Treatment of pediatric intracranial aneurysms: institutional case series and systematic literature review

Numerous studies have established the unique epidemiological and clinical characteristics of pediatric IAs. Most commonly reported is a predilection for males, with ratios in the literature ranging from 1.2:1 to 3:1 [12, 21, 22, 32]. Similarly, our cohort had a male predominance of 1.75. Several authors have also described a bimodal age distribution, in which IAs commonly become symptomatic predominantly by mass effect or SAH within the first two years of life, and again in late adolescence [3, 34]. We did not observe this pattern, although this may be due in part to the inherent rarity of these lesions as well as our limited sample size.

While no single parent artery was clearly predominant, the MCA and ICA were the most represented within our cohort, with 12 and 9 (28.5% and 21.4%) IAs arising from these arteries, respectively. Mehrotra et al. reported MCA/ICA incidence rates of 24.7% & 15.1% respectively; figures which are similar to those by Yasin et al., Garg et al., and others [3, 8, 12, 41].

Children lack many of the risk factors traditionally associated with aneurysm formation in adults, such as hypertension, smoking, and age, and thus many are thought be idiopathic [16]. There may however be a genetic component, as these lesions have been linked with several inherited diseases, such as polycystic kidney disease, aortic coarctation, and sickle cell anemia [26]. Vasculopathies such as Marfan syndrome and Ehlers-Danlos syndrome are implicated in approximately 10% of all pediatric IAs [7, 19]. Pediatric IAs have also been known to affect the posterior circulation at a higher rate than their adult counterparts. While reports vary, most estimates fall between 10% and 40%, consistent with our observed rate of 26.1%, compared to approximately 6% in adults [3, 6, 30].

Pediatric IAs are also structurally distinct. Saccular aneurysms, which encompassed over 90% of adult cases, account for only 30–70% in children, and represented 38.1% of aneurysms in our cohort [9, 11]. Together with fusiform aneurysms (16.7%), these classically described morphologies only comprised just over half of the observed variants in our cohort. This morphological heterogeneity is noted in the literature, many of which report ratios of atypical aneurysms ranging from 30 to 50% [8, 21, 34, 41]. To account for these unique morphologies, Lasjaunias et al. utilized a novel classification system that combined physical architecture with clinical context to classify aneurysms as saccular, fusiform, infectious (mycotic), or traumatic [15, 21]. IAs in children tend to be quite large, with giant aneurysms comprising 3–37% of IAs reported in published case series and 34.3% in our study [21, 25, 31, 33].

Children with IAs commonly present with headache, ranging from 50 to 90% in the literature including patients with SAH, and 50% in our cohort [4, 12, 16]. Headaches are multifactorial, and may be caused by mass effect of large or giant aneurysms, turbulent blood flow, or SAH / hydrocephalus. We had a larger proportion of patients presenting in poor neurological condition (Hunt & Hess grade IV or V) (52.1%, 12/23 patients) than observed in other case series (0–43%) [3, 12, 16, 20, 22, 27, 31, 33, 37, 41]. This is possibly due to our institution’s status as a tertiary and quaternary referral center, routinely caring for patients from Mexico and Guam in addition to Southern California.

Recent years have seen the proliferation of various endovascular techniques to treat aneurysms, including primary coiling, balloon or stent-assisted coiling, and flow diversion. Thus far, these minimally invasive methods show comparable outcomes to traditional open microsurgery for both ruptured and unruptured aneurysms, as reported by numerous meta-analyses and case series [1, 2, 13, 32, 41]. Open treatment via clipping seems to minimize the risk of recurrence, while endovascular intervention is less prone to immediate complications, and may be more suitable for more higher-risk patients with more comorbidities [2, 17, 18, 23, 25]. However, many institutions and individual surgeons/interventionalists prefer one approach, making robust comparisons amongst the literature difficult [35, 40]. In our case series, the majority of aneurysms were treated via an open approach, while the remainder were treated with endovascular methods such as coiling or flow diversion.

Flow-diverting stents are relatively recent additions to the armament of the endovascular neurosurgeon, with emerging literature on their use and long-term efficacy in the pediatric population. Navarro et al. first reported the use of flow diverters in three children to good effect, and later work by Vargas et al., among others, established that these devices, at least within the established follow-up periods, were safe and effective methods of treating complex aneurysms in children [4, 9, 34, 38]. An additional important consideration for the long-term stability and efficacy of PEDs and stents in pediatric patients is ensuring appropriate antiplatelet therapy. While these protocols have been researched extensively in adults, there is little information regarding their applicability to children. In their study assessing flow diverters in children, Barburoglu et al. utilized weight-based dosing extrapolated from adult scales for older, larger children, and modified fractional dosing for younger or lower-weight children [4]. The utilization of flow diverters is also limited in the setting of SAH, which is present at diagnosis for many children, as dual antiplatelets are needed which can be associated with higher rates of hemorrhagic complications, as well as the delayed timeframe of aneurysm thrombosis [24]. However, flow diverters are increasingly being used for adult patients in the setting of rupture, particularly for IAs not amenable to other strategies such as blister aneurysms and may warrant exploration for children as well.

At our institution, open surgery is the preferred treatment modality for pediatric IAs requiring intervention with amenable morphology. The higher risk of recurrence after endovascular intervention is an important factor for children who may have many decades to live. As shown in our case series, saccular IAs were exclusively clipped while IAs with unusual morphologies and aneurysmal malformations were considered for open or endovascular intervention on a case-by-case basis. Large or thrombosed IAs with mass effect, or ruptured IAs with associated intraparenchymal hemorrhages, were preferentially clipped or resected, while fusiform or dissecting lesions were better treated with endovascular approaches. Specifically, we consider flow diversion for irregularly shaped (e.g. dumbbell, fusiform) IAs that are not amenable to deconstructive techniques (i.e., with coils) due to vessel anatomy.

We frequently use DSA as a diagnostic adjunct, even for IAs considered amenable to open treatment. However, DSA is not without risk in the pediatric patient as the reported complication rates range from 0.4 to 6.7% [5, 7]. The majority of complications seen in these studies, however, were transient neurological deficits, with major complications occurring in 3/587 patients and 1/241 patients, respectively [5, 7]. We observed one DSA-related complication, a femoral artery thrombus requiring anticoagulation in a 2-year-old after intraoperative DSA during craniotomy.

As stent-assisted coiling and flow diversion require dual antiplatelet therapy (DAPT), they are rarely utilized in the setting of rupture to avoid hemorrhagic complications. Optimal antiplatelet regimens are not well established in children, particularly in those age less than 10 [8, 28]. We typically use aspirin plus a Plavix elixir for at least 6 months in those who require DAPT. Children are assessed with serial platelet reactivity tests to determine the minimum weight-based dose to achieve therapeutic effect.

Patients undergo post-intervention surveillance with MRA every 6–12 months to minimize tests involving radiation and undergo follow-up DSA if there is a concern for recurrence.

We did not identify any factors significantly impacting survival, possibly due to the high survival rate even among ruptured aneurysm patients in our series of 84%. Within the literature, post-treatment mortality rates range from 2 to 35% [14, 15, 16, 20]. Rebleeding from either de novo or recurrent aneurysms is one of the major factors of perioperative mortality in these patients, which underlines the need for consistent follow-up and surveillance imaging [2]. Recurrence occurred in five (13%) patients in our case series; only one recurrence was noted on surveillance imaging, and the other four presented as delayed post-treatment ruptures in which prior follow-up DSAs had demonstrated obliteration. Amelot et al. noted that rebleeding of IAs were associated with a higher mortality rate compared to unruptured aneurysms in their multivariate analysis (OR = 9.2; p = 0.02) [2]. Yasin et al. and others have also demonstrated an association between SAH severity (as measured by Hunt & Hess grade) and poor functional status and mortality by measure of Glasgow Outcomes Scale [36, 41].

In our series, four (10.5%) patients suffered periprocedural complications; results which are within the range reported in the literature (range 0–54%) [10, 17, 25]. Commonly reported complications in prior studies include intraoperative rupture and bleeding, thrombotic events, transient cranial nerve palsies, and neurocognitive deficits [15, 16, 20]. Recurrent aneurysms or reoperation occurs in a small but significant number of patients. In a study of 51 surgically treated pediatric IA patients, 60% experienced rebleeding, and 23% experienced asymptomatic aneurysm recurrence [2]. Another series of 57 aneurysms found favorable outcomes in the vast majority of cases using endovascular techniques, with complications including three new neurological deficits, two aneurysms recurrences that required retreatment, and one aneurysm rerupture that required retreatment [41]. Regardless of the treatment modality, close follow up is needed as re-treatment may be indicated for aneurysmal neck remnants detected on post-treatment angiograms, de-novo aneurysms that form in patients with connective tissue disorders, or after subsequent (potentially occult) bleeding events.

Limitations of our study include its retrospective design, rarity of the pathology, and the long inclusion period (18 years) during which endovascular technologies have expanded and evolved. Treatment algorithms for rare pathologies such as pediatric IAs rely heavily on resources available at the institution, as well as the treating team’s clinical experience, and thus might not be generalizable to other hospitals. We also acknowledge that the relatively small sample size may have limited power to detect statistical differences between groups.