Original Paper Received: May 17, 2010 Accepted: September 13, 2010 Published online: October 27, 2010
Eur Neurol 2010;64:286–296 DOI: 10.1159/000321162
Combined Use of Pulsed Arterial Spin-Labeling and Susceptibility-Weighted Imaging in Stroke at 3T Magalie Viallon a Stephen Altrichter b Vitor Mendes Pereira a Duy Nguyen b Lucka Sekoranja c Andrea Federspiel d Zsolt Kulcsar b Roman Sztajzel c Rafik Ouared b Christophe Bonvin c Josef Pfeuffer e Karl-Olof Lövblad b
a Departments of Radiology, b Division of Neuroradiology, and c Department of Neurology, Geneva University Hospital, Genève, and d Department of Psychiatric Neurophysiology, University Hospital of Psychiatry, University of Bern, Bern, Switzerland; e Siemens Medical Solutions USA, Inc., Charlestown, Mass., USA
Key Words Arterial spin labeling ⴢ Contrast-enhanced perfusion-weighted imaging ⴢ Susceptibility weighted imaging ⴢ Diffusion tensor imaging
Abstract Background and Purpose: In acute stroke it is no longer sufficient to detect simply ischemia, but also to try to evaluate reperfusion/recanalization status and predict eventual hemorrhagic transformation. Arterial spin labeling (ASL) perfusion may have advantages over contrast-enhanced perfusion-weighted imaging (cePWI), and susceptibility weighted imaging (SWI) has an intrinsic sensitivity to paramagnetic effects in addition to its ability to detect small areas of bleeding and hemorrhage. We want to determine here if their combined use in acute stroke and stroke follow-up at 3T could bring new insight into the diagnosis and prognosis of stroke leading to eventual improved patient management. Methods: We prospectively examined 41 patients admitted for acute stroke (NIHSS 11). Early imaging was performed between 1 h and 2 weeks. The imaging protocol included ASL, cePWI, SWI, T2 and diffusion tensor imaging (DTI), in addition
© 2010 S. Karger AG, Basel 0014–3022/10/0645–0286$26.00/0 Fax +41 61 306 12 34 E-Mail
[email protected] www.karger.com
Accessible online at: www.karger.com/ene
to standard stroke protocol. Results: We saw four kinds of imaging patterns based on ASL and SWI: patients with either hypoperfusion and hyperperfusion on ASL with or without changes on SWI. Hyperperfusion was observed on ASL in 12/41 cases, with hyperperfusion status that was not evident on conventional cePWI images. Signs of hemorrhage or blood-brain barrier breakdown were visible on SWI in 15/41 cases, not always resulting in poor outcome (2/15 were scored mRS = 0–6). Early SWI changes, together with hypoperfusion, were associated with the occurrence of hemorrhage. Hyperperfusion on ASL, even when associated with hemorrhage detected on SWI, resulted in good outcome. Hyperperfusion predicted a better outcome than hypoperfusion (p = 0.0148). Conclusions: ASL is able to detect acutestage hyperperfusion corresponding to luxury perfusion previously reported by PET studies. The presence of hyperperfusion on ASL-type perfusion seems indicative of reperfusion/collateral flow that is protective of hemorrhagic transformation and a marker of favorable tissue outcome. The combination of hypoperfusion and changes on SWI seems on the other hand to predict hemorrhage and/or poor outcome. Copyright © 2010 S. Karger AG, Basel
Karl-Olof Lövblad Service de Radiologie, Hôpitaux Universitaires de Genève HUG 24 rue Micheli-du-Crest CH–1211 Geneva (Switzerland) Tel. +41 22 372 7039, Fax +41 22 372 7072, E-Mail karl-olof.lovblad @ hcuge.ch
In order to improve patient outcome after acute stroke we need to be able to better triage patients for therapy, since only patients with reversible ischemia will benefit from treatment. Image-guided selection of stroke treatment aims to identify these patients and involves: (1) finding the appropriate method and timing for imaging of the ischemic penumbra, i.e. the hypoperfused region, tissue not yet infarcted, but at risk of proceeding to infarction, and (2) finding its exact, but as yet unstated, role in patient triage. According to the current recommendations of the research societies [1], to evaluate the efficacy of reperfusion therapies or other interventions in acute stroke, patients should ideally undergo imaging at 4 time points: (1) detection of the initial parenchymal and vascular state, (2) estimation of the biological effect of the intervention, (3) the occurrence of early hemorrhagic transformation, and (4) the final tissue outcome. Typically, 1–6 h after treatment, patients should undergo a MRI study to assess for recanalization and reperfusion. Arterial occlusion is the first event in the chain of causality that leads to stroke, perfusion and diffusion imaging abnormalities, and ultimately infarction. Documentation of early reperfusion (whether spontaneous or following therapy) is important because it strongly influences the appropriate predictive analysis and maximizes ability to test acute imaging paradigms. Patients who achieve early reperfusion have penumbra distinct from core; nonrecanalizing patients have benign oligemia distinct from penumbra. In addition, a large number of PET studies previously demonstrated that early poststroke hyperperfusion may be beneficial and represents recanalization and subsequent reperfusion [2, 3]. Perfusion is traditionally investigated in MR by performing dynamic T2*-weighted imaging during the injection of gadolinium-based contrast agent (cePWI). While cePWI has been used successfully in the acute setting of ischemia [4–8], its use in the follow-up after stroke is more problematic: blood-brain barrier rupture may render the detection of reperfusion and/or tissue injury difficult. Advanced imaging strategies, like arterial spin labeling (ASL), do not rely on contrast agents but on arterial blood water as an endogenous tracer and may thus be immune from quantification problems related to gadolinium extravasation [9–12] and be more suited for investigating reperfusion and/or collateralization [9, 13]. Since stroke follow-up requires imaging at several times, perfusion studies without contrast agent injection are of great
interest, especially regarding current concerns over nephrogenic systemic fibrosis (NFS). ASL has been applied to stroke in acute phase within 24 h after onset of symptoms and in studies enrolling a small number of patients [9, 14]. However, only one study [9] using continuous ASL reported the case of 1 patient presenting hyperperfusion in the affected territory compared to the normal hemisphere. A ‘safety scan’ – either a noncontrast CT or MRI of the brain – is also recommended to assess the safety of therapies, particularly with respect to the presence and degree of any hemorrhagic transformation. It may be obtained systematically or only in case of clinical worsening, typically between 24 and 72 h after symptom onset. MRI is better able to detect partially bleeding, vascular stasis [15, 16] and blood-brain barrier breakdown phenomena [17]. Susceptibility weighted imaging (SWI) is a new high resolution technique that aims to detect paramagnetic small objects like clots, iron or hemorrhage with a much better sensitivity than conventional T2* gradient echo sequences [18, 19]. SWI uses phase shifts to magnify susceptibility contrast and susceptibility difference between deoxygenated blood (deoxy-Hb) and the surrounding tissue (Oxy-HB). In a preliminary study of hyperacute ischemia, Ida et al. [20] have seen SWI-enhanced cortical veins as well as medullary veins within the hypoperfused area around the core in 14 of 17 patients. They hypothesized that this increased BOLD signal (IBS) reflects relative elevations in concentration of deoxyhemoglobin in the draining veins due to elevation of oxygen extraction fraction (OEF), concluding that this area with IBS was indeed in misery perfusion state. In summary, imaging before and after intervention has several aims: it must exclude hemorrhage, differentiate ischemia versus oligemia, localize occlusion and determine if there is a tissue at risk (‘penumbra’). In the acute phase, the mismatch models obtained by looking at DWI and PWI penumbrae have been used with success. In follow-up, most people have concentrated on studying the effect of treatment as presence or absence of hemorrhage and possible final infarct. While MR techniques are extremely well suited in the acute phase for small lesions and have the advantage of offering full-brain coverage for diffusion and perfusion, it is in follow-up that MRI has to become even more established [21–24]. After the initial few hours, the DWI lesion size will tend to reach a maximum (possibly reverting compared to baseline scan). Only characterization combined with perfusion status can help to assess clinical outcome. Using newer sequences to assess perfusion may allow us to go even further in
Combined PASL and SWI in Stroke at 3T
Eur Neurol 2010;64:286–296
Introduction
287
the evaluation of the ischemic tissue after the initial event. Arterial spin-labeling (ASL) techniques have been in development and available for a number of years [11, 12, 25], but it is the increasing availability of clinical 3T scanners that renders their utilization possible in the daily clinical routine. ASL utility in the follow-up of stroke has not been addressed. Thus, we aim to investigate the benefit of the combined use of ASL perfusion and SWI in follow-up imaging of stroke patients. Materials and Methods Patients The study has been approved by our hospital ethics committee (study 08-049R; NAC 08-012). Patients were admitted to our hospital with a diagnosis of acute stroke. The patients were immediately transferred to our stroke team where a neurologist with experience in cerebrovascular diseases and their management examined them prior to imaging. National Institutes of Health Stroke Scale (NIHSS) was assessed immediately prior to scans on the arrival at the hospital. In the emergency room, a ‘baseline’ CT scan of the head was done. We prospectively examined 41 acute stroke patients (aged 50– 89 years, mean age: 70; 23 female and 18 male) admitted to our institution. The main inclusion criteria were to have an ischemic stroke based on clinical and/or imaging criteria; the visualization of another pathology on imaging would be an exclusion criteria. 35 patients had some kind of risk factor for cardiovascular and/or cerebrovascular diseases (30 had hypertension, 10 were smokers and 6 had diabetes). All patients received conventional antiplatelet treatment or tissue plasminogen activator if eligible for thrombolysis according to the NINDS inclusion and exclusion criteria [26]. Additional inclusion criteria were the possibility to undergo multimodality CT and MR agent, the incapacity to do this was an exclusion criteria. Over the whole patient population, 7 of 41 had thrombolytic treatment (7 patients received XA monitored intraarterial (IA) thrombolysis, 2 of 7 intravenous (IV) thrombolysis, and 5 of 7 combined (IA and IV) thrombolysis), 1 of 41 underwent angioplasty with MCA stenting, and a majority (33 of 41) received conventional antiplatelet treatment. The modified Rankin scale (mRS) was determined as late as possible, ideally on the day of the final late scan. The late clinical examination was done by a neurologist specialized in stroke from the same team as which had done the initial examination.
84 ms, b values of 0 and 1,000 30-gradient directions, 128 ! 128 matrix and 65 slices), 3D time-of-flight (TOF) MR angiogram of the intracranial arteries (TR: 21 ms, TE: 3.48 ms, 512 matrix and 208), T2* SWI (TR: 28 ms, TE: 20 ms; 320 matrix and 104 slices with a flip angle of 15°), T2-fluid attenuated inversion recovery (FLAIR: TR: 9,570 ms TE: 110 ms, 512 matrix and 32 slices), T2weighted precontrast imaging (TR: 4,000 ms, TE: 104 ms, 512 matrix), and T1-weighted postcontrast imaging (TR: 300 ms, TE: 3.4 ms 448 matrix, 32 slices), cePWI and pulsed arterial spin labeling (PASL) perfusion-weighted imaging (TR: 5,000 ms, TE: 15 ms, 16 slices, 64 matrix) providing relCBF images. PASL PWI uses a FAIR (slice-selective and nonslice-selective inversion pulse) alternate labeling scheme combined with a QUIPSSII module for bolus cut-off, as described in Wang et al. [27] and includes in-line prospective motion correction (PACE). Parametric maps were calculated in-line by the MRI scanner for DWI (ADC, FA, Trace) and for PASL (relCBF according to quantification formula as described by Wang et al. [27]). RelCBF, relCBV and MTT maps were calculated off-line for cePWI using the Syngo Perfusion (MR) software (Siemens Healthcare). In order to validate the in-line relCBF calculation implemented in the PASL sequence, we performed a second evaluation using the homemade Matlab program of Wang et al. [27] within spm2 software (http://www.fil.ion.ucl.ac.uk/spm/software/). For cePWI (TR: 1,410 ms, TE 26 ms, 20 slices, with a matrix of 128), half-dose contrast material GD DTPA (Gadovist-Schering, Germany) (equivalent to half-molar agent, 20 ml for 100 kg body weight) was injected at a rate of 4–6 ml/s (Medrad power injector) followed by 20–40 ml saline flush at the same injection rate. IV access 18–20 gauge IV line and right antecubital vein were preferred. During the off-line analysis of cePWI images, the arterial input function (AIF) was chosen by an experienced physicist according to previously published criteria [7, 28]. Based on imaging, the patients were classified as having the following types of infarction: 23 large territorial infarct, 3 watershed lesions, and 11 lacunes. The area of decreased and/or increased perfusion was determined visually by a single interpreter by positioning manually a ROI in the ischemic area and another ROI symmetrically in the unaffected hemisphere on the calculated CBF maps. Due to the relatively small size of the various groups, for statistical assessment we performed an Ansari-Bradley test, which presupposes that the distributions to be compared have the same form and medians but different variants. The perfusion status was correlated with clinical status.
Results Imaging The admission emergency room CT included a noncontrast CT, perfusion CT, CT-angiography (CTA) including the intracranial and cervical arteries and contrast-enhanced CT. MR imaging according to our protocol is then performed within 24 h ideally to 8 days for the safety or third scan. The mean stroke to MR interval is 48 h and a median of 20 h in a range interval of 6 h to 8 days. During follow-up, the patients had a CT or an MRI at discharge as well as an MRI at around 3 months. MR imaging was performed on a 3T (Siemens, Trio Tim System) equipped with a 12-channel receiver head coil. The MRI protocol includes: diffusion-weighted imaging (DWI: TR: 9,000, TE:
288
Eur Neurol 2010;64:286–296
All results are summarized in table 1 classifying our initial stroke population depending on ASL perfusion findings (observed hyper/hypoperfusion/none). Presence/absence of IBS on SWI, or ICH, patient data (received treatment, NIHSS score) and outcome characteristics for each patient (mRS) are also given. Diffusion Imaging. Thirty-seven of the 41 cases had findings compatible with acute lesions (23 had large territorial infarct, 3 presented watershed lesions and 11 laViallon et al.
Table 1. Patient data, perfusion and outcome characteristics
Patient population (n = 41) (F = 23/M = 18) Treatments (n = 7): thrombolysis IV (n = 2)/thrombolysis IA (n = 1)/thrombolysis IV+ IA (n = 4) No hypoperfusion/no hyperperfusion n = 14 Matching cePWI/ASL PWI = 14 Outcome: n+ = 13, n– = 1 SWI: edema = 1
Hypoperfusion ASL n = 15 Matching cePWI/ASL PWI = 14 Outcome: n+ = 9, n– = 5 SWI: ICH(*) = 9/ Embolus(•) = 2/edema = 2
Hyperperfusion ASL n = 12 Matching cePWI/ASL PWI = 1 Outcome: n+ = 12 SWI: ICH(*) = 7/ Embolus(•) = 1/edema = 4
Patient number
NIHSS
mRS
Patient number
NIHSS
mRSS
Patient number
NIHSS
mRS
Stent 1 2 3 4 tPA 5 6 7 8 tPA 9 tPA 10 11• 12 13 14
14 15 5 1 5 2 0 2 1 8 2 2 16 13
0–2 3–6 0–2 0–2 0–2 0–2 0–2 0–2 0–2 0–2 0–2 0–2 0–2 0–2
1* 2*• 3* 4* 5 6 7* 8* 9* 10* 11 12 13 14 tPA 15*•
5 2 19 21 11 6 7 9 18 9 12 12 13 0 19
0–2 0–2 3–6 3–6 0–2 0–2 0–2 0–2 0–2 0–2 3–6 3–6 3–6 0–2 3–6
1* tPA 2* 3 4* 5* 6 tPA 7* tPA 8 9* 10*• 11 12•
16 3 7 5 18 6 19 12 4 3 8 3
0–2 0–2 0–2 0–2 0–2 0–2 0–2 0–2 0–2 0–2 0–2 0–2
The patient population is separated according to perfusion patterns: on the left we have the patients with no hypoperfusion on ASL and SWI, in the middle the patients with hypoperfusion on both techniques on the right the patients with hyperperfusion on ASL. The middle column provides the acute NIHSS score and the right column provides the Rankin outcome scale.
cunar infarct) shown as hyperintensity on DWI with high b value and low ADC. NIHSS score ranged between 0 and 21 with a mean value of 7 and a median of 6. The patients who underwent treatment (n = 2) had a mean NIHSS score of 10 (median of 9) with 6 patients having a final Rankin score of 0–2 and 2 of 3–6; the patients who underwent no special treatment (n = 33) had a mean NIHSS score of 5 (median of 5) with 28 having a Rankin score of 0–2 and 5 of 3–6. Four patients had no lesion visible on the diffusion image or the other sequences that could be classified as a stroke; on follow-up imaging there was no lesion but clinically the final diagnosis was one of stroke despite the absence of imaging findings: this could be due to a size below the resolution of the technique. The smaller lacunar infarctions had lower NIHSS scores and better outcomes and the bigger the territorial infarction the worse were neurological status and outcome.
Contrast-Enhanced Perfusion Imaging. Fourteen of 37 patients with DWI lesions had detectable areas of hypoperfusion on the perfusion maps, 5 of 14 of them had very poor outcome (NIHSS = 3–6) and 9 of 14 had excellent outcome (NIHSS = 0–2). Fourteen of 37 patients had no apparent perfusion deficit with 2 of 14 having poor outcome (NIHSS = 3–6), the others grading NIHSS = 0–2 and having good outcome. Arterial Spin Labeling Perfusion Imaging. ASL evidenced 12 of the 37 cases of hyperperfusion in the affected vascular territory corresponding to the normally perfused area in cePWI. The CT baseline-scan hypodense lesion (or MR hyperintense DWI lesion) when distinct engulfs any observed hyperperfused area. The regions of hyperperfusion were essentially cortical, extensive and widespread with mean ASL relCBF increase of 120 8 83% compared to symmetrical or adjacent normally perfused tissue. One hyperperfusion area located in the putamen expressed a relCBF increase of 400%, and another one a 525% increase
Combined PASL and SWI in Stroke at 3T
Eur Neurol 2010;64:286–296
289
DWI
SWI
ceT1
PASL relCBF
cePWI relCBF
CT
Fig. 1. Images obtained in a 67-year-old male patient after IA + IV thrombolysis. Two days after stroke in the left MCA territory corresponding to right mild hemiparesis and aphasia (NIHSS = 7), DWI image shows an area of ischemia in the basal ganglia. No hemorrhage on SWI image: slight hypersignal leading to a loss of GM/WM contrast in corresponding area probably due to edema (green arrow). Hypoperfusion around lesion 1 (orange arrow) on both cePWI and ASL; hyperperfusion around lesion 2 (green arrow) in ASL, not evidenced by cePWI. Postcontrast T1 shows no BBB leakage. Final infarct on late CT scan (day 28) is small with total recovery seen on lesion 2. Lesion 1 final size equals DWI/ASLPWI match. Patient outcome was excellent (mRS = 0–2).
in cortical area leading to such high standard deviations. There was hemorrhage in 7/12 cases and 2/12 ischemic lesions combined with hyperperfusion and hemorrhage. These areas of hyperperfusion were always larger than the final infarct size (final core detected by chronic-stage CT or MR scan) and the ultimate infarct size never corresponds to hyperfusion (fig. 1, 2). In the apparent hyperperfusion lesions (as in fig. 1 and 2), the hyperperfusion on ASL was always associated with complete tissue recovery. We performed an Ansari-Bradley test to assess the correlations and we found: (1) hyperperfusion versus hypoperfusion: p = 0.0148; (2) no hypoperfusion versus hypoperfusion: p = 0.0423, and (3) hypoperfusion versus hyperperfusion: p = 0.3545. For lacunar infarctions, the spatial resolution of the ASL images is not sufficient to predict local perfusion de290
Eur Neurol 2010;64:286–296
fect when cePWI does. In all other hypoperfused cases, ASL demonstrates hypoperfusion equally as well as cePWI (fig. 1, 2), especially in territorial infarct where the lower spatial resolution of ASL images shows no disadvantage compared to cePWI. SWI Findings. When the stroke to MR interval was short (!6 h), we observed a venous stasis as previously described by Ida et al. [20], with SWI-enhanced cortical veins as well as medullary veins within the hypoperfused area around the core. The IBS vanished later. Figure 3 shows the same patients scanned twice, at 6 and 12 h after symptoms onset. The venous stasis disappears at 12 h and is replaced by a cancellation of grey to white matter contrast with less visible collateral veins (fig. 3: 12 h), or in the case of important edema (fig. 4), a slight signal attenuation (deViallon et al.
DWI
SWI
ceT1
PASL relCBF
cePWI relCBF
CT
Fig. 2. Stroke follow-up in an 80-year-old female patient. Two days after stroke in the left MCA territory associ-
ated with left mild hemiparesis and aphasia (NIHSS = 18), DWI image shows two areas of ischemia in the parietal territory, hypoperfusion around lesion 1 (orange arrow) on both cePWI and ASL, with localized microhemorrhage in SWI; lesion 2 (green arrow) corresponds to hyperperfusion in ASL not evidenced by cePWI, SWI evidenced loss of GM/WM contrast in the corresponding area. Postcontrast T1 shows no BBB leakage. Late CT shows a hypoattenuated area in the exact area predicted by DWI/ASLPWI match. Patient outcome was good (mRS = 0–2).
crease in BOLD signal, DBS). SWI lesions were found in 27 of 41 cases. DBS was visible in 7 of 27 of these cases in tissues within the affected vascular territory, ICH was detected in 16 of 27 cases and SWI was also very efficient in detecting core (n = 3 of 27) within the infarct or emboli within the vessels (orange arrow in fig. 3 and 5). After treatment, 13 of 27 cases with positive SWI lesions correlated with hypoperfusion on ASL (example in fig. 5) and 12 of 27 with ASL hyperperfusion (example in fig. 3). Chronic-stage CT or MR scans indicated poor outcome (mRS = 0–6) in 2 of 27 patients with SWI-identified lesions coupled to hypoperfusion and 5 of 15 patients with hypoperfusion. Chronic scans showed smaller infarcts compared to baseline scans and good outcome (mRS = 0–2) in all patients that were classified with acute-stage hyperperfusion even with SWI lesions.
When looking at neurological scores and outcomes depending on the imaging patterns, for the group with no hypoperfusion the mean NIHSS score was overall 6 with 7 for the treated cases (n = 4) and 6 for the nontreated ones (n = 10); in the cases where there was hypoperfusion with ASL we had patients with a mean NIHSS score of 11; 1 treated patient had an NIHSS score of 19 and the remaining nontreated patients had a mean NIHSS score of 10. In the group with hypoperfusion on ASL, the mean NIHSS score was 9 with 11 in the treated group (n = 3) and 8 in the nontreated group (n = 8). The groups with no hypoperfusion and with hyperperfusion had better outcomes overall with only 1 patient in the no hypoperfusion group with a mRS of 15. The group with hypoperfusion on ASL had a higher NIHSS score initially as well as 6 patients with a higher Rankin score (n = 6).
Combined PASL and SWI in Stroke at 3T
Eur Neurol 2010;64:286–296
291
6h
T2
SWI minIP
SWI
DWI
PASL relCBF
cePWI relCBF
T2
SWI minIP
SWI
DWI
PASL relCBF
cePWI relCBF
T2
DWI
FLAIR
292
Control scan
12 h in a 51-year-old male patient (NIHSS = 3). Control images obtained at 6 months. At 6 h, DWI shows an area of diminished diffusion in the left MCA territory. There is a corresponding area of hypoperfusion on the cePWI as well as on the ASL with a large DWI/PWI penumbra. SWI image evidenced embolus within the vessel and the presence of large hypoattenuated cortical and medullary veins within the hypoperfused area due to IBS reflecting relative elevations in concentration of deoxyhemoglobin in the draining veins due to elevation of OEF. The area with IBS is in misery perfusion state. At 12 h and after IA thrombolysis, the area of diminished diffusion is increased in size with a corresponding area of hyperperfusion on the ASL PWI (luxury perfusion) undetected on the cePWI images. SWI venous stasis observed at 6 h has disappeared and loss of GM/ WM contrast can be appreciated within the hyperperfusion area. Patient outcome was good (mRS = 0–2).
12 h
Fig. 3. Stroke follow-up at 6 and
Eur Neurol 2010;64:286–296
Viallon et al.
T2
DWI
SWI
PASL relCBF
cePWI relCBF
SWI minIP
Fig. 4. Stroke follow-up at 5 days after stoke on a 74-year-old female patient. DWI shows a large area of dimin-
ished diffusion in the right MCA territory and an even larger area of hypoperfusion on the ASL PWI maps that match with the hypoperfusion area detected on the cePWI images and evidence a large DWI/PWI penumbra. SWI image shows a large hypoattenuated area in the left hemisphere that engulfs the hyperintense DWI ischemic lesion and is due to increased water volume in edema and partial volume effect with LCR in the cortical area.
ASL Findings Like others, we have observed that ASL is able to reproduce the hypoperfusion due to ischemia that we see with cePWI, but using no contrast agent so that it is less sus-
ceptible to the natural breakdown of the blood-brain barrier that occurs after ischemia. ASL relCBF maps clearly give complementary information regarding infarct and peri-infarct perfusion status that are compatible with peri-infarct hyperperfusion representing partial recanalization. This finding is coherent with the results recently obtained by Golay’s group [13] which demonstrate that territorial ASL (TASL) is in agreement with collateral circulation evaluated with digital subtraction angiography. In another preliminary study, the same authors used TASL to label the blood of only the vessel feeding the vascular territory including stroke, demonstrating that occurring perfusion could only be explained by collateral flow since the measured vascular territory extended much further than the vascular territory of the labeled vessel [29] . All patients with hyperperfusion on ASL had Rankin
Combined PASL and SWI in Stroke at 3T
Eur Neurol 2010;64:286–296
Discussion
We have found that in the follow-up of cerebral ischemia, the combined use of ASL perfusion and SWI can improve prediction of tissue outcome by not being influenced by the effects of blood-brain barrier breakdown and by demonstrating collateral flow. We also found that ASL can reproduce findings of contrast-enhanced perfusion MR. Both techniques covering the whole brain can be compared and are ideally suited for the evaluation of patients with an ischemic brain lesion.
293
T2
DWI
T2
DWI
SWI
SWI
TOF
SWI
SWI
MTT
SWI
MTT
cePWI relCBF
PASL relCBF
cePWI relCBF
PASL relCBF
Fig. 5. Stroke follow-up in a 72-year-old male patient (NIHSS = 3). Six hours after stoke (left), DWI shows an area of diminished diffusion in the left MCA territory with a corresponding area of hypoperfusion on cePWI and on the ASL images with a large DWI/ PWI penumbra. SWI image evidenced several lesions (orange arrow) showing IBS: embolus within the vessel, core within the infarct and the presence of large hypoattenuated cortical and medul-
294
Eur Neurol 2010;64:286–296
lary veins within the hypoperfused area in misery perfusion state. Right images obtained at 10 days after IA thrombolysis: area of diminished diffusion in the left MCA territory is increased in size on the DWI with increased corresponding area of hypoperfusion on both the ASL PWI and the cePWI images. On the SWI image, the venous stasis described at 6 h is extended and misery perfusion state extends to the wall territory with visible ICH (green arrow).
Viallon et al.
scores compatible with a better outcome. A disadvantage of cePWI is that it relies on the determination of an arterial input function (AIF) that is in practice chosen in the main feeding arteries around the circle of Willis and as a consequence is never correctly chosen to match to the theory of an independent and unique AIF for each pixel. This drawback means ASL can better demonstrate reperfusion and or collateral flow potentially causing hyperperfusion [30]. Recanalization and reperfusion status are considered as a key determinant of tissue outcome and one that can be influenced by treatment. It is therefore very important to have no bias in determining them. Also, ASL provides lower signal intensity than what is obtained with conventional T2* imaging, which may be problematic in some cases. Nevertheless, since the PASL technique used here relies on the correct determination of the transit time of the labeled blood, we could only explore a given range of collateral flow. As a consequence, the generated ASL relCBF map may not be sensitive to collateral flow above a given value. Similarly, in areas with transit times longer than 1,100 ms, relCBF will be underestimated because we do not wait long enough to collect all labels. Some stroke regions can have transit times of 2 s, and no labeled signal will remain due to T1 relaxation. In our study, the fact that the area of hyperperfusion was always much larger than the final infarct size seems to indicate that hyperperfusion in itself does not cause tissue necrosis, a result also observed by Marchal et al. [31]. But in their study, only the patients with no hemorrhage detected on CT were considered, whereas in our study we also included patients who had positive ICH coupled with hyperperfusion area on ASL (at least 7 of 12). Chalela et al. [9] reported the case of a 10% signal increase in one of their 15 patients scanned within 24 h. The values of percentage signal difference found in this study are higher and are closer to an average CBF increase of 74% found in the PET study by Marchal et al. [31]. SWI Findings SWI demonstrates small areas of signal loss and bleeding better than standard T2 and T2* imaging. SWI also showed emboli in the main trunk as hypo intense dots (fig. 3, 5: orange arrow) due to T2* effects in vessels related to the presence of red blood cells in the cardioembolic type of thrombus [32]. These vascular hypo-intensities were reported to be predictors of future hemorrhagic transformation [15, 16] and resulted in poor outcome in these earlier studies. In our patients presenting Combined PASL and SWI in Stroke at 3T
such SWI abnormalities (3 of 15), the outcome was good (mRS = 0–2). It is nevertheless impossible to state that SWI has the ability to show OEF penumbra (misery perfusion) at 3T during follow-up which is indeed the metabolic parameter that relates directly to the cerebral metabolic rate of oxygen consumption (CMRO2) and that differs from the CBF and DWI-PWI penumbra [33]. Depending on the time interval between stroke and MR, SWI will sequentially initially show IBS (venous stasis), followed by DBS after settling of edema and cell swelling. Thus, we see four kinds of imaging patterns based on ASL and SWI: Patients with either hypoperfusion or hyperperfusion on ASL with or without changes on SWI. Patients with hyperperfusion on ASL developed smaller lesions on follow-up images. Hyperperfusion occurs whether there are changes on SWI or not. Hypoperfusion on ASL along with susceptibility changes did not appear to be an indicator of possible further hemorrhagic complications in our study. Patients with hypoperfusion on ASL would develop hemorrhage if there were changes on SWI. ASL seems also to demonstrate quite well the presence of hyperperfusion/collateralization after stroke that seems to protect the tissue from eventual hemorrhagic transformation. All our patients that exhibit hyperperfusion on PWI ASL had excellent outcome, even in the presence of ICH. ASL may differentiate between oligemia and misery perfusion and may therefore be very useful since treatable patients are those without PWI/DWI mismatch, but who have enough collaterals in order to expand the time-totreatment window. There is currently no good triage tool for these patients. SWI and ASL are potentially two new tools to investigate and differentiate stroke patients at the early phase. Since ASL and SWI performances are improved with field strength, they could be valuable tools to be added to the battery of new emerging techniques to identify salvageable tissue such as sodium or oxygen-17 at high field. One needs then to develop on a larger scale how these tools could bring new insights in stroke follow-up and help in tissue specific treatment screening.
Acknowledgements This study was supported by a grant from the Swiss National Science Foundation: multimodality MR imaging in tissue characterization after stroke. SNF grant number 320000-121565.
Eur Neurol 2010;64:286–296
295
References 1 Wintermark M, Albers GW, Alexandrov AV, et al: Acute stroke imaging research roadmap. AJNR Am J Neuroradiol 2008; 29:e23– e30. 2 Marchal G, Young AR, Baron JC: Early postischemic hyperperfusion: pathophysiologic insights from positron emission tomography. J Cereb Blood Flow Metab 1999;19:467– 482. 3 Zaharchuk G, Yamada M, Sasamata M, Jenkins BG, Moskowitz MA, Rosen BR: Is all perfusion-weighted magnetic resonance imaging for stroke equal? The temporal evolution of multiple hemodynamic parameters after focal ischemia in rats correlated with evidence of infarction. J Cereb Blood Flow Metab 2000;20:1341–1351. 4 Karonen JO, Vanninen RL, Liu Y, Ostergaard L, Kuikka JT, Nuutinen J, Vanninen EJ, Partanen PL, Vainio PA, Korhonen K, Perkio J, Roivainen R, Sivenius J, Aronen HJ: Combined diffusion and perfusion mri with correlation to single-photon emission ct in acute ischemic stroke. Ischemic penumbra predicts infarct growth. Stroke 1999; 30: 1583– 1590. 5 Lovblad KO, Baird AE: Actual diagnostic approach to the acute stroke patient. Eur Radiol 2006;16:1253–1269. 6 Ostergaard L, Sorensen AG, Chesler DA, Weisskoff RM, Koroshetz WJ, Wu O, Gyldensted C, Rosen BR: Combined diffusion-weighted and perfusion-weighted flow heterogeneity magnetic resonance imaging in acute stroke. Stroke 2000; 31:1097–1103. 7 Rempp KA, Brix G, Wenz F, Becker CR, Guckel F, Lorenz WJ: Quantification of regional cerebral blood flow and volume with dynamic susceptibility contrast-enhanced mr imaging. Radiology 1994; 193:637–641. 8 Wu O, Koroshetz WJ, Ostergaard L, Buonanno FS, Copen WA, Gonzalez RG, Rordorf G, Rosen BR, Schwamm LH, Weisskoff RM, Sorensen AG: Predicting tissue outcome in acute human cerebral ischemia using combined diffusion- and perfusion-weighted mr imaging. Stroke 2001; 32:933–942. 9 Chalela JA, Alsop DC, Gonzalez-Atavales JB, Maldjian JA, Kasner SE, Detre JA: Magnetic resonance perfusion imaging in acute ischemic stroke using continuous arterial spin labeling. Stroke 2000; 31:680–687. 10 Golay X, Hendrikse J, Van Der Grond J: Application of regional perfusion imaging to extra-intracranial bypass surgery and severe stenoses. J Neuroradiol 2005; 32:321–324. 11 Wong EC: Quantifying CBF with pulsed ASL: technical and pulse sequence factors. J Magn Reson Imaging 2005;22:727–731. 12 Wong EC, Buxton RB, Frank LR: Quantitative imaging of perfusion using a single subtraction (QUIPSS and QUIPSS II). Magn Reson Med 1998;39:702–708.
296
13 Lim CC, Petersen ET, Ng I, Hwang PY, Hui F, Golay X: MR regional perfusion imaging: visualizing functional collateral circulation. Am J Neuroradiol 2007;28:447–448. 14 Siewert B, Schlaug G, Edelman RR, Warach S: Comparison of Epistar and T2*-weighted gadolinium-enhanced perfusion imaging in patients with acute cerebral ischemia. Neurology 1997; 48:673–679. 15 Hermier M, Nighoghossian N, Derex L, Adeleine P, Wiart M, Berthezene Y, Cotton F, Pialat JB, Dardel P, Honnorat J, Trouillas P, Froment JC: Hypointense transcerebral veins at T2*-weighted MRI: a marker of hemorrhagic transformation risk in patients treated with intravenous tissue plasminogen activator. J Cereb Blood Flow Metab 2003;23: 1362–1370. 16 Hermier M, Nighoghossian N, Derex L, Wiart M, Nemoz C, Berthezene Y, Froment JC: Hypointense leptomeningeal vessels at T2*weighted MRI in acute ischemic stroke. Neurology 2005; 65:652–653. 17 Hjort N, Wu O, Ashkanian M, Solling C, Mouridsen K, Christensen S, Gyldensted C, Andersen G, Ostergaard L: MRI detection of early blood-brain barrier disruption: parenchymal enhancement predicts focal hemorrhagic transformation after thrombolysis. Stroke 2008; 39:1025–1028. 18 Haacke EM: Susceptibility weighted imaging (SWI). Z Med Phys 2006;16:237. 19 Sehgal V, Delproposto Z, Haacke EM, Tong KA, Wycliffe N, Kido DK, Xu Y, Neelavalli J, Haddar D, Reichenbach JR: Clinical applications of neuroimaging with susceptibilityweighted imaging. J Magn Reson Imaging 2005;22:439–450. 20 Masahiro I: Diffusion and perfusion MR for acute ischemic stroke: emergency clinical protocol and advances. Int Congr Ser 2006: 37–44. 21 Rohl L, Geday J, Ostergaard L, Simonsen CZ, Vestergaard-Poulsen P, Andersen G, Le Bihan D, Gyldensted C: Correlation between diffusion- and perfusion-weighted MRI and neurological deficit measured by the Scandinavian Stroke Scale and Barthel Index in hyperacute subcortical stroke (! or = 6 h). Cerebrovasc Dis 2001;12:203–213. 22 Rohl L, Ostergaard L, Simonsen CZ, Vestergaard-Poulsen P, Andersen G, Sakoh M, Le Bihan D, Gyldensted C: Viability thresholds of ischemic penumbra of hyperacute stroke defined by perfusion-weighted MRI and apparent diffusion coefficient. Stroke 2001; 32: 1140–1146.
Eur Neurol 2010;64:286–296
23 Sakoh M, Ostergaard L, Rohl L, Smith DF, Simonsen CZ, Sorensen JC, Poulsen PV, Gyldensted C, Sakaki S, Gjedde A: Relationship between residual cerebral blood flow and oxygen metabolism as predictive of ischemic tissue viability: sequential multitracer positron emission tomography scanning of middle cerebral artery occlusion during the critical first 6 hours after stroke in pigs. J Neurosurg 2000;93:647–657. 24 Simonsen CZ, Rohl L, Vestergaard-Poulsen P, Gyldensted C, Andersen G, Ostergaard L: Final infarct size after acute stroke: prediction with flow heterogeneity. Radiology 2002;225:269–275. 25 Wong EC, Buxton RB, Frank LR: A theoretical and experimental comparison of continuous and pulsed arterial spin labeling techniques for quantitative perfusion imaging. Magn Reson Med 1998;40:348–355. 26 National Institute of Neurological Disorders and Stroke RT-PA Stroke Study Group: Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 1995; 333:1581–1587. 27 Wang J, Licht DJ, Jahng GH, Liu CS, Rubin JT, Haselgrove J, Zimmerman RA, Detre JA: Pediatric perfusion imaging using pulsed arterial spin labeling. J Magn Reson Imaging 2003;18:404–413. 28 Ostergaard L, Weisskoff RM, Chesler DA, Gyldensted C, Rosen BR: High resolution measurement of cerebral blood flow using intravascular tracer bolus passages. 1. Mathematical approach and statistical analysis. Magn Reson Med 1996;36:715–725. 29 Chng SM, Petersen ET, Zimine I, Sitoh YY, Lim CC, Golay X: Territorial arterial spin labeling in the assessment of collateral circulation: comparison with digital subtraction angiography. Stroke 2008; 39:3248–3254. 30 Wu O, Ostergaard L, Sorensen AG: Technical aspects of perfusion-weighted imaging. Neuroimaging Clin N Am 2005;15:623–637, xi. 31 Marchal G, Furlan M, Beaudouin V, Rioux P, Hauttement JL, Serrati C, de la Sayette V, Le Doze F, Viader F, Derlon JM, Baron JC: Early spontaneous hyperperfusion after stroke: a marker of favourable tissue outcome? Brain 1996;119:409–419. 32 Cho KH, Kim JS, Kwon SU, Cho AH, Kang DW: Significance of susceptibility vessel sign on T2*-weighted gradient echo imaging for identification of stroke subtypes. Stroke 2005;36:2379–2383. 33 Sobesky J, Zaro Weber O, Lehnhardt FG, Hesselmann V, Neveling M, Jacobs A, Heiss WD: Does the mismatch match the penumbra? Magnetic resonance imaging and positron emission tomography in early ischemic stroke. Stroke 2005; 36:980–985.
Viallon et al.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.