Article Text
PDF
Abstract
Purpose The absence of the susceptibility vessel sign (negative SVS) on gradient-recalled echo or susceptibility-weighted imaging (SWI) in thrombolytic therapy has not been well studied. Since positive and negative SVS may have different components, we aimed to investigate the difference in risk factors for clinical outcome between patients with positive and negative SVS.
Methods We retrospectively examined clinical and imaging data from 85 consecutive patients with acute ischemic stroke with middle cerebral artery occlusion who underwent SWI before intravenous thrombolysis (IVT). We then examined the predictors of negative SVS and the risk factors for a poor outcome (defined as modified Rankin Scale score ≥3) 3 months after IVT in subgroup analysis.
Results Multivariate regression analysis indicated that previous antiplatelet use (OR 0.076; 95% CI 0.007 to 0.847; p=0.036) and shorter time from onset to treatment (OR 1.051; 95% CI 1.003 to 1.102; p=0.037) were inversely associated with poor outcome in patients with negative SVS, while higher baseline National Institutes of Health Stroke Scale (NIHSS) score was associated with poor outcome in patients with positive SVS (OR 1.222; 95% CI 1.084 to 1.377; p=0.001).
Conclusions The risk factors for clinical outcome after IVT in patients with negative SVS may differ from those with positive SVS.
- Thrombolysis
- MRI
Statistics from Altmetric.com
Introduction
Previous studies have postulated that the susceptibility vessel sign (SVS) on gradient-recalled echo (GRE) or susceptibility-weighted imaging (SWI) represents the red clot containing erythrocytes and some fibrin.1 ,2 The susceptibility variation of paramagnetic deoxygenated hemoglobin from the erythrocytes could lead to a fast decay of the local T2* magnet resonance signal due to local inhomogeneity of the magnetic field.3 A recent imaging/pathology correlation study further proved that SVS of the middle cerebral artery (MCA) on GRE sequences represented red blood cell-dominant thrombi.4 The presence of SVS has been reported to be associated with a history of atrial fibrillation or cardioembolic stroke.5 ,6
It has been demonstrated that the morphology of SVS, such as irregular shape and long length (>20 mm), decreases the potential to recanalize the occluded MCA with intravenous thrombolysis (IVT).7 However, in those studies, the rate of negative SVS (the absence of SVS) was still not uncommon, ranging from 26% to 73% in patients with acute ischemic stroke,6–9 and has not been well studied. Based on the imaging/pathology study, one could postulate that positive and negative SVS may represent different components. We therefore systematically investigated the difference in risk factors for clinical outcome between patients with positive and negative SVS, which may be important to improve the medical decision-making for reperfusion therapy.
Methods
Study subjects
We retrospectively reviewed our prospectively collected database for consecutive patients with acute ischemic stroke who received thrombolytic therapy between June 2009 and February 2014. We then enrolled patients who (1) had a diagnosis of acute ischemic stroke confirmed by diffusion-weighted imaging (DWI); (2) received intravenous recombinant tissue plasminogen activator (rtPA); (3) underwent SWI and time-of-flight magnetic resonance angiography (TOF-MRA) before rtPA infusion; (4) had MCA (M1 and M2) occlusion without internal carotid artery involvement demonstrated on baseline TOF-MRA; (5) underwent follow-up TOF-MRA 24 h after rtPA infusion; (6) underwent follow-up SWI or CT 24 h after rtPA infusion; and (7) had a pre-stroke modified Rankin Scale (mRS) score ≤2.10 Patients who were treated with combined endovascular and rtPA therapy and those whose image quality was poor due to motion artifacts were excluded.
We retrieved demographic, clinical, laboratory, and imaging data including age; gender; comorbid conditions such as history of hypertension, diabetes mellitus, hyperlipidemia, or atrial fibrillation (classified as first-detected or chronic);11 previous antiplatelet use; time from onset to rtPA infusion; National Institutes of Health Stroke Scale (NIHSS) score;12 serum glucose, platelet, and international normalized ratio (INR) level before rtPA infusion.
MRI parameters
All subjects underwent MRI on a 3.0 T system (Signa Excite HD, General Electric Medical System, Milwaukee, USA) equipped with an eight-channel phased array head coil. The 3D multi-echo GRE sequence used 11 equally spaced echoes: echo time=4.5 ms (first echo); inter-echo spacing=4.5 ms; repetition time=58 ms; FOV=24×24 cm2; matrix size=256×256; flip angle=20°; slice thickness=2.0 mm with no gap between slices; and in-plane spatial resolution of 0.4688×0.4688 mm/pixel. The DWI sequence was used to measure the infarct volume (TR=4000 ms; TE=69.3 ms; b-value=1000 s/mm2; FOV=24×24 cm2; matrix size=160×160; slice thickness=5.0 mm; interslice gap=1.0 mm; and in-plane spatial resolution of 0.9375×0.9375 mm/pixel), and TOF-MRA was used to evaluate vessel occlusion (TR=20 ms; TE=3.2 ms; flip angle=15°; FOV=24×24 cm2; matrix size=320×224; slice thickness=1.4 mm, 3 slabs).
SVS analysis
A positive SVS was defined as the presence of hypointensity in the MCA with a blooming artifact (ie, the diameter exceeded the hypointense signal in the homologous contralateral vessel diameter on GRE or SWI scans; figure 1C).7 Thus, a negative SVS was defined as the absence of hypointensity in the occluded MCA with a blooming artifact on any visible slices (figure 1F). Two neurologists, blinded to MRA and clinical information, independently assessed the presence of SVS on SWI scans before treatment with rater discrepancies settled by consensus discussion. Acute arterial occlusion of the M1 and M2 segments on initial MRA was identified as invisibility of the artery on MRA with the presenting compatible symptoms.
Illustrations of positive and negative susceptibility vessel sign (SVS). (A) and (D) are reconstructed images of time-of-flight magnetic resonance angiography (TOF-MRA); (B) and (E) are raw slices of TOF-MRA; (C) and (F) are slices of susceptibility-weighted imaging (SWI). (A), (B) and (C) show a patient with left M1 occlusion who had positive SVS (arrow on C); (D), (E) and (F) show a patient with left M1 occlusion who had negative SVS (arrow on F).
Evaluation of outcome
We used the Arterial Occlusive Lesion (AOL) scale to define recanalization or no recanalization based on the presence (grades 2 or 3) or absence (grades 0 or 1) of any downstream flow, according to a recent consensus statement of recommendations on cerebral angiographic revascularization grading standards.13 Hemorrhagic transformation was classified using clinical and imaging criteria as follows: hemorrhagic infarction, parenchymal hemorrhage (PH), and symptomatic intracranial hemorrhage (sICH).14 Outcome at 3 months was assessed using the mRS score and dichotomized into good outcome (0–2) and poor outcome (3–6).
Reliability and validity of the radiological data
The two investigators who jointly evaluated the MRI findings were blinded to the patients’ clinical data. A single trained observer (SY) evaluated the images of all patients twice, at an interval of 3 months. Another observer (YY) independently made the same evaluation. Discrepancies were resolved by consensus.
Statistical analysis
The Fisher exact test was used to compare the dichotomous variables between groups and the independent samples two-tailed t test or Mann–Whitney U test was used for the continuous variables, as appropriate. Variables with a two-tailed p value of <0.1 in univariate analyses were included in the binary logistic regression model to determine predictors for negative SVS, except for baseline NIHSS score as it was impossible for it to be the cause of negative SVS. A history of diabetes mellitus and the serum glucose level were entered into the model separately. All analyses were performed blinded to participant identifying information. Statistical significance was set at a p value of <0.05. All statistical analysis was performed with SPSS V.14.0 for Windows.
Results
The inter-observer and intra-observer reliabilities for the presence of SVS were excellent (kappa values 0.971 and 0.942, respectively). A total of 85 remaining patients were included for the final analysis. Demographic, clinical, and laboratory data were not different between included and excluded subjects. Of the included patients, the median age was 66 years (mean 66±13 years, range 43–94 years) and 28 (32.9%) were women. Follow-up MRA 24 h after rtPA infusion revealed recanalization in 42 (49.4%) patients. On initial SWI scans, we observed the presence of MCA SVS in 61 (71.8%) patients. Negative SVS was found in 31.1% (19/61) of M1 occlusions and 20.8% (5/24) of M2 occlusions.
Multiple regression analysis identified only one independent predictor for PH and poor outcome—namely, baseline NIHSS score (OR 1.152; 95% CI 1.046 to 1.268; p=0.004). For recanalization, multiple regression analysis identified two significant independent predictors: the presence of M1 occlusion (OR 0.271; 95% CI 0.091 to 0.810; p=0.019) and first-detected atrial fibrillation (OR 5.539; 95% CI 1.366 to 22.468; p=0.017). The presence of negative MCA SVS did not predict recanalization (OR 1.466; 95% CI 0.512 to 4.199; p=0.476), PH (OR 1.328; 95% CI 0.270 to 6.521; p=0.727), or poor outcome (OR 1.110; 95% CI 0.341 to 3.610; p=0.863) in patients with MCA occlusion.
Table 1 shows the characteristics of patients with positive and negative MCA SVS for comparison. Patients with negative MCA SVS had a lower NIHSS score and a higher serum glucose and INR level in the univariate analysis. Age (OR 1.046; 95% CI 1.001 to 1.093; p=0.047), serum glucose (OR 1.450; 95% CI 1.097 to 1.916; p=0.009), and INR level (OR 2.277; 95% CI 1.181 to 4.388; p=0.014) were identified as independent predictors for negative MCA SVS, while serum platelet level (OR 1.008; 95% CI 1.000 to 1.017; p=0.053) showed a possible trend towards significance (table 2). When we replaced the serum glucose level with a history of diabetes mellitus, it also independently predicted negative SVS (OR 4.047; 95% CI 1.031 to 15.884; p=0.045).
Comparison of characteristics of patients with positive and negative MCA SVS
Multivariate regression analysis of independent predictors for negative MCA SVS
Concerning the clinical outcome in patients with positive or negative SVS, we found that previous antiplatelet use (OR 0.076; 95% CI 0.007 to 0.847; p=0.036) and shorter time interval from stroke onset to treatment (ONT) (OR 1.051; 95% CI 1.003 to 1.102; p=0.037) were inversely associated with poor outcome in patients with negative SVS, while higher baseline NIHSS score was associated with poor outcome in patients with positive SVS (OR 1.222; 95% CI 1.084 to 1.377; p=0.001) (table 3).
Univariate and multivariate analysis of poor clinical outcome in subgroup analysis
Follow-up SWI was performed in 55 patients, of which 39 (71%) had a positive SVS and 16 (29%) had a negative SVS. Changes in positive SVS on follow-up SWI were consistent with vascular status on follow-up MRA. In 26 patients with recanalization, the positive SVS completely disappeared in 18 patients (69.2%) and partially disappeared in eight (30.8%). In 13 patients without recanalization, the positive SVS partially disappeared in three patients (23.1%) and remained unchanged in 10 (76.9%). For all of the 16 patients with negative SVS, this sign remained ‘invisible’ on follow-up SWI, even in four patients without recanalization. As expected, there were no cases of positive SVS in the remaining 12 patients with recanalization.
Discussion
This is the first study to systematically investigate the formation and prognostic value of the absence of the SVS in patients with MCA occlusion. We found that the risk factors for clinical outcome after IVT were different for patients with and without the SVS. Moreover, the negative SVS remained ‘invisible’ on follow-up SWI in all patients without recanalization.
An autopsy study showed that thromboembolic occlusions can be caused by white, red, or mixed thrombi.15 White thrombi predominantly consist of varying amounts of cellular debris, fibrin, and platelet aggregates, while red thrombi are rich in fibrin and trapped erythrocytes.1 The presence of unpaired electrons in deoxyhemoglobin and hemosiderin gives them paramagnetic properties which produce a non-uniform magnetic field and a rapid dephasing of proton spins, resulting in a loss of signal best seen on GRE scans.3 Thus, red thrombi may be seen as a hypointense signal within occlusive vessels on GRE. Recent pathological findings of occlusive thrombus also showed that the content of the red blood cells determined the appearance of blooming artifact on GRE.4
In the current study we found that the serum glucose and INR level were related to the phenomenon of negative SVS. In patients with diabetes mellitus who have poor control of glycemia, the activation and aggregation of platelets was increased.16 ,17 The prevalence of atherosclerotic thromboembolic disease was also increased in patients with diabetes mellitus.18 We thus could assume that, in patients with a high serum glucose level, white thrombi which mainly consist of platelet aggregates will prevail, without producing the same SVS as in red thrombi. In fact, red thrombi predominantly resulted from activation of the plasma coagulation cascade in areas with reduced blood flow. It is thus also rational to assume that the increased INR level may be associated with decreased formation of red clot. A clot lifespan model analysis also confirmed that clot strength and speed of clot formation increased as INR decreased.19 Future imaging/pathology correlation studies are needed to confirm our assumption. However, thrombus composition might not be the only cause of negative SVS. Some small positive SVS might be missed (false negative SVS) because of susceptibility artifacts at the skull base and partial volume effect of large slice thickness. We therefore selected patients with MCA occlusion without internal carotid artery involvement and used a contiguous thin-slice (2 mm), 3D multi-echo SWI sequence performed with a 3 T MRI unit.
Our findings that previous use of antiplatelet agents and short ONT were associated with good outcome in patients with negative SVS also lend support to the assumption of the prevalence of white thrombi in those patients. Thrombi that consist of more platelet aggregates (white thrombi) might be easier to grow in a short time and thus benefit from prior use of antiplatelet agents. On the other hand, white thrombi composed of platelet-rich material also displayed a relative resistance against IVT.20 From this point of view, the decrease in ONT for thrombolytic therapy is much more urgent in patients with negative SVS. We therefore assumed that the therapy window for patients with negative SVS might be narrower than for those with positive SVS. Further prospective studies need to investigate the different therapy window between these two groups.
Previously, Kimura et al hypothesized that the main component in hyperacute thrombi might be oxyhemoglobin, which made the clot invisible on SWI leading to the identification of negative SVS.7 If this is true, the negative SVS will become positive on follow-up SWI as desaturation of hemoglobin from oxyhemoglobin to deoxyhemoglobin occurs within a few hours. However, in our study the negative SVS remained invisible on follow-up SWI in all patients without recanalization, indicating that this sign may not be related to the component of oxyhemoglobin.
It is worth noting that the formation of a clot is a dynamic process so the initial clot source may be masked by in situ clot aging with structural reorganization. This could make it difficult to relate the visibility of the clot to any stroke etiology or mechanisms. In our study the presence of SVS did not differ between patients with first-detected chronic atrial fibrillation and those without atrial fibrillation, which is different from previous studies which found that the presence of SVS might be an independent predictor of cardioembolic stroke.5 ,6 The histological analysis of thrombi retrieved by endovascular mechanical extraction also failed to demonstrate a clear relationship between thrombus composition and stroke etiology.21
Study limitations
Limitations of the study include the retrospective design, although we prospectively collected data using a stroke registry and MRI protocol, and there might be a potential risk of selection bias. Second, the use of TOF-MRA is somewhat inaccurate for detecting vessel occlusion or stenosis. Third, the sample size was modest and was performed at a single center. Confirmation and extension in large multicenter cohorts is needed. Finally, our results only apply to centers that use MRI as first-line imaging investigation in acute ischemic stroke, leading to limited applicability.
Conclusions
This study shows that negative SVS may be related to high serum glucose and INR levels, while previous use of antiplatelet agents and short ONT are inversely associated with a poor outcome after IVT in these patients, which might indicate the prevalence of white thrombi when no SVS was visible in patients with MCA occlusion. However, future prospective studies with large sample sizes are required to validate our results.
References
Footnotes
Contributors SY: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, acquisition of data, statistical analysis, study supervision. KL: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, acquisition of data. LT, YY, SZ: drafting/revising the manuscript, study concept or design, analysis or interpretation of data. ML: drafting/revising the manuscript, study concept or design, analysis or interpretation of data, contribution of vital reagents/tools/patients, study supervision, obtaining funding.
Funding This work was supported by the Science Technology Department of Zhejiang Province (2013C03043-3) and the National Natural Science Foundation of China (81171095 and 81471170).
Competing interests None declared.
Patient consent Obtained.
Ethics approval Ethics approval was obtained from the human ethics committee of the Second Affiliated Hospital of Zhejiang University, School of Medicine.
Provenance and peer review Not commissioned; externally peer reviewed.