This single-center retrospective study enrolled consecutive patients with AIS who underwent EVT at the Chengdu Second People’s Hospital between January 2023 and May 2025. The inclusion criteria were as follows: (1) aged ≥ 18 years; (2) diagnosis of AIS due to LVO; (3) receipt of reperfusion therapy, including EVT and/or intravenous thrombolysis (IVT); and (4) availability of high-quality imaging without significant motion artifacts. Patients were excluded if: (1) DSA image was incomplete or unavailable; (2) thrombectomy was unsuccessful or performed as a non-emergency procedure. For patients presenting beyond the conventional treatment window, multimodal magnetic resonance imaging (MRI) or computed tomography (CT) perfusion imaging was used to assess eligibility for reperfusion therapy. In accordance with current guidelines, IVT was administered within 4.5 h of symptom onset or last known well time, and EVT was performed within 24 h (20).

Clinical and demographic data were extracted from the stroke center database, including age, sex, and vascular risk factors such as hypertension, diabetes mellitus, hyperlipidemia, coronary artery disease, atrial fibrillation, transient ischemic attacks (TIA), and prior ischemic stroke. Lifestyle factors, including smoking and alcohol consumption, were also recorded. Clinical assessments included admission blood pressure and initial stroke severity, which was evaluated using the National Institutes of Health Stroke Scale (NIHSS) and the modified Rankin Scale (mRS). Both scales were assessed before reperfusion therapy and were defined as the baseline NIHSS and baseline mRS, respectively. Time metrics included onset-to recanalization time and door-to-puncture time (DPT). The site of vessel occlusion and type of reperfusion were documented for each patient. Stroke etiology was determined according to the modified Trial of ORG 10172 in Acute Stroke Treatment (TOAST) classification.

Treatment strategies, including EVT alone or combined EVT and IVT, were determined by the treating clinicians based on individual clinical and imaging profiles. Patients who presented beyond the standard therapeutic time window were considered for reperfusion therapy if multimodal imaging demonstrated a favorable perfusion-diffusion mismatch, in accordance with the 2018 Chinese Guidelines for the Diagnosis and Treatment of Acute Ischemic Stroke. Informed consent was obtained from all patients or their legal representatives prior to treatment, with comprehensive explanation of potential risks, including hemorrhagic complications and financial implications.

DSA was first performed on the contralateral normal side before thrombectomy to identify the occlusion site and evaluate collateral circulation through the circle of Willis, thereby providing a baseline for endovascular intervention. All examinations were conducted using a biplane angiography system (UNIQ Clarity FD20/20, Philips, The Netherlands) equipped with a power injector (8 mL ioversol, injection rate of 6 mL/s, pressure of 200 psi/kg), following a standardized protocol at Chengdu Second People’s Hospital, China. Two experienced neurointerventionalists, each with 10–20 years of clinical experience, independently reviewed all pre-EVT DSA images while blinded to the patients’ clinical data. Collateral circulation was graded according to the American Society of Interventional and Therapeutic Neuroradiology/Society of Interventional Radiology (ASITN/SIR) scale as follows: grade 0: no visible collaterals to the ischemic site; grade 1: slow collaterals to the periphery of the ischemic area with persistent defects; grade 2: rapid collaterals reaching the ischemic boundary with partial defects; grade 3: slow but complete collateral filling of the ischemic bed by the late venous phase; and grade 4: rapid and complete collateral filling of the ischemic bed via retrograde perfusion (21). Posterior circulation collateral status was assessed using the posterior circulation collateral score (PC-CS), which ranges from 0 to 10. For statistical analysis, PC-CS scores were categorized into equivalent ASITN/SIR grades as follows: 0 as grade 0; 1–3 as grade 1; 4–5 as grade 2; 6–8 as grade 3; and 9–10 as grade 4. This categorization represents an equivalent stratification for analytical purposes rather than a direct one-to-one correspondence between the PC-CS and ASITN/SIR grading systems. For further analysis, collateral status was grouped into three categories: poor (grades 0–1), fair (grade 2), and good (grades 3–4). In cases of disagreement between the two neurointerventionalists, a consensus was reached through joint review. Excellent inter-rater reliability was confirmed on a training dataset prior to independent assessments (intraclass correlation coefficient > 0.90). Functional outcomes at 90 days were assessed using the mRS through face-to-face visits or telephone interviews with patients or family members. An mRS score of 0–2 was defined as a good functional outcome.

The reperfusion after EVT was assessed by an experienced neurointerventionist with over 10 years of experience, using the modified Treatment in Cerebral Infarction scale (mTICI, score 0/1/2a/2b/3), and the mTICI 0/1/2a was defined as unsuccessful reperfusion, the mTICI 2b/3 as successful reperfusion (22), indicating restoration of antegrade perfusion in more than 50% of the affected vascular territory.

The study was approved by the Ethics Committee of Chengdu Second People’s Hospital. Written informed consent was obtained from all patients or their legal representatives, confirming their agreement to underdo EVT with or without IVT. All procedures were conducted in accordance with the ethical standards of the institutional research committee and the principles of the Declaration of Helsinki.

The primary outcome has the functional status at 90 days after stroke, assessed by the mRS. Patients were categorized into two groups: good outcome (mRS ≤ 2) and poor outcome (mRS > 2) groups. The secondary outcome was the occurrence of early neurological deterioration (END) within 7 days after reperfusion treatment, defined as an increase of ≥4 points in the NIHSS score or a ≥1 point decrease in the level of consciousness subscore within 7 days post-recanalization (23). Safety outcomes included any intracranial hemorrhage (ICH) and all-cause mortality rate within 90 days.

The normality of continuous variables was assessed using the Kolmogorov–Smirnov test. Normally distributed data were expressed as mean ± standard deviation (SD), whereas non-normally distributed data were reported as median [interquartile range (IQR)]. Categorical variables were compared among collateral grade groups using the chi-square test or Fisher’s exact test when appropriate. Continuous variables were compared using one-way ANOVA or the Kruskal-Wallis test, depending on data distribution. When a significant overall difference was observed across collateral grades, pairwise post hoc comparisons were conducted using the Wilcoxon rank-sum test. Variables with p < 0.05 in univariable analyses were entered into a multivariable logistic regression model (backward stepwise method) to identify independent predictors of favorable outcomes. All statistical tests were two-sided, and p < 0.05 was considered statistically significant. Data analyses were performed using SPSS version 23.0 (IBM, Chicago, IL, USA), and graphical outputs were generated using GraphPad Prism version 9.5.1 (GraphPad Software, San Diego, CA, USA).

A total of 210 patients with AIS who underwent EVT were initially enrolled. After excluding 18 patients with missing or incomplete DSA images, 3 patients with failed thrombectomy, and 4 patients who underwent non-emergency thrombectomy, 185 patients (87 males and 98 females) were included in the final analysis (Figure 1). As shown in Table 1, patients were categorized into three groups based on DSA collateral grades: poor (n = 68, 36.7%), fair (n = 61, 33.0%), and good (n = 56, 30.3%). Significant differences were observed among the three groups for age (78 [69.25, 83.75] vs. 72 [65, 81] vs. 71.5 [65, 80] years, p = 0.016), baseline NIHSS (17 [13, 21] vs. 14 [8.5, 16] vs. 9 [5, 15.75], p < 0.001), baseline mRS (5 [4, 5] vs. 4 [4, 5] vs. 4 [4, 4.75], p < 0.001), DPT (84 [70.25, 135.25] vs. 125 [90.5, 187.0] vs. 145 [112.25, 182.25] minutes, p < 0.001), and TOAST classification (p = 0.010). However, there were no significant differences among the three groups in terms of sex, vascular risk factors (including smoking, drinking, hypertension, diabetes, hyperlipidemia, coronary artery disease, atrial fibrillation, prior ischemic stroke, and TIA), systolic and diastolic blood pressure, occlusion site, reperfusion type, onset-to-reperfusion time, or rate of successful reperfusion (all p > 0.05).

At 90 days after reperfusion therapy, patients with good collaterals had significantly lower median mRS score (1 [0, 3]) compared with those with fair (3 [0, 4]) or poor collaterals (4 [4, 6]) (p < 0.001). The proportion of patients achieving functional independence (mRS 0–2) at 90 days also differed significantly among the good, fair, and poor collateral groups (66.1, 45.9, and 13.2%, respectively; p < 0.001). Conversely, the incidence of 7-day END decreased progressively from poor to good collaterals (23.5, 8.2, and 1.8%; p = 0.001), while the mortality rate at 90 days showed a similar trend (27.9, 11.5, and 1.8%, p < 0.001). However, no significant difference was found in the rate of any ICH among the three groups (19.1, 6.6, and 12.5%, p = 0.106) (Table 1). The distribution of 90-day mRS scores across the three collateral groups in illustrated in Figure 2.

Among the 185 patients who underwent EVT, 74 patients (40%) achieved a good functional outcome (mRS ≤ 2) at 90 days, while 111 patients (60%) had a poor outcome (mRS > 2). Multivariable logistic regression analysis identified several independent predictors of poor outcome: the presence of diabetes (OR 0.220; 95% CI 0.076–0.635, p = 0.005), higher baseline NIHSS score (OR 1.215; 95% CI 1.095–1.348, p < 0.001), lower collateral score (OR 0.512; 95% CI 0.324–0.811, p = 0.004), occlusion location (OR 0.061; 95% CI 0.012–0.304, p = 0.001), and reperfusion type (OR 5.401; 95% CI 2.181–13.374, p < 0.001) (Table 2).

Among the 185 patients undergoing EVT, 22 (11.9%) developed END with 7 days, whereas 163 (88.1%) did not. Multivariate logistic regression revealed that collateral score (OR 0.407; 95% CI 0.232–0.714, p = 0.002) and onset-to-reperfusion time (OR 1.001; 95% CI 1.000–1.003, p = 0.021) were independent predictors of 7-day END (Table 3). At 90 days, 27 patients (14.6%) had died, while 158 (85.4%) were alive. Collateral score (OR 0.443; 95% CI 0.251–0.783, p = 0.005) and reperfusion modality (OR 4.122; 95% CI 1.233–13.778, p = 0.021) were identified as independent factors of 90-day death (Table 4).