How Perfusion Imaging Extended the Stroke Treatment Window

The short answer

For decades, the clock governed stroke care: miss the treatment window and the door closed. Perfusion imaging changed that by letting clinicians see which brain tissue was already dead and which was still salvageable. Two 2018 trials, DAWN and DEFUSE 3, used that distinction to select patients for clot removal 6 to 24 hours after they were last known well, and both were stopped early because the benefit was so large. The lesson was that a picture of the tissue, not just the time on the clock, could identify who still stood to gain.

This article is educational and not medical advice.

Why the clock was the old rule

Ischemic stroke happens when a clot blocks an artery feeding the brain. Downstream, neurons begin to die. But they do not all die at once. At the center of the blockage sits the infarct core, tissue that is already lost. Surrounding it is the penumbra, tissue that is starved but still alive, kept marginally viable by collateral blood vessels. The penumbra is the target of treatment. Restore flow in time and it can recover; wait too long and it collapses into core.

Early trials measured that race with a stopwatch. Clot-dissolving drugs and, later, mechanical thrombectomy (physically pulling the clot out with a catheter) were tested and approved within tight time limits, generally within 6 hours for large vessel occlusion. The reasoning was sound on average: the longer the delay, the more penumbra becomes core. The problem is that averages hide enormous variation. Some people have robust collateral vessels that keep the penumbra alive for many hours. Others lose their penumbra quickly. A fixed clock treats both the same and, in doing so, denies treatment to slow progressors who could still benefit.

What perfusion imaging actually measures

Perfusion imaging estimates blood flow through brain tissue, either with CT perfusion (CTP) or with diffusion and perfusion MRI. Automated software, most prominently the RAPID program used in these trials, processes the scan and produces volume estimates: how much tissue is already infarcted (the core) and how much is underperfused but not yet dead (the penumbra at risk). The gap between the two is the mismatch. A large mismatch means a lot of brain is still salvageable relative to what has been lost.

This reframes the decision. Instead of asking how many hours have passed, the clinician asks how much tissue can still be saved. Two people at the same 12-hour mark can have completely different scans, and the scan, not the clock, predicts who will benefit.

DAWN and DEFUSE 3

The DAWN trial, published in the New England Journal of Medicine in 2018 by Nogueira and colleagues, enrolled patients with occlusion of the intracranial internal carotid artery or proximal middle cerebral artery who were last known to be well 6 to 24 hours earlier. It used a clinical-imaging mismatch: a severe neurological deficit (measured by the NIHSS stroke scale) paired with a small infarct core, with the exact thresholds adjusted for age. The rate of functional independence at 90 days was 49 percent with thrombectomy versus 13 percent with standard care. The trial was halted early for benefit.

DEFUSE 3, reported the same year by Albers and colleagues, took a slightly different route. It enrolled patients 6 to 16 hours out and defined mismatch by imaging alone: an ischemic core under 70 mL, a mismatch ratio greater than 1.8, and an absolute mismatch volume of at least 15 mL. Functional independence at 90 days was 45 percent with thrombectomy versus 17 percent with medical management. It too stopped early.

Both trials point in the same direction. When imaging shows a small core and a large threatened penumbra, opening the artery late still helps, and helps substantially. These are large treatment effects; roughly one additional patient reaches functional independence for every few treated.

What the guidelines concluded

The 2019 update to the American Heart Association and American Stroke Association guideline for early management of acute ischemic stroke incorporated both trials. For selected patients 6 to 16 hours from last known well with anterior circulation large vessel occlusion who meet DAWN or DEFUSE 3 criteria, thrombectomy carries a Class I recommendation, the strongest tier. For selected patients in the 16 to 24 hour window meeting DAWN criteria, it is a Class IIa recommendation, meaning reasonable. The guideline also states that obtaining CTP or diffusion MRI, with or without perfusion MRI, is recommended to aid selection, but only when patients otherwise meet the trial criteria.

That last clause deserves emphasis, and it is where careful appraisal matters.

Reading the evidence honestly

The strength of these trials is also their constraint. The benefit was demonstrated in patients who matched narrow entry criteria: specific vessels, a favorable mismatch profile, defined core-size limits, and severe enough deficits. A patient with a large established core, poor collaterals, or an occlusion in a different location was not represented, and the results should not be stretched to cover them. The trials proved that imaging can find late responders; they did not prove that imaging makes everyone a candidate.

Generalizability carries real caveats. RAPID and similar software produce estimates, not ground truth, and core-volume calculations can vary with scanner, timing, and processing thresholds. The trials enrolled a selected minority of all late-presenting strokes, so the impressive success rates apply to that enriched group, not to the average person arriving 12 hours after symptom onset. There is also active investigation into whether simpler imaging, or looser criteria, can identify additional patients who benefit without the full perfusion workup. The honest summary is that perfusion-based selection is well validated for the populations studied and remains an open research question at the edges.

Still, the conceptual shift is durable. DAWN and DEFUSE 3 replaced a single question, how long has it been, with a better one, how much brain is still alive. That is why the treatment window is now written in tissue as much as in time.