Clinical Neurology and Neurosurgery 121 (2014) 39–45
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Clinical/perfusion CT CBV mismatch as prognostic factor in intraarterial thrombectomy in acute anterior circulation stroke ¨ ¨ Marc Saake ∗ , Lorenz Breuer, Philipp Golitz, Martin Kohrmann, Stefan Schwab, ¨ Arnd Dorfler, Stephan Kloska Department of Neuroradiology, University of Erlangen-Nuremberg, Schwabachanlage 6, 91054 Erlangen, Germany
a r t i c l e
i n f o
Article history: Received 29 October 2013 Received in revised form 23 January 2014 Accepted 6 March 2014 Available online 15 March 2014 Keywords: Stroke Perfusion CT Cerebral blood volume Thrombectomy Mismatch
a b s t r a c t Objective: Contradictory results were reported for the outcome after endovascular recanalization (ERT) in acute anterior circulation ischemic stroke. We assessed whether a clinical/perfusion CT cerebral blood volume (CBV) mismatch concept (CPM) can identify patients who will benefit from reperfusion therapy. Methods: Imaging and clinical data of 58 consecutive ERT cases with acute anterior circulation stroke (ICA, M1, proximal M2) undergoing intraarterial thrombectomy within 4.5 h after symptom onset were analyzed retrospectively. CPM was defined as NIHSS ≥ 8 and PCT CBV ASPECTS ≥ 7. Minor CBV lesion was defined as PCT CBV ASPECTS ≥ 7. Results: All baseline characteristics other than blood glucose did not differ between the paired groups. Revascularization was achieved in 87.9% of all patients without significant difference between the paired groups. Favorable clinical outcome after 3 months (mRS ≤ 2) resulted in 29.3% of all patients, in contrast to 47.4% of the CPM positive and 52.2% of the minor CBV lesion groups. Conclusion: CPM can identify patients who will benefit from reperfusion therapy in acute anterior circulation ischemic stroke. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Stroke is one of the leading causes of morbidity and mortality worldwide and a common cause of depression, epilepsy, and dementia in the elderly. 85% of strokes are ischemic, resulting from an arterial vessel occlusion. Depending on the site of occlusion, mortality rates reach 53–92%, with higher mortality rates in large vessel occlusion. The keystones of acute stroke treatment are fast and accurate detection of brain tissue at risk and time sensitive performance of reperfusion therapy in patients eligible. Beside factors like time from symptom onset, infarct demarcation on noncontrast CT (NCCT) and initial National Institute of Health stroke scale (NIHSS), a beneficial patient selection criterion for reperfusion
Abbreviations: IVT, intravenous thrombolysis; ERT, endovascular reperfusion therapy; NCCT, non-contrast CT; PCT, perfusion CT; CBV, cerebral blood volume; CBF, cerebral blood flow; DWI, diffusion weighted imaging; NIHSS, National Institute of Health Stroke Scale; MMCT, multimodal CT; mRS, modified Rankin Scale; CTA, CT angiography; TTP, time to peak; DSA, digital subtraction angiography; ASPECTS, Alberta Stroke Program Early CT score; MCA, middle cerebral artery; ROC, receiver operating characteristics; PPV, positive predictive value; NPV, negative predictive value; CPM, clinical/PCT CBV mismatch. ∗ Corresponding author. Tel.: +49 91318539388. E-mail address: [email protected] (M. Saake). http://dx.doi.org/10.1016/j.clineuro.2014.03.007 0303-8467/© 2014 Elsevier B.V. All rights reserved.
therapy might be the size of the mismatch between the potentially salvageable tissue (penumbra) and the irreversibly damaged tissue (core) [1,2]. Reperfusion therapy is the only proven causal treatment for acute ischemic stroke. Currently intravenous thrombolysis (IVT) is the only FDA-approved reperfusion strategy. However, IVT has several limitations such as a limited time window and a weak effect in ischemic strokes caused by large vessel occlusions. In these cases, interventional procedures might improve recanalization rates and patient outcome [3]. Significant improvements have been made in the last years in the development of endovascular reperfusion therapies (ERT) [4]. By now, ERT has been demonstrated to be an effective means of achieving reperfusion in stroke patients with proximal cerebral artery occlusions [5]. However, favorable outcome after stroke intervention remains limited, with only 25–50% of patients achieving a modified Rankin Scale ≤ 2 at 90 days [6,7]. Moreover, recent reports suggested that outcome after ERT might be equal to IVT [8,9]. Combined clinical/imaging mismatch concepts compare the clinical symptoms to the imaging findings. Numerous studies have shown that the clinical/MRI diffusion mismatch between the NIHSS score and the infarct size on diffusion weighted imaging (DWI) may identify patients at risk of infarct growth and early neurological deterioration [10–12]. However, CT represents the most
40
M. Saake et al. / Clinical Neurology and Neurosurgery 121 (2014) 39–45
frequently used imaging modality in acute stroke patients. Previous studies have reported poor sensitivity and poor interobserver agreement for the detection of early ischemic changes on NCCT [13]. Several mismatch concepts have been proposed to improve CT stroke detection. A clinical/NCCT approach did not reliably identify patients more or less likely to benefit from intravenous thrombolysis [14]. Perfusion CT (PCT) has been shown to be an accurate predictor of both thrombolytic therapy benefit and clinical outcome [15]. The most frequently used PCT-based approach compares the region of reduced cerebral blood volume (CBV, presumably representing irreversible infarction) and the area of reduced cerebral blood flow (CBF) that lies outside of the CBV abnormality (potentially representing reversible ischemia). The region of reduced CBV has been shown to most accurately describe the infarct core [16] and to correlate with MRI DWI lesions in hyperacute stroke [17]. We hypothesized that clinical/PCT CBV mismatch (CPM) between the patient’s clinical stroke severity defined by the NIHSS score and the PCT CBV lesion size can identify patients who will benefit from reperfusion therapy in acute anterior circulation ischemic stroke.
were: section thickness, 5.0 mm; increment, 5.0 mm; reconstruction kernel, H31s. Primary PCT image reconstruction parameters were: section thickness, 5.0 mm; increment, 5.0 mm; reconstruction kernel, H20f. From this dataset PCT parameter maps (CBV, CBF, and time to peak (TTP)) were calculated using a CT workstation (Multimodality Workplace MMWP; Siemens, Forchheim, Germany) running standard software (Syngo VPCT Neuro; Siemens, Forchheim, Germany). All digital subtraction angiographies (DSA) were performed on a biplane flat-detector angiography system (Axiom Artis dBA; Siemens, Forchheim, Germany) by staff neuroradiologists. Using standard angiographic procedures, the affected internal carotid artery was investigated. For ERT procedures the device vendor guidelines were followed. The selection of the used reperfusion catheter system was to the discretion of the neurointerventionalist. Generally, all ERT systems aim to remove the clot from the cerebral artery, but they differ in technical details. NCCT was repeated at 24–72 h after symptom onset for followup. 2.4. Image analysis
2. Materials and methods 2.1. Study population Institutional review board approval was obtained before the commencement of this retrospective study. Informed consent of the patient or next of kin was obtained according to local laws and regulations. A total of 277 consecutive patients who underwent multimodal stroke CT (MMCT) in our department between June 2010 and July 2012 were analyzed retrospectively for the study. Inclusion criteria were acute anterior circulation stroke, use of MMCT for primary imaging, proximal anterior circulation vessel occlusion (i.e. occlusion of the internal carotid artery, M1, and/or proximal M2 segment of the middle cerebral artery) and performance of ERT due to clinical indication. Exclusion criteria were an age of less than 18 years, incomplete or non-diagnostic examinations, and signs of hemorrhage in the initial NCCT. Additional IVT was not considered an exclusion criterion, as IVT is the primary standard therapy in acute stroke. 2.2. Stroke therapy and follow-up All stroke patients admitted to our hospital are treated at our local stroke unit. According to standard procedures the severity of clinical stroke symptoms is assessed in every patient by the National Institutes of Health Stroke Scale (NIHSS) [18]. All eligible stroke patients receive body weight adapted (0.9 mg tPA/kg) IVT according to standard guidelines [19]. As outcome measure we used the modified Rankin Scale (mRS), a 7-point scoring system that distributes patients from 0 (no symptoms) to 6 (death) [20]. The pre-stroke mRS score was determined by a structured questionnaire during the hospital stay of the patient. Three months after stroke onset, the structured questionnaire was performed again for all patients by telephone to assess the level of independence at home. 2.3. Imaging and endovascular recanalization therapy (ERT) In all acute stroke patients a standardized MMCT protocol, consisting of NCCT, PCT with 9.6 cm whole brain coverage in the z-axis, and supra-aortic CT angiography (CTA) was performed according to hospital guidelines. All examinations were performed on a 128-section CT scanner (Somatom Definition AS+; Siemens, Forchheim, Germany). Image reconstruction parameters for NCCT
All examinations were anonymized. Images were transferred to a workstation (MMWP, Siemens, Forchheim, Germany). Review and analysis of the MMCT and ERT images was performed by a senior neurology resident and a senior neuroradiology resident in consensus. For assessment of the NCCT images and the color-coded PCT parameter maps the Alberta Stroke Program Early CT Score (ASPECTS) was used. The ASPECTS is a semiquantitative approach to infarct sizing on NCCT that has been demonstrated to be both reliable and predictive of outcome following acute stroke treatment [21]. The ASPECTS ranges from 0 to 10. For signs of infarction in each of the 10 parts of the middle cerebral artery (MCA) territory one point is subtracted from 10. Thus, a minor infarction has a high, a major infarction a low ASPECTS. Meanwhile, the ASPECTS has been adapted to the similar assessment of PCT parameter maps [22]. ERT images were scored for recanalization success, using the primary arterial occlusive lesion (AOL) [23] and thrombolysis in cerebral infarction (TICI) [24] scores. 2.5. Data analysis It has been suggested that an NIHSS score of ≥8 is associated with a high rate of neurological deterioration, a low frequency of spontaneous functional recovery, and cortical perfusion deficits [11,25]. So, we used this cut-off point as a clinical indicator of a large volume of ischemic brain tissue. Patients were dichotomized for favorable versus unfavorable outcome at 90 days, with favorable outcome defined as mRS ≤ 2. Receiver operating characteristics (ROC) analysis was performed for PCT CBV ASPECTS in regard of patient outcome. The optimal threshold level for dichotomization of PCT CBV ASPECTS was determined by maximizing the sum of sensitivity and specificity, and minimizing the Euclidean distance of (sensitivity, specificity) to the point (1, 1) of the ROC curve [26]. We divided the study population into the following groups: 1. minor stroke (NIHSS < 8) versus major stroke (NIHSS ≥ 8) groups (NIHSS approach), 2. minor CBV lesion (PCT CBV ASPECT score > threshold) versus major CBV lesion (PCT CBV ASPECT score < threshold) groups (CBV size approach), 3. clinical/PCT CBV matched (CPM negative) versus clinical/PCT CBV mismatched (CPM positive) groups (CPM approach). We compared the paired groups with regard to the following variables: background characteristics including age, gender,
M. Saake et al. / Clinical Neurology and Neurosurgery 121 (2014) 39–45
41
Table 1 Background characteristics, risk factors, vital signs and stroke subtype in the paired groups. All patients
Age, years Gender, female:male Time from onset to CT (h) Hypertension (%) Diabetes mellitus (%) Stroke in history (%) Atrial fibrillation (%) Antithrombotic therapy (%) Systolic blood pressure (mmHg) Diastolic blood pressure, mmHg Body temperature (◦ C) Blood glucose (mg/dL) Lesion side, left:right Stroke subtype classification
Initial NIHSS ≥ 8
All patients
CBV ≥ 7 N = 23
CBV < 7 N = 35
p value
CPM positive N = 19
CPM negative N = 39
p value
CBV ≥ 7 N = 19
CBV < 7 N = 35
p value
69.5 ± 11.0 12:11 2.1 ± 1.3 94.3 30.4 21.7 47.8 34.8 160.2 ± 25.5 87.7 ± 23.3 36.3 ± 0.6 118.7 ± 29.9 13:10 7:7:5:0:4
72.0 ± 11.8 19:16 2.4 ± 1.7 100 28.6 28.6 60.0 42.9 159.8 ± 29.8 84.7 ± 17.6 36.2 ± 0.6 139.9 ± 41.4 19:16 12:17:2:0:4
0.42 1.00 0.46 0.51 1.00 0.76 0.43 0.59 0.96 0.60 0.88 0.03 1.00 0.21
69.5 ± 11.5 11:8 2.1 ± 1.1 100 36.8 15.8 47.4 31.6 157.0 ± 25.9 85.4 ± 22.0 36.3 ± 0.6 116.6 ± 28.8 10:9 6:7:3:0:3
71.7 ± 11.5 20:19 2.3 ± 1.7 94.9 25.6 30.8 59.0 43.6 161.4 ± 29.1 86.1 ± 19.1 36.2 ± 0.6 138.8 ± 40.7 22:17 13:17:4:0:5
0.50 0.78 0.52 0.55 0.54 0.34 0.57 0.41 0.56 0.91 0.48 0.02 1.00 0.91
69.5 ± 11.5 11:8 2.1 ± 1.1 100 36.8 15.8 47.4 31.6 157.0 ± 25.9 85.4 ± 22.0 36.3 ± 0.6 116.6 ± 28.8 10:9 6:7:3:0:3
72.0 ± 11.8 19:16 2.1 ± 1.3 94.3 28.6 28.6 60.0 42.9 159.8 ± 29.8 84.7 ± 17.6 36.2 ± 0.6 139.9 ± 41.4 19:16 12:17:2:0:4
0.46 1.00 0.92 0.54 0.55 0.34 0.40 0.56 0.72 0.90 0.61 0.02 1.00 0.27
Bold text indicates a statistically significant difference with p ≤ 0.05.
time from onset of symptoms to CT, hypertension (past use of antihypertensive agents or blood pressure of >140/90 mmHg at least twice before onset), diabetes mellitus (use of insulin or oral hypoglycemic agents, fasting blood glucose ≥ 126 mg/dL or random blood glucose ≥ 200 mg/dL), stroke in patient history, atrial fibrillation in recent patient history, antithrombotic therapy before onset; vital signs including systolic blood pressure, diastolic blood pressure, body temperature, and blood glucose at entry; side of vessel lesion; ischemic stroke subtype classified according to the TOAST criteria (large-artery atherosclerosis, including large-artery thrombosis and artery-to-artery embolism; cardioembolism; small-artery occlusion; stroke of other determined cause; and stroke of undetermined cause) [27]; pre-stroke mRS; NIHSS score at entry; outcome after 3 months measured by the mRS; mRS deterioration defined by an increase of ≥3 points in mRS after 3 months compared to pre-stroke mRS; number of patients with mRS ≤ 2 after 3 months; baseline NCCT ASPECTS; PCT CBV/CBF/TTP ASPECTS; revascularization of the target vessel defined as an AOL score of II or III and a TICI score of 2a or better; and ASPECTS in follow-up NCCT 24–72 h after onset. For all statistical analysis SPPS 11 (SPSS, Chicago, IL, USA) was used. Student’s t-test or Mann–Whitney test was used in univariate analysis for continuous and interval variables, and Fisher’s exact test for non-continuous variables. A level of p ≤ 0.05 was regarded as statistically significant. Sensitivity, specificity, as well as positive predictive values (PPV) and negative predictive values (NPV) were calculated.
3. NIHSS < 8 and PCT CBV ASPECTS ≥ 7, N = 4 All of the 4 patients with an entry NIHSS score < 8 had a PCT CBV ASPECTS of ≥7. Good outcome was achieved in 3 of these 4 patients (75%). An additional statistical analysis was performed without these 4 patients. Due to the exclusion of patients with a NIHSS < 8, the CBV and CPM approaches were identical for this subgroup analysis. Table 1 describes the background characteristics, risk factors, vital signs, and stroke subtypes in the groups. The only statistically significant difference between the paired groups was blood glucose concentration at admission, which was higher in the CPM negative and the major CBV lesion groups, respectively.
3.2. Stroke therapy Bridging IVT was performed in all patients. In 19 patients thrombolysis was switched from intravenous to intraarterial. The used ERT systems were: Penumbra System in 37 patients (Penumbra Inc., Alameda, USA), Solitaire FR Revascularization Device in 9 patients (ev3 Neurovascular, Irvine, USA), pRESET Thrombectomy Retriever in 11 patients (phenox GmbH, Bochum, Germany), and ReVive SE Thrombectomy Device in 3 patients (Codman & Shurtleff Inc., Raynham, USA). In 26 patients more than one device was used.
3.3. Imaging and clinical outcome 3. Results 3.1. Study population The study population consisted of 58 patients (31 women; mean age 71.0 ± 11.4 years). All patients had clinical signs of acute non-lacunar anterior circulation hemispheric brain ischemia, and proximal vessel occlusion confirmed by CTA. There were 54 patients with an entry NIHSS score ≥ 8 and only 4 patients with an entry NIHSS score < 8. Due to the small number of patients in the second group, no statistically reliable comparison of the two groups from the NIHSS approach could be performed. Optimal PCT CBV ASPECTS threshold level was found to be 6.5. CPM was defined as NIHSS ≥ 8 and PCT CBV ASPECTS ≥ 7. The CPM negative group consists of three subgroups: 1. NIHSS ≥ 8 and PCT CBV ASPECTS < 7, N = 35 2. NIHSS < 8 and PCT CBV ASPECTS < 7, N = 0
Revascularization was achieved in 51 of 58 patients (87.9%). All patients without revascularization (N = 7) had an unfavorable outcome (mRS > 2). In contrast, favorable outcome was found in 17 of 51 patients with successful ERT (33.3%). Table 2 shows the clinical and neuroimaging outcome. The prestroke mRS and the baseline NIHSS score did not significantly differ between the groups. Outcome after 3 months was more unfavorable in the CPM negative group and the major CBV lesion groups. 47.4% of the CPM positive and 52.2% of the minor CBV lesion groups achieved a mRS ≤ 2 after 3 months, equivalent to functional independency. In contrast, only 20.5% and 14.3% of the CPM negative and the major CBV lesion groups reached this outcome level. Table 3 lists sensitivity, specificity, PPV and NPV for all approaches. Sensitivity was highest in the CBV size approach, while specificity, PPV and NPV were comparable for all dichotomization approaches. Figs. 1 and 2 illustrate representative patients from the minor and major CBV lesion groups.
42
M. Saake et al. / Clinical Neurology and Neurosurgery 121 (2014) 39–45
Table 2 Comparisons of stroke scales, outcome, and imaging findings between the paired groups. All patients
Pre-stroke mRS Baseline NIHSS score mRS after 3 months mRS deterioration after 3 months (%) Number of mRS ≤ 2 after 3 months (%) Baseline CT ASPECTS PCT CBV ASPECTS PCT CBF ASPECTS PCT TTP ASPECTS Revascularization (%) CT ASPECTS after 24–72 h
Initial NIHSS ≥ 8
All patients
CBV ≥ 7 N = 23
CBV < 7 N = 35
0.5 ± 0.9 15.0 ± 8.4 3.1 ± 2.0 43.5 52.2 8.6 ± 1.7 7.9 ± 1.1 6.5 ± 2.2 2.7 ± 2.4 87.0 6.2 ± 2.4
0.9 ± 1.2 17.5 ± 2.6 4.2 ± 1.8 62.9 14.3 7.3 ± 2.6 3.0 ± 2.1 1.7 ± 2.0 1.1 ± 1.4 88.6 3.2 ± 3.1
p value 0.17 0.14 0.04 0.18
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Clinical/perfusion CT CBV mismatch as prognostic factor in intraarterial thrombectomy in acute anterior circulation stroke ¨ ¨ Marc Saake ∗ , Lorenz Breuer, Philipp Golitz, Martin Kohrmann, Stefan Schwab, ¨ Arnd Dorfler, Stephan Kloska Department of Neuroradiology, University of Erlangen-Nuremberg, Schwabachanlage 6, 91054 Erlangen, Germany
a r t i c l e
i n f o
Article history: Received 29 October 2013 Received in revised form 23 January 2014 Accepted 6 March 2014 Available online 15 March 2014 Keywords: Stroke Perfusion CT Cerebral blood volume Thrombectomy Mismatch
a b s t r a c t Objective: Contradictory results were reported for the outcome after endovascular recanalization (ERT) in acute anterior circulation ischemic stroke. We assessed whether a clinical/perfusion CT cerebral blood volume (CBV) mismatch concept (CPM) can identify patients who will benefit from reperfusion therapy. Methods: Imaging and clinical data of 58 consecutive ERT cases with acute anterior circulation stroke (ICA, M1, proximal M2) undergoing intraarterial thrombectomy within 4.5 h after symptom onset were analyzed retrospectively. CPM was defined as NIHSS ≥ 8 and PCT CBV ASPECTS ≥ 7. Minor CBV lesion was defined as PCT CBV ASPECTS ≥ 7. Results: All baseline characteristics other than blood glucose did not differ between the paired groups. Revascularization was achieved in 87.9% of all patients without significant difference between the paired groups. Favorable clinical outcome after 3 months (mRS ≤ 2) resulted in 29.3% of all patients, in contrast to 47.4% of the CPM positive and 52.2% of the minor CBV lesion groups. Conclusion: CPM can identify patients who will benefit from reperfusion therapy in acute anterior circulation ischemic stroke. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Stroke is one of the leading causes of morbidity and mortality worldwide and a common cause of depression, epilepsy, and dementia in the elderly. 85% of strokes are ischemic, resulting from an arterial vessel occlusion. Depending on the site of occlusion, mortality rates reach 53–92%, with higher mortality rates in large vessel occlusion. The keystones of acute stroke treatment are fast and accurate detection of brain tissue at risk and time sensitive performance of reperfusion therapy in patients eligible. Beside factors like time from symptom onset, infarct demarcation on noncontrast CT (NCCT) and initial National Institute of Health stroke scale (NIHSS), a beneficial patient selection criterion for reperfusion
Abbreviations: IVT, intravenous thrombolysis; ERT, endovascular reperfusion therapy; NCCT, non-contrast CT; PCT, perfusion CT; CBV, cerebral blood volume; CBF, cerebral blood flow; DWI, diffusion weighted imaging; NIHSS, National Institute of Health Stroke Scale; MMCT, multimodal CT; mRS, modified Rankin Scale; CTA, CT angiography; TTP, time to peak; DSA, digital subtraction angiography; ASPECTS, Alberta Stroke Program Early CT score; MCA, middle cerebral artery; ROC, receiver operating characteristics; PPV, positive predictive value; NPV, negative predictive value; CPM, clinical/PCT CBV mismatch. ∗ Corresponding author. Tel.: +49 91318539388. E-mail address: [email protected] (M. Saake). http://dx.doi.org/10.1016/j.clineuro.2014.03.007 0303-8467/© 2014 Elsevier B.V. All rights reserved.
therapy might be the size of the mismatch between the potentially salvageable tissue (penumbra) and the irreversibly damaged tissue (core) [1,2]. Reperfusion therapy is the only proven causal treatment for acute ischemic stroke. Currently intravenous thrombolysis (IVT) is the only FDA-approved reperfusion strategy. However, IVT has several limitations such as a limited time window and a weak effect in ischemic strokes caused by large vessel occlusions. In these cases, interventional procedures might improve recanalization rates and patient outcome [3]. Significant improvements have been made in the last years in the development of endovascular reperfusion therapies (ERT) [4]. By now, ERT has been demonstrated to be an effective means of achieving reperfusion in stroke patients with proximal cerebral artery occlusions [5]. However, favorable outcome after stroke intervention remains limited, with only 25–50% of patients achieving a modified Rankin Scale ≤ 2 at 90 days [6,7]. Moreover, recent reports suggested that outcome after ERT might be equal to IVT [8,9]. Combined clinical/imaging mismatch concepts compare the clinical symptoms to the imaging findings. Numerous studies have shown that the clinical/MRI diffusion mismatch between the NIHSS score and the infarct size on diffusion weighted imaging (DWI) may identify patients at risk of infarct growth and early neurological deterioration [10–12]. However, CT represents the most
40
M. Saake et al. / Clinical Neurology and Neurosurgery 121 (2014) 39–45
frequently used imaging modality in acute stroke patients. Previous studies have reported poor sensitivity and poor interobserver agreement for the detection of early ischemic changes on NCCT [13]. Several mismatch concepts have been proposed to improve CT stroke detection. A clinical/NCCT approach did not reliably identify patients more or less likely to benefit from intravenous thrombolysis [14]. Perfusion CT (PCT) has been shown to be an accurate predictor of both thrombolytic therapy benefit and clinical outcome [15]. The most frequently used PCT-based approach compares the region of reduced cerebral blood volume (CBV, presumably representing irreversible infarction) and the area of reduced cerebral blood flow (CBF) that lies outside of the CBV abnormality (potentially representing reversible ischemia). The region of reduced CBV has been shown to most accurately describe the infarct core [16] and to correlate with MRI DWI lesions in hyperacute stroke [17]. We hypothesized that clinical/PCT CBV mismatch (CPM) between the patient’s clinical stroke severity defined by the NIHSS score and the PCT CBV lesion size can identify patients who will benefit from reperfusion therapy in acute anterior circulation ischemic stroke.
were: section thickness, 5.0 mm; increment, 5.0 mm; reconstruction kernel, H31s. Primary PCT image reconstruction parameters were: section thickness, 5.0 mm; increment, 5.0 mm; reconstruction kernel, H20f. From this dataset PCT parameter maps (CBV, CBF, and time to peak (TTP)) were calculated using a CT workstation (Multimodality Workplace MMWP; Siemens, Forchheim, Germany) running standard software (Syngo VPCT Neuro; Siemens, Forchheim, Germany). All digital subtraction angiographies (DSA) were performed on a biplane flat-detector angiography system (Axiom Artis dBA; Siemens, Forchheim, Germany) by staff neuroradiologists. Using standard angiographic procedures, the affected internal carotid artery was investigated. For ERT procedures the device vendor guidelines were followed. The selection of the used reperfusion catheter system was to the discretion of the neurointerventionalist. Generally, all ERT systems aim to remove the clot from the cerebral artery, but they differ in technical details. NCCT was repeated at 24–72 h after symptom onset for followup. 2.4. Image analysis
2. Materials and methods 2.1. Study population Institutional review board approval was obtained before the commencement of this retrospective study. Informed consent of the patient or next of kin was obtained according to local laws and regulations. A total of 277 consecutive patients who underwent multimodal stroke CT (MMCT) in our department between June 2010 and July 2012 were analyzed retrospectively for the study. Inclusion criteria were acute anterior circulation stroke, use of MMCT for primary imaging, proximal anterior circulation vessel occlusion (i.e. occlusion of the internal carotid artery, M1, and/or proximal M2 segment of the middle cerebral artery) and performance of ERT due to clinical indication. Exclusion criteria were an age of less than 18 years, incomplete or non-diagnostic examinations, and signs of hemorrhage in the initial NCCT. Additional IVT was not considered an exclusion criterion, as IVT is the primary standard therapy in acute stroke. 2.2. Stroke therapy and follow-up All stroke patients admitted to our hospital are treated at our local stroke unit. According to standard procedures the severity of clinical stroke symptoms is assessed in every patient by the National Institutes of Health Stroke Scale (NIHSS) [18]. All eligible stroke patients receive body weight adapted (0.9 mg tPA/kg) IVT according to standard guidelines [19]. As outcome measure we used the modified Rankin Scale (mRS), a 7-point scoring system that distributes patients from 0 (no symptoms) to 6 (death) [20]. The pre-stroke mRS score was determined by a structured questionnaire during the hospital stay of the patient. Three months after stroke onset, the structured questionnaire was performed again for all patients by telephone to assess the level of independence at home. 2.3. Imaging and endovascular recanalization therapy (ERT) In all acute stroke patients a standardized MMCT protocol, consisting of NCCT, PCT with 9.6 cm whole brain coverage in the z-axis, and supra-aortic CT angiography (CTA) was performed according to hospital guidelines. All examinations were performed on a 128-section CT scanner (Somatom Definition AS+; Siemens, Forchheim, Germany). Image reconstruction parameters for NCCT
All examinations were anonymized. Images were transferred to a workstation (MMWP, Siemens, Forchheim, Germany). Review and analysis of the MMCT and ERT images was performed by a senior neurology resident and a senior neuroradiology resident in consensus. For assessment of the NCCT images and the color-coded PCT parameter maps the Alberta Stroke Program Early CT Score (ASPECTS) was used. The ASPECTS is a semiquantitative approach to infarct sizing on NCCT that has been demonstrated to be both reliable and predictive of outcome following acute stroke treatment [21]. The ASPECTS ranges from 0 to 10. For signs of infarction in each of the 10 parts of the middle cerebral artery (MCA) territory one point is subtracted from 10. Thus, a minor infarction has a high, a major infarction a low ASPECTS. Meanwhile, the ASPECTS has been adapted to the similar assessment of PCT parameter maps [22]. ERT images were scored for recanalization success, using the primary arterial occlusive lesion (AOL) [23] and thrombolysis in cerebral infarction (TICI) [24] scores. 2.5. Data analysis It has been suggested that an NIHSS score of ≥8 is associated with a high rate of neurological deterioration, a low frequency of spontaneous functional recovery, and cortical perfusion deficits [11,25]. So, we used this cut-off point as a clinical indicator of a large volume of ischemic brain tissue. Patients were dichotomized for favorable versus unfavorable outcome at 90 days, with favorable outcome defined as mRS ≤ 2. Receiver operating characteristics (ROC) analysis was performed for PCT CBV ASPECTS in regard of patient outcome. The optimal threshold level for dichotomization of PCT CBV ASPECTS was determined by maximizing the sum of sensitivity and specificity, and minimizing the Euclidean distance of (sensitivity, specificity) to the point (1, 1) of the ROC curve [26]. We divided the study population into the following groups: 1. minor stroke (NIHSS < 8) versus major stroke (NIHSS ≥ 8) groups (NIHSS approach), 2. minor CBV lesion (PCT CBV ASPECT score > threshold) versus major CBV lesion (PCT CBV ASPECT score < threshold) groups (CBV size approach), 3. clinical/PCT CBV matched (CPM negative) versus clinical/PCT CBV mismatched (CPM positive) groups (CPM approach). We compared the paired groups with regard to the following variables: background characteristics including age, gender,
M. Saake et al. / Clinical Neurology and Neurosurgery 121 (2014) 39–45
41
Table 1 Background characteristics, risk factors, vital signs and stroke subtype in the paired groups. All patients
Age, years Gender, female:male Time from onset to CT (h) Hypertension (%) Diabetes mellitus (%) Stroke in history (%) Atrial fibrillation (%) Antithrombotic therapy (%) Systolic blood pressure (mmHg) Diastolic blood pressure, mmHg Body temperature (◦ C) Blood glucose (mg/dL) Lesion side, left:right Stroke subtype classification
Initial NIHSS ≥ 8
All patients
CBV ≥ 7 N = 23
CBV < 7 N = 35
p value
CPM positive N = 19
CPM negative N = 39
p value
CBV ≥ 7 N = 19
CBV < 7 N = 35
p value
69.5 ± 11.0 12:11 2.1 ± 1.3 94.3 30.4 21.7 47.8 34.8 160.2 ± 25.5 87.7 ± 23.3 36.3 ± 0.6 118.7 ± 29.9 13:10 7:7:5:0:4
72.0 ± 11.8 19:16 2.4 ± 1.7 100 28.6 28.6 60.0 42.9 159.8 ± 29.8 84.7 ± 17.6 36.2 ± 0.6 139.9 ± 41.4 19:16 12:17:2:0:4
0.42 1.00 0.46 0.51 1.00 0.76 0.43 0.59 0.96 0.60 0.88 0.03 1.00 0.21
69.5 ± 11.5 11:8 2.1 ± 1.1 100 36.8 15.8 47.4 31.6 157.0 ± 25.9 85.4 ± 22.0 36.3 ± 0.6 116.6 ± 28.8 10:9 6:7:3:0:3
71.7 ± 11.5 20:19 2.3 ± 1.7 94.9 25.6 30.8 59.0 43.6 161.4 ± 29.1 86.1 ± 19.1 36.2 ± 0.6 138.8 ± 40.7 22:17 13:17:4:0:5
0.50 0.78 0.52 0.55 0.54 0.34 0.57 0.41 0.56 0.91 0.48 0.02 1.00 0.91
69.5 ± 11.5 11:8 2.1 ± 1.1 100 36.8 15.8 47.4 31.6 157.0 ± 25.9 85.4 ± 22.0 36.3 ± 0.6 116.6 ± 28.8 10:9 6:7:3:0:3
72.0 ± 11.8 19:16 2.1 ± 1.3 94.3 28.6 28.6 60.0 42.9 159.8 ± 29.8 84.7 ± 17.6 36.2 ± 0.6 139.9 ± 41.4 19:16 12:17:2:0:4
0.46 1.00 0.92 0.54 0.55 0.34 0.40 0.56 0.72 0.90 0.61 0.02 1.00 0.27
Bold text indicates a statistically significant difference with p ≤ 0.05.
time from onset of symptoms to CT, hypertension (past use of antihypertensive agents or blood pressure of >140/90 mmHg at least twice before onset), diabetes mellitus (use of insulin or oral hypoglycemic agents, fasting blood glucose ≥ 126 mg/dL or random blood glucose ≥ 200 mg/dL), stroke in patient history, atrial fibrillation in recent patient history, antithrombotic therapy before onset; vital signs including systolic blood pressure, diastolic blood pressure, body temperature, and blood glucose at entry; side of vessel lesion; ischemic stroke subtype classified according to the TOAST criteria (large-artery atherosclerosis, including large-artery thrombosis and artery-to-artery embolism; cardioembolism; small-artery occlusion; stroke of other determined cause; and stroke of undetermined cause) [27]; pre-stroke mRS; NIHSS score at entry; outcome after 3 months measured by the mRS; mRS deterioration defined by an increase of ≥3 points in mRS after 3 months compared to pre-stroke mRS; number of patients with mRS ≤ 2 after 3 months; baseline NCCT ASPECTS; PCT CBV/CBF/TTP ASPECTS; revascularization of the target vessel defined as an AOL score of II or III and a TICI score of 2a or better; and ASPECTS in follow-up NCCT 24–72 h after onset. For all statistical analysis SPPS 11 (SPSS, Chicago, IL, USA) was used. Student’s t-test or Mann–Whitney test was used in univariate analysis for continuous and interval variables, and Fisher’s exact test for non-continuous variables. A level of p ≤ 0.05 was regarded as statistically significant. Sensitivity, specificity, as well as positive predictive values (PPV) and negative predictive values (NPV) were calculated.
3. NIHSS < 8 and PCT CBV ASPECTS ≥ 7, N = 4 All of the 4 patients with an entry NIHSS score < 8 had a PCT CBV ASPECTS of ≥7. Good outcome was achieved in 3 of these 4 patients (75%). An additional statistical analysis was performed without these 4 patients. Due to the exclusion of patients with a NIHSS < 8, the CBV and CPM approaches were identical for this subgroup analysis. Table 1 describes the background characteristics, risk factors, vital signs, and stroke subtypes in the groups. The only statistically significant difference between the paired groups was blood glucose concentration at admission, which was higher in the CPM negative and the major CBV lesion groups, respectively.
3.2. Stroke therapy Bridging IVT was performed in all patients. In 19 patients thrombolysis was switched from intravenous to intraarterial. The used ERT systems were: Penumbra System in 37 patients (Penumbra Inc., Alameda, USA), Solitaire FR Revascularization Device in 9 patients (ev3 Neurovascular, Irvine, USA), pRESET Thrombectomy Retriever in 11 patients (phenox GmbH, Bochum, Germany), and ReVive SE Thrombectomy Device in 3 patients (Codman & Shurtleff Inc., Raynham, USA). In 26 patients more than one device was used.
3.3. Imaging and clinical outcome 3. Results 3.1. Study population The study population consisted of 58 patients (31 women; mean age 71.0 ± 11.4 years). All patients had clinical signs of acute non-lacunar anterior circulation hemispheric brain ischemia, and proximal vessel occlusion confirmed by CTA. There were 54 patients with an entry NIHSS score ≥ 8 and only 4 patients with an entry NIHSS score < 8. Due to the small number of patients in the second group, no statistically reliable comparison of the two groups from the NIHSS approach could be performed. Optimal PCT CBV ASPECTS threshold level was found to be 6.5. CPM was defined as NIHSS ≥ 8 and PCT CBV ASPECTS ≥ 7. The CPM negative group consists of three subgroups: 1. NIHSS ≥ 8 and PCT CBV ASPECTS < 7, N = 35 2. NIHSS < 8 and PCT CBV ASPECTS < 7, N = 0
Revascularization was achieved in 51 of 58 patients (87.9%). All patients without revascularization (N = 7) had an unfavorable outcome (mRS > 2). In contrast, favorable outcome was found in 17 of 51 patients with successful ERT (33.3%). Table 2 shows the clinical and neuroimaging outcome. The prestroke mRS and the baseline NIHSS score did not significantly differ between the groups. Outcome after 3 months was more unfavorable in the CPM negative group and the major CBV lesion groups. 47.4% of the CPM positive and 52.2% of the minor CBV lesion groups achieved a mRS ≤ 2 after 3 months, equivalent to functional independency. In contrast, only 20.5% and 14.3% of the CPM negative and the major CBV lesion groups reached this outcome level. Table 3 lists sensitivity, specificity, PPV and NPV for all approaches. Sensitivity was highest in the CBV size approach, while specificity, PPV and NPV were comparable for all dichotomization approaches. Figs. 1 and 2 illustrate representative patients from the minor and major CBV lesion groups.
42
M. Saake et al. / Clinical Neurology and Neurosurgery 121 (2014) 39–45
Table 2 Comparisons of stroke scales, outcome, and imaging findings between the paired groups. All patients
Pre-stroke mRS Baseline NIHSS score mRS after 3 months mRS deterioration after 3 months (%) Number of mRS ≤ 2 after 3 months (%) Baseline CT ASPECTS PCT CBV ASPECTS PCT CBF ASPECTS PCT TTP ASPECTS Revascularization (%) CT ASPECTS after 24–72 h
Initial NIHSS ≥ 8
All patients
CBV ≥ 7 N = 23
CBV < 7 N = 35
0.5 ± 0.9 15.0 ± 8.4 3.1 ± 2.0 43.5 52.2 8.6 ± 1.7 7.9 ± 1.1 6.5 ± 2.2 2.7 ± 2.4 87.0 6.2 ± 2.4
0.9 ± 1.2 17.5 ± 2.6 4.2 ± 1.8 62.9 14.3 7.3 ± 2.6 3.0 ± 2.1 1.7 ± 2.0 1.1 ± 1.4 88.6 3.2 ± 3.1
p value 0.17 0.14 0.04 0.18
