Ischemic Stroke and the Ischemic Cascade

Ischemic stroke occurs when a thrombus or embolus occludes a cerebral artery, abruptly reducing cerebral blood flow (CBF) to the tissue it supplies. Normal gray-matter CBF is about 50–60 mL per 100 g of tissue per minute. Neurons are exquisitely dependent on continuous oxygen and glucose delivery because the brain has essentially no fuel reserves and relies almost entirely on oxidative phosphorylation for ATP. When flow drops, tissue fate depends critically on how far it drops and for how long — this is the basis of the flow-dependent injury thresholds described below. The resulting damage is not uniform across the affected territory. At the center of the occluded vascular territory, flow falls to near zero and cells die within minutes — this is the ischemic core, essentially unsalvageable by the time it is imaged. Surrounding the core is the ischemic penumbra, tissue that is functionally silent (it stops firing, and the patient shows a clinical deficit) but structurally alive, kept marginally perfused by collateral vessels from neighboring arterial territories. The penumbra is the entire therapeutic target of acute stroke medicine: it is dying, but slowly, and its fate is reversible if flow is restored or if the injurious cascade is interrupted before structural damage accumulates. The ischemic cascade is the mechanistic backbone connecting "no blood flow" to "dead neuron." It is conventionally divided into phases by their time course — seconds-to-minutes energy failure, minutes-to-hours excitotoxicity and spreading depolarization, hours-to-days inflammation and programmed cell death, and days-to-weeks tissue remodeling. These phases overlap rather than proceeding in strict sequence, and each successive phase both depends on and amplifies the ones before it, which is why interrupting the cascade early yields disproportionately large benefits and why the therapeutic window shrinks rapidly with time.

Step 1 — Energy failure. Oxygen and glucose deprivation halts the mitochondrial electron transport chain and oxidative phosphorylation. ATP, normally maintained at millimolar concentrations, is depleted within roughly 2–4 minutes of complete ischemia. Without ATP, the Na⁺/K⁺-ATPase — which normally pumps 3 Na⁺ out and 2 K⁺ in per cycle, maintaining the steep transmembrane ion gradients that make neurons excitable — fails. Na⁺ and Ca²⁺ leak into the cell down their electrochemical gradients; K⁺ leaks out. The resting membrane potential collapses toward zero (anoxic depolarization).

Step 2 — Ionic collapse and glutamate release. Depolarization opens voltage-gated Ca²⁺ channels and reverses the normal operation of glutamate transporters (EAATs), which begin pumping glutamate out of astrocytes and neurons instead of clearing it. Extracellular glutamate concentrations, normally kept below 1 μM, can rise into the tens to hundreds of micromolar range in ischemic tissue. This glutamate flood over-activates NMDA and AMPA receptors, driving further massive Ca²⁺ influx — the excitotoxic mechanism detailed under Excitotoxicity and Glutamate-Mediated Cell Death.

Step 3 — Calcium-triggered enzymatic destruction. Pathologically elevated intracellular Ca²⁺ (rising from ~100 nM to micromolar levels) activates calpains (proteases that degrade cytoskeletal proteins), phospholipases (which liberate arachidonic acid and generate free radicals and eicosanoids), endonucleases (fragmenting DNA), and nitric oxide synthase (generating NO, which combines with superoxide to form peroxynitrite, a potent oxidant). Mitochondria sequester Ca²⁺ until they are overwhelmed, at which point the mitochondrial permeability transition pore opens, collapsing the membrane potential and releasing cytochrome c — a trigger for apoptotic caspase cascades.

Step 4 — Peri-infarct depolarizations. In the penumbra, cells are too well-perfused to die immediately but poorly enough perfused that they cannot fully repolarize. Waves of spreading depolarization — similar in form to cortical spreading depression — propagate outward from the core across the penumbra, repeatedly forcing already energy-starved cells to spend scarce ATP restoring ion gradients. Each wave transiently worsens the local oxygen/glucose supply-demand mismatch, progressively recruiting penumbral tissue into the core. This is a major mechanism by which the infarct grows over the first many hours.

Step 5 — Inflammation and blood-brain barrier breakdown. Over hours, dying cells release damage-associated molecular patterns (DAMPs) that activate resident microglia and astrocytes. Pro-inflammatory cytokines (TNF-α, IL-1β, IL-6) and matrix metalloproteinases (notably MMP-9) are upregulated; MMP-9 degrades the basal lamina and tight-junction proteins of the blood-brain barrier, permitting neutrophil infiltration and vasogenic edema. Peak brain swelling from edema typically occurs 2–5 days post-stroke and is itself a major cause of secondary injury and, in large infarcts, life-threatening herniation.

Step 6 — Apoptosis and delayed cell death. In less severely affected penumbral tissue, cells activate intrinsic (mitochondrial, cytochrome-c/caspase-9) and extrinsic (death-receptor/caspase-8) apoptotic pathways over hours to days, rather than dying by acute necrosis. This delayed, more "programmed" death is mechanistically distinct from the necrosis of the core and — critically — is in principle interruptible for longer, which is part of why neuroprotection strategies target this phase even after the reperfusion window has closed.

Astrup, Siesjö, and Symon (1977–1981): defining the penumbra by direct measurement. Using microelectrodes and local cerebral blood flow measurement (via hydrogen clearance and later autoradiographic ¹⁴C-iodoantipyrine techniques) in baboon and cat models of focal ischemia, Astrup, Symon, and colleagues systematically reduced CBF and recorded extracellular K⁺ and evoked electrical activity. They found that neuronal electrical activity (measured as cortical evoked potentials and spontaneous EEG) failed when flow dropped below approximately 18–20 mL/100 g/min, while extracellular K⁺ remained normal and cells survived — this was reversible if flow was restored. Only when flow fell further, below roughly 8–10 mL/100 g/min, did extracellular K⁺ rise abruptly (a marker of membrane pump failure and impending cell death) and injury became irreversible. Astrup coined the term "penumbra" (1981, with Siesjö and Symon) for the intermediate zone — functionally silent but structurally intact — precisely because it sat, like the astronomical penumbra of a shadow, in a zone of partial darkness between full illumination and complete shadow. This work established, for the first time with quantitative flow thresholds, that ischemic tissue death is not all-or-none but graded and time-dependent.

Autoradiographic and PET mapping confirmed the penumbra in vivo. Later studies using positron emission tomography (PET) with oxygen-15 labeled water and oxygen (measuring CBF, cerebral metabolic rate of oxygen, and oxygen extraction fraction) in human stroke patients — notably work from groups including Baron and colleagues in the 1980s–1990s — showed that tissue within the penumbra shows elevated oxygen extraction fraction (the brain tries to compensate for low flow by extracting more oxygen from the blood it does receive), a signature distinguishable from the core, where oxygen extraction eventually collapses because there is no viable tissue left to consume oxygen. This gave clinicians a way to image the penumbra directly in patients rather than infer it only from animal thresholds, and it directly validated that human ischemic tissue passes through the same graded flow-dependent zones seen in the animal experiments.

NINDS rt-PA Stroke Study (1995): reperfusion within the therapeutic window changes outcome. The National Institute of Neurological Disorders and Stroke trial randomized patients to intravenous tissue plasminogen activator (tPA) or placebo within 3 hours of symptom onset. Patients treated with tPA were at least 30% more likely to have minimal or no disability at 3 months compared with placebo, despite a higher rate of symptomatic intracerebral hemorrhage (6.4% vs. 0.6%). This trial provided the direct clinical proof that restoring blood flow early — before the cascade completes its work in the penumbra — measurably improves outcome, translating the CBF-threshold biology into an actionable time-limited therapy. Subsequent large trials (ECASS III, and the 2015 wave of endovascular thrombectomy trials — MR CLEAN, ESCAPE, SWIFT PRIME, EXTEND-IA, REVASCAT) extended and refined this evidence, showing mechanical clot retrieval markedly improves outcomes up to 6, and in imaging-selected patients up to 24, hours after onset — because perfusion imaging can identify which patients still have salvageable penumbra even late.

Example 1 — Estimating tissue fate from a perfusion scan. A patient arrives 4 hours after sudden left-sided weakness. CT perfusion imaging shows a small core region with CBF measured at 6 mL/100 g/min and a surrounding region with CBF at 16 mL/100 g/min; the rest of the affected arterial territory shows CBF of 45 mL/100 g/min.

• Using the thresholds established by Astrup and colleagues (structural failure below ~8–10 mL/100 g/min; functional failure below ~18–20 mL/100 g/min):
• The 6 mL/100 g/min region is below the structural threshold — this is core, already infarcted or infarcting irreversibly, and thrombectomy will not recover this tissue.
• The 16 mL/100 g/min region is below the functional threshold but above the structural one — this is penumbra, clinically silent (contributing to the patient's deficit) but still salvageable if flow is restored soon.
• The 45 mL/100 g/min region is above both thresholds — normally perfused, unaffected tissue.
• Clinical logic: a large penumbra-to-core mismatch (a "target mismatch" profile) is exactly the imaging signature used in trials like DEFUSE 3 and DAWN to select patients for late-window thrombectomy, because it indicates substantial salvageable tissue even beyond the traditional window. If this same patient's perfusion scan instead showed a large core and only a thin rim of penumbra, the expected benefit of reperfusion — and the risk/benefit calculation given hemorrhage risk — would be much less favorable.

Example 2 — Quantitative reasoning about the "time is brain" estimate. A commonly cited estimate from stroke epidemiology (Saver, 2006, synthesizing data on infarct growth rates in large-vessel occlusion) is that, on average, an untreated ischemic stroke destroys approximately 1.9 million neurons per minute, based on typical infarct volumes and neuron densities in affected territories, and typical time from occlusion to treatment.

• If a typical large territory infarct (middle cerebral artery distribution) reaches a volume of roughly 100 cm³ over an average 10-hour window without treatment, and cortical gray matter contains on the order of 10¹⁰ neurons per 1,200 cm³ of brain (a standard estimate), then:
• Neurons at risk ≈ 100 cm³ × (10¹⁰ neurons / 1200 cm³) ≈ 8.3 × 10⁸ neurons over 600 minutes.
• That gives ≈ 1.4 × 10⁶ neurons destroyed per minute — in the same order of magnitude as the commonly cited 1.9 million figure, and it illustrates why the figure is large: it is total neurons lost divided by a fairly short time window, not a claim about a constant linear rate (real infarct growth is faster early, when spreading depolarizations are most frequent, and slows as the penumbra is progressively consumed or salvaged).
• Clinical implication: shaving 30 minutes off door-to-needle time is not a marginal administrative win — by this estimate it corresponds to on the order of tens of millions of neurons, a magnitude that helps explain why stroke systems of care are built around minimizing every treatable delay (prehospital recognition, direct-to-CT triage, "drip-and-ship" protocols).

Jerzy (George) Astrup, working with Bo Siesjö and Lindsay Symon, produced the electrophysiological and blood-flow data in the late 1970s that gave the ischemic penumbra its first quantitative, mechanistic definition — turning a vague clinical notion of "tissue at risk" into a specific, measurable pair of blood-flow thresholds. Symon's laboratory in London had already been characterizing regional CBF in primate models of middle cerebral artery occlusion; Astrup and Siesjö brought a biochemical and electrophysiological rigor (extracellular ion measurements, evoked-potential recording) that let the group show, cell by cell and flow-level by flow-level, exactly where function failed and exactly where structure failed. Their 1981 paper synthesizing this work (Astrup, Siesjö, Symon, Stroke) is one of the most cited papers in cerebrovascular neuroscience, precisely because it gave clinicians and later imagers a concrete number to chase.

The core/penumbra concept directly shapes acute stroke medicine: emergency stroke protocols worldwide are organized around minimizing "door-to-needle" and "door-to-groin-puncture" times precisely because the penumbra shrinks with every passing minute. Perfusion-based imaging selection (CT perfusion or diffusion-perfusion MRI mismatch) now extends thrombectomy eligibility to appropriately selected patients up to 24 hours after last-known-well, a direct clinical application of the CBF-threshold physiology described above. Understanding the inflammatory and apoptotic phases of the cascade has driven decades of neuroprotective drug development — NMDA antagonists, free-radical scavengers, anti-inflammatory agents, and hypothermia protocols — most of which succeeded in rodent models but failed in human trials, a translational puzzle that has reshaped how preclinical stroke research is designed (larger animals, comorbid models, blinded multicenter preclinical trials). Blood pressure management, glucose control, and fever control in the acute stroke unit are all guided by the principle that any factor further compromising collateral perfusion or raising metabolic demand can convert penumbra into core. Finally, the eventual fate of surviving penumbral and peri-infarct tissue — its capacity for synaptic reorganization, angiogenesis, and recruitment of endogenous or transplanted progenitor cells — is the direct bridge to Neural Stem Cells and CNS Repair, since the same tissue that the acute cascade threatens to destroy is also the substrate on which post-stroke recovery and rehabilitation depend.

Ask the tutor: How does collateral circulation determine whether a given patient will have a large, slowly-evolving penumbra versus almost none at all?