Predictive Value of Different Computed Tomography Perfusion Software Regarding 90-Day Outcome of Acute Ischemic Stroke Patients After Endovascular Treatment: A Comparison With Magnetic Resonance Imaging : Journal of Computer Assisted Tomography

Secondary Logo

Journal Logo

Neuroimaging: Brain

Predictive Value of Different Computed Tomography Perfusion Software Regarding 90-Day Outcome of Acute Ischemic Stroke Patients After Endovascular Treatment: A Comparison With Magnetic Resonance Imaging

Li, Ling BM∗,†; Jiang, Yun MD, PhD; Wang, Junjie MD§; Chen, Yuhui MD; Cao, Ruoyao MD; Lu, Yao BM∗,†; Wang, Guoxuan BM; Chen, Juan MD, PhD

Author Information
Journal of Computer Assisted Tomography 46(6):p 945-952, 11/12 2022. | DOI: 10.1097/RCT.0000000000001342
  • Free

Abstract

For patients with acute ischemic stroke (AIS), timely reperfusion therapy, which can effectively save recoverable tissues and improve the prognosis of patients, is very important.1,2 Recently, several randomized studies indicated that patients with AIS can be selected for endovascular treatment (EVT) up to 24 hours after symptom onset following the neuroimaging concept of a “tissue window” rather than “time window.”3–5 The ischemic core and penumbra volumes assessed using computed tomography perfusion (CTP) are critical for screening out EVT candidates, which would improve the prognosis and decrease complications.

Several studies have pointed out that there is a good agreement between the ischemic core volume on CTP and final infarct volume on magnetic resonance imaging (MRI).6,7 Currently, several commercial CTP software packages are available for semiautomated determination of ischemic core and penumbra volumes in patients with AIS based on different selected thresholds for different postprocessing algorithms. The main examples are the singular value decomposition plus (SVD+) algorithm, Bayesian algorithm, and maximum slope method.

A few studies have shown that the ischemic core volumes obtained using the SVD+ algorithm and Bayesian algorithm are highly consistent with infarct volume on MRI8,9; however, few studies have analyzed the relationship of core volumes with prognosis. In this regard, we compared the ischemic core volumes obtained using OleaSphere (Bayesian algorithm) and Vitrea software (SVD+ algorithm) with infarct volume on MRI and analyzed their relationship with prognosis in patients with AIS who received EVT.

MATERIALS AND METHODS

Study Population

This was a retrospective study, and all patient information was anonymized. Institutional review board approval was obtained, and the need for informed consent was waived. One hundred five patients from January 2016 to December 2020 were included in this study. All AIS patients with large vessel occlusion in the anterior circulation underwent 1-stop computed tomography angiography–CTP scans at our institution. The inclusion criteria were as follows: (1) patients with AIS within 24 hours of onset, (2) large vessel occlusion in the anterior circulation, (3) EVT performed within 24 hours after symptom onset, and (4) 2- to 7-day MRI (diffusion-weighted imaging [DWI]) data available. The following exclusion criteria were applied: (1) noncontrast CT indicating hemorrhage; (2) severe motion artifacts; (3) comorbid severe cardiac, hepatic, renal, or pulmonary disease; and (4) history of iodine allergy.

Imaging Acquisition

Aquilion ONE (Canon Medical Systems, Otawara, Japan) was used to obtain 1-stop whole-brain dynamic volume computed tomography angiography–CTP data, using a 320 × 0.5-mm detector row CT and covering 160 mm per rotation. Iopamidol or Omnipaque 370 (40–50 mL) was injected according to achieve an iodine concentration per weight of 0.6 mL/kg followed by 30 mL saline, using a 2-channel high-pressure injector. The tube voltage and current were 80 kVp/100 mAs, and Adaptive Iterative Dose Reduction 3-dimensional iterative reconstruction was used. Nineteen phases with a total of 6080 frames of dynamic volume data (320 × 0.5 mm) were acquired and transferred to the workstation (Vitrea, Vital Images, Minnetonka, Minnesota; OleaSphere, Olea Medical, La Ciotat, France) for postprocessing.

Diffusion-weighted imaging sequences were obtained on a Siemens (Erlangen, Germany) 3T or a GE (Chicago, Illinois) 3T MRI with an echo time of 50 to 70 milliseconds, repetition time of 2900 to 8000 milliseconds, b values of 0 and 1000 s/mm2, and slice thickness of 5 mm with a 1.5-mm gap.

CTP and MRI Analysis

Imaging data analysis was semiautomated using OleaSphere and Vitrea software. The ischemic core and penumbra volumes were obtained.

OleaSphere (v3.0) defines the ischemic core as a region with a relative cerebral blood flow less than 25% that of the contralateral side and with time to peak (TTP) >5 seconds that of the contralateral side (dTTP). The penumbra was defined as the area with a TTP >5 seconds than that of healthy tissue (excluding the infarct volume). The arterial input function and venous output function were autoselected by the software. Vitrea (version 6.7.2) defines the ischemic core as a region with relative cerebral blood volume less than 38% that of the contralateral side and with a 5.3-second increase in TTP. The penumbra was defined as the region where the TTP was 5.3 seconds greater than that of the healthy tissue (excluding the ischemic core volume).

OleaSphere software was applied to segment and measure the final infarct volume on MRI (DWI). After importing the data, we used the semiautomatic measurement to segment the infarct areas and obtained the final infarct volume (Fig. 1).

F1
FIGURE 1:
A patient with left internal carotid artery occlusion (white arrow) seen on (A) a maximum intensity projection image. B, Diffusion-weighted imaging shows the resulting region of restricted diffusion (white arrow). The final infarct area was subsequently semiautomatically segmented from the DWI (B) image, using OleaSphere software, which showed a volume of 117.78 mL. Color overlay images from (C) OleaSphere and (D) Vitrea show, in both, the ischemic core as red and the penumbra in yellow. The volume of the ischemic core as determined by the OleaSphere software was 95.54 mL and, as measured by the Vitrea software, was 80.50 mL. The core area on OleaSphere and Vitrea was consistent with the extent and spatial location of the infarct on DWI MRI. Figure 1 can be viewed online in color at www.jcat.org.

Clinical and Imaging Information

Clinical information was collected from medical records. Baseline National Institutes of Health Stroke Scale (NIHSS) score, Alberta Stroke program early CT score (ASPECTS), modified Thrombolysis in Cerebral Infarction (mTICI), and relevant medical history were collected. The modified Rankin Scale (mRS) score acquired after 90 days was used to evaluate the outcome of patients. An mRS score of 0 to 2 indicated a good outcome, whereas an mRS score of 3 to 6 indicated a poor outcome. A score of 6 was recorded if patients died during the follow-up period.

Statistical Analysis

Statistical analysis was performed with SPSS (26.0; IBM Corporation) and MedCalc (19.6; Mariakerke, Belgium). Continuous variables are described as the median (interquartile range [IQR]) for skewed data, and categorical data are represented as frequency distributions. The Mann-Whitney U, Pearson χ2, and Fisher exact tests were used to compare baseline information between the good and poor outcome groups. For ischemic core and penumbra volumes, correlation and agreement between OleaSphere, Vitrea, and MRI were assessed using the Spearman correlation test, intraclass correlation coefficient (ICC), and Bland-Altman plot. A correlation coefficient <0.30 was considered to be poor, 0.31 to 0.50 to be fair, 0.51 to 0.80 to be moderate, and >0.80 to be substantial. An ICC <0.4 was considered to be poor, 0.40 to 0.59 to be fair, 0.60 to 0.75 to be good, and >0.75 to be excellent.

Multivariate logistic regression analysis was used to identify prognostic value. Three models were developed: (1) clinical information (age, NIHSS, ASPECTS, mTICI) and MRI infarct volume, (2) clinical information (age, NIHSS, ASPECTS, mTICI) and OleaSphere ischemic core volume, and (3) clinical information (age, NIHSS, ASPECTS, mTICI) and Vitrea ischemic core volume. The DeLong test was used to test differences in the area under the curve (AUC) of the receiver operating characteristics (ROCs). We calculated the adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for different variables in relation to poor outcomes.

RESULT

Patient Characteristics

Patients' clinical and demographic data are summarized in Table 1 along with the analysis of good (mRS 0–2) and poor (mRS 3–6) outcomes. A total of 105 patients were enrolled in our study, including 59 men (56.2%) and 46 women (43.8%). The median age of the patients was 76.0 years. The median baseline NIHSS and ASPECTS scores were 11.0 (IQR, 7.0–17.0) and 7.0 (IQR, 5.0–8.0), respectively. Occlusion sites were in the internal carotid artery (37 [35.2%]), middle cerebral artery (65 [61.9%]), and anterior cerebral artery (3 [2.9%]). All 105 patients underwent EVT, and 95 of them (90.5%) achieved successful vascular recanalization (with mTICI classification of grade 2b–3).

TABLE 1 - Characteristic and Outcomes of All Patients With AIS
Characteristic All (n = 105) mRS 0–2 (n = 61) mRS 3–6 (n = 44) P
Sex, female 46 (43.8%) 22 (36.1%) 24 (54.5%) 0.060
Age, median (IQR), y 76.0 (63.0–83.0) 71.0 (60.5–80.0) 80.5 (68.0–84.0) 0.003
NIHSS, median (IQR) 11.0 (7.0–17.0) 9.0 (6.0–14.0) 15.0 (10.0–21.0) <0.001
ASPECTS, median (IQR) 7.0 (5.0–8.0) 8.0 (6.0–8.5) 7.0 (4.5–8.0) 0.048
Site of occlusion
 Internal carotid artery 37 (35.2%) 18 (29.5%) 19 (43.2%) 0.182
 Middle cerebral artery 65 (61.9%) 40 (65.6%) 25 (56.8%)
 Anterior cerebral artery 3 (2.9%) 3 (4.9%)
mTICI for thrombectomy cases
 0–2a 10 (9.5%) 2 (3.3%) 8 (18.2%) 0.016
 2b–3 95 (90.5%) 59 (96.7%) 36 (81.8%)
Clinical characteristic
 Atrial fibrillation 46 (43.8%) 21 (34.4%) 25 (56.8%) 0.023
 Hypertension 78 (74.3%) 43 (70.5%) 35 (79.5%) 0.368
 Diabetes mellitus 41 (39.0%) 22 (36.1%) 19 (43.2%) 0.544
 Hyperlipidemia 46 (43.8%) 23 (37.7%) 23 (52.3%) 0.165
 Coronary artery disease 41 (39.0%) 21 (34.4%) 20 (45.5%) 0.312
CTP and MRI findings
 Final MRI infarct volume, median (IQR), mL 31.1 (8.0–77.9) 20.5 (6.8–44.8) 80.3 (14.4–150.7) <0.001
 OleaSphere ischemic core volume, median (IQR), mL 27.5 (7.59–54.7) 22.0 (7.1–39.0) 48.7 (10.8–109.5) 0.001
Vitrea ischemic core volume, median (IQR), mL 26.9 (9.4–55.7) 21.6 (5.6–40.6) 51.1 (14.5–111.9) <0.001
 OleaSphere penumbra volume, median (IQR), mL 59.1 (27.8–93.8) 53.7 (22.2–87.9) 63.2 (40.5–99.5) 0.149
Vitrea penumbra volume, median (IQR), mL 82.4 (40.9–130.4) 83.7 (19.1–127.0) 78.2 (46.7–132.8) 0.465
Time intervals
 Time from onset of stroke to perfusion imaging, median (IQR), min 234.0 (121.0–423.0) 251.0 (127.5–429.5) 197.0 (86.3–411.3) 0.151
 Time from onset of stroke to treatment, median (IQR), min 315.0 (197.5–520.5) 352.0 (215.0–527.5) 262.0 (180.0–508.5) 0.141
 Time from onset of stroke to recanalization, median (IQR), min 455.0 (305.5–666.9) 460.0 (365.0–710.1) 447.0 (296.0–641.0) 0.197

Based on patients' outcomes, we divided them into good (61 patients) and poor (44 patients) outcome groups. As shown in Table 1, significant differences between the 2 groups in many variables were revealed, including age, NIHSS, ASPECTS, mTICI, atrial fibrillation, infarct volume on MRI, and ischemic core volumes on OleaSphere and Vitrea. Patients with good outcomes had a lower median age (71.0 [60.5–80.0] vs 80.5 [68.0–84.0] years, P < 0.01), lower NIHSS score (9.0 [6.0–14.0] vs 15.0 [10.0–21.0], P < 0.001), and higher median ASPECTS score (8.0 [6.0–8.5] vs 7.0 [4.5–8.0], P < 0.05). The rate of successful recanalization in the good outcome group was higher than that in the poor outcome group (96.7% vs 81.8%, P < 0.05).

Correlation and Agreement Analysis of Ischemic Core and Penumbra Volumes for All Patients Who Underwent EVT

Overall, the median core volumes obtained using MRI, OleaSphere, and Vitrea were 31.1, 27.5, and 26.9 mL, respectively. The median ischemic core volumes of OleaSphere and Vitrea were both smaller than the final infarct on MRI, but there was no significant difference (MRI vs OleaSphere, P = 0.362; MRI vs Vitrea, P = 0.502; OleaSphere vs Vitrea, P = 0.760). In the good outcome group, the median infarct volume on MRI was 20.5 mL, and the median ischemic core volumes on OleaSphere and Vitrea were 22.0 and 21.6 mL, respectively. In the poor outcome group, the median infarct volume was 80.3 mL on MRI, 48.7 mL on OleaSphere, and 51.1 mL on Vitrea. Patients with good outcomes had smaller core volumes than those in the poor group (P < 0.001) as measured using all 3 methods. Regarding the penumbra, the median volumes on OleaSphere and Vitrea were 59.1 and 82.4 mL, and there was a significant difference between the 2 methods (P = 0.009). For the 2 CTP software packages, there was no significant difference in penumbra volume between the good and poor groups.

The correlation and agreement between MRI- and CTP-measured volumes are shown in Figures 2A–D and 3A–D. Both ischemic core volumes obtained using OleaSphere (ρ = 0.84, P < 0.001) and Vitrea (ρ = 0.80, P < 0.001) showed substantial correlation with MRI, and the OleaSphere data were more closely correlated with MRI than those of Vitrea. Regarding agreement, OleaSphere (ICC = 0.84) and Vitrea (ICC = 0.83) ischemic core volumes showed excellent agreement with MRI. In addition, the core (ρ = 0.92, P < 0.001; ICC = 0.95) and penumbra (ρ = 0.87, P < 0.001; ICC = 0.77) volumes of OleaSphere and Vitrea had substantial correlation and excellent agreement (Table 2). The Bland-Altman plots confirmed interobserver reliability (Figs. 3A–D).

F2
FIGURE 2:
Correlation between MRI infarct and CTP ischemic core volumes and between penumbra volumes on OleaSphere and Vitrea in all patients (A–D) and patients with successful recanalization (mTICI 2b–3; E–H). Spearman correlation coefficients are shown on each graph. Figure 2 can be viewed online in color at www.jcat.org.
F3
FIGURE 3:
Bland-Altman plots of differences between MRI infarct and CTP ischemic core volumes, as well as differences between penumbra volumes on OleaSphere and Vitrea in all patients (A–D) and patients with successful recanalization (mTICI score, 2b–3; E–H). Figure 3 can be viewed online in color at www.jcat.org.
TABLE 2 - Spearman Correlation Coefficients and ICC on CTP and MRI Volumes
Comparison Software ρ ICC (95% CI)
All mTICI 2b–3 All mTICI 2b–3
Core volume MRI OleaSphere 0.84 0.84 0.84 (0.77–0.89) 0.85 (0.79–0.90)
Vitrea 0.80 0.80 0.83 (0.76–0.88) 0.86 (0.80–0.90)
OleaSphere Vitrea 0.92 0.91 0.95 (0.92–0.96) 0.96 (0.94–0.97)
Penumbra volume OleaSphere Vitrea 0.87 0.87 0.77 (0.68–0.84) 0.79 (0.70–0.85)

Patients Who Underwent Successful Vascular Recanalization (mTICI 2b–3)

In patients who underwent EVT but with poor recanalization (mTICI 0–2a), a larger final infarct volume was seen on MRI than ischemic core volume on CTP. As a result, we performed a subgroup analysis of patients who achieved successful recanalization to explore the relationship between ischemic core volume on CTP software and the final infarct volume on MRI.

The median core volumes obtained using MRI, OleaSphere, and Vitrea were 29.3, 26.9, and 25.9 mL in patients with an mTICI classification of grade 2b–3, which suggests that the OleaSphere and Vitrea ischemic core volumes are closer to the MRI infarct volumes in the successful recanalization group.

The correlation and agreement of MRI and CTP volumes in the successful recanalization group are shown in Figures 2E–H and 3E–H. In terms of Spearman correlation coefficients, the same coefficients were obtained between the successful recanalization group and all patient group, except for the correlation coefficient between OleaSphere and Vitrea (successful recanalization group, ρ = 0.91; all patients group, ρ = 0.92).

The ICC was improved for all parameters in the successful recanalization group (Table 2), with excellent agreement (>0.75) for both the core and penumbra. The parameters in the successful recanalization group showed better agreement, indicating that the difference between the MRI infarct volume and initial CTP ischemic core volume was smaller in the successful recanalization group.

Prognostic Value of Baseline Characteristics and Infarct Volumes on MRI and CTP

Table 3 summarizes the results of 3 multivariable models. The NIHSS score, mTICI, and core volumes were independent predictors of clinical outcome in each model.

TABLE 3 - Three Multivariable Models to Predict 90-Day Poor Outcome
Predictor OR With 95% CI
Model 1 (MRI Infarct) Model 2 (OleaSphere Core) Model 3 (Vitrea Core)
Age (OR per year) 1.03 (0.99–1.08) 1.03 (0.99–1.08) 1.03 (0.99–1.07)
NIHSS (OR per point) 1.13* (1.04–1.23) 1.13* (1.04–1.23) 1.12* (1.03–1.22)
ASPECTS (OR per point) 1.22 (0.88–1.69) 1.07 (0.79–1.45) 1.07 (0.79–1.45)
mTICI 7.37* (1.11–48.77) 8.45* (1.41–50.77) 9.90* (1.65–59.42)
Atrial fibrillation 1.18 (0.37–3.76) 0.87 (0.28–2.66) 0.90 (0.30–2.76)
MRI infarct volume 1.03* (1.01–1.04)
OleaSphere core volume 1.02* (1.01–1.04)
Vitrea core volume 1.02* (1.01–1.04)
Model statistics
 AUC (95% CI) 0.86* (0.78–0.92) 0.84* (0.75–0.90) 0.83* (0.75–0.90)
 DeLong test P value compared with model 1 0.18 0.18
*OR significantly different from 1 (P < 0.05).

Increasing the NIHSS score was associated with a significantly higher risk of poor prognosis (model 1 and model 2: odds ratio [OR], 1.13 [95% CI, 1.04–1.23]; model 3: OR, 1.12 [95% CI, 1.03–1.22]). There was a substantially greater probability of a poor outcome for patients with poor recanalization (mTICI 0–2a; model 1: OR, 7.37 [95% CI, 1.11–48.77]; model 2: OR, 8.45 [95% CI, 1.41–50.77]; model 3: OR, 9.90 [95% CI, 1.65–59.42]). Moreover, larger core volume was associated with an increased risk of poor prognosis (model 1: OR, 1.03 [95% CI, 1.01–1.04]; model 2 and model 3: OR, 1.02 [95% CI, 1.01–1.04]).

Receiver operating characteristic curve analysis (Fig. 4) of the AUC of all 3 models found no statistically significant difference (DeLong test; model 1 vs model 2, P = 0.18; model 1 vs model 3, P = 0.18) between model 1 (AUC, 0.86; 95% CI, 0.78–0.92), 2 (AUC, 0.84; 95% CI, 0.75–0.90) and 3 (AUC, 0.83; 95% CI, 0.75–0.90), suggesting that all 3 models have good predictive performance and were reliable in predicting 90-day unfavorable prognosis in patients with AIS.

F4
FIGURE 4:
Receiver operating characteristic curves of the 3 different multivariable models. Model 1 with MRI infarct volumes, model 2 with OleaSphere core volume, and model 3 with Vitrea core volume. Figure 4 can be viewed online in color at www.jcat.org.

Criterion Values and Coordinates of the ROC Analysis Regarding 90-Day Poor Outcome

Receiver operating characteristic analysis was performed for the statistically significant variables in each model, and the AUC, 95% CI, Youden Index, cutoff value, sensitivity, and specificity were obtained (Table 4). The optimal cutoff points and AUC of each parameter regarding the 90-day mRS are shown in the ROC curves in Figure 5.

TABLE 4 - Criterion Values and Coordinates of the ROC Analysis Regarding 90-Day Poor Outcome
Variable NIHSS MRI Infarct Volume OleaSphere Core Volume Vitrea Core Volume
AUC 0.75 0.73 0.69 0.71
95% CI 0.65–0.83 0.63–0.81 0.59–0.77 0.61–0.79
P <0.001 <0.001 0.001 <0.001
Youden Index 0.38 0.51 0.39 0.39
Cutoff value 13.00 60.12 49.18 45.51
Sensitivity 65.91 59.09 50.00 56.82
Specificity 72.13 91.80 88.52 81.97
DeLong test P value compared with NIHSS 0.81 0.34 0.52

F5
FIGURE 5:
Receiver operating characteristic curves for NIHSS score, MRI infarct volumes, and OleaSphere and Vitrea core volumes. Figure 5 can be viewed online in color at www.jcat.org.

The NIHSS score had the largest AUC (AUC, 0.75; 95% CI, 0.65–0.83), with a cutoff value of 13.00 for predicting a poor outcome. Magnetic resonance imaging infarct volumes (AUC, 0.73; 95% CI, 0.63–0.81; DeLong test, P = 0.81), OleaSphere ischemic volumes (AUC, 0.69; 95% CI, 0.59–0.77; DeLong test, P = 0.34), and Vitrea ischemic volumes (AUC, 0.71; 95% CI, 0.61–0.79; DeLong test, P = 0.52) showed no statistical difference in AUC compared with the NIHSS score. Receiver operating characteristic curve analysis showed that the optimal cutoff values that predicted a poor outcome were 60.12 mL (MRI), 49.18 mL (OleaSphere), and 45.51 mL (Vitrea), respectively (Table 4, Fig. 5).

DISCUSSION

This study analyzed the correlation and agreement of infarct, ischemic core, or penumbra volumes obtained from MRI and 2 CTP software with different postprocessing algorithms.8,9 Substantial correlation and excellent agreement were obtained between CTP ischemic core and MRI infarct volumes. Similarly, there was substantial correlation and excellent agreement between penumbra volumes assessed using OleaSphere and Vitrea. This suggested that ischemic core and penumbra volumes obtained using Bayesian and SVD+ algorithms are reliable, which is consistent with previous studies. We found that the median ischemic core volumes of OleaSphere and Vitrea were both smaller than the final infarct on MRI, which supports the findings of previous studies,10,11 possibly attributable to an increase in tissue volume during the time from the CTP scan to recanalization and to MRI.12,13 In addition, MRI infarct volumes may be overestimated in the subacute phase because of cerebral vasogenic edema caused by the influx of water and cations after the breakdown of the blood-brain barrier.14–16

We noted a significant difference between the median penumbra volume obtained using OleaSphere and Vitrea when using the Mann-Whitney U test, possibly because of differences in the algorithms. Further studies with larger samples are needed to define a “criterion standard” for the penumbra volume by which to compare these results. Nevertheless, we found good agreement between OleaSphere and Vitrea penumbra volumes using the Bland-Altman plot and ICC (Fig. 3D), indicating that either may be used in routine clinical work.8

In this study, patients in the good outcome group were younger, and had lower NIHSS scores, higher ASPECTS scores, a higher proportion with successful recanalization, and smaller final infarct size.17–19 We found that the NIHSS score, mTICI, CTP ischemic core volume, and MRI infarct volume were independent predictors of clinical outcome.20–22 The prognosis of stroke is associated with a variety of risk factors, including age, sex, subtype, hypertension, diabetes, coronary artery disease, hypercholesterolemia, smoking, alcohol consumption, severity, and treatment.23,24 The lack of statistically significant correlation between age, ASPECTS score, and clinical factors in our study may be due to the spatial location of infarcts, sample size, and choice of treatment.

Baseline NIHSS score reflects the degree of neurological impairment, and a higher NIHSS score is associated with a significantly higher risk of poor prognosis.18 The mTICI score is used to evaluate recanalization of the brain vessels after EVT, and TICI 2b/3 is defined as the successful recanalization, which reflects a good outcome.21 Thrombolysis in Cerebral Infarction 0–2a reflects poor recanalization, and affected patients are at a higher risk of poor outcomes. Several studies have revealed that MRI infarct volume is a strong predictor of prognosis in patients with AIS.25,26 Yoo et al27 found that preprocedure DWI volumes ≥70 mL were highly likely to yield poor clinical outcomes regardless of recanalization results, and our results back up their findings. Our study showed that there were no significant differences in the prognostic value of OleaSphere and Vitrea core volumes versus MRI to predict outcomes at 90 days, suggesting that ischemic core volumes obtained using CTP on admission may predict clinical outcomes, and the core volumes predicting poor prognosis were 60.12 mL (MRI), 49.18 mL (OleaSphere), and 45.51 mL (Vitrea). According to the HERMES collaboration meta-analysis,28 each 10-mL increase in baseline ischemic core volume on CTP reduced the likelihood of 90-day functional independence by 20% to 30% (OR, 0.77; 95% CI, 0.69–0.86). Our research supported the aforementioned viewpoints that preintervention CTP parameters were useful in predicting MRI infarct volume and 90-day outcome.

Some patients had MRI infarct volumes larger than CTP ischemic core volumes (Figs. 2, 3) and “outliners” attributable to malignant middle cerebral artery infarction (MMI), a complication in patients with large hemispheric infarction who progress to fatal brain edema. Thomalla's study showed that DWI lesion volume > 82 mL within 6 hours of symptom onset was an independent predictor of malignant middle cerebral artery infarction (sensitivity, 52%; specificity, 98%).29–31

There were some limitations in our study. First, this was a retrospective study with patients from a single stroke center. Second, the infarct volumes may have been overestimated because of the presence of vasogenic and postoperative edema. Third, the study included only patients with EVT.

In conclusion, the ischemic core volumes measured using OleaSphere and Vitrea showed moderate correlation and good agreement with MRI. Moreover, there were no significant differences in the predictive value of the 3 models (MRI, OleaSphere, and Vitrea), suggesting that we can use the core volumes obtained using the 2 CTP software packages with different algorithms to predict the prognosis of patients with AIS.

ACKNOWLEDGMENT

We would like to thank Editage (www.editage.com) for English language editing.

REFERENCES

1. Todo K, Sakai N, Imamura H, et al. Successful reperfusion with endovascular therapy has beneficial effects on long-term outcome beyond 90 days. Cerebrovasc Dis. 2019;47:127–134.
2. Hacke W, Kaste M, Bluhmki E, et al. Thrombolysis with alteplase 3 to 4.5 hours after acute ischemic stroke. N Engl J Med. 2008;359:1317–1329.
3. Campbell BCV, Mitchell PJ, Kleinig TJ, et al. Endovascular therapy for ischemic stroke with perfusion-imaging selection. N Engl J Med. 2015;372:1009–1018.
4. Nogueira RG, Jadhav AP, Haussen DC, et al. Thrombectomy 6 to 24 hours after stroke with a mismatch between deficit and infarct. N Engl J Med. 2018;378:11–21.
5. Albers GW, Marks MP, Kemp S, et al. Thrombectomy for stroke at 6 to 16 hours with selection by perfusion imaging. N Engl J Med. 2018;378:708–718.
6. Koopman MS, Berkhemer OA, Geuskens RREG, et al. Comparison of three commonly used CT perfusion software packages in patients with acute ischemic stroke. J Neurointerv Surg. 2019;11:1249–1256.
7. Xiong Y, Huang CC, Fisher M, et al. Comparison of automated CT perfusion Softwares in evaluation of acute ischemic stroke. J Stroke Cerebrovasc Dis. 2019;28:104392.
8. Rava RA, Snyder KV, Mokin M, et al. Assessment of a Bayesian vitrea CT perfusion analysis to predict final infarct and penumbra volumes in patients with acute ischemic stroke: a comparison with RAPID. Am J Neuroradiol. 2020;41:206–212.
9. Rava RA, Snyder KV, Mokin M, et al. Assessment of computed tomography perfusion software in predicting spatial location and volume of infarct in acute ischemic stroke patients: a comparison of Sphere, Vitrea, and RAPID. J Neurointerv Surg. 2021;13:130–135.
10. Sasaki M, Kudo K, Boutelier T, et al. Assessment of the accuracy of a Bayesian estimation algorithm for perfusion CT by using a digital phantom. Neuroradiology. 2013;55:1197–1203.
11. Ichikawa S, Yamamoto H, Morita T. Comparison of a Bayesian estimation algorithm and singular value decomposition algorithms for 80-detector row CT perfusion in patients with acute ischemic stroke. Radiol Med. 2021;126:795–803.
12. Lin L, Yang J, Chen C, et al. Association of collateral status and ischemic core growth in patients with acute ischemic stroke. Neurology. 2021;96:161–170.
13. Kim SM, Kwon SU, Kim JS, et al. Early infarct growth predicts long-term clinical outcome in ischemic stroke. J Neurol Sci. 2014;347:205–209.
14. Broocks G, Kemmling A, Meyer L, et al. Computed tomography angiography collateral profile is directly linked to early edema progression rate in acute ischemic stroke. Stroke. 2019;50:3424–3430.
15. Broocks G, Faizy TD, Flottmann F, et al. Subacute infarct volume with edema correction in computed tomography is equivalent to final infarct volume after ischemic stroke: improving the comparability of infarct imaging endpoints in clinical trials. Invest Radiol. 2018;53:472–476.
16. Broocks G, Hanning U, Faizy TD, et al. Ischemic lesion growth in acute stroke: water uptake quantification distinguishes between edema and tissue infarct. J Cereb Blood Flow Metab. 2020;40:823–832.
17. Goyal M, Menon BK, Coutts SB, et al. Effect of baseline CT scan appearance and time to recanalization on clinical outcomes in endovascular thrombectomy of acute ischemic strokes. Stroke. 2011;42:93–97.
18. Wirtz MM, Hendrix P, Goren O, et al. Predictor of 90-day functional outcome after mechanical thrombectomy for large vessel occlusion stroke: NIHSS score of 10 or less at 24 hours. J Neurosurg. 2021;134:115–121.
19. Cao R, Qi P, Liu Y, et al. Improving prognostic evaluation by 4D CTA for endovascular treatment in acute ischemic stroke patients: a preliminary study. J Stroke Cerebrovasc Dis. 2019;28:1971–1978.
20. Sucharew H, Khoury J, Moomaw CJ, et al. Profiles of the National Institutes of Health Stroke Scale items as a predictor of patient outcome. Stroke. 2013;44:2182–2187.
21. Yoo AJ, Simonsen CZ, Prabhakaran S, et al. Refining angiographic biomarkers of revascularization: improving outcome prediction after intra-arterial therapy. Stroke. 2013;44:2509–2512.
22. Barrett KM, Ding YH, Wagner DP, et al. Change in diffusion-weighted imaging infarct volume predicts neurologic outcome at 90 days results of the acute stroke accurate prediction (ASAP) trial serial imaging substudy. Stroke. 2009;40:2422–2427.
23. Tsai CF, Anderson N, Thomas B, et al. Risk factors for ischemic stroke and its subtypes in Chinese vs. Caucasians: systematic review and meta-analysis. Int J Stroke. 2015;10:485–493.
24. Matsumoto K, Nohara Y, Soejima H, et al. Stroke prognostic scores and data-driven prediction of clinical outcomes after acute ischemic stroke. Stroke. 2020;51:1477–1483.
25. Albers GW, Thijs VN, Wechsler L, et al; DEFUSE Investigators. 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;60:508–517.
26. Zaidi SF, Aghaebrahim A, Urra X, et al. Final infarct volume is a stronger predictor of outcome than recanalization in patients with proximal middle cerebral artery occlusion treated with endovascular therapy. Stroke. 2012;43:3238–3244.
27. Yoo AJ, Verduzco LA, Schaefer PW, et al. MRI-based selection for intra-arterial stroke therapy: value of pretreatment diffusion-weighted imaging lesion volume in selecting patients with acute stroke who will benefit from early recanalization. Stroke. 2009;40:2046–2054.
28. Campbell BCV, Majoie CBLM, Albers GW, et al. HERMES collaborators. Penumbral imaging and functional outcome in patients with anterior circulation ischaemic stroke treated with endovascular thrombectomy versus medical therapy: a meta-analysis of individual patient-level data. Lancet Neurol. 2019;18:46–55.
29. Treadwell SD, Thanvi B. Malignant middle cerebral artery (MCA) infarction: pathophysiology, diagnosis and management. Postgrad Med J. 2010;86:235–242.
30. Hacke W, Schwab S, Horn M, et al. ‘Malignant’ middle cerebral artery territory infarction: clinical course and prognostic signs. Arch Neurol. 1996;53:309–315.
31. Thomalla G, Hartmann F, Juettler E, et al. Prediction of malignant middle cerebral artery infarction by magnetic resonance imaging within 6 hours of symptom onset: a prospective multicenter observational study. Ann Neurol. 2010;68:435–445.
Keywords:

MRI; CT perfusion; stroke; OleaSphere; Vitrea

Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.