Article Text

PDF

Case series
Predictive value of CT angiography source image ASPECTS in patients with anterior circulation acute ischemic stroke after endovascular treatment: ultimate infarct size and clinical outcome
  1. Jung-Soo Park1,
  2. Jong Myong Lee1,
  3. Hyo-Sung Kwak2,
  4. Gyoung Ho Chung2
  1. 1 Departments of Neurosurgery, Research Institute of Clinical Medicine of Chonbuk National University-Biomedical Research Institute of Chonbuk National University Hospital, Jeonju, Republic of Korea
  2. 2 Department of Radiology, Research Institute of Clinical Medicine of Chonbuk National University-Biomedical Research Institute of Chonbuk National University Hospital, Jeonju, Republic of Korea
  1. Correspondence to Professor Hyo-Sung Kwak, Department of Radiology, Research Institute of Clinical Medicine of Chonbuk National University-Biomedical Research Institute of Chonbuk National University Hospital, Jeonju 561-712, Republic of Korea; kwak8140{at}jbnu.ac.kr

Abstract

Background and purpose The objective of this study was to investigate the predictive value of computed tomographic angiography (CTA) source image Alberta Stroke Program Early CT Score (ASPECTS) on clinical outcome and final infarction extent after endovascular treatment (EVT) in patients with acute ischemic stroke (AIS).

Methods All eligible patients from January 2014 to March 2018 undergoing EVT due to anterior circulation AIS were retrospectively reviewed. The baseline ASPECTS on initial noncontrast CT (NCCT) and the CTA source image were compared with the follow-up MR diffusion-weighted imaging (DWI) ASPECTS. Receiver operating characteristic (ROC) curve analysis was used to assess the predictive value of CTA ASPECTS for prediction of final infarct extent and good clinical outcome.

Results Our sample included a total of 122 eligible patients (64 males, mean age 70 years) with a median baseline National Institute of Health Stroke Scale (NIHSS) score of 12. Baseline ASPECTS on the CTA source image correlated with follow-up MR DWI ASPECTS better than NCCT ASPECTS (P<0.001). ROC curve analysis revealed baseline CTA ASPECTS (area under the curve [AUC] =0.74, 95% CI: 0.65 to 0.83, P<0.001) can better predict favorable functional outcome than NCCT ASPECTS (AUC=0.64, 95% CI: 0.54 to 0.74, P=0.008). Baseline NIHSS score <15, CTA ASPECTS≥8, and successful recanalization were independent predictors of good clinical outcomes.

Conclusion The ASPECTS on the CTA source image provides more information in the prediction of good clinical outcome and final infarction size than NCCT in patients with AIS treated with EVT.

  • ct angiography
  • stroke

Statistics from Altmetric.com

Although magnetic resonance (MR) diffusion-weighted imaging (DWI) is the most accurate method for detection of cerebral infarction in the hyperacute period, noncontrast computed tomography (NCCT) is considered the initial imaging modality for patients with acute ischemic stroke (AIS). Usually, NCCT is not only used to exclude hemorrhagic stroke or tumorous conditions, but it can also detect regions of the brain with irreversible ischemic injury. The Alberta Stroke Program Early CT Score (ASPECTS) is a semiquantitative grading system to detect early ischemic change within the middle cerebral artery (MCA) territory.1 The ASPECTS is scored from 10 divided regions of interest in the MCA territory, with a lower score indicating greater infarct burden. In the field of endovascular treatment (EVT) for AIS, studies on ASPECTS have been focused on the threshold showing good clinical outcome after treatment.2 3 Recent randomized controlled trials on EVT for AIS have been based on strict inclusion criteria, leading to treatment of only those patients with high ASPECTS indicating smaller infarction volume.4–6 However, NCCT is less sensitive to detecting early infarction-related ischemic change.7

The CT angiography (CTA) source image can be rapidly obtained with minimal delay after NCCT. Although CTA has been shown to have value in identifying vessel occlusion, the CTA source image also provides information on collaterals and contrast delivery to the tissue capillaries, and may be useful to predict tissue that is undergoing critical ischemia and is likely to be dead.8 Previous studies showed that the CTA source image was superior to NCCT for predicting final infarction volume.9 10 However, although a few studies exist on the value of CTA-ASPECTS in patients with AIS treated with EVT, NCCT-ASPECTS is still considered as a gold standard for eligible patients' selection for EVT.

In this study, we seek to demonstrate the predictive value and reliability of CTA-ASPECTS compared with NCCT ASPECTS, as well as other known parameters, in patients with anterior circulation AIS treated with EVT.

Materials and methods

Patients

We used a prospectively collected clinical and radiological database to enroll patients with acute anterior circulation ischemic stroke treated with EVT from January 2014 to March 2018. Admission non-contrast computed tomography (NCCT) was obtained, followed by CTA. Patients who did not have available initial CT, CTA, or who did not undergo MRI after EVT were excluded. Patient demographic and initial clinical characteristics were recorded at the time of admission to the emergency room.

Imaging

Standard NCCT was performed with a multislice CT scanner (SOMATOM Sensation 16; Siemens, Erlangen, Germany) and was initially obtained for all patients using 120 kV, 170 mA with 5 mm section thickness. NCCT was followed on the same scanner by CTA with a helical scan technique. CTA was then performed by scanning from the vertex to the aortic arch. Nonionic contrast media (80–120 mL) was administered into the antecubital vein at 3–5 mL/s. The image was autotriggered by the appearance of contrast media in the superior sagittal sinus and the CTA source image was reconstructed at 1 mm thickness in axial planes at 0.7 mm thickness intervals. Magnetic resonance imaging (MRI) (Magnetom Verio, Siemens, Erlangen, Germany) was performed within 24 hours after EVT and used for diffusion weighted imaging (DWI) ASPECTS.

Clinical and imaging analysis

We retrospectively analyzed patients’ medical records to determine the demographic, clinical, and angiographic data. The initial neurological status of each patient was evaluated using the National Institutes of Health Stroke Scale (NIHSS) score at admission. The modified Rankin scale (mRS) was checked on admission and at 3 months for all patients, and a favorable functional outcome was defined as a 3-month mRS score of ≤2. Image review was independently performed on a wide-screen high-resolution monitor with an adequate window and level setting. ASPECTS was calculated by subtracting the number of affected regions from a total score of 10 and assessed on baseline CT, CTA, and follow-up MRI DWI by two neurointerventionists (HSK., JSP.). Early ischemic change on NCCT for calculating NCCT-ASPECTS was defined as tissue hypoattenuation, loss of gray-white matter differentiation, and sulcal effacement. The CTA source image was reviewed at the window and levels where there was maximum contrast between normal and ischemic tissue.11 For assessing CTA-ASPECTS, regions of relatively diminished contrast enhancement were scored as abnormal. ASPECTS on MRI was calculated by subtracting the regions of high signal intensity on DWI from 10.

Statistical analysis

The analysis was performed by using SPSS 18.0 (IBM SPSS, Chicago, IL). Continuous variables were summarized as means with SD or medians with IQR, while categorical data were expressed as counts and percentages. Univariate associations between baseline ASPECTS (on CT and CTA) and follow-up DWI ASPECTS were investigated with the Spearman ρ analysis. Bubble plots were used to visualize these correlation analyses. To compare the predictive value for favorable outcomes (mRS ≤2) with CT ASPECTS and CTA ASPECTS, we used receiver operating characteristic (ROC) curve analysis and the area under the curve (AUC). We also used ROC curve analysis to identify the optimal cut-off point with which to maximize the sensitivity and specificity for discriminating patients with favorable outcomes.

Multivariable logistic regression analyses were performed to identify the ASPECTS (CT and CTA) and other variables as an independent factor for good functional outcomes. Statistically significant differences were defined as P<0.05.

Results

Of 168 patients with anterior circulation occlusion, we identified 122 patients who had complete CT, CTA, and follow-up MR DWI for analysis. Thirty-two patients did not have available CT or CTA, and 14 patients did not have follow-up MRI. Baseline characteristics are summarized in table 1. There were 64 males (52.5%) with mean age of 70.6±13.6 years. Baseline median (IQR) NIHSS score was 12 (9–15), NCCT ASPECTS was 9 (8–10), CTA ASPECTS was 8 (6–9), and follow-up MR DWI ASPECTS was 8 (6–9). Successful recanalization was achieved in 93 patients (76.2%). Correlation analysis showed that baseline CTA ASPECTS correlated better with follow-up MR DWI ASPECTS (r=0.84, 95% CI 0.76 to 0.91) than NCCT ASPECTS (r=0.57, 95% CI: 0.44 to 0.68; figure 1AB; P for comparison of the two coefficients<0.001). The ROC curve analysis revealed that the predictive value of favorable outcome was significantly higher for CTA ASPECTS (AUC=0.74; 95% CI: 0.65 to 0.83 P <0.001) compared with NCCT ASPECTS (AUC=0.64; 95% CI: 0.54 to 0.74 P=0.008) (figure 2). Using ROC analysis, we determined the best cut-off point was 9 for NCCT ASPECTS and 8 for CTA ASPECTS with which to maximize the sensitivity and specificity for discriminating patients with favorable outcomes (table 2). Including these cut-off ASPECTS values, we investigated the variables as an independent prognostic factor for good functional outcome. In univariate analysis, both NCCT ASPECTS≥9 and CTA ASPECTS≥8 were identified as independent predictors of favorable outcome as well as baseline NIHSS <15 and successful recanalization (TICI 2b3). But in multivariable analysis, NCCT ASPECTS≥9 had no statistical significance as an independent predictor (table 3) (figure 3). In patients with CTA SI ASPECTS≥9, more patients had successful recanalization (71/92 (77%) vs 22/30 (73%)), although there was no statistical significant difference (P=0.671).

Figure 1

Correlation of NCCT ASPECTS (A) and CTA ASPECTS (B) with follow-up MR DWI ASPECTS. The area of the bubble is proportional to the number of patient observations at that data point.

Figure 2

ROC curves show sensitivity and specificity values for prediction of good clinical outcome with NCCT ASPECTS and CTA ASPECTS at various levels of cut-off scores. ROC analysis demonstrated that CTA ASPECTS (AUC=0.74; 95% CI: 0.65 to 0.83 P<0.001) can better predict good clinical outcome than NCCT ASPECTS (AUC=0.64; 95% CI: 0.54 to 0.74 P=0.008).

Figure 3

A 56-year-old male admitted with acute right-side weakness and aphasia 105 min after symptom onset. Baseline NCCT (A,B) shows no early ischemic change in the left MCA territory (ASPECTS 10). CTA source images (C,D) shows hypoattenuation with lack of vascularity (arrows, ASPECTS 3). Follow-up MR DWI (E,F) reveals definitive infarction in large amount of MCA territory (ASPECTS 2, sparing only caudate head and internal capsule).

Table 1

Baseline characteristics of the patients

Table 2

ROC curve analysis of CTA and NCCT ASPECTS cut-off value of good clinical outcome (mRS 0–2)

Table 3

Variables associated with good functional outcome (mRS 0–2 at 3 months)

Discussion

In recent years, EVT has been considered a safe and effective treatment modality for large vessel occlusion in the anterior circulation in patients with AIS.12 13 In the decision- making process for EVT in patients with AIS, appropriate patient selection is very important to achieve good results. For this, studies about baseline characters of AIS patients have focused on features that predict a favorable clinical outcome after the procedure. According to previous studies, successful reperfusion, initial high ASPECT score on NCCT, low NIHSS score on admission, low age, and short time interval from symptom onset to recanalization were known as independent predictors of good functional outcomes.14 15 Among these variables, NCCT ASPECTS is a predictor that can be measured through an imaging study before EVT.

In treating AIS, urgent assessment of the clinical and radiological features of the brain insult is essential. In this aspect, an NCCT scan seems an ideal imaging modality with its good accessibility and relatively fast scanning time.16 The NCCT ASPECTS has been shown to correlate with outcomes in patients with AIS in several clinical studies.2 17 However, it is well known that NCCT suffers from limited sensitivity for early infarct-related change.18 Although not available as commonly as NCCT in the acute stage, CTA is useful for confirmation of vessel occlusion for EVT. Moreover, CTA can provide additional information, such as collateral circulation or clot burden length.

In assessing ASPECTS, it is important to recognize the critical ischemic area on NCCT and the CTA source image. A CTA source image region showing a lack of enhancement provides an estimate of cerebral blood volume, whereas NCCT measures shifts in brain tissue water content.19 So, in order to obtain an accurate CTA source image, the appropriate delay time from contrast injection to imaging is very important. Pulli et al reported that significant overestimation occurs with a shortened time from contrast media injection to imaging of the ischemic territory.20 In our study, we triggered CTA source image acquisition by the appearance of contrast media in the superior sagittal sinus, not the commonly used aortic arch, to ensure sufficient delay time. Hypodensity on a CTA source image provides greater demarcation between normal and abnormal tissue, and this finding could be explained by the ability of CTA to detect alterations in cerebral blood volume, as opposed to cytotoxic edema on NCCT, with a threshold insufficient to produce NCCT changes.21 For this reason, previous studies revealed CTA ASPECTS or hypoattenuation on the CTA source image more reliably correlate with ischemic volume on MR DWI or final infarct size.9 19 Likewise, our result also revealed that CTA ASPECTS has a greater correlation with the ultimate infarct compared with NCCT ASPECTS (figure 1).

Although previous studies have shown strong relationships between CTA source image hypoattenuation and infarct volume, they have not demonstrated a correlation with clinical outcomes due to a small sample size or a lack of follow-up data.22 23 In this study, we analyzed the best cut-off point of NCCT and CTA ASPECTS for discriminating patients with favorable outcomes. Including these cut-off values, we also analyzed baseline and reperfusion data associated with good clinical outcome. A good CTA ASPECTS (≥8) was a predictor of favorable outcome in univariate analysis and remains an independent predictor in a multivariable analysis with a baseline NIHSS score <15 and successful recanalization (TICI 2bC). However, although a good NCCT ASPECTS (≥9) was a predictor of good outcome in univariate analysis, it was not statistically significant in multivariable regression. Although larger studies are needed to conclusively demonstrate this result, in light of this result and ROC analysis about the predictive value of a favorable outcome, we suggest that CTA should be considered as a priority in the decision-making process for EVT in patients with AIS.

Our study has limitations. The baseline median NIHSS score was 12, which was relatively lower compared with other studies about EVT in AIS patients. This result may be due to the inclusion in our data of not only internal carotid artery or M1 of MCA occlusion but also ACA or M2 of MCA occlusion. However, in view of a larger data set, this may not have influenced the result. In measuring the NCCT ASPECTS score, ischemic change was defined as not only tissue hypoattenuation and loss of gray-white matter differentiation but also sulcal effacement. However, a previous study showed that brain swelling without hypoattenuation on NCCT does not represent severe ischemic change, but it does suggest ischemic penumbral or oligemic tissue.24 So our NCCT-ASPECTS score may have been underestimated, but this may not have influenced the results.

Conclusion

Baseline CTA ASPECTS correlated better with the definitive infarction area than NCCT ASPECTS. Moreover, the predictability of favorable outcome using ASPECTS on CTA is superior to using NCCT ASPECTS in patients with anterior circulation AIS treated with EVT. These findings suggest that EVT decision-making may be more reliably based on CTA ASPECTS rather than NCCT ASPECTS in patients with anterior circulation AIS.

References

View Abstract

Footnotes

  • Contributors The four authors are justifiably credited with authorship, according to the authorship criteria. J-SP wrote the manuscript. H-SK, JML, and GHC critically reviewed the manuscript. All authors approved the final manuscript.

  • Funding This paper was supported by Fund of Biomedical Research Institute, Chonbuk National University Hospital.

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval Chonbuk National University Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement A technical appendix, statistical code, and dataset are available from the corresponding author. Participants gave informed consent for data sharing.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.