Balloon inflation predicts recanalization of intracranial aneurysms treated with coiling alone

Abstract

Background

Endovascular coil embolization is a common treatment for intracranial aneurysms, but aneurysm recanalization remains a significant problem that may necessitate retreatment. This study aimed to identify patient, aneurysm, and procedural factors associated with recanalization in aneurysms treated exclusively with coil embolization.

Methods

This single center retrospective study assessed intracranial aneurysms treated with coiling-only between 2017 and 2022. Follow-up imaging was reviewed for recanalization with occlusion status graded via a modified Raymond–Roy classification. Univariate analysis assessed the association of clinical, morphological, and procedural factors with clinical complication, aneurysm occlusion, and recanalization status. Stepwise multivariable logistic regression was performed to identify independent predictors of aneurysm recanalization.

Results

Of 163 initially treated aneurysms, 142 were analyzed in complete case analysis for clinical outcomes. Complications occurred in 8 patients and were associated with larger aneurysm neck sizes (3.83 mm vs. 2.92 mm, p = 0.024), increased incidence of coil herniation (63% vs. 10%, p = 0.001), and greater number of coils used per aneurysm (7.13 coils vs. 4.64 coils, p = 0.028). Follow-up angiography in 122 aneurysms showed adequate occlusion in 116 (95.1%) and recanalization in 11.5%. Recanalization was more frequent in aneurysms that had larger initial volumes and in those without balloon inflation during coil deployment (both p < 0.05). On multivariable analysis, balloon inflation during coil deployment was the only independent predictor of aneurysm recanalization (OR 0.18; 95% CI 0.05–0.69; p = 0.009).

Discussion

In this single-center cohort of coiling-only treated aneurysms, the use of the balloon remodeling technique was strongly associated with durable aneurysm occlusion, reducing the odds of aneurysm recanalization. These findings support the routine use of balloon assistance in wide-neck and large aneurysms to achieve complete, stable occlusion and reduce the need for retreatment.

Introduction

The treatment of intracranial aneurysms can be broadly divided into microsurgery and endovascular therapy. With the exception of complex aneurysms that are refractory to endovascular treatment, or those involving the middle cerebral artery (MCA), aneurysm care has increasingly shifted to favor endovascular approaches. ¹ Though newer endovascular treatment modalities such as flow diversion are gaining support, traditional endovascular coiling remains widely used when anatomy is favorable.^2–4 This technique has been shown to be effective in occluding intracranial aneurysms by densely packing detachable coils within the aneurysm sac and demonstrated a survival benefit relative to microsurgery in patients with ruptured aneurysms in the International Subarachnoid Aneurysm Trial (ISAT).^5–7 In anatomically challenging cases where coiling without assistance may not be feasible, adjunctive techniques such as stent-assisted coiling or balloon-assisted coiling (BAC) can also be used to facilitate aneurysm occlusion. ⁸ Despite its use as a mainstream intervention, a known limitation of coiling without stenting is the risk of aneurysm recurrence.^9,10 In some cases, aneurysm recurrence may require retreatment, which tends to be less frequent with microsurgical clipping or stent-assisted coiling (SAC). In these cases, retreatment with additional coils or the use of other adjuvant endovascular or surgical intervention is often warranted.^11–15 Consequently, it is important to better understand the factors that may lead to recanalization in patients treated with only endovascular coiling.

Several studies have described independent factors underlying recanalization risk. These include, but are not limited to, age and smoking status, aneurysm diameter and neck size, aneurysm location, rupture status, post-procedure modified Raymond-Roy (RR) occlusion classification and postoperative time-of-flight magnetic resonance angiography classification.^16–19 Yet, many procedural and morphologic factors remain uninvestigated. Thus, this study aimed to investigate other possible predictors of recanalization among a cohort of patients treated with non-stent assisted endovascular therapy at our institution.

Materials and methods

Ethical considerations

This study was approved by the local Institutional Review Board and the need for informed consent was waived by the local IRB due to the retrospective nature of this study. Wherever possible, reporting for this study was conducted in accordance with the Strengthening Reporting of Observational Studies in Epidemiology (STROBE) checklist (Supplemental 1).

Study design

155 consecutive patients undergoing only endovascular coiling for treatment of an intracranial aneurysm without stenting or flow diversion between August 2017 and April 2022 in a single center were retrospectively reviewed for this study. Clinical and imaging data were collected from the electronic medical record including demographic information, clinical presentation, aneurysm characteristics, antiplatelet regimen, and long-term clinical outcomes. Inclusion criteria included all patients who received coil embolization as the sole modality of treatment whose aneurysms could be successfully segmented with the Siemens aneurysm analysis software. Exclusion criteria included patients treated with stents or flow diverters.

All patients were reviewed by a combined neurovascular team that considered factors including aneurysm location, patient age, and need for antiplatelet therapy prior to treatment. Treatment decisions were also based on interventionalist and patient preference. Standard follow-up occurred with digital subtraction angiography (DSA) at 6 months and 18 months (Figure 1).

Figure 1.

Flowchart of exclusion criteria. A flowchart depicting inclusion and exclusion of participants for analysis within the study is shown.

Aneurysm measurements

Aneurysm measurements were retrieved from 3D aneurysm reconstructions using the Siemens analysis software. In addition, several morphologic parameters were retrieved, including 2D measured dome-to-neck ratio and aneurysm volume. Aneurysm volume was calculated using Siemens volumetric aneurysm analysis and coil packing density was calculated using Angiocalc. RR occlusion classification was used to assess follow-up occlusion status. Recanalization was defined as an increase in the RR occlusion class of at least one grade by time of last follow-up.

Statistical analysis

Continuous variables are presented as mean ± SD, and categorical variables are presented as number (percent). Variables were utilized in a complete case univariate analysis on a by-outcome basis. Patients undergoing follow-up were divided into recanalization and no recanalization. Aneurysm location was also divided into anterior (e.g., anterior communicating artery [AComm], anterior cerebral artery [ACA], middle cerebral artery [MCA], internal carotid artery [ICA]) and posterior (posterior cerebral artery [PCA], posterior communicating artery [PComm], and other posterior circulation vessels) locations. Aspect ratio was defined as aneurysm height divided by neck width.

Mann-Whitney U-test was utilized for non-normally distributed continuous data, t-test for normally distributed continuous data, chi-squared test for categorical data with cell counts ≥ 5, and Fisher's exact test for categorical data with cell counts < 5. Univariable and multivariable logistic regression models were applied to identify the variables associated with aneurysm occlusion and recanalization. All variance inflation factor (VIF) estimates were less than two, indicating no collinearity. Significant variables (p < 0.10) detected in the univariate models were entered into the multivariate logistic regression model. Wald Type II tests were used to determine overall significance for multifactorial variables for stepwise regression. The results are reported as the odds ratio (OR) and 95% confidence interval (CI). All analyses were performed using RStudio version 4.2 (R Foundation for Statistical Computing). For multivariable regression, a two-tailed p-value < 0.05 was considered statistically significant.

Results

Demographics and aneurysm characteristics

163 aneurysms in 155 patients were initially retrieved for analysis (Figure 1). Complete case analysis assessing post-operative clinical complications identified 142 treated aneurysms. Follow-up was obtained for 122 aneurysms, allowing for assessment of recanalization status (Figure 2). Aneurysms treated and assessed for occlusion status were disproportionately in females (77.5% females vs 22.5% males), with a mean age of 58.9 ± 13.7 years. The most common presentation was recently ruptured aneurysm (88.0%). Among unruptured aneurysms, headache (35.3%), syncope or fall (29.4%), incidental imaging findings (29.4%), and seizure (5.9%) were the most common presenting features. Patients had a variety of comorbidities, including hypertension (59.2%), hyperlipidemia (16.9%), coronary artery disease (2.8%), and diabetes mellitus (8.5%). 52 (36.6%) aneurysms were derived from ACA or AComm, 43 (30.3%) from PCA or PComm, 17 (12.0%) from other posterior circulation vessels, 15 (10.6%) from MCA, and 15 (10.6%) from the ICA (non PCOM). Most aneurysms were solitary (67.6%) and located in the anterior circulation (57.7%), with an average volume of 112 mm³, aspect ratio of 1.83, width of 4.54 mm, height of 5.75 mm, and neck size of 2.97 mm.

Figure 2.

(A). Complete occlusion of ruptured right pericallosal aneurysm following balloon assisted coiling (BAC) and (B). Follow up of BAC -treated ruptured pericallosal aneurysm at 6 months showing continued occlusion.

Procedure characteristics and follow-up

The majority of patients were treated with BAC (92.3%) with intraprocedural balloon inflation during coil deployment being used in 86.6% of cases. Aspirin (19.7%), clopidogrel (4.9%), and other antiplatelet use (6.3%) were uncommon post-procedure. Coil packing density averaged 33.4% ± 10.5%. Assessment of procedural complications identified coil herniation in 19 (13.4%) individuals, other technical complications (e.g., thrombus or intra-operative rupture) in 14 (9.9%), and clinical complication in 8 (5.6%). Post-operative complications included clinical stroke in 8 patients, including death in 2 patients due in part to an index ruptured aneurysm. Treated patients were subsequently followed to assess for aneurysm recanalization, with aneurysm recanalization occurring in 14 (11.5%) of the 122 reobserved patients (Table 1). At follow-up, 91 (74.5%) aneurysms were classified as RR1, 25 (20.4%) as RR2, 4 (3.3%) as RR3a, and 2 (1.6%) as RR3b. Adequate occlusion, defined as RR1 or RR2, was seen in 95.1% of patients and incomplete occlusion in 4.9%, with occlusion rates varying according to patient demographics and aneurysmal characteristics (Tables 1 and 2).

Table 1.

Procedural characteristics.

Variable	Clinical Complications			Occlusion			Recanalization
	No, (N = 134)1	Yes, (N = 8)1	p-value2	Complete, (N = 116)1	Incomplete, (N = 6)1	p-value2	Absent (N = 108)1	Present (N = 14)1	p-value2
Aspirin	26 (19%)	2 (25%)	0.657	21 (18%)	1 (17%)	>0.999	21 (19%)	1 (7.1%)	0.461
Plavix	6 (4.5%)	1 (13%)	0.340	3 (2.6%)	1 (17%)	0.185	4 (3.7%)	0 (0%)	>0.999
Anticoagulants	8 (6.0%)	1 (13%)	0.416	8 (6.9%)	0 (0%)	>0.999	8 (7.4%)	0 (0%)	0.595
Coil Type			0.454			0.081			0.137
Balt	33 (25%)	3 (38%)		30 (26%)	0 (0%)		29 (27%)	1 (7.1%)
Microvention Gel	64 (48%)	2 (25%)		55 (47%)	2 (33%)		47 (44%)	10 (71%)
Other/Mixed	37 (28%)	3 (38%)		31 (27%)	4 (67%)		32 (30%)	3 (21%)
Coil Packing Density	33 (10)	38 (21)	0.975	34 (9.8)	29 (6.6)	0.216	34 (10)	35 (8.0)	0.298
Balloon Used	123 (92%)	8 (100%)	> 0.999	111 (96%)	4 (67%)	0 . 038	103 (95%)	12 (86%)	0.184
Balloon Inflated	115 (86%)	8 (100%)	0.598	104 (90%)	3 (50%)	0.024	98 (91%)	9 (64%)	0.015
Coil Herniation	14 (10%)	5 (63%)	0.001	15 (13%)	2 (33%)	0.196	15 (14%)	2 (14%)	>0.999
Total Coils Used	4.64 (2.67)	7.13 (4.67)	0.028	4.7 (2.9)	8.5 (3.3)	0.004	4.8 (3.0)	5.6 (2.9)	0.181

¹n (%), Mean (SD); ²Fisher's exact test, Pearson's Chi-squared test, Wilcoxon rank sum test; RR: Raymond Roy Occlusion Class.

Table 2.

Demographics and aneurysmal characteristics.

Variable	Clinical Complications			Occlusion			Recanalization
	No, (N = 134)1	Yes, (N = 8)1	p-value2	Complete, (N = 116)1	Incomplete, (N = 6)1	p-value2	Absent (N = 108)1	Present (N = 14)1	p-value2
Sex			0.683			>0.999			>0.999
Female	103 (77%)	7 (88%)		93 (80%)	5 (83%)		87 (81%)	11 (79%)
Male	31 (23%)	1 (13%)		23 (20%)	1 (17%)		21 (19%)	3 (21%)
Age	58 (14)	68 (13)	0.069	58 (13)	58 (12)	0.882
HTN	77 (57%)	7 (88%)	0.141	68 (59%)	4 (67%)	>0.999	65 (60%)	7 (50%)	0.466
HLD	22 (16%)	2 (25%)	0.623	18 (16%)	1 (17%)	>0.999	19 (18%)	0 (0%)	0.123
CAD	3 (2.2%)	1 (13%)	0.209	3 (2.6%)	0 (0%)	>0.999	3 (2.8%)	0 (0%)	>0.999
DM	11 (8.2%)	1 (13%)	0.516	11 (9.5%)	0 (0%)	>0.999	11 (10%)	0 (0%)	0.360
Presenting Symptoms			0.649			>0.999			0.837
Incidental	5 (3.7%)	0 (0%)		3 (2.6%)	0 (0%)		3 (2.8%)	0 (0%)
Syncope/Fall	5 (3.7%)	0 (0%)		4 (3.4%)	0 (0%)		4 (3.7%)	0 (0%)
Seizure	1 (0.7%)	0 (0%)		1 (0.9%)	0 (0%)		1 (0.9%)	0 (0%)
Headache	5 (3.7%)	1 (13%)		6 (5.2%)	0 (0%)		5 (4.6%)	1 (7.1%)
Ruptured	118 (88%)	7 (88%)		102 (88%)	6 (100%)		95 (88%)	13 (93%)
Multiple Aneurysms	43 (32%)	3 (38%)	0.714	38 (33%)	1 (17%)	0.663	35 (32%)	4 (29%)	>0.999
Posterior Circulation	55 (41%)	5 (63%)	0.283	47 (41%)	3 (50%)	0.688	41 (38%)	9 (64%)	0.060
Height	5.58 (2.52)	8.64 (7.21)	0.075	5.6 (3.0)	9.6 (3.0)	0 . 003	5.6 (3.1)	6.9 (3.3)	0.119
Neck	2.92 (1.36)	3.83 (1.25)	0.024	2.9 (1.2)	5.5 (2.6)	0.001	2.9 (1.3)	3.5 (2.1)	0.383
Volume	86 (161)	546 (1220)	0.074	106 (355)	298 (336)	0.004	108 (365)	167 (275)	0.024
Width	4.36 (2.29)	7.63 (6.87)	0.206	4.4 (2.7)	7.4 (4.7)	0.021	4.4 (2.9)	5.4 (2.8)	0.088
Aspect Ratio	1.79 (0.95)	2.47 (1.62)	0.220	1.8 (0.99)	2.0 (1.8)	0.582	1.8 (0.94)	2.4 (1.5)	0.122
Dome Neck Ratio	1.65 (1.35)	1.82 (0.83)	0.292	1.6 (1.4)	1.9 (1.0)	0.745	1.6 (1.4)	1.7 (0.74)	0.347

¹n (%), Mean (SD); ²Fisher's exact test, Pearson's Chi-squared test, Wilcoxon rank sum test; HTN: Hypertension; HLD: Hyperlipidemia; CAD: Coronary Artery Disease; DM: Diabetes Mellitus.

Univariate analysis

Several aneurysm and procedural characteristics differed between outcome groups (Tables 1 and 2). Clinical complications were associated with larger aneurysm neck sizes (3.83 mm vs. 2.92 mm, p = 0.024), increased incidence of coil herniation (63% vs. 10%, p = 0.001), and greater total coils used per aneurysm (7.13 coils vs. 4.64 coils, p = 0.028). Incomplete aneurysm occlusion on follow-up was associated with greater aneurysm height (9.6 mm vs. 5.6 mm, p = 0.003), neck size (5.5 mm vs. 2.9 mm, p = 0.001), width (7.4 mm vs. 4.4 mm, p = 0.021), and overall volume (298 mm³ vs. 106 mm³, p = 0.004), as well as increased total coil usage (8.5 coils vs. 4.7 coils, p = 0.004), decreased balloon use (67% vs. 96%, p = 0.038), and lower rates of intraprocedural balloon inflation during coil deployment (50% vs. 90%, p = 0.024). Lastly, aneurysm recanalization was associated with larger aneurysm volumes (167 mm³ vs. 108 mm³, p = 0.024) and lower rates of intraprocedural balloon inflation during coil deployment (64% vs. 91%, p = 0.015).

Recanalization analysis

Independent predictors of incomplete aneurysm recanalization were evaluated using univariable and stepwise multivariable logistic regression models (Table 3). In univariable regression, balloon inflation during coil deployment was associated with decreased recanalization (OR: 0.18, 95% CI: 0.05–0.69, p = 0.009), while aspect ratio (OR: 1.48, 95% CI: 0.95–2.31, p = 0.069), aneurysm location in posterior circulation (OR: 2.94, 95% CI: 0.95–10.1, p = 0.068), and Microvention Gel coil subtype were identified as possible predictors of increased recanalization. In the final multivariable model, only balloon inflation during coil deployment (OR: 0.18, 95% CI: 0.05–0.69, p = 0.009) was significantly associated with recanalization status.

Table 3.

Univariable and stepwise multivariable predictors of incomplete recanalization Status.

	Univariable Regression			Multivariable Regression
Characteristic	OR	95% CI	p-value	OR	95% CI	p-value
Male Sex	1.13	0.24, 4.01	0.861
Ruptured	1.78	0.31, 33.6	0.593
Coil Type
Balt (Reference)	-	-
Microvention Gel	6.17	1.10, 116	0 . 091
Other/Mixed	2.72	0.33, 56.7	0.398
HTN	0.66	0.21, 2.06	0.468
Other Aneurysm	0.83	0.22, 2.69	0.772
Aspirin	0.32	0.02, 1.74	0.283
Coil Herniation	1.03	0.15, 4.31	0.968
Balloon Use	0.29	0.06, 2.19	0.166
Circulation	2.94	0.95, 10.1	0.068
Balloon Inflation	0.18	0.05, 0.69	0.009	0.18	0.05, 0.69	0.009
Age	0.98	0.94, 1.03	0.425
Aspect Ratio	1.48	0.95, 2.31	0.069
Coil Packing Density	1.01	0.95, 1.07	0.678
Dome Neck Ratio	1.03	0.57, 1.40	0.857
Height	1.1	0.94, 1.28	0.179
Neck	1.24	0.87, 1.71	0.187
Total Number of Coils	1.09	0.91, 1.27	0.323
Volume	1	1.00, 1.00	0.573
Width	1.09	0.92, 1.27	0.258

OR: Odds Ratio; CI: Confidence Interval; HTN: Hypertension.

Discussion

In this retrospective study of 163 intracranial aneurysms treated only with endovascular coiling, we identified several univariate predictors of aneurysmal occlusion, recanalization, and perioperative clinical complications. Adequate occlusion at follow-up was seen in 95.1% of aneurysms, recanalization occurred in 11.5% over the study window, and clinical complications occurred in 5.6% of treated patients. In multivariate analysis, balloon inflation during coil deployment emerged as the only independent protective factor against recanalization (OR 0.18, CI: 0.05–0.69, p = 0.009). These results underscore the clinical impact of the balloon remodeling technique in improving the durability of coil embolization, especially for challenging aneurysm anatomies.

These findings reinforce prior observations regarding aneurysm morphology and recurrence risk. Consistent with large multicenter studies, we find that aneurysm size and neck width correlate strongly with long term durability of the coil mass as aneurysms with greater height, width, neck diameter, and overall volume were significantly more likely to have incomplete initial occlusion and to recur over time. ²⁰ The French nationwide ARETA study, for example, identified aneurysm size ≥10 mm and wide neck anatomy as independent predictors of coiling failure and recanalization. ²⁰ These morphological factors reflect the inherent difficulty of achieving durable occlusion in large or broad-necked aneurysm sacs due to factors like incomplete packing and persistent blood flow entry.

We also evaluated the influence of aneurysm shape indices, noting that aspect ratio showed a trend toward association with recanalization (OR: 1.48, 95% CI: 0.95–2.31, p = 0.069). This result reinforces the clinical importance of determining the aspect ratio of an intracranial aneurysm, as other studies have found high aspect ratios to be predictive of rupture if untreated (p < 0.01). ²¹ Given that this metric is intrinsic to a given aneurysm, the aspect ratio should be used to inform selection of treatment methodologies, such as reducing the favorability of coiling-only endovascular treatment for an aneurysm with a high aspect ratio, rather than as a variable to be manipulated prior to intervention. While aneurysm volume as well as other morphological factors were different between the two groups, they were not significant in the final model. Similarly, aneurysm neck was not different between the groups despite prior studies suggesting otherwise. ²⁰ Numerous techniques have been described to facilitate coiling especially in wide neck (>4 mm) aneurysms, which include dual catheter use, balloon remodeling, the Comaneci device, and stent assisted coiling.^22–24 The use of a balloon allows for temporary flow arrest resulting in local thrombosis. ²⁵ Additionally, multiple coils may be deployed within the aneurysm during inflation, thereby stabilizing the coil mass and forming a conglomerate coil mass to rapidly and stably treat challenging aneurysms with challenging angles or neck sizes. ²⁶

The BAC technique was used in the vast majority (92%) of our cases, reflecting our center's practice of proactive balloon support for aneurysms with complex architecture. Notably, even among these BAC cases, we distinguished those in which the balloon was actually inflated during coil deployment (86.6% of total cases) from those in which the balloon was kept deflated in position but not used. This distinction proved critical as aneurysms where the balloon was inflated had significantly lower rates of incomplete occlusion at treatment and markedly lower recanalization on follow-up. Mechanistically, balloon inflation across the neck during coil deployment likely enables denser coil packing and a more complete neck coverage, reducing residual aneurysm filling and subsequent coil compaction. Previous studies have reported that achieving a high coil packing density (>20–25% aneurysm volume) is key to durable occlusion. ²⁷ The balloon remodeling technique facilitates this by preventing coil loops from herniating into the parent artery and by momentarily blocking inflow during coil placement, thus promoting thrombus formation.

Historically, there have been concerns about the safety of balloon usage, including thromboembolic events.^1,2,28 However, previous studies have found no statistical difference between BAC and coiling-only intervention in terms of thromboembolic events (5.4% vs. 6.2%), morbidity (2.3% vs. 2.2%), and mortality (1.4% vs. 0.9%). ²⁰ A further study also found low complication and recanalization rates in BAC (6.6% thromboembolic events and 4.3% recanalization). ²⁹ Our study corroborates these conclusions by showing that balloon inflation during coil deployment correlates with more complete occlusion (p = 0.024) and lower recanalization risk (p = 0.015).

Another crucial point is that BAC achieves these benefits without the need for permanent implantation or prolonged antithrombotic therapy as opposed to SAC. This is especially pertinent in the context of the ruptured aneurysms that predominated our cohort, as the risks of adjunctive devices must be balanced against the urgency of securing an aneurysm. SAC is an alternative strategy for wide-neck aneurysms that has demonstrated improved long-term occlusion in some studies. A propensity-adjusted comparison in ruptured aneurysms showed SAC yielded a higher initial complete occlusion rate and lower 6-month recanalization rate than coiling alone. ³⁰ Similarly, a meta-analysis reported that SAC roughly halved the recurrence rate compared to coiling without adjuncts (approximately 12–15% vs 25–30%). ³¹ However, these advantages come at the cost of requiring dual antiplatelet therapy and incurring device-related risks, including vessel jailing. Stent assistance also has been associated with a higher incidence of thromboembolic complications – in one series, ∼11.6% with SAC vs 2.4% with BAC as well as an increased risk of hemorrhagic complications in the setting of acute subarachnoid hemorrhage due to necessary antiplatelet use. ³² In our series we avoided routine stent use but still achieved a low recanalization rate (11.5%), comparable to SAC literature, using BAC. This highlights a key clinical implication: that balloon remodeling technique can confer many of the durability benefits of stents without exposing patients to the inherent risks of endoluminal implants. BAC may be considered a primary adjunctive strategy, particularly in cases of ruptured aneurysms or when antiplatelet therapy is contraindicated.

Our study reinforces this safety profile of BAC by providing additional data on clinical complications. We observed an overall procedure-related complication rate of 5.6%, in line with the complication rates reported in contemporary endovascular series for both ruptured and unruptured aneurysms. ³³ Most of our complications were thromboembolic events leading to neurological deficit, with a smaller proportion due to intraoperative aneurysm perforation. Notably, we found that complications were associated with aneurysm and procedure complexity as larger neck size, higher total number of coils, and identification of coil herniation into the parent vessel, which were risk factors for an adverse event. Wide-neck aneurysms require more complex catheter and balloon manipulations and carry a risk of coil instability, which can lead to thrombus formation or arterial branch occlusion. ³⁴ Coil protrusion or herniation, when it occurred, was understandably linked to thromboembolic complications in our series; this emphasizes the importance of meticulous technique.

Importantly, BAC was not associated with increased risk of complications in our study. Prior analyses have reported similar safety for BAC compared to standard coiling. For instance, a large series by Waldek et al. noted no significant difference in neurological disturbance between BAC and conventional coiling, despite BAC being used in more complex aneurysms. ³⁵ This may reflect recent advances in balloon catheter technology and the growing operator experience since the introduction of the remodeling technique.

Looking forward, the landscape of endovascular aneurysm therapy is evolving. New adjunctive devices such as intrasaccular flow disruptors and flow-diverting stents are increasingly used for wide-necked aneurysms. These innovations show promise in further reducing recanalization, but each comes with its own considerations of cost, learning curve, and considerations for antiplatelet therapy.^36,37 Our findings highlight that the balloon catheter remains relevant and effective in the current era. Balloon remodeling can be readily combined with newer techniques as well, for example, balloon-assisted WEB deployment or coil plus balloon before a flow diverter, to optimize outcomes.^38,39 These results should be highlighted as preferences for balloon use continue to wane in the interventional realm.^40–43 Even as stent assisted coiling and flow diverters are more often being used in the ruptured setting, we must not retreat from safe and effective interventions like balloon remodeling.

Limitations

Several limitations may affect this study's generalizability. First, as a single-center retrospective study, our findings may not translate directly to other institutions. For example, there is a notable learning curve to mastering the BAC technique which may not translate to institutions with less experience with it. Similarly, our institution's preference for BAC means there was no sizable “simple-coiling” control group. Aneurysms that were treated without balloons may have had specific anatomy which introduces the risk of selection bias in our analysis. Second, while our long study window allowed for adequate follow-up length to assess recanalization for most patients, shorter follow-up duration in more recent patients may have biased them towards recanalization-negative status. Third, we could not directly measure coil packing density or other aneurysm morphologic variables in all cases, leading to a loss of ∼12% of events during complete case analysis. Thus, incomplete documentation might limit the representativeness of this predictor, although consistent findings across the evaluable cases support our conclusions. Moreover, sensitivity analysis retaining coil packing density similarly identified balloon inflation during coil deployment as the sole predictor of aneurysm recanalization in multivariable analysis. Regardless, our sample size and moderate degree of loss to follow-up may have left our study underpowered to detect certain risk factors for complications or follow-up occlusion status. Despite these limitations, our study provides important evidence on modifiable factors affecting the recanalization of coiling-treated intracranial aneurysms. This study's finding that BAC was a significant predictor of non-recanalization status supports the continued use of this technique. Future prospective studies with larger cohorts and a greater range of adjusting regression factors might help establish causality of this phenomenon.

Conclusion

Endovascular coiling of intracranial aneurysms achieves high rates of aneurysm occlusion with an acceptable safety profile. However, aneurysm morphology plays a pivotal role in outcomes. Large, wide-necked aneurysms are more prone to incomplete occlusion and recanalization. Our single-center analysis demonstrates that the balloon remodeling technique may be an effective countermeasure to these challenges. Balloon-assisted coiling, with proper inflation during coil deployment, significantly reduced recanalization in our cohort and was the only independent predictor of durable occlusion. In practice, this means that liberal use of balloon assistance for wide-neck aneurysms may improve long-term treatment success and reduce the need for retreatment, while avoiding the inherent risks of stents or flow diverters in appropriate cases. Routine consideration of balloon assistance in anatomically suitable cases should thus be emphasized, particularly when aneurysm morphology is unfavorable.

Abbreviations

ATENA

Analysis of Treatment by Endovascular Approach of Non-ruptured Aneurysms

DSA

Digital Subtraction Angiography

MRA

Magnetic Resonance Angiography

OR

Odds Ratio

STROBE

Strengthening the Reporting of Observational Studies in Epidemiology

VIF

Variance Inflation Factor