Stroke is due to the occlusion of a major brain vessel by a thrombus or an embolus resulting in an arrest of the blood flow to the brain area supplied by this vessel. Nerve cells located in the concerned brain area are depleted of energetic substrates, i.e. glucose, and oxygen, they become ischemic. Ischemia results in an ATP depletion, loss of ionic homeostasis, Ca++ influx into the neurones and massive release of excitatory aminoacids, mainly glutamate. In turn these events lead rapidly to a necrotic neuronal death. A subpopulation of neurones enters into a process of active cell death: apoptosis. Apoptosis is a programmed cell death that requires protein synthesis and goes through a cascade of events. This active cell death is a normal phenomenon during the embryonic development, also in the CNS, and allows eliminating the excess of neuronal cells.

Therapeutic intervention has up to now mainly aimed at neuroprotection, that means saving neurones by interfering with the mechanisms leading to necrotic or apoptotic neuronal death. Despite promising preclinical results obtained mainly in rodent models of cerebral ischemia clinical trial were not successful. Reasons for failure are diverse; one can site cardiovascular and central side effects, dose-finding problems. The major concern of preclinical scientists and clinicians is the time window of opportunity for treatment. Ischemic stroke is time-dependent dynamic process. In the "core" region where CBF is reduced to levels below 10%, neurones become necrotic within minutes. Whereas, in the surrounding "penumbra" CBF levels are below 50% that is enough to maintain ionic homeostasis but not normal function. These neurones in the penumbra are potential targets for neuroprotection. The core expands through mechanisms like spreading depression and with time penumbral neurones die by necrosis or apoptotsis, finally the necrotic tissue encompasses also the penumbral tissue. Therefore, the time-window for treatment is important. Clinical trials with the tPA, showed that after 3 hours recanalization of the occluded vessel has no beneficial effect. These trials demonstrated that in university hospitals as few as 6% of the patients were treated within 3 hours following the ictus, whereas the proportion was only of 3% in community hospitals.

The loss of population of neurones through an ischemic or traumatic lesion induces neurologic deficits, e.g. sensorimotor or cognitif deficits. These functional impairments are not permanent, in-patients as well as in experimental models of stroke or traumatic brain injury a partial or even complete functional recovery is often observed. Therefore we propose that enhancement of the functional recovery as an alternative to the neuroprotective strategy. Functional recovery in patients can be promoted with physiotherapy or pharmacological treatments like amphetamine and a1 adrenergic agonists. Nevertheless the molecular and cellular mechanisms underlying this phenomenon are not fully understood. Recently it has been shown that after an ischemic insult a subpopulation of neurones re-expresses embryonic cellular markers (cyclin D, nestin). Treatment with staurosporin that stimulates apoptosis decreases the size of the infarct. These results suggest that apoptosis may be beneficial by favouring the reorganisation of the lesioned brain and thus facilitating functional recovery.

We propose to modelize the cortico-thalamic circuitry (figure 1) with a computer program like SIRENE (1). First we are going to use the model to simulate a massive (necrosis) or discrete (apoptosis) neuronal death in the sensorimotor cortex for instance. We will attempt to reproduce the consequences on the descending or ascending cortico-thalamic circuitry (degeneration of the cortico-spinal tract and a secondary neuronal death in the thalamic nuclei) of the lesion and the resulting neurological outcome

In a second step we will integrate in the model known features of neuronal plasticity, e.g. axonal regrowth, functional take-over, in order to modelize the process of functional recovery.

 

(1) Villa A.E.P., Zurita P. (1987) SImulation de REseaux de Neurones : an interactive software written in LISP to study the behaviour of a neural network. Neuroscience 22 : S847