Surpass evolve flow diverter in intracranial aneurysms: results from a single-center cohort with extended follow-up

Abstract

Background

The Surpass Evolve (SE) is a second-generation flow diverter (FD) designed to improve procedural performance compared with earlier devices. This study aimed to evaluate the angiographic efficacy, safety, and clinical outcomes of SE in a single-center cohort with extended follow-up, comparing it to earlier-generation FDs.

Methods

We conducted a single-center comparative cohort study including 120 patients with unruptured intracranial aneurysms treated between 2014 and 2023. The prospective SE cohort (n = 67) was compared to a retrospective control cohort (n = 53; Surpass Streamline [SS] and Pipeline Embolization Device [PED]). Primary outcomes were complete aneurysm occlusion, major stroke, in-stent thrombosis, aneurysm enlargement, and unfavorable functional outcome. Secondary analyses included a composite safety endpoint and procedural parameters. Propensity score matching (PSM) was additionally performed to adjust for baseline imbalances, while multivariate Cox regression was used in the entire cohort to maximize statistical power.

Results

Complete occlusion was achieved in 79.1% of the SE cohort, with major stroke and in-stent thrombosis occurring in 1.5% each, and aneurysm enlargement in 3.0%. At last follow-up, 98.5% of SE patients had favorable functional outcomes (mRS 0–2). Compared with controls, the SE cohort had significantly fewer major strokes (1.5% vs. 13.2%, p = 0.029) and lower aneurysm enlargement (3.0% vs. 22.6%, p = 0.002). These findings remained consistent in the PSM-matched cohort, particularly regarding the prevention of aneurysm enlargement (p = 0.012). Procedural time was shorter (p = 0.001), and balloon angioplasty was less frequent (p = 0.023). SE use was independently associated with reduced aneurysm enlargement (adjusted OR, 0.14; p = 0.019). Cox analysis showed SS had lower occlusion probability (HR, 0.58; p = 0.047) than SE, while PED was similar.

Conclusions

In this single-center comparative cohort, SE demonstrated a favorable safety and efficacy profile, with lower rates of major stroke and aneurysm enlargement and improved procedural efficiency compared with earlier-generation devices.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1007/s00701-026-06952-x.

Introduction

Flow diversion (FD) has become an established therapeutic approach for the management of intracranial aneurysms, particularly those that are large, giant, or otherwise unsuitable for microsurgical clipping or endovascular coiling [2, 12, 15, 16, 18]. The Surpass Evolve (SE; Stryker) is a second-generation device within the Surpass platform, characterized by a reduced profile and enhanced navigability compared with the Surpass Streamline (SS; Stryker). Our previous single-center experience suggested shorter procedural duration, less frequent need for balloon angioplasty, and fewer periprocedural ischemic lesions relative to earlier devices [7]. However, that study was limited by a modest sample size and short follow-up period. The present study was therefore designed to evaluate the angiographic efficacy, safety profile, and clinical outcomes of SE in a larger single-center cohort with extended follow-up, and to compare its performance with earlier-generation FDs.

Methods

Patient selection and data collection

This single-center comparative cohort study was approved by the institutional review board (IRB File No. 2021–01–176). Patient data for the SE cohort were prospectively collected between July 2014 and August 2023. The control cohort comprised patients treated with other commercially available flow diverters (Pipeline Embolization Device [PED; Flex and Flex with Shield Technology, Medtronic] or SS) between July 2014 and August 2023, whose data were retrospectively reviewed. Only patients with unruptured intracranial aneurysms were included, encompassing both anterior and posterior circulation lesions.

Interventions and follow-up

Treatment decisions were made through multidisciplinary discussion. Device selection was at the discretion of the operator. All patients received dual antiplatelet therapy (aspirin and clopidogrel, or ticlopidine in hyporesponders) prior to treatment, maintained for 6 months, followed by single antiplatelet therapy for an additional 6 months. Procedures were performed under general anesthesia with the intent to deploy a single flow diverter. Additional stents were placed if complete neck coverage could not be achieved. Post-deployment angiography and rotational flat-panel Computed Tomography (Dyna CT) were performed to assess stent apposition, and balloon angioplasty was used when necessary. Post-procedural magnetic resonance imaging (MRI) was obtained within 5 days. Clinical and radiological follow-up with MRI or CT angiography was scheduled at 6, 12, 18, and 24 months, and annually thereafter.

Outcome measures

Main outcomes were complete aneurysm occlusion, major stroke, in-stent thrombosis, aneurysm enlargement, and unfavorable functional outcome. Complete occlusion was defined as angiographic obliteration of the target aneurysm without significant (> 50%) parent artery stenosis or need for retreatment. Major stroke was defined as hemorrhagic or ischemic stroke resulting in a ≥ 4-point increase in the National Institutes of Health Stroke Scale (NIHSS) score. In-stent thrombosis was defined as partial or complete thrombus formation within the stent, with or without infarction. Aneurysm enlargement was defined as an increase in aneurysm volume at follow-up exceeding 125% of baseline, based on a threshold previously validated in our institutional cohort as being strongly associated with major adverse events and unfavorable functional outcomes following flow diversion [8]. Unfavorable functional outcome was defined as a modified Rankin Scale (mRS) score of 3–6 at the final clinical follow-up. The same definitions for all outcome measures were applied uniformly across all three device groups (SE, SS, and PED).

Detailed outcomes included a composite safety endpoint (any hemorrhagic stroke, major ischemic stroke, in-stent thrombosis, or mortality) and procedural parameters (number of devices used, need for balloon angioplasty, and procedural time). Procedural time was defined as the interval from the first angiographic run of the target artery to femoral sheath removal; for a secondary analysis, procedures were dichotomized as ≥ 90 min or < 90 min. Angiographic course variables (time to complete occlusion and duration of angiographic follow-up) were also recorded.

Statistical analyses

Statistical analyses were conducted using R software (version 3.4.2; R Foundation for Statistical Computing, Vienna, Austria). Two-tailed p-values < 0.05 were considered statistically significant. Categorical variables were compared using the chi-square test with continuity correction, and continuous variables were analyzed using the t-test.

To mitigate potential selection bias stemming from baseline differences in aneurysm size and clinical characteristics, propensity score matching (PSM) was performed, resulting in 41 pairs of matched Surpass Evolve and control cases. A logistic regression model was constructed with the device type as the outcome variable and covariates including age, sex, aneurysm location, maximal aneurysm diameter, neck diameter, non-saccular morphology, presence of an incorporated branch, and ARU to generate propensity scores. “Nearest-neighbor” matching was performed without replacement to assign matched pairs. Baseline demographics, procedural details, and clinical outcomes were re-evaluated for these matched pairs using the same statistical methods as for the total cohort.

Associations between clinical variables and complete aneurysm occlusion were assessed through univariate binary logistic regression. Variables with a p-value < 0.20 were included in a multivariate logistic regression analysis to identify independent predictors. Results were expressed as odds ratios (ORs) with 95% confidence intervals (CIs). Additionally, time-dependent factors influencing complete aneurysm occlusion were analyzed using a Cox proportional hazards model, with a significance level of 0.20. Hazards ratios (HRs) and 95% CIs were calculated for the identified risk factors. For three-arm comparisons by device type, categorical variables were analyzed using chi-square or Fisher's exact test, and continuous variables using one-way ANOVA. Pairwise comparisons used Fisher's exact test or t-test with Bonferroni correction.

Results

Baseline characteristics

A total of 120 patients were included in the study, with 67 patients in the SE group and 53 patients in the Control group. The Control group included the SS (31 patients, 58.5%) and the PED (22 patients, 41.5%). Baseline characteristics are presented in Table 1. Patients in the SE group were more often female (76.1% vs. 56.6%, p = 0.038) and had smaller aneurysm diameters (13.3 ± 6.3 mm vs. 17.8 ± 7.1 mm, p < 0.001). Aneurysms with incorporated branches were less common in the SE group (17.9% vs. 39.6%, p = 0.015). Although not statistically significant, the SE group tended to include fewer previously treated aneurysms, more internal carotid artery lesions, and lower proportions of non-saccular and wide-neck aneurysms.

Table 1.

Baseline characteristics of patients treated with flow diverters

	Surpass Evolve (N = 67)	Control (N = 53)	Total (N = 120)	p-value
Device				< 0.001
Surpass Evolve	67 (100.0%)	0 (0.0%)	67 (55.8%)
Surpass Streamline	0 (0.0%)	31 (58.5%)	31 (25.8%)
Pipeline Flex	0 (0.0%)	22 (41.5%)	22 (18.3%)
Age (years)	58.9 ± 12.4	59.3 ± 13.9	59.1 ± 13.0	0.859
Female	51 (76.1%)	30 (56.6%)	81 (67.5%)	0.038
Previous treatment	2 (3.0%)	6 (11.3%)	8 (6.7%)	0.175
Coil embolization	1 (1.5%)	3 (5.7%)	4 (3.3%)
Endoluminal stent reconstruction	1 (1.5%)	0 (0.0%)	1 (0.8%)
Clipping	0 (0.0%)	2 (3.8%)	2 (1.7%)
Bypass/Distal occlusion	0 (0.0%)	1 (1.9%)	1 (0.8%)
Anterior circulation	50 (74.6%)	35 (66.0%)	85 (70.8%)	0.409
Location				0.067
ICA	48 (71.6%)	27 (50.9%)	75 (62.5%)
ACA	1 (1.5%)	3 (5.7%)	4 (3.3%)
MCA	1 (1.5%)	5 (9.4%)	6 (5.0%)
PCA	1 (1.5%)	0 (0.0%)	1 (0.8%)
BA	4 (6.0%)	2 (3.8%)	6 (5.0%)
VA	12 (17.9%)	16 (30.2%)	28 (23.3%)
Non-saccular type	21 (31.3%)	24 (45.3%)	45 (37.5%)	0.169
Aneurysm diameter (mm)	13.3 ± 6.3	17.8 ± 7.1	15.3 ± 7.0	< 0.001
< 10	19 (28.4%)	4 (7.5%)	23 (19.2%)	0.006
10–25	43 (64.2%)	39 (73.6%)	82 (68.3%)
≥ 25	5 (7.5%)	10 (18.9%)	15 (12.5%)
Neck diameter (mm)	8.4 ± 4.2	9.6 ± 5.3	8.9 ± 4.7	0.167
Incorporated branch	12 (17.9%)	21 (39.6%)	33 (27.5%)	0.015
Pre-treatment DAPT				0.181
On DAPT	2 (3.0%)	2 (3.8%)	4 (3.3%)
Loading dose	2 (3.0%)	6 (11.3%)	8 (6.7%)
Scheduled	63 (94.0%)	45 (84.9%)	108 (90.0%)
Pre-treatment DAPT regimen				0.906
Aspirin + Cilostazol	0 (0.0%)	1 (1.9%)	1 (0.8%)
Aspirin + Clopidogrel	67 (100.0%)	52 (98.1%)	119 (99.2%)
ARU	429.8 ± 58.5	464.2 ± 74.4	445.0 ± 67.9	0.005
PRU	178.2 ± 69.9	175.8 ± 70.1	177.1 ± 69.7	0.852
Post-treatment DAPT				0.488
Aspirin + Cilostazol	0 (0.0%)	1 (1.9%)	1 (0.8%)
Aspirin + Clopidogrel	56 (83.6%)	42 (79.2%)	98 (81.7%)
Aspirin + Ticlopidine	11 (16.4%)	10 (18.9%)	21 (17.5%)
Additional coil	2 (3.0%)	1 (1.9%)	3 (2.5%)	1.000

Values are presented as number (percentage) or mean ± standard deviation, unless otherwise indicated. ICA internal carotid artery, ACA anterior cerebral artery, MCA middle cerebral artery, PCA posterior cerebral artery, BA basilar artery, VA vertebral artery, DAPT dual antiplatelet therapy, ARU aspirin resistance units, PRU P2Y12 reaction units

To minimize the potential bias from baseline characteristic differences, a 1:1 PSM was performed, resulting in 41 pairs (Table 2). After matching, significant differences in baseline variables were mitigated, including female gender (75.6% vs. 70.7%, p = 0.803), aneurysm diameter (13.7 ± 6.7 mm vs. 13.3 ± 5.3 mm, p = 0.753), the presence of incorporated branches (19.5% vs. 22.0%, p = 1.000), and ARU levels (426.1 ± 60.1 vs. 422.1 ± 43.6, p = 0.734).

Table 2.

Matched patients baseline characteristics

	Matched Surpass Evolve (N = 41)	Matched Control (N = 41)	p-value
Age (years)	59.4 ± 12.2	57.4 ± 13.1	0.466
Female	31 (75.6%)	29 (70.7%)	0.803
Anterior circulation	30 (73.2%)	29 (70.7%)	1.000
Aneurysm diameter (mm)	13.7 ± 6.7	13.3 ± 5.3	0.753
Neck diameter (mm)	8.2 ± 4.8	7.7 ± 2.4	0.553
Non-saccular type	16 (39.0%)	15 (36.6%)	1.000
Incorporated branch	8 (19.5%)	9 (22.0%)	1.000
ARU	426.1 ± 60.1	422.1 ± 43.6	0.734
PRU	165.7 ± 69.5	187.3 ± 57.9	0.131

Values are presented as number (percentage) or mean ± standard deviation, unless otherwise indicated. ARU aspirin resistance units, PRU P2Y12 reaction units

Main outcomes

In the SE cohort (N = 67), complete aneurysm occlusion was achieved in 53 patients (79.1%). Major stroke occurred in one patient (1.5%), which was an ischemic event; no hemorrhagic strokes were observed. In-stent thrombosis was documented in one patient (1.5%), and aneurysm enlargement was identified in two patients (3.0%). At the last clinical follow-up (mean, 20.3 ± 9.9 months), only one patient (1.5%) had an unfavorable functional outcome. When compared with the control cohort, the SE group demonstrated significantly fewer major strokes (p = 0.029) and a lower incidence of aneurysm enlargement (p = 0.002) (Tables  3).

Table 3.

Main outcomes after flow diversion treatment

	Surpass Evolve (N = 67)	Control (N = 53)	p-value
Complete aneurysm occlusion*	53 (79.1%)	34 (64.2%)	0.106
Major stroke	1 (1.5%)	7 (13.2%)	0.029
Hemorrhagic stroke	0 (0.0%)	4 (7.5%)	0.076
Major ischemic stroke†	1 (1.5%)	3 (5.7%)	0.453
In-stent thrombosis	1 (1.5%)	3 (5.7%)	0.453
Aneurysm enlargement‡	2 (3.0%)	12 (22.6%)	0.002
Unfavorable functional outcome§	1 (1.5%)	4 (7.5%)	0.235

Values are presented as number (percentage) unless otherwise indicated

*Complete aneurysm occlusion was defined as the angiographic occlusion of the target aneurysm at the end of follow-up, without significant (> 50%) parent artery stenosis or the need for additional treatment

†Major ischemic stroke was defined as an increase of ≥ 4 points in the National Institutes of Health Stroke Scale score

‡Aneurysm enlargement was defined as a follow-up aneurysm volume exceeding 125% of the initial aneurysm volume

§Unfavorable functional outcome was defined as a modified Rankin Scale score of 3–6 at the last clinical follow-up

The clinical and angiographic outcomes for the matched cohorts are summarized in Table 4. Even after adjusting for potential confounders through PSM, several key clinical trends remained evident. Most notably, the SE group demonstrated a significantly lower rate of aneurysm enlargement (2.4% vs. 19.5%, p = 0.012). The incidence of major stroke (2.4% vs. 12.2%, p = 0.115) and hemorrhagic stroke (0.0% vs. 7.3%, p = 0.112) also remained lower in the SE group. These results suggest that the favorable safety profile of the SE device is observed even when baseline differences are balanced.

Table 4.

Matched cases main outcomes after flow diversion treatment

	Matched Surpass Evolve (N = 41)	Matched Control (N = 41)	p-value
Complete aneurysm occlusion*	32 (78.0%)	27 (65.9%)	0.224
Major stroke	1 (2.4%)	5 (12.2%)	0.115
Hemorrhagic stroke	0 (0.0%)	3 (7.3%)	0.112
Major ischemic stroke†	1 (2.4%)	2 (4.9%)	1
In-stent thrombosis	1 (2.4%)	2 (4.9%)	1
Aneurysm enlargement‡	1 (2.4%)	8 (19.5%)	0.012
Unfavorable functional outcome§	1 (2.4%)	3 (7.3%)	0.306

Values are presented as number (percentage) unless otherwise indicated

*Complete aneurysm occlusion was defined as the angiographic occlusion of the target aneurysm at the end of follow-up, without significant (> 50%) parent artery stenosis or the need for additional treatment

†Major ischemic stroke was defined as an increase of ≥ 4 points in the National Institutes of Health Stroke Scale score

‡Aneurysm enlargement was defined as a follow-up aneurysm volume exceeding 125% of the initial aneurysm volume

§Unfavorable functional outcome was defined as a modified Rankin Scale score of 3–6 at the last clinical follow-up

Detailed outcomes

In the SE group, composite safety outcomes occurred in three patients (4.5%), including one covered perforator infarction, one in-stent thrombosis without infarction, and one death. No hemorrhagic complications were observed. Compared with the control group, the incidence of composite safety outcomes was not significantly different (p = 0.092). Significant (> 50%) parent artery stenosis, which by definition precluded classification as complete occlusion, occurred in 2 patients (3.0%) in the SE group, 3 patients (9.7%) in the SS group, and 2 patients (9.1%) in the PED group, with no statistically significant difference between SE and the combined control group (p = 0.239).

Procedural characteristics in the SE cohort included the use of a single device in most patients (97.0%) and balloon angioplasty in 16 cases (23.9%). The mean procedure time was 79.3 ± 33.2 min, with 32.8% of procedures lasting 90 min or longer. The mean time to complete occlusion was 13.3 ± 10.5 months, and the mean duration of angiographic follow-up was 19.2 ± 9.4 months.

When compared with controls, balloon angioplasty was less frequently required in the SE group (p = 0.023), and both mean procedure time (p = 0.001) and the proportion of procedures lasting ≥ 90 min (p = 0.005) were significantly lower (Table 5).

Table 5.

Detailed outcomes after flow diversion treatment

	Surpass Evolve (N = 67)	Control (N = 53)	Total (N = 120)	p-value
Composite safety outcomes*	3 (4.5%)	8 (15.1%)	11 (9.2%)	0.092
Hemorrhagic stroke	0 (0.0%)	4 (7.5%)	4 (3.3%)	0.076
Remote ICH	0 (0.0%)	2 (3.8%)	2 (1.7%)	0.197
Delayed rupture, SAH	0 (0.0%)	1 (1.9%)	1 (0.8%)	0.455
Cortical siderosis	0 (0.0%)	1 (1.9%)	1 (0.8%)	0.455
Ischemic stroke	1 (1.5%)	3 (5.7%)	4 (3.3%)	0.320
Downstream embolic infarction	0 (0.0%)	1 (1.9%)	1 (0.8%)	0.455
Covered perforator infarction	1 (1.5%)	0 (0.0%)	1 (0.8%)	0.343
In-stent thrombosis with infarction	0 (0.0%)	2 (3.8%)	2 (1.7%)	0.197
In-stent thrombosis without infarction	1 (1.5%)	1 (1.9%)	2 (1.7%)	0.879
Mortality	1 (1.5%)	0 (0.0%)	1 (0.8%)	0.343
Number of flow diverters used				0.788
1	65 (97.0%)	50 (94.3%)	115 (95.8%)
2	2 (3.0%)	3 (5.7%)	5 (4.2%)
Balloon angioplasty	16 (23.9%)	24 (45.3%)	40 (33.3%)	0.023
Procedural time (min)†	79.3 ± 33.2	106.9 ± 52.8	91.5 ± 44.9	0.001
Procedural time ≥ 90 min	22 (32.8%)	32 (60.4%)	54 (45.0%)	0.005
Mean time to complete occlusion (months)	13.3 ± 10.5	16.6 ± 20.2	14.7 ± 15.6	0.287
Mean angiographic follow-up (months)	19.2 ± 9.4	34.8 ± 23.8	26.1 ± 18.9	< 0.001

Values are presented as number (percentage) or mean ± standard deviation, unless otherwise indicated. ICH intracerebral hemorrhage, SAH subarachnoid hemorrhage

*Composite safety outcomes include any hemorrhagic or ischemic stroke, in-stent thrombosis, or mortality

†Procedural time was estimated as the time between the first angiography of the target artery and femoral sheath removal

Values are expressed as odds ratios (OR) with 95% confidence intervals (CI)

Figure 1 provides a graphical summary of the main clinical outcomes, demonstrating the overall rates of complete occlusion, composite safety events, major stroke, in-stent thrombosis, unfavorable functional outcome, and aneurysm enlargement in the SE and control groups.

Fig. 1.

Clinical outcomes in patients treated with Surpass Evolve versus other flow diverters. Bar charts demonstrate the proportion of patients experiencing each outcome: complete aneurysm occlusion, composite safety outcomes, major stroke, in-stent thrombosis, mRS > 2, and aneurysm enlargement. Patients in the Surpass Evolve group showed a significantly lower incidence of major stroke (p = 0.029) and aneurysm enlargement (p = 0.002) compared with the control group (* indicates statistical significance, p < 0.05). mRS, modified Rankin Scale

Regarding procedural parameters for the results after PSM, the SE group showed a significantly shorter mean procedural time compared to the matched control group (80.5 ± 34.2 min vs. 106.3 ± 46.5 min, p = 0.008), with 65.9% (27/41) of cases completed within 90 min. The mean angiographic follow-up duration for the matched cohort was 19.4 ± 10.6 months. These procedural values align with the patterns observed in the entire cohort (Table 5), particularly the trend toward reduced procedural time in the SE group. The fact that these favorable outcomes—encompassing both the shortened procedural time and the favorable clinical outcomes such as a lower incidence of major stroke and the prevention of aneurysm enlargement—persist in the matched analysis confirms that the benefits of the SE device are independent of baseline clinical and anatomical variations (Table 4).

Three-arm comparison by device type

To address the possibility that pooling the control group may obscure device-specific differences, we performed a three-arm analysis comparing SE, SS, and PED individually (Supplementary Table 1). After Bonferroni correction for pairwise comparisons, SE was significantly superior to SS in major stroke (p = 0.035), aneurysm enlargement (p = 0.001), balloon angioplasty (p = 0.002), and procedural time (p < 0.001), whereas no SE-versus-PED comparison reached significance. The poorer outcomes in the combined control group were therefore driven predominantly by SS, while PED showed outcomes numerically intermediate between SE and SS but statistically comparable to SE—consistent with the Cox analysis (Fig. 2).

Fig. 2.

Kaplan–Meier curves for time to complete occlusion in patients treated with Surpass Evolve (reference), Surpass Streamline, and Pipeline Flex. In Cox proportional hazards analysis, the probability of occlusion with Pipeline Flex was comparable to SE (HR, 1.30; 95% CI, 0.74–2.28; p = 0.363), whereas Streamline was associated with a significantly lower hazard of occlusion (HR, 0.58; 95% CI, 0.34–0.99; p = 0.047)

Predictors of outcomes

In univariate analysis, SE treatment showed an odds ratio of 2.12 (95% CI, 0.94–4.85; p = 0.071) for complete aneurysm occlusion. The odds ratio for composite safety outcomes was 0.26 (95% CI, 0.06–0.97; p = 0.058). Major stroke was less frequent in the SE cohort (OR, 0.10; 95% CI, 0.01–0.59; p = 0.034).

Multivariable analysis was performed including female sex, aneurysm diameter, neck diameter, and ARU as covariates, selected based on baseline characteristics and univariate screening results. In this model, SE treatment was associated with a lower odds of major stroke (adjusted OR, 0.15; 95% CI, 0.01–0.93), although this did not reach statistical significance (p = 0.084). For aneurysm enlargement, the odds ratios were 0.11 (95% CI, 0.02–0.41; p = 0.004) in univariate analysis and 0.14 (95% CI, 0.02–0.61; p = 0.019) after adjustment. No significant associations were found for in-stent thrombosis or unfavorable functional outcome (Table 6).

Table 6.

Association between use of Surpass Evolve and clinical outcomes: results of univariate and multivariate analyses

Outcome	Unadjusted OR (95% CI)	p-value	Adjusted OR (95% CI)	p-value
Complete aneurysm occlusion	2.12 (0.94–4.85)	0.071	–	–
Composite safety outcomes	0.26 (0.06–0.97)	0.058	–	–
Major stroke	0.10 (0.01–0.59)	0.034	0.15 (0.01–0.93)	0.084
In-stent thrombosis	0.41 (0.03–5.73)	0.507	–	–
Aneurysm enlargement	0.11 (0.02–0.41)	0.004	0.14 (0.02–0.61)	0.019
Unfavorable functional outcome	0.19 (0.02–1.71)	0.137	–	–

Unadjusted ORs were obtained from univariate logistic regression analyses

Adjusted ORs were obtained from multivariate logistic regression analyses including variables with p < 0.10 in the univariate analysis

A dash (–) indicates that the variable was not retained in the multivariate model after adjustment

Cox proportional hazards analysis for complete aneurysm occlusion was conducted using the SE group as the reference. The cumulative probability of occlusion in the PED group was comparable to that of SE (HR, 1.30; 95% CI, 0.74–2.28; p = 0.363), whereas the SS group demonstrated a significantly lower hazard of occlusion (HR, 0.58; 95% CI, 0.34–0.99; p = 0.047) (Fig. 2).

Discussion

In this single-center comparative cohort study, the SE demonstrated a favorable safety and efficacy profile in the treatment of unruptured intracranial aneurysms. Complete occlusion was achieved in the majority of patients, while the incidence of major stroke and aneurysm enlargement was significantly lower compared with earlier-generation devices. Procedural efficiency was also improved, with reduced procedure time and less frequent need for balloon angioplasty. These findings suggest that SE represents a meaningful advancement in flow diversion therapy.

Previous retrospective studies and meta-analyses have consistently demonstrated the safety and efficacy of the SE in the treatment of intracranial aneurysms [5, 11, 13, 14, 17]. Jazayeri et al. reported a complete occlusion rate of 67.1% and an adequate occlusion rate of 86.6%, with an ischemic stroke rate of 1.9% and a device-related mortality of 0.5%, whereas Rodriguez-Calienes et al. demonstrated a pooled complete occlusion rate of 69% and a favorable occlusion rate of 91%, with overall neurological complications occurring in approximately 6% of patients [5, 17].

In our single-center comparative cohort, the SE group achieved a complete occlusion rate of 79.1%, with major stroke occurring in only 1.5% and favorable functional outcome observed in 98.5% of patients, which is largely consistent with the pooled results of these meta-analyses. Importantly, our study also identified a significantly lower rate of aneurysm enlargement in the SE group compared with controls, an outcome not routinely emphasized in prior meta-analyses.

One methodological difference that should be noted is that our study evaluated only complete aneurysm occlusion as the primary angiographic endpoint, whereas prior meta-analyses often incorporated adequate or favorable occlusion (e.g., O'Kelly Marotta C–D or Raymond-Roy I–II) as outcome measures [5, 17]. The concept of adequate occlusion was introduced to account for the progressive nature of flow diversion, in which substantial but incomplete reduction of aneurysm filling may indicate a favorable trajectory toward eventual cure. This broader definition was particularly useful in studies with relatively short or mid-term follow-up, allowing earlier interpretation of treatment success. In contrast, our cohort was characterized by extended follow-up, and we deliberately focused on complete occlusion alone, reasoning that any residual filling, even if reduced, still carries a potential risk of aneurysm regrowth or rupture and therefore should not be regarded as a definitive endpoint [8]. Consequently, our findings may represent a more stringent assessment of angiographic efficacy, while still remaining aligned with the overall conclusions of previous evidence supporting the favorable performance of SE.

The reason aneurysm enlargement was included as a main outcome in the present study is its strong association with poor clinical outcomes. Our previous study involving patients treated with flow diversion found that aneurysm enlargement occurred in 16.4% of all aneurysms [8]. In that cohort, the enlargement group had a significantly higher rate of major adverse events and unfavorable functional outcomes than the no-aneurysm filling or stable remnant filling groups (p < 0.001). Notably, a major adverse event occurred in 80.0% of patients in the Enlargement group. In our current study, the rate of aneurysm enlargement in the SE group was observed to be low at 3.0% compared to our previous study results, and this incidence was also significantly lower than that of the control group (22.6%; p = 0.002) [8]. Given the potential long-term influence of enlargement on poor clinical outcomes, this remarkable decrease in aneurysm enlargement within the SE group of the present study is considered a significant difference.

Our PSM analysis reaffirms that the clinical benefits of the SE device are not merely a result of patient selection bias. By balancing anatomical factors such as aneurysm diameter and incorporated branches, we demonstrated that the SE group still maintains favorable safety profiles, particularly in significantly preventing aneurysm enlargement (p = 0.012). While we utilized PSM to compare outcomes within a balanced framework to validate these findings, we conducted subsequent analyses, including multivariate Cox regression, using the entire cohort to maximize statistical power and evaluate the independent impact of the device across the full spectrum of treated cases.

Our results align with prior reports that highlighted the improved deliverability and reduced procedural complexity of the SE compared with the Surpass Streamline [7]. Notably, the rate of major ischemic or hemorrhagic stroke was markedly lower in the SE cohort, consistent with the hypothesis that better navigability and lower-profile design reduce the risk of periprocedural complications. In contrast, the Pipeline Embolization Device demonstrated comparable efficacy to SE in terms of complete occlusion, whereas the Surpass Streamline was associated with inferior occlusion rates, highlighting device-specific differences in long-term angiographic performance.

A critical consideration is whether the improved outcomes with SE reflect genuine device advantages or a learning-curve effect. PED, which was used across the entire study period (2014–2023) and overlapped substantially with SE, showed shorter procedural time than SS but longer than SE, and a balloon angioplasty rate similar to SE rather than SS. Because PED and SE were used concurrently, this pattern is more consistent with device profile than with chronology. The lower crossing profile and optimized braid geometry of SE plausibly account for its improved wall apposition, reduced balloon angioplasty need, and lower aneurysm enlargement. Our previous direct SS–SE comparison within overlapping time periods supports this interpretation [7]. We acknowledge, however, that learning-curve effects cannot be entirely excluded for the SE-versus-SS comparison given their limited temporal overlap.

The Pipeline Embolization Device has been extensively studied in large prospective trials. The Premier trial reported 83.5% complete occlusion at one year in small/medium aneurysms with 2.1% major morbidity/mortality [3], PUFS reported 73.6% occlusion at 180 days in large/giant aneurysms with 5.6% neurological morbidity/mortality [1], and the IntrePED registry documented 8.4% neurological morbidity/mortality across 793 real-world cases [10]. The 72.7% PED occlusion rate in our cohort falls within this range, particularly given our substantially larger mean aneurysm size (17.8 ± 7.1 mm in our control cohort vs. 5.0 mm in Premier). The principal SE-specific improvements therefore lie in procedural efficiency and aneurysm enlargement prevention rather than in superior occlusion relative to PED.

An important consideration in interpreting the PSM results is the reduction in statistical power for rare events when sample size is halved. Effect sizes for major stroke (2.4% vs. 12.2%) and hemorrhagic stroke (0.0% vs. 7.3%) in the matched cohort remained similar to those in the full cohort, but significance was not reached (p = 0.115 and p = 0.112). The persistent direction of effect, together with the surviving aneurysm enlargement difference (p = 0.012), suggests limited power rather than absence of a true safety benefit.

This study has several limitations. First, although patient data for the SE group were prospectively collected, the control group was retrospectively assembled, which may introduce selection bias. Second, baseline differences such as aneurysm size and morphology between groups may have influenced outcomes despite multivariable adjustment. Our center has consistently reported research results related to flow diversion outcomes, and the efficacy and safety outcomes have improved over time [6–9]. Growing experience with the flow diversion procedure has led to enhanced technical proficiency, contributing to a reduced average procedural time [4, 7]. Furthermore, the development of devices and the refinement of flow diversion indications are thought to be the cause of the improving outcomes. Early in our experience, the criteria for selecting cases for FD treatment were broader, and our understanding of predictors influencing outcomes was limited. Over time, important factors such as aneurysm diameter, incorporated branches, and parent vessel angle were identified, and this refinement of treatment indications allowed us to select patients who were more likely to benefit from FD [8]. The refinement of indications likely resulted in some differences in the baseline characteristics between the SE group which was used more recently and the control group [9]. Although we performed multivariate analysis to adjust for this, the advancement of these indications may have made the results of the SE group appear better. Third, the sample size, particularly for subgroup comparisons, was relatively modest, and the study was conducted at a single institution, which may limit the generalizability of the results.

Conclusion

In this study, the SE demonstrated a favorable safety and efficacy profile, characterized by lower rates of major stroke and aneurysm enlargement, as well as shorter procedural times compared with earlier-generation devices. These results indicate that SE is an effective device for the treatment of intracranial aneurysms.

Supplementary information

Below is the link to the electronic supplementary material.

Abbreviations

ACA

Anterior cerebral artery

ARU

Aspirin resistance units

BA

Basilar artery

CI

Confidence interval

CT

Computed Tomography

DAPT

Dual antiplatelet therapy

FD

Flow diversion/flow diverter

HR

Hazard ratio

ICA

Internal carotid artery

ICH

Intracerebral hemorrhage

IRB

Institutional Review Board

MCA

Middle cerebral artery

MRI

Magnetic resonance imaging

mRS

Modified Rankin Scale

NIHSS

National Institutes of Health Stroke Scale

OR

Odds ratio

PCA

Posterior cerebral artery

PED

Pipeline Embolization Device

PRU

P2Y12 reaction units

PSM

Propensity Score Matching

SAH

Subarachnoid hemorrhage

SE

Surpass Evolve

SS

Surpass Streamline

VA

Vertebral artery

Author contributions

T.K.J. and H.W.C. contributed equally to this work as co-first authors. T.K.J., H.W.C.: Conceptualization, Methodology, Data curation, Formal Analysis, Writing – Original Draft, Writing –Review and Editing; P.J.: Conceptualization, Methodology, Data Curation, Formal Analysis, Supervision, Visualization, Writing – Review and Editing; J.Y.Y., K.H.K., J-S.K.: Conceptualization, Data curation, Writing – Review and Editing.

Funding

This study was supported by a research grant from Stryker.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval

This study is based on recorded data and authors submitted an application to the institutional review board for the retrospective study.

This study was approved by the institutional review board. All procedures performed in the studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

Consent to participate

A waiver of consent was approved by the institutional review board. This study involves no more than minimal risk to the subject and the waiver will not adversely affect the rights and welfare of the subjects.

Competing interests

This study was supported by a research grant from Stryker. The corresponding author is the sole recipient of the institutional research grant from Stryker that supported this study. Stryker had no role in study design, data collection, analysis, interpretation, manuscript preparation, or the decision to publish.