Cervical Internal Carotid Artery Reocclusion After Endovascular Thrombectomy in Tandem Occlusion: A Meta-Analysis

Abstract

BACKGROUND:

Reocclusion after endovascular thrombectomy in tandem occlusions may contribute to disability, but its mechanism and the role of emergent carotid artery stenting remain unclear. We conducted a meta-analysis to evaluate the impact of cervical internal carotid artery (c-ICA) reocclusion on functional outcomes, recurrent ischemic stroke, and the potential benefit of emergent stenting.

METHODS:

We systematically searched Medline, EMBASE, and Scopus for studies reporting c-ICA reocclusion after endovascular thrombectomy for tandem occlusions. The primary outcome was c-ICA reocclusion incidence; secondary outcomes included recurrent ischemic stroke and 90-day functional outcome (modified Rankin Scale score, 0–2). Random-effects models were used. Metaregression, sensitivity, and power analyses were performed.

RESULTS:

Fifty-one studies (4998 patients: 3444 received emergent carotid artery stenting and 1554 did not) were included. Overall, c-ICA reocclusion incidence was 10.8% (95% CI, 8.0%–14.3%). In patients treated with emergent carotid artery stenting, reocclusion occurred in 10.9% versus 28.2% without stenting (risk difference, –14.7% [95% CI, –20.2 to –9.2]). c-ICA treatment approach was not associated with recurrent ischemic stroke (odds ratio, 1.00 [95% CI, 0.33–3.01]), whereas c-ICA reocclusion was associated with worse functional outcome (odds ratio, 0.45 [95% CI, 0.28–0.71]). Metaregression identified no significant moderators. Sensitivity analyses confirmed adequate power to detect moderate associations.

CONCLUSIONS:

c-ICA reocclusion after endovascular thrombectomy in tandem occlusions is common and associated with worse functional outcome, likely due to infarct progression. Emergent carotid artery stenting reduces reocclusion rates but does not modify the risk of ischemic stroke. Future studies should refine procedural strategies and integrate physiological markers to better characterize the risk of reocclusion and outcomes.

CLINICAL PERSPECTIVE.

What Is New?

• In tandem occlusion, early cervical internal carotid artery reocclusion after endovascular thrombectomy occurs in ≈1 in 10 patients and is associated with worse 90-day functional outcome.

• Emergent carotid artery stenting lowers reocclusion compared with nonstenting strategies, but available data do not show a clear reduction in recurrent ischemic stroke.

What Are the Clinical Implications?

• Maintaining cervical carotid patency may be a clinically relevant procedural target, but the association between reocclusion and outcome should not be interpreted as causal and requires confirmation in randomized trials with standardized imaging follow-up and antithrombotic protocols.

Tandem occlusions (TOs), defined as simultaneous cervical internal carotid artery (c-ICA) stenosis or occlusion with ipsilateral intracranial large vessel occlusion (typically internal carotid artery, M1, or M2), account for ≈15% of anterior circulation ischemic strokes (ISs).¹ These patients often have worse outcomes and show limited response to intravenous thrombolysis compared with isolated large vessel occlusion.²

Endovascular thrombectomy (EVT) improves outcomes in TOs, but most pivotal EVT trials did not target this population, and management strategies for cervical lesions varied widely.^3–8 To date, no randomized controlled trial has specifically addressed the acute management of TOs.

Meta-analyses of retrospective studies have suggested better functional outcomes with emergent carotid artery stenting (eCAS) compared with no stenting; however, this remains observational, and the benefit of stenting remains uncertain.^9,10 Recently, a subanalysis of 340 patients with TOs from the IRIS collaboration also suggested better functional outcomes in patients treated with eCAS, and this association was not modified by intravenous thrombolysis.¹¹ Importantly, the mechanism of potential benefit is also unclear. Stenting may improve outcomes by maintaining acute vessel patency and supporting reperfusion, as well as by means of secondary stroke prevention. Without stenting, up to one-quarter of patients experience early c-ICA reocclusion compared with ≈15% with eCAS.¹²

In this meta-analysis, we evaluated whether eCAS reduces early reocclusion and recurrent IS and whether reocclusion is associated with worse functional outcomes. Our findings aim to clarify the clinical relevance of reocclusion and inform both current practice and the design and interpretation of ongoing trials. We aimed to address these gaps through a comprehensive meta-analysis of c-CA reocclusion, recurrent IS, and functional outcomes after EVT in TOs.

Methods

Data Availability

This systematic review and meta-analysis were based on previously published studies. The extracted data and analysis code are being organized into an open-source repository, which will be made available through GitHub and Zenodo as part of an ongoing initiative to promote reproducibility. Data sharing will follow institutional and journal policies upon completion.

We conducted this systematic review and meta-analysis in accordance with the recommendations from the PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analyses; Supplemental Material).¹³ The protocol was registered with PROSPERO (International Prospective Register of Systematic Reviews; CRD42024581676).

Ethics Approval

Ethics approval was not required as this study is a systematic review and meta-analysis of previously published studies.

Search Strategy and Selection Criteria

We conducted an extensive search of the Medline, EMBASE, and Scopus electronic databases. We executed a comprehensive literature search using a combination of MeSH terms and free text for the concepts of tandem occlusion, thrombectomy, stent, acute stroke, carotid stenosis, occlusion OR occluded, and carotid artery disease. A detailed description of the search strategy is provided in Table S1. We included studies from January 2015 to July 2024. Data from the study by Alvarado-Bolanos et al¹² were available for analysis during the study period although it was only published in April 2025.

We searched for studies assessing patients presenting with acute high-grade stenosis (70%–99%) or occlusion of the c-ICA with ipsilateral occlusion of the distal internal carotid artery and middle cerebral artery who were treated with EVT and endovascular treatment of the c-ICA (eCAS versus no-eCAS). Any cause for the TO was considered eligible for consideration. Study designs accepted for inclusion were randomized controlled trials, cohort and cross-sectional studies, case series with ≥10 patients, and case-control studies. We excluded studies conducted in animal models, protocols, reviews, studies with <10 patients, case reports, and other meta-analyses. We only included fully published studies with full text in English.

Two independent reviewers (A.A.-B. and A.M.), with training in neurointervention, initially screened all identified records by reading the titles and abstracts using the online systematic review software Covidence (Veritas Health Innovation, Melbourne, Australia; available at www.covidence.org). Potentially relevant articles were reviewed as full text. The reviewers performed data extraction from these studies and cross-checked the extracted data. Disagreements in any of these steps were resolved through discussion or, when necessary, with the assistance of a third senior reviewer (S.F.).

We ultimately included studies reporting the incidence of c-ICA reocclusion and recurrent IS after EVT, stratified according to the treatment of the c-ICA lesion (ie, eCAS versus no-eCAS). Last, we included studies reporting modified Rankin Scale scores stratified according to c-ICA patency (ie, occluded versus patent).

Data Collection and Outcomes

From every included study, we collected the number of participants, age, sex, and cardiovascular comorbidities. We also collected the underlying cause (ie, atherosclerotic or dissection) for the TO, the degree of c-ICA stenosis/occlusion, stroke workflow metrics, EVT procedural data (first-pass effect, modified Thrombolysis in Cerebral Infarction score ≥2b, and use of angioplasty), and periprocedural use of antiplatelets (single-antiplatelet therapy, dual-antiplatelet therapy, or GpIIb/IIIa [glycoprotein IIb/IIIa receptor] inhibitors). We collected the exact number of patients treated with eCAS or no-eCAS and those who developed c-ICA reocclusion. No-eCAS patients included cases where the cervical lesion was treated with balloon angioplasty, aspiration, or no specific treatment.

The primary outcome was the incidence of c-ICA reocclusion during follow-up after EVT for TO on any diagnostic modality at any time during the available follow-up period. Any mechanism of c-ICA reocclusion was accepted, and both symptomatic and asymptomatic c-ICA reocclusions were included. Results were stratified according to treatment approach into eCAS or no-eCAS.

Secondary outcomes included the incidence of recurrent IS, as defined by the study, stratified by treatment modalities for the c-ICA lesion. We dichotomized the 90-day modified Rankin Scale score into good (0–2) and poor (3–6) outcomes and stratified them according to c-ICA patency. For analyses relating c-ICA patency to functional outcome, patency status was defined radiographically irrespective of clinical symptoms; therefore, asymptomatic reocclusions were included when reported.

Study Quality and Risk of Bias Assessment

The quality of the studies was assessed depending on the study’s design. For randomized controlled trials, we used the RoB 2 tool¹⁴; for nonrandomized studies with a comparator, we used the ROBINS-I tool¹⁵; and for studies without a comparator, we used the Minors score.¹⁶ Ultimately, the overall risk of bias was rated as low, moderate, or serious.

Statistical Analysis

All analyses were conducted using R (version 4.4.2). A complete, reproducible version of the analysis code is available on GitHub/Zenodo.

Primary Analyses

We conducted meta-analyses to estimate the incidence and comparative effects of eCAS in patients with TO undergoing EVT. The primary outcome was the rate of cervical internal carotid artery reocclusion. Secondary outcomes included recurrent IS, and a good functional outcome at 90 days, defined as a modified Rankin Scale score of 0 to 2.

We used random-effects models to account for between-study heterogeneity. Dichotomous outcomes were summarized using pooled odds ratios or risk differences, and single-arm incidence outcomes were modeled using logit-transformed proportions. Between-study heterogeneity was assessed with the I² statistic and τ². Prediction intervals were reported when applicable.

Supplemental Analyses

We conducted several prespecified sensitivity and exploratory analyses. Analytical and simulation-based power estimates assessed the ability to detect an odds ratio of 0.45 across varying baseline event rates. Metaregression analyses included year of publication, mean age difference, and sex difference between groups. Subgroup analyses evaluated reocclusion rates by occlusion type, timing of reocclusion, and patient characteristics. Influence diagnostics included leave-one-out analyses, and graphical heterogeneity was explored using GOSH analysis. All tests were 2-sided, with P<0.05 considered statistically significant.

Results

Characteristics of Included Studies and Bias Assessment

Our search detected 2011 citations, of which 186 were potentially eligible for inclusion (Figure 1).

Figure 1.

PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram for study selection. Flow diagram summarizing identification, screening, eligibility assessment, and inclusion of studies in the systematic review and meta-analysis according to PRISMA guidelines.

We included 51 studies encompassing 4998 patients with TOs treated with EVT, of which 3444 received eCAS and 1554 did not. Of these studies, only 1 study was a subanalysis from a randomized controlled trial, 32 studies had a control group (ie, eCAS versus no-eCAS or c-ICA reocclusion versus patent), and 18 studies had no control group.

Of all 51 included studies, only 17 reported the c-ICA cause. Atherosclerosis represented the cause in 36.9% (n=1845) and carotid dissection in 6.8% (n=343), and the cause was not specified in 56.2% (n=2810) of included patients.

The substantial heterogeneity observed in several pooled estimates is consistent with the predominance of nonrandomized study designs, variation in patient selection and stenting indications, and inconsistency in antithrombotic regimens and use of thrombolysis, as well as heterogeneous reporting of outcomes, covariates, and imaging follow-up protocols across studies (Tables S2 through S4).

Incidence of c-ICA Reocclusion

Forty-eight studies reported the incidence of c-ICA reocclusion (n=4370; Table S5). The overall pooled estimate for the incidence of c-ICA reocclusion was 10.8% (95% CI, 8.0%–14.3%), with substantial heterogeneity (I²=88.0%; Figure 2).

Figure 2.

Meta-analysis of the overall incidence of cervical internal carotid artery (c-ICA) reocclusion. Forest plot displaying pooled incidence of c-ICA reocclusion following endovascular thrombectomy for tandem occlusions across included studies.

When stratified by treatment approach, the reocclusion rate among patients who underwent eCAS was 10.9% (95% CI, 8.5%–13.8%; 48 studies, 3029 patients; I²=66.8%). In contrast, studies in which eCAS was not performed and angioplasty alone or no cervical treatment was used showed a reocclusion incidence of 28.2% (95% CI, 20.5%–37.5%; 14 studies, 568 patients; I²=71.4%). Among 14 comparative studies, the pooled risk difference for reocclusion comparing eCAS versus no-eCAS was –14.7% (95% CI, –20.2 to –9.2), favoring eCAS. Heterogeneity was low (I²=0%; Figure 3).

Figure 3.

Meta-analysis comparing the incidence of cervical internal carotid artery (c-ICA) reocclusion between emergent carotid artery stenting (eCAS) and no-eCAS strategies. Forest plot illustrating pooled risk difference (RD) in c-ICA reocclusion between treatment approaches.

IS Recurrence According to c-ICA Treatment Approach

Only 5 studies (759 patients) reported the risk of 90-day recurrent IS in relation to c-ICA treatment approach (eCAS versus no-eCAS; Table S6). The pooled odds ratio was 1.00 (95% CI, 0.33–3.01; I²=18.0%), indicating no association between treatment approach and recurrent IS (Figure 4).

Figure 4.

Meta-analysis of recurrent ischemic stroke comparing emergent carotid artery stenting (eCAS) vs no-eCAS. Forest plot showing pooled odds ratio (OR) for 90-day recurrent ischemic stroke according to cervical carotid treatment strategy.

Association Between c-ICA Reocclusion and Functional Outcome

Twelve studies (1355 patients) evaluated the association between c-ICA reocclusion and 90-day functional outcome (Table S7). The pooled odds ratio for achieving a favorable outcome (modified Rankin Scale score, 0–2) in patients with reocclusion versus those without was 0.45 (95% CI, 0.28–0.71; I²=44.2%), indicating worse outcomes in those with reocclusion (Figure 5).

Figure 5.

Meta-analysis of proportions of 90-day modified Rankin Scale score of 0 to 2 of studies directly comparing cervical internal carotid artery (c-ICA) reocclusion and patency. Forest plot depicting pooled odds ratio (OR) for favorable outcome (modified Rankin Scale score, 0–2) in patients with c-ICA reocclusion vs patent c-ICA.

Baseline Characteristic According to c-ICA Treatment Approach

Baseline characteristics were variably reported across included studies. Both treatment approaches showed similar vascular risk profiles though atrial fibrillation was more frequent in nonstenting cohorts, while severe stenosis and carotid occlusion were reported more often in stented patients. The mean age was around 69 years, and most patients were male. A full description is depicted in Table S8.

Timing of Reocclusion and Metaregression

A mixed-effects metaregression was conducted in 44 studies that reported the timing of reocclusion. Timing was categorized as (1) within 7 days, (2) 7 days to 3 months, and (3) after 3 months. There was no significant difference in the proportion of reocclusion across time intervals (QM [df=2]=1.57; P=0.456). The largest number of studies (n=29) defined reocclusion within the first 7 days, limiting between-group comparisons.

Discussion

Overall, ≈10% of patients with TOs experience early reocclusion after thrombectomy, and this proportion is higher when eCAS is not performed. Some observational studies have reported an association between post-EVT c-ICA patency and better functional outcomes in TOs, whereas others have not. To our knowledge, this is the first meta-analysis confirming this association and suggesting that maintaining c-ICA patency may be one mechanism by which eCAS has been associated with better functional outcomes in observational cohorts and prior meta-analyses. Importantly, the observed association between c-ICA reocclusion and functional outcome should not be interpreted as causal, given the predominance of nonrandomized evidence and the potential for residual confounding (eg, more complex cervical pathology, antithrombotic variability, or procedural complexity).^17–19

As demonstrated by our meta-analysis, recurrent IS is an outcome that has seldom been reported after EVT for TO. Outside the setting of EVT for TO, the risk of IS recurrence among patients with symptomatic c-ICA occlusion has been estimated to be around 15% to 24% during the first 2 years of follow-up.^20,21 The benefit of carotid revascularization among these patients remains an area of major controversy. A meta-analysis of pivotal randomized controlled trials failed to show the benefit of carotid revascularization in patients with occlusion or carotid near-occlusion,²² but other large meta-analyses, including observational studies, have suggested some benefit.^23,24

Our findings carry significant procedural implications. When the primary aim of EVT is to ensure short-term patency during the initial event, rather than achieving a durable anatomic reconstruction, procedural strategies may prioritize technical sufficiency over maximum optimization. For instance, minimizing vessel manipulation, including avoiding eCAS or angioplasty, may mitigate the risks of hyperperfusion injury and hemorrhage, especially in patients who received intravenous thrombolysis. On the other hand, a minimalistic approach can result in a lost opportunity to revascularize the c-ICA in as high as nearly one-fourth of patients not treated with eCAS.

Monitoring c-ICA reocclusion and its associated consequences during the early recovery phase might be relevant but poses substantial challenges. Conventional clinical examination may lack the reliability needed to detect recurrent embolic events or evolving ischemia, as new deficits might be obscured by residual symptoms stemming from the index stroke. Imaging markers may offer a more nuanced understanding of these processes: magnetic resonance imaging with diffusion-weighted imaging can uncover new embolic lesions, perfusion imaging can indicate ongoing hemodynamic compromise, and carotid Doppler may reveal evolving cervical thrombosis. Furthermore, transcranial Doppler can provide continuous assessment of embolic signals, hypoperfusion, or hyperperfusion, offering dynamic physiological monitoring during this critical window. By leveraging these modalities, clinicians can better differentiate between infarct growth, silent recurrent embolism, or delayed recovery. Nevertheless, the practicality, timing, and safety of these modalities remain topics of uncertainty, particularly due to concerns regarding cost, procedural delays, contrast exposure, and determining the optimal monitoring window. A recently published large multicenter registry of patients with TOs, including 4053 patients, found an increased risk of early neurological deterioration in patients not treated with eCAS.²⁵ The mechanism, however, remains unknown, and the authors did not report the rate of c-ICA reocclusion.

These considerations can impact both ongoing registry initiatives and the design of future clinical trials. Recent observational studies increasingly use imaging and physiological markers to enhance the characterization of reocclusion and assess its functional consequences.^26,27 Future trials focused on stenting strategies, antiplatelet regimens, or adjunctive therapies should consider integrating these markers to refine event adjudication and improve outcomes. While functional outcomes remain the most clinically relevant end points, assessments based on imaging of infarct progression and early physiological monitoring may yield important complementary insights necessary for accurately defining treatment effects.

The cause of c-ICA occlusion was seldom reported across included studies and was unspecified in >50% of included patients, with <10% of cases being secondary to dissection. These proportions seem to be in keeping with other large series that have reported dissection as the cause for TOs in 10% to 20% of cases.^25,28 We think that our results can largely be applied to patients with atherosclerotic c-ICA occlusion, but to what extent they can be applied to patients with dissection is unclear.

Our study has several limitations. A key limitation is that the meta-analysis was largely based on observational studies, with only 1 subanalysis derived from a randomized controlled trial, making the pooled estimates susceptible to selection bias and residual confounding. Similarly, significant heterogeneity was observed among included studies, which weakens the precision of the pooled estimates despite the use of random-effects modeling and limits the generalizability of the findings. In addition, although the procedural and outcome data were generally well reported, the completeness of reporting for demographic, clinical, and imaging covariates varied considerably across studies. Similarly, there was significant heterogeneity in the reporting of antithrombotic regimen and the cause of the c-ICA occlusion across studies (ie, mostly atherosclerosis and dissection). This might be relevant as natural history and recurrent IS rates might be different across antithrombotic regimens and vary between the 2 causes. Nonetheless, several large studies have only found a small or marginal effect of antithrombotic regimen on functional outcomes.^25,29 These inconsistencies hamper the capacity to investigate heterogeneity through metaregression. On the other hand, sensitivity analyses affirmed that the study was adequately powered to detect moderate associations between reocclusion and stroke recurrence, even across a variety of baseline event rates.

This analysis is further constrained by the inconsistent reporting of reocclusion across different studies and the lack of systematic follow-up imaging in most cohorts. A limited number of studies evaluated cervical patency beyond the acute phase, which restricts our ability to examine temporal patterns of reocclusion and their association with outcomes. In addition, variability in stenting techniques, antiplatelet regimens, and imaging protocols may exacerbate heterogeneity among findings. Similarly, studies used different definitions for recurrent IS. Although clinical implications might be different, our results did not make any distinction between symptomatic or asymptomatic recurrent IS or between symptomatic or asymptomatic c-ICA reocclusion. Ongoing randomized trials, such as TITAN (Thrombectomy in Tandem Lesion), EASI-TOC (Endovascular Acute Stroke Intervention – Tandem Occlusion), and CASES (Carotid Artery Stenting During Endovascular Treatment of Acute Ischemic Stroke Versus Deferred Treatment of Carotid Artery Stenosis; the Netherlands/Belgium), are anticipated to yield more standardized data concerning reocclusion events, procedural strategies, and outcomes. Our study provides important insights, suggesting that maintaining c-ICA patency is a relevant outcome. First, c-ICA reocclusion occurs in about 1 in 10 patients with TOs following thrombectomy. Second, the rate of c-ICA occlusion can be as high as 1 in 4 if eCAS is not performed. Third, eCAS does not seem to be associated with lower rates of recurrent IS compared with no-eCAS. Fourth, c-ICA reocclusion is associated with worse functional outcomes, and this association could be primarily driven by infarct growth and progression. Until prospective evidence becomes available, it appears reasonable to favor eCAS as a means of preventing c-IC reocclusion and to obtain better functional outcomes.

ARTICLE INFORMATION

Author Contributions

Drs Alvarado-Bolanos and Fridman were involved in study conception and design. Drs Alvarado-Bolanos and Mascarenhas were involved in data collection. Dr Fridman was involved in statistical analysis. All authors were involved in interpretation of data. Drs Alvarado-Bolanos and Fridman were involved in manuscript drafting. All authors were involved in critical revision of manuscript.

All authors were involved in final approval.

Sources of Funding

None.

Disclosures

None.

Supplemental Material

Supplemental Materials and Methods

Tables S1–S8

Search Strategy

Risk of Bias

Sensitivity, Subgroup, Meta-Regression, Influence, and GOSH Analyses

GOSH Plots and Influence Diagnostics

Leave-One-Out and Sensitivity Analyses

Data and Code Availability

PRISMA Checklist