Article Text

Original research
Post-reperfusion hyperperfusion after endovascular stroke treatment: a prospective comparative study of TCD versus MRI
  1. Markus Kneihsl1,2,
  2. Nicole Hinteregger2,
  3. Oliver Nistl2,
  4. Hannes Deutschmann2,
  5. Susanna Horner1,
  6. Birgit Poltrum1,
  7. Simon Fandler-Höfler1,
  8. Isra Hatab1,
  9. Melanie Haidegger1,
  10. Daniela Pinter1,
  11. Alexander Pichler1,
  12. Karin Willeit3,
  13. Micheal Knoflach3,
  14. Christian Enzinger1,2,
  15. Thomas Gattringer1,2
  1. 1 Department of Neurology, Medical University of Graz, Graz, Austria
  2. 2 Division of Neuroradiology, Vascular and Interventional Radiology, Department of Radiology, Medical University of Graz, Graz, Austria
  3. 3 Department of Neurology, Medical University of Innsbruck, Innsbruck, Austria
  1. Correspondence to Professor Thomas Gattringer, Department of Neurology, Medical University of Graz, Graz, Steiermark, Austria; thomas.gattringer{at}medunigraz.at

Abstract

Background Increased middle cerebral artery (MCA) blood flow velocities on transcranial duplex sonography (TCD) were recently reported in individual patients after successful mechanical thrombectomy (MT) and were related to intracranial hemorrhage and poor outcome. However, the retrospective study design of prior studies precluded elucidation of the underlying pathomechanisms, and the relationship between TCD and brain parenchymal perfusion still remains to be determined.

Methods We prospectively investigated consecutive patients with stroke successfully recanalized by MT with TCD and MRI including contrast-enhanced perfusion sequences within 48 hours post-intervention. Increased MCA flow on TCD was defined as >30% mean blood flow velocity in the treated MCA compared with the contralateral MCA. MRI blood flow maps served to assess hyperperfusion rated by neuroradiologists blinded to TCD.

Results A total of 226 patients recanalized by MT underwent post-interventional TCD and 92 patients additionally had perfusion MRI. 85 patients (38%) had increased post-interventional MCA flow on TCD. Of these, 10 patients (12%) had an underlying focal stenosis. Increased TCD blood flow in the recanalized MCA was associated with larger infarct size, vasogenic edema, intracranial hemorrhage and poor 90-day outcome (all p≤0.005). In the subgroup for which both TCD and perfusion MRI were available, 29 patients (31%) had increased ipsilateral MCA flow velocities on TCD. Of these, 25 patients also showed parenchymal hyperperfusion on MRI (sensitivity 85%; specificity 62%). Hyperperfusion severity on MRI correlated with MCA flow velocities on TCD (rs=0.379, p<0.001).

Conclusions TCD is a reliable bedside tool to identify post-reperfusion hyperperfusion, correlates well with perfusion MRI, and indicates risk of reperfusion injury after MT.

  • MRI
  • Stroke
  • Ultrasound
  • Thrombectomy
  • Blood Flow

Data availability statement

Data are available upon reasonable request. Study data are available from the corresponding author on reasonable request.

Statistics from Altmetric.com

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.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Increased middle cerebral artery (MCA) blood flow velocities on transcranial duplex sonography (TCD) have been identified in individual patients after successful mechanical thrombectomy (MT) and were associated with intracranial hemorrhage (ICH) and poor outcome. However, the underlying mechanisms and the relationship with parenchymal brain perfusion still remained to be determined.

WHAT THIS STUDY ADDS

  • This prospective study identified post-reperfusion hyperperfusion as a frequent mechanism explaining elevated MCA blood flow velocities on TCD after MT. TCD correlated well with perfusion MRI and indicated an increased risk of post-interventional reperfusion injury (ICH, vasogenic brain edema).

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • TCD can serve as a reliable bedside tool to monitor cerebral hemodynamic changes after MT and has the potential to serve as an early prognostic tool in this setting.

Introduction

Endovascular stroke treatment with mechanical thrombectomy (MT) has become the standard of care for stroke due to large intracranial artery occlusion.1 2 Despite high recanalization rates of up to 90%, a relevant proportion of patients treated with MT still have an unfavorable clinical course and poor outcome.3 To individualize treatment in the vulnerable phase after MT, early predictors of post-interventional complications are of high clinical relevance. In this context, increased middle cerebral artery (MCA) blood flow velocities on bedside transcranial duplex sonography (TCD) were detected after successful MT in individual patients and associated with intracranial hemorrhage (ICH) and poor 90-day outcome.4–7 However, the retrospective design of these studies did not allow elucidation of the underlying pathomechanisms (ie, hyperperfusion vs localized vasospasm/vessel stenosis). Moreover, the relationship between TCD and parenchymal perfusion abnormalities as evidenced by cerebral MRI still remains to be determined. A recent single case report demonstrated the concordance between bedside TCD and perfusion MRI to identify hemodynamic changes/cerebral hyperperfusion after MT.8 However, systematic data on this important topic are still lacking.

We therefore prospectively explored changes in MCA blood flow on TCD after successful MT and their predictive value for post-interventional complications and outcome. In a further step, we aimed to correlate TCD findings with brain perfusion as assessed by MRI.

Methods

Study design and data collection

In the years 2018–2021 we prospectively identified all patients with ischemic stroke who had been successfully treated with MT for acute large vessel occlusion in the anterior cerebral circulation at our primary and tertiary care university hospital. Demographics, cerebrovascular risk factors, details of the MT procedure as well as post-interventional complications and outcome were assessed and documented in an electronic database. MT was conducted by experienced interventional radiologists via stent retrievers or clot aspiration systems. The Thrombolysis in Cerebral Infarction (TICI) score was used to grade recanalization status after MT on DSA at the end of the intervention.9 Successful recanalization was defined as TICI grades 2b–3. After MT, all patients were treated at our stroke unit or, if still intubated and ventilated, at our neurointensive care unit. Target blood pressure levels were at least below 180/105 mmHg according to international guideline recommendations.10

To exclude a major influence on TCD flow velocities, patients with hemodynamically significant carotid artery stenosis/MCA branch re-occlusion were not included in the final study cohort (figure 1).

Figure 1

Flow diagram of included study participants. MCA, middle cerebral artery; TCD, transcranial duplex sonography; TICI, Thrombolysis in Cerebral Infarction.

Functional neurological outcome was rated according to the modified Rankin Scale (mRS) at stroke unit discharge and at 90 days after stroke during a visit to our stroke outpatient department. If a physical consultation was not possible, a telephone interview was performed instead.

Transcranial duplex sonography

All patients underwent TCD (devices: GE Healthcare Vivid E9 or Vivid I; handheld transducers 3.3–3.6 MHz; Chalfont St Giles, UK) within 48 hours after MT by trained technicians blinded to clinical data and had a TCD follow-up 90 days after MT.

According to our study protocol, MCA mean blood flow (MBF) velocities were analyzed in the treated (ipsilateral) and contralateral MCA. To account for interindividual differences (eg, blood pressure) and in agreement with recently published studies,8 MCA MBF velocity index (ipsilateral/contralateral MCA MBF) was calculated. Similar depths of insonation were used to ensure optimal comparison of bilateral MCA blood flow velocities. To differentiate focal blood flow velocity alterations (ie, vasospasm/vessel stenosis) from hemispheric hemodynamic changes, blood flow velocities were assessed in all visible parts of the MCA M1/M2 segments in all patients.

Angle correction was applied if a straight MCA segment of at least 1.5 cm was visualized.11 MCA pulsatility indices were calculated from both MCAs according to the Gosling method:

Embedded Image .12

In case of an absent temporal bone window, ultrasound contrast agents (Sonovue, Bracco International, Amsterdam, The Netherlands) were used to determine intracranial blood flow.

Brain imaging

All included patients underwent post-interventional brain imaging (CT or MRI) at 24 hours after MT or at any time in case of clinical deterioration. MRI was performed on a 3.0 T Magnetom Prisma (Siemens Healthineers, Munich, Germany). Final infarct size and intracranial bleeding complications (according to the Heidelberg Bleeding Classification) were rated on the last available neuroimaging modality (mostly MRI, <7 days after admission).13 Symptomatic intracranial hemorrhage (ICH) was defined as clinical symptom deterioration according to a National Institutes of Health Stroke Scale (NIHSS) score increase of >2 points in one category or >4 points in total.14 To identify residual stenosis or vasospasm after MT, DSA scans at the end of the intervention were reviewed as pre-specified for this prospective study. Moreover, all patients with sonographic signs of focal stenosis after MT underwent intracranial MR angiography.

If MRI was performed within 48 hours after MT, vasogenic brain edema was rated as a surrogate marker of blood–brain barrier (BBB) disruption and contrast-enhanced MR perfusion imaging was added to the standard protocol to evaluate perfusion abnormalities.

Perfusion imaging studies included cerebral blood flow and volume as well as mean transit time and time to peak maps and were interpreted visually. In concordance with recent studies, the brain perfusion pattern on the side of MT was compared with the visually corresponding region on the contralateral hemisphere.15 To quantify the extent of hyperperfusion, the affected areas were rated according to the Alberta Stroke Program Early CT Score (ASPECTS) regions resulting in an MRI-perfusion score of 0–10.16

All MRI scans were rated by experienced neuroradiologists (NH, ON) who were unaware of the TCD data. Twenty randomly selected cases were reviewed by both neurologists regarding inter-rater reliability of perfusion ASPECTS rating.

Statistical analysis

Statistical analysis was performed using the IBM SPSS Statistics, version 27. According to established ultrasound scores (ie, Thrombolysis in Brain Ischemia Score, TIBI), increased MCA flow velocities on TCD were defined as >30% MBF velocity increase in the treated compared with the contralateral MCA (MCA MBF velocity index >1.3).17 18 Consequently, patients with increased MCA MBF velocity indices were compared with those with MCA MBF velocity indices ≤1.3. For subgroup analyses, a χ2 test was used to compare categorical variables. Depending on Gaussian distribution, Student’s t-tests or Mann–Whitney U tests were performed for continuous variables. Multivariable regression analysis was performed to identify variables that were independently associated with an increased MCA MBF velocity index. The model included age, sex and stroke severity at admission as well as factors that were associated with a MCA MBF velocity index >1.3 in univariable analysis (final infarct size, post-interventional ICH, 90-day outcome). In a further step, a confusion matrix for sensitivity and specificity (with 95% CI) of increased MCA blood flow on TCD for the diagnosis of parenchymal hyperperfusion on MRI was calculated.

Results

All 226 included patients (mean age 71.7±12.9 years, 58% female) had successful recanalization and underwent TCD at a mean time interval of 8.7±10.2 hours after MT. Increased MCA MBF velocity indices >1.3 were identified in 85 patients (38%). Of these, 10 patients had an underlying focal MCA stenosis (n=8) or a vasospasm (n=2), additionally confirmed by MRA.

Compared with patients with MCA MBF velocity indices ≤1.3 (n=141), the residual 75 patients with increased MCA blood flow velocities had larger post-interventional infarcts (median ASPECTS 4.5 vs 6.5, p<0.001), a higher percentage of infarcts affecting ≥2/3 of the MCA territory (39% vs 10%, p<0.001), and more often had vasogenic brain edema on MRI (48% vs 19%, p=0.005). Moreover, an increased MCA MBF velocity index was associated with any post-interventional ICH (52% vs 34%, p=0.008), more severe post-interventional ICH according to Heidelberg Bleeding Classification types hemorrhagic transformation 2 (HT2)–parenchymal hemorrhage 2 (PH2) (43% vs 20%, p<0.001) and a worse 90-day outcome (mRS 0–2: 29% vs 50%, p=0.004). MCA pulsatility index was lower in patients with increased MCA blood flow velocities (0.95±0.60 vs 1.01±0.19, p<0.001) while demographics, NIHSS at presentation, pretreatment ASPECTS and intravenous thrombolysis rate did not differ between the two subgroups (table 1).

Table 1

Demographics, medical history, clinical/imaging parameters and outcome in patients with ischemic stroke successfully recanalized by mechanical thrombectomy

After adjustment for age, sex and initial NIHSS, increased MCA MBF velocity indices >1.3 remained predictive for larger infarct size (≥2/3 of MCA territory, OR 4.3, 95% CI 1.9 to 9.4, p<0.001) and post-interventional ICH (HT2–PH2, OR 2.7, 95% CI 1.4 to 5.3, p=0.004), while the association with a worse 90-day functional outcome was not statistically significant in multivariable analysis (OR 1.5, 95% CI 0.8 to 3.1, p=0.212).

Correlation between TCD and perfusion MRI

Ninety-two of the study participants (41%) additionally underwent contrast-enhanced perfusion MRI within 48 hours after MT. In this subgroup, 29 patients (31%) had an increased MCA MBF velocity index >1.3 on post-interventional TCD, 25 of whom had parenchymal hyperperfusion on MRI (sensitivity 86%, 95% CI 0.64 to 0.95; specificity 62%, 95% CI 0.49 to 0.74; diagnostic accuracy 76%; figure 2). Hyperperfusion severity on MRI (median hyperperfusion ASPECTS 7, range 2–10) correlated with absolute MCA flow velocities on TCD (Spearman rank correlation (rs)=0.379, p<0.001). Inter-rater agreement of perfusion ASPECTS rating was 90% (kappa 0.8). Within the analyzed post-interventional time period of 48 hours, we could not identify any significant correlations between diagnostic time intervals from recanalization to MRI/TCD and hyperperfusion (either on MRI or on TCD; p>0.1). No adverse events associated with MRI/TCD were observed.

Figure 2

Representative images of a patient with acute occlusion of the right middle cerebral artery (MCA) M1 segment (A), arrow). The MCA was successfully recanalized 3 hours after symptom onset with mechanical thrombectomy (Thrombolysis in Cerebral Infarction grade 3 (B). Post-interventional transcranial Duplex sonography (TCD) shows twice increased MCA mean blood flow velocities in the whole recanalized M1 segment compared with the contralateral MCA, indicating post-reperfusion hyperperfusion (C). Brain MRI 24 hours after thrombectomy shows the infarct together with vasogenic edema in the corresponding MCA territory on fluid attenuated inversion recovery (FLAIR) (D) and apparent diffusion coefficient sequences (ADC, (E), arrows; vasogenic edema evidenced by high ADC ie, hyperintensity (upper arrow) and ischemia evidenced by lowered ADC, ie, hypointensity (lower arrow), respectively). Related ICH is shown on susceptibility-weighted imaging (parenchymal hemorrhage 1, (F), arrow). Perfusion-weighted cerebral blood flow map confirms post-interventional hyperperfusion in the right MCA territory (G), arrows) while time-of-flight magnetic resonance angiography excludes intracranial vessel stenosis (H).

Of note, increased MCA blood flow velocities on TCD would have identified severe hyperperfusion on MRI (hyperperfusion ASPECTS <6) in 12 of 15 cases (sensitivity 80%, specificity 78%). Severe parenchymal hyperperfusion was unlikely in patients with normal TCD blood flow velocities in the treated MCA (3 of 60, negative predictive value 95%).

Compared with patients with normal perfusion MRI (n=45), cerebral hyperperfusion on MRI (n=47) was associated with large final infarct size (≥2/3 of the MCA territory, 23% vs 2%, p=0.003), vasogenic brain edema (40% vs 16%, p=0.011), poor 90-day outcome (mRS 0–2, 47% vs 71%, p=0.021) and there was a trend for an increased rate of any post-interventional ICH (60% vs 47%, p=0.057) (table 2).

Table 2

Demographics, medical history, clinical/imaging parameters and outcome according to post-interventional MRI perfusion status

Discussion

In this prospective study on cerebral hemodynamics after endovascular stroke treatment, we demonstrate good diagnostic accuracy between post-interventional bedside TCD and perfusion MRI to identify post-reperfusion hyperperfusion and consequently patients at risk for post-interventional complications and poor outcome.

Increased MCA blood flow velocities on TCD have been associated with ICH and poor outcome after MT in recent studies, but the underlying mechanisms and correlation with parenchymal perfusion status still remain to be determined.4–7 19 20 Although some authors assumed hyperperfusion to be the underlying mechanism,6 7 others hypothesized that a residual focal stenosis, luminal narrowing based on intimal hyperplasia or a transient vasospasm could have caused increased post-interventional MCA blood flow.4–7 19–21 However, those studies were limited by absent post-interventional angiographic data and did not investigate parenchymal brain perfusion status.4–7 19 20

Our study was specifically designed to clarify this ambiguity, as we had DSA data at the end of the intervention and MRI-based angiography mostly within 48 hours after MT available. Moreover, MCA blood flow velocities were measured in all accessible parts of the MCA M1/M2 segments to differentiate focal from generalized blood flow velocity increase. As only ~10% of patients who underwent MT with increased MCA blood flow velocity had a local stenosis/blood flow increase (additionally confirmed by MRA), we can exclude a major contribution of residual/de novo stenosis or transient vasospasm of the MCA in our prospective study cohort. Moreover, MRI-based parenchymal hyperperfusion in the treated hemisphere was identified in nearly 90% of patients with increased post-interventional MCA blood flow velocity on TCD. This underlines the presumed impairment of cerebrovascular autoregulation leading to post-thrombectomy hyperperfusion in a relevant proportion of patients who undergo MT. Endothelial damage during ischemia and formation of reactive oxygen species leading to arteriolar vasodilation and disruption of the BBB after reperfusion are the most likely mechanisms behind such blood flow alterations.22 23 In this context, we could detect vasogenic brain edema, a surrogate marker of BBB disruption and reperfusion injury, in almost every second patient with signs of hyperperfusion on TCD and MRI. Moreover, observed reduced MCA pulsatility indices in patients with hyperperfusion further support a reduced peripheral arterial resistance based on vasodilation in such patients.

The good correlation between TCD blood flow velocities and parenchymal hyperperfusion qualifies ultrasound as a bedside screening tool for hemodynamic changes after MT. Of note, severe parenchymal hyperperfusion was unlikely in patients with normal TCD blood flow in the treated MCA (negative predictive value 95%).

ICH and a poor outcome might be a consequence of post-thrombectomy hyperperfusion.4–7 However, hyperperfusion-associated hemorrhage mostly remains small and we could also identify only three patients (1.4%) who clinically deteriorated due to a parenchymal hemorrhage in the phase after MT. Although this challenges its direct effect on patient outcome, even small ICH can prolong or prohibit the use of anticoagulants/antithrombotics which would be necessary for early prevention of recurrent ischemic stroke.

Moreover, further pathomechanisms might contribute to the worse 90-day outcome. While previous TCD studies did not report on post-interventional infarct size,4–7 we here present an independent association of increased MCA blood flow and large infarcts within 48 hours after MT in multivariable analysis. Post-reperfusion hyperperfusion might therefore not only indicate patients at risk for hemorrhagic transformation of brain infarction but could also be predictive for large post-interventional infarct size including edema formation. This notion is in line with a recent small MRI perfusion-based study that found an association between post-interventional cerebral hyperperfusion and larger infarct volume at 24 hours after the MT procedure.15

Although the understanding of the underlying cause-effect relationship requires further research, our results support the prognostic importance of cerebral hemodynamics in the early phase after stroke MT.

As this study focused on the pathomechanisms behind increased MCA blood flow after MT, we excluded hemodynamically significant vessel stenosis or occlusions of intra- and extracranial vessels to avoid a major influence on TCD data. This study was therefore not designed to investigate the prognostic value of post-interventional hypoperfusion or no-reflow phenomenon.24 25

A further limitation was that we had no consistent short-term TCD follow-up examinations available. Although we showed in accordance with a recent TCD study that MCA blood flow normalizes within 90 days after the intervention,6 further research is needed to characterize and quantify the course of hemodynamic changes in the first days after MT. Larger sample sizes will also be needed to sufficiently report on the association between hyperperfusion and post-interventional ICH as the statistical power of this study is limited by the low number of symptomatic bleeding complications after MT. Finally, we had no quantitative perfusion measurements available but instead relied on visual rating of hyperperfusion on brain MRI, and pragmatically scored its extent based on ASPECTS regions. While slight abnormalities on perfusion maps might not be detected with qualitative visual rating, our approach seems feasible in routine clinical practice also supported by a high inter-rater agreement of 90%.

Conclusions

TCD is a reliable tool to identify post-reperfusion hyperperfusion and correlates with perfusion MRI. Moreover, abnormal hemodynamics after MT might indicate patients at risk for reperfusion injury (ICH, vasogenic brain edema) and larger infarct size as well as poor functional outcome.

Supplemental material

Data availability statement

Data are available upon reasonable request. Study data are available from the corresponding author on reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by the Ethics Committee of the Medical University of Graz, Austria (EC-Number: 30-254 ex 17/18). Participants gave informed consent to participate in the study before taking part.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Contributors MK: study design, acquisition and interpretation of data, manuscript preparation. NH: acquisition of data. OL: acquisition of data. HD: acquisition of data, critical revision of the manuscript content. SH: critical revision of the manuscript content. BP: acquisition of data. SF-H: acquisition of data, critical revision of the manuscript content. IH: acquisition of data. MH: acquisition of data. DP: acquisition of data. AP: acquisition of data. CE: study design, critical revision of the manuscript content. KW: critical revision of the manuscript content. MiK: critical revision of the manuscript content. TG: study design, acquisition and interpretation of data, manuscript preparation, critical revision of the manuscript. All authors have read and approved the final manuscript and agreed to be accountable for all aspects of the work. MK is responsible for the overall content as guarantor.

  • Funding This study was funded by Austrian Science Fund (FWF).

  • Competing interests None declared.

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

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.