Outcomes of early targeted embolization for ruptured arteriovenous malformation-associated aneurysms: A single-center study

Introduction

Cerebral arteriovenous malformations (AVMs) are abnormal vascular formations characterized by direct arteriovenous shunting caused by the absence of capillary beds and disorganized vascular structures.[1] AVMs are frequently associated with significant clinical complications, most notably hemorrhagic stroke, their predominant mode of presentation.[1,2,3] The markedly increased risk of rebleeding in ruptured AVMs with associated aneurysms highlights the need for effective therapeutic interventions to prevent catastrophic outcomes.[4] Nevertheless, the optimal approach and timing for treating ruptured AVMs with aneurysms remain controversial due to lesion variability and hematoma effects.[5,6,7]

Targeted embolization offers an alternative strategy by providing an immediate intervention to selectively treat high-risk vascular structures, such as aneurysms and fistulae. This approach aims to prevent rebleeding while reducing the morbidity and mortality associated with AVM management.[8,9] The American Heart Association guidelines classify targeted embolization for ruptured AVM-associated aneurysms as a Class IIb recommendation, suggesting that it can reduce rebleeding risk.[10]

We retrospectively evaluated the efficacy and safety of targeted embolization for ruptured AVM-associated aneurysms, hypothesizing high occlusion rates and low complication and rebleeding risks.

Subjects and Methods

Patient data

From April 2007 to March 2024, 11 consecutive patients with ruptured AVM-associated aneurysms identified as the source of hemorrhage underwent 12 targeted embolization procedures. The following parameters were evaluated: Demographics (age and sex), AVM characteristics (Spetzler–Martin [SM] grade), type of hemorrhage, aneurysm classification, aneurysm size, feeding artery, embolization method, postprocedural occlusion status, complications, rebleeding events, subsequent AVM treatment, and clinical outcomes.[11] Ethical approval was obtained from the Institutional Review Board of Saitama Medical University International Medical Center (approval number: 14–196, July 15, 2022). The study was conducted in accordance with the principles of the Declaration of Helsinki.

Management of patients

Ruptured AVMs were diagnosed using a computed tomography (CT) scan and CT angiography. Digital subtraction angiography (DSA) was performed in all patients.

Aneurysms were classified as flow-related aneurysms (FRA) and intranidal aneurysms (INA). FRA was defined as an aneurysm located on the feeder of the AVM, whereas INA referred to an aneurysm within the AVM nidus.[12] Cases in which AVM-associated aneurysms were identified as the source of hemorrhage were included. The feasibility of achieving a cure through multimodal approaches, including surgical resection, stereotactic radiosurgery (SRS), and embolization, was thoroughly evaluated. For cases in which curative treatment was the primary goal, targeted embolization was performed for AVM-associated aneurysms to reduce the risk of hemorrhage during the waiting period for surgical resection or SRS. In cases in which curative treatment was deemed unfeasible, targeted embolization was chosen as a palliative therapy.

Targeted embolization for ruptured arteriovenous malformation-associated aneurysms

For aneurysms located on feeders involved in normal perfusion, coiling was performed to preserve the parent vessel by directly guiding the microcatheter into the aneurysm. Conversely, for feeders not involved in normal perfusion, INA, or cases where microcatheter navigation into the aneurysm was challenging, embolization with N-butyl cyanoacrylate (NBCA) was employed. Embolization was terminated once the aneurysm disappeared without targeting the nidus. The degree of occlusion was assessed immediately after the procedure with DSA.

Complication and clinical outcome evaluation

Hemorrhagic and ischemic complications were evaluated. Hemorrhagic complications were defined as a new high-density area on CT imaging within 24 h posttreatment, while ischemic complications were defined as high-intensity areas on diffusion-weighted magnetic resonance imaging. If subsequent AVM treatment was performed after managing the associated aneurysms, the final follow-up period was considered the time of this additional treatment. Clinical outcomes were assessed using the modified Rankin Scale (mRS) 6 months after the embolization of AVM-associated aneurysms.

Clinical trial registry

This work is a retrospective analytical study. No clinical trials were involved.

Results

Patient data

The study included 11 consecutive patients who underwent targeted embolization for ruptured AVM-associated aneurysms, with one patient receiving two separate procedures, yielding a total of 12 procedures for evaluation. A summary is provided in Table 1.

Table 1.

Patient demographics, arteriovenous malformation characteristics, and treatment outcomes of patients 1–5

Patient number
1	1	2	3	4	5
Case number	1*	2*	3	4	5	6
Age, sex	77, female	74, female	65, female	65, male	61, male	59, female
AVM location	Left frontoparietal	Left frontoparietal	Left cerebellum	Right occipital	Right cerebellum	Right frontal
SM grade	IV	IV	I	III	I	IV
Size	3	3	1	1	1	3
Eloquent	1	1	0	1	0	1
Venous drainage	0	0	0	1	0	0
Bleeding type (IPH, IVH, SAH)	IVH	IVH	SAH	IVH	IPH	IPH
Aneurysm type (FRA, INA)	FRA	FRA	FRA	FRA	FRA	FRA
Aneurysm size (mm)	5.7	1.5	4.2/2 (2 aneurysms on the same feeder)	4.4	4.5	3.1
Feeding artery to aneurysms	Posterior choroidal artery	Anterior pericallosal artery	PICA	Parieto-occipital artery of PCA	Rt. SCA	Rt. posterior internal frontal artery of ACA
Aneurysm treatment	NBCA	NBCA	NBCA	Coil	NBCA	NBCA
Result of immediate postoperative angiography	CO	CO	CO	CO	CO	CO
Complication	None	None	Infarction	None	None	None
Rebleeding	No	Yes	No	No	No	No
Subsequent treatment for AVM	None	None	EVT + SRS	EVT + surgical removal	EVT + SRS	EVT + surgical removal
mRS on admission	2	2	3	5	5	5
mRS at 6 months	2	2	0	5	4	4
Patient demographics, arteriovenous malformation characteristics, and treatment outcomes of patients 6–11
	Patient number
	6	7	8	9	10	11
Case number	7	8	9	10	11	12
Age, sex	59, female	50, female	47, female	45, male	38, male	16, male
AVM location	Right frontal	Left frontal	Right cerebellum	Right cerebellum	Left parietal	Left cerebellum
SM grade	IV	IV	III	I	I	III
Size	3	2	2	1	1	2
Eloquent	0	1	1	0	0	0
Venous drainage	1	1	0	0	0	1
Bleeding type (IPH, IVH, SAH)	SAH	IPH	SAH	SAH	IPH + IVH	IPH + SAH
Aneurysm type (FRA, INA)	INA	FRA	FRA	FRA	INA	FRA
Aneurysm size (mm)	2.8	5.5	2.1	7.1	1.8	3.8
Feeding artery to aneurysms	N/A	Left LSA	Left PICA	Reft SCA	N/A	Left SCA
Aneurysm treatment	NBCA	NBCA	Coil	Coil	NBCA	NBCA
Result of immediate postoperative angiography	CO	CO	CO	CO	CO	CO
Complication	None	None	None	None	None	None
Rebleeding	No	No	No	No	No	No
Subsequent treatment for AVM	None	EVT	EVT	EVT + surgical removal	None	EVT + surgical removal
mRS on admission	2	4	4	3	3	4
mRS at 6 months	1	4	0	1	1	1

*Procedures 1 and 2 were performed on the same patient. This table summarizes patient clinical outcomes, AVM and aneurysm characteristics, and treatment details. Key parameters include demographics, Spetzler–Martin grade, hemorrhage type, treatment modality, complications, and follow-up outcomes. The same parameters (patient demographics, AVM characteristics, aneurysm details, treatment modalities, and outcomes) for patients 6–11. AVM: Arteriovenous malformation, SM: Spetzler-Martin, IPH: Intraparenchymal hematoma, IVH: Intraventricular hematoma, SAH: subarachnoid hemorrhage, FRA: Flow-related aneurysm, INA: Intranidal aneurysm, mRS: Modified Rankin Scale, PICA: Posterior inferior cerebellar artery, PCA: Posterior cerebral artery, SCA: Superior cerebellar artery, ACA: Anterior cerebral artery, LSA: Lenticulostriate artery, N/A: Not applicable, NBCA: n-butyl-cyanoacrylate, CO: Complete obliteration, EVT: Endovascular treatment, SRS: Stereotactic radiosurgery

The median age of the cases was 59 years (range: 47–65 years), and five cases were male. At the time of hemorrhage onset, three cases had a favorable mRS score of 0–2. AVMs were most commonly located in the cerebellum (42%), with 50% involving eloquent regions. Regarding SM grade, four cases (33%) were classified as I–II and eight as III–IV (67%). Hemorrhage types were classified as intraparenchymal hematoma in three cases (25%), intraventricular hemorrhage (IVH) in three cases (25%), subarachnoid hemorrhage in four cases (33%), and mixed types (17%).

Treatment for arteriovenous malformation-associated aneurysms

The median time from onset to targeted embolization was 2.5 days (interquartile range: 0–11.5 days). A summary of treatment details is provided in Table 1. Among the AVM-associated aneurysms, ten were classified as FRAs (cases 1–6, 8–10, and 12), and two were INAs (cases 7 and 11). The median aneurysm size was 3.8 mm (range: 2.1–4.5 mm). Embolization techniques included feeding artery occlusion using NBCA in nine cases (cases 1–3, 5–8, 11, and 12) and aneurysmal coiling in three cases (cases 4, 9, and 10).

Subsequent treatment for arteriovenous malformation

A summary of subsequent treatments is provided in Table 1. Following targeted embolization for AVM-associated aneurysms, additional AVM management was performed in eight cases (67%). Treatment modalities included embolization combined with surgical resection in four cases (33%), embolization alone in two cases (17%), and embolization followed by SRS in two cases (17%).

In four cases (33%), treatment was completed with targeted embolization alone. Among these, three cases (cases 1, 2, and 7) involved SM Grade IV AVMs, for which curative treatment was deemed unfeasible. In contrast, case 11 involved a small SM Grade I AVM, and curative treatment was achieved through embolization targeting the AVM-associated aneurysm.

Clinical outcome

A summary of clinical outcomes is shown in Table 1. In all cases, immediate posttreatment DSA confirmed complete aneurysm occlusion. Over a median follow-up period of 11 months (range: 3–19 months), one patient (case 2) experienced hemorrhage from a de novo FRA, but no rebleeding was observed from embolized AVM-associated aneurysms.

In Case 3, an infarction occurred, but the patient’s symptoms improved during hospitalization, and her mRS score at 6 months was 0. At the 6-month follow-up, eight cases (67%) achieved a favorable mRS score of 0–2. In addition, eight cases (67%) demonstrated clinical improvement, while four (33%) remained stable compared with their condition at admission.

Illustrative cases

Patient 1

A 74-year-old woman presented with a diffuse, unresectable fronto-parietal AVM involving an eloquent area. CT scan at initial presentation revealed IVH [Figure 1a]. Three-dimensional (3D)-DSA demonstrated an FRA located in the left anterior pericallosal artery [Figure 1b]. Embolization with NBCA resulted in complete occlusion of the FRA, as confirmed by postprocedural DSA [Figure 1c and d]. Twenty-six months after the initial treatment, the patient experienced recurrent IVH [Figure 1e]. Subsequent 3D-DSA revealed a de novo FRA arising from the left posterior choroidal artery [Figure 1f]. Targeted embolization was performed again using NBCA, and complete feeder occlusion was achieved [Figure 1g and h].

Figure 1.

Frontoparietal arteriovenous malformation with two flow-related aneurysms (FRAs) (Patient 1). (a and e) Computed tomography showing intraventricular hematoma. (b-d and f-h) Three-dimensional-digital subtraction angiography (DSA) and DSA showing FRAs in anterior pericallosal and posterior choroidal arteries, successfully occluded using N-butyl cyanoacrylate

Discussion

This study highlights the effectiveness of early targeted embolization in preventing rebleeding and reducing morbidity, reinforcing its role in multimodal and palliative treatment strategies for ruptured AVMs. Treating ruptured AVM-associated aneurysms is critical due to the increased risk of hemorrhage posed by these lesions.[3,13,14,15,16] Interventions for ruptured AVMs have reduced the annual re-rupture risk from 11.34% to 1.7%.[5] Moreover, ruptured AVMs represent a heterogeneous condition, as hemorrhage may originate from an FRA, the AVM nidus itself, or an unknown source, emphasizing the need for case-specific interventions.[17]

The optimal timing for intervention in ruptured AVMs remains undetermined.[9,10] Curative embolization in the acute phase may increase the risk of delayed recurrence due to potential underestimation of the nidus size and morphology caused by mass effect from adjacent hematomas or vascular distortions.[9] Delayed intervention has benefits but carries a significant rebleeding risk. The re-rupture rate in the 1^st month after initial hemorrhage is 1.3% and 0.6% per month after that.[5] Early therapeutic intervention for ruptured AVMs, particularly when associated with aneurysms is detected, may reduce the risk of rebleeding and improve overall prognosis.[9]

Targeted embolization demonstrated a high rate of complete occlusion and a low incidence of major complications.[12,18,19] A systematic review of embolization for AVM-associated aneurysms reported a complete occlusion rate of 89% (122/137 cases) at the last angiographic follow-up.[18] In two retrospective case series, immediate postprocedure angiography confirmed complete occlusion in 100% (8/8 cases) and 93.8% (15/16 cases), respectively.[12,19] In addition, the systematic review, which included 216 patients undergoing targeted embolization for AVM-associated aneurysms, reported periprocedural complications in 11.6%, the majority of which were transient.[18] Ischemia, characterized by ischemic symptoms or radiographic infarction, was the most common complication (6.9%), followed by intracranial hemorrhage (2.3%).

In contrast, a registry-based study reported that curative and adjunctive AVM embolization achieved complete occlusion rates of 30% and 12%, with major complications occurring in around 25% of cases in both groups, primarily due to symptomatic hemorrhage.[20] These findings suggest that AVM embolization aimed at a curative outcome is often incomplete and carries a significant risk of symptomatic hemorrhagic complications, even when used as an adjunctive therapy.

In our study, complete occlusion of AVM-associated aneurysms was achieved in all cases, as confirmed by immediate postprocedure angiography. Complications occurred in 8.3% (1 out of 12 cases), involving a case of infarction. These results align with previous reports on targeted embolization for AVM-associated aneurysms, supporting the feasibility and effectiveness of this approach.

Hemorrhagic complications may also arise in the early to subacute phase following AVM embolization owing to premature occlusion of draining veins before complete nidus occlusion, abrupt changes in intranidal flow dynamics after partial or staged embolization, or normal perfusion pressure breakthrough.[9] We employed a strategy, in which embolization with NBCA was terminated once the aneurysm was no longer visualized on angiography, which minimizes abrupt hemodynamic changes, thereby reducing the risk of hemorrhagic complications. Indeed, no hemorrhagic complications were observed in our study.

In AVM embolization, ischemic complications may occur due to catheter-or procedure-related thromboembolism, nontarget arterial occlusion, and venous penetration.[9] In our study, ischemic complications were observed in one case involving parent artery sacrifice, due to the difficulty in navigating the microcatheter into the FRAs. While parent artery occlusion with liquid embolization achieves high rates of complete occlusion, ischemic complications can arise if the feeder contributes to normal perfusion.[18] Therefore, a meticulous preprocedural evaluation of vascular structures and microcatheter accessibility through cerebral angiography is crucial.

The prognosis for targeted embolization of AVM-associated aneurysms is generally favorable.[12,21] A retrospective case series of 14 patients who underwent targeted embolization for ruptured AVM-associated aneurysms reported that coil embolization was performed for FRAs, while feeding artery sacrifice was primarily chosen for dissecting aneurysms and INAs.[12] At 6 months to 1 year of follow-up, 13 out of 14 patients (93%) achieved a Glasgow Outcome Scale (GOS) score of 5, reflecting a good recovery, while one patient (7%) had a GOS score of 4. In our study, during the median follow-up period of 11 months, no patients experienced worsening clinical outcomes. These findings suggest that targeted embolization for ruptured AVM-associated aneurysms is effective and associated with favorable outcomes.

This study has some limitations. First, the small sample size and single-center design limited the statistical power and generalizability of the findings. Second, the patient cohort exhibited heterogeneity across several clinical characteristics, including AVM location, SM grade, and type of hemorrhage, which may also limit the generalizability of the results. Third, the lack of a control or matched comparison group prevented direct evaluation of the relative efficacy of early targeted embolization. Fourth, the follow-up duration was relatively short in some cases, which may have led to an underestimation of delayed complications.

This study demonstrates that early-phase targeted embolization is an effective and safe treatment option for ruptured AVM-associated aneurysms, achieving high rates of aneurysm occlusion with a low incidence of complications. The absence of rebleeding from treated aneurysms during the follow-up period highlights the potential of this approach in preventing further hemorrhagic events. However, the occurrence of a de novo aneurysm underscores the necessity of long-term follow-up to identify and manage new aneurysm formations posttreatment. Future multicenter studies are warranted to validate these findings and to establish standardized treatment strategies.

Author contributions

KO: Concepts, experimental studies, data acquisition, data analysis, manuscript preparation and guarantor; KA: Data acquisition; MY: Data analysis, manuscript editing and review; HK: Design; SK: Design, manuscript editing and review.

Ethical policy and institutional review board statement

Ethical approval was obtained from the Institutional Review Board of Saitama Medical University International Medical Center (approval number: 14–196, July 15, 2022). The study was conducted in accordance with the principles of the Declaration of Helsinki. The requirement for individual patient consent was waived by the Institutional Review Board because this retrospective study used fully anonymized data that contain no identifiable patient information or images.

Data availability statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.