Clinical Imaging Factors Associated With Infarct Progression in Patients With Ischemic Stroke During Transfer for Mechanical Thrombectomy | Cerebrovascular Disease | JAMA Neurology | JAMA Network
[Skip to Navigation]
Sign In
Figure.  Collateral Grading System
Collateral Grading System

Collateral status was defined as follows: 1, absent collateral blood flow or less than 50% of the middle cerebral artery territory (no or poor); 2, collateral blood flow less than that on the contralateral side but more than 50% of the middle cerebral artery territory (moderate); 3, collateral blood flow equal to that of the contralateral side (adequate); and 4, collateral blood flow greater than that on the contralateral side (augmented).

Table 1.  Baseline Characteristics of the Study Patientsa
Baseline Characteristics of the Study Patientsa
Table 2.  Multivariable Analysis of Variables Associated With Increasing Odds of ASPECTS Decay
Multivariable Analysis of Variables Associated With Increasing Odds of ASPECTS Decay
Table 3.  Multivariable Analysis of Variables Associated With Increasing Odds of ASPECTS Decay
Multivariable Analysis of Variables Associated With Increasing Odds of ASPECTS Decay
1.
Berkhemer  OA, Fransen  PSS, Beumer  D,  et al; MR CLEAN Investigators.  A randomized trial of intraarterial treatment for acute ischemic stroke.  N Engl J Med. 2015;372(1):11-20.PubMedGoogle ScholarCrossref
2.
Campbell  BCV, Mitchell  PJ, Kleinig  TJ,  et al; EXTEND-IA Investigators.  Endovascular therapy for ischemic stroke with perfusion-imaging selection.  N Engl J Med. 2015;372(11):1009-1018.PubMedGoogle ScholarCrossref
3.
Jovin  TG, Chamorro  A, Cobo  E,  et al; REVASCAT Trial Investigators.  Thrombectomy within 8 hours after symptom onset in ischemic stroke.  N Engl J Med. 2015;372(24):2296-2306.PubMedGoogle ScholarCrossref
4.
Goyal  M, Demchuk  AM, Menon  BK,  et al; ESCAPE Trial Investigators.  Randomized assessment of rapid endovascular treatment of ischemic stroke.  N Engl J Med. 2015;372(11):1019-1030.PubMedGoogle ScholarCrossref
5.
Saver  JL, Goyal  M, Bonafe  A,  et al; SWIFT PRIME Investigators.  Stent-retriever thrombectomy after intravenous t-PA vs. t-PA alone in stroke.  N Engl J Med. 2015;372(24):2285-2295.PubMedGoogle ScholarCrossref
6.
Bracard  S, Ducrocq  X, Mas  JL,  et al; THRACE investigators.  Mechanical thrombectomy after intravenous alteplase versus alteplase alone after stroke (THRACE): a randomised controlled trial.  Lancet Neurol. 2016;15(11):1138-1147.PubMedGoogle ScholarCrossref
7.
Pexman  JHW, Barber  PA, Hill  MD,  et al.  Use of the Alberta Stroke Program Early CT Score (ASPECTS) for assessing CT scans in patients with acute stroke.  AJNR Am J Neuroradiol. 2001;22(8):1534-1542.PubMedGoogle Scholar
8.
Powers  WJ, Derdeyn  CP, Biller  J,  et al; American Heart Association Stroke Council.  2015 American Heart Association/American Stroke Association focused update of the 2013 guidelines for the early management of patients with acute ischemic stroke regarding endovascular treatment: a guideline for healthcare professionals from the American Heart Association/American Stroke Association.  Stroke. 2015;46(10):3020-3035.PubMedGoogle ScholarCrossref
9.
Smith  EE, Schwamm  LH.  Endovascular clot retrieval therapy: implications for the organization of stroke systems of care in North America.  Stroke. 2015;46(6):1462-1467.PubMedGoogle ScholarCrossref
10.
Mokin  M, Gupta  R, Guerrero  WR, Rose  DZ, Burgin  WS, Sivakanthan  S.  ASPECTS decay during inter-facility transfer in patients with large vessel occlusion strokes.  J Neurointerv Surg. 2017;9(5):442-444.PubMedGoogle ScholarCrossref
11.
Fiorelli  M, Bastianello  S, von Kummer  R,  et al.  Hemorrhagic transformation within 36 hours of a cerebral infarct: relationships with early clinical deterioration and 3-month outcome in the European Cooperative Acute Stroke Study I (ECASS I) cohort.  Stroke. 1999;30(11):2280-2284.PubMedGoogle ScholarCrossref
12.
Maas  MB, Lev  MH, Ay  H,  et al.  Collateral vessels on CT angiography predict outcome in acute ischemic stroke.  Stroke. 2009;40(9):3001-3005.PubMedGoogle ScholarCrossref
13.
Tang  W, Hu  J, Zhang  H, Wu  P, He  H.  κ Coefficient: a popular measure of rater agreement.  Shanghai Arch Psychiatry. 2015;27(1):62-67.PubMedGoogle Scholar
14.
von Elm  E, Altman  DG, Egger  M, Pocock  SJ, Gøtzsche  PC, Vandenbroucke  JP; STROBE Initiative.  The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies.  Lancet. 2007;370(9596):1453-1457.PubMedGoogle ScholarCrossref
15.
Shuaib  A, Butcher  K, Mohammad  AA, Saqqur  M, Liebeskind  DS.  Collateral blood vessels in acute ischaemic stroke: a potential therapeutic target.  Lancet Neurol. 2011;10(10):909-921.PubMedGoogle ScholarCrossref
16.
Miteff  F, Levi  CR, Bateman  GA, Spratt  N, McElduff  P, Parsons  MW.  The independent predictive utility of computed tomography angiographic collateral status in acute ischaemic stroke.  Brain. 2009;132(pt 8):2231-2238.PubMedGoogle ScholarCrossref
17.
Leslie-Mazwi  TM, Hirsch  JA, Falcone  GJ,  et al.  Endovascular stroke treatment outcomes after patient selection based on magnetic resonance imaging and clinical criteria.  JAMA Neurol. 2016;73(1):43-49.PubMedGoogle ScholarCrossref
18.
Desilles  J-P, Consoli  A, Redjem  H,  et al; ETIS (Endovascular Treatment in Ischemic Stroke) Research Investigators.  Successful reperfusion with mechanical thrombectomy is associated with reduced disability and mortality in patients with pretreatment diffusion-weighted imaging–Alberta Stroke Program Early Computed Tomography Score ≤6.  Stroke. 2017;48(4):963-969.PubMedGoogle ScholarCrossref
19.
Gilgen  MD, Klimek  D, Liesirova  KT,  et al.  Younger stroke patients with large pretreatment diffusion-weighted imaging lesions may benefit from endovascular treatment.  Stroke. 2015;46(9):2510-2516.PubMedGoogle ScholarCrossref
20.
Sun  C-HJ, Connelly  K, Nogueira  RG,  et al.  ASPECTS decay during inter-facility transfer predicts patient outcomes in endovascular reperfusion for ischemic stroke: a unique assessment of dynamic physiologic change over time.  J Neurointerv Surg. 2015;7(1):22-26.PubMedGoogle ScholarCrossref
21.
Fuentes  B, Alonso de Leciñana  M, Ximénez-Carrillo  A,  et al; Madrid Stroke Network.  Futile interhospital transfer for endovascular treatment in acute ischemic stroke: the Madrid Stroke Network experience.  Stroke. 2015;46(8):2156-2161.PubMedGoogle ScholarCrossref
22.
Oppenheim  C, Samson  Y, Manaï  R,  et al.  Prediction of malignant middle cerebral artery infarction by diffusion-weighted imaging.  Stroke. 2000;31(9):2175-2181.PubMedGoogle ScholarCrossref
23.
Olivot  JM, Mlynash  M, Inoue  M,  et al; DEFUSE 2 Investigators.  Hypoperfusion intensity ratio predicts infarct progression and functional outcome in the DEFUSE 2 cohort.  Stroke. 2014;45(4):1018-1023.PubMedGoogle ScholarCrossref
24.
Kamalian  S, Morais  LT, Pomerantz  SR,  et al.  Clot length distribution and predictors in anterior circulation stroke: implications for intra-arterial therapy.  Stroke. 2013;44(12):3553-3556.PubMedGoogle ScholarCrossref
25.
Riedel  CH, Zimmermann  P, Jensen-Kondering  U, Stingele  R, Deuschl  G, Jansen  O.  The importance of size: successful recanalization by intravenous thrombolysis in acute anterior stroke depends on thrombus length.  Stroke. 2011;42(6):1775-1777.PubMedGoogle ScholarCrossref
26.
Campbell  BCV, Christensen  S, Tress  BM,  et al; EPITHET Investigators.  Failure of collateral blood flow is associated with infarct growth in ischemic stroke.  J Cereb Blood Flow Metab. 2013;33(8):1168-1172.PubMedGoogle ScholarCrossref
27.
Man  S, Aoki  J, Hussain  MS,  et al.  Predictors of infarct growth after endovascular therapy for acute ischemic stroke.  J Stroke Cerebrovasc Dis. 2015;24(2):401-407.PubMedGoogle ScholarCrossref
28.
de Havenon  A, Haynor  DR, Tirschwell  DL,  et al.  Association of collateral blood vessels detected by arterial spin labeling magnetic resonance imaging with neurological outcome after ischemic stroke.  JAMA Neurol. 2017;74(4):453-458.PubMedGoogle ScholarCrossref
29.
Liebeskind  DS, Jahan  R, Nogueira  RG, Jovin  TG, Lutsep  HL, Saver  JL; SWIFT Investigators.  Early arrival at the emergency department is associated with better collaterals, smaller established infarcts and better clinical outcomes with endovascular stroke therapy: SWIFT study.  J Neurointerv Surg. 2016;8(6):553-558.PubMedGoogle ScholarCrossref
30.
Cheng-Ching  E, Frontera  JA, Man  S,  et al.  Degree of collaterals and not time is the determining factor of core infarct volume within 6 hours of stroke onset.  AJNR Am J Neuroradiol. 2015;36(7):1272-1276.PubMedGoogle ScholarCrossref
31.
Woitzik  J, Pinczolits  A, Hecht  N,  et al.  Excitotoxicity and metabolic changes in association with infarct progression.  Stroke. 2014;45(4):1183-1185.PubMedGoogle ScholarCrossref
32.
Garcia  JH, Yoshida  Y, Chen  H,  et al.  Progression from ischemic injury to infarct following middle cerebral artery occlusion in the rat.  Am J Pathol. 1993;142(2):623-635.PubMedGoogle Scholar
33.
Hakimelahi  R, Vachha  BA, Copen  WA,  et al.  Time and diffusion lesion size in major anterior circulation ischemic strokes.  Stroke. 2014;45(10):2936-2941.PubMedGoogle ScholarCrossref
34.
Lansberg  MG, Cereda  CW, Mlynash  M,  et al; Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution 2 (DEFUSE 2) Study Investigators.  Response to endovascular reperfusion is not time-dependent in patients with salvageable tissue.  Neurology. 2015;85(8):708-714.PubMedGoogle ScholarCrossref
35.
Abou-Chebl  A.  Endovascular treatment of acute ischemic stroke may be safely performed with no time window limit in appropriately selected patients.  Stroke. 2010;41(9):1996-2000.PubMedGoogle ScholarCrossref
36.
Bang  OY, Saver  JL, Kim  SJ,  et al; UCLA-Samsung Stroke Collaborators.  Collateral flow averts hemorrhagic transformation after endovascular therapy for acute ischemic stroke.  Stroke. 2011;42(8):2235-2239.PubMedGoogle ScholarCrossref
37.
Goyal  M, Menon  BK, Coutts  SB, Hill  MD, Demchuk  AM; Penumbra Pivotal Stroke Trial Investigators, Calgary Stroke Program, and the Seaman MR Research Center.  Effect of baseline CT scan appearance and time to recanalization on clinical outcomes in endovascular thrombectomy of acute ischemic strokes.  Stroke. 2011;42(1):93-97.PubMedGoogle ScholarCrossref
38.
Minnerup  J, Wersching  H, Teuber  A,  et al; REVASK Investigators.  Outcome after thrombectomy and intravenous thrombolysis in patients with acute ischemic stroke: a prospective observational study.  Stroke. 2016;47(6):1584-1592.PubMedGoogle ScholarCrossref
39.
Menon  BK, Sajobi  TT, Zhang  Y,  et al.  Analysis of workflow and time to treatment on thrombectomy outcome in the ESCAPE randomized controlled trial.  Circulation. 2016;133(23):2279-2286.PubMedGoogle ScholarCrossref
40.
Liang  JW, Stein  L, Wilson  N, Fifi  JT, Tuhrim  S, Dhamoon  MS.  Timing of vessel imaging for suspected large vessel occlusions does not affect groin puncture time in transfer patients with stroke.  [published online January 24, 2017].  J Neurointerv Surg. 2017;neurintsurg-2016-012854. doi:10.1136/neurintsurg-2016-012854PubMedGoogle Scholar
41.
McTaggart  RA, Yaghi  S, Cutting  SM,  et al.  Association of a primary stroke center protocol for suspected stroke by large-vessel occlusion with efficiency of care and patient outcomes.  JAMA Neurol. 2017;74(7):793-800. doi:10.1001/jamaneurol.2017.0477PubMedGoogle ScholarCrossref
42.
Khoury  N, Fahed  R, Guilbert  F,  et al.  Early CT changes in patients admitted for thrombectomy: intra- and interrater agreement and systematic review of the literature.  Neurology. 2016;86(16 suppl):S47.001.Google Scholar
Original Investigation
November 2017

Clinical Imaging Factors Associated With Infarct Progression in Patients With Ischemic Stroke During Transfer for Mechanical Thrombectomy

Author Affiliations
  • 1Stroke Research Center, Massachusetts General Hospital, Harvard Medical School, Boston
  • 2Department of Neuroradiology, Université Paris-Descartes, INSERM U894, Centre Hospitalier Sainte-Anne, Paris, France
  • 3Department of Neuroradiology, Goethe University, Frankfurt, Germany
  • 4Stroke Service, Massachusetts General Hospital, Harvard Medical School, Boston
  • 5Neuroendovascular Program, Massachusetts General Hospital, Harvard Medical School, Boston
JAMA Neurol. 2017;74(11):1361-1367. doi:10.1001/jamaneurol.2017.2149
Key Points

Question  Among patients with ischemic stroke transferred for mechanical thrombectomy, what baseline clinical imaging factors portend an unfavorable evolution to a point at which the patient may not derive clinical benefit from mechanical thrombectomy at arrival?

Findings  This cohort study of prospectively collected data found that, along with initial clinical severity, poor collateral blood vessel status was the most determinant factor of evolution to an unfavorable imaging profile during transfer.

Meaning  In certain subgroups of patients with ischemic stroke, vascular imaging at the referring hospitals may play a critical role in determining the benefits of transfer for thrombectomy.

Abstract

Importance  When transferred from a referring hospital (RH) to a thrombectomy-capable stroke center (TCSC), patients with initially favorable imaging profiles (Alberta Stroke Program Early CT Score [ASPECTS] ≥6) often demonstrate infarct progression significant enough to make them ineligible for mechanical thrombectomy at arrival. In rapidly evolving stroke care networks, the question of the need for vascular imaging at the RHs remains unsolved, resulting in an important amount of futile transfers for thrombectomy.

Objective  To examine the clinical imaging factors associated with unfavorable imaging profile evolution for thrombectomy in patients with ischemic stroke initially transferred to non-TCSCs.

Design, Setting, and Participants  Data from patients transferred from 1 of 30 RHs in our regional stroke network and presenting at our TCSC from January 1, 2010, to January 1, 2016, were retrospectively analyzed. Consecutive patients with acute ischemic stroke initially admitted to a non–thrombectomy-capable RH and transferred to our center for which a RH computed tomography (CT) and a CT angiography (CTA) at arrival were available for review.

Main Outcomes and Measures  ASPECTS were evaluated. The adequacy of leptomeningeal collateral blood flow was rated as no or poor, decreased, adequate, or augmented per the adapted Maas scale. The main outcome was an ASPECTS decay, defined as an initial ASPECTS of 6 or higher worsening between RH and TCSC CTs to a score of less than 6 (making the patient less likely to derive clinical benefit from thrombectomy at arrival).

Results  A total of 316 patients were included in the analysis (mean [SD] age, 70.3 [14.2] years; 137 [43.4%] female). In multivariable models, higher National Institutes of Health Stroke Score, lower baseline ASPECTSs, and no or poor collateral blood vessel status were associated with ASPECTS decay, with collateral blood vessel status demonstrating the highest adjusted odds ratio of 5.14 (95% CI, 2.20-12.70; P < .001). Similar results were found after stratification by vessel occlusion level.

Conclusions and Relevance  In patients with ischemic stroke transferred for thrombectomy, poor collateral blood flow and stroke clinical severity are the main determinants of ASPECTS decay. Our findings suggest that in certain subgroups vascular imaging, including collateral assessment, can play a crucial role in determining the benefits of transfer for thrombectomy.

Introduction

The recent results of 6 randomized clinical trials1-6 have provided compelling evidence that thrombectomy improves outcome in selected patients with acute ischemic stroke and large vessel occlusion. Subgroup analyses from 3 of those trials3-5 indicate that patients with lower Alberta Stroke Program Early CT Scores (ASPECTSs)7 were less likely to benefit from mechanical clot removal. On this basis, the American Heart Association’s updated guideline recommends endovascular treatment for clinically eligible patients with a large vessel occlusion if a favorable imaging profile is present, defined as an ASPECTS of 6 or higher.8

A significant proportion of patients with ischemic stroke are initially evaluated at referring hospitals (RHs) that are prepared for acute stroke cases. These patients may receive intravenous thrombolysis before being transferred (drip and ship) to a thrombectomy-capable stroke center (TCSC) for endovascular treatment evaluation.9 For those patients clinically eligible at the RH (ie, favorable imaging profile, National Institutes of Health Stroke Score [NIHSS]≥6, and able to reach a TCSC within 6 hours of symptom onset), clinical stroke severity10 is known to be a determinant of infarct progression. Other potential clinical or imaging factors that determine infarct progression are poorly understood. Adequacy of leptomeningeal collateral blood vessels is known to influence infarct progression in patients with ischemic stroke, but their role in predicting eligibility for mechanical thrombectomy has not been studied in the common setting of transferred patients. The ability to determine infarct evolution has the potential to allow optimized resource use within stroke networks by better selecting patients most likely to remain eligible and limiting futile transfers for thrombectomy.9 We sought to determine the clinical and imaging factors associated with evolution to an unfavorable imaging profile in patients transferred from RHs to a TCSC and to specifically evaluate the influence of the adequacy of leptomeningeal flow on infarct progression in these patients.

Methods
Study Design

We performed an observational, single-center, retrospective cohort study of prospectively acquired data to evaluate consecutive patients transferred from RHs to a tertiary care TCSC. Informed consent requirements were waived for the retrospective use of Get With The Guidelines–Stroke data. Institutional review board approval was obtained from the Partners Human Research Committee for all prospective data collection and for retrospective medical record review and analysis. Data were not deidentified.

Study Population

We defined infarct progression as an ASPECTS decay over time, with the critical threshold of a score of 6. Using our local Get With The Guidelines–Stroke database from January 1, 2010, to January 1, 2016, we identified adult patients with stroke transferred from 1 of the 30 RHs in our regional stroke network for which (1) the referring hospital computed tomography (CT) and (2) the CT angiography (CTA) at TCSC arrival were available for review. Patients with posterior strokes and established infarcts (ASPECTS<6) at RH arrival were excluded from the analysis.

Imaging

The CTs acquired at RHs and at the TCSC were reviewed by 2 of us (G.B., A.L.). The following variables were analyzed: ASPECTS, presence and localization of a dense vessel sign, presence and localization of a vessel occlusion on CTA, degree of leptomeningeal collateral blood flow, and presence and classification of interval hemorrhagic transformation according to the European Cooperative Acute Stroke Study criteria.11

At our center, CTA was performed from vertex to aortic arch after injection of 65 to 140 mL of a nonionic contrast agent (Isovue, Bracco Diagnostics) at 3 to 4 mL/s. Median parameters were a 1.25-mm section thickness and 220-mm reconstruction diameter. Degree of collateral blood flow was defined using previously validated criteria according to Maas and colleagues’ rating scale12 modified as follows: 1, absent collateral blood flow or less than 50% of the middle cerebral artery territory (no or poor); 2, collateral blood flow less than that on the contralateral side but more than 50% of the middle cerebral artery territory (moderate); 3, collateral blood flow equal to that on the contralateral side (adequate); and 4, collateral blood flow greater than that on the contralateral side (augmented) (Figure).

Outcome Definition and Group Assignment

ASPECTS decay was defined as a decrease between the RH and the TCSC, with a total initial score of 6 or higher worsening to a score of less than 6, making the patient ineligible for thrombectomy at arrival according to American Heart Association/American Stroke Association guidelines. Interval time indicates the time between the CTs at the RH and the TCSC.

In sensitivity analyses, we added to the main sample those patients presenting at the RH with an ASPECTS already below 6 (ie, established infarct). For those patients, we imputed an ASPECTS of 10 at the time of symptom onset and calculated the interval time as the time since symptom onset to obtaining the CT from the RH. In both analyses sets, patients were dichotomized into those with ASPECTS decay vs those without.

Statistical Analysis

The χ2 test, Fisher exact test, 2-tailed, unpaired t test, and Mann-Whitney test were used for the univariate analysis, with P < .05 as the threshold for statistical significance. Multivariable logistic regression models were used to determine factors that were independently associated with ASPECTS decay. Variables associated with the outcome in univariate analysis (P ≤ .10) were entered in the nominal logistic model, adjusting for age and sex, and then backward elimination was used to remove nonsignificant variables (P > .05), resulting in a minimal adjusted model. Sensitivity analyses were performed using ordinal regression models. Variables with significant associations (predetermined P < .10) in univariable analyses were tested using a fully adjusted ordinal regression model (predetermined to be adjusted for sex, age, and time to qualifying CT). Interrater agreement statistics were calculated using the Cohen κ.13 All analyses were performed using JMP Pro software, version 12 (SAS Institute Inc). This report was prepared according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.14

Results

Among 508 patients screened for eligibility, 123 were excluded because they did not undergo a noncontrast CT at our center within 18 hours, 9 because they did not undergo CTA, 46 because of a final diagnosis of posterior circulation stroke, and 14 because of an established infarct at RH arrival. The final study population consisted of 316 patients with anterior circulation ischemic stroke. Compared with patients included in the analysis, those excluded had proximal dense vessels at the RH less often (16 [8.2%] vs 112 [35.4%], P = .002) but did not differ in terms of baseline demographic characteristics and medical history. Interrater agreement was excellent for collateral blood vessel status (0.85; 95% CI, 0.76-0.91) and good for ASPECTS (dichotomized) (0.78; 95% CI, 0.72-0.83).

Univariable subgroup analyses revealed that patients with ASPECTS decay were more likely to be female (36 [58.1%] vs 1201 [39.8%], P = .01), not to be taking anticoagulants (3 [5.0%] vs 37 [15.3%], P = .05), to have higher initial NIHSSs (median [interquartile range], 19 [15-22] vs 11 [6-17]; P < .001), to have a proximal dense vessel sign on initial CT (45 [72.6%] vs 67 [26.4%], P < .001), and to have poor collateral blood flow on CTA at the TCSC (55 [88.7%] vs 48 [18.9%], P < .001) (Table 1 and eFigure 1 in the Supplement). Median (interquartile range) interval time between CTs did not differ between patients with or without ASPECTS decay (3.25 [2.7-4.0] vs 3.17 [2.65-4.20], P = .76). Increasing interval time between CTs did not appear to be associated with ASPECTS change in unadjusted analysis (eFigure 2 in the Supplement).

To better understand the factors associated with progression to ineligibility for thrombectomy on arrival at a TCSC, we built a series of prespecified hypotheses-driven multivariable models. The first model included only information available at the time of initial evaluation at the RH. Additional models applied information gathered at the TCSC to impute values that could potentially be captured at the RH. In the primary multivariable model (model 1) that used only information available at the RH, higher NIHSS, lower initial ASPECTS, and a proximal dense vessel sign were independent predictors of ASPECTS decay (Table 2). A second model (model 2) incorporated information gathered after transfer to the TCSC when a CTA was acquired. We assumed that if a proximal occlusion was found at the TCSC, it would have been present if the CTA had been performed before transfer, and thus, we assigned the CTA proximal occlusion and collateral status to the patient while at the RH. We then used this information to derive a second model (model 2) that attempts to model the scenario in which CTA is performed routinely at the RH. In this model, the addition of CTA information had little influence on the independent predictors of ASPECTS decay from model 1 (initial NIHSS and ASPECTS), but poor collateral blood flow was significant, with collateral status demonstrating the highest adjusted odds ratio of 5.14 (95% CI, 2.20-12.70; P < .001) (Table 3).

Several sensitivity analyses were performed to test interactions within the variables. The significance of collateral blood vessels was maintained after stratification of patients by site of occlusion (eTable 1 in the Supplement) and adjustment for prestroke anticoagulant use. When we extended the sample of patients with ASPECTS decay to patients with initially unfavorable imaging profiles (n = 330), we found similar results.

In additional sensitivity analyses that used ordinal regression models, we explored the effect of variables included in model 2 on the absolute decrease in ASPECTS coded as an ordinal variable instead of a dichotomous variable with a threshold score of 6. The variables that were significantly associated with ASPECTS decrease were age, proximal occlusion, and presence of poor collateral blood flow (adjusted odds ratio per point decrease in ASPECTS, 2.25; 95% CI, 1.78-2.87; P < .001). Of importance, interval time between RH and TCSC imaging was not (eTable 2 in the Supplement).

Discussion

Our study provides important new information about the nature of infarct progression on CT and the factors associated with infarct growth for patients transferred to the TCSC. Although “time is brain” is a frequent phrase in acute stroke care, each patient has a different rate of stroke evolution as a result of factors such as adequacy of leptomeningeal collateral blood vessels.15,16 Our study found that poor leptomeningeal collateral blood flow as assessed by CTA is a predictor of an unfavorable imaging profile for thrombectomy and ASPECTS decay during transfer. In our sample, 62 (19.6%) of 316 patients experienced a decrease in ASPECTS significant enough to be less likely to benefit from endovascular thrombectomy according to the American Heart Association/American Stroke Association guidelines.8 This finding is consistent with prior work17 from our center in which evolution to a large infarct on neuroimaging excluded 20% of transfer patients from endovascular therapy. Of note, recent reports18,19 have found that patients with large infarct core may also benefit from endovascular clot removal, with the possibility that the definition of an ASPECTS decay as used in this study will evolve.

Previous studies10,20-23 have assessed ASPECTS decay in the population with ischemic stroke across various settings. For example, Sun et al20 suggested that patients (eventually treated with thrombectomy) with less ASPECTS decrease had higher odds of good functional outcome. Mokin et al10 found that the only predictor of ASPECTS decay was the initial clinical severity, but the authors did not report vascular or collateral assessments. Finally, in the Madrid Stroke Network, Fuentes et al21 reported that the second most frequent reason of futile transfer (after clinical improvement) was the worsening of the imaging profile. In this data set, the authors found that approximately 41% of transfers were futile in regard to thrombectomy, but they could not identify characteristics predictive of this futility. Although there is wide variability in clinical imaging determinants of infarct progression,22,23 such factors are less well defined for CT, the most widely used modality in the acute phase of stroke.

Our finding that the initial NIHSS and the presence of a dense vessel on RH imaging identify infarct progression reflects the rate of cerebral tissue loss in stroke from large vessel occlusion. The median NIHSS among patients with M1 occlusions is 14, with scores greater than 9 indicating an 81% positive predictive value for proximal clot.12 Hyperdense vasculature on noncontrast CT represents an imaging signature of clot burden. Thin-section noncontrast CT at admission can identify clot presence and clot length greater than 8 mm,24 a length threshold that typically indicates failure to recanalize with intravenous tissue plasminogen activator alone.25 These features are readily obtained from current RH clinical and imaging data with minimal modification to existing protocols and can help identify patients in need of rapid transfer or those who might not tolerate prolonged transfer times.

A more compelling finding from these data was the effect of measuring collateral blood vessels with CTA. The importance of leptomeningeal collateral blood flow in ischemic stroke outcome is well established.12,26-28 However, its influence on eligibility for thrombectomy has been scantly reported, especially in the setting of transferred patients.29,30 This influence of collateral blood flow on eligibility for mechanical thrombectomy in transferred patients is particularly relevant given the increasing attention paid to reorganization of stroke care delivery in light of the demonstrated benefit of endovascular therapy. Appropriate triage and transfers to TCSCs are paramount in these proposed new paradigms to ensure the most cost-effective care and to reduce the risk of overwhelming the capacity of the TCSC. Our study suggests that in a stroke network setting, vascular imaging to identify intracranial occlusions and assess collateral blood vessels may provide an additional tool for determining triage decisions when transferring a patient to a TCSC. These decisions include the mode of transportation (air for patients at greatest risk of rapid ASPECTS decay), the most appropriate destination center, and the utility of transfer at all (for patients unlikely to have adequate viable tissue volume by the time they reach the destination center). The role of collateral blood vessels as assessed by CTA may be a predictor of infarct growth in the setting of transferred patients. Our findings therefore support the use of CTA for selected patients at the acute phase of ischemic stroke in RHs and community hospitals.

An additional finding of our study was the absence of a definite association between interval time and ASPECTS decay or absolute ASPECTS decrease in our population that was likely influenced by the collateral blood vessel status demonstrated in the various analyses conducted. Infarct progression is caused not only by hypoperfusion but also by its complex excitotoxic and metabolic consequences.31,32 However, hypoperfusion is a powerful driver of penumbral tissue loss. Although infarct growth occurs in a time-dependent fashion, the influence of the time delay on infarct progression is variable based on collateral status.33 It is likely that in patients with poor leptomeningeal collateral blood flow, the time needed for infarct evolution was so short that it was not well evaluated in our sample; for instance, only 17 patients (5.4%) had a transfer time equal to or less than 90 minutes. In our analysis of the influence of time in our sample stratified by dichotomized collateral blood vessel status (eg, no or poor vs moderate or good), we determined that transfer time was mildly associated with ASPECTS decay only in patients with better collateral circulation. The size of our sample did not allow for analysis in each subgroup of collateral status.

Our results therefore suggest the importance of minimizing delays in patients with no or poor collateral blood flow. The Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution (DEFUSE) 2 trial34 found that in patients with salvageable tissue, the association between attenuation of infarct growth and reperfusion was not time dependent, further strengthening the central role of collateral blood vessels in sustaining a favorable imaging profile. Furthermore, mechanical thrombectomy may be effective without time limit in selected patients with robust collateral blood flow,35 and symptomatic hemorrhagic transformation is more frequent among patients with poor collateral blood vessel status who are undergoing mechanical ventilation.36 Both findings have potential implications for patients’ triage at RHs.

Implementing refinements of patient triage through vascular and collateral imaging at the RH represents an organizational challenge.9 Such implementation would need to be carefully evaluated with the perspective of better value-based health care policies, tailored to the specificities of each regional stroke network. The possibility of delays from implementation of imaging changes, potentially deleterious to patients eligible for mechanical thrombectomy,37-39 should be minimized using standardized protocols and centralized reading. A recent study40 found that the timing of vascular imaging (CTA) implementation did not affect onset to groin puncture time in a specific setting and offered the possibility to enhance more efficient triage strategies at the RH. Similarly, a recent study41 found that the implementation of a protocol that includes vascular imaging at the RH and cloud-based image sharing to a TCSC for patients with suspected large vessel occlusion who were first seen at a non-TCSC was associated with a reduction in time to groin puncture and better clinical outcomes.

Strengths and Limitations

The strength of our study derives from a large sample of patients, allowing a robust analysis of the predictors of infarct growth. Limitations include that the study population derives from a single academic center, which may limit generalization. In addition, the readings were not performed in the acute care setting but in an optimal research environment. Concerns about poor interrater agreement for ASPECTS42 were not substantiated in this study, in which the interrater agreement for the dichotomized score (≥6 vs <6) was good between our readers. Finally, our results derive from simulated situations in which we translated findings at the TCSC to apply them to the RH. Because the evolution of collateral blood flow is not well known, modifications of collateral blood flow may have occurred during the interval time, thus introducing a systematic bias in our approach (ie, if collateral blood vessels deteriorated in transit in some patients, this would weaken the association between collateral blood vessels assessed at the RH compared with those assessed at the TCSC). Furthermore, one of the limitations of static CTA is that collateral blood vessel quality may be underestimated and specificity may be improved by dynamic acquisitions (multiphase CTA or CT perfusion).

Conclusions

For patients with ischemic stroke transferred for thrombectomy, collateral blood flow, stroke severity, and proximal vascular occlusion are the main determinants of ASPECTS decay. The added value of vascular imaging and collateral blood vessel assessment for decision making regarding transfers requires further evaluation and adequate tailoring to local resources and needs.

Back to top
Article Information

Corresponding Author: Lee H. Schwamm, MD, Stroke Research Center, Massachusetts General Hospital, Harvard Medical School, 55 Fruit St, Ste ACC720, Boston, MA 02114 (lschwamm@mgh.harvard.edu).

Accepted for Publication: May 30, 2017.

Published Online: September 25, 2017. doi:10.1001/jamaneurol.2017.2149

Author Contributions: Dr Boulouis had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Boulouis, Siddiqui, Charidimou, Leslie-Mazwi, Schwamm.

Acquisition, analysis, or interpretation of data: Boulouis, Lauer, Siddiqui, Charidimou, Regenhardt, Viswanathan, Rost.

Drafting of the manuscript: Boulouis, Siddiqui, Leslie-Mazwi.

Critical revision of the manuscript for important intellectual content: Boulouis, Lauer, Siddiqui, Charidimou, Regenhardt, Viswanathan, Rost, Schwamm.

Statistical analysis: Boulouis, Lauer, Siddiqui, Charidimou.

Administrative, technical, or material support: Siddiqui, Regenhardt, Rost, Leslie-Mazwi.

Supervision: Siddiqui, Viswanathan, Rost, Schwamm.

Conflict of Interest Disclosures: None reported.

Funding/Support: Dr Boulouis was supported by Fulbright and Monahan Foundation grants.

Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.

References
1.
Berkhemer  OA, Fransen  PSS, Beumer  D,  et al; MR CLEAN Investigators.  A randomized trial of intraarterial treatment for acute ischemic stroke.  N Engl J Med. 2015;372(1):11-20.PubMedGoogle ScholarCrossref
2.
Campbell  BCV, Mitchell  PJ, Kleinig  TJ,  et al; EXTEND-IA Investigators.  Endovascular therapy for ischemic stroke with perfusion-imaging selection.  N Engl J Med. 2015;372(11):1009-1018.PubMedGoogle ScholarCrossref
3.
Jovin  TG, Chamorro  A, Cobo  E,  et al; REVASCAT Trial Investigators.  Thrombectomy within 8 hours after symptom onset in ischemic stroke.  N Engl J Med. 2015;372(24):2296-2306.PubMedGoogle ScholarCrossref
4.
Goyal  M, Demchuk  AM, Menon  BK,  et al; ESCAPE Trial Investigators.  Randomized assessment of rapid endovascular treatment of ischemic stroke.  N Engl J Med. 2015;372(11):1019-1030.PubMedGoogle ScholarCrossref
5.
Saver  JL, Goyal  M, Bonafe  A,  et al; SWIFT PRIME Investigators.  Stent-retriever thrombectomy after intravenous t-PA vs. t-PA alone in stroke.  N Engl J Med. 2015;372(24):2285-2295.PubMedGoogle ScholarCrossref
6.
Bracard  S, Ducrocq  X, Mas  JL,  et al; THRACE investigators.  Mechanical thrombectomy after intravenous alteplase versus alteplase alone after stroke (THRACE): a randomised controlled trial.  Lancet Neurol. 2016;15(11):1138-1147.PubMedGoogle ScholarCrossref
7.
Pexman  JHW, Barber  PA, Hill  MD,  et al.  Use of the Alberta Stroke Program Early CT Score (ASPECTS) for assessing CT scans in patients with acute stroke.  AJNR Am J Neuroradiol. 2001;22(8):1534-1542.PubMedGoogle Scholar
8.
Powers  WJ, Derdeyn  CP, Biller  J,  et al; American Heart Association Stroke Council.  2015 American Heart Association/American Stroke Association focused update of the 2013 guidelines for the early management of patients with acute ischemic stroke regarding endovascular treatment: a guideline for healthcare professionals from the American Heart Association/American Stroke Association.  Stroke. 2015;46(10):3020-3035.PubMedGoogle ScholarCrossref
9.
Smith  EE, Schwamm  LH.  Endovascular clot retrieval therapy: implications for the organization of stroke systems of care in North America.  Stroke. 2015;46(6):1462-1467.PubMedGoogle ScholarCrossref
10.
Mokin  M, Gupta  R, Guerrero  WR, Rose  DZ, Burgin  WS, Sivakanthan  S.  ASPECTS decay during inter-facility transfer in patients with large vessel occlusion strokes.  J Neurointerv Surg. 2017;9(5):442-444.PubMedGoogle ScholarCrossref
11.
Fiorelli  M, Bastianello  S, von Kummer  R,  et al.  Hemorrhagic transformation within 36 hours of a cerebral infarct: relationships with early clinical deterioration and 3-month outcome in the European Cooperative Acute Stroke Study I (ECASS I) cohort.  Stroke. 1999;30(11):2280-2284.PubMedGoogle ScholarCrossref
12.
Maas  MB, Lev  MH, Ay  H,  et al.  Collateral vessels on CT angiography predict outcome in acute ischemic stroke.  Stroke. 2009;40(9):3001-3005.PubMedGoogle ScholarCrossref
13.
Tang  W, Hu  J, Zhang  H, Wu  P, He  H.  κ Coefficient: a popular measure of rater agreement.  Shanghai Arch Psychiatry. 2015;27(1):62-67.PubMedGoogle Scholar
14.
von Elm  E, Altman  DG, Egger  M, Pocock  SJ, Gøtzsche  PC, Vandenbroucke  JP; STROBE Initiative.  The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies.  Lancet. 2007;370(9596):1453-1457.PubMedGoogle ScholarCrossref
15.
Shuaib  A, Butcher  K, Mohammad  AA, Saqqur  M, Liebeskind  DS.  Collateral blood vessels in acute ischaemic stroke: a potential therapeutic target.  Lancet Neurol. 2011;10(10):909-921.PubMedGoogle ScholarCrossref
16.
Miteff  F, Levi  CR, Bateman  GA, Spratt  N, McElduff  P, Parsons  MW.  The independent predictive utility of computed tomography angiographic collateral status in acute ischaemic stroke.  Brain. 2009;132(pt 8):2231-2238.PubMedGoogle ScholarCrossref
17.
Leslie-Mazwi  TM, Hirsch  JA, Falcone  GJ,  et al.  Endovascular stroke treatment outcomes after patient selection based on magnetic resonance imaging and clinical criteria.  JAMA Neurol. 2016;73(1):43-49.PubMedGoogle ScholarCrossref
18.
Desilles  J-P, Consoli  A, Redjem  H,  et al; ETIS (Endovascular Treatment in Ischemic Stroke) Research Investigators.  Successful reperfusion with mechanical thrombectomy is associated with reduced disability and mortality in patients with pretreatment diffusion-weighted imaging–Alberta Stroke Program Early Computed Tomography Score ≤6.  Stroke. 2017;48(4):963-969.PubMedGoogle ScholarCrossref
19.
Gilgen  MD, Klimek  D, Liesirova  KT,  et al.  Younger stroke patients with large pretreatment diffusion-weighted imaging lesions may benefit from endovascular treatment.  Stroke. 2015;46(9):2510-2516.PubMedGoogle ScholarCrossref
20.
Sun  C-HJ, Connelly  K, Nogueira  RG,  et al.  ASPECTS decay during inter-facility transfer predicts patient outcomes in endovascular reperfusion for ischemic stroke: a unique assessment of dynamic physiologic change over time.  J Neurointerv Surg. 2015;7(1):22-26.PubMedGoogle ScholarCrossref
21.
Fuentes  B, Alonso de Leciñana  M, Ximénez-Carrillo  A,  et al; Madrid Stroke Network.  Futile interhospital transfer for endovascular treatment in acute ischemic stroke: the Madrid Stroke Network experience.  Stroke. 2015;46(8):2156-2161.PubMedGoogle ScholarCrossref
22.
Oppenheim  C, Samson  Y, Manaï  R,  et al.  Prediction of malignant middle cerebral artery infarction by diffusion-weighted imaging.  Stroke. 2000;31(9):2175-2181.PubMedGoogle ScholarCrossref
23.
Olivot  JM, Mlynash  M, Inoue  M,  et al; DEFUSE 2 Investigators.  Hypoperfusion intensity ratio predicts infarct progression and functional outcome in the DEFUSE 2 cohort.  Stroke. 2014;45(4):1018-1023.PubMedGoogle ScholarCrossref
24.
Kamalian  S, Morais  LT, Pomerantz  SR,  et al.  Clot length distribution and predictors in anterior circulation stroke: implications for intra-arterial therapy.  Stroke. 2013;44(12):3553-3556.PubMedGoogle ScholarCrossref
25.
Riedel  CH, Zimmermann  P, Jensen-Kondering  U, Stingele  R, Deuschl  G, Jansen  O.  The importance of size: successful recanalization by intravenous thrombolysis in acute anterior stroke depends on thrombus length.  Stroke. 2011;42(6):1775-1777.PubMedGoogle ScholarCrossref
26.
Campbell  BCV, Christensen  S, Tress  BM,  et al; EPITHET Investigators.  Failure of collateral blood flow is associated with infarct growth in ischemic stroke.  J Cereb Blood Flow Metab. 2013;33(8):1168-1172.PubMedGoogle ScholarCrossref
27.
Man  S, Aoki  J, Hussain  MS,  et al.  Predictors of infarct growth after endovascular therapy for acute ischemic stroke.  J Stroke Cerebrovasc Dis. 2015;24(2):401-407.PubMedGoogle ScholarCrossref
28.
de Havenon  A, Haynor  DR, Tirschwell  DL,  et al.  Association of collateral blood vessels detected by arterial spin labeling magnetic resonance imaging with neurological outcome after ischemic stroke.  JAMA Neurol. 2017;74(4):453-458.PubMedGoogle ScholarCrossref
29.
Liebeskind  DS, Jahan  R, Nogueira  RG, Jovin  TG, Lutsep  HL, Saver  JL; SWIFT Investigators.  Early arrival at the emergency department is associated with better collaterals, smaller established infarcts and better clinical outcomes with endovascular stroke therapy: SWIFT study.  J Neurointerv Surg. 2016;8(6):553-558.PubMedGoogle ScholarCrossref
30.
Cheng-Ching  E, Frontera  JA, Man  S,  et al.  Degree of collaterals and not time is the determining factor of core infarct volume within 6 hours of stroke onset.  AJNR Am J Neuroradiol. 2015;36(7):1272-1276.PubMedGoogle ScholarCrossref
31.
Woitzik  J, Pinczolits  A, Hecht  N,  et al.  Excitotoxicity and metabolic changes in association with infarct progression.  Stroke. 2014;45(4):1183-1185.PubMedGoogle ScholarCrossref
32.
Garcia  JH, Yoshida  Y, Chen  H,  et al.  Progression from ischemic injury to infarct following middle cerebral artery occlusion in the rat.  Am J Pathol. 1993;142(2):623-635.PubMedGoogle Scholar
33.
Hakimelahi  R, Vachha  BA, Copen  WA,  et al.  Time and diffusion lesion size in major anterior circulation ischemic strokes.  Stroke. 2014;45(10):2936-2941.PubMedGoogle ScholarCrossref
34.
Lansberg  MG, Cereda  CW, Mlynash  M,  et al; Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution 2 (DEFUSE 2) Study Investigators.  Response to endovascular reperfusion is not time-dependent in patients with salvageable tissue.  Neurology. 2015;85(8):708-714.PubMedGoogle ScholarCrossref
35.
Abou-Chebl  A.  Endovascular treatment of acute ischemic stroke may be safely performed with no time window limit in appropriately selected patients.  Stroke. 2010;41(9):1996-2000.PubMedGoogle ScholarCrossref
36.
Bang  OY, Saver  JL, Kim  SJ,  et al; UCLA-Samsung Stroke Collaborators.  Collateral flow averts hemorrhagic transformation after endovascular therapy for acute ischemic stroke.  Stroke. 2011;42(8):2235-2239.PubMedGoogle ScholarCrossref
37.
Goyal  M, Menon  BK, Coutts  SB, Hill  MD, Demchuk  AM; Penumbra Pivotal Stroke Trial Investigators, Calgary Stroke Program, and the Seaman MR Research Center.  Effect of baseline CT scan appearance and time to recanalization on clinical outcomes in endovascular thrombectomy of acute ischemic strokes.  Stroke. 2011;42(1):93-97.PubMedGoogle ScholarCrossref
38.
Minnerup  J, Wersching  H, Teuber  A,  et al; REVASK Investigators.  Outcome after thrombectomy and intravenous thrombolysis in patients with acute ischemic stroke: a prospective observational study.  Stroke. 2016;47(6):1584-1592.PubMedGoogle ScholarCrossref
39.
Menon  BK, Sajobi  TT, Zhang  Y,  et al.  Analysis of workflow and time to treatment on thrombectomy outcome in the ESCAPE randomized controlled trial.  Circulation. 2016;133(23):2279-2286.PubMedGoogle ScholarCrossref
40.
Liang  JW, Stein  L, Wilson  N, Fifi  JT, Tuhrim  S, Dhamoon  MS.  Timing of vessel imaging for suspected large vessel occlusions does not affect groin puncture time in transfer patients with stroke.  [published online January 24, 2017].  J Neurointerv Surg. 2017;neurintsurg-2016-012854. doi:10.1136/neurintsurg-2016-012854PubMedGoogle Scholar
41.
McTaggart  RA, Yaghi  S, Cutting  SM,  et al.  Association of a primary stroke center protocol for suspected stroke by large-vessel occlusion with efficiency of care and patient outcomes.  JAMA Neurol. 2017;74(7):793-800. doi:10.1001/jamaneurol.2017.0477PubMedGoogle ScholarCrossref
42.
Khoury  N, Fahed  R, Guilbert  F,  et al.  Early CT changes in patients admitted for thrombectomy: intra- and interrater agreement and systematic review of the literature.  Neurology. 2016;86(16 suppl):S47.001.Google Scholar
×