Abstract
Background
Flow diverters (FDs) are the first line of treatment for specific intracranial aneurysms. However, aneurysm persistence at follow‐up presents in up to 25%. Occlusion after flow diversion in older patients seems less effective due to clinical, anatomical, and physiological characteristics. We aimed to study the effect of age on aneurysm occlusion mediated by intraprocedural technical events.
Methods
We conducted a pooled analysis of 2 cohorts, including patients with unruptured saccular aneurysms in the internal carotid artery, treated with the Surpass Streamline FD. Multivariable logistic regression was used to identify predictors of complete occlusion at 12‐month follow‐up. A mediation analysis was performed to assess the role of intraprocedural technical events (eg, fish‐mouthing of the distal end, poor device opening, FD twisting, foreshortening, excess friction of the FD and the delivery system during deployment, and delivery system kink) in the relationship between age and occlusion rates.
Results
A total of 316 patients (mean age 59.4 ± 11.2 years) were included. Complete aneurysm occlusion was achieved in 82% of cases at 12 months. Increasing age was associated with lower odds of occlusion (adjusted odds ratio = 0.962, P<0.001) and a higher incidence of intraprocedural technical events (adjusted odds ratio = 1.088, P<0.001). Intraprocedural technical events were inversely associated with occlusion (adjusted odds ratio = 0.265, P = 0.004), and mediation analysis revealed that 16.3% of the effect of age on aneurysm occlusion was mediated by these events.
Conclusion
Intraprocedural technical events partially mediate the effect of age on complete aneurysm occlusion after FD treatment. Identifying additional mechanisms that influence occlusion could improve procedural outcomes, particularly in older patients.
Keywords: endovascular, flow diverter, intracranial aneurysm, subarachnoid hemorrhage
Nonstandard Abbreviations and Acronyms
- FD
flow diverter
- SCENT
Safety and Effectiveness of an Intracranial Aneurysm Embolization System for Treating Large or Giant Wide Neck Aneurysms
- SESSIA
Safety and Efficacy of the Surpass Streamline for Intracranial Aneurysms
1. Clinical Perspective
What Is New?
Increasing patient age is associated with a higher incidence of intraprocedural technical events during flow diverter implantation, which in turn reduces the likelihood of complete aneurysm occlusion.
What Are the Clinical Implications?
Intraprocedural technical events explain 16% of the effect of age on aneurysm occlusion, highlighting the need for tailored procedural strategies in older patients.
Identifying and minimizing intraprocedural events could improve long‐term angiographic outcomes, particularly in anatomically challenging cases often seen with aging vasculature.
Flow diversion is considered the first‐line therapy for the treatment of selected intracranial aneurysms due to its safe and effective profile. 1 , 2 , 3 , 4 This technique has expanded gradually from proximal unruptured large complex aneurysms to small distal lesions and, in some cases, even acutely ruptured cases. 5 The efficacy of flow diverters (FDs) depends on several factors, including the characteristics of the device, patient comorbidities, aneurysm complexity, parent vessel tortuosity, and operator experience. Unlike coils or intrasaccular devices, the implantation of FDs produces endoluminal reconstruction and blood flow redirection, which leads to aneurysm occlusion by intrasaccular thrombosis over time. 6 The rate of complete aneurysm occlusion in previously published studies has been shown to increase with longer follow‐up times. 7
Despite the well‐established use and effectiveness of FDs, there are still a significant number of cases in which complete occlusion is not achieved even after a 12‐month follow‐up. 7 , 8 , 9 Several studies have evaluated and found that some factors that influence occlusion include sex, smoking status, aneurysm location, aneurysm size, presence of a side‐branch artery, parent artery diameter, and the use of adjunctive techniques (coils and stents). 10 , 11 , 12 , 13 , 14 Notably, the rates of aneurysm occlusion after flow diversion in older patients are less effective, possibly due to an increased burden of atherosclerosis and a diminished endothelial repair and migration capacity. 10 , 14 , 15 , 16 , 17 Furthermore, the navigation and implantation of FDs in stiff and tortuous blood vessels, commonly found in aging patients, may be technically challenging and, in some cases, unfavorable for optimal FD performance. As life expectancy continues to increase, assessing the effect of increasing age in aneurysm occlusion after flow diversion might offer valuable insights in favor of improving the clinical and angiographical outcomes. Hence, we investigated whether intraprocedural technical events mediated the relationship between age and complete aneurysm occlusion by analyzing pooled patient data from 2 cohort studies (1 prospective and 1 retrospective) of patients with intracranial aneurysms treated with the Surpass Streamline FD.
Methods
The data that support the findings of this study are available from the corresponding author upon reasonable request. Access to the data may be subject to institutional or ethical restrictions.
Study Design and Participants
We analyzed pooled data from 2 databases of previously published studies. 18 , 19 Both studies included patients treated with the Streamline FD (Streamline; Stryker Neurovascular, Fremont, CA, USA). Briefly, the SCENT (Safety and Effectiveness of an Intracranial Aneurysm Embolization System for Treating Large or Giant Wide Neck Aneurysms) trial was a multicenter, prospective, single‐arm, nonrandomized, interventional trial of 180 patients with 180 target aneurysms located in the intracranial internal carotid artery extending from the petrous segment up to the carotid terminus. SESSIA (Safety and Efficacy of the Surpass Streamline for Intracranial Aneurysms) was a multicenter retrospective cohort study that evaluated consecutive patients with intracranial aneurysms. In SCENT, the outcomes were core‐lab adjudicated, whereas in SESSIA, they were locally adjudicated by independent investigators. For both studies, approval from the local institutional review boards was obtained in all participating centers, and all participants (or legal representatives) provided written informed consent for the procedure. For the current study, we exclusively included patients with unruptured saccular aneurysms located in the intracranial internal carotid artery from the petrous segment up to the carotid terminus (Figure 1). This study is reported according to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines. 20
Figure 1.
Study population flow chart. ICA indicates internal carotid artery; PCA, posterior cerebral artery; PICA, posterior inferior cerebellar artery; SCENT, Safety and Effectiveness of an Intracranial Aneurysm Embolization System for Treating Large or Giant Wide Neck Aneurysms; and SESSIA, Safety and Efficacy of the Surpass Streamline for Intracranial Aneurysms. * From the petrous segment up to the carotid terminus.
Outcome
The outcome of interest was complete aneurysm occlusion at the last follow‐up. Aneurysm occlusion was evaluated using the Raymond–Roy occlusion scale, where Class 1 represents complete occlusion. 8 In the SESSIA study, the angiographic follow‐up was performed at 1–3 months, 6–12 months, and >12 months. 18 On the other hand, in the SCENT study, follow‐up evaluations were performed on 1 day, 1 month, 6 months, 1 year, 2 years, 3 years, 4 years, and 5 years. The outcomes reported in the SCENT trial were determined from the analysis of the 12‐month follow‐up data. 19
Mediator
Intraprocedural technical events were assessed by a core lab in the SCENT trial and by independent neurointerventionalists in the SESSIA study, with definitions based on the respective study protocols. These events included fish‐mouthing of the distal end, poor device opening, FD twisting, foreshortening, excess friction of the FD and the delivery system during deployment, delivery system kinking, and mandatory change of the delivery system to achieve successful implantation. A summary of the general criteria is provided in Table S1.
Covariates
We performed a multivariable logistic regression model to identify predictors of complete occlusion. Covariates were selected based on preliminary univariable statistical significance, previous literature reports, and clinical relevance, considering their association and potential effect on angiographic outcome. 18 The following potential covariates were identified: previous treatment, number of devices implanted, diameter of the implanted FD (in mm), balloon‐assisted angioplasty, and the aneurysm neck width (in mm). Information on the mechanisms underlying the relationships was derived from empirical associations between specific characteristics, as well as theory.
Mediation Model
Figure 2 depicts the hypothetical causal model where age is associated with intraprocedural technical events during the procedure and where intraprocedural technical events mediate complete aneurysm occlusion at follow‐up. To assess the intraprocedural technical events as mediators of the relationship between patient age and complete aneurysm occlusion, mediation analysis was performed using the method described by Baron and Kenny, 21 as well as the method presented by Vanderweele and Vansteelandt. To perform mediation analysis, it is necessary to test 3 pathways: pathway c, the association between the covariate of interest (age) and the outcome (aneurysm occlusion); pathway a, the association between the covariate of interest (age) and intraprocedural technical events (mediator); and pathway b, the association between intraprocedural technical events (mediator) and the outcome (aneurysm occlusion). Once confirmation is obtained for all 3 associations, the establishment of mediation (indirect effect) occurs in a fourth pathway through the assessment of the direct causal relationship (pathway c', see Figure 2). Mediation is absent when not all pathways are satisfied. All pathways were tested using multivariable logistic regression analysis and reported as adjusted odds ratios (aORs). The multivariable modeling included the covariates listed above.
Figure 2.
Model of the hypothetical causal pathway in patients with intracranial aneurysms treated with the Surpass Streamline. Total effect (c') = direct effect (c) + indirect effect (ab). Indirect effect was computed as the difference between regression coefficients.
The proportion of the causal relationship between age and aneurysm occlusion that is mediated through intraprocedural technical events was estimated by dividing the log ORs of indirect effect (pathways ab) by the log OR of the total effect (pathway c).
Predicted probabilities of complete aneurysm occlusion were computed using the full model (pathway c') for the 2 intraprocedural technical events status using the mean values of the diameter of the implanted FD and the aneurysm neck width, over a range of age values. Results are plotted against age.
Two hypothetical patients were defined as someone with the mean values of the diameter of the implanted FD, the aneurysm neck width, and the age, and with different statuses of intraprocedural technical events. Predicted probabilities of complete aneurysm occlusion were computed for them based on the full model (pathway c'). Parametric bootstrap on the full model (pathway c') was used to compute 95% CI as quantiles and P value as proportion of bootstrapped difference being less than or equal to the difference using the original data. 22
Statistical Analysis
Descriptive statistics were used to summarize continuous and categorical variables. Based on data distribution, continuous variables were reported as mean±SD or median with the interquartile range as appropriate. Categorical variables were reported as frequencies and percentages. The normality of distributions was assessed by the Shapiro–Wilk test.
To optimize the utility of the available outcome data, we conducted a multiple imputation procedure of the missing aneurysm neck width values (no other variables had missing values). Results were pooled together using Rubin's rules. 23 Next, the steps were tested from the imputed sample using univariable (steps 1 and 2) and multivariable (step 3) logistic regressions following the guidelines given by Baron and Kenny. 21 For the adjusted model, we performed a multivariable logistic stepwise regression to identify predictors of complete occlusion. The candidate variables (previous treatment, number of devices implanted, diameter of the FD, balloon‐assisted angioplasty, and neck width) were selected based on preliminary univariate statistical significance, previous literature reports, and clinical relevance. 18 For predictor selection, cases with missing observations were not included. All OR and 95% CI were estimated using a logistic mixed‐effects model with a random effect for center to account for between‐center variabilities. The proportion of the effect of age on complete aneurysm occlusion mediated by intraprocedural technical events was estimated as log OR of the indirect effect divided by log OR of the total effect. Finally, we estimated the probability of aneurysm occlusion using adjusted predicted probability curves and stratifying by the occurrence of technical events. The curves were constructed across different patient ages using the unimputed mean diameter of the Streamline and the mean neck width. The P value between curves was computed using bootstrap samples.
All the statistical analyses were considered significant at a 2‐sided alpha level of P≤0.05 set as statistical significance. Statistical analyses were conducted using R version 4.1.3 (R Foundation for Statistical Computing, Vienna, Austria). Data will be made available upon reasonable request to the corresponding author.
Results
Baseline Characteristics and Outcomes
A total of 457 patients with intracranial aneurysms who underwent flow diversion with Streamline were initially assessed. We excluded 141 patients with aneurysms previously ruptured (n = 43), located in the posterior circulation (n = 30), nonsaccular morphology (n = 44), and without outcome data (n = 24) (Figure 1). A total of 316 patients with 316 aneurysms were included in the analysis.
Baseline characteristics are shown in Table 1. Mean age was 59.4 ± 11.2 years, and 265 patients (84%) were female. The median aneurysm size was 7.2 mm (mean 8.8 ± 5.8 mm), with a median neck size of 4.3 mm (mean 4.9 ± 2.8 mm). The most common location was the paraophthalmic segment of the internal carotid artery in 149 (47%) cases, followed by the petrocavernous (85 cases, 27%) and supraclinoid segments (79 cases, 25%). Previous treatment was evidenced in 10 (3%) aneurysms, of which stent with and without coiling was the most common in 5 (2%). A single Streamline was implanted in 278 (88%) patients, and multiple FDs were used in 38 (12%). The mean number of Streamlines implanted per patient was 1.1 ± 0.3 (range 1–3). Balloon‐assisted angioplasty was performed in 153 (48%) patients, and adjuvant coils were used in 14 (4%).
Table 1.
Baseline Characteristics
| Characteristic | Value |
|---|---|
| Patient demographics (n = 316) | |
| Age, y, mean ± SD | 59.4 ± 11.2 |
| Female, n (%) | 265 (84) |
| Baseline mRS score, mean ± SD (range) | 1 ± 0.7 (1–4) |
| Race or ethnicity, n (%) | |
| White | 196 (62) |
| Black | 54 (17) |
| Hispanic | 41 (13) |
| Others * | 25 (8) |
| Comorbidities, n (%) | |
| Hypertension | 180 (57) |
| Hyperlipidemia | 113 (36) |
| Diabetes | 50 (16) |
| Sleep apnea | 14 (5) |
| Alcohol consumption | 86 (27) |
| Smoking history or active | 159 (50) |
| Clinical presentation, n (%) | |
| Asymptomatic/incidental finding | 191 (60) |
| Symptomatic† | 125 (40) |
| Aneurysm characteristics (n = 316) | |
| Aneurysm dimensions, median (mean ± SD) | |
| Size, mm | 7.2 (8.8 ± 5.8) |
| Dome width, mm | 6.0 (7.8 ± 5.6) |
| Neck width, mm | 4.3 (4.9 ± 2.8) |
| Dome‐to‐neck ratio | 1.8 (2.2 ± 1.5) |
| Aneurysm location, n (%) | |
| Paraophthalmic segment | 149 (47) |
| Petrocavernous segment | 85 (27) |
| Supraclinoid segment | 79 (25) |
| Internal carotid artery terminus | 3 (1) |
| Previous treatment, n (%) | |
| Primary coiling | 4 (1) |
| Stent or stent+coil | 5 (2) |
| Coils+clipping | 1 (0.3) |
| Procedure characteristics | |
| FDs used per patient, mean ± SD (range) | 1.1 ± 0.3 (1–3) |
| Patients with >1 FD, n (%) | 38 (12) |
| FD length (mm), median (mean±SD) | 20.0 (14.9 ± 12.0) |
| FD diameter (mm), median (mean±SD) | 4.0 (3.8 ± 0.5) |
| Balloon‐assisted angioplasty, n (%) | 153 (48) |
| Adjunctive coils, n (%) | 14 (4) |
| Adjunctive stent, n (%) | 19 (6) |
FD indicates flow diverter; and mRS, modified Rankin Scale.
11 Asian, 14 not reported or unknown.
Headaches, mass effect, neurologic deficit, stroke, transient ischemic attack.
Characteristics of the outcomes are shown in Table 2. A total of 29 intraprocedural technical events were encountered. Fish‐mouthing of the distal end of the FD was the most common event in 6 (1.9%) patients, followed by kinking of the delivery system in 5 (1.6%), poor FD opening in 4 (1.2%), change in the delivery system in 4 (1.2%), and foreshortening in 3 (0.9%). At a median follow‐up of 12.4 months, complete occlusion was achieved in 258 (82%) patients. Target aneurysm retreatment was performed in 6 (2%) patients. More than 50% parent vessel stenosis and in‐stent occlusion were observed in 9 (3%) patients, respectively. The follow‐up modified Rankin Scale score was 1.2±0.6. Major ischemic events were encountered in 6 (2%) patients, and target aneurysm rupture occurred in 1 (0.3%). One (0.3%) patient died after successful flow diversion due to subdural hematoma secondary to a traumatic fall.
Table 2.
Characteristics of the Outcomes
| Characteristic | Value |
|---|---|
| Procedural outcomes | |
| Intraprocedural technical events (n = 29), n (%) | |
| Fish‐mouthing of the distal end of the FD | 6 (1.9) |
| Kinking of the delivery system | 5 (1.6) |
| Change in delivery system | 4 (1.3) |
| Poor FD opening | 4 (1.2) |
| Foreshortening | 3 (0.9) |
| Excess friction during deployment | 3 (0.9) |
| FD migration during deployment | 2 (0.6) |
| FD twisting | 2 (0.6) |
| Angiographic outcomes at 12 mo | |
| Final follow‐up time (mo), median (mean ± SD) | 12.4 (12.8 ± 4.3) |
| Raymond–Roy occlusion scale, n (%) | |
| Class I | 258 (82) |
| Class II | 22 (7) |
| Class III | 36 (11) |
| Follow‐up outcomes related to the FD implantation, n (%) | |
| Target aneurysm retreatment | 6 (2) |
| Significant (>50%) parent vessel stenosis | 9 (3) |
| Asymptomatic FD occlusion | 9 (3) |
| Clinical outcomes | |
| Follow‐up mRS score, mean ± SD (range) | 1.2 ± 0.6 (1–6) |
| Major ipsilateral ischemic stroke | 6 (2) |
| Minor ipsilateral ischemic stroke | 5 (2) |
| Total mortality | 1 (0.3) |
| Hemorrhagic events * | 1 (0.3) |
FD indicates flow diverter; and mRS, modified Rankin Scale.
Intraoperative target aneurysm rupture.
Mediation Analysis
Results are presented in Table 3. Increasing age was independently associated with lower odds of complete aneurysm occlusion (pathway c: aOR, 0.962; 95% CI, 0.934–0.990; P<0.001) and with the occurrence of intraprocedural technical events (pathway a: aOR, 1.088; 95% CI, 1.036–1.142; P<0.001). Intraprocedural technical events were inversely associated with aneurysm occlusion (pathway b: aOR, 0.265; 95% CI, 0.106–0.661; P = 0.004). After adjusting for intraprocedural technical events, increasing age remained associated with lower odds of complete aneurysm occlusion (pathway c': aOR, 0.968; 95% CI, 0.939–0.997; P<0.001).
Table 3.
Proportion of Association of Age With Complete Aneurysm Occlusion Mediated by Device‐related Technical Events
| Pathways * | Unadjusted | Adjusted † | ||||
|---|---|---|---|---|---|---|
| Measure | Value (95% CI) | P value | Measure | Value (95% CI) | P value | |
| a | OR | 1.081 (1.031–1.132) | <0.001 ‡ | aOR | 1.088 (1.036–1.142) | <0.001 ‡ |
| b | OR | 0.296 (0.124–0.710) | <0.001 ‡ | aOR | 0.265 (0.106–0.661) | 0.004 ‡ |
| c | OR | 0.958 (0.930–0.987) | <0.001 ‡ | aOR | 0.962 (0.934–0.990) | <0.001 ‡ |
| c' | OR | 0.963 (0.935–0.993) | 0.017 ‡ | aOR | 0.968 (0.939–0.997) | <0.001 ‡ |
aOR indicates adjusted odds ratio; and OR, odds ratio.
Path a, the association between the patients’ ages (covariate of interest) and intraprocedural technical events (mediator); path b, the association between complete aneurysm occlusion (outcome) and the mediator; path c, the association between the covariate of interest and the outcome; path c', the association between the covariate of interest and the outcome, controlling for the mediator.
Adjusted for the diameter of the implanted flow diverter and the aneurysm neck width.
Statistical significance P<0.05.
In the mediation analysis, all the pathways were satisfied, and the magnitude of the effect of the covariate of interest in pathway c (absolute value of regression coefficient, 0.042) was larger than in pathway c' (absolute value of regression coefficient: 0.036). We found a partial mediation of intraprocedural technical events on the association of age with complete aneurysm occlusion due to the maintained association between age and occlusion after controlling for the mediator. After adjusting for confounders, including the diameter of the implanted FD and aneurysm neck width, the mediator (intraprocedural technical events) explained 16.3% of the association between patient age and aneurysm occlusion. In the unadjusted model, this proportion was 13.6%.
Finally, after adjusting for the diameter of the FD and neck width, we estimated the probability of aneurysm occlusion across increasing patient ages stratified by the occurrence or not of intraprocedural technical events (Figure 3). The predicted probability of occlusion in a simulated 59.4‐year‐old patient was significantly lower when intraprocedural technical events were encountered (64.3% with events versus 85.4% no events, P = 0.02).
Figure 3.
Predicted probability of occlusion by age. Model fitted for patients without (green line) or with (red line) intraprocedural technical events. PPO indicates predicted probability of occlusion. * Simulated 59.4‐year‐old patient. After adjusting for flow diverter diameter and aneurysm neck width, the PPO was significant when no technical events were encountered.
Discussion
Our pooled analysis from the SCENT and SESSIA studies, including patients with unruptured saccular aneurysms in the internal carotid artery treated with Streamline, showed that intraprocedural technical events partially mediate the relationship between age and complete aneurysm occlusion at 1‐year follow‐up. A 16.3% of the aneurysm occlusion could be attributed to the effect of technical events during the procedure, which suggests that there are additional mechanisms that affect the effectiveness of flow diversion in achieving complete occlusion.
Previously published studies have focused on identifying the predictors of aneurysm occlusion after FD implantation. 14 , 24 , 25 Although several patient‐ and aneurysm‐related factors have been identified, age has been consistently reported as a predictor of incomplete occlusion and has concordance with our results. 24 The detrimental effect of increasing age can arise from different physiological and anatomical characteristics that influence the implantation, safety, and effectiveness of FDs. Considering vessel tortuosity, it increases with age and has been associated with aneurysm presence in any of the intracranial segments. 26 A successful FD implantation relies mainly on the delivery system (which should provide safe and stable access to the target site) and device. However, despite the evolution in design and technology of FDs and their delivery systems, the amount of vessel tortuosity has been shown to influence the procedural complexity and technical success of FDs. 27 The fact that older patients have increased tortuosity of the cerebral vasculature certainly influences the implantation and effectiveness of FDs. This is supported by our results demonstrating that intraprocedural technical events partially mediate the association between age and occlusion. However, multiple factors related to the rheology of the blood and the integrity of the vessels further contribute to achieving final occlusion after FD implantation.
The occurrence of intraprocedural technical events is often associated with factors such as challenging vascular anatomies, complex aneurysmal locations and morphologies, limited neurointerventionalist experience, and inadequate selection/performance of the delivery systems or devices. 28 Furthermore, due to the mechanism of action of FDs, their performance is influenced directly by the position of implantation (amount of aneurysm neck coverage) and the degree of wall apposition. 29 As discussed previously, in aging patients, the vascular tortuosity tends to increase, and the platelets and endothelium function differently. Taking all into consideration, defining the complete causal model that explains the effect of flow diversion on aneurysm occlusion is complex and will require the integration of additional mechanisms and factors that are not commonly measured.
A key element to evaluate the efficacy of flow diversion is the assessment of aneurysm occlusion after treatment. To date, several classifications have been proposed to measure the degree of occlusion after flow diversion. 30 However, these classifications have limitations because they were not originally designed to measure the degree of occlusion after flow diversion or require a dynamic evaluation of aneurysm filling, contrast stasis, and parent artery status. Additionally, vessel remodeling and aneurysm thrombosis (theoretical mechanisms of action of FDs) take weeks to months to happen, and it is not uncommon that aneurysms with branches (arising from the sack or neck) or in bifurcations take longer to occlude. 13 Due to these limitations, new occlusion classifications for aneurysms treated with flow diversion have been developed. 13 Furthermore, these classifications aim to introduce the concept of dynamic aneurysm remodeling, which aims to confer protection against rupture (despite persistent aneurysm filling). 30 As shown by our results, unrecognized mechanisms and/or factors may influence the occurrence of complete occlusion. Furthermore, the radiographic modalities used to evaluate complete aneurysm occlusion can introduce non‐negligible limitations, which, to different degrees, influence the identification of complete occlusion after flow diversion.
Limitations
Our study has several limitations. First, as a post hoc analysis, it carries inherent limitations related to retrospective data interpretation. A significant limitation is the categorization of aneurysm location, as we grouped petrous and cavernous aneurysms due to differences in anatomical classifications between the SCENT trial and the SESSIA study. Second, we only included cases treated with the Streamline device, which restricts the generalizability of our findings to other FDs. However, despite the introduction of newer devices with improved delivery systems, the occlusion rates and reported technical events are comparable to the latest data on the Pipeline Flex device. Third, only 40% of the obliteration rates (outcome) and intraprocedural technical events (mediator) were adjudicated by the core lab, introducing potential heterogeneity. However, the rates observed in both real‐world registries and single‐arm trials remain comparable. Fourth, although we controlled for potential confounding variables in the mediation analysis, other unmeasured confounders, such as aneurysm location and branching location, may still exist to explain the remaining variance. Theoretically, unmeasured confounders must not exist between parameters in the hypothetical causal mode (Figure 2) to establish a trustworthy mediation. Therefore, considering that there are additional factors that influence the final complete aneurysm occlusion, the interpretation of our findings must be made carefully. Additionally, Baron and Kenny's mediation model has several limitations, including assumptions about the linearity of relationships, the necessity of temporal sequencing between the independent and dependent variables, and the dependence on sample size, which may affect the robustness of our estimates. 31 , 32 Fifth, our analysis lacks information on the morphological characteristics of the vessels (such as tortuosity index and presence of atherosclerotic plaque) and biological factors, like age, which may influence re‐endothelialization and vessel remodeling. As a result, we are unable to accurately assess the impact of these factors on final aneurysm occlusion rates. Finally, due to the low frequency of individual intraprocedural technical events, we created a composite mediator assuming a shared direction of effect based on theoretical rationale and operator experience; however, we were unable to determine the independent weight or directional consistency of each event, limiting individual understanding of each event within the composite.
Conclusion
In this pooled, patient‐level mediation analysis, intraprocedural technical events accounted for a significant portion (16%) of the relationship between increasing age and reduced rates of complete aneurysm occlusion following flow diversion. This finding highlights the complex interplay between patient age, procedural challenges, and the effectiveness of flow diversion. Identifying additional factors that mediate or modify aneurysm occlusion rates—beyond technical events—could provide critical insights for improving procedural strategies, particularly in older patients. Future studies should focus on further elucidating these mechanisms to optimize outcomes and enhance the safety and efficacy of flow diversion therapy.
Conflict of Interest Statement
J. Vivanco‐Suarez – None.
A. Rodriguez‐Calienes – None.
Y. Lu – None.
R. Hanel – None.
J.A. Singer – Consulting fees: Stryker Neurovascular, Medtronic. Honoraria: Cerenovus, Penumbra, Nico.
K. Bekelis – None.
K. Javed – None.
D.J. Altschul – Consulting fees: Microvention, Stryker Neurovascular.
J.T. Fifi – Consulting fees: Microvention, Stryker Neurovascular, Cerenovus, Penumbra. DSMB lead: MIVI medical. Board member of the Society of NeuroInterventional Surgery. Member of the editorial board: Journal of Neurointerventional Surgery. Stock: Imperative Care.
S. Matsoukas – None.
P.M. Meyers – None.
J. Cooper – None.
F. Al‐Mufti – Consulting fees: Stryker Neurovascular, Penumbra, Cerenovus.
B. Gross – None.
B. Jankowitz – None.
P.T. Kan – consultant for Stryker, Imperative Care, Cerenovus, and Microvention; research grant from NIH, Siemens, Joe Niekro, and Medtronic; Journal of NeuroInterventional Surgery editorial board.
M. Hafeez – None.
M. Malaga – None.
E. Orru – None.
M. Galecio‐Castillo – None.
A.L. Coon – None.
A.K. Wakhloo – Grants: Philips, Medtronic fellowship grant, Microbot. Consulting fees: Stryker Neurovascular, Cerenovus, Philips, Microbot. Scientific committee member: Surpass Evolve Study. Leadership: Deinde Med, Neurofine, InNeuroCo., Prometheus. ThrombX, NovaSignal, Neurostream. Stock: Corvista, Neurostream, Medtronic RIST, Prometheus, InNeuroCo., ThrombX, NovaSignal, Neurofine, Neurostream, Hyperion.
S. Ortega‐Gutierrez – Grants: NIH‐NINDS (R01NS127114‐01, RO3NS126804‐01), Stryker, Medtronic, Microvention, Methinks, Viz.ai. Consulting fees: Medtronic, Stryker Neurovascular.
Funding
Stryker funded this study through an investigator‐initiated grant (Grant/Award Number: 2021‐HEM‐034).
Disclosures
Johanna Fifi is an Associate Editor for S:VIN and was not involved in the handling or final disposition of this article. Disclosures provided by Johanna Fifi in compliance with American Heart Association's annual Journal Editor Disclosure Questionnaire are available at https://www.ahajournals.org/editor‐coi‐disclosures. Bradley Gross is an Associate Editor for S:VIN and was not involved in the handling or final disposition of this article. Disclosures provided by Bradley Gross in compliance with American Heart Association's annual Journal Editor Disclosure Questionnaire are available at https://www.ahajournals.org/editor‐coi‐disclosures. Brian Jankowitz is an Associate Editor for S:VIN and was not involved in the handling or final disposition of this article. Disclosures provided by Brian Jankowitz in compliance with American Heart Association's annual Journal Editor Disclosure Questionnaire are available at https://www.ahajournals.org/editor‐coi‐disclosures. Santiago Ortega‐Gutierrez and Ricardo A. Hanel serve on the Editorial Board of S:VIN. Editorial Board Members are not involved in the handling or final disposition of submissions.
Supporting information
Table S1: Definitions of Intraprocedural Technical Events Involving Flow Diverter Deployment
Acknowledgments
None.
This manuscript was sent to Dr. Andrei V. Alexandrov, Guest Editor, for review by expert referees, editorial decision, and final disposition.
References
- 1. Houdart E. Commentary about a 20th meta‐analysis. J Neurointerv Surg. 2021;13:e19. 10.1136/neurintsurg-2020-017234 [DOI] [PubMed] [Google Scholar]
- 2. Brinjikji W, Murad MH, Lanzino G, Cloft HJ, Kallmes DF. Endovascular treatment of intracranial aneurysms with flow diverters: a meta‐analysis. Stroke. 2013;44:442–447. 10.1161/strokeaha.112.678151 [DOI] [PubMed] [Google Scholar]
- 3. Arrese I, Sarabia R, Pintado R, Delgado‐Rodriguez M. Flow‐diverter devices for intracranial aneurysms: systematic review and meta‐analysis. Neurosurgery. 2013;73:193–199; discussion 199–200. 10.1227/01.neu.0000430297.17961.f1 [DOI] [PubMed] [Google Scholar]
- 4. Houdart E. Meta‐analysis as a symptom: the example of flow diverters. AJNR Am J Neuroradiol. 2020;41:E51. 10.3174/ajnr.A6594 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Chancellor B, Raz E, Shapiro M, Tanweer O, Nossek E, Riina HA, Nelson PK. Flow diversion for intracranial aneurysm treatment: trials involving flow diverters and long‐term outcomes. Neurosurgery. 2020;86:S36–S45. 10.1093/neuros/nyz345 [DOI] [PubMed] [Google Scholar]
- 6. Dandapat S, Mendez‐Ruiz A, Martinez‐Galdamez M, Macho J, Derakhshani S, Foa Torres G, Pereira VM, Arat A, Wakhloo AK, Ortega‐Gutierrez S. Review of current intracranial aneurysm flow diversion technology and clinical use. J Neurointerv Surg. 2021;13:54–62. 10.1136/neurintsurg-2020-015877 [DOI] [PubMed] [Google Scholar]
- 7. Becske T, Kallmes DF, Saatci I, McDougall CG, Szikora I, Lanzino G, Moran CJ, Woo HH, Lopes DK, Berez AL, et al. Pipeline for uncoilable or failed aneurysms: results from a multicenter clinical trial. Radiology. 2013;267:858–868. 10.1148/radiol.13120099 [DOI] [PubMed] [Google Scholar]
- 8. Roy D, Milot G, Raymond J. Endovascular treatment of unruptured aneurysms. Stroke. 2001;32:1998–2004. 10.1161/hs0901.095600 [DOI] [PubMed] [Google Scholar]
- 9. Darsaut TE, Gentric JC, McDougall CM, Gevry G, Roy D, Weill A, Raymond J. Uncertainty and agreement regarding the role of flow diversion in the management of difficult aneurysms. AJNR Am J Neuroradiol. 2015;36:930–936. 10.3174/ajnr.A4201 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Adeeb N, Moore JM, Wirtz M, Griessenauer CJ, Foreman PM, Shallwani H, Gupta R, Dmytriw AA, Motiei‐Langroudi R, Alturki A, et al. Predictors of incomplete occlusion following pipeline embolization of intracranial aneurysms: is it less effective in older patients? AJNR Am J Neuroradiol. 2017;38:2295–2300. 10.3174/ajnr.A5375 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Bender MT, Colby GP, Lin LM, Jiang B, Westbroek EM, Xu R, Campos JK, Huang J, Tamargo RJ, Coon AL. Predictors of cerebral aneurysm persistence and occlusion after flow diversion: a single‐institution series of 445 cases with angiographic follow‐up. J Neurosurg. 2018;130:259–267. 10.3171/2017.11.Jns171738 [DOI] [PubMed] [Google Scholar]
- 12. Guédon A, Thépenier C, Shotar E, Gabrieli J, Mathon B, Premat K, Lenck S, Degos V, Sourour N, Clarençon F. Predictive score for complete occlusion of intracranial aneurysms treated by flow‐diverter stents using machine learning. J Neurointerv Surg. 2021;13:341–346. 10.1136/neurintsurg-2020-016748 [DOI] [PubMed] [Google Scholar]
- 13. Hanel RA, Monteiro A, Nelson PK, Lopes DK, Kallmes DF. Predictors of incomplete aneurysm occlusion after treatment with the Pipeline Embolization Device: PREMIER trial 1 year analysis. J Neurointerv Surg. 2022;14:1014–1017. 10.1136/neurintsurg-2021-018054 [DOI] [PubMed] [Google Scholar]
- 14. Maragkos GA, Ascanio LC, Salem MM, Gopakumar S, Gomez‐Paz S, Enriquez‐Marulanda A, Jain A, Schirmer CM, Foreman PM, Griessenauer CJ, et al. Predictive factors of incomplete aneurysm occlusion after endovascular treatment with the pipeline embolization device. J Neurosurg. 2019;132:1598–1605. 10.3171/2019.1.Jns183226 [DOI] [PubMed] [Google Scholar]
- 15. Kühn AL, Kan P, Henninger N, Srinivasan V, de Macedo Rodrigues K, Wakhloo AK, Gounis MJ, Puri AS. Impact of age on cerebral aneurysm occlusion after flow diversion. J Clin Neurosci. 2019;65:23–27. 10.1016/j.jocn.2019.04.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Gennaro G, Ménard C, Michaud S‐É, Rivard A. Age‐dependent impairment of reendothelialization after arterial injury. Circulation. 2003;107:230–233. 10.1161/01.CIR.0000050652.47145.4C [DOI] [PubMed] [Google Scholar]
- 17. Xia Y, Yang Y. RMSEA, CFI, and TLI in structural equation modeling with ordered categorical data: the story they tell depends on the estimation methods. Behav Res Methods. 2019;51:409–428. 10.3758/s13428-018-1055-2 [DOI] [PubMed] [Google Scholar]
- 18. Vivanco‐Suarez J, Mendez‐Ruiz A, Farooqui M, Bekelis K, Singer JA, Javed K, Altschul DJ, Fifi JT, Matsoukas S, Cooper J, et al. Safety and efficacy of the surpass streamline for intracranial aneurysms (SESSIA): a multi‐center US experience pooled analysis. Interv Neuroradiol. (2023);29:589–598. 10.1177/15910199221118148 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Meyers PM, Coon AL, Kan PT, Wakhloo AK, Hanel RA. SCENT trial. Stroke. 2019;50:1473–1479. 10.1161/strokeaha.118.024135 [DOI] [PubMed] [Google Scholar]
- 20. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61:344–349. 10.1016/j.jclinepi.2007.11.008 [DOI] [PubMed] [Google Scholar]
- 21. Baron RM, Kenny DA. The moderator‐mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51:1173–1182. 10.1037//0022-3514.51.6.1173 [DOI] [PubMed] [Google Scholar]
- 22. Preacher KJ, Hayes AF. SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behav Res Methods Instrum Comput. 2004;36:717–731. 10.3758/BF03206553 [DOI] [PubMed] [Google Scholar]
- 23. Barnard J, DB Rubin. Small‐sample degrees of freedom with multiple imputation. Biometrika. 1999;86:948–955. [Google Scholar]
- 24. Adeeb N, Moore JM, Wirtz M, Griessenauer CJ, Foreman PM, Shallwani H, Gupta R, Dmytriw AA, Motiei‐Langroudi R, Alturki A, et al. Predictors of incomplete occlusion following pipeline embolization of intracranial aneurysms: is it less effective in older patients? Am J Neuroradiol. 2017;38:2295. 10.3174/ajnr.A5375 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Jabbour P, Chalouhi N, Tjoumakaris S, Gonzalez LF, Dumont AS, Randazzo C, Starke RM, Hasan D, Chitale R, Singhal S, et al. The Pipeline Embolization Device: learning curve and predictors of complications and aneurysm obliteration. Neurosurgery. 2013;73:113–120; discussion 120. 10.1227/01.neu.0000429844.06955.39 [DOI] [PubMed] [Google Scholar]
- 26. Kliś KM, Krzyżewski RM, Kwinta BM, Stachura K, Gąsowski J. Tortuosity of the internal carotid artery and its clinical significance in the development of aneurysms. J Clin Med. 2019;8:237. 10.3390/jcm8020237 [DOI] [Google Scholar]
- 27. Lin LM, Colby GP, Jiang B, Uwandu C, Huang J, Tamargo RJ, Coon AL. Classification of cavernous internal carotid artery tortuosity: a predictor of procedural complexity in Pipeline embolization. J Neurointerv Surg. 2015;7:628–633. 10.1136/neurintsurg-2014-011298 [DOI] [PubMed] [Google Scholar]
- 28. Al‐Mufti F, Cohen ER, Amuluru K, Patel V, El‐Ghanem M, Nuoman R, Majmundar N, Dangayach NS, Meyers PM. Bailout strategies and complications associated with the use of flow‐diverting stents for treating intracranial aneurysms. Interv Neurol. 2020;8:38–54. 10.1159/000489016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Aquarius R, de Korte A, Smits D, Gounis M, Verrijp K, Driessen L, Leenders W, de Vries J. The importance of wall apposition in flow diverters. Neurosurgery. 2019;84:804–810. 10.1093/neuros/nyy092 [DOI] [PubMed] [Google Scholar]
- 30. Hanel RA, Cortez GM, Lopes DK, Saatci I, Cekirge HS. Brain aneurysm and parent vessel remodeling after flow diversion treatment: a proposed modification for Cekirge‐Saatci classification (mCSC). J Neurointerv Surg. 2023;15:102–104. 10.1136/jnis-2022-019757 [DOI] [PubMed] [Google Scholar]
- 31. Zhao X, Lynch JG, Jr , Chen Q. Reconsidering Baron and Kenny: myths and truths about mediation analysis. J Consum Res. 2010;37:197–206. 10.1086/651257 [DOI] [Google Scholar]
- 32. Imai K, Keele L, Tingley D. A general approach to causal mediation analysis. Psychol Methods. 2010;15:309–334. 10.1037/a0020761 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Definitions of Intraprocedural Technical Events Involving Flow Diverter Deployment