Article Text

other Versions

Case series
Impact of immediate post-reperfusion cooling on outcome in patients with acute stroke and substantial ischemic changes
  1. Yang-Ha Hwang1,2,
  2. Ji-Su Jeon1,
  3. Yong-Won Kim1,2,3,
  4. Dong-Hun Kang2,3,4,
  5. Yong-Sun Kim2,3,
  6. David S Liebeskind5
  1. 1Department of Neurology, Kyungpook National University School of Medicine and Hospital, Daegu, South Korea
  2. 2Cerebrovascular Center, Kyungpook National University School of Medicine and Hospital, Daegu, South
  3. 3Department of Radiology,  Kyungpook National University School of Medicine and Hospital, Daegu, South Korea
  4. 4Department of Neurosurgery, Kyungpook National University School of Medicine and Hospital, Daegu, South Korea
  5. 5UCLA Stroke Center, University of California, Los Angeles, California, USA
  1. Correspondence to Dr David S Liebeskind, UCLA Department of Neurology, 635 Charles E Young Drive South, Suite 225 Los Angeles, CA 90095-7334, USA; davidliebeskind{at}


Background In patients with acute stroke and an extensive ischemic burden at baseline, the prognosis is usually poor despite timely reperfusion.

Objective To overcome universally poor outcomes in such patients, by applying immediate ‘post-reperfusion cooling’ in order to reduce reperfusion-related complications, and to describe the clinical and imaging characteristics.

Methods Patients having (1) an acute anterior large vessel occlusive stroke within 4.5 h since last known well, (2) Alberta Stroke Program Early CT Score (ASPECTS) ≤5 on baseline imaging, and (3) targeted temperature management with endovascular cooling after confirmed reperfusion were included in this study.

Results Eighteen patients (mean±SD age 59.5±10.9 years, median National Institutes of Health Stroke Scale score of 17, and median ASPECTS of 3) were analyzed. Median lesion volumes at baseline and after treatment were 130.2 and 110.6 mL, respectively. Median time from onset to the start of hypothermia and hypothermia duration were 213 min and 51 h, respectively. Favorable outcome (modified Rankin Scale ≤2) at 3 months was observed in 10 (55.6%) patients. Symptomatic intracranial hemorrhage, malignant brain edema, and pneumonia were observed in 2, 6, and 8 patients, respectively.

Conclusions The use of post-reperfusion cooling as a rescue treatment in patients with substantial ischemia at baseline might improve clinical outcome.

Statistics from


In treating patients with acute ischemic stroke in the anterior circulation, timely reperfusion using thrombectomy devices has become the key strategy to minimize infarct size and improve clinical outcome.1–5 However, reperfusion itself does not always guarantee good functional outcome, and patients with substantial ischemic changes (defined here as patients with an Alberta Stroke Program Early CT Score (ASPECTS) ≤5 on baseline imaging) may have poor outcomes regardless of treatment.6–9 Often, patients with extensive ischemic burden at baseline are excluded from reperfusion therapy owing to the high risk of reperfusion-related complications and subsequent poor prognosis.

Immediate post-reperfusion cooling can be a promising strategy to minimize reperfusion-related complications. In experimental stroke models, targeted temperature management (TTM) has been shown to be helpful as a potent neuroprotectant, especially in ischemic-reperfusion models.10 ,11 Also, recent clinical studies have shown the promising role of TTM together with endovascular reperfusion in patients with modest ischemic changes (mostly ASPECTS ≥5).12 ,13 However, no study has examined the effect of TTM in patients with substantial ischemic burden (defined here as ASPECTS ≤5 on baseline imaging). The purpose of this study was to describe the details of clinical and imaging characteristics in patients with substantial ischemic burden after TTM with endovascular cooling.


Between May 2010 and November 2014, patients were screened from a prospectively maintained acute stroke registry at our institution. Criteria for inclusion in this study were (1) anterior circulation stroke with proximal vessel occlusions (internal carotid artery (ICA)-T or -L (T and L indicate the shape of the occlusive lesion), middle cerebral artery (MCA) M1, or ICA/M1 tandem occlusion); (2) arrival at our emergency center within 4.5 h from last known well; (3) extensive ischemic changes on baseline scans, specifically ASPECTS ≤5 on diffusion-weighted imaging (DWI), or ≤4 on CT in MR-ineligible cases; (4) sufficient angiographic reperfusion (Thrombolysis In Cerebral Infarction (TICI) ≥2 in cerebral angiography or MR angiography) using endovascular treatment and/or intravenous thrombolysis; and (5) informed consent for TTM with endovascular cooling. The local institutional review board approved this study for retrospective analysis.

Endovascular treatment was considered if the treatable ICA or MCA occlusion persisted, and was initiated under conscious sedation. Treatment strategies were selected based on available therapies at the time of angiography, which included intra-arterial thrombolytic infusion, mechanical thrombectomy (including forced arterial suction thrombectomy or Solitaire thrombectomy), rescue intracranial/extracranial stenting, or a combination of these.14 ,15

‘Post-reperfusion cooling’ in this study was defined as immediate TTM in patients with acute anterior circulation stroke following confirmed reperfusion of proximal vessels. In patients eligible for post-reperfusion cooling, an Icy catheter (Zoll, Chelmsfold, Massachusetts, USA) was placed in the inferior vena cava via a femoral venous sheath, which was done in the angio suite or the intensive care unit (ICU) immediately after active reperfusion treatment. The patient was immediately taken to the ICU where the cooling catheter was connected to the cooling machine (CoolGuard system). Immediate cooling was started with the patient in the awake or drowsy state, depending on whether or not sedative drugs were used, to 34°C for more than 24 h, followed by gradual rewarming for more than 12 h at 0.1°C/h. The duration of maximal cooling and gradual rewarming was determined by the attending physician's decision based on the patient's neurological and systemic conditions. Shivering was prophylactically controlled with buspirone (10–30 mg every 8 h), meperidine (25–100 mg intravenous bolus), and counterwarming techniques consisting of air circulating blankets (BAIR Hugger, Arizant Healthcare, Minnesota, USA) warmed to the maximal temperature setting (43°C) during TTM. Additionally, mild sedation with midazolam (loading dose 10–40 µg/kg, maintenance 20–100 µg/kg/h), propofol (loading dose 0.5–1 mg/kg; maintenance 0.3–5.0 mg/kg/h), or dexmedetomidine (initial dose 0.5 µg/kg/h; maintenance, 0.2–1.0 µg/kg/h) was used when the initial measures were unsuccessful in controlling shivering. Measurement of shivering was based on the Bedside Shivering Assessment Scale (BSAS),16 specifically a BSAS score of 2 or 3. Temperature was recorded every hour from thermometers at the tip of the rectal catheter. The time of hypothermia initiation, target temperature, and rewarming initiation were also recorded. Hypothermia duration was defined as time between the start of hypothermia and the initiation of rewarming. Induction time was defined as the total time in minutes from hypothermia initiation to target temperature.

All patients were managed in accordance with recent guidelines using the resources available at our institute. In cases of refractory ischemic brain edema despite maximal medical treatment, such as head elevation and osmotherapy, surgical decompression was planned and performed in selected cases.17 ,18 All medical complications occurring during the therapeutic period were recorded. Pneumonia was recorded as present if there were any clinical findings at auscultation, along with new pulmonary infiltrate on the chest radiograph (persistent on follow-up for more than 48 h), with an increase in serum white cell count. Patients presumed to have pneumonia were actively treated with appropriate antibiotics. Electrolyte imbalance was recorded as present if any abnormal findings, including hypernatremia, hyponatremia, hyperkalemia, or hypokalemia, were present. Deep vein thrombosis (DVT) was screened if clinical suspicion by the attending physician, and ultrasonography was performed if DVT was strongly suspected.

Clinical and imaging parameters were collected and analyzed. Baseline ASPECTS and lesion volume were measured using baseline DWI (CT in one MR-ineligible case). To measure post-treatment lesion volume, 5–7-day fluid-attenuated inversion recovery images were used; this was available in 12 cases. In the remaining six cases, 1–3-day CT, excluding CT scans after decompressive craniectomy, was used. Lesion volume was calculated using open source image-analysis software (OsiriX Imaging Software; Patients had a CT or MRI scan within 48 h if clinical evidence of deterioration was observed to assess for edema or hemorrhagic changes. Malignant brain edema was defined as a midline shift of the infarcted area on follow-up CT or MRI. Symptomatic intracranial hemorrhage was defined as any type of hemorrhage associated with an increase in the National Institutes of Health Stroke Scale (NIHSS) score of ≥4 with 24 h of treatment.19 Functional status was assessed using the modified Rankin Scale (mRS) at 3 months.


A total of 18 patients (12 men and 6 women) with a mean age of 59.5±10.9 years, median NIHSS score of 17 (IQR 15–20.3), and median ASPECTS of 3 (IQR 2–4) were analyzed. Details of clinical and imaging features are summarized in figure 1. Median lesion volume at baseline and after treatment was 130.2 mL (IQR 105.4–163.4) and 110.6 mL (75.5–236.9), respectively (figure 2). Median time from onset of symptoms to hospital arrival was 74.5 min (IQR 41.5–129.5). There were 11 (61.1%) MCA and 7 (38.9%) ICA occlusions at baseline. Angiographic reperfusion was achieved using mechanical thrombectomy in 15/18 patients (83.3%). Reperfusion status before TTM with endovascular cooling was evaluable on cerebral angiography in 17/18 patients: 7 (41.2%) with TICI 2a, 5 (29.4%) with TICI 2b, and 5 with TICI 3. The American Society of Interventional and Therapeutic Neuroradiology/Society of Interventional Radiology collateral-flow grade was evaluable in 13/17 patients:20 11 (84.6%) with grade 0–1, 1 (7.7%) with grade 2, and 1 with grade 4. Median time from angiographic reperfusion to cooling initiation was 81 min (IQR 70.3–104.5). Median time spent for post-reperfusion cooling was 51 h (IQR 45.5–78.8). The parameters related to cooling are described in table 1.

Table 1

Post-reperfusion cooling parameters (n=18)

Figure 1

Summary of clinical and imaging features. Ischemic brain edema was classified as follows: grade 0, no edema or effacement of the cortical sulci only; grade 1, ventricular collapse; and grade 2, midline shift. ASPECTS, Alberta Stroke Program Early CT Score; HD, hypothermia duration; HI, hemorrhagic infarction; IAT, intra-arterial therapy; IBE, ischemic brain edema; ICAO, internal carotid artery occlusion; ICH, intracranial hemorrhage; IVT, intravenous thrombolysis; MCAO, middle cerebral artery occlusion; mRS, modified Rankin Scale; NA, not applicable; NIHSS, National Institutes of Health Stroke Scale; OTC=onset to cooling initiation; OTR, onset to reperfusion; PH, parenchymal hematoma.

Figure 2

Lesion volume changes in relation to clinical outcome. *Cases in which post-treatment lesion volumes were calculated using non-contrast CT of 1–3 days. In case 8, both pre- and post-treatment lesion volumes were calculated using non-contrast CT owing to MR ineligibility. mRS, modified Rankin Scale.

Favorable outcome (mRS 0 to 2) at 3 months was seen in 10/18 (55.6%) patients. Pneumonia was seen in (44.4%) patients, and they were actively treated with appropriate antibiotics. Malignant brain edema was seen in 6 (33.3%) patients during the maintenance phase of TTM. Of these patients, three underwent decompressive craniectomy, but they were functionally dependent at 3 months. The reasons for neurosurgical intervention were malignant edema due to expanding parenchymal hematoma in two cases, and malignant ischemic edema only in one case. Surgery was carried out within 2 days, except in one case (79 h from onset) where delayed expansion of parenchymal hematoma was shown. The remaining three patients were managed conservatively, and one patient died from associated malignant edema, while two patients recovered to the level of functional independence. Symptomatic intracranial hemorrhage was observed in two patients, and they underwent neurosurgical intervention, including decompressive craniectomy, and the outcome was dismal (mRS 4 and mRS 5 at 3 months, respectively).


Management of acute stroke in patients with substantial ischemic burden (defined here as ASPECTS ≤5 at baseline) often excludes reperfusion, as it often results in a poor outcome. In such cases, immediate post-reperfusion cooling could be a promising strategy to improve outcomes. The main result of this study was that slightly more than half our patients (55.6%) with a substantial ischemic burden undergoing reperfusion and post-reperfusion cooling had a favorable outcome (mRS≤2) at 3 months.

Most studies of TTM in acute stroke focus on two main beneficial effects (modulation of brain swelling and neuroprotection), separately.21 However, both mechanisms of TTM should be considered simultaneously in patients who show sizable ischemic changes at baseline despite early presentation, as in our case series. Two previous studies evaluated the effect of TTM after active reperfusion, and showed it to be safe and effective (table 2).12 ,13 However, the imaging inclusion criteria were modest in terms of the ischemic burden (mostly ASPECTS ≥5 in both studies), so the duration of hypothermia was relatively short, focusing on neuroprotective mechanisms of TTM.21 In our case series, the hypothermia duration was relatively long compared with the aforementioned studies (table 2), with the mechanism of TTM focusing on both neuroprotection and control of intracranial pressure.

Table 2

Main findings of relevant studies, including current case series involving patients with substantial ischemic changes at baseline

The applicability of TTM in patients with substantial ischemic changes has so far been unknown. Our case series is the first to report the outcome of therapeutic cooling following reperfusion for patients with a low ASPECTS at baseline. Most previous cohorts of patients with a low ASPECTS have shown poor outcomes despite reperfusion.6–8 Moreover, one study showed that it is very difficult to attain full reperfusion in such cases (with a TICI 2a reperfusion rate of 41.2% for patients with a low ASPECTS).22 Therefore, the moderate success found in our study is encouraging. Also, the angiographically defined collateral-flow status was mostly poor in our case series, which coincided with the previous report underlining the role of collaterals in determining baseline ischemic burden.23

Some complications occurred—the rate of pneumonia, one of the most feared complications of TTM, was high in our study (44.4%), although patients were actively treated with appropriate antibiotics. The rate of symptomatic hemorrhage was 11.2%, which was similar to that found in previous studies.12 ,13 Malignant brain edema was seen in 6 (33.3%) patients; and one patient died from associated malignant edema. Nevertheless, a favorable outcome was obtained in more than half of our cases, which might highlight the role of post-reperfusion cooling in patients with extensive ischemia and inadequate collaterals.

The limitations of this study include its small sample size, retrospective design, and patient selection bias despite using a prospective registry, any of which might have affected our results. Also, two cases were not treated with TTM of sufficient duration, because they developed malignant edema and/or intracranial hemorrhage. So, neurosurgical intervention was applied and TTM was stopped early, which diluted the beneficial effects of TTM in these cases. Another limitation was that the age and sex distribution of the patients in our study differed from those of the general stroke population, which might have affected the results.


Selecting patients predicted to benefit from reperfusion therapy does not imply that those excluded are beyond help. In treating patients having substantial ischemic changes at baseline, the application of post-reperfusion cooling could be a viable option in improving functional outcome. Therefore, a future, prospectively designed trial should be repeated to confirm these findings.


View Abstract


  • Twitter Follow Yang-Ha Hwang at @yangha73

  • Contributors Study concept and design: Y-HH and DSL. Analysis and interpretation of data: Y-HH, J-SJ, Y-WK, D-HK. Drafting of the manuscript: Y-HH. Critical revision of the manuscript for important intellectual content: Y-SK, DSL.

  • Competing interests None declared.

  • Patient consent No.

  • Ethics approval Kyungpook National University Hospital.

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

  • Data sharing statement Not applicable.

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.