Info

The Big Heart Disease Lie

Natural Cardiovascular Disease Cure and Treatment

Get Instant Access

patients with larger mismatches tend to demonstrate more lesion growth. , It should be noted that, in two of these studies,37,84 the perfusion parameter used to define the mismatch was not CBF or MTT, but instead the time it took for contrast concentration to reach peak concentration in each image voxel after contrast injection ("time to peak'' or TTP). TTP measurements are often used as rough approximations of MTT measurements because calculation of CBF and MTT are somewhat complex, requiring a mathematical process called "deconvolution." The details of deconvolution are beyond the scope of this chapter, and the reader is referred to other sources for further explanation.86,87 In many clinical settings, maps of parameters like TTP that do not require deconvolution may be available much more quickly than those that do require deconvolution. TTP is less specific than MTT in detecting underperfused tissue88 because it does not distinguish between delayed contrast arrival time (such as that related to perfusion via collateral vessels) and truly prolonged intravascular transit time.

The second hypothesis, that patients should be selected for thrombolysis depending on whether or not they exhibit a diffusion-perfusion mismatch, may have enormous implications for stroke therapy in the near future, and is one of the most actively investigated and debated subjects in neuroimaging.

A group of studies investigating intravenous thrombolysis in acute stroke, considered together, provide indirect support for this hypothesis. In the National Institute of Neurological Disorders and Stroke (NINDS) rt-PA study,16 patients who presented within 3 hours of stroke onset were treated with either intravenous rt-PA or placebo, based on clinical and NCCT criteria only, irrespective of whether or not they had diffusion-perfusion mismatch. In this study, patients who received the drug had significantly better outcomes after 3 months. However, the European Cooperative Acute Stroke study (ECASS), ECASS-II, and Alteplase Thrombolysis for Acute Noninter-ventional Therapy in Ischemic Stroke (ATLANTIS) studies, which used treatment windows of 0-6, 0-6, and 3-5 hours, respectively, found that thrombolysis resulted in worse outcomes than placebo.89-91 These four studies, none of which used the presence of a diffusion-perfusion mismatch as an eligibility criterion, provided support for the Food and Drug Administration's (FDA) approval for intravenous thrombolysis for acute stroke patients, but only when those patients were known to be without symptoms no more than 3 hours before the time of initiation of treatment.

However, several important studies have shown that intravenous thrombolysis may be beneficial more than 3 hours after stroke onset, provided that only patients with a significant diffusion-perfusion mismatch are treated. In one such study, Ribo et al.92 found that patients with a significant diffusion-perfusion mismatch could be treated safely and effectively in the 3-6-hour time period. In phase II of the desmo-teplase in acute stroke (DIAS) trial, patients with diffusion-perfusion mismatch were treated with desmoteplase up to 9 hours after stroke onset, and showed better outcomes than patients given placebo, with only a minimal incidence of symptomatic hemorrhage.22 Similar success was achieved in the same time window by the dose escalation study of desmoteplase in acute ischemic stroke (DEDAS).93

Another recent study94 compared the outcomes of two groups of acute stroke patients who received intravenous or intra-arterial thrombolysis. In one group of patients, thrombolysis was initiated less than 6 hours after a known time of stroke onset. In the other group, the actual time of onset was not known, but thrombolysis was initiated within 6 hours of the time at which the patient became aware of his or her stroke. This was generally far more than 6 hours after the time at which the patient was last seen without symptoms. Patients in this second group were allowed to receive thrombolytic therapy only if an initial MRI examination showed a significant diffusion-perfusion mismatch. Their outcomes were actually slightly better than those in the group who were treated within 6 hours of onset, although the difference did not reach statistical significance.

These studies raise the possibility that, one day, imaging-based treatment protocols may allow for intravenous thrombolysis in patients well outside of the now-accepted 3-hour window, provided they demonstrate substantial diffusionperfusion mismatch. Such protocols could allow for treatment of a vastly larger number of patients than are currently treated. It has been estimated that only 1-7% of acute stroke patients currently receive thrombolytic medication,95-98 and that, in up to 95% of cases, they are ineligible because they present outside of the 3-hour time window.99 As many as 80% of patients who present 6 hours after stroke onset may demonstrate a significant diffusion-perfusion mismatch.100

The echoplanar imaging thrombolysis evaluation trial (EPITHET) is the first large study designed specifically to assess whether the existence of a diffusion-perfusion mismatch should be an eligibility criterion for thrombolysis. Preliminary results published by the EPITHET investigators88 failed to show a significant correlation between the volume of diffusion-perfusion mismatch and the extent of infarct expansion. The study is ongoing at the time of this writing.

Perfusion Imaging: Comparison of CTP and MRP

CTP is a relatively recent development in acute stroke imaging that is already in routine clinical use in many centers. CTP and MRP are similar in that both techniques are based on rapid serial image acquisition during intravenous injection of a bolus of contrast material. In both techniques, measurements of density over time (for CTP) or signal intensity over time (for MRP) are converted to contrast agent-versus-time curves, and these are processed in similar ways to yield the same perfusion measurements (most often CBV, CBF, and MTT). Example CTP images are shown in Figure 2.12.

Despite these similarities, CTP and MRP have some significant differences. Chief among CTP's advantages is its widespread availability and accessibility. As discussed above, CT scanners are far more widely available than MRI scanners in or near North American emergency departments, particularly after hours, when many MRI scanners are not operational. Furthermore, although some investigators have proposed that acute stroke patients may be safely directed to an MRI scanner without an initial CT scan,2, 101 the clinical reality in most centers is that patients with suspected acute stroke undergo CT examination as a first study. Therefore, CTP offers the possibility of performing an examination that includes perfusion

FIGURE 2.12 CT perfusion images. CTP images were acquired in this acute stroke patient who was unable to undergo MRI. A map of CBV (a) shows a well-defined region of decreased blood volume in the left frontal lobe. Because DWI images are not available, this region is presumed to represent the core of the infarct. MTT maps (b) show a much larger region of prolonged MTT, reflecting tissue at risk of infarction. In a follow-up CT scan (c), most but not all of the threatened tissue has progressed to infarction. Note that in some of the tissue that demonstrates low CBV, perfusion is so severely impaired that the amount of contrast agent that arrives is so small that MTT cannot be measured accurately, resulting in a noisy ''speckled'' appearance in the MTT map.

FIGURE 2.12 CT perfusion images. CTP images were acquired in this acute stroke patient who was unable to undergo MRI. A map of CBV (a) shows a well-defined region of decreased blood volume in the left frontal lobe. Because DWI images are not available, this region is presumed to represent the core of the infarct. MTT maps (b) show a much larger region of prolonged MTT, reflecting tissue at risk of infarction. In a follow-up CT scan (c), most but not all of the threatened tissue has progressed to infarction. Note that in some of the tissue that demonstrates low CBV, perfusion is so severely impaired that the amount of contrast agent that arrives is so small that MTT cannot be measured accurately, resulting in a noisy ''speckled'' appearance in the MTT map.

imaging without having to move to a second modality. As discussed above, CT allows for scanning of patients with pacemakers and other ferromagnetic implants, as well as monitoring of patients with ferromagnetic equipment that cannot be brought into an MRI scanner room.

Aside from CTP's use of potentially nephrotoxic contrast material and relatively large doses of ionizing radiation,102 one of the main disadvantages of the technique is its limited coverage of the brain. The degree of coverage is highly dependent on the scanner being used, with multislice scanners affording much greater coverage. For example, our institution's current protocol for 16-slice CT scanners allows for imaging of two separate 2-cm axial slabs, resulting in coverage of 4 cm of the brain. However, our protocol for 64-slice scanners allows for imaging of two separate 4-cm slabs, which together cover most of the brain. By comparison, our current MRP protocol allows for acquisition of 16 slices of any desired thickness and orientation. As we usually choose to acquire axial MRP slices that are 5 mm thick and separated by 1 mm, this results in coverage of a 9.5-cm axial slab. Thus, with a 64-slice scanner, brain coverage with CTP approximates that of MRP, although evaluation of the posterior fossa may be somewhat compromised with CTP due to beam hardening artifacts at the skull base.

For both CTP and MRP, perfusion measurements are based on detection of a nondiffusible contrast agent that is confined to the 2-5% of each image voxel that is occupied by blood vessels.103-106 Because CTP directly measures the quantity of contrast material in each image voxel, confinement of the contrast agent within vessels places an intrinsic limit on the degree of density change that can be measured by CTP, which is usually on the order of 10% or less. This fact, in conjunction with the intrinsically lower contrast-to-noise ratio of CT imaging, means that CTP maps are much noisier than MRP maps. Typically, CTP postprocessing algorithms perform extensive spatial averaging, in order to reduce noise by sacrificing some of CT's considerably superior spatial resolution.

MRP maps are less noisy than CTP maps because MRP detects the passage of gadolinium using susceptibility effects, which "bloom" out of each vessel, extending through a space whose radius is roughly proportional to the radius of the vessel. Thus, the susceptibility effect related to gadolinium in microscopic vessels blooms out of those vessels, reducing signal arising from all parts of each voxel, and resulting in a much larger measurable signal change as the gadolinium passes through brain, in the range of 20-40% in the gradient-echo images that are most often used for MRP. This blooming effect accounts for the superior contrast-to-noise ratio of MRP maps, and also for the fact that, unlike CTP maps, most MRP maps disproportionately weigh the presence of contrast in larger vessels.107

MRP can also be performed with spin-echo pulse sequences, which results in sensitivity to contrast in vessels of all sizes that more closely (but not perfectly)

Was this article helpful?

0 0
Your Heart and Nutrition

Your Heart and Nutrition

Prevention is better than a cure. Learn how to cherish your heart by taking the necessary means to keep it pumping healthily and steadily through your life.

Get My Free Ebook


Post a comment