INTRODUCTION
Currently, endovascular thrombectomy (EVT) is the standard of care for treating acute ischemic stroke (AIS) due to large vessel occlusion [1]. The current American Stroke Association guidelines recommend EVT for patients with AIS who present within 24 hours of symptom onset with a National Institutes of Health Stroke Scale (NIHSS) score ≥6 [2]. Despite the relatively significant effect of EVT, approximately 50% of patients who have successful recanalization fail to attain functional independence at 90 days [3]. Notably, several parameters, including onset-to-door time, door-to-computed tomography scan time, door-to-puncture time, and recanalization status, have been recognized as critical factors associated with clinical outcomes and serve as performance measures to improve the quality of care [4]; however, procedure time (PT) is relatively unexplored. PT, defined as the time between groin access and vessel recanalization, is a recently identified predictor of poor outcomes after EVT [5-8]. PT within 60 minutes, known as the “golden hour,” has been associated with improved outcomes after EVT [6,9]. However, large population studies based on real-world data on the association between PT and AIS outcomes are still lacking, particularly regarding associated factors of prolonged PT. This makes it difficult to answer the question of what we can do to optimize PT and how long an intervention should last for the best clinical efficacy.
Therefore, we performed a retrospective analysis based on a nationwide representative acute stroke registry to determine the association between PT and post-thrombectomy vascular status and clinical outcomes. Furthermore, we examined factors significantly associated with prolonged PT.
MATERIALS AND METHODS
Patient Population
Study participants were selected from a prospectively maintained Korean Stroke Registry (KSR) database, including patients who received EVT for AIS between January 2018 and December 2022 at 50 stroke centers [10]. Patients aged >18 years with anterior and posterior circulation stroke (PCS) were included. Individuals with disability (modified Rankin Scale [mRS] score >1) before index AIS were excluded (Supplementary Fig. 1). We included patients who underwent EVT using modern thrombectomy devices such as a stent retriever, direct aspiration technique, or a combination, which were used in previous randomized controlled trials [11-13]. Intravenous thrombolysis (IVT) was used in accordance with current guidelines and independent of the decision to have EVT.
Data Collection
Data collected included demographic variables, NIHSS scores on admission, use of IVT, pre-stroke mRS scores, and functional outcomes at discharge as represented by mRS scores. Stroke location (anterior or posterior circulation) and the grade of angiographic revascularization classified based on the modified Thrombolysis in Cerebral Infarction (TICI) score were determined by the neurointerventionist performing the procedure [14]. Additionally, the KSR collected information on the occurrence of symptomatic intracranial hemorrhage (sICH) after EVT in 2020 and beyond. We collected information on the occurrence of sICH after EVT in 29 centers that performed brain imaging within 3 days of EVT and registered information on the occurrence of sICH.
The onset-to-door time was computed from the onset time of stroke symptoms (first abnormal time: the time when the symptoms were first noticed by the patient or observed by a witness) to the time of arrival at the hospital, as documented in the KSR. Door-to-puncture time was calculated as the time from when the patient presented to the hospital until the thrombectomy was started. PT was defined as the time from groin access to either successful recanalization (TICI score ≥2b) or termination of the procedure if the procedure was stopped before ≥TICI 2b was achieved. We also checked whether the patient visited on a working day or on a weekend or public holiday, and whether they visited the hospital during working hours (9 AM–6 PM) or during night duty hours (6 PM–9 AM).
Statistical Methods
Demographics and clinical features were compared between participants with PT <60 minutes and those with PT ≥60 minutes. For continuous measures, means and standard deviations, or medians with interquartile ranges are presented, and P-values were calculated with independent samples t-test or Wilcoxon rank sum test. Frequencies and percentages are presented for categorical measures, and P-values were calculated with a chi-square test or Fisher’s exact test. Univariate and multivariate analysis used a logistic regression model with functional stroke outcomes at discharge, successful recanalization, and PT ≥60 minutes as the dependent variables. We included those reported in previous studies or clinically relevant to each dependent variable as independent variables [15-17]. Good functional outcome was defined as mRS score of 0–2 at discharge. For good functional outcome and successful recanalization, a multivariate logistic regression model was developed with age, sex, initial stroke severity, PT <60 minutes, vascular risk factors (history of previous stroke or transient ischemic attack, coronary artery disease, hypertension, diabetes, dyslipidemia, smoking, and atrial fibrillation), previous medication use (anticoagulants, antiplatelets, and statins), use of IVT, onset type of stroke (clear onset or not), level of the hospital (tertiary referral hospital or others), location of the stroke, and admission time at the hospital (working days or weekends, working hours or not) as independent variables. We also examined the association between PT <60 minutes and functional outcome in the group of patients with complete reperfusion (TICI 3) after EVT and with PCS using multivariate logistic regression analysis.
For PT delays greater than 60 minutes, we included age, sex, stroke severity, time to visit after onset, time to puncture after visit, IVT use, vascular risk factors, previous medication use, stroke onset type, level of the hospital, location of stroke, and admission time at the hospital as confounding factors. Factors identified as P-value <0.10 in the univariate analysis were included as covariates in the multivariate analysis. Furthermore, we performed a 3-knot (knots=50th, 75th, 95th percentiles) restricted cubic spline analysis to evaluate the nonlinear association and logistic regression to explore the optimal cutoff point in the relationship between the PT and good functional outcomes.
All statistical analyses were performed using SAS statistical software (version 9.4; SAS Institute) and R statistical software (version 4.1.2; R Foundation for Statistical Computing). Statistical significance was set at a 2-sided P-value <0.05.
RESULTS
Baseline Characteristics
We analyzed 4,703 patients (mean age: 69.5±11.9, 60.3% male) who underwent EVT. The mean PT was 54.6±36.7 minutes. Patients with PT <60 minutes were younger (mean age: 69.2±12.0 vs. 70.1±11.5 years, P=0.017), had clear-onset stroke more often (57.1% vs. 50.3%, P<0.001), and had shorter onset-to- door time (138.9±139.9 vs. 170.9±168.4 minutes, P<0.001) and door-to puncture time (98.7±38.0 vs. 107.0±41.9 minutes, P<0.001) than those with PT ≥60 minutes. However, regarding risk factors, the PT <60 minutes group was significantly less likely to have a history of smoking (33.2% vs. 38.0%, P=0.001) and more likely to have atrial fibrillation (49.7% vs. 46.2%, P=0.025). Furthermore, the history of anticoagulation (13.5% vs. 11.2%, P=0.030) and use of IVT (47.3% vs. 42.8%, P=0.004) was significantly higher in the PT <60 minutes group. Finally, there were significantly more PT ≥60 minutes cases among patients with PCS than those with anterior circulation stroke (Table 1).
Clinical and Angiographic Outcomes
When assessed for functional status at discharge, the PT <60 minutes group had significantly better outcomes compared with the PT ≥60 minutes group in all ranges (mRS: 0–1 [28.0% vs. 19.8%, P<0.001], mRS: 0–2 [45.1% vs. 36.4%, P<0.001], and mRS: 0–3 [61.9% vs. 52.9%, P<0.001]). There was no significant difference in death during hospitalization between either group. However, bedridden status at discharge (mRS: 5–6) was significantly higher in the PT ≥60 minutes group than in the PT <60 minutes group (20.9% vs. 26.0%, P<0.001). When TICI grade ≥2b was defined as the criterion for successful recanalization, it was significantly achieved in the PT <60 minutes group (94.0% vs. 90.2%, P<0.001). Moreover, the incidence of sICH after EVT was significantly lower in the PT <60 minutes group than in the PT ≥60 minutes group (2.7% vs. 9.6%, P<0.001) (Table 2, Supplementary Fig. 2).
Association of Procedure Time and Functional Outcomes
To evaluate whether PT <60 minutes independently predicts good functional outcomes (mRS: 0–2) in AIS patients, we performed multivariate logistic regression, adjusting for established prognostic factors. The results confirmed that PT <60 minutes was significantly associated with good functional outcomes (adjusted odds ratio [aOR]: 1.40, 95% confidence interval [CI]: 1.22–1.59). Multivariate analysis also identified age ≥65 years, initial NIHSS score, use of IVT, diabetes mellitus, clear-onset stroke, and tertiary referral hospital as factors significantly associated with stroke outcome in addition to PT (Table 3). In the other ranges, PT <60 minutes remained significantly associated with favorable outcomes after multivariate analysis (mRS: 0–1, aOR: 1.54, 95% CI: 1.32–1.79; mRS: 0–3, aOR: 1.34, 95%CI: 1.17–1.55). In contrast, PT ≥60 minutes independently predicted bedridden status after EVT (mRS: 5–6, aOR: 1.30, 95% CI: 1.12–1.50) (Supplementary Tables 1–3). The restricted spline curve shows the timewise ORs for achieving good functional outcomes in terms of PT. The result suggested that there was a nonlinear relationship between the PT and good functional outcome (P=0.046). Especially, aOR decreased sharply until PT 60 minutes and later slightly declined (Fig. 1A). In addition, the probability of achieving good functional outcomes was increased with a shorter PT (Fig. 1B).
Association of Procedure Time and Functional Outcomes in Patients Who Achieved Complete Reperfusion (TICI 3) after Endovascular Thrombectomy
We performed a multivariate logistic regression analysis on the group of patients who had fully reperfused after EVT (TICI 3 group). Consequently, PT <60 minutes was found to be significantly associated with good functional outcomes (aOR: 1.49, 95%CI: 1.23–1.79). In addition, PT <60 minutes was found to be significantly associated with neurological function at discharge, with mRS 0–1 and mRS 0–3 (mRS: 0–1, aOR: 1.68, 95% CI: 1.36–2.06; mRS: 0–3, aOR: 1.50, 95% CI: 1.24–1.82) (Supplementary Tables 4–6).
Association of Procedure Time and Functional Outcomes in Patients with Posterior Circulation Stroke
Furthermore, a multivariate logistic regression analysis was conducted, with the study population consisting of patients with PCS. The analysis revealed that PT <60 minutes was significantly associated with good functional outcomes (aOR: 1.76, 95% CI: 1.17–2.67). Additionally, PT<60 minutes was found to be significantly associated with mRS 0–1 (aOR: 1.78, 95% CI: 1.11–2.83) and mRS 0–3 (aOR: 1.47, 95% CI: 1.03–2.18) functional status at discharge (Supplementary Tables 7–9).
Association of Procedure Time with Successful Recanalization
To determine the association between PT and successful recanalization of the affected vessel, we performed a multivariate analysis with factors associated with recanalization reported in previous studies as covariates and successful post-thrombectomy reperfusion status represented by TICI grade ≥2b as the dependent variable. The results showed that PT <60 minutes was significantly associated with successful recanalization, even after adjusting for confounding variables (aOR: 1.66, 95% CI: 1.33–2.07) (Supplementary Table 10).
Factors Associated with Delayed Procedure Time ≥60 Minutes
We conducted a multivariate analysis with PT ≥60 minutes as the dependent variable to identify factors associated with prolonged PT. After adjusting for confounding variables, age ≥65 years (aOR: 1.20, 95% CI: 1.05–1.38), onset-to-door time (aOR: 1.03, 95% CI: 1.01–1.04), door-to-puncture time (aOR: 1.05, 95% CI: 1.03–1.06), PCS (aOR: 1.13, 95% CI: 1.02–1.28), and smoking (aOR: 1.24, 95% CI: 1.09–1.45) independently predicted a delay in PT ≥60 minutes. However, clear-onset stroke (aOR: 0.75, 95% CI: 0.66–0.86) was significantly and negatively associated with PT ≥60 minutes (Table 4).
Cutoff Point for the Procedure Time
Finally, to decide and confirm the cutoff point for the PT from puncture to reperfusion, 5 cutoffs at 15-minute intervals were selected (30, 45, 60, 75, and 90 minutes). A multivariate logistic regression model was developed for each cutoff time with a good stroke outcome of mRS 0–2 at discharge as the dependent variable. The highest aOR was observed in the 60-minute cutoff model (aOR: 1.45, 95% CI: 1.27–1.67, P<0.001) (Supplementary Table 11).
DISCUSSION
Using a nationwide representative stroke registry, this study showed that PT was associated with successful recanalization of the occluded vessel and clinical outcomes. Furthermore, older age, prolonged onset-to-door and door-to-puncture time, smoking, PCS, and mode of stroke onset (clear-onset stroke) were identified as factors associated with delays in PT ≥60 minutes, which has been reported as the “golden hour” in previous studies. Finally, our study identified 60 minutes as the cutoff time most strongly associated with functional stroke outcomes.
In previous studies, PT has been reported to be associated with functional outcomes in clinical trial populations and real- world data [6,9,18]. Similarly, in our study, PT was significantly associated with functional outcomes in patients with AIS, even after adjusting for patient baseline characteristics and comorbidities as covariates. Based on the KSR (2018–2022), our analysis reaffirms this relationship and provides new insights using a large East Asian cohort. This setting allows evaluation of the generalizability of prior findings in a different demographic context, where stroke characteristics and care systems may differ (Table 5) [5,6].
Our data support the role of PT as a predictor of functional recovery, as well as a reflection of procedural complexity for the refractory occlusion in response to mechanical thrombectomy. Even in patients with complete reperfusion (TICI 3), PT <60 minutes remained a significant predictor of good outcomes. This suggests that PT contributes to neurological recovery—likely by minimizing ischemic burden, avoiding vascular injury from repeated attempts, and enabling earlier post-procedural care [9]. Additionally, in patients with PCS, PT <60 minutes was significantly associated with better outcomes despite anatomical challenges, underscoring its value as a potentially modifiable factor. In contrast, sICH was more frequently observed in the PT ≥60 minutes group, suggesting that prolonged PT may increase the risk of post-EVT hemorrhagic complications. This finding aligns with previous studies indicating that longer PT may be associated with higher rates of vascular injury, endothelial damage, and microvascular reperfusion injury [6,8].
Furthermore, as illustrated by the restricted spline curve (Fig. 1A), shorter PT is consistently linked to better outcomes, reinforcing the need to minimize PT when feasible. However, PT reduction should not come at the expense of procedural safety. Optimization efforts should include streamlining workflow, selecting appropriate devices, and making rapid clinical decisions. As a standardized and intuitive metric, PT enables comparisons across institutions and supports its role as a performance indicator in stroke care.
We identified factors associated with delayed PT ≥60 minutes. Notably, the multivariate analysis recognized onset-to-door time and door-to-puncture time after adjusting for confounding variables. These 2 time-related factors have been reported in previous studies as independent prognostic predictors after EVT in patients with AIS [19-22]. The earlier a patient presents to the hospital after the onset of AIS, the sooner the EVT is performed, the better the patient’s prognosis. Similarly, it can be inferred that a quick thrombectomy after a blockage is associated with successful recanalization and shorter PT. While previous studies have suggested that prolonged onset-to-admission and door-to-puncture times could alter thrombus composition, increasing its fibrin content and adherence to the vessel wall [23-25], this hypothesis remains uncertain due to limited direct evidence. Alternatively, differences in the etiology of strokes may be a contributing factor to variations in prehospital delay and PT duration. For instance, in cases of underlying intracranial atherosclerosis- related occlusion, patients may experience initial symptoms of a milder nature, which may result in delayed hospital arrival and delayed clinical decision-making, consequently leading to prolonged PT. Additionally, underlying atherosclerotic stenosis may necessitate multiple EVT attempts due to re-occlusion, which could contribute to an extended PT duration. Further studies incorporating detailed stroke etiology data and thrombus characteristics are needed.
Patient-specific, non-modifiable factors associated with a delay in PT ≥60 minutes may be considered when deciding when to stop EVT. Older age and smoking are associated with prolonged PT due to increased vascular tortuosity and atherosclerosis burden [26-28]. The longer the PT, the more the thrombus becomes entrapped in the vessel, the more vascular micro-injuries occur, and the higher the incidence of complications such as intracerebral hemorrhage [29]. Therefore, neurointerventionists need to decide whether to stop the procedure after a few attempts or after a certain amount of time if successful recanalization is challenging. These factors must be considered in conjunction with other complex clinical circumstances when deciding to continue or discontinue a procedure.
Notably, several possible explanations exist for the significant delay in PT in PCS. It has been reported that PCS is often subject to more futile recanalization than anterior circulation [30,31]. In PCS, the thrombectomy process can be prolonged due to anatomical features, including the small diameter of the vertebral artery [32]. Previous studies have also shown that PT is prolonged in patients with PCS, especially in cases of intracranial artery stenosis with atherosclerosis [33]. In addition, PCSs commonly present with vague symptoms, such as dizziness, which can delay detection and diagnosis [34]. This may cause late hospital arrival or in-hospital decision, which may delay PT for more than 60 minutes due to the late beginning of EVT. In fact, the onset-to-door time was significantly prolonged in PCS compared with anterior circulation stroke in this study’s patient population (Supplementary Table 12). However, previous studies have shown that rapid PT can be achieved in the posterior circulation in patients with embolic strokes such as cardioembolism [33]. Hence, a tailored EVT strategy—one that considers stroke mechanism and vascular anatomy—is essential for optimizing PT and outcomes in PCS.
This study has some limitations. First, details regarding the thrombectomy process, such as first-pass recanalization and techniques, are lacking. Second, information on stroke mechanisms was unavailable, although patients with a history of atrial fibrillation had a shorter PT. Despite these limitations, the strength of our analysis lies in its scale and real-world relevance: a large, nationally representative multicenter cohort reflecting current thrombectomy practices in the post-2018 guideline era [35].
CONCLUSION
This study confirms that PT <60 minutes was independently associated with successful recanalization and positive functional outcomes after EVT. Identifying and addressing factors that delay PT should be a priority in optimizing EVT care. Continuous monitoring of PT as a performance metric, combined with strategies for early hospital arrival and individualized procedural planning, especially in elderly and high-risk patients, will be crucial in enhancing stroke care outcomes.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement
Print
