Perfusion imaging in acute ischemic stroke: utile or futile?

Science Editorials and Reviews - Edited by S. Andrew Josephson, MD

February 21, 2008


By Maarten G. Lansberg, MD, PhD

This editorial was developed for, which is publishing expert opinions on a variety of hot topics in neurology.

Maarten G. Lansberg, MD, PhD
Department of Neurology and Neurological Sciences
Stanford University, Stanford Stroke Center
Author Disclosure Statement

The Premise of Perfusion

The premise is simple: tissue plasminogen activator (tPA) and other therapies aimed at restoring blood flow only benefit patients who have hypoperfused yet salvageable brain tissue called an ischemic penumbra. It is not straightforward to determine if an acute stroke patient has an ischemic penumbra, but several perfusion imaging techniques have been evaluated in terms of their ability to visualize this area. Despite significant advances in this area of research, it remains to be determined which perfusion imaging modality is optimal and whether perfusion imaging can lead to better clinical outcomes of acute stroke patients. The ischemic penumbra is brain tissue that is hypoperfused but that will not go on to infarction if blood flow is restored.

The ischemic core consists of brain tissue that has already undergone irreversible injury from hypoperfusion. Restoration of blood flow to the ischemic core will not reduce the infarct volume.

The transition from reversible to irreversible ischemic damage is a function of cerebral blood flow (CBF) and duration of ischemia. A large number of animal and human studies have tried to identify CBF thresholds that differentiate the ischemic core from the penumbra and the penumbra from normal brain tissue.

  • Classic studies in humans have shown that normal CBF is around 50 ml/100g/min, reversible ischemia occurs when CBF drops below 20 ml/100g/min, and cerebral ischemia becomes irreversible when CBF drops below 8 ml/100g/min.
  • These thresholds are all approximate; they vary greatly by brain region (with gray matter requiring much higher CBF than white matter), duration of ischemia, and temperature. This complicates the assessment of the ischemic core and penumbra in acute stroke patients.

Further limiting the assessment of core and penumbra in the acute stroke setting is the lack of technology to rapidly produce high resolution CBF maps with accurate quantitative CBF values. As a result, there are currently no consensus criteria for imaging assessment of the ischemic penumbra.

Perfusion Imaging Modalities

Several perfusion imaging modalities exist. To better understand the various technologies it is helpful to divide them into the following two main categories:

Diffusible-tracer perfusion imaging: Diffusible tracers pass the blood-brain-barrier and are taken up by the brain parenchyma. The concentration of the tracer in the brain parenchyma is therefore correlated with the amount of blood flow to the brain parenchyma. With knowledge of the tracer concentration in the arterial blood supply, absolute CBF values can be calculated. Perfusion imaging techniques that use diffusible tracers include Xenon-CT, Single Photon Emission CT (SPECT) and positron emission tomography (PET). PET is generally viewed as the gold-standard perfusion imaging technique, but it is used little in clinical practice due to its high cost, limited availability, and the poor resolution of PET perfusion images.

Bolus-tracking perfusion imaging: This technique uses a non-diffusible tracer and is the most widely used perfusion imaging technique. The two imaging modalities that utilize bolus-tracking are CT-perfusion imaging (CTP) which uses iodinated CT contrast as the tracer and perfusion-weighted MRI (PWI) which uses gadolinium contrast. These contrast agents change the signal intensity of the brain parenchyma when they reach the capillary bed. CT contrast enhances the CT signal, whereas gadolinium decreases the MRI signal. The effect of the contrast on the signal intensity is assessed over time by obtaining serial scans after administration of the contrast bolus. Typically 30-40 sequential scans are obtained over one to two minutes. When blood flow to a certain brain region is reduced due to a vessel occlusion or stenosis it will affect several features of the contrast going through the hypoperfused capillary bed:

  • Less contrast will reach these capillaries
  • The contrast will arrive later
  • It will take longer for the contrast to pass through the capillaries compared to regions with normal blood flow.

The total amount of contrast that reaches the capillary bed is visualized on the cerebral blood volume (CBV) map;

  • The delay in contrast arrival time on the time to peak (TTP) or time-to-maximal enhancement (Tmax) map, and
  • The time for contrast to pass through the hypoperfused vessels on the mean-transit time (MTT) map. Finally, a map of cerebral blood flow (CBF) can be mathematically generated based on the CBV and MTT maps. Which of these maps is best suited to assess the ischemic penumbra has not yet been determined.

Perfusion Imaging of Infarct and Penumbra

MRI is a commonly used method for assessment of the ischemic core and penumbra. The diffusion-weighted MRI (DWI) lesion is generally assumed to reflect the ischemic infarct, whereas the PWI lesion includes both infarct and penumbra; hence the potential for perfusion/diffusion mismatch. The high conspicuity of acute ischemic lesions on DWI is a clear advantage of using DWI for assessment of the ischemic infarct. A limitation of DWI is that it may occasionally include penumbral tissue. Several studies have now demonstrated that DWI lesions can partially reverse with early restoration of blood flow. The main challenge with using PWI for assessment of the ischemic penumbra is choosing the appropriate PWI map and the optimal PWI threshold. Commonly used thresholds include tissue that has a prolonged MTT of at least four seconds or a prolonged Tmax of at least two seconds. Both methods appear to be too sensitive as the PWI lesions assessed with these methods tend to include tissue that is hypoperfused but not at risk for infarction (benign oligemia).

Identification of infarct and penumbra on CTP is typically based on a combination of CBF and CBV criteria. The CTP signature of the ischemic infarct is a combination of low CBF (or prolonged MTT) and low CBV, whereas the penumbra is characterized by low CBF (or prolonged MTT) and high to normal CBV values. As with MR imaging, optimal thresholds for assessment of infarct and penumbra have not yet been established. The main advantages of CTP over MRI for assessment of the ischemic infarct and penumbra are its wider availability, lower cost, and shorter imaging time. The need to evaluate two maps and the poor lesion conspicuity are disadvantages of CT compared to DWI for assessment of the infarct. Limited coverage of the brain parenchyma is another drawback of CTP (4 slices on CT vs 12 slices on MRI).

Figure Legend

The figure illustrates the perfusion/diffusion mismatch. In this patient the lesion on DWI (representing the ischemic core) is much smaller than the lesion on PWI. This patient thus has a large ischemic penumbra and is expected to benefit from prompt restoration of blood flow to the penumbral area. Magnetic resonance angiography is notable for an abrupt cut-off of the left middle cerebral artery (arrow), indicating that occlusion of the left MCA caused the stroke. The PWI image in this figure is a Tmax perfusion map.

Infarct Penumbra
CT Very low CBF (or prolonged MTT) Low CBF (or prolonged MTT)
Low CBV High to normal CBV
MRI DWI high signal intensity DWI normal signal intensity
MTT or Tmax prolonged

Acute Stroke Trials That Have Used Perfusion Imaging

Promising preliminary data regarding the use of PWI have been reported in three prospective clinical studies. The DEFUSE study was a prospective open-label pilot study of intravenous tPA administered in the three to six hour time-window. It supports the PWI/DWI mismatch hypothesis by demonstrating that patients with PWI/DWI mismatch are likely to benefit from restoration of blood flow whereas patients without mismatch are not. DIAS and DEDAS were phase II double-blind placebo-controlled safety trials of desmoteplase (a thrombolytic agent) administered in the three to nine hour time-window. Only patients with a PWI/DWI mismatch were eligible for enrollment in these trials. In both studies reperfusion was associated with favorable clinical outcome.

Despite these promising pilot studies, perfusion imaging has not led to major breakthroughs in clinical research protocols. DIAS-2, a phase III double-blind randomized-controlled study, which selected patients based on either a PWI/DWI mismatch or a CTP mismatch, did not demonstrate improved clinical outcomes in patients treated with desmoteplase. Future trials are thus needed to better define the role of perfusion imaging in the selection of acute stroke patients for therapies aimed at restoring blood flow. Results of EPITHET, an Australian double-blind placebo-controlled trial of intravenous tPA administered in the three to six hour time-window, will help achieve this goal. MRIs were obtained prior to and three to five days following tPA administration in EPITHET. Results of this study will be presented at the February 2008 AHA Stroke meeting in New Orleans. Results of MR-RESCUE—an ongoing placebo controlled-trial of mechanical thrombectomy in patients selected based on PWI/DWI mismatch—will provide further evidence of the utility or futility of perfusion imaging in acute stroke.


Perfusion imaging has great promise in selecting patients for acute stroke treatment. However, before it is ready for prime time, the benefit of perfusion imaging needs to be demonstrated in an acute stroke trial.

Detailed reviews of acute stroke perfusion imaging:
Kohrmann M, Juttler E, Huttner HB, Nowe T, Schellinger PD. Acute stroke imaging for thrombolytic therapy--an update. Cerebrovascular diseases. 2007;24:161-169 Muir KW, Buchan A, von Kummer R, Rother J, Baron J-C. Imaging of acute stroke. The Lancet Neurology. 2006;5:755-768

Author Disclosure

Disclosure: Dr. Lansberg has received personal compensation to attend an advisory meeting for Nuvelo, maker of alfimeprase. He has served as an expert witness for medical legal reviews regarding stroke cases. He has received research support from the National Institute of Neurological Disorders and Stroke (K23 career development award) and an internal Stanford University grant (Bio-X).