|Home | About | Journals | Submit | Contact Us | Français|
Craniocervical arterial dissection (CCAD) in childhood usually presents with symptoms of acute ischemic stroke (AIS) or transient ischemic attack (TIA). It occurs in 2.5 children per 100,000 per year . Separation between the intimal layers of the vessel wall creates an area of damaged endothelium with exposure of collagen, activated tissue factor, and exposed von Willebrand factor. These factors generate secondary fibrin and platelet adhesion, leading to thrombus propagation. Once a clot has formed, ischemia occurs from vessel occlusion at the site of dissection or from clot embolus downstream . Aneurysmal dilatation, which can occur secondary to impaired integrity of the vessel wall and persistent arterial pressure occlusion, frequently appears in the C1-C2 vertebral circulation in children [2,3].
Risk factors for dissection in children include head and neck injury, connective tissue disorders (such as Ehlers-Danlos syndrome), and male gender [4–6, Class IV]. Other well-known childhood AIS risk factors (e.g., thrombophilia) may theoretically contribute to risk of AIS in the presence of CCAD. Patients with arterial abnormalities have a high risk of recurrent AIS [7, Class III].
There are two types of CCAD in childhood: extracranial dissection and intracranial dissection. These entities have differing risk factors and management. Extracranial dissections account for 5% to 25% of childhood-onset AIS [4,8••, 9••] and are often preceded by trauma [4–6, Class IV]. Typically, anterior circulation dissection presents with focal neurologic symptoms such as hemiparesis or aphasia. Posterior circulation events from vertebral or basilar dissection are more challenging to diagnose because their symptoms and signs can range from dizziness to coma. Clues to this diagnosis include history of recent trauma and/or cranial nerve abnormalities. Early evidence suggests that dissection is more prevalent in the posterior than the anterior circulation of children with AIS [8••, Class IV]. Interestingly, although neck pain is a common sign of dissection in adult AIS, diffuse headache is more common in children [4,6].
Because of imprecise classification and challenges in definitive diagnosis, the prevalence of intracranial dissection in childhood is unknown. Recent literature in childhood AIS has focused upon intracranial focal cerebral arteriopathy, which occurs in up to 80% of previously healthy children with AIS [10, Class IV]. Although many of these transient narrowings within the intracranial cerebral vasculature are likely to be associated with an infectious or parainfectious phenomenon rather than dissections [11•, Class IV], recent case reports suggest that some of these lesions may indeed be dissections .
Diagnosis of childhood CCAD and AIS relies upon the appropriate clinical suspicion. Improvement in clinical suspicion and community awareness will help address the concerning observation that median time to diagnosis for childhood-onset AIS is 25 hours from time of symptom onset .
The International Pediatric Stroke Study defines CCAD as “(1) angiographic double lumen, intimal flap, or pseudo aneurysm, or, on axial T1 fat saturation MRI images, a “bright crescent sign” in the arterial wall [as seen in Fig. 1, Case 1]; (2) cervical or cranial trauma, or neck pain, less than 6 weeks preceding angiographic findings of segmental arterial narrowing (or occlusion) located in the cervical arteries; (3) angiographic segmental narrowing (or occlusion) of the vertebral artery at the level of the C2 vertebral body, even without known traumatic history” [14, Class IV].
Spontaneous CCAD occurs with no preceding history of significant trauma, or after seemingly innocuous trauma (examples from our experience include neck torsion during rugby, chiropractor manipulation, climbing over a wall, landing on the head when playing on a trampoline, and falling off of a skateboard). At the same time, it is appreciated that many children without dissection also report a recent prior history of minor trauma, and some “background” level of minor trauma is expected in the healthy pediatric population. Well-designed case-control studies are lacking on this issue. In one series of CCAD in children, minor trauma occurred prior to presentation in 25% . Most children with spontaneous CCAD present with nonspecific, often transient, neurologic symptoms including headache, vomiting, dizziness, vertigo, diplopia, confusion, and neck pain. Other presenting symptoms and signs include altered level of consciousness (25%), Horner’s syndrome, and/or seizures (12.5%) .
Traumatic CCAD, in contrast, occurs after generalized major trauma, such as a motor vehicle accident. It can present with similar signs and symptoms but is increasingly recognized via screening imaging studies in asymptomatic patients with major trauma.
The cornerstone of diagnosing CCAD is imaging of the cervical and intracranial vasculature. The imaging modalities available for diagnosis of CCAD—conventional angiography (CA), CT angiography (CTA), magnetic resonance angiography (MRA), and Doppler ultrasound (DUS)—possess complementary strengths and weaknesses. Our suggested algorithm for diagnostic evaluation is shown in Figure 2, but it should be recognized that individual clinical circumstances warrant careful, case-by-case consideration.
In the past 10 to 20 years, advances in technology have made MRI/MRA and CT/CTA increasingly sensitive and specific, although CA remains the gold standard for equivocal cases. In general, imaging modalities can be divided into those that can analyze the arterial lumen (CA), and those that can depict both the lumen and mural arterial thrombus (MRA with MRI, CTA, and DUS).
Some authors have hypothesized that the extracranial carotid and vertebral arteries are more vulnerable to dissection than other arteries in the body because of their mobility and proximity to bony projections of the cervical spine . The vertebral artery is particularly vulnerable at its tortuous course around the C1-C2 lateral masses and through the transverse foramina . As a result, two thirds of vertebral artery CCADs occur at these sites . The most common location for carotid CCAD is 2 to 3 centimeters above the carotid bulb .
CA is still widely considered the gold standard for diagnosis of adult and childhood CCAD, but the risks of this technique may outweigh its benefits in many clinical scenarios [18••]. CA depicts intraluminal findings of CCAD with very high spatial resolution through direct intra-arterial injection of contrast. The most common findings of extracranial CCAD on CA are arterial stenosis, aneurysm formation, or occlusion [19,20]. An intimal flap or double lumen is indicative of CCAD, but these findings are detected in fewer than 10% of dissected arteries, and less commonly in the vertebral arteries . A relative drawback of CA is that intramural hematoma or periarterial findings cannot be directly visualized .
Increasingly, CA is supplanted by MRI/MRA for the primary diagnosis of CCAD [21–23]. This trend is due to increased availability of MRI/MRA, potential complications of CA (e.g. femoral hematoma, femoral arterial pseudoaneurysm, recurrent AIS, and radiation exposure), and the need for sedation in CA . Additionally, as noninvasive techniques like MRI/MRA become more prevalent, fewer physicians are trained in CA, resulting in fewer experienced angiographers . Although concern for CA complications is one of the primary factors in limiting CA, several recent papers have documented an excellent safety record for CA in children in the hands of experienced angiographers [25,26]. However, because CA carries a higher risk of complications in patients with connective tissue disease, especially Ehlers-Danlos or an undiagnosed collagen abnormality, CA should be used with caution in this population .
In most centers, MRI/MRA has become the first-line imaging modality for patients with suspected dissection [21,22]. MRI/MRA is noninvasive, uses no radiation, and simultaneously images for dissection and stroke.
Arterial luminal findings of CCAD on either time of flight (TOF) MRA or contrast-enhanced MRA are similar to findings of CA in both adults and children, including arterial stenosis, intimal flap, dissecting aneurysm, or occlusion [4,28,29]. A tapered stenosis (“flame sign”) or a thin, segmented stenosis (“string sign”) is a less common sign of CCAD [28,30]. Intimal flaps and dissecting aneurysms are two specific luminal findings for CCAD, but they are infrequently visualized on MRI/MRA [28,31,32].
Although CA is the gold standard for diagnosis, one advantage of MRI/MRA (TOF or contrast-enhanced) over CA is the ability to directly visualize the intramural hematoma with T1 or T2 fat-saturated imaging as a crescentic hyperintensity along the vessel wall [28,33] (as seen in Fig. 1, Case 1).
Intramural hematomas have been reported in up to 76% to 91% of dissected vessels [28,31,32]. On MRI, the appearance of the intramural hematoma in CCAD changes over time: it has been reported as isointense to nearby tissues for the first day or two, then T1 isointense and T2 hyperintense, and finally T1 hyperintense after several more days [21,28,33,34]. T1 and T2 hyperintensity in an intramural hematoma can persist for months [21,33]. The combination of mural hematoma and flowing blood on TOF MRA can lead to an apparent increase in vessel diameter on both source images and maximum intensity projection (MIP) images. This increase in vessel diameter on TOF MRA was 99% specific for CCAD in one study of carotid and vertebral CCAD .
Several studies of CCAD in adults that were published between 1994 and 2002 compared MRI/MRA(TOF) versus CA. MRI/MRA had 50% to 100% sensitivity and 29% to 100% specificity [33,35–39]. In a more recent study of CCAD in adults, contrast-enhanced MRA was 89% sensitive for detection of CCAD, versus 50% for TOF MRA . This finding mimics our clinical experience. MRI/MRA can perform favorably compared with CA, and contrast-enhanced MRA is superior to TOF MRA. Compared with TOF MRA, contrast-enhanced MRA possesses several advantages in detecting arterial stenosis and dissection: imaging of the entire course of the cervical arteries in one acquisition , decreased overestimation of stenosis or occlusion due to slow flow , and fewer motion artifacts .
Limited studies in childhood AIS have demonstrated a 100% correlation between MRA (TOF) and CA in large-artery intracranial abnormalities [43, Class III]; other studies have shown that CA is more sensitive in intracranial abnormalities than extracranial abnormalities of the large vessels [44, Class IV]. Pediatric literature remains limited.
High-resolution MRI with specialized cervical surface coils is on the horizon for diagnosis of intramural hematoma in cases of CCAD in adults [48,49••,50, Class IV]. This promising technique could increase sensitivity for intramural hematoma and overcome the artifact related to confounding perivertebral venous plexus enhancement.
Recent studies report a high sensitivity of CTA (98–100%) for diagnosis of spontaneous and traumatic CCAD in adults [51–53]. Other advantages of CTA include widespread availability, speed of examination, noninvasiveness, and ability to be performed as part of an initial trauma screening. The great disadvantage of CTA, especially for children, is the high radiation burden . Furthermore, the diagnostic performance of CTA for CCAD in children has not been extensively studied.
The signs of dissection on CTA are similar to those on both MRA and CA. As in those modalities, the most common findings are stenosis or occlusion at a typical location [45,53,55,56]. The characteristic CCAD luminal findings of intimal flaps and dissecting aneurysms can also be visualized on CTA [55,56] more often than on MRA [45,57].
Artifacts associated with CTA include bone artifact near the skull base and artifact from dental amalgam [55,56]. High-quality CTA examinations also require excellent bolus timing and high injection rates, both of which can be more difficult to accomplish in children. References on CTA techniques in children have been published that address bolus timing and injection rates .
In adults, DUS in the diagnosis of spontaneous CCAD has reported sensitivities ranging from 66% to 96% [51,59–62]. Advantages of DUS are that it is relatively inexpensive, uses no radiation, and can be performed emergently at the bedside, but its sensitivity is highly dependent on the operator .
Compared with the anterior circulation, DUS is less sensitive in evaluating the vertebral arteries because of their location within the transverse foramina , their deeper and more irregular course, and their small diameter . DUS also has reduced sensitivity for focal disease of the carotid arteries near the skull base because of the limited window .
Because of its limited utility in evaluating the vertebral arteries, DUS has not yet been widely studied in children with CCAD.
Recent adult trauma literature cites a prevalence of blunt cervical arterial injury (BCAI) following major trauma as high as 1.1% to 1.2% [65,66], although a recent survey of the National Pediatric Trauma registry estimated the prevalence in pediatric trauma at 0.03%. Presumably, BCAI is either underdiagnosed in pediatric patients or its prevalence is lower than in adults .
Recent adult surgical and trauma literature advocates CTA as a screening tool for BCAI following major blunt trauma [65,66]. Controversy exists regarding the sensitivity of CTA in this setting, with sensitivities that range from 54% to 97.7% [52,68]. One pediatric study reported a sensitivity of 88% and a specificity of 100% in CTA in this setting .
Given the current recommendations in the trauma literature and the advantages of CTA in this setting (speed, 24/7 availability, and ability to scan multiple body parts at once), CTA is often used for BCAI screening in both children and adults, but given the relative lack of data on BCAI specifically in children and concerns over radiation exposure, MRI/MRA should be considered as an alternative screening tool when appropriate .
Diagnosis of craniocervical arterial dissection (CCAD) in children begins with a careful history and physical in a child with a transient ischemic attack (TIA) or arterial ischemic stroke (AIS). The extent of radiologic evaluation for suspected CCAD is based upon careful consideration of the risks associated with the best imaging techniques, weighed against the benefits of enhanced vascular imaging with better diagnostic sensitivity. Although conventional angiography (CA) and CT angiography (CTA) have a higher sensitivity than magnetic resonance angiography (MRA), they are accompanied by risks: for CA, femoral hematoma, femoral arterial pseudoaneurysm, recurrent AIS, and radiation exposure; for CTA, radiation.
For children (non-neonates) with suspected CCAD, MRI with MRA is recommended as the first-line imaging study. MRI usually includes diffusion-weighted, FLAIR, and T1 images of the brain, and T1 or T2 fat-saturation axial imaging through the neck. MRA should include 3D time-of-flight MRA of the head and neck (from the aortic arch through the circle of Willis). Contrast-enhanced MRA should be highly considered in neck imaging. If MRI/MRA is equivocal, CCAD is strongly suspected but not detected on MRI/MRA (especially in the posterior circulation), or the child has recurrent events, additional imaging of the craniocervical vasculature is likely warranted. Individual clinical circumstances warrant careful, case-by-case consideration.
Treatment of CCAD in children is challenging and differs for intracranial and extracranial dissections. In extracranial CCAD, we most commonly use anticoagulation for 6 weeks to 6 months in patients with TIA or AIS. Typically, unfractionated heparin is used in the acutely ill patient at heightened risk for bleeding (because of its short half-life), whereas low-molecular-weight heparin (LMWH) or warfarin are reserved for the stable patient. If the history is suspicious for dissection (head and neck trauma, recent cervical chiropractic manipulation, recent car accident, or neck pain), we consider treatment for dissection even with normal MRI/MRA. For patients with CCAD with a stroke size greater than one third to one half of the middle cerebral artery territory (or other bleeding risk factors) and extracranial CCAD, in whom there is concern about heightened risk for hemorrhagic conversion, we commonly use aspirin therapy during the acute phase. Regardless of their treatment in the initial weeks to months, we subsequently treat all patients with aspirin for 1 year after their event, and sometimes longer if they have other risk factors. Interventional techniques, such as extracranial cerebral arterial stent placement or selective occlusion, are understudied in children. Interventional techniques are typically reserved for patients who fail aggressive medical management and have recurrent TIA or AIS.
The diagnosis and treatment of intracranial dissection is extraordinarily challenging in children, in whom inflammatory intracranial arteriopathies are common. When intracranial arteriopathy is clearly associated with dissection, the clinician should look for the presence of subarachnoid hemorrhage and/or dissecting aneurysm. Treatment decisions should be made by a multidisciplinary pediatric stroke team, given the lack of data in this area. Intracranial cerebral artery stent placement carries high risk and is not recommended for intracranial CCAD in children.
Most importantly, we educate all children with CCAD and their parents about the paucity of evidence in the treatment of this disease, the risks of enhanced imaging techniques such as CTA or CA, and the challenges involved in weighing the risks of aggressive therapies and interventions against the costs of unclear diagnosis and potentially ineffective treatments. We also educate our patients with CCAD about the signs and symptoms of recurrence and the importance of emergent evaluation.
No potential conflicts of interest relevant to this article were reported.
Papers of particular interest, published recently, have been highlighted as:
• Of importance
•• Of major importance