You are here

Article Text

Article menu

Download PDFPDF

Cerebrovascular disease
Original research
Novel selection paradigms for endovascular stroke treatment in the extended time window
Free
  1. Mehdi Bouslama1,2,
  2. Diogo C Haussen1,2,
  3. Gabriel Rodrigues1,2,
  4. Clara Barreira1,2,
  5. Michael Frankel1,2,
  6. Raul G Nogueira1,2
  1. 1 Neurology, Emory University, Atlanta, Georgia, USA
  2. 2 Marcus Stroke and Neuroscience Center, Grady Memorial Hospital, Atlanta, GA, USA
  1. Correspondence to Dr Raul G Nogueira, Neurology and Interventional Neuroradiology, Emory University, Atlanta, GA 30303, USA; rnoguei{at}emory.edu

Abstract

Background and purpose The optimal selection methodology for stroke thrombectomy beyond 6 hours remains to be established.

Methods Review of a prospectively collected database of thrombectomy patients with anterior circulation strokes, adequate CT perfusion (CTP) maps, National Institute of Health Stroke Scale (NIHSS)≥10 and presenting beyond 6 hours from January 2014 to October 2018. Patients were categorised according to five selection paradigms: DAWN clinical-core mismatch (DAWN-CCM): between age-adjusted NIHSS and CTP core, DEFUSE 3 perfusion imaging mismatch (DEFUSE-3-PIM): between CTP-derived perfusion defect (Tmax >6 s lesion) and ischaemic core volumes and three non-contrast CT Alberta Stroke Program Early CT Score (ASPECTS)-based criteria: age-adjusted clinical-ASPECTS mismatch (aCAM): between age-adjusted NIHSS and ASPECTS, eloquence-adjusted clinical ASPECTS mismatch (eCAM): ASPECTS 6–10 and non-involvement of the right M6 and left M4 areas and standard clinical ASPECTS mismatch (sCAM): ASPECTS 6–10.

Results 310 patients underwent analysis. DEFUSE-3-PIM had the highest proportion of qualifying patients followed by sCAM, eCAM, aCAM and DAWN-CCM (93.5%, 92.6%, 90.6%, 90% and 84.5%, respectively). Patients meeting aCAM, eCAM, sCAM and DAWN-CCM criteria had higher rates of 90-day good outcome compared with their non-qualifying counterparts(43.2% vs 12%,p=0.002; 42.4% vs 17.4%, p=0.02; 42.4% vs 11.2%, p=0.009; and 43.7% vs 20.5%, p=0.007, respectively). There was no difference between patients meeting DEFUSE-3-PIM criteria versus not(40.8% vs 31.3%,p=0.45). In multivariate analysis, all selection modalities except for DEFUSE-3-PIM were independently associated with 90-day good outcome.

Conclusions ASPECTS-based selection paradigms for late presenting and wake-up strokes ET have comparable proportions of qualifying patients and similar 90-day functional outcomes as DAWN-CCM and DEFUSE-3-PIM. They also might lead to better outcome discrimination. These could represent a potential alternative for centres where access to advanced imaging is limited.

Data availability statement

Data are available upon reasonable request. The data that support the findings of this study are available from the corresponding author upon reasonable request.

http://dx.doi.org/10.1136/jnnp-2020-325284

Statistics from Altmetric.com

    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.

    Introduction

    The benefit of thrombectomy in large vessel acute ischaemic stroke (LVOS) in late presenting and wake-up strokes is now well established. The landmark DAWN (clinical mismatch in the triage of wake up and late presenting strokes undergoing neuro-intervention with Trevo)1 and DEFUSE 3 (endovascular therapy following imaging evaluation for ischemic stroke 3)2 trials have proven the superiority of endovascular therapy (ET) to medical management in patient presenting between 6 and 24 hours from last known well. However, they used different selection methodologies leaving the question on optimal selection paradigms wide open.3 Moreover, both relied on advanced imaging to determine eligibility for ET, which may limit access to treatment in centres where those imaging modalities are not readily available.

    We sought to evaluate different simple non-contrast CT (NCCT)-based paradigms and assess their selection rates and ability to discriminate outcomes as compared with the DAWN and DEFUSE 3 criteria.

    Methods

    Patients selection and measures of outcomes

    We retrospectively reviewed our prospectively collected database at a tertiary care academic centre for all mechanical thrombectomies performed between January 2014 and October 2018. Our institutional protocol for endovascular stroke treatment in the extended window does not include any prespecified imaging criteria other than definite loss of gray-white matter differentiation involving large areas of eloquent cortex. Specifically, we do not employ any specific Alberta Stroke Program Early CT Score (ASPECTS) or CT perfusion (CTP) parameter cut-offs in isolation to exclude patients. In the current analysis, we included all consecutive acute anterior circulation LVOS patients presenting beyond the 6-hour window with who had adequate CTP imaging maps. In addition, we limited the baseline stroke severity to National Institute of Health Stroke Scale (NIHSS)≥10, as patients with a lower NIHSS have not yet been definitely shown to benefit from treatment.2 4 Patients were then retrospectively categorised according to following clinical-imaging criteria:

    1. DAWN clinical core mismatch (DAWN-CCM) positive was defined by one of the following:

      • NIHSS≥10 and core infarct <31 cc (and age <80 years old).

      • NIHSS≥20 and core infarct <51 cc (and age <80 years old).

      • NIHSS≥10 and core infarct <21 cc (and age ≥80 years old).

    2. DEFUSE 3 perfusion imaging mismatch (DEFUSE-3-PIM) positive patients met all the following criteria:

      • CTP-derived infarction core lesion <70 cc.

      • CTP mismatch (Tmax >6 s lesion – infarction core lesion)≥15 cc and ratio >1.8.

    3. Age-adjusted clinical ASPECTS mismatch (aCAM) positive was defined as ASPECTS 6–10 on non-contrast CT and one of the following:

      • NIHSS≥10 and 0–1 cortical-ASPECTS (M1-6 areas) involvement (any age).

      • NIHSS≥10 and 0–2 cortical-ASPECTS (M1-6 areas) involvement (and age <80 years old).

      • NIHSS≥20 and 0–3 cortical-ASPECTS (M1-6 areas) involvement (and age <80 years old).

    4. Eloquence-adjusted clinical ASPECTS mismatch (eCAM) positive was defined as ASPECTS 6–10 on non-contrast CT and non-involvement of the right M6 and left M4 areas regardless of age and NIHSS strata.5

    5. Standard clinical ASPECTS mismatch (sCAM) positive was defined as ASPECTS 6–10 on non-contrast CT regardless of age and NIHSS strata.

    The primary outcome measures included the proportion of qualifying patients per the different selection paradigms and the ability of the latter to discriminate 90-day good outcome (modified Rankin scale, mRS 0–2). Secondary outcomes included rates of successful reperfusion (mTICI 2b-3). Safety parameters comprised 90-day mortality and rates of any parenchymal haematoma (any PH).

    Imaging protocol

    All patients included in the study underwent the same institutional imaging protocol, including NCCT and CTP. Imaging acquisition parameters were the same for all patients included in the study. Large vessel occlusion was documented on CT angiography or conventional angiography for all patients.

    CT perfusion

    CTP encompassing 8 cm of brain coverage was evaluated with by a fully automated software (RAPID V.4.5.0, iSchemaView, Menlo Park, California). The ischaemic core was defined by a voxel relative cerebral blood flow of <30% of the contralateral normal tissue. The total hypoperfused volume was defined by a>6 s delay in the time to maximum of the tissue residue function (Tmax), and the volume of at-risk tissue defined by the difference between total hypoperfused and ischaemic core tissue estimates. The RAPID software is used as part of our centre’s clinical stroke protocol.

    e-Stroke Suite ASPECTS

    The ASPECTS score was retrospectively calculated on baseline NCCT by the e-ASPECTS tool (e-Stroke Suite V.8.0; Brainomix, Oxford, UK, www.brainomix.com). e-ASPECTS is the NCCT module of the e-Stroke Suite and preferentially processes thin slice non-contrast CT images (<2 mm slice thickness), resampling and standardising the Digital Imaging and Communications in Medicine data. A machine learning classifier, which has been trained on historical stroke and negative control data, segments the image and creates a voxel-wise probability map of the early ischaemic change in the MCA territory. This map is then used to calculate the equivalent ASPECT score. The e-Stroke Suite software was provided for free under a limited research license.

    Statistical analysis

    Normality was assessed with Shapiro-Wilk test and continuous variables were reported as mean±SD or median (IQR), as appropriate. Categorical variables were reported as proportions. Between groups, comparisons for continuous variables were made with Student’s t-test or Mann-Whitney U, as appropriate. Categorical variables were compared by χ2 test, Fisher’s exact test, or McNemar test for discordant pairs, as appropriate.

    The ability of the individual selection tool to predict 90-day good outcome was assessed by constructing separate binomial logistic regression models with each selection paradigm as predictor variable and controlling for variables for variables at the 0.1 level of significance on univariate analysis. Significance was set at p<0.05 and all p values were two sided. Statistical analysis was performed using IBM SPSS Statistics 24 (IBM).

    Results

    Primary analysis

    A total of 310 patients fit inclusion criteria (figure 1). Baseline characteristics are summarised in table 1. The median (IQR) age was 67 (55–76), 20% were 80 years old or older and 46.8% of the patients were male. The median (IQR) baseline NIHSS and e-ASPECTS were 17 (14–21) and 8 (7–9), respectively. The median (IQR) time from last seen normal to arterial puncture was 645 (478–900) min with 15 patients (4.8%) presenting beyond 24 hours.

    Figure 1
    Figure 1

    Study sample. ASPECTS, Alberta Stroke Program Early CT Score; LVOS, large vessel acute ischaemic stroke; NCCT, non-contrast CT.

    View this table:
    Table 1

    Overall baseline characteristics and outcome measures of the study population

    DEFUSE-3-PIM (290/310, 93.5%) had the highest numbers of qualifying patients followed sCAM (287/310, 92.6%), eCAM (281/310, 90.6%); aCAM (279/310, 90%) and DAWN-CCM (262/310, 84.5%). Agreement/disagreement between the different selection modalities is summarised in table 2 and figure 2. DAWN-CCM and sCAM had statistically different selected patients when compared with the other paradigms two by two.

    View this table:
    Table 2

    Disagreement between the different selection paradigms

    Figure 2
    Figure 2

    Venn diagrams for the different selection modalities. ASPECTS, Alberta StrokeProgram Early CT score; aCAM indicates age-adjusted clinical-ASPECTS mismatch; eCAM, eloquence-adjusted clinical-ASPECTS mismatch; sCAM, standard clinical-ASPECTS mismatch.

    Ninety-day outcome data were available in 254/310 patients (81.9%). The rate of 90-day good outcomes for the entire cohort was 40.2% and overall mortality was 24.8%. The aCAM, eCAM, sCAM and DAWN-CCM paradigms were all associated with 90-day good outcomes (when comparing patients who met the selection criteria vs those that did not (absolute difference, Δ): 43.2% vs 12% (Δ+31.2%), p=0.002; 42.4% vs 17.4% (Δ+25%), p=0.02; 42.4% vs 11.2% (Δ+31.2%), p=0.009; and 43.7% vs 20.5% (Δ+23.2%), p=0.007, respectively). However, there was no difference between those that did qualify for DEFUSE-3-PIM criteria versus not: 40.8% versus 31.3% (Δ+9.5%), p=0.45.

    In addition, aCAM was associated with lower rates of any PH and 90-day mortality (8.6% vs 22.6%; Δ −14%, p=0.02% and 22.7% vs 44%; Δ −21.3%, p=0.02, respectively).

    There was a non-significant trend towards lower rates of any PH in the other ASPECTS-based selection criteria (eCAM: 8.9% vs 20.7%; Δ −12%, p=0.05; and sCAM: 9.1% vs 21.7%; Δ −12.6%, p=0.07) while there was no difference between DEFUSE-3-PIM, DAWN-CCM and eCAM positive versus negative patients in terms of mortality (figure 3).

    Figure 3
    Figure 3

    Outcome measures per selection paradigm. Significant results are highlighted in green. mRS indicates modified Rankin Scale; mTICI, modified Thrombolysis in Cerebral Infarction; PH, parenchymal haematoma.

    In multivariate analysis, adjusting for age, NIHSS, glucose and atrial fibrillation, all selection modalities except for DEFUSE-3-PIM were independent predictors of 90-day good outcomes: aCAM(+) (aOR 6.93, 95% CI [1. 58–30.51], p=0.01), DAWN-CCM(+) (aOR 3.34, 95% CI [1.39 to 8.03], p=0.007), eCAM(+) (aOR 4.15, 95% CI [1.16 to 14.84], p=0.029), sCAM(+) (aOR 11.63, 95% CI [1.55 to 86.97], p=0.017) and DEFUSE-3-PIM(+) (aOR 2.25, 95% CI [0.69 to 7.38], p=0.181).

    Additionally, after adjusting for intravenous thrombolysis, only aCAM was associated with any PH (aOR 0.36, 95% CI 0.16 to 0.86). None of the studied selection paradigms were associated with 90-day mortality after adjusting for potential confounders.

    Sensitivity analysis

    For sensitivity analysis, only patients with an intracranial ICA, MCA M1 or M2 occlusion and complete 90-day outcomes were included (n=205). DEFUSE-3-PIM and sCAM had the highest numbers of included patients (192/205, 93.7%) followed by eCAM (187/310, 91.2%); aCAM (186/310, 90.7%) and DAWN-CCM (176/205, 85.9%).

    In terms of outcomes only DAWN-CCM and aCAM paradigms were associated with 90-day good outcomes when comparing patients who met the selection criteria vs those who did not (absolute difference, Δ) (42% vs 20.7% (Δ+21.3%), p=0.029% and 41.4% vs 15.8% (Δ+25.6%), p=0.029, respectively). There was a trend of favourable outcome in patients fulfilling sCAM criteria (40.6% vs 15.4% (Δ+25.2%), p=0.071) while there was no difference between those that did qualify for DEFUSE-3-PIM and eCAM selection criteria versus not (39.1% vs 38.5% (Δ+0.6%), p=0.97% and 40.6% vs 22.2% (Δ+18.4%), p=0.126, respectively).

    In multivariate analysis, adjusting for age, NIHSS, glucose and HTN, all selection modalities except for DEFUSE-3-PIM and eCAM were independent predictors of 90-day good outcomes: aCAM(+) (aOR 5.23, 95% CI 1.38 to 24.09, p=0.034), DAWN-CCM(+) (aOR 2.91, 95% CI 1.07 to 7.92, p=0.037), eCAM(+) (aOR 2.94, 95% CI 0.78 to 11.05, p=0.11), sCAM(+) (aOR 8.59, 95% CI 1.08 to 68.27, p=0.047) and DEFUSE-3-PIM(+) (aOR 2.28, 95% CI 0.36 to 4.6, p=0.702). Notably, 12 (92.3%) out of the 13 DEFUSE-3-PIM(−) patients fulfilled criteria by at least one selection modality that was independently associated with good outcomes (eg, aCAM, DAWN-CCM or sCAM). On the other hand, only five patients did not meet any of the aCAM, DAWN-CCM or sCAM criteria, of which 4 DEFUSE-3-PIM (+) with only one having a good outcome at 90 days.

    None of the selection modalities were associated with any PH or 90-day mortality in multivariate analysis.

    Discussion

    Our study shows that, when comparing patients that met the different paradigm criteria versus those that did not, selection in late presenting or wake-up strokes based on itemised ASPECTS resulted in similar rates of good procedural, functional and safety outcomes as compared with the DAWN and DEFUSE 3 criteria. Moreover, all ASPECTS paradigms had a high proportion of qualifying patients and performed better in terms of outcome discrimination when compared with DEFUSE 3. In fact, patients without a perfusion imaging mismatch had similar clinical outcomes as those who met the DEFUSE 3 criteria. This held true in multivariate analysis adjusting for potential confounders where DEFUSE-3-PIM did not show to be independently associated with good outcome at 90-days suggesting that PIM might be a poor discriminator of treatment response. On the other hand, the DAWN-CCM and the ASPECTS based paradigms were independently associated with 90-day good outcome.

    The AHA/ASA guidelines extending the thrombectomy treatment window state that either the DAWN or DEFUSE 3 trial criteria could be used to determine eligibility for ET in that patient population.3 It is important to recognise that both required the use of MRI or CTP to assess infarct core volume (DAWN and DEFUSE 3) and hypoperfusion/imaging mismatch (DEFUSE 3). These ‘high tech’ approaches are not typically available in clinical settings where resources are more limited. Therefore, in the absence of simpler selection paradigms, ET may not be offered to many patients that could still benefit from treatment. Additionally, there is still debate with regards to whether the use of advanced imaging is associated with improved clinical outcomes post thrombectomy. A meta-analysis of 10 clinical trials by Tsivgoulis et al showed that the use of advanced neuroimaging was associated with higher treatment effects as compared with conventional imaging,6 whereas a more recent analysis of the Trevo registry failed to reveal any benefit of an imaging selection modality over another in the early and extended treatment windows.7 Moreover, given the potential higher radiation, contrast-induced kidney injury and longer times to treatment,8 identifying and validating selection modalities relying on simpler imaging tools become paramount.

    There is growing evidence that using NCCT and ASPECTS might be safe and effective in selecting late presenting stroke patients for ET.9–11 Recently, Santos et al in a cohort of 249 patients found that patients with wake-up or late presenting strokes selected with a clinical-ASPECTS mismatch paradigm had similar outcomes as those that presented within 6 hours from last known well.9 Similarly, Nagel et al did not find any differences in clinical outcomes between patients with and without a perfusion imaging mismatch. Moreover, in multivariate analysis, the performance of MRI or CTP parameters did not influence outcome and ASPECTS was the only imaging parameter associated with good outcome.10 Our results are also in agreement with a recent study demonstrating that the prevalence of DAWN-CCM among patients within the same ASPECTS categories (9–10 vs 6–8 vs 0–5) does not decline over time, further reinforcing an ASPECTS-based paradigm for patient selection in the extended window.12

    ASPECTS is a simple and widely used tool in clinical practice for patient selection. Several studies have shown it to be a reliable predictor of outcomes.13–15 However, despite its speed and ubiquity, ASPECTS has major shortcomings including poor ability to predict baseline infarction volumes16 and low inter-rater agreement.17 While these are valid concerns and could be problematic when caring for patients presenting early in the treatment window, ASPECTS has been shown to be reliable in identifying infarct core in late presenting strokes.18 Similarly, the inter-rater agreement improves as time elapses.19A major flaw in relation to the ASPECTS concept derives from the fact that ASPECTS is not a linear scale since, as compared with the cortical regions, the deep areas of the brain are lower in both volume and eloquence (the posterior limb of the internal capsule representing an important exception).20–22 Our two newly proposed ASPECTS based selection paradigms have therefore adjusted for this critical issue and should be further explored in future larger prospective studies.

    In our study, for consistency, we relied on an automated ASPECTS assessment (e-ASPECTS, e-Stroke Suite V.8.0; Brainomix, Oxford, UK, www.brainomix.com). The software has been proven to be non-inferior to expert reader’s performance23 and e-ASPECTS were found to be correlated with stroke severity and clinical outcomes.23 Given that automated software may surpass the human eye in neuroimaging assessment,24 the use of an automated ASPECTS software allowed us to focus on the merit of each selection paradigm and limit the confounding effect of ‘human’ versus ‘machine’ measurement. Our approach, using the itemised score, incorporated stroke topology which we believe refines selection criteria. As discussed above, the anatomic regions covered by ASPECTS have been reported to be unequally weighted25 and topological information has been shown to affect clinical outcome over and above the effect of infarct size.5 Another interesting finding is the potential superiority of aCAM to the eCAM ASPECTS selection tool as demonstrated by the larger absolute difference in good outcomes between those meeting the criteria and their negative counterparts. Age and core volume have been repeatedly identified as strong predictors of outcomes and we have shown in previous studies not only that age-adjusted selection paradigms were superior to non-adjusted approaches in outcome discrimination26 but also that final infarct volume cut-offs to predict outcomes decreased with age.27 28 In the current analysis, we failed to demonstrate any definite advantage of these adjusted paradigms over the simpler sCAM in predicting 90-day good outcome. Moreover, our data hint that the adoption of an aCAM might lead to refined haemorrhagic transformation (any PH) and mortality discrimination.

    Our study has several limitations mostly inherent to its retrospective design and its relatively limited sample size. Additionally, the absence of a control group comprising untreated patients limits out ability to measure treatment effects in comparison to each selection-positive group. However, we believe this absence underestimates the discernment between favourable and unfavourable outcomes, as all patients in this study did receive ET, increasing the likelihood of favourable outcomes and biasing analysis towards the null. Another limitation associated with our study is that the increase in familywise error rate across the reported statistical analysis was not corrected. We consider our research hypothesis generating as only prospective randomised controlled trials will be able to properly address this important clinical dilemma. Despite these limitations, our analysis suggests that using ASPECTS-based selection methods for late presenting and wake-up strokes might result in similar proportions of treated patients, with comparable rates of good outcomes to the DAWN and DEFUSE 3 criteria, and refined outcome discrimination. Fortunately, the hypothesis that simpler imaging paradigms can be used in the selection for endovascular treatment in the late window is currently being tested in two randomised studies (MR CLEAN LATE, https://www.mrclean-late.nl; RESILIENT-Extend, ClinicalTrials.gov Identifier: NCT04256096).

    Conclusions

    Non-contrast CT ASPECTS-based selection paradigms for late presenting and wake-up strokes ET have comparable proportions of qualifying patients and similar 90-day functional outcomes as DAWN-CCM and DEFUSE-3-PIM. They also might lead to better outcome discrimination as compared with DEFUSE-3 criteria. These could represent a potential alternative for centres where access to advanced imaging is limited. Future prospective studies are warranted.

    Data availability statement

    Data are available upon reasonable request. The data that support the findings of this study are available from the corresponding author upon reasonable request.

    Ethics statements

    Patient consent for publication

    Ethics approval

    Local institutional review board (IRB)/Emory University IRB.

    References

    1. ↵
      2. Nogueira RG ,
      3. Jadhav AP ,
      4. 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.doi:10.1056/NEJMoa1706442 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29129157
      OpenUrlCrossRefPubMed
    2. ↵
      2. Albers GW ,
      3. Marks MP ,
      4. Kemp S , et al
      . Thrombectomy for stroke at 6 to 16 hours with selection by perfusion imaging. N Engl J Med 2018;378:708–18.doi:10.1056/NEJMoa1713973 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29364767
      OpenUrlCrossRefPubMed
    3. ↵
      2. Powers WJ ,
      3. Rabinstein AA ,
      4. Ackerson T , et al
      . 2018 guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American heart Association/American stroke association. Stroke 2018;49:e46–110.doi:10.1161/STR.0000000000000158 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29367334
      OpenUrlAbstract/FREE Full Text
    4. ↵
      2. Goyal M ,
      3. Menon BK ,
      4. van Zwam WH , et al
      . Endovascular thrombectomy after large-vessel ischaemic stroke: a meta-analysis of individual patient data from five randomised trials. Lancet 2016;387:1723–31.doi:10.1016/S0140-6736(16)00163-X pmid:http://www.ncbi.nlm.nih.gov/pubmed/26898852
      OpenUrlCrossRefPubMed
    5. ↵
      2. Rangaraju S ,
      3. Streib C ,
      4. Aghaebrahim A , et al
      . Relationship between lesion topology and clinical outcome in anterior circulation large vessel occlusions. Stroke 2015;46:1787–92.doi:10.1161/STROKEAHA.115.009908 pmid:http://www.ncbi.nlm.nih.gov/pubmed/26060248
      OpenUrlAbstract/FREE Full Text
    6. ↵
      2. Tsivgoulis G ,
      3. Katsanos AH ,
      4. Schellinger PD , et al
      . Advanced neuroimaging in stroke patient selection for mechanical thrombectomy. Stroke 2018;49:3067–70.doi:10.1161/STROKEAHA.118.022540 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30571421
      OpenUrlCrossRefPubMed
    7. ↵
      2. Nogueira RG ,
      3. Haussen DC ,
      4. Liebeskind D , et al
      . Stroke imaging selection modality and endovascular therapy outcomes in the early and extended time windows. Stroke 2021;52:491–7.doi:10.1161/STROKEAHA.120.031685 pmid:http://www.ncbi.nlm.nih.gov/pubmed/33430634
      OpenUrlCrossRefPubMed
    8. ↵
      2. Sheth KN ,
      3. Terry JB ,
      4. Nogueira RG , et al
      . Advanced modality imaging evaluation in acute ischemic stroke may lead to delayed endovascular reperfusion therapy without improvement in clinical outcomes. J Neurointerv Surg 2013;5 Suppl 1:i62–5.doi:10.1136/neurintsurg-2012-010512 pmid:http://www.ncbi.nlm.nih.gov/pubmed/23076268
      OpenUrlPubMed
    9. ↵
      2. Santos T ,
      3. Carvalho A ,
      4. Cunha AA , et al
      . NCCT and CTA-based imaging protocol for endovascular treatment selection in late presenting or wake-up strokes. J Neurointerv Surg 2019;11:200–3.doi:10.1136/neurintsurg-2018-014051 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30097484
      OpenUrlAbstract/FREE Full Text
    10. ↵
      2. Nagel S ,
      3. Herweh C ,
      4. Pfaff JAR , et al
      . Simplified selection criteria for patients with longer or unknown time to treatment predict good outcome after mechanical thrombectomy. J Neurointerv Surg 2019;11:559–62.doi:10.1136/neurintsurg-2018-014347 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30368472
      OpenUrlAbstract/FREE Full Text
    11. ↵
      2. Motyer R ,
      3. Thornton J ,
      4. Power S , et al
      . Endovascular thrombectomy beyond 12 hours of stroke onset: a stroke network's experience of late intervention. J Neurointerv Surg 2018;10:1043–6.doi:10.1136/neurintsurg-2017-013575 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29459367
      OpenUrlAbstract/FREE Full Text
    12. ↵
      2. Desai SM ,
      3. Tonetti DA ,
      4. Molyneaux BJ , et al
      . Interaction between time, aspects, and clinical mismatch. J Neurointerv Surg 2020;12:911–4.doi:10.1136/neurintsurg-2020-015921 pmid:http://www.ncbi.nlm.nih.gov/pubmed/32245842
      OpenUrlAbstract/FREE Full Text
    13. ↵
      2. Barber PA ,
      3. Demchuk AM ,
      4. 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;355:1670–4.doi:10.1016/s0140-6736(00)02237-6 pmid:http://www.ncbi.nlm.nih.gov/pubmed/10905241
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      2. Demchuk AM ,
      3. Hill MD ,
      4. Barber PA , et al
      . Importance of early ischemic computed tomography changes using aspects in NINDS rtPA stroke study. Stroke 2005;36:2110–5.doi:10.1161/01.STR.0000181116.15426.58 pmid:http://www.ncbi.nlm.nih.gov/pubmed/16166579
      OpenUrlAbstract/FREE Full Text
    15. ↵
      2. Hill MD ,
      3. Rowley HA ,
      4. Adler F , et al
      . Selection of acute ischemic stroke patients for intra-arterial thrombolysis with pro-urokinase by using aspects. Stroke 2003;34:1925–31.doi:10.1161/01.STR.0000082483.37127.D0 pmid:http://www.ncbi.nlm.nih.gov/pubmed/12843342
      OpenUrlAbstract/FREE Full Text
    16. ↵
      2. Han M ,
      3. Choi JW ,
      4. Rim N-J , et al
      . Cerebral infarct volume measurements to improve patient selection for endovascular treatment. Medicine 2016;95:e4702. doi:10.1097/MD.0000000000004702 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27583902
      OpenUrlCrossRefPubMed
    17. ↵
      2. Farzin B ,
      3. Fahed R ,
      4. Guilbert F , et al
      . Early CT changes in patients admitted for thrombectomy: intrarater and interrater agreement. Neurology 2016;87:249–56.doi:10.1212/WNL.0000000000002860 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27316243
      OpenUrlCrossRefPubMed
    18. ↵
      2. Bal S ,
      3. Bhatia R ,
      4. Menon BK , et al
      . Time dependence of reliability of noncontrast computed tomography in comparison to computed tomography angiography source image in acute ischemic stroke. Int J Stroke 2015;10:55–60.doi:10.1111/j.1747-4949.2012.00859.x pmid:http://www.ncbi.nlm.nih.gov/pubmed/22974504
      OpenUrlCrossRefPubMed
    19. ↵
      2. Naylor J ,
      3. Churilov L ,
      4. Chen Z , et al
      . Reliability, reproducibility and prognostic accuracy of the Alberta stroke program early CT score on CT perfusion and Non-Contrast CT in hyperacute stroke. Cerebrovasc Dis 2017;44:195–202.doi:10.1159/000479707 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28810259
      OpenUrlCrossRefPubMed
    20. ↵
      2. Haussen DC ,
      3. Dehkharghani S ,
      4. Rangaraju S , et al
      . Automated CT perfusion ischemic core volume and noncontrast CT aspects (Alberta stroke program early CT score): correlation and clinical outcome prediction in large vessel stroke. Stroke 2016;47:2318–22.doi:10.1161/STROKEAHA.116.014117 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27507858
      OpenUrlAbstract/FREE Full Text
    21. ↵
      2. Sheth SA ,
      3. Malhotra K ,
      4. Liebeskind DS , et al
      . Regional contributions to poststroke disability in endovascular therapy. Interv Neurol 2018;7:533–43.doi:10.1159/000492400 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30410533
      OpenUrlCrossRefPubMed
    22. ↵
      2. Nogueira RG ,
      3. Ribó M
      . Endovascular treatment of acute stroke. Stroke 2019;50:2612–8.doi:10.1161/STROKEAHA.119.023811 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31340728
      OpenUrlCrossRefPubMed
    23. ↵
      2. Nagel S ,
      3. Wang X ,
      4. Carcel C , et al
      . Clinical utility of electronic Alberta stroke program early computed tomography score software in the enchanted trial database. Stroke 2018;49:1407–11.doi:10.1161/STROKEAHA.117.019863 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29777016
      OpenUrlAbstract/FREE Full Text
    24. ↵
      2. Psychogios M-N ,
      3. Sporns PB ,
      4. Ospel J , et al
      . Automated perfusion calculations vs. visual scoring of collaterals and CBV-ASPECTS : has the machine surpassed the eye? Clin Neuroradiol 2020. doi:doi:10.1007/s00062-020-00974-3. [Epub ahead of print: 20 Nov 2020].pmid:33216157
    25. ↵
      2. Phan TG ,
      3. Donnan GA ,
      4. Koga M , et al
      . The aspects template is weighted in favor of the Striatocapsular region. Neuroimage 2006;31:477–81.doi:10.1016/j.neuroimage.2005.12.059 pmid:http://www.ncbi.nlm.nih.gov/pubmed/16504541
      OpenUrlCrossRefPubMed
    26. ↵
      2. Bouslama M ,
      3. Bowen MT ,
      4. Haussen DC , et al
      . Selection paradigms for large vessel occlusion acute ischemic stroke endovascular therapy. Cerebrovasc Dis 2017;44:277–84.doi:10.1159/000478537 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28877524
      OpenUrlCrossRefPubMed
    27. ↵
      2. Ribo M ,
      3. Flores A ,
      4. Mansilla E , et al
      . Age-adjusted infarct volume threshold for good outcome after endovascular treatment. J Neurointerv Surg 2014;6:418–22.doi:10.1136/neurintsurg-2013-010786 pmid:http://www.ncbi.nlm.nih.gov/pubmed/23832414
      OpenUrlAbstract/FREE Full Text
    28. ↵
      2. Bouslama M ,
      3. Haussen DC ,
      4. Rodrigues GM , et al
      . Age-Adjusted infarct volume cut-off points improve stroke outcome prognostication beyond modeling with age and infarct volume. J Neurointerv Surg 2021;50:neurintsurg-2020-017066. doi:10.1136/neurintsurg-2020-017066 pmid:http://www.ncbi.nlm.nih.gov/pubmed/33722959
      OpenUrlPubMed

    Footnotes

    • Twitter @BouslamaMd, @DiogoHaussen

    • Contributors Study design: MB, RGN. Drafting the original manuscript: MB. Data collection, analysis and interpretation; revising the work critically for important intellectual content; final approval; agreement to be accountable for all aspects of the work: all authors.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests RGN: Stryker Neurovascular (Trevo-2 Trial PI - modest, DAWN Trial PI – no payment, Trevo Registry Steering Committee – no payment, Physician Advisory Board/ Consultant - significant), Covidien/Medtronic (SWIFT and SWIFT-PRIME Steering Committee - modest, STAR Trial Core Lab - significant), and Penumbra (3-D Separator Trial Executive Committee – no payment), Neuravi/ Cerenovus (ARISE-II Trial Steering Committee – no payment, Physician Advisory Board - modest), Phenox (Physician Advisory Board/ Consultant – modest), Genentech (Physician Advisory Board/ Consultant – modest).

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