Revascularization Treatment of Basilar Artery Occlusion in Patients with Mild‐to‐Moderate Severe Stroke

Abstract

BACKGROUND

The benefit of endovascular thrombectomy (EVT) in patients with basilar artery occlusion and severe neurological deficits is well established. However, its effectiveness in those with mild‐to‐moderate deficits remains uncertain. This study compared outcomes of EVT(±intravenous thrombolysis [IVT]) versus IVT alone in patients with basilar artery occlusion and mild‐to‐moderate stroke severity.

METHODS

We used data from the international multicenter EVA‐TRISP (Endovascular Treatment and Thrombolysis for Ischemic Stroke Patients) and TRISP (Thrombolysis for Ischemic Stroke Patients) collaboration. All patients with radiologically confirmed basilar artery occlusion, National Institutes of Health Stroke Scale score <10, and a time to first treatment within 6 hours were included. Main outcomes of interest were favorable (functional) outcome (modified Rankin Scale [mRS] score 0—2), overall distribution of mRS, mortality at 3 months, and symptomatic intracranial hemorrhage. We applied binary logistic and ordinal regression using covariate adjustment and inverse probability of treatment weighting.

RESULTS

Among 274 patients from 18 centers, 176 (64.3%) received EVT (mean age 68±15 years, 38% female, median [interquartile range] National Institutes of Health Stroke Scale score 5 [3—8], 34% with bridging IVT) and 98 (35.8%) received IVT alone (mean age 70±13 years, 43% female, median National Institutes of Health Stroke Scale score 5 [4–8]). Favorable outcome occurred in 63.6% of patients with EVT(±IVT) and in 64.3% of patients with IVT alone (adjusted odds ratio [OR] 0.89, 95% CI 0.46–1.72). There was an association of EVT(±IVT) with unfavorable distribution of the mRS (adjusted OR 1.83, 95% CI 1.10–3.06), and mortality was higher in the EVT(±IVT) group (15.9% versus 6.1%, adjusted OR 3.38, 95% confidence interval 1.30–8.75). Rates of symptomatic intracranial hemorrhage did not differ between groups (2.0% versus 0%). The results remained unchanged after additional inverse probability of treatment weighting analyses.

CONCLUSIONS

In this multicenter observational cohort study, EVT(±IVT) in patients with basilar artery occlusion with mild‐to‐moderate stroke, was not associated with improved clinical outcome but higher mortality compared with IVT‐treatment. Our findings underscore equipoise and the need for prospective trials in this population.

Stroke due to basilar artery occlusion (BAO) accounts for 1%–2% of all ischemic strokes. ¹ , ² , ³ BAO is a life‐threatening condition and is associated with a high rate of morbidity and mortality. ¹ Intravenous thrombolysis (IVT) has been used for decades to treat strokes in the posterior circulation. ⁴ Endovascular treatment (EVT) has shown strong benefit in several randomized controlled trials (RCTs) in anterior circulation stroke. ⁵ In the posterior circulation, particularly in BAO, the benefit of adding EVT to IVT alone remained controversial for a long time. Earlier RCTs did not demonstrate a benefit of EVT over IVT alone in patients with BAO. ⁶ , ⁷ However, a subgroup analysis of the BASICS (Basilar Artery International Cooperation Study) trial suggested that patients with severe neurological deficits, defined as a National Institutes of Health Stroke Scale (NIHSS) score ≥10, may have a better outcome with EVT compared with best medical treatment, whereas in patients with NIHSS score <10, no difference between treatment arms was observed. ⁷ More recent RCTs conducted in Chinese populations (BAOCHE [Basilar Artery Occlusion Chinese Endovascular], ⁸ ATTENTION [Endovascular Treatment for Acute Basilar‐Artery Occlusion] ⁹ ) have demonstrated a benefit of EVT over best medical therapy (<50% with IVT) for the treatment of BAO. However, patients with mild‐to‐moderate symptoms (NIHSS score <10) were either not enrolled (ATTENTION) or underrepresented (BAOCHE, only 12/217 patients) in these studies. ⁸ , ⁹ A recently published individual patient‐data meta‐analysis of the landmark RCTs demonstrated a robust benefit of EVT. However, there was evidence of a treatment interaction with no sustained benefit of EVT in patients with NIHSS score <10. ¹⁰ Accordingly, the 2024 European Stroke Organization guidelines for the management of BAO recommend EVT(±IVT) if NIHSS score is ≥10 and recommend IVT only in patients with NIHSS score <10. ¹¹

Taken together, although the efficacy of EVT for BAO has been shown in severely affected patients, the potential benefit of EVT in patients with mild‐to‐moderate stroke (NIHSS score <10) remains unclear. In a large European cohort study, we aimed to analyze the association of EVT with or without IVT compared to IVT alone in patients with BAO and mild‐to‐moderate deficits with functional outcome, distribution of disability, mortality, and symptomatic intracranial hemorrhage (sICH).

METHODS

Study Design and Study Population

This retrospective, multinational cohort study was conducted based on prospectively collected data from the EVA‐TRISP (Endovascular Treatment and Thrombolysis for Ischemic Stroke Patients) and TRISP (Thrombolysis for Ischemic Stroke Patients) collaborations. ¹² Whereas the investigator‐driven TRISP registry collects information on patients with ischemic stroke predominantly treated with IVT (using alteplase and not tenecteplase), the EVA‐TRISP registry has been developed from the former TRISP collaboration by focusing on EVT‐treated patients. TRISP and EVA‐TRISP are independent from industry and nonprofit collaborations. All data were collected locally in participating centers (25 centers in 11 countries, ongoing recruitment), anonymized, and centrally pooled at the coordinating stroke center in Basel, Switzerland. The methodology, design, structure, and policy of the TRISP and EVA‐TRISP registry have previously been published. ¹² , ¹³ Each center has received necessary official approval from its respective local authorities and/or ethical committees according to national and local rules. This manuscript conforms to the Strengthening the Reporting of Observational Studies in Epidemiology reporting guideline and checklist. Data from this study are available from the corresponding author upon reasonable request.

Inclusion criteria for this analysis were documented BAO with and without additional occlusion of vertebral artery or posterior cerebral artery, baseline stroke severity <10 points according to the NIHSS score, onset to treatment time (OTT) ≤6 hours, and available information about functional outcome according to the modified Rankin Scale (mRS) at 3 months (see Figure 1). Patients with concomitant vessel occlusion in the anterior circulation and unknown symptom onset were excluded. Data were collected until April 2024 (inclusion periods of individual centers are shown in Table S1).

Figure 1.

Flow chart of patient selection. BAO indicates basal arterial occlusion; EVT, endovascular treatment; IVT, intravenous thrombolysis; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; OTT, onset‐to‐treatment time; sICH, symptomatic intracerebral hemorrhage; and TRISP/EVATRISP, Thrombolysis for Ischemic Stroke Patients/Endovascular Treatment and Thrombolysis for Ischemic Stroke Patients.

Nonstandard Abbreviations and Acronyms

BAO

basilar artery occlusion

EVT

endovascular thrombectomy

IVT

intravenous thrombolysis

mRS

modified Rankin Scale

NIHSS

National Institutes of Health Stroke Scale

OTT

onset to treatment time

sICH

symptomatic intracranial hemorrhage

VA

vertebral artery

CLINICAL PERSPECTIVE

What Is New?

• In patients with basilar artery occlusion and mild‐to‐moderate stroke severity (National Institutes of Health Scale score <10), endovascular treatment was not associated with better outcomes than intravenous thrombolysis alone.

What Are the Clinical Implications?

• These findings highlight clinical equipoise and emphasize the urgent need for randomized trials to guide treatment decisions in this understudied population.

Data Collection

The data were prospectively collected according to prespecified data definitions. Baseline parameters such as demographics, laboratory results, OTT, NIHSS score on admission, and treatment modality (EVT±IVT or IVT alone) were assessed, which are key baseline variables and already available in the EVA‐TRISP data set. ¹² Functional outcome at 3 months was obtained by mRS‐certified clinicians by telephone calls, postal questionnaire, or outpatient visits. ¹²

For this particular project, additional variables included exact location of BAO (proximal basilar [from the confluence of the vertebral arteries to the origin of the anterior inferior cerebellar artery], midbasilar [from the origin of the anterior inferior cerebellar artery to the origin of the superior cerebellar arteries], and distal [including top‐of‐the‐basilar occlusion]), ⁷ exact segments of concomitant occlusion (vertebral artery, posterior cerebral artery), level of consciousness according to NIHSS subitem 1a and evidence of symptom fluctuation before the initiation of revascularization treatment were collected retrospectively by asking each center to reassess these variables.

Outcomes of Interest

Primary outcomes of interest were a favorable functional outcome according to the mRS score 0–2 at 3 months and different dichotomizations of the mRS at 3 months (mRS score 0–3 [fair outcome, corresponding to independent ambulation] and mRS 0–1 [excellent outcome]). Another outcome of interest was a shift in the distribution toward increased disability on the mRS score for functional outcome at 3 months. ¹⁴ Additionally, we assessed mortality at 3 months as well as sICH (according to the European Cooperative Acute Stroke Study II criteria). ¹⁵

Statistical Analysis

Categorial variables are presented as absolute numbers and percentages. Continuous variables were estimated as either means with SD or medians with interquartile range and tested for normality. Univariate analyses were performed using a t‐test or Mann–Whitney U test. The significance level for all tests was set at P<0.05.

We investigated associations between treatment modalities (EVT±IVT or IVT alone) and outcomes by calculating odds ratios (OR) with 95% CI, using both binary and ordinal logistic regression models. In the next step, we adjusted for age, sex, OTT, prior stroke, diabetes, atrial fibrillation, additional vessel occlusion, year of treatment after 2014, NIHSS score on admission, prestroke disability (mRS score >2), and location of BAO. These variables were chosen because they are known strong predictors of unfavorable outcome after revascularization therapies, as well as variables that differed between the treatment groups in the univariable comparison (P<0.2). ¹⁶ , ¹⁷ Due to low event rates, the analysis regarding 90‐day mortality was adjusted for age, NIHSS score, and additional vessel occlusion, and in a second model, also for the previously mentioned variables. We further compared the overall distribution of mRS scores after 3 months using an ordinal logistic regression model (shift analysis), with an OR of > 1 implicating reduced disability and <1 for increased disability after EVT(±IVT) (see Figure 2).

Figure 2.

mRS distribution of functional outcome at 3 months. Distribution of the 90‐day mRS in each group. The adjusted odds ratio was 1.83 [1.10–3.06]. Adjusted for age, sex, prior stroke, independent mRS score, diabetes, onset‐to‐treatment time, atrial fibrillation, additional vessel occlusion, year of treatment after 2014, National Institutes of Health Stroke Scale score on admission, and location of basal arterial occlusion. EVT indicates endovascular thrombectomy; IVT, intravenous thrombolysis; and mRS, modified Rankin Scale.

In the next step, we performed a logistic regression analysis following inverse probability of treatment weighting (IPTW) using the aforementioned variables to compare treatment modalities (EVT±IVT or IVT alone). The weights were calculated as average treatment effects from a regression model for the treatment received and were truncated at 10. The fit of the logistic regression was assessed using McFadden's Pseudo‐R². Details on covariates and effective sample size after IPTW and a love plot of standardized mean differences in covariates before and after inverse IPTW (see Table S3, Figures S1 and S2).

Further, we added sensitivity analyses that included patients with isolated BAO and no additional occlusions in vertebral artery/posterior cerebral artery regarding our main outcomes (see Table 3) and a secondary analysis with patients who underwent EVT(±IVT) only to evaluate factors associated with a favorable outcome at 3 months (mRS 0—2), which may provide a foundation for inclusion/exclusion criteria for future trials (see Table S4).

Table 3.

Sensitivity Analysis for Isolated BAO and Outcomes After 3 Months

Isolated BAO n = 222	EVT±bridging IVT (% (n))	IVT (% (n))	Unadjusted OR 95% [CI]	Adjusted OR 95% [CI]
Favorable outcome (mRS 0–2)	49.2 (64)	58.7 (54)	0.68 [0.40–1.17]	0.52 [0.27–1.00]
Excellent outcome (mRS 0–1)	66.9 (87)	65.2 (60)	1.08 [0.61–1.90]	0.84 [0.43–1.66]
Fair outcome (mRS 0–3) *	78.5 (102)	82.6 (76)	0.77 [0.34–1.52]	0.62 [0.28–1.34]
Mortality (3 mo)	15.4 (20)	6.5 (6)	2.60 [1.00–6.77]	3.60 [1.15–11.20] 3.21 [1.20–8.60] †
Symptomatic intracranial hemorrhage (sICH, ECASS‐II criteria)	0.8 (1)	0.0 (0)	‐	‐

BAO indicates; ECASS, European Cooperative Acute Stroke Study; EVT, endovascular treatment; IVT, intravenous thrombolysis;mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio; and sICH, symptomatic intracranial hemorrhage.

^* Independent ambulation.

Adjusted for age, sex, prior stroke, independent mRS score, diabetes, onset‐to‐treatment time, atrial fibrillation, year of treatment after 2014, NIHSS score on admission, and location of BA.

^† Adjusted for age, NIHSS score.

In the last step, we conducted a sensitivity analysis to account for prestroke anticoagulation as a formal contraindication for IVT but not EVT (Table S2). In this analysis, we included only patients without any type of prestroke anticoagulation (novel oral anticoagulants, vitamin K antagonists, other). Data were statistically analyzed using Statistical Package for Social Sciences (IBM SPSS Statistics, version 27) and R (GNU R 4.4.1).

RESULTS

From all 31 235 registered patients in the TRISP and EVA‐TRISP registries, 778 patients with stroke due to BAO were identified (2.5%, inclusion period: 2006–2023). Of these, 392 had a baseline NIHSS score <10. After exclusion of patients with missing information on OTT (n = 27) or with OTT >6 hours (n = 75) and patients with missing information regarding the outcomes of interest (n = 19), 274 patients were finally included in the analysis. For details on patient selection, see Figure 1.

Of these 274 patients, 176 patients (64.2%) received EVT (34.4% with bridging IVT), and 98 patients (35.8%) received IVT only. Details on individual contributions per center are shown in Table S1.

Baseline patients’ characteristics of the 2 treatment groups are shown in Table 1. Age, sex, admission NIHSS score, prestroke dependency, time to treatment, and level of BAO were equally distributed between groups. Compared with IVT alone, EVT(±IVT) was more often performed after the year 2014. Patients receiving EVT(±IVT) had a concomitant vertebral artery or posterior cerebral artery occlusion more often and received a prior oral anticoagulant significantly more often. Symptom fluctuation before treatment occurred more often in patients with IVT (44.2% versus 32.4%); however, this difference did not reach statistical significance (P = 0.07). Level of consciousness tended to be higher in EVT(±IVT) patients (alert: 71.5% versus 65.3%, P = 0.06).

Table 1.

Comparison of Baseline Characters Between Patients With EVT(± IVT) and IVT Alone

Characteristics	EVT±bridging IVT (n = 176)	IVT alone (n = 98)	P value
Admission data
Age, y, mean [±SD]	68 (15)	70 (13)	0.23
Female sex, % (n)	38.1 (67)	42.9 (42)	0.44
NIHSS score on admission, median [IQR]	5 [3–8]	5 [4–8]	0.33
NIHSS score ≤5 (minor stroke), % (n)	52.9 (91)	45.9 (45)	0.27
RR systolic on admission in mm Hg, median [IQR]	150 [130–170]	150 [137–160]	0.32
RR diastolic on admission in mm Hg, median [IQR]	80 [71–91]	80 [70–90]	0.75
Glucose on admission, median [IQR]	7.2 [6.2–8.9]	7.2 [6.1–8.3]	0.96
OTT, min, median [IQR]	168 [114–235]	171 [125–216]	0.89
DTT min, median [IQR]	40 [26–65]	57 [40–75]	0.35
Year of treatment after 2014, % (n)	96.6 (170)	66.3 (65)	<0.001
Patients’ history
Prestroke independence, % (n)	93.1 (163)	89.7 (87)	0.32
Prestroke mRS score, median [IQR]	0 [0.0–1.0]	0 [0.0–1.0]	0.04
Prior ischemic stroke, % (n)	19.9 (35)	16.3 (16)	0.47
Prestroke anticoagulants			0.01
Vitamin K antagonists, % (n)	7.4 (13)	0 (0)
Novel oral anticoagulants, % (n)	5.7 (10)	2.0 (2)
Other, % (n)	1.1 (2)	0 (0)
Atrial fibrillation, % (n)	28.4 (50)	20.4 (20)	0.15
Diabetes, % (n)	15.9 (28)	15.3 (15)	0.9
Hypertension, % (n)	69.9 (123)	72.4 (71)	0.66
Hypercholesterolemia, % (n)	44.9 (79)	53.1 (52)	0.19
Current smoking (or stopped <2 y), % (n)	19.4 (34)	17.3 (17)	0.67
Stroke etiology
Modified TOAST			0.12
Cardioembolic (incl PFO), % (n)	31.8 (56)	26.5 (26)
Large‐artery occlusion, % (n)	31.8 (56)	31.6 (31)
Small artery occlusion, % (n)	1.7 (3)	8.2 (8)
Other, % (n)	7.4 (13)	4.1 (4)
>1, % (n)	2.8 (5)	6.1 (6)
Undetermined, % (n)	23.9 (42)	22.4 (22)
BAO characteristics
Location of BAO			0.78
Proximal, % (n)	15.9 (28)	16.3 (16)
Midbasilar, % (n)	17.0 (30)	13.3 (13)
Distal, % (n)	58.0 (102)	59.2 (58)
Multisegmental, % (n)	9.1 (16)	11.2 (11)
Level of consciousness (NIHSS subitem 1a)			0.06
Alert, % (n)	71.5 (118)	65.3 (62)
Minor stimulation, % (n)	24.8 (41)	22.1 (21)
Major stimulation, % (n)	3.0 (5)	10.5 (10)
Unresponsive, % (n)	0.6 (1)	2.1 (2)
Evidence of symptom fluctuation before treatment, % (n)	32.4 (48)	44.2 (42)	0.07
Concomitant occlusion of the vertebral artery or posterior cerebral artery % (n)	26.1 (46)	6.1 (6)	<0.001

BAO indicates basilar arterial occlusion; DTT, door to treatment; EVT, endovascular treatment; IVT, intravenous thrombolysis; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; OTT, onset to treatment; PFO, patent foramen ovale; RR, blood pressure; TOAST, Trial of ORG 10172 in Acute Stroke Treatment; and VO, vessel occlusion.

P values <0.05 that are considered statistically significant are depicted in a bold font.

Missing data: functional independence n = 2; current smoking (or stopped <2 y) n = 1; DTT n = 17; systolic blood pressure on admission n = 31; diastolic blood pressure on admission n = 32; Prestroke mRS scoren = 1; glucose on admission n = 23; level of consciousness according to NIHSS subitem 1a n = 14; evidence of symptom fluctuation before treatment n = 31.

Compared with IVT alone, EVT(±IVT) was not associated with a favorable functional outcome at 3 months (mRS score 0–2; 63.6% versus 64.3%, adjusted OR 0.89 [0.46–1.72]). Using fair outcome criteria as in the BAO RCTs (mRS score 0–3), corresponding to independent ambulation, there was also no difference between the treatment groups (75.0% versus 81.7%, adjusted OR 0.59 [0.28–1.26]), whereas after IPTW, the aOR_IPTW of 0.56 [0.35–0.89] indicating higher odds of worse outcome following EVT(±IVT). Considering an excellent outcome (mRS score 0–1), no significant difference was observed either (47.1% versus 57.1%, adjusted OR 0.56 [0.30–1.05]). IPTW analysis also revealed an association of EVT(±IVT) with worse outcome (OR_IPTW 0.63 [0.43–0.92]). With regard to the overall distribution of disability across the mRS score at 3 months, we identified a significant shift toward higher mRS scores in the EVT(±IVT) group (adjusted OR = 1.83 [1.10–3.06]). Mortality at 3 months was higher in patients with EVT(±IVT) compared with IVT alone (15.9% versus 6.1%, adjusted OR 3.38 [1.30–8.75]. Rates of sICH did not differ between groups (2.0% versus 0%). An overview of the association of treatment with the outcomes of interest is shown in Table 2. Due to the higher number of tandem occlusions (see Table 1) in patients with EVT(±IVT), we added a sensitivity analysis that included only patients with isolated BAO (n = 222, see Table 3). Only mortality was found to be significantly less favorable to EVT(±IVT). The remaining results were not noticeably affected (favorable outcome at 3 months (mRS score 0–2) 66.9% versus 65.2%, adjusted 0.84 [0.43–1.66]); fair outcome at 3 months (mRS score 0–3) 78.5% versus 82.6%, adjusted 0.62 [0.28–1.34]); excellent outcome at 3 months (mRS score 0–1) 49.2% versus 58.7%, adjusted 0.52 [0.27–1.00]; mortality 15.4% versus 6.5%, adjusted OR 3.60 [1.15–11.20], sICH 0.8% versus 0.0%).

Table 2.

Comparison of Outcomes Between EVT(±IVT) Versus IVT

Outcomes	EVT±bridging IVT (% (n))	IVT (% (n))	Unadjusted OR 95% [CI]	Adjusted OR 95% [CI]	IPTW 95% [CI]
Favorable outcome (mRS score 0–2)	63.6 (112)	64.3 (63)	0.97 [0.58–1.63]	0.89 [0.46–1.72]	1.01 [0.68–1.50]
Excellent outcome (mRS score 0–1)	47.1 (83)	57.1 (56)	0.67 [0.41–1.10]	0.56 [0.30–1.05]	0.63 [0.43–0.92]
Fair outcome * (mRS score 0–3)	75.0 (132)	81.7 (80)	0.68 [0.36–1.25]	0.59 [0.28–1.26]	0.56 [0.35–0.89]
Unfavorable shift in mRS scores	‐	‐	1.48 [1.09–2.29]	1.83 [1.10–3.06]	‐
Mortality (3 mo)	15.9 (28)	6.1 (6)	2.90 [1.16–7.27]	4.55 [1.45–14.26] 3.38 [1.30–8.75] †	4.11 [2.09–8.65]
Symptomatic intracranial hemorrhage (sICH, ECASS‐II criteria)	2.0 (2)	0.0 (0)	‐	‐	‐

ECASSindicates European Cooperative Acute Stroke Study; EVT, endovascular treatment; IVT, intravenous thrombolysis; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; OR, odds ratio; and sICH, symptomatic intracranial hemorrhage.

^* independent ambulation.

Adjusted for age, sex, prior stroke, functional independence prior to stroke, diabetes, onset‐to‐treatment time, atrial fibrillation, additional vessel occlusion, year of treatment after 2014, NIHSS score on admission, location of BAO.

^† Adjusted for age, NIHSS score on admission, and additional vessel occlusion.

Missing: functional independence prior to stroke, n = 2.

In an additional sensitivity analysis, we excluded patients with prestroke anticoagulation (n = 247). The overall results were similar (favorable outcome at 3 months [mRS score 0—2] 64.4% versus 63.5%, adjusted 1.02 [0.53–1.99]; fair outcome at 3 months [mRS score 0—3] 77.7% versus 81.3%, adjusted 0.69 [0.32–1.49]; excellent outcome at 3 months (mRS score 0–1) 52.6% versus 56.3%, adjusted 0.70 [0.38–1.33]; mortality 12.1% versus 6.3%, adjusted OR 3.97 [1.25–12.66], sICH 0.8% versus 0.0%).

DISCUSSION

In this retrospective multicenter observational study of prospectively collected data of patients with acute BAO receiving EVT(±IVT) or IVT alone, we observed the following findings: First, there was no difference in favorable functional outcome (defined as a score on the mRS of ≤2) between patients treated with EVT (with or without prior IVT) and patients treated with IVT alone. Second, with regard to the entire mRS scale, EVT(±IVT) was significantly associated with a shift toward increased disability. Third, mortality after EVT(±IVT) was increased when compared with IVT alone.

EVT in the posterior circulation is often considered riskier due to anatomical factors, clinical vulnerability (involving brain stem functions), technical challenges, or periprocedural risks. ¹⁰ , ¹⁸ This study does not indicate a high risk of periprocedural complications, although the databases lacked information on iatrogenic perforations or dissections. Nonetheless, recent evidence from multiple clinical trials suggests EVT is safe and effective in severely affected patients with BAO. ⁶ , ⁷ , ⁸ , ⁹ At present, data on functional outcomes of BAO patients with mild‐to‐moderate disability are scarce and often limited by small sample sizes. Across the 4 published RCTs investigating the efficacy and safety of EVT involving 988 BAO patients, a patient‐level meta‐analysis (VERITAS [Vertebrobasilar Flow Evaluation and Risk of Transient Ischemic Attack and Stroke]) reported only 97 patients with an NIHSS score <10. ¹⁰ A trial‐level meta‐analysis of these 4 RCTs reports data on clinical outcomes in patients with BAO and an NIHSS score <10 with similar findings and reporting no differences between favorable functional outcomes (mRS score 0–2) with approximately 60% in both EVT and best medical treatment groups. ¹⁹ Additionally, pooled data analysis revealed an overall rate of sICH of 5.4% versus 0.8% and the mortality rate was 36% versus 45%. ¹⁹ In contrast, a propensity score matched analysis from French centers observed a more than 3‐fold higher odds of a favorable outcome in mildly affected patients with BAO treated with bridging EVT versus IVT alone. ²⁰ However, the groups were small (n<30 per group) and limited to very mild deficits NIHSS score <5. ²⁰ A recent pooled analysis of 2 French registries (n = 127) observed numerically higher rates of excellent functional outcome (mRS score 0–1) in patients treated with EVT, but also numerically more patients with mRS score >3. ²¹ A recent cohort study from Helsinki including patients with BAO treated with revascularization treatments within 48 hours of symptom onset, IVT alone was associated with a higher odds of fair functional outcome (mRS score 0–3) compared with EVT and there was no evidence of an interaction between treatment modality and NIHSS score (n = 90 with NIHSS score <10). ²² Given these controversial findings, our large multicenter observational study adds valuable evidence. Based on our results, EVT in patients with BAO and mild‐to‐moderate stroke does not seem to have a benefit when compared with IVT alone. This confirms existing pooled data from RCTs and strengthens the current guideline recommendations. Of note, we observed a 2‐ to 3‐fold higher 90‐day mortality in patients with BAO with NIHSS score <10 who underwent EVT. This observation remained robust after IPTW and was not driven by the occurrence of symptomatic ICH. Existing data confirm similar 90‐day mortality rates in this population, with approximately 17% in the meta‐analysis of RCT data mentioned previously. ¹⁹ Taken together, these findings do not support EVT instead of IVT alone in mildly affected patients with BAO.

Although our findings are consistent with pooled randomized trial data, understanding the potential reasons behind these results is essential for interpreting their clinical relevance. Patients with mild‐to‐moderate BAO may have sufficient collateral circulation to maintain adequate perfusion until spontaneous or IVT‐induced recanalization occurs, reducing the potential incremental benefit of EVT. In addition, patients chosen for EVT despite low NIHSS scores may have had imaging features (eg, poor collateral flow, which were not captured in our data set) indicating a perceived higher risk for deterioration, which could inherently predispose to worse outcomes and may introduce selection bias.

This study comprises a relatively large sample size of this distinct subgroup of patients with BAO with NIHSS score <10, from a well‐established multicenter registry. Limitations include the observational design with the potential of bias by indication. Unfortunately, we lack data on the extent of early ischemic changes (measured by posterior circulation Alberta Stroke Program Early CT [Computed Tomography] Score), which are strong predictors of functional outcome and may have influenced the decision to perform or omit EVT or IVT. Without a randomized study design, radiological data such as posterior circulation Alberta Stroke Program Early CT Score, collaterals, infarct size, topography, and reperfusion success are important covariates for decision‐making, which could not be considered in our study. ²³ The NIHSS is considered inadequate for posterior circulation stroke due to failure to recognize dysphagia or gait disturbance due to trunk or limb ataxia. ²⁴ It is possible that EVT was performed as rescue therapy in some patients. However, the fact that we did not find higher rates of symptom fluctuation and lower rates of depressed conscious state in EVT patients makes this seem unlikely. Furthermore, there was no difference in OTT, location of BAO, and the results remained robust even after several ways of adjustment. Because the registry includes only patients who received EVT and/or IVT, our results are limited to this group. Thus, conclusions regarding the use of EVT versus best medical treatment in patients who did not receive IVT cannot be drawn from our data.

CONCLUSION

In this multicenter analysis of patients with BAO with mild‐to‐moderate stroke severity (NIHSS score <10) did not demonstrate a benefit of EVT(±IVT) compared with IVT alone. Further, in this subgroup with BAO, there was a trend toward higher rates of disability and mortality in patients who received EVT(±IVT). Taken together, our findings underscore the existing clinical equipoise and highlight the urgent need for RCT studies in this patient population.

Sources of Funding

None.

Disclosures

P.R. reports speaker honoraria from Boehringer Ingelheim, Bayer and Pfizer and honoraria for scientific advisory board from Boehringer Ingelheim, and Bayer, outside the submitted work, V.P. reports travel or speaker honoraria from Boehringer Ingelheim; honoraria from scientific advisory board from Medtronic, outside the submitted work, P.S. reports speaker honoraria from Amicus Pharma, outside the submitted work, R.R.L. reports receiving speaker honoraria from IschemaView, Boehringer Ingelheim, Pfizer, Jansen, Biogen, Medtronic and Abott, steering committee honoraria from Bayer and advisory board honoraria from Filterlex, Y.B. reports personal fees from BMS, Pfizer, Medtronic, Amgen, Servier, NovoNordisk, Novartis, Argenx, outside the submitted work, M.R.H. reports grants from SITEM Research Support Funds, Swiss National Science Foundation and Swiss Heart Foundation, all not directly related to this work, A.Z. reports consulting and speaker fees from Angels Initiative, Boehringer‐Ingelheim, Alexion, Daiichi Sankyo, Pfizer, fees for Advisory Board from Daiichi Sankyo, Bayer and Astra Zeneca, PIAM all outside the submitted work. The following authors have no disclosures to report: S.L., S.H., C.R., C.H.N, J.F.S., C.I., I.M., V.L.A., H.G., S.T.E., S.M.T., G.D., R.P., M.Z., L.P., A.P., J.P.M., G.B., G.N., H.J.A., J.K., S.W., C.C., M.M., V.M.F., N.W., F.R., I.I., P.J.N., A.S., C.C., M.T., Z.B., S.S., G.S.

Supporting information

Table S1: Participating Endovascular Treatment and Thrombolysis in Ischemic Stroke Patients (EVA‐TRISP/TRISP) Centers.

Table S2: Sensitivity analysis for pre stroke anticoagulation and outcomes after 3 months.

Table S3: logistic regression analysis after IPTW.

Table S4: Sensitivity analysis for EVT (± Bridging IVT) and favorable outcome (mRS 0‐2) after 3 months.

Supplemental figure S1 Distribution of the inverse probability weights calculated.

Supplemental figure S2 Love plot of Standardized Mean Differences in Covariates before and after inverse probability of treatment weighting (IPTW). Before IPTW red and after IPTW blue.