more remarkable neuroprotection than the administration of a single dose given before the insult (Tables 1 and 2). These data show the importance of continuous long-term administration after ischemic damage in clinical trials to achieve the maximal beneficial effects of neuroprotection by nimesulide.

Unfortunately, a large number of promising neuroprotective compounds identified from preclinical experiments have failed in clinical trials in stroke patients 3454647. Although several factors may contribute to these disappointing results, an important issue is the 'therapeutic time window of protection', defined as the time period after the onset of ischemia during which administration of treatment is effective 4849. Most of the agents that confer protection in experimental animal models of stroke when given before or a short period after cerebral ischemia have failed in clinical studies 4550. Thus, the assessment of the therapeutic time window of protection is of paramount importance in pursuing future therapies to treat stroke victims.

Therefore, our next experiments were conducted to evaluate the effects of nimesulide when administered in a delayed treatment schedule in order to establish the therapeutic time window of protection of this COX-2 inhibitor in pMCAO, thus increasing predictive outcome in the clinic. Interestingly, reduction in infarct size and neurological deficits and improvement of rotarod performance were still observed when nimesulide treatment was delayed until 2 h after ischemia (Fig. 2A, Table 3).

It is important to compare our present results in pMCAO with those previously obtained in transient ischemia 12. In the model of transient focal ischemia, the time window of nimesulide's neuroprotection extends over a 24 h period 12, and in other models of cerebral ischemia, the time window of protection of nimesulide is similarly wide 102251. These results have been also obtained with other COX-2 inhibitors (e.g., NS-398, SC58125 and rofecoxib) in models of transient ischemic stroke 452 and global cerebral ischemia 5354. These studies suggest that although the protective effects of COX-2 inhibitors are more beneficial when administered early after the ischemic insult, COX-2 selective inhibitors show a wide therapeutic window for the prevention of neuronal death in both focal and global ischemia.

However, our present results suggest that in permanent stroke, COX-2 inhibition by nimesulide is not as protective as in transient models (39 % of infarct reduction with pre-treatment in pMCAO vs. 60 % lesion reduction in transient ischemia with immediate treatment) and the therapeutic time window is narrower as compared to temporary occlusion models (2 h in pMCAO vs. 24 h in transient ischemia) as the present results (Tables 1 and 3, Fig. 2) and our recent studies indicate 12. The vascular inaccessibility of nimesulide into the ischemic/infarcted region could be a plausible explanation for these findings considering that, unlike transient ischemia, in permanent ischemic stroke the protective effects of any drug/agent depend, in large part, on the ability of the compound to reach the ischemic areas mainly through passive diffusion. Although these findings on nimesulide's effects on stroke tempted us to conclude that COX-2 selective inhibitors are less protective in permanent than in transient stroke models, these results should be interpreted with caution since a structurally similar COX-2 inhibitor (NS-398) reduced permanent stroke damage in mice when the treatment started 24 h after MCA occlusion 55 and the same COX-2 inhibitor also reduced lesion size when administered starting 6 h after pMCAO in another previous study 5. Apparent discrepancies between our present results and these two reports 555 might be due to different methods to induce pMCAO (intraluminal vs. distal MCAO involving craniectomy), the specific COX-2 inhibitor used, treatment paradigm or animal species. This emphasizes the importance of conducting more preclinical studies with COX-2 selective inhibitors before these agents could be used in clinical trials in stroke patients. Another issue that needs urgent consideration in future studies with COX-2 inhibitors in cerebral ischemia is the effect of long-term treatment since anti-inflammatory interventions could interfere with nervous regeneration/plasticity and recovery as demonstrated in some types of neuronal injury 5657.

In summary, the present study has evaluated for the first time the neuroprotective effects of the COX-2 inhibitor nimesulide in permanent focal cerebral ischemia, showing beneficial effects on reduction of infarct volume and improvement of functional recovery. This ability of nimesulide to diminish permanent ischemic damage is observed even when the first treatment was delayed 2 h after the ischemic episode. Taken together, these results have important implications for the therapeutic potential of using the COX-2 selective inhibitor nimesulide in the treatment of cerebral ischemia.

COX-2, cyclooxygenase-2; pMCAO, permanent middle cerebral artery occlusion; MCA, middle cerebral artery; TTC, 2,3,5-triphenyltetrazolium chloride; ANOVA, analysis of variance

The author(s) declare that they have no competing interests.

ECJ carried out the surgical procedures to induce stroke, participated in the design of the study and in the statistical analysis, reviewed the data and drafted the manuscript. NHM performed the evaluation of neurological deficits and rotarod performance. AGF and MGC performed the calculation of the infarct volumes and participated in the statistical analysis of the data. OSL, EM and BLF participated in the design and coordination of the study, reviewed the data, provided consultation and helped to draft the manuscript. OSL and BLF share senior authorship. All authors read and approved the final manuscript.