SEARCH

SEARCH BY CITATION

Keywords:

  • computed tomography;
  • CT angiography;
  • predictive model;
  • recanalization;
  • stroke

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References
  8. Supporting Information

Background and purpose

The study aims to assess the recanalization rate in acute ischemic stroke patients who received no revascularization therapy, intravenous thrombolysis, and endovascular treatment, respectively, and to identify best clinical and imaging predictors of recanalization in each treatment group.

Methods

Clinical and imaging data were collected in 103 patients with acute ischemic stroke caused by anterior circulation arterial occlusion. We recorded demographics and vascular risk factors. We reviewed the noncontrast head computed tomographies to assess for hyperdense middle cerebral artery and its computed tomography density. We reviewed the computed tomography angiograms and the raw images to determine the site and degree of arterial occlusion, collateral score, clot burden score, and the density of the clot. Recanalization status was assessed on recanalization imaging using Thrombolysis in Myocardial Ischemia. Multivariate logistic regressions were utilized to determine the best predictors of outcome in each treatment group.

Results

Among the 103 study patients, 43 (42%) received intravenous thrombolysis, 34 (33%) received endovascular thrombolysis, and 26 (25%) did not receive any revascularization therapy. In the patients with intravenous thrombolysis or no revascularization therapy, recanalization of the vessel was more likely with intravenous thrombolysis (P = 0·046) and when M1/A1 was occluded (P = 0·001). In this subgroup of patients, clot burden score, cervical degree of stenosis (North American Symptomatic Carotid Endarterectomy Trial), and hyperlipidemia status added information to the aforementioned likelihood of recanalization at the patient level (P < 0·001). In patients with endovascular thrombolysis, recanalization of the vessel was more likely in the case of a higher computed tomography angiogram clot density (P = 0·012), and in this subgroup of patients gender added information to the likelihood of recanalization at the patient level (P = 0·044).

Conclusion

The overall likelihood of recanalization was the highest in the endovascular group, and higher for intravenous thrombolysis compared with no revascularization therapy. However, our statistical models of recanalization for each individual patient indicate significant variability between treatment options, suggesting the need to include this prediction in the personalized treatment selection.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References
  8. Supporting Information

Previous studies have demonstrated that penumbral information extracted from imaging in patients with acute ischemic stroke symptoms requires recanalization status to be known so that it can be interpreted appropriately. Indeed, a large penumbra can hallmark a favorable outcome in the setting of early reperfusion, while it can result in a large final infarct and a negative outcome in the absence of early reperfusion [1]. This observation is in agreement with the results of the DEFUSE, DEFUSE 2, and EPITHET trials [2-4]. Upon admission of the patient in the emergency room, imaging can be obtained to determine the presence (or absence) and extent of arterial occlusion, ischemic core, and penumbra. However, recanalization status is not known at this time, and thus needs to be predicted so that the penumbral information can be integrated in the treatment decision making.

A number of studies have attempted to predict recanalization. Some reported clinical predictors, such as time to treatment, National Institutes of Health Stroke Scale, and vascular risk factors [5-8]. Other studies assessed imaging predictors, including M1 susceptibility vessel sign on T2*, density and length of thrombus on admission noncontrast head computed tomography (NCT), perfusion-weighted imaging-derived collateral flow index, etc [9-13]. Most of these studies have focused on acute ischemic stroke patients treated with intravenous (IV) tissue plasminogen activator (tPA) thombolysis, and some on patients who received endovascular therapy [11]. There are a few studies assessing the rate of spontaneous recanalization in untreated patients.

The goal of this study was to assess the recanalization rate in acute ischemic stroke patients who received no revascularization therapy, IV thrombolysis, and endovascular treatment, respectively, and to evaluate reported clinical and imaging predictors to determine the best biomarkers of recanalization in each treatment group.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References
  8. Supporting Information

Study patients

The data presented in this study belong to a previously reported repository created from the data collected as part of the standard clinical stroke care at three participating institutions, including the University of Virginia Health System (Charlottesville, VA), the Centre Hospitalier Universitaire Vaudois (Lausanne, Switzerland), and the University of Pittsburgh Medical Center (Pittsburgh, PA) [1, 14]. Only completely anonymous data, stripped of HIPAA information, were contributed to the repository, with the approval of the respective institutional review boards of the three contributing institutions. At these institutions, patients with symptoms suggestive of acute stroke undergo a stroke CT imaging work-up which includes an NCT and CT angiogram (CTA) of the cervical and intracranial arteries. Treatment protocols used at all institutions are similar.

We retrospectively identified all consecutive patients admitted to these institutions with signs and symptoms suggesting hemispheric stroke, who fulfilled the following inclusion criteria: (1) acute ischemic stroke with internal carotid artery (ICA), M1, M2, A1, or/and A2 occlusion (one or more sites allowed) confirmed by CTA, and without evidence of intracranial hemorrhage; (2) completion of a stroke CT protocol including NCT and CTA upon admission within 12 h of symptom onset; (3) completion of recanalization imaging between 1 and 48 h, consisting of either CTA, MR angiography (MRA), and/or digital subtraction angiography (DSA).

The following demographic and clinical variables were recorded: age, gender, type of revascularization therapy: none, IV tPA, or endovascular therapy (if any), time from symptom onset to baseline imaging, time from baseline imaging to revascularization therapy (if any), time from baseline imaging to recanalization imaging, vascular risk factors (hypertension, diabetes mellitus, hyperlipidemia, atrial fibrillation, coronary heart disease), TOAST classification of stroke mechanism, blood glucose upon admission, and systolic and diastolic blood pressure upon admission. We excluded those patients who received both IV and intra-arterial (IA) therapies because we wanted to evaluate the recanalization rate and its predictors in each treatment category, and wanted to avoid interactions between different types of treatments. Of note, the literature reports the rate of recanalization with combined IV/IA therapies to be similar to that of IA alone, suggesting the IA component of combined therapy to be the determinant of the final recanalization rate [15].

CTA image acquisition

The CTA studies of the cervical and intracranial arteries were obtained on 64-slice CT scanners (GE Healthcare, Milwaukee, WI, USA), using the following image acquisition protocol: helical acquisition, 0·5–0·8 s gantry rotation, pitch: 1–1.375:1, slice thickness: 0·625–1·25 mm, reconstruction interval: 0·5–1 mm, and acquisition parameters: 120 kVp/200–300 mAs. A caudocranial scanning direction was selected covering the mid-chest to the vertex of the brain. Raw data and reconstructed CTA images were available for review.

Image processing and interpretation

Images were transferred to a Mac Pro workstation (Apple, Cupertino, CA, USA). A neuroradiologist (M. W.) with 14 years of experience and a neurologist (G. Z.) with 12 years of experience reviewed the NCT and CTA images using OsiriX v 5·6 (Pixmeo, Geneva, Switzerland). Both were blinded to the initial clinical interpretation of these studies, to the type of treatment received by each patient, and to the success of this treatment, or lack thereof, in terms of recanalization. They were provided only with the body side of the signs and symptoms.

The presence or absence of a hyperdense middle cerebral artery sign (HMCAS) was assessed on NCT. The length and density of HMCAS were measured and recorded. The NCT relative density of HMCAS was measured as the ratio between the absolute density of HMCAS and the density of the ipsilateral vitreous. The ASPECTS score on the NCT was also recorded [16].

The CTA images were used to evaluate the site and degree of arterial occlusion and calculate the clot burden score, to characterize the collateral circulation, to assess the degree of cervical internal carotid artery stenosis and to calculate the density of the ICA and/or A1/M1 clots on the raw CTA images. The site of occlusion was recorded as ICA, M1, A1, and/or M2/A2. The degree of arterial occlusion was scored with using a modified Thrombolysis in Myocardial Ischemia (TIMI) scheme: 0 = complete occlusion, 1 = severe luminal narrowing, 2 = mild or moderate luminal narrowing, and 3 = patent artery [17]. Recanalization imaging, which consisted either of CTA, MRA, or DSA, was reviewed for a second TIMI measurement. Recanalization was defined as an improvement in the TIMI score from baseline to recanalization of ≥2 points [1, 14, 18]. For patients who underwent DSA, TICI 2b and 3 were used to define recanalization [19].

The clot burden score (CBS) is a scoring system to characterize the extent of thrombus found in the proximal anterior circulation. It ranges from 0 to 10. A 10-point score indicates the absence of thrombus. Two points are subtracted for thrombus found on CTA in the supraclinoid ICA and each of the proximal and distal halves of the MCA trunk. One point is subtracted for thrombus found in the infraclinoid ICA, A1 segment, and for each affected M2 branch [20].

Absence of collateral flow to the ischemic territory was graded as 0, whereas collateral flow in <50%, >50%, and 100% of the vessels filling in the ischemic territory was graded as 1, 2, or 3 respectively [21, 22]. The North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria were used to quantify the degree of cervical carotid arterial stenosis [23].

The density of the clot in the ICA and/or M1/A1 was assessed on raw, thin-slice CTA images. CTA relative clot density was measured as the ratio between the absolute density of the clot and the density of the ipsilateral vitreous.

The BASIS classification methodology was used to classify the major and minor strokes [24]. Patients were classified as having a major stroke if a proximal cerebral artery occlusion was identified, or when a large infarct was identified in a patient with patent major cerebral arteries, or the presence of parenchymal abnormalities was identified with noncontrast CT or diffusion MR imaging. Patients not meeting these criteria were classified as having a minor stroke.

Statistical analyses

Descriptive statistics

Continuous scaled variables (e.g. age, absolute density, length of HMCAS) were summarized by the mean and the standard deviation of the measurement distribution. Categorical data (e.g. gender) were summarized as frequencies and percentages.

Multivariate modeling overview

We used a two-step modeling process to develop multivariate statistical models to predict recanalization at the patient level. In step 1 of the modeling process, we used information extracted from NCT and CTA imaging to predict recanalization at the vessel level; and in step 2 of the modeling process, we used the results from step 1 of the modeling process, along with patient-specific clinical and imaging information to predict recanalization at the patient level. The details of the modeling process are provided in their entirety in Appendix S1.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References
  8. Supporting Information

Study patients

We identified 103 patients (female gender 48·5%, 50/103) who fulfilled our inclusion criteria. The mean patients' age was 64·3 ± 11·9 years; range 32–86 years. Other clinical characteristics, as well as the values of NCT and CTA variable,s are summarized in Tables 1-3. Forty-three (42%) patients received IV thrombolysis, 34 (33%) received endovascular thrombolysis, and 26 (25%) did not receive any revascularization therapy (outside time window, wake-up strokes, and contraindication to revascularization therapy). The modified Rankin scores (mRS) at 3 months are reported in Table 1.

Table 1. Demographic and clinical variables
VariableValueP valuesa
AllNo revascularization treatment (n = 26)IV thrombolysis (n = 43)Endovascular therapy (n = 34)
  1. a

    Kruskal–Wallis tests were used for continuous variables while chi-square tests were used for categorical variables.

  2. b

    n = 95.

  3. c

    n = 26.

  4. BASIS, Boston Acute Stroke Imaging Scale; IV, intravenous; mRS, modified Rankin score; NIHSS, National Institutes of Health Stroke Scale; TOAST, Trial of Org 10172 in Acute Stroke Treatment.

Age (years)64·3 ± 11·960·5 ± 11·264·0 ± 10·867·5 ± 13·20·038
Female48·5% (50/103)57·7% (15/26)41·9% (18/43)50·0% (17/34)0·434
Hypertension (=Yes)44·7% (46/103)42·3% (11/26)27·9% (12/43)67·7% (23/34)0·002
Diabetes mellitus (=Yes)21·4% (22/103)11·5% (3/26)18·6% (8/43)32·4% (11/34)0·127
Atrial fibrillation (=Yes)52·4% (54/103)65·4% (17/26)62·8% (27/43)29·4% (10/34)0·005
Hyperlipidemia (=Yes)52·4% (54/103)61·5% (16/26)65·1% (28/43)29·4% (10/34)0·004
Admission systolic blood pressure (mmHg)157·3 ± 30·4b158·7 ± 32·0163·0 ± 31·5146·4 ± 24·5c0·233
Admission diastolic blood pressure (mmHg)86·5 ± 20·6b89·2 ± 21·992·5 ± 20·274·0 ± 14·2c<0·001
Admission blood glucose (mmol/L)8·0 ± 3·46·9 ± 1·57·5 ± 2·39·5 ± 4·90·039
Time from symptom onset to baseline imaging (h)5·4 ± 4·56·4 ± 5·33·0 ± 1·86·6 ± 2·9<0·001
Time from baseline imaging to revascularization therapy – if any (h)   2·2 ± 1·8
Time from baseline imaging to recanalization imaging (h)21·8 ± 6·334·6 ± 6·129·4 ± 5·42·2 ± 1·8<0·001
NIHSS scores16·4 ± 5·215·4 ± 5·017·1 ± 5·916·2 ± 4·40·341
TOASTLarge artery atherosclerosis37·7% (29/77)31·3% (5/16)51·6% (16/31)26·7% (8/30)0·015
Cardioembolic stroke33·8% (26/77)31·3% (5/16)25·8% (8/31)43·3% (13/30)
Small vessel0% (0/77)0% (0/16)0% (0/31)0% (0/30)
Other determined11·7% (9/77)25·0% (4/16)16·1% (5/31)0% (0/30)
Undetermined16·9% (13/77)12·5% (2/16)6·5% (2/31)30·0% (9/30)
BASISMajor stroke100% (103/103)100% (26/26)100% (43/43)100% (34/34)
Minor stroke0% (0/103)0% (0/26)0% (0/43)0% (0/34)
mRS at 3 months≤247·6% (49/103)46·2% (12/26)39·5% (17/43)58·8% (20/34)0·189
>252·4 (54/103)53·8% (14/26)60·5% (26/43)41·2% (14/34)
Table 2. Imaging variables
VariableValueP valuesa
AllNo revascularization treatment (n = 26)IV thrombolysis (n = 43)Endovascular therapy (n = 34)
  1. a

    Kruskal–Wallis tests were used for continuous variables while chi-square tests were used for categorical variables.

  2. CBS, clot burden score; CTA, computed tomography (CT) angiogram; HMCAS, hyperdense middle cerebral artery sign; ICA, internal carotid artery; IV, intravenous; NASCET, North American Symptomatic Carotid Endarterectomy Trial; NCT, noncontrast head computed tomography (CT); TIMI, Thrombolysis in Myocardial Ischemia.

Admission NCTHMCAS (=Yes)70·9% (73/103)69·2% (18/26)76·7% (33/43)64·7% (22/34)0·502
Length (mm)1·2 ± 1·11·2 ± 1·01·3 ± 1·10·9 ± 1·10·279
Relative clot density (%)6·2 ± 4·45·5 ± 3·45·3 ± 3·85·8 ± 4·80·358
Admission CTACollateral flow score=08·8% (9/103)11·5% (3/26)11·6% (5/43)2·9% (1/34)0·007
=121·4% (22/103)15·4% (4/26)30·2% (13/43)14·7% (5/34)
=231·1% (32/103)42·3% (11/26)25·6% (11/43)29·4% (10/34)
=330·1% (31/103)23·1% (6/26)16·3% (7/43)52·9% (18/34)
=48·7% (9/103)7·7% (2/26)16·3% (7/43)0% (0/34)
Site of occlusionICA5·8% (6/103)11·5% (3/26)4·7% (2/43)2·9% (1/34)0·444
M1/A157·3% (59/103)57·7% (15/26)51·2% (22/43)64·7% (22/34)
Both ICA and M1/A136·9% (38/103)30·8% (8/26)44·2% (19/43)32·4% (11/34)
CBSMean4·9 ± 2·25·0 ± 2·44·2 ± 2·45·8 ± 1·50·007
≤679·6% (82/103)76·9% (20/26)83·7% (36/43)76·5% (26/34)
>620·4% (21/103)23·1% (6/26)16·3% (7/43)23·5% (8/34)
Baseline TIMI089·3% (92/103)92·3% (24/26)81·4% (35/43)97·1% (33/34)0·231
110·7% (11/103)7·7% (2/26)18·6% (8/43)2·9% (1/34)
20000
30000
NASCET (%)47·2 ± 47·749·6 ± 49·055·7 ± 48·234·5 ± 44·60·112
Relative clot density (%)6·5 ± 4·35·0 ± 3·86·6 ± 4·45·7 ± 3·40·935
Table 3. Recanalization status
 AllNo revascularization treatment (n = 26)IV thrombolysis (n = 43)Endovascular therapy (n = 34)P valuesa
  1. a

    Kruskal–Wallis tests were used for continuous variables while chi-square tests were used for categorical variables.

  2. IV, intravenous; TIMI, Thrombolysis in Myocardial Ischemia.

Recanalization imagingImprovement of TIMI036·9% (38/103)57·7% (15/26)39·5% (17/43)17·7% (6/34)0·030
17·8% (8/103)7·7% (2/26)2·3% (1/43)14·7% (5/34)
236·9% (38/103)19·2% (5/26)41·9% (18/43)44·1% (15/34)
318·5% (19/103)15·4% (4/26)16·3% (7/43)23·5% (8/34)
Recanalization (=Yes)55·3% (57/103)34·6% (9/26)58·1% (25/43)67·6% (23/34)0·034

Patients with no revascularization and patients with IV thrombolysis

At the vessel level, 111 arteries were assessed and 43·2% (48/111) showed recanalization. Of the five predicting variables that were included in the vessel-level recanalization multivariate model (Table S1a), only IV thrombolysis (P = 0·046) and site of arterial occlusion (P = 0·001) were unique predictors for vessel recanalization. Recanalization was more likely with IV thrombolysis [adjusted OR: 2·57, 95% CI (1·01, 6·32)] and when M1/A1 was occluded [adjusted OR: 5·45, 95% CI (1·96, 15·12)]. The model C-statistic for the reduced regression model was 0·71 [95% CI (0·65, 0·79)] (Table S1b). This reduced model was utilized to derive the vessel recanalization index values (i.e. VRI) values that were used in the patient-level recanalization analyses.

At the patient level, 69 patients were evaluated and 49·3% (34/69) showed recanalization. In Table S2a, we list the patient-specific clinical and imaging predictor variable rankings associated with the added variable selection. In addition to the VRI, NASCET, hyperlipidemia, and the CBS were the predictor variables that were selected as the best set of predictor variables for predicting patient-level recanalization (P < 0·001) (Table S3a). Recanalization was more likely in patients who had a lower degree of stenosis (NASCET) of cervical arteries, had a high CBS (restricted extent of clot), had a high VRI, and had hyperlipidemia. Of note, the time to baseline imaging information and the time to the recanalization imaging information did not add significant prediction information to the prediction model either individually (P = 0·590 and P = 0.121, respectively) or together (P = 0·260).

A receiver operator curve (ROC) analysis showed the prediction model in Table 4 to have a ROC equal to 81%. Our model predicts a percentage of likelihood of recanalization. Using an optimum probability cut-point value of 60% to predict presence vs. absence of recanalization, the diagnostic sensitivity of the model's predictions was 60% for the IV thrombolysis patient group, and 75% for the no treatment patient group, while the diagnostic specificity of the model's predictions was 100% for the IV thrombolysis patient group and 89% for the no treatment patient group (Table 4).

Table 4.  Diagnostic classification summaries based on the LR model-predicted probabilities for patient-level recanalization
Study groupOptimum probability cut-pointaDiagnostic sensitivity (%)Diagnostic Specificity (%)PPV (%)NPV (%)Overallb agreement (%)
  1. a

    The patient-level recanalization probability cut-point that minimized the overall misclassification error rate (i.e. minimized the false positive and false negative errors).

  2. b

    Percentage of model-based patient-level recanalization classifications that agreed with the patient recanalization classification, when patient recanalization required a ≥ 2 point improvement in all of the patient's vessel TIMI scores from baseline to recanalization imaging.

  3. PPV denotes the positive predictive value of the diagnostic classification.

  4. NPV denotes the negative predictive value of the diagnostic classification.

  5. IA, intra-arterial; IV, intravenous; TIMI, Thrombolysis in Myocardial Ischemia.

IV60%601001007481
No treatment7589758984
IA60%8373866779

Patients with endovascular revascularization

At the vessel level, 53 arteries were assessed and 75·7% (40/53) showed recanalization. Of the four predicting variables that were included in the vessel level recanalization multivariate model (Table S1c), only CTA relative clot density (P = 0·013) was a unique predictor for vessel recanalization. Recanalization was more likely if the CTA relative clot density was high [adjusted OR: 2·03, 95% CI (1·16, 3·54)]. The model C-statistic for the reduced regression (Table S1d) model was 0·75 [95% CI (0·66, 0·85)]. This reduced model was utilized to derive the VRI values that were used in the patient-level recanalization analyses.

At the patient level, 34 patients were evaluated and 67·6% (23/34) showed recanalization. In Table S2b, we list the patient-specific clinical and imaging predictor variable rankings associated with the added variable selection. Gender was the only variable that was selected to be included in the prediction model with the VRI to predict patient-level recanalization (P = 0·044) (Table S3b). Recanalization was more likely if the patient was male and if the patient's VRI was high. Again, note that the time to baseline imaging information and the time to the recanalization imaging information did not add significant prediction information to the LR prediction model, either individually (P = 0·164 and P = 0·160, respectively) or together (P = 0·161).

A ROC analysis showed the prediction model in Table 4 to have a ROC equal to 79%. At the optimum probability cut-point value of 60%, the diagnostic sensitivity of the model's predictions was 83%, while the diagnostic specificity of the model's predictions was 73% (Table 4).

Theoretical response of patients to different treatment types

As part of a final exploratory step, the predictive models calculated for each treatment category (no treatment, IV thrombolysis, endovascular therapy) were applied to each study patient, independently of the actual treatment category they belonged to (Fig. 1a–c). These calculations can be performed on an iPhone/iPad app, ‘iStrokeMD’ (Fig. 2), available on the Apple Store. From these predictions (Fig. 1), one can observe that the probability of recanalization varies quite widely from patient to patient, but also depending on the treatment category. Also, although endovascular therapy tended overall to be associated with the higher probability of recanalization, the highest probability of recanalization was not always observed in the endovascular therapy category, and the relative probability from treatment category to treatment category was also variable (sometimes same probability in the three categories, and sometimes very different probabilities in the three categories).

figure

Figure 1. Prediction of the probability of recanalization (ranging from 0 to 1, or 0% to 100%, on the X axis) in the study patients. The patients are grouped based on the type of revascularization treatment they received: (a) patient who received no treatment; (b) patient who received intravenous (IV) treatment; (c) patient who received intra-arterial (IA) treatment. Three hypothetical situations are simulated: the hypothetical situation if they had received endovascular therapy (triangles), the hypothetical situation if they had received had received IV thrombolysis (circles), and the hypothetical situation if these patients had not received any revascularization therapy (squares). The recanalization status observed in these patients is indicated with the squares/circles/triangles being full (recanalization) or empty (no recanalization). This simulation demonstrates that there is a significant individual patient variation in terms of the relative rates of recanalization of endovascular therapy, IV thrombolysis (and no treatment).

Download figure to PowerPoint

figure

Figure 2. Illustration of the iStrokeMD app, available on the Apple Store for iPhone and iPad. On the middle panel, the user inputs the value of the relevant predictive factors. Based on these values, the far-right panel demonstrates the probability of recanalization, in %, should the patient receive no revascularization therapy, IV tPA, and endovascular treatment. IA, intra-arterial; IV, intravenous; NASCET, North American Symptomatic Carotid Endarterectomy Trial; tPA, tissue plasminogen activator.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References
  8. Supporting Information

Although pretreatment and final infarct volumes are pivotal biomarkers to predict acute ischemic stroke patients' clinical outcome [25, 26], early reperfusion of ischemic brain tissue remains the strongest predictor of neurologic improvement [27]. Prognostic value of the penumbral information can be assessed accurately only if recanalization status is known or can be predicted [1].

In the present study, we systematically reviewed the clinical and imaging biomarkers reported in the literature as predictive of the likelihood of recanalization. Our goal was to determine which ones are relevant in patients who do not receive revascularization therapy, in patients treated with IV thrombolysis, and in patients who received endovascular therapy. Being able to compare the probability of recanalization with IV or endovascular revascularization therapy compared with the rate of spontaneous recanalization in untreated patients is particularly important to determine whether the risk associated with treatment is warranted by the incremental likelihood of recanalization. Our study is unique in the sense that it provides values for this spontaneous rate of recanalization in the absence of revascularization treatment.

Our results indicate that treatment is the best predictor of recanalization. In general, IA therapy was associated with higher rates of recanalization, followed by the more limited rates of recanalization achieved by IV thrombolysis. This finding is in agreement with a large meta-analysis [15] and more recent articles [28, 29]. Our analysis (Fig. 1), however, demonstrates that there is a significant individual patient variation in terms of the relative rates of recanalization of endovascular therapy, IV thrombolysis, and no treatment.

The rate of recanalization associated with IV thrombolysis varies depending on the site of occlusion. Higher rates are observed when M1/A1 are occluded compared with ICA. IV tPA and endovascular revascularization had similar recanalization rates for M1/A1 occlusions, while endovascular revascularization had significantly higher recanalization rates compared with IV tPA for ICA occlusions. The more limited efficacy of IV tPA in patients with ICA occlusion has previously been reported, and attributed to larger or less ‘friable’ clots typically observed in the ICA [5, 30, 31]. Our study also finds CBS and NASCET to be determinants of vessel recanalization in patients receiving IV thrombolysis, in agreement with previous studies [21, 32]. A severe extracranial ICA stenosis represents an obstacle to recanalization, and a lower CBS, corresponding to a more extensive thrombus, is associated with lower recanalization rates. Patients harboring these imaging biomarkers may need more aggressive, endovascular recanalization treatment.

Our study found men to be more likely to recanalize than women, which is at odds with a previous study showing a trend toward lower recanalization rates in men [32]. Finally, we were surprised by the higher rate of recanalization observed in patients with hyperlipidemia. One possible hypothesis for this observation is that the higher rate of recanalization is not related to hyperlipidemia itself, but to the statins that the patients with hyperlipidemia are more likely to receive prior to the index stroke [33].

In patients who received endovascular therapy, the only variable that predicted the likelihood of recanalization was the relative clot density on CTA. Recanalization was more likely when the relative clot density was high, in agreement with other studies [11].

We acknowledge a number of limitations to our study. Our study was retrospective in nature, the study sample size was relatively small, and the timing and imaging modalities used for the follow-up were heterogenous. Also, our end-point was recanalization, which an important but not the only predictor of outcome. Outcome in acute ischemic stroke patients depends on a number of factors, including, but not limited to, recanalization, delay to recanalization, ischemic core, and penumbra.

The timing of the recanalization imaging assessment could also have represented a confounder for our analysis, because of the wide range we had in our study, including some patients with relatively late recanalization imaging assessment. To address this limitation, the timing of the recanalization imaging assessment was considered in our statistical analysis, but did not turn out to be a significant variable.

We focused on recanalization (TIMI) rather than reperfusion [measured using, for instance, the modified TICI [34, 35]] because of the type of vascular imaging available in our registry. Recanalization may happen even in the absence of reperfusion, resulting in poor outcome. Reperfusion is overall a better predictor of outcome then recanalization [36]. Also, we did not distinguish between different types of devices in the IA group.

In conclusion, our study is proposing models to predict the likelihood of recanalization in acute ischemic stroke patients receiving no revascularization therapy, IV thrombolysis, and endovascular treatment. The predictive factors were different in each of these three categories. The overall likelihood of recanalization was the highest in the endovascular group, and higher for IV thrombolysis compared with no revascularization therapy. However, our theoretical predictions for each individual patient indicate a significant variability in the relative likelihood of recanalization for each treatment group, suggesting the need to include this factor in the personalized treatment selection. This approach needs to be validated in future studies with prospective design.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References
  8. Supporting Information
  • 1
    Zhu G, Michel P, Aghaebrahim A et al. Prediction of recanalization trumps prediction of tissue fate: the penumbra: a dual-edged sword. Stroke 2013; 44:10141019.
  • 2
    Albers GW, Thijs VN, Wechsler L et al. Magnetic resonance imaging profiles predict clinical response to early reperfusion: the diffusion and perfusion imaging evaluation for understanding stroke evolution (DEFUSE) study. Ann Neurol 2006; 60:508517.
  • 3
    Davis SM, Donnan GA, Parsons MW et al. Effects of alteplase beyond 3 h after stroke in the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET): a placebo-controlled randomised trial. Lancet Neurol 2008; 7:299309.
  • 4
    Lansberg MG, Straka M, Kemp S et al. MRI profile and response to endovascular reperfusion after stroke (DEFUSE 2): a prospective cohort study. Lancet Neurol 2012; 11:860867.
  • 5
    Kimura K, Iguchi Y, Shibazaki K et al. Early stroke treatment with IV t-PA associated with early recanalization. J Neurol Sci 2010; 295:5357.
  • 6
    Lansberg MG, Dabus G. Interaction between time to treatment and reperfusion therapy in patients with acute ischemic stroke. J Neurointerv Surg 2013; 5:4851.
  • 7
    Kufner A, Nolte CH, Galinovic I et al. Smoking-thrombolysis paradox: recanalization and reperfusion rates after intravenous tissue plasminogen activator in smokers with ischemic stroke. Stroke 2013; 44:407413.
  • 8
    Hermier M, Nighoghossian N, Adeleine P et al. Early magnetic resonance imaging prediction of arterial recanalization and late infarct volume in acute carotid artery stroke. J Cereb Blood Flow Metab 2003; 23:240248.
  • 9
    Aoki J, Kimura K, Shibazaki K et al. DWI-aspects as a predictor of dramatic recovery after intravenous recombinant tissue plasminogen activator administration in patients with middle cerebral artery occlusion. Stroke 2013; 44:534537.
  • 10
    Kimura K, Iguchi Y, Shibazaki K et al. M1 susceptibility vessel sign on T2* as a strong predictor for no early recanalization after IV-t-PA in acute ischemic stroke. Stroke 2009; 40:31303132.
  • 11
    Moftakhar P, English JD, Cooke DL et al. Density of thrombus on admission ct predicts revascularization efficacy in large vessel occlusion acute ischemic stroke. Stroke 2013; 44:243245.
  • 12
    Shobha N, Bal S, Boyko M et al. Measurement of length of hyperdense MCA sign in acute ischemic stroke predicts disappearance after IV tPA. J Neuroimaging 2014; 24:710.
  • 13
    Nicoli F, Lafaye de Micheaux P, Girard N. Perfusion-weighted imaging-derived collateral flow index is a predictor of MCA M1 recanalization after i.v. thrombolysis. AJNR Am J Neuroradiol 2013; 34:107114.
  • 14
    Zhu G, Michel P, Aghaebrahim A et al. Computed tomography workup of patients suspected of acute ischemic stroke: perfusion computed tomography adds value compared with clinical evaluation, noncontrast computed tomography, and computed tomography angiogram in terms of predicting outcome. Stroke 2013; 44:10491055.
  • 15
    Rha JH, Saver JL. The impact of recanalization on ischemic stroke outcome: a meta-analysis. Stroke 2007; 38:967973.
  • 16
    Barber PA, Demchuk AM, Zhang J et al. Validity and reliability of a quantitative computed tomography score in predicting outcome of hyperacute stroke before thrombolytic therapy. Aspects study group. Alberta Stroke Programme early CT score. Lancet 2000; 355:16701674.
  • 17
    Passamani E, Hodges M, Herman M et al. The thrombolysis in myocardial infarction (TIMI) phase II pilot study: tissue plasminogen activator followed by percutaneous transluminal coronary angioplasty. J Am Coll Cardiol 1987; 10:51B64B.
  • 18
    Smith WS, Sung G, Starkman S et al. Safety and efficacy of mechanical embolectomy in acute ischemic stroke: results of the merci trial. Stroke 2005; 36:14321438.
  • 19
    Zaidat OO, Yoo AJ, Khatri P et al. Recommendations on angiographic revascularization grading standards for acute ischemic stroke: a consensus statement. Stroke 2013; 44:26502663.
  • 20
    Puetz V, Dzialowski I, Hill MD et al. Intracranial thrombus extent predicts clinical outcome, final infarct size and hemorrhagic transformation in ischemic stroke: the clot burden score. Int J Stroke 2008; 3:230236.
  • 21
    Tan IY, Demchuk AM, Hopyan J et al. Ct angiography clot burden score and collateral score: correlation with clinical and radiologic outcomes in acute middle cerebral artery infarct. AJNR Am J Neuroradiol 2009; 30:525531.
  • 22
    Tan JC, Dillon WP, Liu S et al. Systematic comparison of perfusion-CT and CT-angiography in acute stroke patients. Ann Neurol 2007; 61:533543.
  • 23
    Moneta GL, Edwards JM, Chitwood RW et al. Correlation of North American Symptomatic Carotid Endarterectomy Trial (NASCET) angiographic definition of 70% to 99% internal carotid artery stenosis with duplex scanning. J Vasc Surg 1993; 17:152157, discussion 157–159.
  • 24
    Torres-Mozqueda F, He J, Yeh IB et al. An acute ischemic stroke classification instrument that includes CT or MR angiography: the boston acute stroke imaging scale. AJNR Am J Neuroradiol 2008; 29:11111117.
  • 25
    Yoo AJ, Chaudhry ZA, Nogueira RG et al. Infarct volume is a pivotal biomarker after intra-arterial stroke therapy. Stroke 2012; 43:13231330.
  • 26
    Kruetzelmann A, Kohrmann M, Sobesky J et al. Pretreatment diffusion-weighted imaging lesion volume predicts favorable outcome after intravenous thrombolysis with tissue-type plasminogen activator in acute ischemic stroke. Stroke 2011; 42:12511254.
  • 27
    Bang OY, Liebeskind DS, Buck BH et al. Impact of reperfusion after 3 hours of symptom onset on tissue fate in acute cerebral ischemia. J Neuroimaging 2009; 19:317322.
  • 28
    Abou-Chebl A. Intra-arterial therapy for acute ischemic stroke. Interventional Neurology 2012; 1:100108.
  • 29
    Mazighi M, Serfaty JM, Labreuche J et al. Comparison of intravenous alteplase with a combined intravenous-endovascular approach in patients with stroke and confirmed arterial occlusion (recanalise study): a prospective cohort study. Lancet Neurol 2009; 8:802809.
  • 30
    Kimura K, Sakamoto Y, Aoki J et al. Clinical and MRI predictors of no early recanalization within 1 hour after tissue-type plasminogen activator administration. Stroke 2011; 42:31503155.
  • 31
    Zangerle A, Kiechl S, Spiegel M et al. Recanalization after thrombolysis in stroke patients: predictors and prognostic implications. Neurology 2007; 68:3944.
  • 32
    Mendonca N, Rodriguez-Luna D, Rubiera M et al. Predictors of tissue-type plasminogen activator nonresponders according to location of vessel occlusion. Stroke 2012; 43:417421.
  • 33
    Georgiadis AL, Hussein HM, Vazquez G et al. Premorbid use of statins is associated with higher recanalization rates in patients with acute ischemic stroke undergoing endovascular treatment. J Neuroimaging 2009; 19:1922.
  • 34
    Kallmes DF. TICI: if you are not confused, then you are not paying attention. AJNR Am J Neuroradiol 2012; 33:975976.
  • 35
    Goyal M, Fargen KM, Turk AS et al. 2C or not 2C: defining an improved revascularization grading scale and the need for standardization of angiography outcomes in stroke trials. Journal of neurointerventional surgery 2013; 6:8386.
  • 36
    Soares BP, Tong E, Hom J et al. Reperfusion is a more accurate predictor of follow-up infarct volume than recanalization: a proof of concept using CT in acute ischemic stroke patients. Stroke 2010; 41:E383.
  • 37
    Hardin JW, Hilbe JM. Generalized Estimating Equations. Wiley encyclopedia of clinical trials. New York, John Wiley & Sons, Inc., 2007.
  • 38
    Harrell F. Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis. New York, Springer, 2001.
  • 39
    Obuchowski NA. Receiver operating characteristic curves and their use in radiology. Radiology 2003; 229:38.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. References
  8. Supporting Information
FilenameFormatSizeDescription
ijs12312-sup-0001-si.docx48K

Table S1. (a–d) Generalized estimating equation regression models for predicting the log-odds for vessel-level recanalization among patients who received IV tPA or no treatment (panels a), and among patients who received IA revascularization therapy (panels c). Vessel recanalization index values (VRI) that were used in the patient-level recanalization analyses were computed based on the regression coefficients of the reduced models that are presented in panels (b) and (d), respectively, for the IV or no treatment patient subgroup, and for the IA patient subgroup.

Table S2. (a, b) Added variable rankings for the IV and no treatment subgroup of patients (panel a), and added variable rankings for the IA revascularization subgroup of patients (panel b). Note that the order ranks were computed based subtracting the expect likelihood ratio chi-squared statistic (i.e. the model degrees of freedom) under the null hypothesis that no patient-level recanalization prediction information is provided by the model predictor variables from the model likelihood ratio chi-square statistic. Note that VRI was a common predictor in all models.

Table S3. (a, b). Logistic regression models for predicting the log-odds for patient-level recanalization among patients who received IV tPA or no treatment (panel a), and among patients who received IA revascularization therapy (panel b).

Appendix S1. The appendix describes the analytical details for step 1 and step 2 of the statistical modeling process (37–39).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.