Radiation-associated intracranial aneurysms: A systematic review of clinical presentation, morphology, and treatment outcomes

Abstract

Background

Radiation-associated intracranial aneurysms (RAIs) are a rare but increasingly recognized late complication of cranial and cervical radiotherapy, particularly among long-term survivors of head and neck tumors. This study aims to provide a comprehensive review of the clinical, anatomical, and therapeutic characteristics of RAIs.

Methods

We conducted a systematic review of published RAI cases (1984–2024), collecting data on patient demographics, oncologic history, aneurysm morphology and location, latency from radiotherapy, clinical presentation, and treatment outcomes. The review followed Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines.

Results

A total of 103 patients with 142 intracranial aneurysms were included. The mean latency between radiotherapy and aneurysm diagnosis was 11.3 years (range: 2–21 years). The mean age at radiotherapy was 36.7 years (range: 4 months to 79 years), and the mean age at aneurysm diagnosis was 47.8 years (range: 6–90 years). Aneurysms were most commonly located in the internal carotid artery (32%) and posterior circulation (23%). Morphologically, 45.1% were dissecting or nonsaccular. Half of the aneurysms presented with subarachnoid hemorrhage. Approximately 65% underwent treatment, with about two-thirds managed via endovascular approaches—primarily coiling and stent-assisted coiling. Adequate occlusion was achieved in 66% of aneurysms overall, with an even higher rate of complete/near-complete occlusion—73.1%—among endovascularly treated aneurysms, compared to 48.5% for those treated surgically. The mean radiological follow-up period was 19.5 months.

Conclusion

Radiation-associated intracranial aneurysms are rare vascular lesions with distinct anatomical and clinical features. Early recognition and sustained long-term monitoring are crucial to enable timely intervention. Further research is needed to establish evidence-based strategies for screening and managing this high-risk population.

Introduction

Radiation therapy (RT) remains a cornerstone in the multidisciplinary treatment of head and neck malignancies, as well as various benign and malignant intracranial conditions. Advances in radiation delivery techniques have enabled more precise targeting of tumor volumes while minimizing exposure to surrounding eloquent brain structures.^1,2 Nonetheless, the inclusion of skull base vasculature within the high-dose radiation field is often unavoidable. With the growing population of long-term cancer survivors, increasing attention is being paid to delayed neurological complications of cranial irradiation, including neurocognitive impairment, radiation-induced vasculopathies, and, more rarely, radiation-associated intracranial aneurysms (RAIs).^3–5 Radiation-associated intracranial aneurysms are infrequently encountered but potentially life-threatening vascular lesions. The current understanding of their pathophysiology and natural history is limited, as most of the available literature consists of isolated case reports or small series. Unlike sporadic saccular aneurysms, RAIs frequently arise in atypical locations and often exhibit nonsaccular morphologies, including fusiform or dissecting configurations. Their development is thought to result from progressive radiation-induced vascular injury, involving endothelial dysfunction, medial necrosis, adventitial fibrosis, and chronic inflammation, ultimately weakening the vessel wall and predisposing to aneurysm formation and rupture.

Emerging evidence suggests that subarachnoid hemorrhage (SAH) presentation of RAIs is not uncommon, raising concerns about their inherent fragility and the need for timely detection and treatment. However, due to their rarity, variable latency period, and atypical presentation, RAIs pose diagnostic and therapeutic challenges, and consensus on surveillance and management strategies is lacking.^6–10 To address this knowledge gap, we conducted a systematic review of the literature and assembled the largest cohort of RAIs reported to date. The objective of this study is to provide a comprehensive characterization of the demographic, anatomical, morphological, and therapeutic features of RAIs, with the aim of improving clinical recognition, informing treatment strategies, and guiding long-term follow-up for this uncommon but clinically significant complication of radiotherapy.

Materials and methods

Study design and patient selection

We conducted a comprehensive literature search of PubMed and Google Scholar for studies involving patients who developed intracranial aneurysms (IAs) after receiving RT for head and neck lesions. The search was conducted for studies published between 1984 and 2024. PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-analyses) were followed. ¹¹ Inclusion criteria were as follows: (1) history of cranial or cervical radiotherapy for neoplastic or non-neoplastic lesions; (2) diagnosis of at least one IA not present prior to radiotherapy, confirmed by CT angiography, MR angiography, or digital subtraction angiography; and (3) availability of demographic, clinical, radiological, and treatment-related data.

Exclusion criteria included: (1) studies reporting radiotherapy to anatomical regions other than the head, neck, or the vascular territory of the aneurysm; (2) review articles; (3) publications not written in English (at least abstract in English language); (4) in vitro or animal studies; and (5) series focusing on traumatic or infectious aneurysms.

In cases of overlapping patient populations, only the study with the largest sample size or the most comprehensive dataset was included. Article selection was performed independently by two reviewers, who screened each study in its entirety to assess eligibility. Disagreements were resolved by a third author.

Risk of bias and quality assessment

The methodological quality and risk of bias of the included case reports and case series were assessed using the tool proposed by Murad et al., ¹² specifically developed for evaluating noncomparative studies. This tool evaluates four domains: selection, ascertainment, causality, and reporting. Each study was independently assessed across eight signaling questions, including clarity of patient selection, adequacy of exposure and outcome reporting, consideration of alternative causes, temporal relationship, dose–response effect (if applicable), rechallenge (if applicable), and the level of detail provided.

Each item was scored as “Yes” (1 point), “No,” or “Unclear” (0 points), resulting in a total score ranging from 0 to 8. Studies were categorized as having high (7–8), moderate (5–6), or low (≤4) methodological quality based on the total score. Disagreements between reviewers were resolved by consensus.

Data collection

We extracted the following data: patient demographics (age, sex), details of the underlying lesion and RT (type of lesion, technique used, radiation field), latency from RT to aneurysm diagnosis, aneurysm characteristics (number, size, location, morphology, rupture status), and clinical presentation leading to aneurysm discovery. Aneurysm size was classified as small (<5 mm), medium (5–10 mm), large (10–25 mm), and giant (>25 mm). Morphology was categorized as saccular and dissecting/nonsaccular. Aneurysm location was defined according to standard vascular territories (internal carotid artery [ICA], middle cerebral artery [MCA], anterior cerebral artery [ACA], ACom, PCom, vertebrobasilar system, etc.). For aneurysm receiving treatment, the treatment strategy was classified into surgical, endovascular, or combined. Outcomes after treatment were extracted and defined as adequate aneurysm occlusion (complete occlusion or neck remnant).

Radiotherapy characteristics

Radiotherapy modality was classified as conventional, stereotactic, brachytherapy, or mixed.

Head and neck lesions were either vascular or neoplastic lesions and were classified based on the nature of lesion.

Treatment and outcome assessment

Information on aneurysm treatment modality (microsurgical, endovascular, combined, or conservative) was recorded. Endovascular strategies included coiling, coiling with parent artery occlusion (PAO), stent-assisted coiling, glue embolization, and trapping. Surgical strategies included clipping, trapping, ligation, and other techniques. Aneurysm occlusion was classified as adequate (complete or near-complete) or incomplete, based on posttreatment imaging.

Clinical outcomes were assessed using the modified Rankin Scale (mRS), where available. A good outcome was defined as mRS 0–1, and poor outcome as mRS ≥2. Radiological follow-up was reviewed to assess for rebleeding and aneurysm occlusion status.

Statistical analysis

Descriptive statistics were used to summarize the data. Continuous variables were expressed as mean ± standard deviation or median and range, depending on distribution. Categorical variables were expressed as counts and percentages with the 95% confidence interval for each outcome. Data were analyzed using SPSS v.27.

Ethical considerations

This study was conducted in accordance with the Declaration of Helsinki and approved by the local ethics committee. Informed consent was waived due to the nature of the data collection.

Results

Literature review

Studies included in our systematic review are summarized in Supplemental Table 1. The PRISMA search flow diagram is shown in Supplemental Figure 1. The search strategy was developed by combining keywords and free-text terms related to the topic of interest. Specifically, the terms “radiation induced intracranial aneurysms,” “radiotherapy,” and “intracranial aneurysms” were used and combined using the Boolean operators AND and OR to broaden and refine the search for relevant evidence.

The full search strategy was as follows:

• PubMed: (((radiation induced[Title/Abstract]) AND (intracranial aneurysms[Title/Abstract])) OR ((intracranial aneurysms) AND (radiotherapy))) OR ((radiation) AND (intracranial aneurysms)), yielding 714 results.

• Ovid EMBASE and Ovid MEDLINE: ‘radiotherapy’ AND (‘intracranial aneurysm’ OR ‘cerebrovascular malformation'/exp OR ‘radiation induced malformation'/exp), yielding 1495 results.

After removal of duplicates, title and abstract screening, and exclusion of papers due to missing data or publication in a non-English language, a total of 86 studies reporting on 142 IA were selected and included in the review.

Risk of bias and study quality

Risk of bias and assessment of study quality is reported in Supplemental Table 2.

Out of the 86 studies included:

• 74 studies (86%) were rated as high quality (score 7 or 8), indicating good methodological rigor and a low risk of bias.

• 12 studies (14%) were rated as having moderate quality (score 6), often due to limitations in domains such as rechallenge, dose–response, or completeness of reporting.

Population characteristics

A total of 103 patients (63 males, 38 females, unknown in two cases); female-to-male ratio (0.6:1) with 142 IAs were included in the study (Table 1). Most patients (88/103, 85.4%; 95% CI: 77.5%–91.2%) had a single aneurysm, while 15 (14.6%; 95% CI: 8.8%–22.5%) presented with multiple aneurysms. The mean age at the time of radiotherapy was 36.7 years (range: 4 months to 79 years), and the mean age at IA diagnosis was 47.8 years (range: 6–90 years), resulting in a mean latency period of 11.3 years (range: 2–21 years) between radiotherapy and aneurysm detection.

Table 1.

Population characteristics.

Variables	Row numbers	95% CI	Number of articles
N of patients/NR aneurysm	103/142		86
Pts with single/multiple intracranial aneurysms single	88 (85.4%)	77.5%–91.2%	86
Multiple	15 (14.6%)	8.8%–22.5%
Mean age at radiotherapy (years)	36.7	(Range in months 4–79)	84
Mean age at IA (years)	47.8	(Range in years 6–90)
Mean RAIs latency (years)	11.3	(Range in years 2–21)	84
Female/Male ratio	38/63 (0.6:1)		84
Mean aneurysm size(mm)	8.5		44
Small	29/62 (46.8%)	34.2%–59.6%
Medium	15/62 (24.2)	14.3%–36.7%
Large	11/62 (17.7%)	9.2%–29.6%
Giant	7/62 (11.3%)	4.7%–21.9%
NR	80
Chemotherapy			55
Yes	21/67 (31.3%)	20.8%–43.6%
No	46/67 (68.7%)	56.4%–79.2%
NR	35
Aneurysm location			86
ICA	44/138 (31.9%)	24.2%–40.4%
MCA	16/138 (11.6%)	6.8%–18.1%
ACA	7/138 (5.1%)	2.1%–10.2%
Acom	6/138 (4.3%)	1.6%–9.1%
Pcom	3/138 (2.2%)	0.5%–6.3%
AICA	17/138 (12.3%)	7.3%–19.0%
Basilar	7/138 (5.1%)	2.1%–10.2%
PCA	10/138 (7.2%)	3.5%–13.0%
SCA	5/138 (3.6%)	1.2%–8.2%
PICA	5/138 (3.6%)	1.2%–8.2%
Other	14/138 (10.1%)	5.6%–16.3%
NR	4/138 (2.8%)
Aneurysm morphology			63
Saccular	34/142 (34.3%)	24.9%–44.8
Nonsaccular/dissecting	64/142 (64.6%)	54.3%–74.0
NR	43/142 (30.3%)
Ruptured status/total	71/142 (50%)	41.5%–58.5%	86
Incidentally discovered aneurysms	36/142 (25.4%)	18.5%–33.3%	86
Number of multiples aneurysms	39/142 (27.5%)	20.4%–35.5%	15

The mean aneurysm size was 8.5 mm (95% CI: 6.2–10.8 mm). Of the aneurysms with size data available (n = 80), 29 were small (<5 mm), 15 medium (5–9 mm), 11 large (10–24 mm), and 7 classified as giant (≥25 mm). Chemotherapy was administered in 21 patients, while 46 did not receive chemotherapy.

Anatomically, aneurysms were most frequently located in the ICA (44/138; 31.9%; 95% CI: 24.2%–40.4%), followed by the anterior inferior cerebellar artery (AICA, 17/138; 12.3%; 95% CI: 7.3%–19.0%), the MCA (16/138; 11.6%; 95% CI: 6.8%–18.1%), and the posterior cerebral artery (PCA, 10/138; 7.2%; 95% CI: 3.5%–13.0%). Other documented locations included the ACA (7/138; 5.1%; 95% CI: 2.1%–10.2%), anterior communicating artery (6/138; 4.3%; 95% CI: 1.6%–9.1%), posterior communicating artery (3/138; 2.2%; 95% CI: 0.5%–6.3%), basilar artery (7/138; 5.1%; 95% CI: 2.1%–10.2%), superior cerebellar artery (5/138; 3.6%; 95% CI: 1.2%–8.2%), and posterior inferior cerebellar artery (PICA, 5/138; 3.6%; 95% CI: 1.2%–8.2%).

In terms of morphology, 64 out of 99 aneurysms with reported classification (64.6%; 95% CI: 54.3%–74.0%) were nonsaccular or dissecting, while 35 (35.3%; 95% CI: 24.9%–44.8%) were saccular. Morphological data were not reported for 43 aneurysms (30.3%).

At presentation, 71 out of 142 aneurysms (50.0%; 95% CI: 41.5%–58.5%) were ruptured. Multiple aneurysms were present in 39 of 142 patients (27.5%; 95% CI: 20.4%–35.5%).

Head and neck lesion characteristics

Primary head and neck lesions preceding radiotherapy were heterogeneous in histology (Table 2). The most frequent diagnoses were nasopharyngeal carcinoma (n = 16, 18.6%; 95% CI: 10.4%–26.8%), vestibular schwannoma (n = 14, 16.3%; 95% CI: 8.5%–24.1%), glioma (n = 13, 15.1%; 95% CI: 7.5%–22.7%), and pituitary adenoma (n = 12, 14.0%; 95% CI: 6.6%–21.3%). Other common lesions included medulloblastoma (n = 10, 11.6%; 95% CI: 4.9%–18.4%) and arteriovenous malformation (n = 9, 10.5%; 95% CI: 4.0%–16.9%). Less frequent conditions were craniopharyngioma and trigeminal neuralgia (each 7.0%), followed by various rare lesions such as germinoma, chordoma/chondrosarcoma, Ewing sarcoma, meningioma, other head and neck tumors, neurocytoma, and retinoblastoma, all accounting for less than 5% of cases individually.

Table 2.

Head and neck lesion characteristics and treatment modality.

Variables	Row numbers	95% CI	Number of articles
Type of lesion (female/male/NR)			86
AVM (2/7/0)	9 (10.5%)	4.0%–16.9%
Brain metastases (2/0/0)	2 (2.3%)	0%–5.5%
Craniopharyngioma (2/3/1)	6 (7%)	1.6%–12.4%
Chordoma/chondrosarcoma (0/2/0)	2 (2.3%)	0%–5.5%
Ewing sarcoma (0/2/0)	2 (2.3%)	0%–5.5%
Germinoma (1/1/0)	2 (2.3%)	0%–5.5%
Glioma (6/7/0)	13 (15.1%)	7.5%–22.7%
Head and neck tumor (1/2/0)	3 (3.5%)	0%–7.4%
Nasopharyngeal carcinoma (2/14/0)	16 (18.6%)	10.4%–26.8%
Medulloblastoma (1/9/0)	10 (11.6%)	4.9%–18.4%
Meningioma (2/1/0)	3 (3.5%)	0%–7.4%
Neurocytoma (0/1/0)	1 (1.2%)	0%–3.4%
Pituitary adenoma (8/3/1)	12 (14%)	6.6%–21.3%
Retinoblastoma (0/1/0)	1 (1.2%)	0%–3.4%
Vestibular schwannoma (7/7/0)	14 (16.3%)	8.5%–24.1%
Trigeminal neuralgia (2/4/0)	6 (7%)	1.6%–12.4%
Undeterminated	1 (1.2%)	0%–3.4%
RT technique			80
Conventional	43/80 (53.8%)	42.8%–64.7%
Stereotactic	39/80 (48.8%)	37.8%–59.7%
BRT	3/80 (3.8%)	0%–7.9%
Mixed	10/80 (12.5%)	5.3%–19.7%

Sex distribution varied across tumor types, with a marked male predominance observed in nasopharyngeal carcinoma (2 females, 14 males) and medulloblastoma (1 female, 9 males). For some tumor types, sex data were incomplete or not reported.

Radiotherapy modality

Among patients with available data (n = 80), conventional radiotherapy was the most frequently adopted modality (n = 43, 53.8%; 95% CI: 42.4%–65.0%), followed closely by stereotactic radiotherapy (n = 39, 48.8%; 95% CI: 37.5%–60.2%). Brachytherapy was employed in 3 cases (3.8%; 95% CI: 0.8%–10.6%), while a mixed radiotherapy approach was reported in 10 patients (12.5%; 95% CI: 6.2%–21.7%). The mean RT dose was 56 Gy and 32 Gy, in conventional and stereotactic treatment, respectively.

Aneurysm treatment and outcomes

Among the 142 IAs reviewed, 92 (64.8%) underwent active treatment (Table 3). Endovascular therapy was the most frequently adopted modality (n = 52, 56.5%; 95% CI: 46.0%–66.6%), followed by surgical treatment (n = 33, 36.0%; 95% CI: 26.5%–46.4%) and a combined surgical-endovascular approach (n = 7, 7.6%; 95% CI: 3.1%–15.0%).

Table 3.

Aneurysm treatment modality and treatment-related outcomes.

Variables	Row numbers	95% CI	Number of articles
N of aneurysms receiving treatment/ Total of N of aneurysms	92/142 (64.8%)		74
Type of treatment			74
Surgical	33 (36%)	26.5%–46.4%
Endovascular	52 (56.5)	46.0%–66.6%
Surgical + endovascular	7 (7.6%)	3.1%–15.0%
Endovascular strategy
Coiling	28 (53.8%)	40.1%–67.1%
PAO	9 (17.3%)	8.3%–30.4%
Stenting-associated techniques	6 (11.5%)	4.3%–23.4%
Endovascular NOS	9 (17.3%)	8.3%–30.4%
Surgical strategy
Clipping	17 (51.5%)	34.9%–67.8%
Surgical PAO	11 (33.3%)	18.6%–51.0%
Wrapping	1 (3.0%)	0.1%–15.3%
Surgery NOS	4 (12.1%)	3.4%–28.2%
N of patients achieving adequate occlusion (complete/near complete)	61/92 (66%)	56.5%–76.1%
Mean radiological follow-up (months)	19.5		32
Clinical outcome			81
Full recovery	43/103 (41.7%)	32.1%–51.6
Minor signs/symptoms (Rankin 1) Deteriorated or major	32/103 (31.1%)	22.3%–40.9%
signs/symptoms (Rankin 2–3)	5/103 (4.9%)	1.6%–11.1%
Died for SAH	14/103 (13.6%)	7.7%–21.8%
Died for cancer	3/103 (2.9%)	0.6%–8.3%
Not reported	6/103 (5.8%)	2.2%–12.1%

Within the endovascular group, the most common strategy was coiling (n = 28, 53.8%; 95% CI: 40.1%–67.1%), followed by PAO in 9 cases (17.3%; 95% CI: 8.3%–30.4%), stent-assisted techniques in 6 (11.5%; 95% CI: 4.3%–23.4%), and other or nonspecified endovascular techniques in 9 patients (17.3%; 95% CI: 8.3%–30.4%).

Surgical management consisted of clipping in 17 cases (51.5% of surgeries; 95% CI: 34.9%–67.8%), surgical PAO in 11 (33.3%; 95% CI: 18.6%–51.0%), wrapping in 1 (3.0%; 95% CI: 0.1%–15.3%), and other surgical strategies in 4 patients (12.1%; 95% CI: 3.4%–28.2%).

Adequate aneurysm occlusion—defined as complete or near-complete—was achieved in 61 of the 92 treated aneurysms (66.0%; 95% CI: 56.5%–76.1%). Specifically, this included 38 of the 52 aneurysms treated endovascularly (73.1%; 95% CI: 58.1%–84.7%), 16 of the 33 treated surgically (48.5%; 95% CI: 31.7%–65.4%), and all 7 aneurysms treated with a combined surgical-endovascular approach (100%; 95% CI: 59.0%–100.0%). The mean radiological follow-up duration was 19.5 months, reported in 32 cases.

Clinical outcome data were available for 81 patients. Full recovery was observed in 43 (41.7%; 95% CI: 32.1%–51.6%), while 32 (31.1%; 95% CI: 22.3%–40.9%) had minor symptoms or remained stable (modified Rankin score [mRS] = 1). Deterioration or major neurological deficits (mRS 2–3) occurred in 5 patients (4.9%; 95% CI: 1.6%–11.1%). Fourteen patients (13.6%; 95% CI: 7.7%–21.8%) died due to SAH, and 3 (2.9%; 95% CI: 0.6%–8.3%) died due to cancer progression.

Case presentation

Figures 1 and 2 illustrate the case of a 49-year-old woman who, in 2009, underwent a single session of proton therapy for a left-sided cavernous sinus osteosarcoma.

Figure 1.

(a) Axial T1-weighted gadolinium-enhanced brain MRI from 2009 showing the osteosarcoma involving the left cavernous sinus (yellow arrows). This region was treated with a single session of proton therapy. The basilar artery appears normal in diameter (red arrow). (b) Isodose curve map showing the irradiated brain region; the basilar artery is located centrally within the high-dose radiation area. (c) Axial T2-weighted gadolinium-enhanced MRI performed in 2019, following a posterior fossa stroke without large vessel occlusion; the basilar artery remains normal in diameter (yellow arrow). (d) In 2022, 13 years after radiotherapy, follow-up MRI revealed a 16-mm side-wall aneurysm of the basilar artery. Because the isodose distribution encompassed the aneurysm site, this lesion was suspected to be a radiation-induced intracranial aneurysm.

Figure 2.

(a) Lateral and (b) anteroposterior angiographic views showing a large side-wall aneurysm of the basilar artery. (c–d) Endovascular treatment with LEO stent placement and coiling was performed, resulting in immediate complete occlusion of the lesion. (e–f) Follow-up brain MRI with time-of-flight (TOF) angiography at 12 months confirmed persistent complete occlusion of the aneurysm.

In 2019, she experienced a posterior fossa stroke without large vessel occlusion. At that time, no aneurysm was detected. In 2023—13 years after radiation exposure—MRI follow-up revealed a large side-wall aneurysm of the basilar artery, suspected to be a radiation-induced aneurysm. The lesion was successfully treated with stent-assisted coiling, resulting in complete occlusion.

Discussion

Radiation-induced IAs represent a rare but increasingly recognized vascular complication following cranial or cervical irradiation, particularly in long-term survivors of head and neck tumors. This study represents the largest to date analyzing the clinical, radiological, and treatment-related characteristics of RIAs. Our findings provide important insights into the latency, distribution, morphology, and treatment-related outcomes of these lesions.

Latency and demographics

Our study underlines several relevant demographic data. First, the mean latency between radiotherapy and aneurysm diagnosis in our series was 11.3 years, consistent with prior literature suggesting a delayed but potentially progressive vascular insult following RT. This latency is even longer than that reported by Huang et al., ¹³ who found a mean latency of 6.8 years, reinforcing the notion that vascular damage postradiation may evolve insidiously over time and highlights the necessity for long-term surveillance.

Second, the mean age at diagnosis of RAIs in our series was 48 years, and the female-to-male ratio was 0.60:1, indicating a clear predominance in male patients. These findings contrast with the demographics reported in the ISUIA cohort, ¹⁴ which analyzed the natural history of unruptured IA. In that group, the mean age was 56.0 years—slightly older than in our series—and 71.1% of patients were female, reinforcing the well-established female predominance in sporadic aneurysm populations.

This divergence suggests that RAIs may represent a distinct clinical and biological entity, characterized by unique risk profiles and pathophysiological mechanisms. The male predominance observed in our cohort aligns with findings from the multicenter study by Huang et al., which reported a male-to-female ratio of 2.4:1 in their RAI population. A plausible explanation lies in the oncologic background of these patients. Among the four most common primary lesions—nasopharyngeal carcinoma, vestibular schwannoma, glioma, and medulloblastoma—a consistent male predominance was observed. Collectively, these entities accounted for more than 50% of all cases in our cohort, with male patients representing approximately 70% of this subgroup. This pattern reflects the known male predominance in the incidence of these tumors, potentially contributing to the sex distribution observed in our RAI cohort.^15–17

Aneurysm morphology and location

Radiation-induced IAs in our cohort showed a notable predominance in the ICA (32%) and in the posterior circulation (23%), particularly within the AICA, PICA, and PCA territories. The proportion of posterior circulation aneurysms was higher than that reported in the ISUIA cohort (15%), ¹⁴ suggesting a distinct distribution pattern in radiation-associated cases. This increased frequency may be explained by the anatomical proximity of the posterior fossa vessels and the ICA to skull base structures, which are often included in the high-dose radiation field during treatment of posterior fossa and skull base tumors. Chronic radiation-induced endothelial injury, inflammation, and progressive vessel wall remodeling may contribute to the delayed formation of IAs. Emerging evidence suggests that aneurysms in irradiated patients result from a cascade of vascular damage, including endothelial apoptosis, extracellular matrix degradation, and blood–brain barrier disruption. As reviewed by Guo et al., ¹⁵ ionizing radiation activates matrix metalloproteinases, further weakening the vessel wall. These mechanisms reinforce the hypothesis that vascular territories exposed to high-dose irradiation—particularly near tumor margins—are more susceptible to long-term structural damage. This highlights the importance of long-term vascular surveillance in these patients.

Morphologically, nearly half of the aneurysms were classified as nonsaccular or dissecting (45.1%), which is considerably higher than in the general population. Dissecting and fusiform IAs are uncommon and often underrecognized entities. Clinical series suggest they account for less than 2% of all IA, ¹⁸ as reported in studies such as the aneurIST trial and other hospital-based cohorts. However, these figures likely underestimate the true prevalence, as many lesions are asymptomatic and go undetected. Interestingly, this rarity aligns with findings from the largest series on RIAs, in which sidewall aneurysms were significantly more frequent in the radiation-associated group compared to controls (46.5% vs. 32.3%, p = 0.048). ¹³ This observation supports the hypothesis that radiation exposure may promote aneurysm formation along the vessel wall, likely due to progressive wall weakening and altered hemodynamic stress in irradiated vascular territories.

Clinical presentation and detection

In our study, 50% of aneurysms were diagnosed following rupture, a proportion noticeably higher than that typically reported in screening of incidental aneurysms.^14,19 Additionally, 35% of patients were diagnosed incidentally, underscoring the heterogeneity of clinical presentation in RAIs. This trend is consistent with findings from Huang et al., ¹³ who reported that 88.2% of RAIs presented with SAH, significantly more than in the nonradiation control group. Although our data do not allow us to conclusively demonstrate a higher rupture risk in RAIs, the elevated proportion of ruptured presentations, both in our cohort and in the literature,^7,11,20–22 supports the hypothesis that RAIs may be more fragile and prone to rupture.

This presumed fragility may be attributed to several radiation-induced pathological changes, including vessel wall necrosis, chronic inflammation, and endothelial dysfunction, which compromise structural integrity over time. ⁶ Given these considerations, early detection and structured long-term follow-up should be strongly considered for patients at risk of RAIs. While the optimal timing and modality of screening have yet to be defined, our findings—showing a mean latency of 11.3 years between radiotherapy and aneurysm diagnosis—suggest that prolonged surveillance with noninvasive brain MRI may be warranted, particularly in younger patients with long life expectancy.

Treatment strategies and outcomes

Management of aneurysms in the setting of prior radiotherapy remains particularly challenging due to the potential fragility of the irradiated vasculature and the frequent presence of comorbidities. In our cohort, approximately 65% of aneurysms underwent active treatment. Among these, endovascular therapy was the most commonly adopted approach, used in nearly 60% of cases, followed by microsurgical treatment or a combined surgical-endovascular strategy in the remainder. Given that this retrospective review spans over three decades, it is important to acknowledge that treatment strategies have evolved substantially during this period. Endovascular approaches were primarily based on simple coiling and stent-assisted coiling, together accounting for over 60% of endovascular cases. Parent artery occlusion was employed in approximately 20% of lesions. Advanced techniques such as flow diversion or intrasaccular devices were rarely represented in this series, likely due to their relatively recent introduction, but they may represent valuable options in selected patients moving forward.

Microsurgical treatment was performed in roughly 35% of cases, with clipping being the most frequently utilized technique. However, due to the high prevalence of nonsaccular (dissecting or fusiform) aneurysms, PAO was required in approximately one-third of surgically treated patients. Overall, adequate aneurysm occlusion was achieved in 66% of treated lesions. This rate is notably lower than the 80–95% occlusion rates typically reported in contemporary series of unirradiated endovascularly or surgically treated aneurysm,^23–25 depending on aneurysm type, size, and location. Nevertheless, the endovascular occlusion rate in our cohort appeared relatively good, with a complete or near-complete occlusion achieved in 73.1% of endovascularly treated aneurysms. This discrepancy may reflect the inherent difficulties of treating radiation-induced aneurysms, which often present with atypical morphology, involve nonstandard anatomical locations, and are associated with histopathological vessel wall changes such as fibrosis, necrosis, and chronic inflammation. Furthermore, the high proportion of hemorrhagic presentations in RAIs increases the complexity and urgency of treatment.

These findings underscore the importance of long-term follow-up, even in treated cases, to monitor for recurrence or delayed rebleeding. Few studies in our review^26–28 have reported cases of rebleeding in previously ruptured and treated RAIs, as well as rupture in previously unruptured and treated lesions, highlighting the need for vigilant surveillance and readiness to reintervene when necessary.

Limitations

This study is subject to several limitations, primarily related to the heterogeneity of clinical data.^29–105 Imaging and treatment protocols were not standardized across all cases, and the definition of occlusion and clinical outcome relied on available documentation. The true prevalence of RIAs remains difficult to assess due to the absence of systematic screening and possible underdiagnosis in asymptomatic individuals. Finally, long-term outcomes beyond the radiological follow-up window remain largely unknown.

Conclusions

Radiation-associated IAs represent a distinct and challenging vascular entity, characterized by atypical morphology, unusual locations, and a high rate of hemorrhagic presentation. Our findings highlight the need for heightened clinical awareness, individualized treatment planning, and structured long-term follow-up, given the delayed onset and potential fragility of these lesions. Although therapeutic outcomes remain suboptimal compared to nonirradiated aneurysms, advancements in endovascular and surgical techniques may improve future management. Further prospective studies are needed to optimize screening protocols and long-term surveillance strategies in this high-risk population.