Adjunctive Coiling in Flow Diverter Treatment Does Not Prevent Delayed Rupture: A Nationwide Survey
Article information
Abstract
Purpose
Delayed rupture (DR) can occur even after successful deployment of flow diverters (FDs). Although coiling is often added to reduce the risk of rupture in high-risk intracranial aneurysms, its effectiveness remains unproven. To assess current practice in Japan, a questionnaire was distributed to evaluate the effect of coiling on rupture prevention.
Materials and Methods
A retrospective survey was sent to 124 institutions with qualified FD practitioners, receiving 76 responses (61%). A total of 5,527 patients treated with FDs were included in the study, and clinical records of 5,211 aneurysms were analyzed.
Results
DRs (excluding intraprocedural accidents) occurred in 36 cases (0.7%). Of 1,286 aneurysms treated with FD and coiling, 9 (0.7%) ruptured. Ruptured aneurysms in the FD with coiling group were located in the supraclinoid segment (4 cases), paraclinoid internal carotid artery (ICA) (3 cases), and basilar artery (2 cases), of which 7 were classified as giant (diameter >20 mm). In the FD-alone group, 55% (15/27) of ruptured aneurysms were located in the cavernous ICA. Time to rupture ranged from 1 to 2,220 days with no significant difference between groups, except for 1 exceptionally delayed case. Outcomes were worse in the FD with coiling group, where 67% (6/9) had a modified Rankin Scale score of 6. The DR rate was identical between the FD-alone and FD with coiling groups.
Conclusion
Coiling does not appear to prevent rupture after FD treatment. Aneurysms treated with coiling tended to be larger, to rupture earlier, and to have worse outcomes, probably due to selection bias. Routine loose coiling may provide psychological reassurance but lacks proven efficacy as a preventive measure.
INTRODUCTION
A flow diverter (FD) is a device designed to cover the neck of an aneurysm with a mesh, promoting thrombus formation within the aneurysm by streamlining blood flow in the parent artery and inducing stagnation inside the aneurysm. This process naturally leads to thrombus organization and subsequent resorption, resulting in the shrinkage of the aneurysm without the need for additional intervention. As such, FDs have been considered to provide complete protection against rupture, and they also offer the added benefit of alleviating the mass effect of large aneurysms [1,2].
However, clinical reports have increasingly documented cases of delayed rupture (DR) in some aneurysms following FD deployment [3,4]. Possible causes include rapid thrombus formation within the aneurysm, hemodynamic changes leading to intra-aneurysmal pressure imbalances, or the formation of localized jets impacting the aneurysm wall [5]. To enhance intra-aneurysmal flow disruption and promote early thrombus formation, a technique involving the placement of a few coils within the aneurysm as “seeds” has been advocated [6].
Despite these recommendations, the rupture-prevention efficacy of placing a small number of coils, with an embolization rate of approximately 10%, has not been conclusively established. Additionally, there is no consensus on the optimal coil volume or target application sites. A decade has passed since FDs were introduced into Japan’s healthcare system, making FD treatment widely accessible. This study retrospectively investigates the incidence of DR following FD treatment in Japan and evaluates the effectiveness of coiling in preventing such ruptures.
MATERIALS AND METHODS
A survey was conducted among facilities with neurointerventionists qualified to perform FD procedures. The survey focused on the number of FD cases and the incidence of DR. For DR cases, detailed patient and aneurysm information was reviewed. The survey collected data on patient demographics (age, sex), aneurysm characteristics (location, size), coil usage, the number of coils in combination cases, time to rupture, additional treatments after rupture, and final outcomes. A total of 124 institutions were approached, and responses were received from 76 facilities (response rate: 61%).
A total of 5,527 FD-treated cases were enrolled, of which 5,211 cases with complete clinical records were included in the analysis. Cases were divided into 2 groups for comparison: those treated with FD alone (no coil group: group A) and those treated with FD combined with coiling (coil group: group C). Statistical differences between the 2 groups were evaluated using Student’s t-test.
RESULTS
Of the 5,211 cases, 3,925 were in group A and 1,286 were in group C. DR occurred in 36 cases (group A: 27 cases, group C: 9 cases), with an incidence of 0.7% in both groups. Detailed information on the rupture cases is shown in Table 1. The mean age of the patients was 63.8 years (range: 30–86 years) and 83% (30 cases) were female. There were no significant differences in age or sex between the groups.
Summary of DR cases
However, the location of the aneurysms differed between the 2 groups. In group A, a higher proportion of aneurysms were located in the carotid cavernous sinus, whereas group C had no carotid cavernous aneurysms and a higher proportion of paraclinoid aneurysms. Seven cases of basilar artery aneurysms were large (group A: 5 cases, group C: 2 cases), all located on the basilar trunk.
Giant aneurysms (diameter >20 mm) accounted for 18 cases (50%), with a higher proportion in group C. Small aneurysms (<10 mm) ruptured in only 1 case. The time to rupture varied widely, ranging from the day after surgery to 2,220 days postoperatively. Median values suggested that group A had later ruptures compared to group C, although this was influenced by an outlier in group A with a rupture occurring 6 years postoperatively. Excluding this outlier, the mean durations were similar (group A: 56.9 days, group C: 59.0 days). Distribution analysis (Table 2) showed that 53% (19 cases) ruptured within the first month and 94% (34 cases) ruptured within 4 months.
Duration to rupture
Post-rupture treatments included coiling in 11 cases and FD overlap in 5 cases for group A and trapping in 1 case and FD overlap in 3 cases for group C. However, many cases (41% in group A, 55% in group C) received no post-rupture treatment. Clinical outcomes showed that 25% (9 cases) had a complete recovery, but poor outcomes were common (modified Rankin Scale 3–6: 58% [21 cases]). Death occurred in 12 cases in group A and 6 in group C, with a particularly high mortality rate in group C (67%). Mortality rates were particularly high for intracranial aneurysms (71%) and basilar artery aneurysms (86%).
Criteria for coil use varied widely among institutions. Some centers did not use coils at all, while others used coils in more than half of their cases. Common criteria for coil use included symptomatic large/giant aneurysms, narrow necks, or the presence of jet flow. Most centers used 3–5 coils, and some based their criteria on embolization rates (10–15%).
DISCUSSION
One of the purposes of adding coils to FDs is to reduce turbulent flow within the aneurysm, creating a stagnant state that promotes early thrombosis.7-12 Numerous studies have reported higher rates of complete aneurysm occlusion when coils are used compared to when coils are not employed. However, other studies have shown that, despite higher occlusion rates, using coils can lead to more complications, particularly ischemic complications.7,8 Additionally, there are reports of increased perioperative stroke-related or hemorrhagic complications in cases where coils were used [9,10].
Another expectation from using coils is the prevention of DR following FD treatment, which has been widely discussed in the literature [3,4,7,13-16]. DR is believed to be caused by factors such as: 1) intrasaccular pressure elevation due to flow stagnation [5,17], 2) local jet formation due to heterogeneous thrombus formation [5,18], and 3) inflammation of the aneurysm wall caused by intrasaccular thrombosis [19,20]. The criteria for coil application generally include giant aneurysms, highly symptomatic aneurysms, acutely expanded aneurysms, and aneurysms with wall enhancement. However, these criteria often do not take into account the thinning or weak areas of the aneurysm wall, leading to a cautious or preemptive approach, akin to a “precautionary measure.” There is no strict standard for the number of coils used, as it is believed that a few coils can provide prevention [18]. In this study, 3–5 coils, or an embolization rate of 10–15%, was most commonly used, but these criteria are not evidence-based, and full prevention has not been achieved. Moreover, studies show no significant difference in prevention between loose packing and dense packing methods [10]. In this study, rupture rates were 0.7% in both the coil group and the non-coil group, aligning with other reports ranging from 0.7% to 2.0% [7,13,14]. This is not significantly lower than the DR rate (1.2–1.5%) [21,22] after standard coil embolization.
Post-rupture treatment varied; in cases of carotid cavernous aneurysms, additional coils were often used, but these cases were typically treated via transvenous sinus packing, similar to traditional treatments for direct carotid-cavernous fistulas, not through trans-cell approaches from the FD strut. For most other ruptured aneurysms, no treatment was administered, likely because the condition had already progressed irreversibly. Rupture outcomes were poor, which aligns with previous reports [13,16]. Notably, intracranial aneurysms, particularly basilar artery aneurysms, had very poor outcomes. In treatment decision-making, including indications, further consideration and reevaluation are necessary.
Comparisons of outcomes between neck-bridge stent cases and coil-using cases are difficult due to different aneurysm profiles. However, the question remains whether current inadequate coil packing leads to sufficient thrombus formation and rupture prevention. Further investigation is required.
This study has the following limitations. First, it is a simple retrospective review of real-world conditions, not a randomized controlled trial. Second, the number of cases at the same site is unknown, making it difficult to assess site-specific rupture rates. Third, aneurysm shape and symptoms already determine coil placement to some extent, and physician decisions introduce bias, leading to inconsistent criteria-based comparisons. And fourth, different institutions have different policies regarding coil placement, making it impossible to assess outcomes based on a standardized protocol.
To overcome these issues and conduct accurate and useful efficacy evaluations, future studies should include uniform coil application criteria and standardized coil usage across multiple centers. Furthermore, perioperative complications arising from simultaneous use of multiple catheters in coil-assisted procedures need to be considered.
CONCLUSION
Coil usage in FD treatment is widely applied to prevent DR in high-risk aneurysms, but its effectiveness is not adequate. In this survey, there were no differences in DR rates between FD alone and FD with coiling groups. Coiling does not appear to be a foolproof method for preventing rupture after FD treatment. To avoid devastating DRs, it is necessary to explore how effective coil use is and what embolization methods should be optimized to achieve better results.
SUPPLEMENTARY MATERIALS
Supplementary materials related to this article can be found online at https://doi.org/10.5469/neuroint.2025.00129.
Delayed rupture cases treated with flow diverter alone.
Delayed rupture cases treated with flow diverter+coiling.
Notes
Acknowledgments
The authors appreciate all Japanese investigators of expert physicians for flow diverters who collaborated this study.
Fund
None.
Ethics Statement
This study is approved by the ethical committee of Aichi Medical University (approved no. 2019-167). We anonymized patient information that could identify an individual.
Conflicts of Interest
The authors have no conflicts to disclose.
Author Contributions
Concept and design: SM. Analysis and interpretation: SM, RK. Data collection: SM. Writing the article: SM. Critical revision of the article: SM. Statistical analysis: SM, RK. Obtained funding: none. Overall responsibility: SM.