Blood Pressure Management Before, During, and After Endovascular Thrombectomy for Acute Ischemic Stroke

Abstract

There is an absence of specific evidence or guideline recommendations on blood pressure management for large vessel occlusion stroke patients. Until randomized data is available, the peri-procedural blood pressure management of patient undergoing endovascular thrombectomy can be viewed in two phases relative to the achievement of recanalization. In the hyper-acute phase, prior to recanalization, hypotension should be avoided to maintain adequate penumbral perfusion. The American Heart Association guidelines should be followed for the upper end of pre-thrombectomy blood pressure: ≤185/110 mm Hg, unless post-tPA administration when the goal is <180/105 mm Hg. After successful recanalization (TICI 2b-3) we recommend a target of a maximum SBP <160 mm Hg, while the persistently occluded patients (TICI <2b) may require more permissive goals up to <180/105. Future research should focus on generating randomized data on optimal blood pressure management both before and after endovascular thrombectomy, to optimize patient outcomes for these divergent clinical scenarios.

Introduction

Historically, the effect of blood pressure (BP) on stroke outcome has been shown to have a U-shaped curve, with extremes in BP associated with worse outcomes.¹ Provision of adequate cerebral perfusion in acute ischemic stroke (AIS) patients with large vessel occlusion (LVO) represents a critical aspect of patient care in both the peri- and intraprocedural settings as it relates to endovascular thrombectomy (EVT). Intuitively, reductions in systemic BP before revascularization may result in a cerebral perfusion decrement, compromised collateral flow, and consequent infarct progression.² This could be particularly important in penumbral territory where a pressure-passive system exists in which arteries and arterioles maximally dilate in response to tissue ischemia and no longer exhibit the capacity for pressure autoregulation. In this manner, cerebral blood flow could be directly dependent on systemic BP.³ However, hypertension could be damaging to the stroke and penumbra by promoting edema, cytotoxic mediators, and hemorrhagic conversion.^4–6

Despite the inherent complexity to BP management in the peri-EVT period, the current guideline treatment paradigm of these patients is identical to the management of non-occlusive AIS. The AHA/ASA guidelines suggest permissive hypertension, specifically recommending that “hypotension and hypovolemia be corrected.”⁷ Unless intravenous tissue plasminogen activator (tPA) is administered, hypertension up to a BP of ≤185/110 mm Hg can be tolerated prior to EVT.⁸ This 2019 AHA guideline (Table 1) is a revision of the prior 2018 guideline that allowed BP up to 220/120 mm Hg.⁷ Following tPA administration, the goal is <180/105 mm Hg (Table 1) with similar goals for the post-procedural period. These guidelines do not distinguish between stroke etiology, subtype, or severity. Ultimately, BP management surrounding EVT is deferred to the treating physician, and there is considerable variability in practice.⁹ The intent of this review article is to help guide BP management in the peri-EVT period. Ultimately, further research will be needed to better define specific BP goals in this patient population; taking into account the stage of treatment, patient-level variables, and post-procedural recanalization status.

Table 1.

American Heart Association guidelines for peri-thrombectomy blood pressure.⁸

Clinical Scenario	Blood Pressure Goal	Class of Recommendation	Level of Evidence
Pre-EVT, no tPA	≤185/110	IIa	B-NR
Pre-EVT, post-tPA administration	<180/105	I	B-R
Peri-EVT, no tPA	≤180/105	IIa	B-NR
Pre-EVT, post-tPA administration	<180/105	I	B-R
Post-EVT, no tPA and TICI 0–2a	≤180/105	IIa	B-NR
Post-EVT, post-tPA OR TICI 2b-3	<180/105	IIb	B-NR

^*Class of Recommendation: I=strong, IIa=moderate, IIb=weak; Level of Evidence: B-R=randomized, B-NR=nonrandomized

Preprocedural BP

A recent multicenter study showed that, for patients undergoing EVT, a ≥15% decrease or increase in their mean arterial pressure from pre-EVT baseline to post-procedural mean had the worst outcome in an adjusted analysis.¹⁰ These findings suggest that lowering BP pre-EVT may lead to infarct expansion and worsening ischemia, while, at the same time, an increase in BP may be directly harmful to the brain or a sign of worsening compensatory mechanisms such as cerebral collateral failure.^6,11–13 Interventional trials to alter BP prior to EVT or intravenous tPA are few, but there are ample data on blood pressure reduction in the hours and days after stroke onset using a variety of antihypertensive medications. The Prehospital Transdermal Glyceryl Trinitrate in Patients with Ultra-acute presumed stroke (RIGHT-2) failed to show benefit in the hyperacute window.^14(p2) The Controlling Hypertension and Hypotension Immediately Post Stroke (CHHIPS) and Evaluation of Acute Candesartan Cilexetil Therapy in Stroke Survivors (ACCESS) reported a mortality benefit, but no functional benefit of BP lowering during the acute phase of stroke.^4,15

BEST, a trial evaluating beta-blockers in AIS, showed increased mortality in patients randomized to early beta-blocker use.¹⁶ SCAST, a randomized study lowering BP in AIS patients using candesartan, suggested a higher risk of poor outcome in the treatment group.¹⁷ Accordingly, a Cochrane review of 26 trials concluded that there is “insufficient evidence that lowering BP during the acute phase of stroke improves functional outcome.”¹⁸ The ENCHANTED (Enhanced Control of Hypertension and Thrombolysis Stroke Study) trial applied intensive BP lowering (target systolic BP 130 to 140 mm Hg within 1 hour) versus guideline recommended BP lowering (target SBP < 180 mm Hg) to IV tPA eligible AIS patients. Functional status at 90 days did not differ between the groups.¹⁹ The cumulative impact of these studies is that there is no tangible benefit to dramatically reducing blood pressure immediately after stroke onset.

This is further complicated by the fact that BP tends to be spontaneously elevated in patients presenting with AIS for adaptive physiologic perfusion, and typically declines over time without intervention.^20,21 The initial hypertensive response presumably occurs in an effort to increase the collateral flow in order to salvage the penumbral region.⁶ Despite the neutral findings of the Head Positioning in Acute Stroke Trial (HeadPoST), which compared lying flat or being allowed to sit up in the days following stroke, some patients do improve with head-of-bed flattening or elevating BP, so-called induced hypertension.²² This suggests that the ischemic penumbra’s viability can be dependent on BP level.²³ It follows that inducing elevation of BP in the hyperacute phase could be beneficial to an ischemic brain dependent on collaterals. Not surprisingly, animal models have suggested the safety and efficacy of short-term induced hypertension, even if there is an increased risk of vasogenic edema with long-term use.²⁴ Drummond et al showed that phenylephrine induced hypertension can acutely improve local cerebral blood flow and reduce the areas of critical hypoperfusion (<15ml/100g/min) in rats with brief middle cerebral artery occlusion.²⁵ In a primate model of stroke, this improvement in local cerebral blood flow with induced hypertension resulted in restitution of neuronal function as assessed by evoked potentials and improvement of focal neurologic deficits.^26,27

Because LVO patients have a relatively larger amount of penumbral tissue compared to other stroke etiologies, they may particularly benefit from BP elevation prior to recanalization. However, the concept of induced hypertension has yet to be substantiated by adequate prospective randomized studies. In a small prospective study using phenylephrine to induce higher BP, Rordorf et al concluded that patients with “large extracranial or intracerebral vessel stenosis or occlusion” seem to have the most benefit from induced BP.²⁸ Hillis et al induced higher BP in a small number of patients with perfusion-diffusion mismatch and reported a “strong, statistically significant association between improved function and improved perfusion.”²⁹ In Hillis’s study, intravenous phenylephrine was titrated to achieve a 10–20% increase in MAP over 1–8 hours, and this target was maintained for at least 24 hours. Each patient in the intervention group showed some degree of improvement on the NIHSS and a cognitive battery after the treatment target was reached. On day 3 the NIHSS score had improved by 4.2±1.0 in the intervention group and 1.2±3.0 in the control group. Based on these and other studies,^30–32 induced hypertension may be a reasonable approach to the pre-thrombectomy LVO patient, especially if there is a prolonged period between medical care access and endovascular recanalization, such as in “hub and spoke” stroke systems of care.³³

That said, the studies cited above have sample sizes less than 50 patients and methodological weaknesses. The current AHA/ASA guidelines do not offer specific guidance on the practice of induced hypertension, noting instead that its “usefulness in patients with AIS is not well established.”⁸ Ultimately, in the absence of compelling data, the AHA recommends permissive hypertension (≤185/110 mm Hg), or <180/105 mm Hg if IV tPA has been administered, prior to EVT.⁸ The goal of ≤185/110 mm Hg is new in the 2019 guideline. The authors base this goal off criteria for the recent randomized clinical trials of EVT, which had an exclusion for BP >185/110 mm Hg.⁸ While this is a reasonable suggestion, there is not evidence for improved outcome with this goal. Future research may allow for more patient-specific BP interventions aimed at improving the survival of the ischemic penumbra.

Intraprocedural BP

Although anesthetic considerations in patients undergoing EVT for LVO are widely studied, there is a relative dearth in the literature or guidelines regarding intraprocedural BP management. A 2014 guideline by the Society of Neuroscience in Anesthesiology and Critical Care recommended that systolic BP be maintained between 140 and 180 mm Hg during EVT.³⁴ The American Heart Association recommends BP be ≤180/105 mm Hg during EVT (Table 1).⁸ These guidelines, although motivated by presumed mitigation of deleterious effects of extremes of intraprocedural BP, lack strong supporting evidence. Given the absence of research specific to intraprocedural BP, we can look at the literature addressing anesthesia in this context. Jumaa et al. were the first to note improved clinical outcomes and smaller final infarct volumes among patients who underwent EVT with conscious sedation as compared with general anesthesia.³⁵ Subsequently, observational studies have highlighted better outcomes in stroke patients undergoing conscious sedation during EVT.^36–39 Several explanations have been postulated to account for the difference in outcomes, including (1) time associated with intubation, (2) hypotension associated with anesthesia induction, and (3) vasodilation caused by use of anesthetic gases resulting in a steal phenomenon, whereby cerebral blood flow is shunted to healthy brain tissue from the penumbra. However, randomized trials have failed to show a difference in outcomes of patients treated with general anesthetic versus conscious sedation.^40,41 One key factor that putatively explains the better outcomes with general anesthesia in observational studies versus in clinical trials may be the prompt treatment of hypotensive episodes for anesthetized patients in clinical trials.⁴²

Recent publications have studied the impact of hypotension on patient outcomes after EVT, independent of anesthetic choice. By strictly analyzing their experience with patients who underwent EVT under general anesthesia, Lowhagen Henden et al. were able to control for the confounding effect of anesthesia type and noted that intraprocedural mean BP reductions of >40% from baseline were predictive of poor outcome.⁴³ Similarly, Whalin et al. studied patients who were treated via conscious sedation and showed that for every 10 mm Hg decrease in intra-procedural mean arterial BP below 100 mm Hg, there was 28% higher likelihood of poor outcome.⁴⁴ Most recently, Petersen et al. demonstrated that before vessel recanalization, larger intraprocedural BP reductions and sustained relative hypotension were both independently associated with worsened functional outcome, regardless of reperfusion grade.² For every 10 mm Hg reduction in the mean BP during EVT as compared with admission pressure, there was a 22% greater likelihood of unfavorable modified Rankin scale score (3–6) at 90 days. In addition, larger intraprocedural BP reductions were also predictive of infarct growth and final infarct volume: every relative 10 mm Hg reduction in the mean BP during EVT demonstrated a 4.1-ml increase in infarct volume. Conversely, a post-hoc analysis of the SIESTA trial did not find any association between drops in BP from baseline with functional outcomes, likely due to strictly protocolized BP management interprocedurally.⁴²

Thus, intraprocedural BP management during EVT may represent a critical factor in the care of AIS patients with LVO. Regardless of the manner of anesthesia provided during the case, careful monitoring of BP, avoidance of hypotension, and potentially, treatment of hypotensive episodes, appear to be associated with better patient outcomes. In an era of improved LVO patient throughput, as well as increasingly effective and rapid recanalization techniques, focus on this key aspect of patient care may represent an important step forward in further improving outcomes.

Postprocedural BP

Support for BP control following EVT comes from retrospective series.^45–47 Mistry et al. identified a significant correlation between maximum SBP and worse outcomes on 90 day modified Rankin Score and hemorrhagic complications following EVT in a three center study with over 200 patients undergoing EVT.⁴⁵ This correlation between maximum SBP and 90 day functional outcomes held when the authors performed a sub-group analysis of patients undergoing successful (TICI 2b/3) versus unsuccessful recanalization (TICI 0–2a). Similarly, Goyal et al. examined patients with LVOs following successful EVT.⁴⁶ Patients who died at 3 months from onset had higher maximum SBP levels compared to those who did not (184 ± 24 mm Hg vs 167 ± 21 mm Hg; p<0.001), whereas patients who were functionally independent at 3 months had a lower maximum SBP level (163 ± 20 mm Hg vs 179 ± 23 mm Hg; p<0.001). Interestingly, BP management was not associated with the development of symptomatic intracerebral hemorrhages (sICH), although the overall rates of sICH were relatively low (6.5%). Anadani, et al. also identified BP following EVT as a determinant in outcomes at three months.⁴⁷ In their analysis of patients undergoing EVT for LVO, a lower average SBP was associated with improved functional outcomes (mRS 0–2) at three months. Interestingly, maximum SBP following EVT was not associated with improved outcomes.

Conversely, the SIESTA investigators did not identify changes in post-procedural BP to be associated with neurological improvement or long-term functional outcome in a post-hoc analysis.⁴² However, there were some distinct differences between SIESTA and the previously discussed retrospective series. The SIESTA trial had a target SBP range of 140–160 mm Hg for EVT patients, which essentially targeted both hyper- and hypotension. Given the conflicting results of the existing data, BP management following EVT continues to be an area of considerable practice variation. A recent survey of StrokeNet sites identified significant heterogeneity of post-thrombectomy BP management.⁹ The majority (52%) of the 58 responding centers did not use a standard protocol preferring to set individualized targets. The recently published guidelines from the AHA recommend BP ≤180/105 mm Hg for 24 hours following EVT (Table 1), unless there the patient is post-tPA or has successful recanalization (TICI 2b-3) in which case the goal is <180/105.⁸ Based the retrospective data, some authors have advocated for goal SBP <180/105 mm Hg in patients with unsuccessful recanalization and SBP <140 or <160 mm Hg in patients with successful recanalization following EVT, which appears reasonable until more data is available.⁴⁸

BP variability post thrombectomy

In patients with acute ischemic stroke, increased BP variability (BPV), measured by coefficient of variation, successive variation, standard deviation, or more sophisticated methods, is associated with worse neurologic outcome.^49,50 The increase in BPV may be due to the impaired autoregulation that can accompany stroke,⁵¹ which could in turn harm the ischemic penumbra. Despite the growing number of articles on BPV after stroke, there is little data on the effect of BPV on outcome in patients receiving EVT.

One study included patients with acute ischemic stroke receiving EVT and found there was an association between increased BPV and worsening functional outcome, and that this association was strongest in patients without successful recanalization as opposed to those who achieved successful reperfusion.⁵² Another study showed that BPV was associated with early neurological deterioration and worse outcome which again was particularly the case in patients with incomplete recanalization.⁵³ It remains unclear whether in patients receiving EVT, BPV is the direct cause of poor outcome or whether increased BPV is an epiphenomenon of infarct growth or other physiologic stressors.

Discussion

Randomized controlled trial data are lacking to guide management of peri-EVT BP, but it has been common practice to allow permissive hypertension to recruit collaterals in the setting of vessel occlusion and then reduce BP following recanalization.^54–57 Eventually, guidelines were released advocating for SBP ≤180/105 mm Hg for the first 24 hours after treatment in patients undergoing EVT.⁷ Concurrently, evidence emerged that lowering BP, at least post-EVT could provide some benefit.^46,58,59 With the conceptual progress offered by these studies, it further became clear that not all post-thrombectomy patients should be treated equally, and survey data indicate that most practitioners decide treatment on a per-patient basis.⁶⁰ The most important factor affecting post-thrombectomy BP management is recanalization status. In patients with successful recanalization (TICI 2b-3), more favorable outcomes have been found among those with lower SBP.^56,61 In successfully recanalized patients, consensus is moving toward SBP goals such as <160 mm Hg; others advocate for even more aggressive measures, particularly for patients with potentially larger infarct cores as seen with the <140 mm Hg SBP goal in the protocol for the DAWN trial.^46,54,62

For patients in whom satisfactory recanalization cannot be obtained (TICI 2a or lower), less evidence is available to guide practice, but it seems reasonable to continue permissive hypertension.⁶⁰ Many advocate for a SBP goal ≤180/105 mm Hg in such patients.^46,61,63 Additionally, BP variability appears particularly detrimental in patients with no or incomplete recanalization.^52,64–66 Special consideration should also be given to patients in whom IV tPA has been administered. Guidelines for patients receiving IV tPA suggest BP goals of <180/105 mm Hg regardless of thrombectomy status.^7,67 In the NINDS trial that led to FDA approval of IV tPA, hemorrhage rates nearly doubled (12.4% vs. 6.4%) in patients with SBPs >180 mm Hg.⁶⁸ Application of tPA-specific data to patients with LVO seems viable in the context of them having received tPA.

In light of the absence of consensus of BP management following EVT, the need for nuance persists and each patient should be considered individually. For instance, it is important to account for a patient’s baseline BP; overzealous reduction of SBP in a chronically hypertensive patient could expand core infarct.^69,70 Most practitioners favor nicardipine as the optimal choice for intravenous anti-hypertensive in these patients.^56,61 Labetalol may be preferable if a patient is tachycardic, but it otherwise proves less easily titratable and often causes bradycardia.⁷¹ Nearly all high quality data guiding management of LVO patients only involves anterior circulation occlusions. Further data are needed for posterior circulation patients, which, for instance, accounted for 18% of thrombectomy cases in the cohort of Goyal et al.⁴⁶ Finally, patients with tandem disease or underlying stenosis are now more commonly undergoing angioplasty and stent placement of both extra- and intracranial disease in conjunction with EVT. These patients merit additional consideration for reperfusion syndrome in addition to BP control guidelines driven by EVT data.⁷²

Future Research Directions

Future research should focus on optimizing not only biomarker outcomes such as infarct volume but also functional outcomes of EVT-treated patients. Data-driven and actionable BP targets need to be generated in the pre-, peri- and post-EVT period. Large observational studies with sophisticated data-science methodologies need to be deployed in search of these targets, and need to address how patient-specific variables such as core size, quality of collaterals, site of occlusion, and recanalization status affect the target BP. These targets need to be tested in randomized trials with rigorous adjudication of safety and efficacy. Lastly, the peri-EVT BP management needs to be viewed as a continuum that follows the pathophysiology of cerebral autoregulation relative to the recanalization status of an LVO.

For the pre-EVT period, one research topic that has particular appeal is induced hypertension. Under normal circumstances, the cerebral vasculature has the intrinsic ability to maintain relatively constant cerebral blood flow despite changes in systemic blood pressure, a mechanism known as cerebral autoregulation.⁷³ This mechanism ensures that the cerebral blood flow matches the brain’s metabolic demands and protects it from hypo- or hyperperfusion. After large-vessel intracranial occlusion, the downstream perfusion pressure in that vessel will be reduced below the lower autoregulatory limit. Exhaustion of compensatory vasodilatory capacity and the loss of intrinsic autoregulatory function in the ischemic tissue leads to pressure passivity.^74,75 This means that changes in systemic blood pressure are directly transmitted to the brain and by increasing blood pressure, a corresponding increase in cerebral blood flow to the ischemic tissue occurs.

Animal experiments have shown that even a short interval of induced hypertension effectively augments cerebral blood flow in LVO via an abrupt opening of collapsed collateral vessels.⁷⁶ Worsening of cerebral edema and intracerebral hemorrhage were not seen in these experiments and have only been described with prolonged ischemia and extreme hypertension.⁷⁷ In humans, the direct relationship between blood pressure and cerebral blood flow can be observed in-vivo using monitoring of high-frequency of physiological data or neuroimaging.⁷⁸ Blood pressure augmentation with induced hypertension may prolong the viability of the ischemic penumbra via enhancement of collateral flow. However, despite compelling conceptual reasons and pilot data, we will need rigorous clinical trial evidence to support blood pressure augmentation as a potential neuroprotective strategy after acute stroke. Likewise, we will need clinical trial data to support BP treatment strategies post-EVT, when blood pressure reduction could be beneficial. The existing data supports studies that would use calcium channel blockers or nitric oxide donors as opposed to beta-blockers or ACE inhibitors, given the potentially negative effects of those medications in the acute stroke period.^14,16–18

Conclusions

In conclusion, there is an absence of solid evidence on BP management specific to LVO patients and further research is urgently needed. Until compelling data is available, the peri-procedural BP management of an LVO patient undergoing EVT can be viewed in two phases relative to the achievement of recanalization (Table 1). In the hyper-acute phase prior to recanalization, both in the pre-EVT and intraprocedural periods, hypotension should be avoided to maintain adequate penumbral perfusion. The utility of induction of hypertension in order to maintain collateral blood flow is not yet well established. Likewise, more research is needed on the potentially deleterious impact of increased BPV in AIS patients. Although the majority of prior observational studies have noted an association between higher post-recanalization BP and worse functional outcomes, no randomized data is available regarding the safety and efficacy of antihypertensive treatment following recanalization. Prior to EVT, current AHA guidelines should be followed (Table 1): ≤185/110 mm Hg, unless post-tPA administration when the goal is <180/105 mm Hg. After successful recanalization (TICI 2b-3), we believe that, in addition to the AHA goal of ≤185/105, a target of a maximum SBP <160 mm Hg appears reasonable, while the persistently occluded patients (TICI <2b) may require a more permissive goal of ≤180/105 mm Hg. Future research should focus on generating randomized data on optimal BP management both before and after EVT, to improve patient outcomes for these divergent clinical scenarios.