| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Review |
1 Department of Radiology, Duke University Medical Center, Box 3808, Durham, NC
27710.
2 Departments of Radiology, Biomedical Engineering and Medicine, Emory
University School of Medicine, Atlanta, GA.
3 Department of Radiology, University of California, San Francisco, San
Francisco, CA.
Received December 22, 2007;
accepted after revision May 1, 2008.
Address correspondence to J. M. Provenzale.
Abstract
 Top Abstract IntroductionSummary of Important Therapeutic...Controversies in Stroke Imaging...unresolved Issues Requiring...SummaryReferences |
|---|
CONCLUSION. Advanced MRI and CT play a growing role in selection of patients for therapy. Numerous clinical trials have shed light on the efficacy of various reperfusion therapies; disparities in trial design have also left unanswered questions.
Keywords: brain • CT • ischemia • MR • perfusion • stroke
|
TopAbstractIntroductionSummary of Important Therapeutic...Controversies in Stroke Imaging...unresolved Issues Requiring...SummaryReferences |
|---|
The intent of this review is to examine various clinical trials of reperfusion agents and address the key imaging issues in future studies. At the Advanced Neuroimaging for Acute Stroke Treatment meeting held in Washington, DC, on September 7 and 8, 2007, leading members of the stroke imaging community from academic medical centers, the National Institute of Neurological Disorders and Stroke, the National Institute of Biomedical Imaging and Bioengineering, industry representatives, and members of the U.S. Food and Drug Administration (FDA) met to discuss the role of advanced neuroimaging in the management of acute stroke patients. This review provides the background for the recommendations discussed in that meeting.
Summary of Important Therapeutic Trials
|
TopAbstractIntroductionSummary of Important Therapeutic...Controversies in Stroke Imaging...unresolved Issues Requiring...SummaryReferences |
|---|
DIAS Trial
The Desmoteplase in Acute Ischemic Stroke (DIAS) trial was a multicenter,
randomized, double-blind, placebo-controlled phase 2 trial that concentrated
on patients in the 3–9 hour time window after stroke onset
[1]. Entry criteria included a
National Institutes of Health Stroke Scale (NIHSS) score between 4 and 20 and
MR evidence of a substantial (i.e., at least 20%) mismatch between PWI and
DWI. The PWI measurement used was that of the mean transit time (MTT) which
was performed at baseline and again in a 4–8 hour period after
treatment. The degree of arterial stenosis or occlusion was assessed on the
basis of an adaptation of the Thrombolysis in Myocardial Infarction (TIMI)
grading scheme (using MR angiography), which was as follows: 0, complete
occlusion; 1, severe stenosis; 2, mild to moderate stenosis; and 3, normal
arterial caliber. Patients were randomized to one of three doses of
desmoteplase or placebo. The original trial design planned to investigate
three fixed doses of 25 mg, 37.5 mg, and 50 mg desmoteplase versus placebo in
four parallel groups of 30 patients each. However, after development of
intracranial hemorrhage in four patients, the trial design was restructured
with a 25-mg dose and placebo. After an excess intracranial hemorrhage rate
was again seen, the trial was again restructured with a placebo-controlled
bodyweight-adjusted dose-escalation design starting at a dose of 62.5
µg/kg, followed by 90 µg/kg and 125 µg/kg. Reperfusion was defined as
either 
30% reduction of volume of MTT abnormality or 2 points
improvement on the adapted TIMI grading scheme. As will be discussed in more
detail later, for many reasons, recanalization is not equivalent to
reperfusion. Thus, using degree of arterial patency—that is,
recanalization—as an outcome measure reflecting reperfusion can be
considered a shortcoming of this study.
The first 47 patients were randomized to one of three fixed doses of desmoteplase or placebo, but after an unacceptable rate of intracranial hemorrhage was seen, doses were weight-adjusted in the remaining patients to one of the following doses: 62.5 µg/kg, 90 µg/kg, or 125 µg/kg. The study showed statistically significantly higher rates of reperfusion and favorable 90-day clinical outcomes in patients receiving the highest dose. Furthermore, early reperfusion correlated favorably with clinical outcome. Although these findings appear well defined, it is worth noting that the DIAS study underwent many protocol changes during the trial, which makes interpretation of trial results more difficult than for trials that do not suffer such shortcomings.
In the DIAS trial, patients in the low-dose and placebo groups had a 10% favorable outcome rate. It is interesting to note that even patients who received the middle or high desmoteplase dose but who were not seen to reperfuse on the second PWI study had a 50% favorable outcome rate. Such results would not be expected given the absence of reperfusion. One potential explanation for the high rate of favorable outcome despite the absence of reperfusion is that the second PWI scan overestimated the actual degree of PWI abnormality (and underestimated the rate of reperfusion), possibly because of the (too early) timing of the scan. Of note, no single generally accepted time interval between initial and subsequent posttherapy scanning has been advanced. In fact, time intervals between scanning have differed between major clinical trials (please see Number and Timing of Imaging Studies in Stroke Treatment Trials later in this article) perhaps underscoring the need for more emphasis on determination of optimal imaging intervals and uniformity across studies. Alternatively, perhaps the threshold of a 30% decrease in size of perfusion abnormality was inappropriately stringent and masked degrees of reperfusion that were clinically relevant.
DEDAS Trial
The Dose Escalation of Desmoteplase for Acute Ischemic Stroke (DEDAS) trial
was a randomized, dose-escalation study that had three treatment arms:
low-dose (90 µg/kg) desmoteplase, high-dose (125 µg/kg) desmoteplase,
and placebo [2]. A perfusion
deficit of > 2 cm in diameter and involving the cerebral cortex in addition
to a mismatch between PWI and DWI of at least 20% were the entry criteria.
Primary efficacy end points were a reduced mismatch at 4–8 hours as well
as clinical outcome at 90 days. A higher rate of reperfusion (using the same
definition as in the DIAS trial) was seen in high-dose-treated patients
compared with the other arms, and reperfusion was strongly correlated with
good clinical outcome. The results of the DEDAS trial generally supported the
results of the DIAS trial by providing further evidence that IV thrombolysis
with desmoteplase 3–9 hours after stroke onset is safe in patients
selected by a mismatch between PWI and DWI. Nonetheless, reperfusion did not
reach statistical significance compared with placebo, although there was a
strong trend with high-dose therapy.
In the DEDAS study, the investigators recognized that some patients were incorrectly thought to have mismatches between PWI and DWI and incorrectly enrolled in the intent-to-treat component of the study on the basis of the perceived mismatch. Thus, even in a well-designed study, inappropriate enrollment of a substantial proportion of the study population was seen on the basis of incorrect analysis of perfusion data, emphasizing that, in the real world of clinical applications of a procedure, mistakes will be made. To the extent that automated perfusion analysis tools can be designed and made widely available, one might expect that the number of such errors would be diminished.
DIAS-2 Trial
DIAS-2 was a prospective, multicenter, ran domized, double-blind,
placebo-con trolled study investigating the efficacy and safety of two doses
of desmoteplase, 90 µg/kg and 125 µg/kg, given as a single IV bolus
[3]. Whereas the original DIAS
trial was a phase II trial intended to provide proof of concept, DIAS 2 was a
phase IIIa trial needed prior to approval by the FDA. Patients were eligible
for DIAS-2 if they could be treated within 3 and 9 hours after the onset of
stroke symptoms, had an NIHSS score between 4 and 20, and had a distinct
ischemic penumbra of at least 20% established by MR (PWI/DWI) or perfusion CT.
The primary outcome was clinical improvement at day 90, defined as having
achieved all of the following: an improvement in NIHSS score of 8 points or
more (or an NIHSS score of < 1), Barthel index score of 75–100, and a
modified Rankin scale score of 0–2.
Using intention-to-treat analysis, investigators found no significant difference between the groups in clinical response rates, with numbers that contrasted sharply with their previous findings with this agent in the DIAS and DEDAS trials. The clinical response rate was 46.0% in the placebo group, 47.4% in the 90-µg/kg group, and 36.4% in the 125-µg/kg group. All-cause mortality at 90 days was 6.3% in the placebo group, 5.3% in the 90-µg/kg group, and 21.2% in the 125-µg/kg group. Of the 14 deaths in the 125-µg/kg group, 10 were considered by the investigators to be unrelated to the treatment, nine were not neurologic in origin, and nine were late—that is, occurring more than 10 days after administration of the study drug. There was no excess in systemic bleeding with treatment; however, and although there were intracranial hemorrhage (ICH) cases in the two treatment groups (3.5% in the 90-µg/kg group and 4.5% in the 125-µg/kg group), the rate was not high enough to explain the difference in mortality.
DEFUSE Trial
The Diffusion and Perfusion Imaging Evaluation for Understanding Stroke
Evaluation (DEFUSE) trial was designed to determine whether a mismatch between
perfusion and diffusion could be used to predict clinical outcome in patients
with early reperfusion after treatment with recombinant tissue plasminogen
activator (rt-PA) during the 3–6 hour time window after stroke onset
[4]. DEFUSE was a multicenter
open-label study of IV rt-PA given 3–6 hours after stroke onset.
Enrollment criteria included a NIHSS > 5, no evidence of hemorrhage on CT,
and ability to undergo an MRI with diffusion and perfusion imaging. All
patients received a standard dose of rt-PA. Repeat MRI was performed 3–6
hours after rt-PA administration. A perfusion abnormality was defined as a
delay of 2 seconds on a PWI Tmax map.
A mismatch between PWI and DWI was defined as a perfusion abnormality
volume of at least 20% and 10 mL larger than the diffusion lesion volume.
Such a mismatch was seen in 54% of 74 patients enrolled. The investigators
defined reperfusion as at least a 30% reduction in the perfusion abnormality
volume between baseline and follow-up MRI (and thus used a different
definition than that used in the DIAS trial). The study found that patients
with a baseline mismatch between PWI and DWI of at least 30% and a reduction
in perfusion abnormality volume of at least 10 mL had a better clinical
outcome. The trial thus showed that baseline MRI findings could be used to
identify groups of patients likely to benefit from thrombolytic therapy and,
potentially, other forms of reperfusion therapy.
Patients with a mismatch between PWI and DWI for whom early reperfusion (n = 18) was seen were more likely to have a favorable clinical response than patients with a mismatch who did not have early reperfusion (n = 16). A total of only 11 patients were seen who had no mismatch between PWI and DWI, including only four who had no mismatch but had early reperfusion, making comparison with the patients with a mismatch and early reperfusion difficult from a statistical standpoint. Interestingly, the seven patients who had a matched lesion but not early reperfusion had a better clinical outcome than all other patients. The authors of that study explain this unexpected, counter-intuitive result as a result of the small sample size, reflecting one limitation of the study. Another limitation of the study is the lack of a placebo population.
EPITHET Trial
The Echoplanar Imaging Thrombolysis Evaluation Trial (EPITHET) trial was a
phase 2 multicenter, randomized, double-blind, controlled trial that
investigated whether patients with a mismatch between PWI and DWI who were
treated with IV rt-PA within 3–6 hours after stroke onset would have
attenuation of infarct growth (defined by the difference in size of DWI
abnormality volume on initial imaging and final infarct volume)
[5,
6]. The same definition of
mismatch as that in the DEFUSE trial was used. Researchers at 15 centers in
Australia, Belgium, New Zealand, and Scotland screened 1,224 patients
[7]. Of 101 randomized
patients, a mismatch between PWI and DWI was present in 86%. Of those, 37
patients received rt-PA and 43 received a placebo saline solution. Average
time-to-treatment was about 5 hours from stroke onset. The primary outcome
measure was the effect of late application of rt-PA (i.e., beyond 3 hours
after stroke onset) on infarct growth. The late effect on infarct growth was
not significant. However, a strong trend toward infarct reduction was noted.
In addition, rt-PA was associated with significant restoration of blood flow
at 3–5 days and improved functional outcomes at 90 days. The
investigators of EPITHET suggested that these results provide support for
further studies on extending the time window for IV rt-PA and proposed that a
sample size of 400 patients (200 in the rt-PA arm and 200 in the placebo arm)
would be adequate to show the efficiency of rt-PA in this setting.
Critics of the EPITHET trial have noted a number of study limitations [8]. First, patients were not chosen for inclusion on the basis of MRI criteria, but instead all ischemic stroke patients within 3–6 hours who did not have contraindications to rt-PA were included. Second, reperfusion was assessed relatively late (at 3–5 days) when a substantial pro portion of patients will have already spon taneously recanalized, thereby limiting assessment of effects of the drug.
In the EPITHET trial, the number of patients with a mismatch between PWI and DWI (86%) was much greater than in DEFUSE (54%) despite the same definition for mismatch. The authors of the EPITHET trial explain this higher proportion of mismatch by the greater baseline neurologic impairment and lower proportion of small lesions in patients in the EPITHET trial. However, another possible explanation is a difference in postprocessing techniques, resulting in different volumes of ischemic penumbra despite similar definitions of the tissue at risk. If true, this explanation represents yet another limitation of imaging trials and emphasizes the importance of standardization of perfusion imaging postprocessing.
MR RESCUE Trial
The MR and Recanalization of Stroke Clots Using Embolectomy (MR RESCUE)
trial is an ongoing, multicenter, randomized, blinded, controlled trial of
patients with large-artery occlusion who can undergo mechanical embolectomy
with a balloon catheter and clot retrieval device within 8 hours of stroke
onset [9]. The study intends to
determine whether DWI can identify patients who might benefit from therapy and
whether embolectomy is clinically more effective than standard medical
management. For this trial, a program has been developed to fully automate the
infarct prediction process. This program incorporates DWI and PWI images via
DICOM transfer and performs apparent diffusion coefficient (ADC) computation,
brain segmentation, determination of arterial input function, computation of
perfusion measures, co registration of DWI and PWI images, and prediction of
infarct size (Schaewe et al., presented at the 2006 annual meeting of the
International Society of Magnetic Resonance in Medicine).
Controversies in Stroke Imaging and Management
|
TopAbstractIntroductionSummary of Important Therapeutic...Controversies in Stroke Imaging...unresolved Issues Requiring...SummaryReferences |
|---|
In the past decade, various operational definitions of the ischemic penumbra have been proposed. On MRI, the ischemic penumbra has generally been defined as the area of mismatch between PWI and DWI on initial MRI (Fig. 1A, 1B). In MRI terms, the penumbra is the region that has solely altered hemodynamics on PWI and no diffusion abnormality on DWI. On CT perfusion im aging, many investigators consider the region of markedly reduced cerebral blood volume (CBV) to represent the region of infarcted tissue. In addition, the same investigators consider the region having prolonged MTT and reduced cerebral blood flow but preserved CBV to represent the ischemic penumbra, or tissue at risk [10, 11]. Thus, on CT perfusion imaging, the ischemic penumbra has generally been considered to be the difference between the size of the MTT abnormality and the CBV abnormality.
|
|
The effect of these various operational definitions of the ischemic
penumbra on introducing further heterogeneity into perfusion imaging clinical
trials is readily seen in the following examples. First, the definitions of
perfusion abnormality differed between studies. Thus, DIAS and DEDAS used
visual assessment of normalized first-moment MTT maps to define the perfusion
abnormality, whereas DEFUSE and EPITHET used a quantitative threshold of
2 seconds on Tmax maps. These differences impair translation of results from
one study to another, leading to difficulties with comparison of
therapies.
Just as the definition of perfusion deficit has differed among studies, so
has that of reperfusion. In the DIAS and DEDAS trials, reperfusion was defined
as either 30% reduction of volume of MTT of abnormality or 2 points
improvement on the adapted TIMI scale on MR angiography. However, in the
DEFUSE trial, reperfusion was defined as a 30% and 10 mL or more
reduction in PWI lesion volume, without reference to appearance of the
arterial lumen on MR angiography. When one recognizes that studies may have
also differed with regard to postprocessing techniques used to analyze
perfusion abnormalities, some of which may not be widely commercially
available, the issues of generalizing from one study to another become even
more apparent.
Although DWI abnormality is usually accepted as a hallmark of irreversible infarct core, and the mismatch between DWI and PWI as a marker of the penumbra, or tissue at risk, cases of DWI reversibility have been reported [15]. Because measurements of mismatch between DWI and PWI can be time-dependent, the use of a simple mismatch is unsatisfactory in such cases. For this reason, more complex, multiparametric, models have been proposed as an alternative to the simple DWI and PWI concept. These models combine DWI and PWI values with values from other MR sequences to extract information about tissue viability [16]. The MR RESCUE trial uses one of these multivariate models to assess brain perfusion in candidates for the trial [9].
Key Factors That May Influence Interpretation of Results of Trials
One confounding factor in assessment of reperfusion therapy has been lack
of uniformity in various studies to determine whether the artery supplying an
ischemic region is patent. The degree of patency of the artery supplying the
ischemic region is a very important vari able in assessment of response to
rt-PA therapy because it may influence management choice as well as degree of
benefit
[17–19].
For instance, in a study in which one patient population receives rt-PA and
another receives a placebo, if the group receiving the placebo has a much
higher proportion of patients in whom the artery subserving the infarct is
patent at the time at which the placebo is administered, equivalent outcome in
the two groups could erroneously be attributed to failure of therapy. It is
worth noting the distinction between the importance of arterial patency before
therapy and the much lesser importance of arterial patency after therapy.
Knowing the degree of patency of the parent artery supplying an ischemic
region is important because it provides an assessment of whether therapy is
capable of reaching the infarct. However, as will be discussed later, knowing
the patency of an artery after therapy and using improvement of the degree of
patency as an outcome measure are less important be cause the parent artery
may recanalize without reduction in ultimate infarct size.
One important confounding factor is early reperfusion (whether spontaneous or after ther apy). The information gained from follow-up imaging studies is important in determining perfusion thresholds that are meaningful for prediction of outcome in treatment protocols. The information gained from patients who reperfuse is important because it allows one to develop perfusion thresholds on baseline imaging studies that help to distinguish infarct core from penumbra, assuming that the final infarct size represents the infarct core at initial imaging.
|
|
|
|
|
|
|
|
Another factor that may influence interpretation of trial results is that even patients who experience ischemia in the same vascular territory and have the same degree of arterial patency can have different outcomes based on the presence or absence of extensive collateral blood flow. Thus, the issue of collateral circulation is an important one that needs to be taken into consideration in clinical trials of stroke therapy to fully evaluate the effects of therapy. The degree of adequacy of collateral circulation is difficult to assess without using catheter angiography [21] (Fig. 3A, 3B, 3C). Although catheter angiography has the advantage of high temporal resolution, its use in the acute stroke setting is increasingly being replaced by noninvasive techniques such as CT angiography and MR angiography. Reliable, although static, ways to assess collateral circulation on CT angiography and MR angiography have been reported [22] (Figs. 4A, 4B and 5A, 5B), but these are not yet in general use.
|
|
|
|
|
|
|
Despite the importance of knowing the degree of collateral circulation for understanding treatment effects, visualization of collateral blood vessels on catheter angiography cannot generally be used to guide therapy because it is not known how the presence of collaterals correlates with reperfusion imaging. However, it is possible that such knowledge will be used in the future to select patients for intervention.
Clinical Scores Used to Assess Therapy
Many therapeutic trials are oriented toward clinical improvement as
measured by a standard score of neurologic function. However, like other
aspects of therapeutic trials, clinical scores used in stroke trials have some
limitations. For instance, some scores for clinical outcome (e.g., Rankin
score) are global and nonspecific. Other scores, for example, the NIHSS, are
optimally suited for certain types of infarcts (e.g., left hemisphere
infarcts) and are less than optimal for infarcts in other locations.
Several groups have proposed imaging as a surrogate end point for clinical outcome. The limitation of this approach is that the impact of a given infarct on the clinical presentation depends on the infarct volume but is also strongly influenced by infarct location. To the extent that infarcts in different arterial territories are included in a stroke therapy trial, the degree of clinical benefit may vary between patients solely on the basis of infarct location, rather than on the therapy provided and its success (or lack thereof). For this reason, some therapeutic trials include ischemic lesions in solely one arterial territory, for example, the middle cerebral artery territory. Another alternative would be to develop an imaging scoring system integrating infarct location and size to predict outcome and monitor the efficacy of stroke treatment.
unresolved Issues Requiring Further Study
|
TopAbstractIntroductionSummary of Important Therapeutic...Controversies in Stroke Imaging...unresolved Issues Requiring...SummaryReferences |
|---|
Number and Timing of Imaging Studies in Stroke Treatment Trials
Brain imaging plays a major role in not only understanding the amount of
tissue at risk to proceed to infarction but also understanding the effects of
therapy. However, the exact number of times to perform imaging—and the
timing of the imaging examinations relative to therapy—has varied from
one therapeutic trial to another, challenging direct comparisons between the
results of the different trials. As noted in the discussion of the DIAS trial,
premature follow-up scanning could potentially result in an overestimation of
the PWI deficit and in an underestimation of the degree of reperfusion.
To assess the effects of therapy, at the very least, an imaging study immediately before therapy and one performed at a time at which the full effect of therapy can be determined are necessary. The baseline study, which may be CT or MRI, would ideally show infarcted tissue and tissue at risk for infarction (by perfusion imaging), presence of hemorrhage (using gradient-echo recalled sequences on MRI), and degree of artery patency (by CT or MR angiography). The imaging study for final assessment of therapeutic effect should, at a minimum, depict final infarct volume (considered by many investigators to mean imaging many weeks after therapy).
In addition to the two studies outlined, a study in the early posttreatment phase (e.g., in the range of 1–6 hours after treatment) that uses the same imaging technique (e.g., CT or MR) as the pretreatment study is appropriate to assess recanalization and reperfusion. As such, it is important that this imaging study includes a perfusion imaging sequence and a noninvasive angiography sequence. Finally, because thrombolytic therapies are associated with a risk of hemorrhage, an additional posttherapeutic study that is sensitive to detection of hemorrhage should be considered, especially in cases of clinical worsening.
Widely Accepted Definitions for Important Perfusion Parameters
One of the most important unresolved issues in stroke trials is that of
delineation of standard definitions for such important PWI concepts as the
ischemic penumbra. As an example, a single designation of the degree of
difference between size of PWI abnormality and DWI abnormality that should be
considered as a mismatch would eliminate one of the important differences
between trials. In addition, uniform definitions for PWI parameters, for
example, degree to which MTT must be prolonged to be considered abnormal, must
be established by a consensus of leading investigators in the stroke imaging
community. Furthermore, validation of imaging analysis techniques is needed,
so that the medical community widely adopts standard methods for processing of
PWI data.
|
TopAbstractIntroductionSummary of Important Therapeutic...Controversies in Stroke Imaging...unresolved Issues Requiring...SummaryReferences |
|---|
|
TopAbstractIntroductionSummary of Important Therapeutic...Controversies in Stroke Imaging...unresolved Issues Requiring...SummaryReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Sheikh, M. J. Schipper, and E. G. Hoeffner Feasibility of Superficial Temporal Artery as the Input Artery for Cerebral Perfusion CT Am. J. Roentgenol., June 1, 2009; 192(6): W321 - W329. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
