Introduction

Intracranial aneurysm (IA) is a rare condition in the pediatric population accounting for less than 5% of all IA [3, 22]. The onset of ruptured IA is still responsible for a high mortality and morbidity rate up to 10 [26] and 17%, respectively [27]. IA in children are different in many ways compared with adults with a predominance of IA located on internal carotid artery [36] [48] and more frequent in male population [18]. Moreover, in children, IA tend to be giant, non-saccular and at higher risk of bleeding [27, 41]. Consequently, the treatment strategies are specific and cannot be based only on data from adult series. Due to the scarcity of pediatric IA and the growing place of endovascular techniques in the recent years, we are lacking evidence structuring clear recommendations in the management of pediatric IA, especially to guide our choice between endovascular and surgical treatment. The aim of our study was to report the characteristics, management, and outcomes of children treated for IA in two institutions over the last 17 years and to perform a literature review on this rare pathology.

Materials and methods

Reporting on this cohort study is in accordance with the STROBE statements. It was reviewed and approved by the Institutional Ethics Committee of Rouen University Hospital (Rouen, France).

Setting and study design

This study is based on data provided by two registries from two tertiary neurosurgical units located in Normandy (University Hospital of Rouen and University Hospital of Caen, France). Normandy covers an area of around 3.5 M inhabitants. These registries retrospectively included all children younger than 18 years who were treated in our institutions for intracranial aneurysm between 2001 and 2018. Medical records and available imaging and operative reports were retrospectively reviewed for each child. Collected data concerned general characteristics (age, sex, medical personal and family history), clinical presentation (subarachnoid hemorrhage, headaches, asymptomatic), morphology, etiology of aneurysm (fusiform, saccular, giant, mycotic), and location. Fusiform IA were defined as nonsaccular dilatations involving the entire vessel for a short distance [34]. Giant IA were defined as having a diameter ≥ 25 mm [49]. Modalities of treatment either surgical (simple clipping or bypass) or endovascular (coiling, stenting, vessel occlusion) and their complications (need for shunting, vasospasm, recanalization) were collected. Vasospasm was defined by a sudden onset of neurological deficit, transient or not, associated with angiographic confirmation of vessel narrowing in the corresponding territory [54].

All treated IA were considered in this study. Aneurysms associated with arteriovenous malformations were excluded from the analysis. Aneurysms were diagnosed by contrast-enhanced imaging and/or angiography. Surgical or endovascular treatment was available, and the choice of treatment method was made by a multidisciplinary team including neurosurgeons and interventional neuroradiologists. Clinical signs of rupture, blood localisation and morphological characteristics were considered in decision-making.

After treatment, children were monitored in pediatric intensive care unit. Clinical evaluation and regular imaging were performed to detect complications like rebleeding, vasospasm, or hydrocephalus. Once clinical condition was stable, children were hospitalized in a conventional unit. Later, the follow-up consisted of regular consultation with a neurosurgeon or neuroradiologist to determine other complications, neurological deficits, discovery of new aneurysm, and final autonomy (Glasgow Outcome Scale).

Patients’ characteristics were compared using Fisher’s exact test for discrete variables and Wilcoxon-Mann Whitney test were used for continuous variables. Statistical significance was set at the 0.05 level.

Literature review

The PubMed/MEDLINE databases were used to perform an online search of reported studies using the following keywords and MeSH terms: “neonate,” “intracranial,” “aneurysm,” “cerebral,” and “pediatric.” Only full-text articles in English were chosen. Articles were selected from January 2000 to January 2020 to offer a recent up-to-date review of the literature. For each article, we collected the number of IA, the mean age of the cohort, the sex ratio, the rate of subarachnoid hemorrhage, the rate of good grade at presentation with a Hunt and Hess score ≤ 3, morphological data of IA, the rate of infectious, giant, traumatic and multiple IA, the location, the proportion of surgical and endovascular treatment, the rate of good outcome grade defined by a Glasgow Scale Outcome ≥ 4, and the mortality.

According to the best match and relevance, the search results were refined. Only cases of spontaneous IA in children < 18 years were included. Aneurysm associated with arteriovenous malformation was excluded from the review.

Results

Patient characteristics

During the study period, 22 pediatric patients with IA (mean age 12.6 years ± 5.3 years—range: from 0.1 to 18 years) were treated in our hospitals. We have excluded two cases of IA associated with arteriovenous malformation and two other cases for incomplete data. In total, we have analyzed eighteen children. The main results of our cohort are presented in Table 1. We observed a sex ratio male/female of 3.5 and only three patients (n°11, n°12, and n°15) were younger than 10 years. Three patients (n°2, n°10, and n°17) had a medical condition like Noonan syndrome, high blood pressure, and aortic valvulopathy. We have also noted a family history of IA in two children (n°5 and n°6). Intracranial aneurysms were multiple in three patients (cases n°2, 3 and 10). Thirteen children (72.2%) were admitted with subarachnoid hemorrhage (SAH). In these patients, clinical status was poor (Hunt and Hess (HH) grading ≥ IV) in 6 children. The remaining patients (7/13) with ruptured IA were in good clinical condition at admission with HH grading from I to III. For child n°5, IA was discovered during screening as regards family history.

Table 1 Clinical and radiological characteristics, management and outcomes of patients presenting with pediatric intracranial aneurysm in our cohort
Full size table

Aneurysm characteristics

On initial computerized tomography (CT) imaging, 11/13 patients presented with a Fisher IV SAH (6 tetraventricular hemorrhage and 5 intracerebral hematoma). The remaining 2 others had a Fisher 3 SAH. In patients with unruptured IA, brain imaging was motivated by headaches in 4 and by family history of cerebral aneurysm for the rest (case n°5). The mean size of IA was 13.6 mm ± 15.1 mm (range: from 1.5 to 50 mm). Giant aneurysms were present in 4 male patients. With regard to aneurysm morphology, 12 IA were saccular (67%) and the other 4 were fusiform. The most common location was the internal carotid artery (38%,7/18). The characteristics and location of IA are listed in A 1.

Treatment modalities

From all IA, 9 were treated by endovascular means and 9 with surgery. Concerning ruptured IA (n = 13), seven were treated by endovascular approach and six by surgical exclusion. The mean age was similar between the two groups of treatment, but when regarding the functional outcome criterion evaluated by the hunt and hess score, a non-statistically significant trend in favor of surgical treatment is recorded (2.1 versus 3.3 in embolization group; p = 0,47). With regard to unruptured IA treatment (n = 5), surgery was used in 3 cases (two clipping and one surgical bypass) and embolization for the two others (cases n°7 and 16). An endovascular approach was preferably used for all IA located in posterior circulation. Five out of 7 ICA aneurysms (71%) were treated by surgery and 3 out of 5 (60%) of MCA ones. The treatment of giant IA (n = 4) consisted in three endovascular treatments by coils (cases n°7, 8, and 16) and a surgical intracranial bypass for the biggest (case n°6). Compared with children with endovascular treatment, children in surgical group seemed older (13.8 versus 11.5 years old; p = 0,27) with smaller aneurysms (10.8 mm versus 16.3 mm; p = 0,65).

Three patients were treated for multiple IA (cases n°2, 3 and 10). In the case of n°2, the patient presented with bilateral IA on the posterior cerebellar arteries and with an aneurysm on the superior cerebellar artery. All these aneurysms were treated by simple coiling. In the case of n°3, the child had a mirror aneurysm on the right internal carotid artery, treated by a surgical approach. Finally, in case n°10 initially treated for a left ICA aneurysm, a middle cerebral artery was discovered and treated using surgical approach. In our institution, we have decided, if possible, to use the same treatment modalities in a context of multiple IA so as to simplify follow-up.

Outcomes

In the 13 children presenting with SAH, one died quickly after clipping and receiving medical care (case n°11). After IA treatment, six patients (6/13, 46%) developed hydrocephalus with the need for an external ventricular shunt. Two children required permanent shunt (cases n°1 and 6). Two children presented an early aneurysm recanalization after endovascular procedure (cases n°2 and 8) with rebleeding. In contrast to embolized aneurysms, we report complete exclusion without recanalization after surgical clipping at 1 year follow-up. We had no procedural morbidity in the endovascular group and one transient hemiparesis after surgical bypass for the treatment of giant IA (case n°6). Six children received calcium-blockers treatment for cerebral vasospasm (cases n°1 to 3 and 10 to 12) with good outcomes in all except one (case n°11). After a mean follow-up of 1 year, 2 patients were hemiplegic (cases n°8 and10), one presented a mutism (case n°2) and epilepsy (case n°3). Two children (cases n°1 and 7) had cognitive impairment. In total, the morbidity rate among survivors was of 35.3% (6/17). We did not note difference in morbidity with IA morphology and types of treatment. All children with unruptured IA had a GOS 5 except one (case n°6).

Discussion

Intracranial aneurysm occurring during childhood is very uncommon. Herein, we report the experience of two tertiary referral centers of this rare pathology and perform an updated literature review (Table 2). After applying the above-mentioned criteria, the literature review yielded 972 IA through 26 articles. The selected cases of pediatric IA and summarized results of these articles described in the methods section are presented in Table 2.

Table 2 Literature review of all studies reporting pediatric intracranial aneurysm from 2000 to 2020
Full size table

Demographic data

In our series, we observed a prevalence at onset of IA at early adolescence (mean age 12.6 years ± 5.3 years). These results are concordant with literature data in which the majority of the patients age range from 10 to 15 years old [22, 46]. The occurrence of IA in the neonatal period is exceptionally rare [32, 55] and the incidence of IA in children increases with age. In childhood, there is a male predominance in the onset of IA [28, 36] compared with adults’ series [22, 33]. The sex ratio is dependent of age: Lasjaunias et al. [28] have reported a sex ratio male/female of 3/2 in their entire cohort but below 2 years of age, this sex ratio female/male was of 5/1. In our series, 3 out of 18 patients presented with predisposing factors (Aortic valvulopathy, Noonan syndrome and high blood pressure). Unlike IA in adulthood, pediatric IA are more frequently related to infectious or traumatic context [22], collagen mutation syndrome [18, 35] or other vascular condition [1].

IA characteristics

Similarly to the recent literature [5, 27], the mean size of IA in our cohort was > 10 mm, and the incidence of giant IA is higher compared with adult series [25, 37] up to 30% of cases [29, 42]. We have observed in the literature [25, 46] a predominance of IA located in the anterior circulation, preferentially in the internal carotid artery (ICA) and the proportion of IA of the posterior circulation is higher compared with adult population (21% versus less than 5%) [33, 38]. The incidence of fusiform IA is higher compared with adult series (30% versus less than 15%) [5, 7]. The high prevalence of fusiform IA in children can be explained by the vascular conditions more frequently associated with pediatric IA [28].

Clinical presentation

Subarachnoid hemorrhage is the main presentation of pediatric IA accounting for more than 70% of clinical presentation, followed by headaches, seizures, and focal deficit [5, 23, 28]. The proportion of patients presenting with SAH is lower compared with adult series in which around 80–90% of IA are presenting with SAH [33]. Giant IA in pediatric population is responsible for mass effect and diagnosed with headaches [37]. Epileptic seizures in patients with IA are also more frequent in childhood compared with adulthood [15]. Two explanations are suggested: the immaturity of the cortex in children makes it prone to seizures and the larger proportion of IA in children, related to infectious processes [8, 13].

Treatment modalities

When pooling the results of our literature review concerning treatment modalities, it appeared that the management of saccular pediatric IA was divided between endovascular and surgical procedures in a balanced way but with huge discrepancies between centers [27, 28]. There is no recommendation concerning the management of pediatric IA between both surgical or endovascular procedures. Nevertheless, we have observed an increased incidence of endovascular procedures in our review compared with former studies performed before 2000 [47].

We believe that the factors influencing the choice of treatment should be the personal and team expertise, the age of the child, the location of IA, and clinical context. In the literature [20, 57], surgery is preferentially proposed for IA from the anterior circulation with wide necks. On the contrary, endovascular procedures are preferred in the case of narrow neck’ IA from the posterior circulation. In pediatric patients, IA are more frequently giants. Giants IA are prone to present residual circulation or recanalization with endovascular procedures. As a consequence, younger children present a higher risk of long-term recanalization. In the work from Amelot et al. [5], the risk of annual IA recurrence is estimated around 2.6%. By definition, due to their age, children experience a higher risk of recanalization compared with adults.

We also note specificities in children cerebral circulation. The total cerebral blood-flow and vascular velocities are increased compared with adults which could lead to IA genesis. This is particularly true before 7 years of age and decline after [52, 57]. It has been shown [11, 12] that flow-diverter may increase the pressure in IA, particularly in the case of giants IA and could lead to IA bleeding. As children are prone to present with giants IA, and due to small caliber of the vessels, the use of flow-diverter is at higher risk in children. Moreover, in neonates, most IA concern the distal circulation, mainly the MCA artery, and are responsible for a higher rate of ICH compared with adults. This could emphasize the role of surgical treatment in this population [32].

Outcomes

In our series, 6 children (46%) developed a cerebral vasospasm. Evolution was good for 5 of them (83%) after medical treatment. The reported incidence of vasospasm following SAH in children is around 25% [43, 44] but the impact of vasospasm is minor in children compared with adult population [4] with good clinical outcomes even in the absence of medical treatment. This is due to the existence of peripheral recirculation in children. As regards vasospasm treatment, the use of nimodipine following SAH is based on adult’s series. Few series [21, 45] have reported their experience with discordant results and risk of hypotension. As a consequence, nimodipine in child can be discussed as a case by case situation and is not a standard of care. Another management of vasospasm is the use of mechanic balloon angioplasty [43]. In our study, no patient received this treatment.

Outcomes appeared favorable in the literature in up to two thirds of the patients [5, 27, 36]. Outcome was determined by the clinical status at admission, severity of the hemorrhage, age of the patients, and type of IA [22, 36]. The mortality ranges from 1.3 to 28% [22, 23] reflecting the large heterogeneity between cohorts and pathologies. Unfavorable outcomes were mostly in neonates and patients with giant or dissecting IA [28, 32, 37]. On the contrary, when matched with clinical status at admission, outcomes in pediatric patients appeared better compared with adult population [25, 27, 35].

Pediatric specificities

The characteristics of our cohort are similar to data from the literature but differ from IA in adult population. Although women are at higher risk of aneurysm formation in adult population [12], pediatric IA are more often found in boys with an average sex ratio M/F of 1.4:1. One explanation is the role of female hormone in IA genesis which appears to be a risk factor of IA [12]. Then, this could explain why, before puberty, females are not at risk of IA compared with boys. On the contrary, some conditions predisposing to IA in children may be overrepresented in male populations.

The mechanisms leading to aneurism development in adults were not very well known, but we can expose hypothesis to explain differences between IA in children and adults [12]. First, risk factors in adults are mainly atherosclerosis due to high blood pressure, smoking, or alcohol exposure. In children, IA are more frequently due to dissection or predisposing condition such as connective tissue mutations, explaining the higher prevalence of fusiform or giant IA [9]. Another condition predisposing to IA are mycotic IA which are more frequently found in pediatric population. The treatment and presentation of these IA differ from classical IA.

Concerning clinical features, a significant part of pediatric IA are diagnosed in the presence of headaches, most often in a context of giant IA. We also note that, in Krishna et al. [25], epileptic seizure is significantly more frequent in childhood compared with adulthood, maybe as neural networks are not totally mature. In the same way, children experience better outcomes compared with adults with the same initial status, certainly due to brain plasticity in children.

Regarding location [17, 25], IA are mostly on ICA in children whereas the main location is ACoA in adults. We also find more IA on posterior circulation in pediatric population compared with adults, and the same observation goes for fusiform aspect. Although there are more fusiform IA and more posterior location in childhood, treatment is divided between surgical clipping and embolization. Yet, endovascular treatment is, today, predominant in most centers in adult population. For example, Gawlitza et al. [17] have reported a 77.2% embolization rate. Nevertheless, pediatric IA is a separate entity and treatment modalities may differ from adults. Firstly, the development of endovascular techniques is relatively recent and therefore, due to the low incidence of pediatric IA, large series are often surgical. Secondly, due to the small size of the vessels, there is an increased risk of perforation or thrombosis in pediatric patients. Then, the risk of recanalization is higher in pediatric patients. Childhood is a population particularly vulnerable to recanalization due to an early treatment and a higher life expectancy [58]. In our study, 2 of the 9 children (22%) treated by embolization had an IA recanalization. The risk of recanalization is lower for surgery compared with endovascular techniques as reported by Costa et al. [14] (2 vs. 27%; P = 0.0008). IA recanalization requires new treatment in order to avoid a rebleeding. All these reasons could explain the lower use of endovascular treatment in pediatric population compared with adult population.

Conclusion

Intracranial aneurysm in pediatric patients remains a rare condition distinct from those in adults. Population characteristics are age related with a majority of male patients with IA after the age of 3 years old. The clinical presentation is mostly intracranial hemorrhage and headaches. Aneurysms are more frequently large or giant and located in the internal carotid artery. As regards treatment, we may encounter age-related difficulties in this very diverse population. Surgical procedure remains a therapeutic option with good outcomes in two-third of surgically treated pediatric aneurysm patients. Due to its scarcity, a multidisciplinary discussion is the cornerstone of the management of pediatric IA.