Abstract
Objective: To describe the clinical characteristics of dissections of cerebral arteries in children. Methods: Searches of MEDLINE (1966–2000) and bibliographies were systematically performed for English-language publications that described patients <18 years old with anterior circulation arterial dissections (ACAD) or posterior circulation arterial dissections (PCAD). Results: A total of 2,027 studies were reviewed; 118 patients were identified in 79 studies. Seventy-four percent of patients with ACAD (n = 73) and 87% with PCAD (n = 47) were male (p < 0.0001). When patients with preceding trauma were excluded, this male predominance persisted. All patients had evidence of cerebral ischemia at the time of diagnosis. Headache was reported in approximately half of patients. Sixty percent of ACAD were intracranial. ACAD with no preceding trauma were more commonly intracranial than those preceded by significant trauma (86 vs 25%, p = 0.002). The most common location for PCAD was the vertebral artery at the level of the C1–C2 vertebral bodies (53%). Recurrent ischemic events after the diagnosis of dissection were reported in 15% of PCAD and 5% of ACAD cases. None of the PCAD group and 10% of the ACAD group had recurrent dissections. Conclusions: There is a marked male predominance among children with cerebral arterial dissections that is not explained by trauma. Unlike adult ACAD, childhood ACAD are most commonly intracranial. Spontaneous ACAD, in particular, tend to be intracranial, while post-traumatic ACAD are more often extracranial. The vertebral artery segment most susceptible to dissection is similar between children and adults.
Cerebral arterial dissection has long been considered a rare cause of ischemic stroke in children, while “idiopathic” remains the most commonly reported etiology.1-10⇓⇓⇓⇓⇓⇓⇓⇓⇓ Historically, diagnosis of arterial dissections has depended on conventional angiography, a technique often avoided in children given its invasive nature. Advances in neuroimaging, particularly in noninvasive techniques such as MR angiography (MRA), have improved our ability to diagnose dissections of the cerebral arteries. As these techniques become more widely used in the evaluation of pediatric stroke, it is likely that many arterial dissections will be identified in cases that otherwise would have been deemed idiopathic.
The natural history of cerebral arterial dissections in childhood remains poorly understood. The literature regarding this disorder consists of case reports and small series, with the largest study including 18 patients.11 Some authors have suggested that pediatric cerebral arterial dissections have certain characteristics that differ from those in adults: a higher frequency of intracranial dissections11 and a male predominance attributed to a higher incidence of trauma among boys.12-18⇓⇓⇓⇓⇓⇓ The objective of our study was to better characterize this disorder through a systematic review of the literature.
Methods.
Data sources.
We performed a search for reported cases of cerebral arterial dissections in children younger than 18 years. The MEDLINE database (January 1964 through December 2000) was searched using the keywords “dissecting” and “dissection,” combined sequentially with the keywords “vertebral,” “basilar,” “carotid,” “cerebral,” “cervical arteries,” “cervical artery,” “cervical arterial,” and “cervicocephalic.” English language publications only were included. The title or abstract of each study was scanned; irrelevant articles were excluded. The full text of the remaining articles was retrieved. The reference lists of each of these publications, as well as of review publications, were searched to identify additional relevant reports.
Study selection.
Studies were included if they reported cases of cerebral arterial dissection in children younger than 18 years. They were excluded if they did not provide an individual case description, if an unrelated comorbid diagnosis contributed significantly to the patient’s presentation and outcome, or if the diagnosis of an arterial dissection was questionable. The diagnosis of arterial dissection was accepted if the case met one of the following criteria: 1) pathologic diagnosis of arterial dissection, 2) evidence for dissection on conventional angiography or MRA (presence of contrast in a false lumen in the wall of the artery, tapering or irregular narrowing of the artery, narrowing of the artery with distal dilatation, or presence of a pseudoaneurysm),19 3) evidence for dissection on fat-saturation MRI images (“crescent sign” on axial cuts), or 4) occlusion of the artery with a history of preceding trauma and with no other condition predisposing to arterial occlusion. Dissections included solely based on the last criteria were defined “presumptive”; the data were analyzed both with and without the presumptive cases.
Data abstraction.
Details regarding clinical presentation, location of dissection, treatment, and outcome were extracted from each study and tabulated. Dissections of the internal carotid artery (ICA) and its branches—middle cerebral artery (MCA) and anterior cerebral artery (ACA)—were included as anterior circulation arterial dissections (ACAD). Dissections of the vertebral artery (VA), basilar artery (BA), or posterior cerebral artery (PCA) were categorized as posterior circulation arterial dissections (PCAD).
A history of trauma was abstracted when available. “Trivial” trauma was defined using a modified definition of “mild head injury”: a blow to the head with no or brief loss of consciousness (less than 10 minutes), a Glasgow Coma Scale (GCS) score of 15 on admission to the emergency department, and no evidence of a skull fracture on physical examination.20 New focality on neurologic examination did not exclude patients from this group. Whiplash injury and other minor injuries of the neck were also included as trivial trauma. Trauma was defined as “significant” if the patient had fractures of the skull or spine, head injury with prolonged loss of consciousness (longer than 10 minutes), new traumatic abnormalities of the brain or skull on head CT, penetrating trauma to the neck such as a gunshot wound, or other significant injury to the head and neck region. Patients whose dissections were associated with “activity” were separately tabulated; this included patients whose neurologic symptoms began during or soon after physical activity such as playing soccer. Dissections were defined as “spontaneous” if the report explicitly stated that the patient had no history of preceding trauma or physical activity.
Location of the dissection was defined by the origin of the dissection on histopathology, when available, or by the location of the most proximal stenosis or occlusion on cerebral angiogram or MRA. Segments of the internal carotid artery were defined as follows: C1, cervical (bifurcation of common carotid artery to carotid canal); C2, petrous; C3, cavernous; and C4, supraclinoid. Standard definitions of the segments of the vertebral artery were used.21
When possible, outcome was divided into mild, moderate, and severe deficits. “Mild” was defined as subtle deficits on examination that do not interfere with activities of daily living (ADL). “Moderate” was defined as deficits that are more readily apparent or that interfere to some degree with ADL. “Severe” was defined as deficits that significantly impair the patient’s ability to perform ADL, or would be expected to result in a low quality of life.
Statistical analysis.
All comparisons were analyzed using χ2 tests, or Fisher’s exact tests when any value was less than five. Sex was compared with a hypothetical population of 50% males.
Results.
Literature search.
The MEDLINE search resulted in a list of 2,027 references. Of these, 54 studies met all inclusion criteria. Only one study was excluded owing to a comorbid diagnosis that contributed significantly to the patient’s outcome. The reference lists of these publications and review articles led to the identification of an additional 25 relevant publications reporting on children with cerebral arterial dissections.
A total of 79 studies describing 118 patients with 132 cerebral arterial dissections were included (additional material related to this article can be found on the Neurology Web site; go to www.neurology.org and scroll down the Table of Contents to find the title link for this article). Nine patients had dissections in two different vessels at initial presentation, and five had recurrent arterial dissections at a later time. A recurrent dissection in a patient was considered a different “case.” Of the 118 patients, 71 had ACAD, 45 had PCAD, and two had both.
Identification of cases.
In the 118 patients, a total of 153 diagnostic studies were performed: 103 conventional cerebral angiograms, 16 MRA studies, eight fat saturation MRI studies, and 26 postmortem examinations. Doppler vascular studies were not reported in any of the reviewed cases. A total of 19 patients (seven in the ACAD group, 12 in the PCAD group) with a “presumptive” diagnosis of dissection were included.
Presentation.
The mean age at initial presentation was 10.4 years in the ACAD group (range, 6 days to 17 years) and 8.6 years in the PCAD group (range, 9.5 months to 17 years). A male predominance was found with both types of dissections (p < 0.0001 for both ACAD and PCAD; table 1). It persisted when patients with preceding trauma or physical activity were excluded from the analysis (p = 0.13 for ACAD and p = 0.02 for PCAD).
Sex ratio in 118 children with cerebral arterial dissections
Presence or absence of preceding trauma was similar between the ACAD and PCAD groups; the combined data are shown in table 2. The most common time interval from trauma to onset of neurologic symptoms was 1 to 7 days.
History of preceding trauma in children with cerebral arterial dissection
Three patients in the ACAD group and three in the PCAD group had conditions that could predispose to arterial dissection: one had a premorbid diagnosis of a connective tissue disorder; four had angiographic changes consistent with fibromuscular dysplasia (FMD) noted at the time of the diagnosis of arterial dissection; and one had abnormal internal elastic lamina noted in cerebral and extracerebral arteries on postmortem pathology.
Signs or symptoms of focal cerebral ischemia existed in all patients at the time of diagnosis. Hemiparesis was the most common presenting sign or symptom among patients with ACAD (97%) and PCAD (57%). Headache was reported in 54% of patients with ACAD and 53% of patients with PCAD. Neck pain was an uncommon complaint, reported in 1% of ACAD and 11% of PCAD cases. Nine percent of PCAD cases and no ACAD cases reported neck stiffness. Seizures in the acute phase were described in 21% of ACAD and 9% of PCAD cases.
Subarachnoid hemorrhage was noted in two patients at the time of presentation: one with an extracranial ICA dissection and severe head trauma, and one with a PCA dissection with a 1.8-cm pseudoaneurysm. We found no reported cases of subarachnoid hemorrhage with vertebral pseudoaneurysms.
Location/angiographic findings.
Specific information regarding the most proximal location of the dissection was provided in 73 of the 77 ACAD cases and all of the 47 PCAD cases (table 3).
Location of cerebral arterial dissections in children
The relationship between history of trauma and location of dissection was further analyzed in cases where information for both was available (n = 56 for ACAD). Of ACAD following significant trauma, 25% were intracranial in location, compared with 58% of cases following trivial trauma, and 86% of cases with no history of preceding trauma (p = 0.002, Fisher’s exact test comparing differences between groups). In the PCAD group, there was no association between trauma and location of dissection. There was no apparent relationship between sex and location of dissection in either the ACAD or PCAD group.
Of the 118 patients, 10 (8%) had multiple dissections at their initial presentation: one with left MCA and ACA dissections; two with an ICA dissection and a contralateral vertebral dissection; six with bilateral vertebral dissections; and one with a combination of new and old ICA, ACA, MCA, and PCA dissections. The latter patient had an unspecified underlying connective tissue disorder. Another patient with multiple dissections had evidence of FMD on angiography. History regarding presence or absence of prior trauma was given in eight of the 10 cases: significant trauma in one, trivial trauma in one, and no history of trauma in six.
Two patients with ACAD and four with PCAD were noted to have pseudoaneurysms on their initial diagnostic imaging studies.
Treatment.
Postdissection treatment was reported in 73 of 77 cases of ACAD, and in 43 of 47 cases of PCAD. Of the ACAD cases, 12% were treated with antiplatelet agents alone, 22% with anticoagulation alone, and 66% with neither. Of the PCAD cases, 23% were treated with antiplatelet agents alone, 23% with anticoagulation alone, 14% with both, and 40% with neither. The treatment of PCAD in children has changed over time. Before 1990 (n = 14), no patient received anticoagulation therapy, and only three (21%) received antiplatelet therapy. From 1990 and beyond (n = 29), 24% were treated with antiplatelet agents alone, 34% with anticoagulation alone, 21% with both, and 21% with neither. Similar increased usage of antithrombotic agents was noted for patients with ACAD, with anticoagulation usage increasing from 15 to 29% from before 1990 to after 1990, and usage of antiplatelet agents increasing from 3 to 26% before 1990 to after 1990.
Two patients had complications from anticoagulation therapy: one had a fatal intracranial hemorrhage, and one had a massive gastrointestinal bleed necessitating multiple blood transfusions, but with a good outcome. No patients treated with antiplatelet therapy had hemorrhagic complications.
Outcome.
Outcome was reported in 118 cases (table 4). Two percent of the PCAD group and 33% of the ACAD group died as a result of their dissections. History regarding treatment was available in 24 fatal cases: 23 deaths (96%) occurred in patients not receiving anticoagulation, whereas one death (4%) occurred in a patient receiving heparin. In the ACAD group, the mortality among reported cases was higher for patients with intracranial than extracranial dissections (p = 0.001; see table 4). Mortality was lower among cases reported in or after 1990 compared with those reported before 1990 (p = 0.03).
Mortality from cerebral arterial dissections in children
Time to last follow-up among survivors was an average of 5.2 years (range, 1 week to 21 years) in the ACAD group and 10.8 months (range, 6 days to 4.5 years) in the PCAD group. Outcome of survivors was similar for the ACAD and PCAD groups. Combined, 37% had complete recovery, 33% mild deficits, 8% moderate deficits, 6% severe deficits, and 15% had residual deficits, severity not further defined. None of the PCAD group and 2% of the ACAD group had chorea. Eight percent of the ACAD group and none of the PCAD group had epilepsy.
Recurrences.
We analyzed recurrences both in terms of recurrent arterial dissection, as well as recurrent ischemic events. None of the PCAD survivors and five (10%) of the 49 ACAD survivors had a single recurrent cerebral arterial dissection after their initial diagnosis; no patient had multiple recurrent dissections. The initial dissection in these cases was intracranial in four patients and extracranial in one. The recurrent dissection occurred in the vertebral artery in one and in the intracranial anterior circulation in four. The recurrences occurred 3 weeks to 1 year after the initial dissection. Only one patient had a condition that could predispose to arterial dissection: abnormal internal elastic lamina on postmortem pathology.
We found two categories of patients with recurrent ischemic events: those with a history of recurrent ischemic events by the time of their initial diagnosis of dissection, and those with recurrent ischemic events occurring after their initial diagnosis. A total of 16 patients (seven in the ACAD group, nine in the PCAD group) fell into the former category. Of these, 11 had TIA and five had strokes. The time interval from initial ischemic event to diagnosis of dissection ranged from hours to 1 year (median, 2 weeks). Treatment of these patients varied (anticoagulation in four, antiplatelet agents in three); none were reported to have any additional ischemic events after diagnosis.
Eleven patients fell into the latter category (those with recurrent ischemic events occurring after their initial diagnosis of dissection): four (5%) of 73 patients with ACAD, and seven (15%) of 47 patients with PCAD. Five patients had a single recurrent ischemic event in the acute phase (3 to 11 days) after the initial ischemic event; none was receiving either antiplatelet or anticoagulation therapy. Six patients had a total of eight recurrent ischemic events in the chronic phase (3 weeks to 15 months) after initial presentation. Five recurrences occurred while taking no medical therapy, one while taking antiplatelet therapy, and two while taking anticoagulants. Among all cases of dissection for which both a history of treatment and outcome was provided, recurrent ischemic events (acute and chronic) occurred in two (7%) of 28 cases treated with anticoagulation, compared with 10 (21%) of 48 cases not treated with anticoagulation.
A total of 39 patients had follow-up angiography: 37 had conventional angiograms, two had MRA. Ten patients (26%) had complete resolution of the previous abnormality, 16 (41%) showed improvement, nine (23%) were unchanged, and four (10%) showed progression. Of the four that progressed, one had an enlarging pseudoaneurysm, two had increased stenosis, and one revealed new occlusion of an arterial branch distal to the dissection. Of patients with dissection treated with anticoagulation, nine had follow-up angiography: five showed improvement, two were unchanged, and two showed progression.
Discussion.
In this systematic review of the literature, we identified reports of 118 children with cerebral arterial dissection. The annual incidence of spontaneous dissections of the cervical ICA in the population of a community has been estimated at 2.5 per 100,000; no pediatric cases were identified in that study.22 We suspect, however, that dissections in children are more common than previously recognized, and that current, less invasive imaging techniques will lead to an increase in the recognition of this disorder.
The authors of the largest reported series of childhood cerebral dissections recognized a male predominance among children with this disorder.11 This observation was confirmed in our review, with a particularly striking male predominance in the PCAD group. This contrasts with the adult dissection literature, in which most studies report no sex predilection,23-26⇓⇓⇓ or even report a female predominance.27,28⇓ We found that the male predilection in children could not be explained by trauma, suggesting a sex influence on the pathogenesis of spontaneous dissections. Of note, a number of studies of ischemic stroke in children have reported a 60 to 70% male predominance.3,6,7,29,30⇓⇓⇓⇓ Whether the disparity in rates of dissection between boys and girls contributes to sex differences in childhood stroke cannot be answered by this review.
All the children in our study presented with signs or symptoms of cerebral ischemia, most commonly hemiparesis. Pain was not a prominent presenting feature, with headache reported in only half of the patients, and neck pain rarely noted. This contrasts to adults, in whom pain is often noted to be the most common presenting feature, reported in 60 to 94% of patients, whereas signs of cerebral ischemia at presentation are noted in only 49 to 79%.25,27,31,32⇓⇓⇓ In adults, pain often serves as an important “warning symptom,” facilitating early diagnosis of cerebral arterial dissection. Our observation of a 100% rate of cerebral ischemia at diagnosis suggests a failure to diagnose cerebral arterial dissections in children before ischemia occurs. This could reflect a lower incidence of pain heralding a dissection, a lower index of suspicion for this diagnosis in children presenting with head or neck pain, or underreporting of cases of dissection without ischemic stroke.
Intracranial ACAD are uncommon in adults,25,26,33,34⇓⇓⇓ yet were seen in 60% of the ACAD cases in our study. The authors of the largest series of childhood dissections recognized a relatively high proportion of intracranial location in their series of ACAD (included in our review).11 Interestingly, adults with intracranial dissections have a younger age at onset than those with extracranial dissections.35,36⇓ In addition, a slight male predominance has been noted in adults with intracranial ACAD.35,36⇓
In our review, PCAD appeared to affect the vertebral artery most commonly at the level of the C1–C2 vertebral bodies. The predilection of this location to dissection has been well described in adults,24,26,27,37,38⇓⇓⇓⇓ and has been attributed to mechanical factors at this level.39,40⇓ Interestingly, we found that the propensity for dissection at this location was seen both in patients with a history of significant trauma, as well as in those with spontaneous dissections. This suggests that the amount of mechanical force necessary to cause an arterial dissection may be so slight as to be unrecognized by the patient or observers, or that a regional arterial predisposition to dissection exists at this level.
A history of preceding trauma is thought to be more likely to cause an extracranial, as opposed to intracranial, dissection.26 Indeed, we found that spontaneous ACAD were more likely to be intracranial, and ACAD after significant trauma were more likely to be extracranial. A similar relationship, however, was not found in our PCAD group, in which an extracranial location was most common regardless of history of trauma.
The findings in adults that intracranial ACAD are associated with young age, male sex, and spontaneous etiology substantiate our findings in children and suggest the following conclusions. First, intracranial ACAD in children and young adults may share a common pathophysiology. Second, this pathophysiology may be distinct both from that of ACAD in older adults, and from that of dissections of the posterior circulation. Interestingly, very few children reported had an identifiable underlying condition that could predispose to arterial dissection. Thus, in the majority of cases, the etiology of childhood dissection is unknown. Studies in adults have suggested that patients with dissection may have an underlying arteriopathy.41 Whether similar pathology exists in children is unknown.
The utility of antiplatelet agents and anticoagulation in the treatment of children with cerebral arterial dissections remains unclear. In our study, fewer recurrent ischemic events were reported in patients receiving anticoagulation therapy compared with those not receiving such therapy. In addition, the majority of fatalities reported occurred in patients not receiving anticoagulation. However, given the nature of this retrospective review, these findings must be interpreted with caution. Although many authors have recommended the use of anticoagulation therapy in adult patients with dissections,25,42 43⇓⇓ no prospective randomized study has been performed and the use of anticoagulation in this setting remains unproven. Our study offers no conclusive evidence to help with this dilemma. Anticoagulation presumably reduces a patient’s risk of a thromboembolic event after a dissection. In our review, the rate of recurrent ischemic events after dissection was relatively low (5% in ACAD cases, and 15% in PCAD cases), but may have been higher had none of these patients been treated with anticoagulation. Hemorrhagic complications of anticoagulation, although rare, were reported.
We found the mortality among reported cases of childhood ACAD to be much higher than that of PCAD (33 vs 2%). Only two of 24 fatalities due to ACAD had a history of prior significant trauma, suggesting that trauma cannot account for this high percentage of mortality. Intracranial ACAD in particular had a high mortality (51%), and accounted for 21 of 24 deaths due to ACAD. Poor outcome after intracranial ACAD has been previously reported, but remains controversial.11,26⇓ As suggested by others, the high mortality rate may in part reflect a bias toward reporting cases with a histopathologic diagnosis, as well as the difficulty of making an antemortem diagnosis of an intracranial dissection. We observed that cases of ACAD published in or after 1990 had a lower mortality than those reported before 1990. This may reflect improvements in supportive care, or increased use of treatment modalities such as anticoagulation. A more likely explanation, perhaps, is that our improved ability to diagnose dissection has led to the increased recognition of this disorder in patients with less severe lesions.
Our study is clearly limited in that it is a retrospective analysis of reported cases. The use of terms such as “dissection,” “stroke,” “trauma,” and “good” vs “bad” neurologic outcome is inconsistent between studies. Studies vary in the detail they offer regarding clinical history, neurologic examination, treatment, and outcome. Length of follow-up is highly variable. Major biases likely exist toward reporting complex cases and cases with pathologic (postmortem) confirmation of the diagnosis. Antemortem diagnosis of arterial dissection, particularly intracranial, is difficult, and many cases may be unrecognized. Finally, although an underlying condition that could predispose to arterial dissection (such as FMD or a connective tissue abnormality) was identified in some patients, the majority was not evaluated for such conditions.
With these limitations in mind, we offer the following conclusions. First, an ischemic stroke in any child warrants evaluation for an underlying dissection. Second, whereas head or neck pain often heralds an arterial dissection in adults, these symptoms are rarely observed in the pediatric population. Third, childhood arterial dissections are distinct in that they occur more commonly 1) in boys, independent of trauma, and 2) in an intracranial location. Although reports of recurrent ischemic events after arterial dissection were uncommon in patients treated with anticoagulation, because of the nature of this study no conclusions can be made regarding the efficacy of this treatment.
Acknowledgments
Acknowledgment
The authors thank Dane Chetkovich, MD, PhD, Yvonne Wu, MD, MPH, Donna Ferriero, MD, and Michael Aminoff, MD, for their assistance in reviewing this manuscript.
Footnotes
-
Additional material related to this article can be found on the Neurology Web site. Go to www.neurology.org and scroll down the Table of Contents for the October 9 issue to find the title link for this article.
- Received March 15, 2001.
- Accepted June 15, 2001.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Topics Discussed
Alert Me
Recommended articles
Headache and neck pain in spontaneous internal carotid and vertebral artery dissections
Incidence and outcome of cervical artery dissection
Risk of stroke and recurrent dissection after a cervical artery dissection
Long-term follow-up of aneurysms developed during extracranial internal carotid artery dissection