Penumbral tissue is subjected to blood flow reduction above the viability threshold. Neurons remain alive while they produce enough energy to maintain their membrane potential, at expenses of suppressing other energy-consuming functions, such as protein synthesis. Disturbances can be transient and not affect tissue viability, for instance protein synthesis inhibition can still reverse after 4 h. Yet, this causes cellular stress as evidenced by strong induction of heat-shock proteins. Besides energy restriction, the penumbra is also subjected to dynamic changes generated by a progressively distorted environment. Alterations within the ischemic core propagate to the neighboring penumbra through various mechanisms, including spreading depression, release of soluble pro-inflammatory mediators to the extracellular space, and direct cell-matrix and cell-cell interactions. These factors may exacerbate ongoing changes in the penumbra and promote cell death. Subsequently, viable penumbral neurons become at risk of death if essential cellular functions do not recover in time and cell death programs are activated. The question that arises is: how long penumbral cells are viable? Penumbra viability likely depends on intrinsic cellular vulnerability as neurons are more vulnerable than other cells, and certain neurons are more vulnerable than others. Type of neurotransmitter receptors, cell size, morphology, cytoarchitectural features, networking connections, neuroprotective resources (e.g. antioxidant capacity, calcium binding proteins), and species peculiarities, among others, might contribute to selective neuronal vulnerability. Nonetheless, the duration of the ischemic insult highly determines penumbra viability. Reperfusion can rescue tissue at risk beyond the 3-hour time window established for thrombolysis with rt-PA. The duration of ischemia also determines cell death characteristics. Molecular and biochemical features of programmed cell death occur at core margins in inverse relationship to the duration of ischemia, as signs of apoptosis are more abundant than necrosis after short ischemic episodes. Fifteen to thirty-minutes middle cerebral artery (MCA) occlusion in rats is a good example of extensive signs of apoptosis. Also, after more severe ischemia programmed cell death is higher in regions with rich collateral circulation, whereas it is negligible in areas subjected to intense and prolonged blood flow restriction. Neuronal death can be delayed in several hours in the cortex compared with the striatum after transient MCA occlusion in rats. Heterogeneity in the neuronal reaction to moderate ischemic conditions might cause necrotic and apoptotic-like responses in neighboring cells. Moreover, the possibility that apoptotic-like signals might arise in cells that also show mild necrotic features cannot be excluded, but in this case inhibiting programmed cell death is unlikely to be beneficial. In transient focal ischemia, signs of programmed cell death have been reported from 4 to 24 h, and they can persist during days. There is possibly a range of time-windows at which tissue at risk can still be rescued, and the efficacy of treatments targeting cell death programs will depend on whether some tissue remains salvageable. Imaging techniques approach the identification of tissue at risk with the diffusion-perfusion ‘mismatch’ concept, but cell death can occur long after reperfusion is restored. This calls for new imaging criteria aimed to identify salvageable tissue at protracted time.
