Article Text

Download PDFPDF

Post-thrombectomy management of the ELVO patient: Guidelines from the Society of NeuroInterventional Surgery
  1. Thabele Leslie-Mazwi1,
  2. Michael Chen2,
  3. Julia Yi3,
  4. Robert M Starke4,
  5. M Shazam Hussain5,
  6. Philip M Meyers6,
  7. Ryan A McTaggart7,
  8. G Lee Pride8,
  9. Sameer A Ansari9,
  10. Todd Abruzzo10,
  11. Barbara Albani11,
  12. Adam S Arthur12,
  13. Blaise W Baxter13,
  14. Ketan R Bulsara14,
  15. Josser E Delgado Almandoz15,
  16. Chirag D Gandhi16,
  17. Don Heck17,
  18. Steven W Hetts18,
  19. Richard P Klucznik19,
  20. Mahesh V Jayaraman7,
  21. Seon-Kyu Lee20,
  22. William J Mack21,
  23. J Mocco22,
  24. Charles Prestigiacomo23,
  25. Athos Patsalides24,
  26. Peter Rasmussen5,
  27. Peter Sunenshine25,
  28. Donald Frei26,
  29. Justin F Fraser27
  30. on behalf of the Standards and Guidelines committee of the Society of NeuroInterventional Surgery (SNIS)
  1. 1 Neurointerventional Service, Massachusetts General Hospital, Boston, Massachusetts, USA
  2. 2 Department of Neurological Sciences, Rush University Medical Center, Chicago, Illinois, USA
  3. 3 University Illinois at Chicago, Chicago, Illinois, USA
  4. 4 Department of Neurosurgery and Radiology, University of Miami, Miami, Florida, USA
  5. 5 Cerebrovascular Center, Cleveland Clinic, Cleveland, Ohio, USA
  6. 6 Columbia Presbyterian Hospital, New York, New York, USA
  7. 7 Warren Alpert Medical School of Brown University, Providence, Rhode Island, USA
  8. 8 Department of Neuroradiology, University of Texas Southwestern, Dallas, Texas, USA
  9. 9 Departments of Radiology, Neurology, and Neurological Surgery, Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA
  10. 10 Department of Neurosurgery, University of Cincinnati, Cincinnati, Ohio, USA
  11. 11 Department of Neurointerventional Surgery, Christiana Care Health Systems, Newark, Delaware, USA
  12. 12 Semmes-Murphey Clinic, Memphis, Tennessee, USA
  13. 13 Department of Radiology, Erlanger Medical Center, Chattanooga, Tennessee, USA
  14. 14 Department of Neurosurgery, Yale University School of Medicine, New Haven, Connecticut, USA
  15. 15 Department of Neurointerventional Radiology, Abbott Northwestern Hospital, Minneapolis, Minnesota, USA
  16. 16 Department of Neurosurgery, Rutgers-New Jersey Medical School, Newark, New Jersey, USA
  17. 17 Department of Radiology, Forsyth Medical Center, Winston Salem, North Carolina, USA
  18. 18 Department of Radiology, University of California in San Francisco, San Francisco, California, USA
  19. 19 Houston Methodist Neurological Institute, Houston Methodist Hospital, Houston, Texas, USA
  20. 20 The University of Chicago, Chicago, Illinois, USA
  21. 21 Keck School of Medicine, University of Southern California, Los Angeles, California, USA
  22. 22 Mount Sinai School of Medicine, Mount Sinai Health System, New York, New York, USA
  23. 23 East Hanover, New Jersey, USA
  24. 24 New York Presbyterian Hospital, Weill Cornell Medical College, New York, New York, USA
  25. 25 Banner Good Samaritan Medical Center, Phoenix, Arizona, USA
  26. 26 Swedish Medical Center, Denver, Colorado, USA
  27. 27 Department of Neurological Surgery, University of Kentucky, Lexington, Kentucky, USA
  1. Correspondence to Dr Justin F Fraser, Department of Neurological Surgery, University of Kentucky, Lexington, KY 40536, USA; Jfr235{at}

Statistics from

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.


Recent trials have proven the benefits of endovascular treatment for patients with stroke from emergent large vessel occlusions (ELVOs).1–5 Optimal management of these patients involves attention to pre-procedural, intra-procedural, and post-procedural elements. However, many of the ideal treatment approaches following endovascular stroke therapy remain controversial. This document synthesizes current recommendations from the best available evidence to provide guidance in the post-procedural management of a patient undergoing stroke thrombectomy.

Materials and methods

This document was constructed by the Standards and Guidelines Committee of the Society for NeuroInterventional Surgery, a multidisciplinary committee composed of practitioners with backgrounds including neuroradiology, vascular neurosurgery, stroke neurology, and neurocritical care. We reviewed electronic databases for publications related to the management of acute stroke patients post-procedure, using both broad and narrow search terms. We subsequently evaluated those results for papers with randomized clinical results, which were given the highest priority. The remaining papers were assessed on the basis of individual methodology, and recommendations were made based on the data available. In the absence of supporting adequate clinical trial evidence, the committee made consensus recommendations. Each recommendation is graded, where possible, with a level of evidence utilizing the American Heart Association/American Stroke Association grading system.6 This document represents one of a continuum related to acute stroke intervention, including other documents on prehospital management, training standards for thrombectomy, and management of ELVO patients.7–9


Post-thrombectomy care environment

ELVO patients require careful monitoring in a stroke unit or intensive care unit. Stroke units provide dedicated, specialized, multidisciplinary inpatient care for ELVO patients. Patients treated in this environment are more likely to survive, regain independence, and return home than those receiving less organized service.10 Stroke units are characterized by protocol guided care, adherence to guidelines, and coordination of care provided by various services.11–14 Furthermore, a dedicated stroke unit is preferable to a mobile consultative stroke service.10 While there is ongoing debate about the relative superiority of stroke care at designated comprehensive stroke centers, their requirements for designation as comprehensive stroke centers embody some vital qualities for stroke units in general.15 Higher staffing models to accommodate weekend coverage, along with uninterrupted access to advanced imaging modalities and intensive care provide vital assets with proven effects on reducing stroke mortality.16–20


  • Patients undergoing stroke thrombectomy are more likely to regain functional independence if they receive post-procedure inpatient care in a designated stroke unit with access to a coordinated multidisciplinary team. This benefit applies to patients irrespective of age, sex, or stroke severity. (Class I, level of evidence B)

  • Of the various levels of stroke care, comprehensive stroke centers or the local equivalent are most likely to provide post-procedural care that is specific of the needs of patients after ELVO treatment. Stroke units should therefore be prioritized for ELVO post-procedural care disposition. (Class I, level of evidence B)

Preserving threatened tissue (failed or incomplete recanalization)

The reduction in regional cerebral blood flow (CBF) caused by an ELVO is demarcated by regions of severe reduction (core) and moderate reduction (penumbra). The penumbra may remain viable for many hours, as collateral perfusion maintains tissue above the threshold for infarction, estimated at 10 mL/100 g/min of CBF.21 22 ‘Collateral circulation’ refers to the vascular channel network that compensates for CBF reduction when flow in the principal conduits fails. These connections may be divided into primary and secondary channels. Primary collaterals provide immediate diversion of CBF to ischemic regions through existing anastomoses (such as the anterior or posterior communicating artery). Secondary collaterals, such as leptomeningeal anastomoses, may be anatomically present, although their capacity to meet metabolic needs may require time to develop (as the leptomeningeal vessels enlarge and remodel over time).23 Successful acute stroke intervention preserves that penumbral tissue by restoring flow through the principal conduit. However, a variety of techniques have been evaluated to improve collateral support to penumbra, optimizing tissue preservation after unsuccessful or partially successful recanalization.

Therapeutic hypertension

The most common mechanism of hemodynamic augmentation is hypertensive therapy for collateral optimization, which has a sound pathophysiological basis. In the presence of ischemia, physiologic cerebral autoregulation is impaired, and cerebral vessels become pressure passive (ie, they fail to demonstrate a vasoconstrictive response to decreasing perfusion pressure).24 Elevation of mean arterial pressure (MAP) can augment both cerebral perfusion pressure and mean flow velocity in this context.25 26 Therefore, blood pressure elevation should not only enhance CBF, oxygen delivery, and removal of metabolic waste in penumbral tissue, but also augment collateral channels through increased hydrostatic pressure.27 However, defining specific blood pressure goals has been challenging in the setting of small retrospective studies, potential selection bias, and significant variability in therapeutic agents, blood pressure (BP) targets, and management protocols. Exposure to intravenous tissue plasminogen activator (IV tPA) is not a contraindication, but does limit the upper extent of BP elevation. Patients with intracranial hemorrhage, congestive heart failure, active coronary ischemia (changes on ECG or troponin leak), or systolic BP >200 are generally excluded.28 29

The therapeutic aim should be to increase MAP by 10–20% above baseline (or increase systolic BP to a maximum of 185 mmHg if IV tPA has been given). This is accomplished by withdrawing all antihypertensive medications, providing fluid boluses (eg, 500 mL–1 liter of normal saline, then maintenance fluids), patient positioning, and vasopressor therapy, as indicated. Phenylephrine is the preferred agent because it has been most studied; norepinephrine is a reasonable alternative.28 29 There is less evidence for the use of other vasopressors as a primary agent. Patients should be admitted to the intensive care unit with continuous arterial pressure monitoring, and with central venous access if vasopressors are being used. Monitoring the potential dangers of induced hypertension includes evaluation of chest X-rays (urgent echocardiogram in selected cases), ECG, central venous pressure if a central line is implanted, urine output, and daily labs, including cardiac enzymes. Major concerns would be intracerebral hemorrhage, pulmonary edema, and cardiac arrhythmias, but these are uncommon.30 31

Assessing clinical improvement during this period of hemodynamic collateral augmentation has been accomplished by measuring improvement in the National Institutes of Health Stroke Scale (NIHSS) score (thresholds of 4 or greater are reported in the literature).28 29 Functionally significant improvement (eg, improved fine motor function of the hand or more fluent language output) should also be considered. Duration of collateral support should be long enough to enable adaptation of these collateral connections to durably accommodate increased blood flow, estimated to require 24 hours at least. Particular attention should be given to periods of sleep, when blood pressure tends to be physiologically lower and collaterals could be insufficient. After 24 hours, weaning hemodynamic support is individualized, but should be accomplished gradually and with a return to prior support levels if the patient becomes symptomatic. In some patients, hemodynamic support may be required for several days. For example, vasopressor dosages can be decreased every 2–4 hours with revised systolic BP or MAP parameters, followed by a reduction in IV fluids, followed by gradual orthostatic challenges (eg, sitting in bed, progressing to chair position in bed, progressing to seated out of bed, progressing to standing). For the occasional patient who is unable to tolerate weaning hemodynamic support, longer term support may be needed with addition of an oral hypertensive agent such as fludrocortisone or midodrine.

Cerebral perfusion to threatened penumbral tissue can also be readily modified with patient head positioning. For patients able to protect their airway, the supine position may improve cerebral perfusion by increasing CBF, perhaps by as much as 15–20%.32 33 For those at risk of aspiration, the head of the bed should be positioned at 30°.34

Other methods to sustain the ischemic penumbra that have had limited success include vasodilators, modifications to blood viscosity,35 circulating volume expansion,36 albumin,37 and flow augmentation using intra-aortic balloon catheter placement.38 None of these interventions has been proven to improve patient outcomes.


  • Sustaining ischemic penumbral tissue through hemodynamic support should be considered in patients with unsuccessful or partially successful recanalization. Induced hypertension and patient positioning are likely to be most helpful, although optimized protocols remain to be determined. Collateral support should continue for at least 24 hours before attempts to wean. (Class IIb, level of evidence B)

Complication avoidance and management

Reperfusion injury and hemorrhagic transformation

Reperfusion to salvage the penumbra can produce cerebral damage through reperfusion injury and hemorrhagic transformation. Hemorrhagic transformation refers to bleeding within the bed of infarcted cerebral tissue. Reperfusion injury and hemorrhage can be viewed as a continuum, essentially determined by pathologic permeability of the blood–brain barrier.39 Ischemia leads to increased permeability of the blood–brain barrier through activation of a variety of enzymes,40 the most prominent being the matrix metalloproteinases (MMP), in particular MMP-9.41 This process is exacerbated by circulating tPA.42 Reperfusion of ischemic cerebral arteries contributes to the risk of hemorrhage,43 with the highest risk among hypertensive extremes (>220 mm Hg/105 mm Hg).44 Mechanical thrombectomy is associated with higher reperfusion rates, but recognized to have low rates of reperfusion injury and hemorrhagic conversion (approximating 5.7% compared with 5.1% for IV tPA alone in pooled analysis of randomized mechanical thrombectomy data; p=0.56).42 Further, these rates do not appear to be exacerbated by IV tPA administration.45 However, studies have demonstrated some predictors of hemorrhagic change, including hyperglycemia, exposure to tPA, hypertension, coagulopathy, and large pretreatment infarct core.46 Typically, patients present with neurological deterioration, even after initial improvement. For some patients this decline is catastrophic. Risk of bleeding declines throughout the first 24 hours, and fatal hemorrhage is rare after that time point unless external causes are present (eg, anticoagulation).

The perceived risk of hemorrhage determines the role and frequency of follow-up imaging in all patients after endovascular therapy. Conventionally, post-thrombectomy patients are imaged within 24 hours or immediately on neurological deterioration. Several classification systems for hemorrhagic transformation exist, with varying definitions of ‘symptomatic intracranial hemorrhage’. The National Institute of Neurological Disorders and Stroke (NINDS),47 Safe Implementation of Thrombolysis in Stroke–Monitoring Study (SITS-MOST),48 and the European Cooperative Acute Stroke Study II (ECASS-II)49 definitions are most commonly used. An important limitation is the inability to distinguish between hemorrhage and neurological deterioration from other causes (seizure, infection, metabolic factors, hemodynamic failure, etc.). The ECASS-II grading scale (table 1) is favored because of its simplicity and prior validation.50 Symptomatic intracranial hemorrhage is further defined as neurological worsening producing an increase of 4 points of more in the post-procedure NIHSS score occurring in the first 7 days after treatment (most occur within 36 hours). While any symptomatic intracranial hemorrhage decreases the probability of eventual functional independence, patients with parenchymal hematoma type 2 (PH-2) hemorrhage are most consistently adversely affected50; for this reason, PH-2 is most commonly used as a primary safety outcome measure in many ischemic stroke studies and hospital metrics.

Table 1

ECASS-II hemorrhage grades for post-procedure imaging evaluation. The scale was developed to assess hemorrhage related to intravenous thrombolysis, but has broader applicability

Hyperattenuated lesions are commonly seen on post-procedure CT imaging, reportedly in up to 84.2%. Women with a higher NIHSS score on admission, exposure to thrombolytics, hyperglycemia, and a significant contrast load more often have this finding.51 52 Distinguishing between contrast extravasation and tissue hematoma can be challenging. Some strategies include serial non-contrast CT imaging, MRI with gradient echo evaluation, and dual energy CT. The sensitivity, specificity, and accuracy of dual energy CT in the identification of hemorrhage were 100% (6 of 6 areas), 91% (20 of 22 areas), and 93% (26 of 28 areas), respectively, in an early study evaluating this modality.53 Even contrast staining, however, is not benign. Renu et al recently published findings related to the clinical significance of contrast staining compared with hemorrhage on dual energy CT scans after stroke treatment. Both contrast staining and hemorrhage were associated with worse outcomes (OR 11.3 and 10.4, respectively), and contrast staining was more likely to be associated with delayed hemorrhagic transformation (OR 4.5).54 Despite the significance of CT hyperdensity remaining debatable, with conflicting findings in non-homogenous samples,51 52 55 the finding of contrast staining should inform the post-procedural approach to blood pressure management and other factors that may impact a patient’s hemorrhagic risk. However, the impact on patient outcomes requires further study.

While idealized blood pressure management remains a topic of intense debate for stroke patients in general,56 it seems likely that persistently elevated arterial pressures in the context of recent endovascular recanalization and existing ischemic injury may be detrimental. This is certainly the case following hemorrhagic transformation which may lead to continued hemorrhage, repeat hemorrhage, and edema. Maintaining a systolic BP <140 mm Hg or 160 mm Hg is reasonable following detection of or suspicion for hemorrhage; the use of rapidly acting intravenous agents (nicardipine, labetalol, enalapril, etc)57 helps to balance the risk for hemorrhage against maintaining adequate cerebral perfusion.

Reversal of tPA is also an important consideration in hemorrhaging after stroke with tPA treatment. The post-tPA hypocoagulable state outlasts the duration of drug action, and can results in bleeding (intracranial and systemic) in 6–10% of patients.58–60 For patients who receive IV tPA, any abnormal bleeding within the first 24 hours after administration should prompt consideration of tPA reversal. If the tPA infusion is still running when bleeding is detected, the thrombolytic should be stopped. There are no widely accepted approaches to reversal of tPA related hemorrhage, so institutional protocols may vary. However, certain concepts are common: emergent imaging, laboratory analysis (coagulation parameters, fibrinogen level, hemoglobin and hematocrit, type and screen), infusion of fresh frozen plasma or cryoprecipitate to replenish fibrinogen levels, and consideration of platelet administration or addition of antifibrinolytic agents (eg, tranexamic or aminocaproic acid).


  • Postoperative imaging to monitor for post-thrombectomy hemorrhage is appropriate and useful. (Class IIb, level of evidence C)

  • Post-reperfusion blood pressure management should take into account the patient’s baseline blood pressure ranges, balancing reperfusion needs against the risk of hemorrhage. (Class IIb, level of evidence C)

  • Reversal of tPA can and should be considered in post-thrombectomy patients experiencing significant bleeding complications from tPA. (Class IIB, level of evidence C)

Malignant cerebral edema

Cerebral edema is a significant post-procedural concern in the care of ELVO patients. The edema that results from cerebral ischemia is caused by intracellular accumulation of sodium and other ions in neurons and glial cells, resulting in cellular swelling.61 This produces cytotoxic edema, cerebral swelling, and tissue shifts. For patients with proximal occlusions the volume of tissue affected may be large enough to produce potentially life threatening (‘malignant’) cerebral or cerebellar edema. Brain herniation can then result with regional changes in intracranial pressure (ICP).62 Transtentorial herniation accounts for almost 80% of deaths in the first week after supratentorial infarction and almost 1 in 3 deaths during the first 30 days.63 64 Certain features increase the risk of malignant cerebral edema after an ELVO. Extensive pre-procedure infarction, procedural delays, and unsuccessful recanalization increase the risk of post-procedural challenges from edema. Younger patients have less cerebral atrophy and cannot accommodate the space occupying effect of cerebral edema as readily as the elderly.65 As such, younger patients are deemed higher risk for malignant cerebral edema after ELVO. Isolated middle cerebral artery (MCA) branch occlusions do not typically produce a sufficient stroke volume to cause malignant edema,62 so concern should be high in ELVO patients with more proximal occlusions. Inadequate collaterals further identify patients at risk, particularly those with isolated circulations (eg, truncated circle of Willis, fetal posterior cerebral artery) and poor pial collaterals.66 Malignant cerebral edema after MCA occlusion has been estimated to occur in 20% of cases. Diffusion weighted MRI lesion volume >80 mL predicts a malignant MCA infarct profile with high specificity (0.98, 95% CI 0.94 to 1.00), but low sensitivity (0.52, 0.32–0.71).67 Malignant edema for posterior fossa strokes is less predictable, but likely lower in incidence,68 with only 50% with radiographic mass effect showing clinical deterioration.69 For those patients not recanalized, early MRI may help stratify risk for cerebral edema by determining infarct volume.

The greatest risk for cerebral edema is within the first 48–72 hours, with up to 30% of patients decompensating within 24 hours.70 If decompensation occurs abruptly, hemorrhagic transformation should be suspected. More delayed decompensation is likely due to progressive loss of ischemic penumbra with subsequent swelling in delayed fashion, and can occur as late as 8 days after initial presentation.71 There should be concern for patients with a post-procedure NIHSS of ≥15 in the non-dominant or ≥20 in the dominant hemisphere.72 Clinical thresholds in posterior fossa stroke are less clear because significant space occupying cerebellar infarction can be present with relatively low NIHSS scores. The most concerning clinical sign is a decrease in level of arousal. This is best assessed through repeat bedside neurological examination, although there is no standardized clinical variable that reliably quantifies this change. Diminishing consciousness may reflect compression of the activating reticular formation at the level of the midbrain and components of the thalamus.73


  • Clinical and imaging data should be utilized to identify patients at high risk for malignant cerebral or cerebellar edema. (Class I, level of evidence B)

  • Regular frequent neurological examinations to follow the patient’s level of arousal, pupillary findings, and motor responses must be monitored in patients at risk of malignant cerebral or cerebellar edema. (Class I, level of evidence C)

  • Early imaging findings that demonstrate significant tissue involvement within the first 6 hours predict significant cerebral edema. These include MCA hypodensity in one-third of the MCA territory, diffusion weighted imaging volume >80 mL, or midline shift. (Class 1, level of evidence B)

Interventions for malignant cerebral edema can include ICP monitoring, head of bed positioning, osmotherapy, cooling, hyperventilation, and surgical decompression. With regard to ICP monitoring, malignant cerebral infarcts are more often the cause of clinical deterioration via herniation syndromes, as opposed to global ICP elevations. Therefore, there is no clear role for ICP monitoring in a patient with ELVO.74 Elevation of the head of bed decreases cerebral perfusion pressure and, by extension, the degree of cerebral edema, although data are limited.72 75 Elevation of the head of the bed to between 30 and 45° is reasonable for the patient at risk of malignant cerebral edema.

Osmotic therapies extract intracellular and interstitial fluid along an osmotic gradient into the blood vessel. With high reflection coefficients, these agents typically do not cross the blood–brain barrier. The most commonly used agents are either hypertonic saline or mannitol, which has gained widespread acceptance despite a paucity of published data.76 Hypertonic saline does appear superior to mannitol, perhaps because of its higher reflection coefficient (1 as opposed to 0.9).77 78 Mannitol has the distinct advantage of peripheral administration, which is not true for hypertonic saline concentrations above 3%, which risk venous sclerosis if infused peripherally. During osmotherapy with mannitol, it is crucial to maintain a hyperosmotic euvolemic state through adequate fluid balance. The role of single dose, recurrent bolus therapy, gradual elevation of systemic osmoles, or combinations of these remains unclear in terms of preferred method and dose of agent.79 80 Hypertonic saline is available as 1.5%, 3%, 7.5%, and 23%, while mannitol has been used at doses of 0.5–1 g/kg up to every 4–6 hours. Typically, serial sodium and osmolarity levels are followed to trend the effect of the medications.81 Limited data are available on the specifics of appropriate thresholds, and on the most appropriate frequency of monitoring. Most centers evaluate relevant bloodwork every 6–8 hours in ELVO patients at risk of malignant cerebral edema. Other agents evaluated in the context of acute stroke include hypertonic saline–hydroxyethyl starch,82 tromethamine buffer, corticosteroids,83 and barbituates.84 These agents are not currently standard of care for malignant cerebral edema in ELVO patients.

Hypothermia has been extensively evaluated in the management of cerebral ischemia.85 86 Therapeutic hypothermia is potentially neuroprotective through multiple mechanisms, including decreasing metabolic demand, inhibition of excitatory neurotransmitters and free radicals, and decreasing ICP through reductions in cerebral blood volume and edema.87 While initial results in adults have shown promise, controversies persist regarding methods of cooling, duration of cooling, and protocols for rewarming. Durations as long as 48 hours88 and temperatures as low as 33° F89 have been trialed. At this point, no clear evidence exists to support its generalized use, although further trials are needed.

Hyperventilation is most often used as a bridge between failing medical management and urgent/emergent surgical intervention. Hyperventilation (with a target PCO2 of 30 mm Hg) produces a respiratory alkalosis which in turn yields cerebral vasoconstriction.90 The effect is short lived (usually 1–3 hours) because cerebral interstitial buffers eventually restore a more normal pH. There is concern about hyperventilation exacerbating herniation because this effect is more pronounced in normal tissue. Hyperventilation may be most helpful while a herniating patient is being transported to the operating room.

Decompressive surgery reduces mortality for ELVO patients who develop malignant cerebral edema. In the face of tissue swelling, surgery aims to remove the unyielding barrier of the skull, allowing for accommodation of the expanding brain. Decompressive hemicraniectomy can reduce ICP to a greater degree when compared with any other treatment option.91 A generous bone flap is generally required,92 with smoothed bone edges, sparing of the midline, and extending to the floor of the middle cranial fossa. Parenchymal hemorrhage at the edges of the hemicraniectomy site is an ominous finding, suggesting too small a craniectomy, or sharp bone defects.93 The use of decompressive hemicraniectomy is supported by evidence. It became standard of care in 2007 following the publication of a pooled analysis from three randomized controlled trials94; HAMLET (Hemicraniectomy After Middle Cerebral Infarction With Life-threatening Edema Trial),95 DESTINY (Decompressive Surgery for the Treatment of Malignant Infarction of the Middle Cerebral Artery),96 and DECIMAL (Decompressive Craniectomy in Malignant Middle Cerebral Artery Infarcts)97 trials. These studies recruited patients aged 18–60 years with malignant MCA infarction. A strong survival benefit was found, with survival increasing from 25% to 80% in the surgical arm with a number needed to treat of 2. Survival with a modified Rankin Scale (mRS) score of ≤3 doubled (number needed to treat of 4), but 45% of survivors had moderate to severe disability at 1 year (surviving with an mRS score of 4 increased 10 fold). The side of the affected hemisphere should not be a factor in the decision to pursue hemicraniectomy.98 The utility of this treatment in patients older than 60 years has been assessed in two further randomized trials, both finding the mortality benefit to be preserved, but with limited functional improvement.99 100 With nearly uniform poor functional outcome in elderly patients, the benefit of hemicraniectomy continues to be debated. Ideal timing and imaging criteria for decompression remains unclear. Decisions based exclusively on age and infarct size without concern for clinical examination raises the concern of unnecessary procedures. However, delaying decompression until clinically evident neurological changes are observed risks decompressing patients too late to realize the benefit of the therapy. Studies identifying patients at high risk for whom prophylactic intervention would benefit are ongoing.101 Many institutions use early mental status changes or aggressive edema and tissue shift on early serial imaging as absolute criteria for operative intervention. A further consideration for the ELVO patient is the use of antiplatelet therapy post-intervention, and whether this affects the surgical risks associated with hemicraniectomy. No data are available to guide such a decision, and cases should be managed individually. In the posterior fossa, concern should be given to any patient with >50% of a posterior inferior or superior cerebellar artery vascular territory affected. If there is hydrocephalus, an external ventricular drain may be required, but this is infrequently done in isolation. Approximately 40% of patients with massive cerebellar strokes require suboccipital decompression, and up to 74% of these patients survive with an mRS score ≤2.102 This is because of the relative lack of eloquence of the cerebellum compared with other regions of the brain. It has functional redundancy, and even large infarcts can be well compensated for. Therefore, suboccipital decompression should be aggressively pursued if indicated.


  • Intracranial pressure monitoring has no defined role in ELVO. (Class III, level of evidence C)

  • Hyperosmolar agents may benefit patients with cerebral edema following large volume stroke. (Class IIa, level of evidence C)

  • Hyperventilation has short term benefit on acute herniation; it should be used as a bridging therapy just prior to definitive surgical management. (Class III, level of evidence B)

  • Prophylactic hyperventilation is not recommended. (Class III, level of evidence B). Hypothermia and use of other neuroprotectant strategies may be beneficial, but data are insufficient to support clinical use in adults after large vessel stroke. (Class III, level of evidence C)

  • Decompressive craniectomy with dural expansion should be considered in all patients <60 years of age with large volume infarctions who decompensate, or who are at imminent risk of decompensation despite medical management. (Class I, level of evidence B)

  • Decompressive craniectomy may be considered in patients >60 years of age, however the mortality benefit may not be accompanied by functional recovery. (Class IIb, level of evidence C)

  • External ventricular drain placement and suboccipital craniectomy with dural expansion should be pursued for patients with cerebellar stroke who deteriorate or are at imminent risk of decompensation despite medical management. (Class I, level of evidence B)

  • Family should be informed that despite decompressive craniectomy up to half of survivors will remain severely disabled. In contrast, following suboccipital craniectomy the majority of patients make a favorable neurologic recovery. (Class IIb, level of evidence C)

Access site complications

Arterial access complications often manifest while ELVO patients are in the post-procedure care unit. ELVO patients can be a particularly high risk for access site complications, due to the inherent risk factors of that disease population: coagulopathy (underlying or drug related), usage of tPA, prior groin access, existing peripheral vascular disease, obesity, advanced age, underlying coagulopathy, size of the access sheath, and the location of the arteriotomy. Women are more likely than men to have complications, for unclear reasons.103 104 Decreasing the sheath size may reduce the incidence of groin complications.105 While poorly defined in the stroke population, data on these complications, including length of hospital stay, morbidity, and mortality are well established in the cardiac literature.104 Complications include hemorrhage (usually either a hematoma contained in the tissues of the thigh, or distributed in the retroperitoneal compartment), dissection, pseudoaneurysm development, arteriovenous fistula formation, peripheral arterial thrombosis or embolism, and rarely infection. These complications can be compounded by the use of large bore systems (7 F or larger) in patients with concomitant IV tPA. Hemorrhage and hematoma are usually evident within 6 hours of concluding the stroke treatment. Signs and symptoms include access site pain, site swelling, falling hematocrit, and hemodynamic insufficiency. Other potential concerns such as pseudoaneurysm formation and an arteriovenous fistula may not be detected for days to weeks afterward. There should be a low threshold for imaging with ultrasound or CT angiography, with vascular surgery consultation if site complication is suspected.103

Debate exists about the optimal method of achieving access site hemostasis post-thrombectomy. Manual compression is time consuming and requires prolonged immobility. Longer duration of compression is directly correlated to decreased complications.106 Arterial access closure devices are now widely used and available, and include mechanical seals and collagen plugs.107 108 However, these devices have reported risks, including a risk of groin hematoma, pseudoaneurysm formation, and infection (as high as 0.3% in one series).109–115 In immunocompromised patients, periprocedural antibiotics are recommended in addition to strict adherence to aseptic technique. Prior to the use of a closure device an angiographic images of the femoral artery should be obtained to determine the suitability of using a closure device as well as provide a reference if a complication were to occur.116


  • Closure devices are useful in the appropriate clinical context, with similar complication rates. There is a modest advantage of immediate hemostasis that may allow for faster patient mobilization post-procedure. (Class I, level of evidence C)

  • Significant complications can develop acutely or subacutely at the access site, with investigations and interventions that may be urgently or emergently needed. Appropriate and standardized monitoring strategies should be used to detect these complications in the post-procedural setting. (Class I, level of evidence C)

Post-procedural rehabilitation

All patients admitted with acute stroke should have an initial assessment by multidisciplinary rehabilitation professionals (physical, occupational, and speech therapy) as soon as possible after admission, preferably within the first 24–36 hours. Combination therapies are often needed to optimize stroke recovery. An increasing role for consultation services from physical medicine and rehabilitation physicians has emerged.117 118 Apart from directed guidance of therapists, physical medicine and rehabilitation expertise extends to arousal, delirium management, communication aids, appropriate positioning and splinting, spasticity and salivary management, and assistance with prognostication and managing family expectations, among other roles.

Most ELVO patients will require varying degrees of inpatient rehabilitation, which is correlated with improved outcomes.119Patients discharged to a skilled nursing facility have a significantly lower probability of achieving a good neurological outcome than those discharged to an acute rehabilitation environment.120 Case management and social work teams are particularly helpful with this goal. Patients should be discharged with clear follow-up plans and with targets for risk factor modification. Finally, 90 day outcomes should be assessed for all ELVO patients post-thrombectomy by telephone or clinic visit to measure the effect of current therapies and to aid quality control measures.


  • All patients admitted with acute stroke should have an initial assessment by multidisciplinary rehabilitation professionals (physical, occupational, and speech therapy) as soon as possible after admission, preferably within the first 24–36 hours. (Class II, level of evidence B)

  • Due to impact on outcome, every effort should be made to establish aggressive rehabilitation placement for this patient population. (Class II, level of evidence B)

  • Ninety day outcome assessment after thrombectomy is a reasonable and appropriate standard follow-up. (Class II, level of evidence C)

Additional considerations

There are additional aspects of post-thrombectomy care that are currently either beyond the scope of this manuscript or areas of novel research. The ELVO patient with severe disability despite aggressive treatment may still have a poor outcome. These patients have a high prevalence of palliative care needs, and communication with families is paramount.121–124 This area requires further development and study. Similarly, utilization of neuroprotective therapies to augment outcome from thrombectomy and to promote neurorepair may be important in the future. Examples of compounds under study currently include MMP-9 inhibitors, free radical scavengers, calcium channel blockers, magnesium sulfate, and PSD95 inhibitors.125–135


Post-thrombectomy management of the ELVO patient is complex. Vital aspects of patient care that require monitoring and treatment include optimization of reperfusion, post-reperfusion hemorrhage, cerebral edema, access site complications, and rehabilitation efforts. Careful attention to these aspects is vital to outcome optimization.



  • Contributors JFF, as senior author, coordinated the reformatting, revising, and collaboration with the primary authors for review to the Journal of NeuroInterventional Surgery. RMS, MSH, PMM, RAM, GLP, JFF, SA, TA, BA, ASA, BWB, KRB, JEDA, CDG, DH, SWH, RPK, SKL, WJM, JM, CP, AP, PR, PS, and DF represent the combined members of the SNIS Executive Board and the Standards and Guidelines Committee (in addition to the other authors). These two bodies work together to evaluate published literature in the context of clinical practice, to determine the relevance and strength of that literature, and to provide expert opinion in clinical practice guidelines. Each of these authors reviewed the document, made revisions, and offered suggestions as their role as topic experts.

  • Disclaimer This literature review (‘Review’) is provided for informational and educational purposes only. Adherence to any recommendations included in this Review will not ensure successful treatment in every situation. Furthermore, the recommendations contained in this Review should not be interpreted as setting a standard of care, or be deemed inclusive of all proper methods of care nor exclusive of other methods of care reasonably directed to obtaining the same results. The ultimate judgment regarding the propriety of any specific therapy must be made by the physician and the patient considering all the circumstances presented by the individual patient, and the known variability and biological behavior of the medical condition. This Review and its conclusions and recommendations reflect the best available information at the time the Review was prepared. The results of future studies may require revisions to the recommendations in this Review to reflect new data. SNIS does not warrant the accuracy or completeness of the Review and assumes no responsibility for any injury or damage to persons or property arising out of or related to any use of this Review or for any errors or omissions.

  • Competing interests MC reports being a consultant for Medtronic, Penumbra, Stryker and Genentech. JEDA reports being a consultant for Medtronic, Penumbra, Sequent Medical and Accriva Diagnostics. JFF is an equity interest holder for Fawkes Biotechnology and a consultant for Stream Biomedical. DH reports being a consultant for Stryker Neurovascular. SWH reports being a consultant for Medina and Neuravi, as well as research contracts with Stryker Neurovascular, Siemens, MicroVention Terumo. Hirsch reports being a consultant for Medtronic. RK reports being a proctor and speaker for Medtronic. JM reports the following: consultant: Rebound Therapeutics, TSP, Cerebrotech, Lazarus Effect, Pulsar, Medina; investor: Blockade Medical, TSP, Lazarus effect, Medina; principal investigator (PI)/co-PI (CO-PI) for the following trials: THERAPY (PI), FEAT (PI), INVEST (Co-PI), COMPASS (Co-PI), LARGE (Co-PI), COAST (Co-PI), POSITIVE (Co-PI). Steering committee for the MAPS trial. CP reports the following: Consultant, Codman Neurovascular (serving on DSMB). GLP reports the following: Consultant, Sequent Medical (DSMB for Web-IT study). PR reports the following: Blockade Medical—Investor, Scientific Advisory Board and Stock Holder; Medtronic—Consultant/Honorarium; Nervive Medical—Scientific Advisory Board and Stock Holder; Perflow Medical—Scientific Advisory Board and Stock Holder; Stryker Neurovascular—Scientific Advisory Board. PS reports protctoring for Medtronic and is an investigator in STRATIS (Medtronic) and CARE (Penumbra) studies.

  • Provenance and peer review Commissioned; internally peer reviewed.

  • Correction notice Since this article was first published the initial A has been added to the author name Sameer Ansari. Dr Ansari’s affiliation has been updated also.