Endovascular clot retrieval for early ischaemic stroke

This highly promising strategy for acute stroke management is sufficiently interesting from the exam standpoint that it should merit its own page, particularly as it had recently become the topic of Question 4 from the first paper of 2020. There, trainees were expected to briefly describe its "indications, post-procedure management and procedural complications". Though this question did not ask for a critical appraisal of the practice and the evidence behind it, that stylised approach to discussing it was felt to have some merit, as it in the event that such a question ever comes up its answer will already be available.

Systemic thrombolysis, which is still the standard of care in those Australian institutions which happen to be further from the beach,  is sufficiently interesting to be discussed on its own in another chapter. Supportive management for stroke has also been discussed elsewhere. The topic here will be mechanical endovascular embolectomy.

In summary:

In terms of published evidence, there are several excellent articles which can be recommended to the time-poor trainees unwilling to wade through hundreds of pages of stroke literature. An excellent ICU-centric paper by Ma, Moynihan and Donaldson (2019) covers the post procedure management very well. Patient selection and technique is well described by Silva et al (2020), and the AAN has decided not to paywall it for now. Evidence regarding the efficacy of this method is analysed in a systematic review by Rodrigues et al (2016). These sources and others have been blended together into something that might look like a CICM Part II pass-level answer, if one does not look at it too closely. 

Early history of mechanical embolectomy for stroke

The initial experience was very promising, but only applicable to a limited extent. The investigator's interest in endovascular stroke management was considerable. For instance, the MERCI trial investigators enrolled all-comers to receive embolectomy (within 8 hours!) by use of the Merci Retriever device. As a marketing exercise for this device, the trial was a success; in spite of a 7.1% serious complication rate the technique was viewed as "safe" in the conclusion. In fact there are numerous such devices, and the campaign to sell each one has yielded some equally promising propaganda material. Unfortunately, like in the MERCI trial, the outcomes of all these studies compared poorly with historical controls.

In 2009, the American Journal of Neuroradiology has published a summary of their opinions and experiences regarding endovascular stroke management.

They concluded:

"despite higher rates of recanalization, the mechanical thrombectomy studies have demonstrated substantially lower rates of good outcomes compared with IV and/or intra-arterial thrombolytic trials."

This was sad. There is something heroic and noble-sounding about accessing the cerebral vessels to suck out the clot and save the brain. In spite of this rather negative reception, as well as consistently depressing outcome data, enthusiasts continued to experiment, and the technology matured. Rather than attempting to dress it up as a viable alternative to thrombolysis, the investigators have recently focused on identifying a patient group which would derive the greatest benefit from the technique.

Rationale for endovascular embolectomy in stroke

The discussion of "rationale" for getting rid of the clot in a disease state where an errant clot is doing all the harm seems somewhat indirect. Surely, the rationale is obvious. The real question is, what is the rationale for performing this interventional neuroradiology procedure, as opposed to the systemic thrombolysis?

In summary:

• Management of ischaemic and embolic stroke is characterised by a time-pressured need to revascularise the occluded cerebral vessel.
• The "normal" approaches include conservative management with BP control / dual antiplatelet agents,  and active management with systemic intravenous thrombolysis.
• The former carries a high risk of poor outcome and closely resembles doing nothing, whereas the latter may achieve revascularisation at the cost of significant bleeding complications.
• A targeted endovascular approach may have the effect of revascularising the culprit vessel without systemic bleeding complications, and with much greater effect than conservative therapy.
• Complaints regarding the expense of maintaining the equipment and expertise to perform this boutique inner-city intervention should be countered with the expense of maintaining the equipment and expertise required to care for a large population of highly dependent stroke patients with unresolved neurodeficits. In other words, one might suggest that the cost of the intervention is much lower than the long-term cost of caring for a disabled survivor.

Indications for endovascular clot retrieval in stroke 

For local trainees preparing for their exam in the advanced medical thought incubator that is NSW Health, the Agency for Clinical Innovation had published an excellent document regarding patient eligibility for ECR. Their indications for referral include:

• Ischaemic stroke with confirmed large vessel occlusion
• Good level of premorbid independence
• Substantial neurological deficit, NIHSS ≥ 5
• Timeframe criteria:
  • Less than six hours: "broad clinical and imaging criteria"
  • 6-24 hour window: "significant volume of salvageable tissue"
• The criteria are flexible:
  • If your deficit fluctuates across the NIHSS=5 threshold, you should still be considered for this
  • If your deficit initially met the criteria but improved while you were waiting for the radiologist to get out of bed, you should still have the procedure
  • "Good level of premorbid independence" is obviously flexible, to make sure your service can still offer this procedure to privately funded 90-year-old amputees with dementia.

This NSW agency had got their recommendations from the Australian Stroke Foundation clinical guidelines,  for which the most recent available version at the time of writing is 2019. These were influenced by the contemporary AHA guidelines (Powers et al, 2019) and the European guidelines (ESO / ESMINT, 2019).

Volume of salvageable tissue and the intervention window

The most interesting element of the indications for endovascular clot retrieval is the 6-24 hour window. The 6 hour window is pretty uncontroversial, and closely resembles the four hour window you have come to know and love from the IV t-PA stroke management protocol. The 6-24 hour window where some "volume of salvageable tissue" can still be rescued is a completely different animal. In support of this timeframe, the Stroke Foundation refers to Nogueira et al (2018) and Albers et al (2018), both of which are soiled with the grime of heavy industry involvement.

Nogueira et al (the DAWN trial) looked at a group of ICA or proximal MCA strokes who had a major difference between the severity of their clinical deficit and the radiological appearance; specifically, the clinical features of the stroke were very severe but the CT or MRI volume of the stroke was comparatively small, ~ 10ml. These patients, the investigators argued, have a large amount of threatened brain tissue (hence symptoms) which has not yet completely infarcted (hence radiology), and therefore these are the people who will benefit most from a cath. And if one were going to use a cath anyway, why not choose this lovely Trevo device from my very good friends at Stryker? Literally everybody among the investigators was on the payroll of a manufacturer who stood to gain from a positive result. The list of affiliation disclosures spans across most of the last page of this study.

Similarly, Albers et al (the DEFUSE-3 trial) was a study of large vessel strokes which happened to have occurred within mere minutes of a fully stocked highly advanced neurointerventional radiology lab with proprietary sponsor software installed into its imaging system. This trial was stopped prematurely when the DAWN trial was published, and data were analysed with a much smaller group of patients, mainly because equipoise was lost. 

In short, the data collected by these trials was highly promising. Of their carefully selected patient population, with an overbearing abundance of interventional expertise and equipment available, functional independence at 90 days  was achieved by 49% of embolectomised patient vs 13% of the conventional therapy group in DAWN, and by 45% vs 17% in DEFUSE-3. Though there are obviously serious concerns of impartiality raised by the taint of sponsorship, one cannot be dismissive of the methodology or the obviously promising results. Even bloggers who have no vested interest in protecting corporate whoring have spoken out in favour of these studies, and the changes in AHA guidelines which they have produced. The promise of rescuing at least a couple of people from perpetual hemiplegia is too good to pass by.

Post-procedure management in the ICU

• Airway:  these people, often periprocedurally uncooperative or comatose, will often arrive intubated. Beyond the crude need to defend their airway against pooled secretions, there should be no reason to keep them that way. An association of poorer outcomes with prolonged intubation is mentioned by Ma et al as likely reflective of stroke severity rather than being any fault of the overcautious intensivist.
• Ventilation:  It would be disingenuous to make bland recommendations here (eg. normoxia and normocapnia), as most people would naturally arrive at the conclusion that stroke patients need normal levels of oxygen ad carbon dioxide. There is no specific need for any bizarre gas exchange targets or ventilator strategies.
• Blood pressure control is very important. For those whose stroke has been incompletely corrected (i.e. the vessel is still tight or partially thrombosed) the target for blood pressure should probably be more generous. In fact the Society of Neurointerventional Surgery recommend a MAP about 10-20% higher than the baseline, so long it does not max out higher than an SBP of 185 mmHg. For those in whom the clot has been successfully sucked out, the BP targets are less clear; but common sense recommends that their cerebral perfusion pressure should be made as close as possible to their pre-stroke baseline. The AHA authors, in the absence of anything sensible from the evidence, opine that "it seems rational to set a systolic upper limit of 140 to 160 mmHg post-complete or near-complete reperfusion". 
• Sedation has no specific neuroprotective goals and might even be counterproductive if one's intention is to serially reassess the patient. One can make a strong argument for ceasing all sedation. Analgesic requirements should be minimal, unless the interventional radiologist did something unconventional during the groin puncture.
• Electrolyte management focuses on prevention of AF, and is as basic as one can get. No specific electrolyte related recommendations can be made.
• Fluid therapy  focuses on maintaining a neutral balance, and is also not something which has been investigated well enough to make recommendations about.
• Nutritional support is probably not an essential part of management in the first 24 hours, as the patient is unlikely to be coming off a nutritionally depleted state (i.e. they were often eating and drinking just fine right up until the stroke). Moreover, hyperglycaemia is known to be a negative outcome-modifying factor. In short, to commence NG feeding in these people immediately upon ICU admission is not essential. Of greater importance is the speech and swallowing assessment.
• Anticoagulation and antiplatelets following endovascular clot retrieval (which often follows systemic thrombolysis) is logically something to be deferred. If thrombolysis has been given, one might conclude that the blood is as thin as it's going to get, in which case no additional agents are needed until the next day at least. If the patient has not had any thrombolysis, the situation is a little more foggy. Van de Graaf et al (2018) performed a highly unsatisfying systematic review of the published data and concluded that nobody could reach any conclusion on the basis of what they had to work with. Overall, everybody seems more confident to poison the platelets when they have a nice normal follow-up CT to look at.  In short:
  • Antiplatelets for all, 24 hours after the thrombolysis (when the repeat CT has excluded haemorrhage).
  • Antiplatelets for some (those who have had stents) within the first 24 hrs 
• DVT prophylaxis can be mechanical in the short term for patients receiving thrombolysis, and chemicals can wait until after the first CT for those who have not. Of the published guidelines, only the Germans (Torbey et al, 2015) have the balls to recommend low molecular weight heparin to start immediately, and they did not specifically target the post-embolectomy population.
• Anticoagulation for patients with AF-related stroke can probably wait up to 14 days, as the repeat stroke risk in this group is low in the first 15 days. This suggestion is based on very indirect evidence (stroke risk in patients with mechanical valves whose warfarin has been withheld).
• Management of fever seems important, as hyperthermia seems to have an adverse effect on stroke outcomes. Most guideline writers stop short of actually recommending active cooling; paracetamol seems to be where they draw the line.
• Early discharge from the ICU is probably beneficial. These patients are known to benefit from the increased access to allied health and specialist nursing staff which is afforded by a dedicated stroke-focused environment. You really keep them in the ICU mainly to detect and manage early complications from thrombolysis and radiologically guided intracerebral vessel perforation. The risk of these complications is probably maximal at the time of admission, and peters off by the next day, so that if one is able to extubate them, one is often able to discharge them within 24 hours.

Disadvantages of endovascular thrombectomy for stroke

• Previously it was mentioned that this procedure should be safer because it may decrease the need for systemic thrombolysis, but in fact most bodies (and certainly the Australian Stroke Foundation) strongly recommend that you give systemic thrombolysis anyway, while waiting for the endovascular procedure.
• The need for specialist imaging and interventional services sets up an unjust class system of stroke referrals, with inner urban communities significantly advantaged over rural and regional centres (particularly across the vast diffusely populated Martian landscape of Australia).

Post-procedure complications

Bleeding: Hao et al (2017) looked through all the major publications on the outcomes following endovascular clot retrieval and reported that among the 1,499 reported patient cases,  the risk of intracranial haemorrhage was substantially higher in the endovascular group. That's right, medical management actually carried a much lower risk of intracranial haemorrhage, only 19% - whereas with the endovascular approach the risk was 35%. However, this is only surprising if one assumes that "medical management" is systemic thrombolysis. In actual fact, Hao et al included trials which grouped "best medical therapy" together with intravenous t-PA, which probably dropped the risk of ICH somewhat. Moreover, even with this caveat, the extra intracranial blood in the interventional groups appears to have gone largely unnoticed by the patients and their families: when only "symptomatic"  stroke was included in the analysis, the numbers were much lower, and very similar (5.6% vs 5.2%).  

Cerebral oedema can develop, with an ensuing rise in the intracranial pressure and adverse effects of midline shift. Malignant MCA-like management, including decompressive craniectomy, is then floated as an idea. One might counter this enthusiasm by pointing out that, if the brain has swollen quite as bad as this, then one might make the argument that the volume of salvageable tissue was not rescued, and the outcome for this patient will be more like the outcome for untreated stroke (i.e. bad no matter what you do). 

Vascular access site complications are seen in a substantial proportion of these patients as they are often full of antiplatelet agents at the time of interventional access sheath removal.