Article Text
PDF
Abstract
Background Transcranial Doppler (TCD) is a non-invasive, bedside tool that allows for real-time monitoring of the patient’s hemodynamic status following mechanical thrombectomy (MT). This systematic review and meta-analysis aims to evaluate the predictive value of TCD parameters following successful MT (Thrombolysis in Cerebral Infarction 2b–3).
Methods In July 2024, we searched PubMed, Embase, and Scopus, to identify observational studies in which TCD parameters were measured within 48 hours of MT. Using random-effects models, we compared four TCD parameters (mean flow velocity (MFV), MFV index, pulsatility index (PI), and peak systolic velocity (PSV) among groups with vs without hemorrhagic transformation (HT) and favorable vs poor functional recovery (modified Rankin Scale 0–2 vs 3–6).
Results Eleven studies comprising 1432 patients (59% male; mean age range: 63–73 years) were included. The MFV and MFV index were higher in patients with HT (Hedges' g=0.42 and 0.54, P=0.015 and 0.005, respectively). Patients with MFV index ≥1.3 showed a higher risk of all HT (RR 1.97; 95% confidence interval (CI) 1.28 to 3.03, P=0.002), symptomatic HT (RR 4.68; 95% CI 1.49 to 14.65, P=0.008), and poor functional status at 90 days (RR 1.65; 95% CI 1.27 to 2.14, P=0.029), respectively. There was no difference in mean PSV (P=0.1) and PI (P=0.3) among groups with and without HT.
Conclusion This study underscores the prognostic value of the MFV index in predicting HT, symptomatic HT, and poor functional recovery after successful MT in the anterior circulation. Large-scale, multi-center studies are necessary to confirm these findings and to validate the MFV index as a reliable predictor for improving post-thrombectomy care.
- Thrombectomy
- Ultrasound
- Hemorrhage
Data availability statement
Data are available upon reasonable request.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Transcranial Doppler (TCD) is a non-invasive bedside and safe modality to assess the patient’s hemodynamic status in the affected vessels after acute ischemic stroke.
WHAT THIS STUDY ADDS
Certain TCD parameters can predict symptomatic hemorrhagic transformation and poor outcomes following successful mechanical thrombectomy.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The mean flow velocity (MFV) index could be used as a tool for risk stratification in clinical settings. Patients with a higher MFV index could be monitored more closely for potential complications.
Introduction
Mechanical thrombectomy (MT) has revolutionized stroke care in recent decades and is currently the standard of care for eligible patients with acute ischemic stroke due to large vessel occlusions (AIS-LVO).1 Despite its efficacy, 20–40% of patients experience hemorrhagic transformation (HT)2 and up to 50% of patients do not achieve favorable functional outcomes, even after successful recanalization of the occluded artery.3 There is increasing evidence that post-MT favorable outcome depends not only on prompt reperfusion therapy but also on post-MT hemodynamic status and at this point an optimal systemic blood pressure target is still a matter of debate in such cases.4 Prompt identification of complications following MT enables the timely initiation of life-saving treatment to enhance neurological outcomes. Transcranial Doppler (TCD) is a non-invasive, bedside, and safe modality to assess the hemodynamic status in the affected vessels.5 It offers several advantages over other imaging modalities, including cost-effectiveness, providing real-time hemodynamic parameters, and the ability to be administered periodically without any radiation risk posed to patients. Recent studies have reported the potential of TCD in predicting HT and functional recovery after successful recanalization with MT.6–9 However, the evidence on this topic remains scattered and has not been systematically reviewed or consolidated. This study aims to synthesize the available evidence and analyze the prognostic utility of TCD for predicting HT and functional recovery in patients who had successful recanalization of anterior circulation AIS-LVO following MT.
Materials and methods
This systematic review and meta-analysis followed the guidelines provided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).10 PRISMA checklists for the abstract and main document are included in the supplementary materials. The Shiny app in R was used to produce a PRISMA 2020-compliant flow diagram and checklists.11 The protocol of this review is registered prospectively in PROSPERO (registration number: CRD42024575381).
Search strategy
A comprehensive search of the literature was conducted without any restriction on language or geographical location. The databases included PubMed, Embase, and Scopus, spanning records published between January 1, 2000 to July 25, 2024. This timeline was chosen to encompass studies published in the last two decades, aligning with the period when MT and TCD became widely used in clinical research and practice. For our PubMed search, we utilized a variety of Medical Subject Headings (MeSH) terms, complemented by relevant titles and text words. The search syntax was customized to other databases to meet their specific requirements. The complete search syntax for all databases is provided in the supplementary materials (online supplemental table 1). Once we compiled the list of included studies, we manually reviewed their reference lists and conducted a citation search on Google Scholar to identify additional articles.
Supplemental material
Eligibility criteria and selection process
Studies were included if they met the following criteria: (1) Design: retrospective or prospective observational studies or clinical case series with >10 patients; (2) Population: adult patients (≥18 years) with acute ischemic stroke in anterior circulation with successful recanalization based on Thrombolysis in Cerebral Infarction (TICI) scale ≥2 b after treatment; (3) Investigation: TCD or transcranial color-coded duplex sonography (TCCD) using a low-frequency transducer (1.6–4 MHz), performed at the depth of 45–65 mm to insonate the first segment of middle cerebral artery (MCA)-M1 segment, within the first 48 hours after MT; (4) Comparison: at least one of the following TCD parameters: mean flow velocity (MFV) in the ipsilateral MCA (iMCA), MFV index (MFV of iMCA divided by unaffected MCA), peak systolic velocity (PSV) or pulsatility index (PI) (the difference between the peak systolic flow and minimum diastolic flow velocity, divided by MFV); (5) Outcome: functional status data using modified Rankin Scale score (mRS) at 90 days and HT during follow-up. The following studies were excluded: (1) Studies that used intravenous (IV) thrombolysis without MT; (2) Studies in which TCD was done more than 48 hours after MT; and (3) Case series <10 patients, conference abstracts, letters, editorials, book chapters, non-human studies, and reviews.
Two authors (SBJ and RBS) independently reviewed the titles and abstracts using eligibility criteria. If there was any disagreement, a third author (VR) was brought in to reach a consensus. The same two authors independently evaluated the full texts of all abstracts that fulfilled the inclusion criteria and checked the references of reviews. One author (SBJ) performed citation checking of included studies in Google Scholar. To prevent the inclusion of studies with overlapping populations, we thoroughly checked the final studies for similarities in authors, patient characteristics, and results. In cases where study populations overlapped or were duplicated, we selected the most comprehensive study with the largest sample size and the most extensive statistical models adjusted for a higher number of covariates. Included studies were reviewed to ensure institutional review board approval was obtained.
Data extraction
A standardized data collection form was designed, including the first author’s name and year of publication, country, period of observation, TCD protocol (frequency, depth, device, and timing relative to MT), baseline patient characteristics (age, sex, occlusion site, severity, and infarct size), TCD parameters, and patient outcomes. Afterward, two authors (SBJ and RBS) independently conducted data extraction. Any discrepancies or differences in data extraction were resolved through discussion and consensus. We also contacted the corresponding authors of all included studies, asking them to provide us with coded-blind raw materials for their research studies. In case of no reply from the corresponding authors, a follow-up email was sent 1 week after the first correspondence.
Quality assessment
Two independent reviewers (SBJ and RBS) assessed the quality of studies using an 8-point assessment tool based on the level of evidence criteria for prognostic studies by Oxford University. This tool was previously developed to measure the study quality of the predictive value of TCD after IV thrombolytic therapy by Stolz et al., and, therefore, was compatible with our aim.12 Eight criteria were considered, each receiving 1 point if the criterion was met and otherwise receiving a score of 0 for a maximum possible score of 8. Studies scoring 7–8 were considered low risk, 5–6 indicated some concern, and scores below 5 were deemed high risk of bias.
Statistical analysis
TCD parameters
We extracted the following TCD parameters: (1) PSV of MCA in the affected vessel (ipsilateral); (2) MFV (cm/s) in MCA M1 of the affected side (ipsilateral); (3) MFV index; and (4) PI of the affected side (ipsilateral). Pulsatility index is a TCD parameter that mirrors the distal cerebrovascular flow resistance.13 Because MFV depends on individual factors such as age, gender, and metabolic conditions, the MFV index better accounts for interindividual differences.14 We extracted these parameters directly from the study reports or calculated them based on the data of the individual studies.
Measures
Successful recanalization was defined based on TICI scale ≥2 b. The TICI score was based on angiography findings: 0: no perfusion and no anterograde flow beyond the point of occlusion, 1: penetration with minimal perfusion, 2: partial perfusion of less than 50% of the territory (A) or more than 50% (B); and 3: complete reperfusion.15 For two studies we used patient-level data review to exclude patients with TICI 2a from analysis to ensure the homogeneity of the study populations.
Functional recovery was assessed by mRS at 90 day follow-up. Favorable functional recovery was defined as mRS 0–2 and a poor outcome was defined as mRS 3–6. mRS is the most widely used scale for assessing functional recovery relative to pre-stroke level of activity and has 6 points: running from perfect health without symptoms to death: 0 No symptoms; 1 No significant disability; 2 Slight disability; 3 Moderate disability; 4 Moderately severe disability; 5 Severe disability; and 6 Death.16
Hemorrhagic transformation was assessed based on follow-up computed tomography (CT) or magnetic resonance imaging (MRI) of patients. For this analysis, all HT refers to any type of intracerebral hemorrhage (both symptomatic and asymptomatic) evident on follow-up imaging. Symptomatic HT was defined as neurological deterioration, defined as an increase in baseline National Institutes of Health Stroke Scale (NIHSS) score, in addition to HT.17 18
Data synthesis
We examined the association of each TCD parameter with patient outcomes. Two tables were designed to summarize the results of each TCD parameter and its association in relation to the outcomes (table 1 and table 2). For each comparison, a P-value is depicted in the table showing between-group comparisons. Depending on Gaussian distribution, Student’s t-tests or Mann–Whitney U tests were performed. Then, depending on data availability we performed meta-analyses to pool the data across studies. A cut-off of 1.25–1.3 was used for the MFV index to categorize patients into two groups of low vs high MFV index. This cut-off was selected based on our literature review.14 19 20 Due to the low number of studies, we were unable to dichotomize patients into high vs low PI groups. Additionally, due to the limited number of studies, we were unable to pool data on PSV, MFV, and MFV index in patients with different functional recovery.
TCD parameters in patients with and without HT.
TCD parameters in patients with favorable and poor functional recovery at 3 month follow-up.
For studies containing individual data, we calculated mean and standardized deviation (SD) of each TCD metric and compared them between groups using IBM SPSS Statistics version 26.0 (IBM Corp, Armonk, NY). Meta-analyses were conducted using R software version 4.3.2 (R Project for Statistical Computing) meta package version 6.5–0 and metafor package version 4.4–0. For dichotomous variables, we calculated risk ratios (RRs) and their corresponding 95% confidence intervals (CI) using a random-effects model (Generalized linear mixed models) to account for methodological heterogenities.12 21 22 For continuous variables, we calculated the standardized mean difference (SMD) between groups using Hedge’s g as effect size with a 95% CI using a random effects-model.23
Heterogeneity was assessed using Q statistic and I2 test, with I2 greater than 50% or P<0.05 considered significant. In the case of significant heterogeneity, a sensitivity analysis was planned with the removal of outlier studies to bring the heterogeneity to an insignificant level. Outliers were detected using the method previously described in the literature.24 Given that fewer than 10 studies were included in the analysis, it was not possible to conduct the Egger test for publication bias or to perform a meta-regression.25
Results
Search and screening results
The initial search retrieved 1025 records including 221 duplicates. After removing duplicates, the title and abstracts of 804 remaining records were screened, of which, 770 were excluded and 34 full texts underwent further checks. We excluded 25 full-texts because of the following reasons: Nine studies assessed TCD parameters in patients treated with IV thrombolytics only; six studies assessed TCD after 48 hours following MT; four studies did not report the pre-defined outcomes (HT or mRS); two studies did not report the desired TCD parameters; three studies were review articles and one study included posterior circulation strokes. Two additional studies were identified via citation checking of included articles. Finally, 11 studies were determined to satisfy the inclusion and exclusion criteria with the appropriate report of outcomes of interest6 8 14 26–33 (figure 1).
PRISMA flowchart detailing search and study selection.
Baseline study characteristics
The studies included AIS-LVO patients who underwent MT with MCA or ICA involvement, assessed using TCD (n=7) or TCCD (n=4) within 48 hours post-MT, with all TCD parameters measured in the MCA (M1 segment). All studies included patients with successful recanalization except two, from which, we excluded the TICI 2a patients and re-calculated rates based on data available in supplemental materials of the study28 or from raw data provided by the corresponding author.30 All studies were published between 2017–2023. Five studies were conducted in European countries, four in China, and two in North America; United States (n=1) and Canada (n=1) and all were in English language. All studies were single-center and half of them (5/11) were prospective. Overall, the studies included 1432 patients (range: 15–226 patients), including 841 (59%) men. The mean age range of patients was between 63–73 years. IV thrombolysis was used in 33–60% of patients among studies. Two studies were found to have overlapping populations.31 32 Only the most recent study with a larger sample size was included in the meta-analysis. A summary of study characteristics is presented in table 3.
Baseline study information
Risk of bias
All studies received a quality assessment score of 5 or higher (table 3). No study had high risk of bias. Ten studies had some concerns and one study had low risk of bias. Details of quality scores is presented in online supplemental table 2.
Qualitative analysis
In total, nine studies contained information regarding the association of TCD parameters and the risk of HT, and seven studies assessed TCD parameters with 90 day mRS results (online supplemental table 3). Six studies divided patients into two groups, with and without HT (HT+ and HT-, respectively), and then compared TCD parameters between them. The other three studies divided patients into categories of high vs low MFV index (n=2) or PI (n=1) and compared the rate of HT between them. The summary of TCD parameters for HT+ vs HT- and mRS 0–2 vs 3–6 is presented in tables 1 and 2 respectively.
Three studies compared mean PSV values in patients with and without HT. In the study by Katsanos et al., HT+ patients showed significantly higher PSV in both the symptomatic and asymptomatic MCA, compared with those without HT in follow-up CT scans. However, in two other studies, the difference in PSV was not statistically significant, although in both studies the PSV was higher in the HT+ group. Two studies by Baracchini et al., showed a higher PSV ratio (PSV in affected MCA/unaffected MCA) soon after MT and in 48 hour follow-up visits among patients with HT+. A higher PSV ratio was found to be an independent predictor of HT in their multivariate analysis. No other studies evaluated the PSV ratio6 (table 1).
Four studies compared mean MFV values in patients with and without HT. In the study by Kneihsl et al. (n=123, mean age: 63, male: 60%, mean time to TCD: 6.6±2.3 hours) HT+ patients exhibited significantly higher MFV in the ipsilateral MCA compared with patients without HT. However, no significant difference in MFV was observed in the contralateral MCA. The other three studies did not find a significant difference in MFV among patients with and without HT.
Quantitative analysis (meta-analysis)
The meta-analysis of MFV encompassing 409 patients (349 HT- and 63 HT+) from four studies showed higher MFV in the HT+ group (Hedges' g=0.42, 95% CI=0.08–0.76, P=0.015) (figure 2A). There was no significant heterogeneity among studies (I2=20%, P=0.38). The pooled analysis of the same patients showed a higher MFV index among HT+ patients (Hedges' g=0.54, 95% CI=0.19–0.88, P=0.003) with no significant heterogeneity among studies (I2=23%, P=0.3) (figure 2B). Furthermore, the meta-analysis of the MFV index encompassing 352 patients (110 MFV index ≥1.3 and 242 MFV index <1.3) from three studies showed a higher risk of all HT and symptomatic HT for patients with higher MFV index (RR=2.01, 95% CI=1.27–3.17, P=0.003) and (RR=4.68, 95% CI=1.49–14.65, P=0.008), respectively (figure 3A,B). There was no significant heterogeneity among studies (I2=36%, 0%, P=0.2, 0.9, respectively). Risk of poor functional outcome was higher among patients with MFV index ≥1.3 (RR=1.65, 95% CI=1.27–2.14, P=0.015), with no significant heterogeneity among studies (I2=20%, P=0.38) (figure 4).
The forest plot of MFV (A) and MFV index (B) indicates that patients with HT have significantly higher MFV and MFV index compared with those without HTMFV: Mean Flow Velocity, HT: Hemorrhagic Transformation.
The forest plot of any HT (A) and symptomatic HT (B) indicates a higher risk in patients with a high MFV index compared with those with a low MFV index.
Higher risk of unfavorable functional outcome (mRS 3–6) among patients with MFV index ≥1.3. MFV; Mean Flow Velocity.
The meta-analysis of PSV encompassing 299 patients (244 HT- and 45 HT+) from three studies showed no difference in mean PSV between groups (Hedges' g=0.41, 95% CI = −0.09–0.9, P=0.1) (online supplemental figure 1). There was no significant heterogeneity among studies (I2=36%, P=0.25).
The meta-analysis of PI encompassing 592 patients (501 HT- and 91 HT+) from five studies showed no difference in mean PI between groups (Hedges' g=0.12, 95% CI=-0.11–0.36, P=0.3) (online supplemental figure 2). There was no heterogeneity among studies (I2=0%, P=0.7).
Discussion
The findings of our study support the MFV index as a predictor of all HT, symptomatic HT, and unfavorable functional outcome after successful thrombectomy for AIS-LVO in anterior circulation. Our results suggest the use of TCD as a noninvasive tool to assess post-MT hemodynamic status with major prognostic and therapeutic implications.
Currently MT, with or without IV thrombolysis, is considered the gold standard of care for patients with AIS-LVO. However, despite successful recanalization, approximately 25–55% of patients experience futile recanalization, where there is no significant clinical improvement or even a poor outcome despite restored blood flow.34 35 Therefore, it is important to identify an inexpensive and effective monitoring system based on brain physiology to identify patients at higher risk of complications, such as HT and poor functional recovery after initial successful recanalization. Previous studies have shown that significant changes in MCA hemodynamic status occur following successful recanalization in a substantial number of patients, and these alterations are closely associated with subsequent patient outcomes.36
In a case-control study, Ng et al.,37 observed an abnormal PI (>1.2) in 47% of patients after successful thrombectomy compared with 22% in controls. Notably, PI was asymmetrically increased in the affected MCA in 28% of patients compared with 5% in controls and remained elevated even after controlling for infarct volume. Their follow-up data indicated that patients with higher microvascular resistance (asymmetrically higher PI) demonstrated a lower rate of favorable functional outcomes at 90 days. However, there was no significant association between this finding and the occurrence of symptomatic HT. We found no other studies that specifically evaluated the PI ratio, but this finding underscores the potential role of vascular resistance, particularly in relation to the presence or absence of active collaterals, in predicting patient outcomes. Further research could provide valuable insights into how these factors influence recovery and guide post-recanalization care.
Recently, Xu et al., monitored patients for a week in four sessions at 24 hours, 48 hours, 72 hours, and 7 days after MT and observed that ipsilateral MFV and the MFV index generally increase after MT but return to normal during a week, however, patients with persistent high MFV index (≥1.2) 7 days after thrombectomy, had seven times higher risk of poor functional outcome at 90 days.27 Gomes-Escalonilla et al., showed that 50% increase in the recanalized MCA MFV compared with the contralateral MCA MFV (MFV index >1.5) is a predictor of larger infarct volume.38 Another TCD study in the subacute phase of stroke (10–14 days after successful MT) suggested that hemodynamic changes in the recanalized vessel can predict 90 day outcomes,9 however more studies in the subacute phase are needed to validate their findings. Qi et al., in 2023, assessed the prognostic value of combining TCD parameters with quantitative electroencephalography and observed that combination of PI, NIHSS and pairwise derived brain symmetry index has better predictive value for mRS 4–6 at 3 months than any of the single items.8 In our analysis, MFV and the MFV index were higher among patients who developed HT. Patients with an MFV index ≥1.3 were at greater risk for HT, symptomatic HT, and unfavorable functional outcomes at 90 days. This suggests that elevated MFV and MFV index could serve as important indirect TCD markers for identifying patients at higher risk of complications following successful recanalization.
Different types of HT, such as parenchymal hematoma (PH) and hemorrhagic infarction (HI), have varying impacts on patient outcomes.39 PH is generally associated with worse functional outcomes compared with HI.39 40 The differentiation between types of HT is crucial, as it can influence the management and prognosis of patients post-recanalization. Unfortunately, the lack of data on the type of HT according to the European Cooperative Acute Stroke Study (ECASS) classification in our analysis limits our understanding of these differential impacts.
To the best of our knowledge, this study was the first meta-analysis of TCD parameters in predicting outcomes after MT. However, our study has several limitations. First, we only investigated four predefined TCD-parameters in the setting of post-thrombectomy including MFV, MFV index, PSV and PI. However, other TCD parameters such as resistive index,37 dynamic cerebral autoregulation41–43 and microembolic signals44 45 have also been studied. Assessment of all TCD parameters in literature was out of the scope of this study, however we believe these parameters may also have the potential to have predictive utility and should be evaluated in future studies. Although the heterogeneity in meta-analysis was low in all outcomes, our findings come from single-center observational studies that lack randomization and proper blinding, therefore the results should be treated with caution. We could not assess the impact of pre-stroke mRS scores due to the lack of reporting in most of the studies included in our analysis. Our risk of bias assessment highlighted the need for more prospective and multicenter studies in the field of TCD, with sufficient blinding of clinical assessors to TCD data. Future multicenter studies with a core outcome assessment are recommended. Access to individual-level data can enhance the power of analysis, however it was not possible in the present study. Finally, we were not able to perform meta-analysis for functional recovery dichotomized as mRS 0–2 vs 3–6 for variables of MFV, PSV and PI, due to a paucity of data. We suggest future TCD studies measure these metrics and provide data to enable future analyses to assess the predictive value of these parameters.
Conclusion
This study showed that MFV and MFV index are higher among acute ischemic stroke patients who experienced HT after successful reperfusion with MT. In addition, the MFV index ≥1.3 is a predictor of symptomatic HT and poor clinical outcomes. We suggest care providers plan more frequent follow-up visits and more intense care for the subgroup of patients with an MFV index≥1.3.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
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 SBJ: analysis and writing, BS and RBS: research idea and supervision, YP, VR, JCP: methodology design and data interpretation, JB, RM, KA, TH, BCM, DM: data aquisition and analysis. All authors drafted the work or revised it for significant intellectual content; All authors approved the final version of the manuscript and agree to be accountable for all aspects of the work, including its accuracy and integrity, RBS is the guarantor.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
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.