This multicenter, retrospective cohort study analyzed consecutive patients with AIS due to underlying ICAD-LVO who underwent EVT between June 2022 and December 2024. The study was conducted at two comprehensive stroke centers in China: The First Affiliated Hospital of Army Medical University and Zigong Third People’s Hospital of Sichuan Province.

Patients were included if they were ≥ 18 years of age, had no evidence of intracranial hemorrhage on baseline non-contrast CT or MRI, and had an intracranial LVO confirmed by CTA, MRA, or DSA. Occlusions of the vertebral artery V4 segment were included only in the presence of contralateral vertebral artery hypoplasia or occlusion. The underlying etiology of ICAD-LVO was determined using a positive microcatheter first-pass effect (15) (Figure 1). Additional inclusion criteria consisted of a baseline National Institutes of Health Stroke Scale [NIHSS] score ≥6, indicating moderate to severe deficits, and an onset-to-puncture time within 24 h. Key exclusion criteria covered procedural termination before implementation of the initial approach, pre-stroke functional dependence (defined as mRS > 1), and the presence of ≥70% stenosis or occlusion proximal to the occluded vessel suggestive of an artery-to-artery embolism mechanism. We also excluded patients with refractory hypertension (>185/110 mmHg despite medication), severe dysglycemia (blood glucose <2.7 or >22.2 mmol/L), coagulopathy (platelet count <40 × 10⁹/L, aPTT >50 s, or INR > 3), and those with incomplete clinical or angiographic data. These criteria were applied to minimize confounding related to comorbidities, technical variability, and data reliability issues. The final analysis included 161 patients with ICAD-LVO, who were classified into two groups according to the initial endovascular strategy: the AS group and the MT group (Figure 2).

The study protocol received ethical approval from the institutional review boards of both participating hospitals (Approval No. B-KY2025079). Informed consent was waived due to the retrospective design and use of anonymized clinical data. All procedures were conducted in accordance with the ethical principles of the Declaration of Helsinki.

Diagnosis of underlying ICAD-LVO in this study was primarily determined through preprocedural clinical data, preprocedural CTA, and intraprocedural DSA imaging. First, suspicion for ICAD-LVO arose if patients had preexisting atherosclerotic risk factors (hypertension; diabetes; smoking) (16, 17) or exhibited trunk occlusion rather than branch occlusion (18). Operators subsequently verified the diagnosis intraprocedurally using the microcatheter first-pass effect (15). To evaluate this effect, the microcatheter was advanced with the microguidewire through the occlusion into the distal patent segment and then withdrawn to a position proximal to the occlusion while the microguidewire remained in place, after which angiography was performed via an intermediate catheter. A positive result was defined as the angiographic demonstration of both transient antegrade flow and an underlying severe stenosis at the occlusion site, a finding indicative of an underlying ICAD-LVO. The absence of flow (persistent occlusion) was considered a negative result, suggestive of an embolic etiology (Figure 1). Diagnosis was made intraoperatively by the operator and prospectively recorded in the database. Subsequently, another neurointerventionalist reviewed the diagnosis. Discrepancies were resolved through unanimous consensus.

All endovascular procedures were performed under general anesthesia via a femoral artery approach. All operators had ≥10 years of experience in acute stroke neurointervention, and the initial strategies were selected based on individual expertise. All endovascular devices were approved by China’s National Medical Products Administration for intracranial interventional procedures. All patients followed a standardized heparinization protocol: a 3,000-unit intravenous bolus was given immediately after femoral artery puncture, followed by a maintenance dose of 1,000 units per hour. No heparin was administered to patients who had received preoperative intravenous thrombolysis.

Postprocedural management adhered to standardized stroke unit protocols: systolic blood pressure maintained <140 mmHg, serial neurological assessments (every 60 min for 24 h), and individualized secondary prevention. A uniform antiplatelet regimen was administered: continuous intravenous tirofiban infusion for 24 h, transitioning to oral aspirin (100 mg/day) plus clopidogrel (75 mg/day) for 3 months. Intravenous tirofiban was generally initiated immediately after angioplasty or mechanical thrombectomy at a maintenance infusion rate of 0.2–0.25 mg/h. In cases of recurrent occlusion attributed to endothelial injury, an additional intracatheter (intra-arterial) bolus of 0.25–0.5 mg could be administered locally. For patients who had undergone prior intravenous thrombolysis or exhibited contrast extravasation on post-EVT imaging, antiplatelet agents were adjusted—either discontinued or reduced—following multidisciplinary consensus on hemorrhagic transformation risk.

Functional independence at 90 days, defined as mRS scores 0 to 2, served as the primary endpoint. Neurologists prospectively collected outcome data through structured electronic health record abstraction and validated telephone-based mRS assessments. To ensure methodological rigor and reliability, all assessments were performed by neurologists with specialized expertise in stroke outcomes.

Secondary outcomes included successful recanalization (defined as modified Thrombolysis in Cerebral Infarction [mTICI] grade 2b–3) (19), symptomatic intracranial hemorrhage (SICH) within 48 h (classified according to the Heidelberg bleeding classification) (20), and 90-day all-cause mortality.

Data analyses were conducted using SPSS Statistics (version 29.0; IBM). Baseline characteristics were summarized as proportions (%) for categorical variables, and as mean ± standard deviation (SD) or median (interquartile range [IQR]) for continuous variables, based on normality assessed by Shapiro–Wilk tests. Group comparisons utilized chi-square or Fisher’s exact tests for categorical variables, and independent t-tests or Mann–Whitney U tests for continuous variables, as appropriate. Logistic regression models were applied to evaluate associations between first-line recanalization strategies and clinical outcomes. Variables with p < 0.10 in univariate analyses alongside prespecified clinical covariates (age, diabetes, baseline NIHSS) were included in adjusted models. Results are presented as crude and adjusted odds ratios (ORs) with 95% confidence intervals (CIs). Multicollinearity was verified using variance inflation factors (VIF < 2.0), and model calibration was confirmed via Hosmer-Lemeshow tests (p > 0.05). Statistical significance was defined as a two-tailed p < 0.05.

Among 614 consecutive patients undergoing emergency EVT for LVO, 37.6% (231/614) were diagnosed with underlying ICAD-LVO. After applying the inclusion and exclusion criteria, the final analysis included 161 patients (107 males and 54 females) with a mean age of 67 years (range, 34–96 years), as detailed in Figure 2. Baseline demographic and clinical characteristics are summarized in Table 1.

Most baseline characteristics were well balanced between the two groups, including age, sex, occlusion sites, wake-up stroke, intravenous thrombolysis, baseline NIHSS and Alberta Stroke Program Early CT Score [ASPECTS], and procedural time metrics (all p > 0.05). However, the AS group had significantly higher rates of hyperlipidemia (48.9% vs. 32.8%, p = 0.041), progressive stroke onset (51.1% vs. 23.9%, p < 0.001), and tirofiban administration (92.6% vs. 82.1%, p = 0.043).

The technical details of the initial and rescue interventions are presented in Table 2. In the AS group, balloon angioplasty alone was performed in 33 patients (35.1%), while 61 patients (64.9%) received adjunctive stent implantation. In the MT group, initial strategy was aspiration thrombectomy in 24 patients (35.8%) and a combined approach (aspiration with stent retriever) in 43 patients (64.2%).

Successful initial recanalization without rescue therapy was achieved in 69 patients (73.4%) in the AS group compared to 25 patients (37.3%) in the MT group. Rescue interventions were initiated in 23 patients (24.5%) from the AS group and 37 patients (55.2%) from the MT group. Among patients in the AS group requiring rescue, subsequent strategies included stent-retriever/aspiration after balloon (78.3%, 18/23), aspiration after stenting (4.3%, 1/23), and balloon re-dilation after stenting (17.4%, 4/23). All rescue cases in the MT group received balloon or stent angioplasty (100%, 37/37). Successful rescue recanalization was observed in 22 patients (95.7%) in the AS group and 36 patients (97.3%) in the MT group.

Clinical outcomes are summarized in Table 3. In unadjusted analysis, the AS group demonstrated a higher proportion of functional independence compared to the MT group (63.8% vs. 47.8%, OR = 1.930, 95% CI: 1.020–3.653, p = 0.043). The multivariate logistic regression model was adjusted for covariates identified in univariate analyses (baseline NIHSS score, stroke-to-puncture time, and lesion location; threshold: p < 0.1) and prespecified clinically relevant confounders, including age, baseline NIHSS score, tirofiban administration, and diabetes mellitus. After adjustment, this association remained statistically significant (adjusted OR = 2.886, 95% CI: 1.290–6.736, p = 0.011).

No statistically significant differences were observed across secondary outcomes between the AS and MT groups. Recanalization rates were comparable (96.8% vs. 91.0%, p = 0.132), with no association identified after adjustment (adjusted OR = 2.958, 95% CI: 0.703–15.449, p = 0.155). Similarly, SICH within 48 h (13.8% vs. 13.4%, p = 0.942) and all-cause mortality at 90 days (14.9% vs. 13.4%, p = 0.794) remained comparable between groups, with adjusted OR of 1.009 (95% CI: 0.391–2.702, p = 0.985) and 1.151 (95% CI: 0.425–3.273, p = 0.784).