|ECR 2019 / C-3147|
|Neuroimaging findings in children with abusive head trauma|
Findings and procedure details
In most institutions unenhanced CT is the initial examination for children with suspected AHT. Initial CT should include soft tissue and bone settings. Volume-rendered CT images and multiplanar reformatting can increase the conspicuity of subtle fracture lines.
CT of the head should be performed in all children who present signs of abuse in combination with signs of possible neurotrauma or intraocular haemorrhages.
MRI is not the first imaging tool in AHT. The most important reason is a lower sensitivity for acute haemorrhage compared to CT. Secondly, the long scan time makes it more difficult to perform successfully in children, unless general anaesthesia is used.
Unenhanced MRI with DWI and susceptibility- weighted imaging (SWI) is sensitive for parenchymal injury and small volumes of haemorrhage (such as in diffuse axonal injury) and can provide valuable prognostic information.
Diffusion-weighted imaging (DWI) sequences can detect cytotoxic edema within 20–30 minutes of injury with excellent anatomic definition, and T2 and FLAIR sequences are positive within 4 hours from the injury.
Considering HII is more often seen in patients with AHT than in patients with accidental head trauma, when clinical suspicion for AHT is high, it could be beneficial to obtain an early MRI with DWI sequences regardless of an unremarkable CT scan.
Subdural haematoma is the most commonly observed intracranial lesion, in up to 90%, especially in children less than one year old. The specificity of SDH for AHT is increased when associated with RH and underlying diffuse parenchymal injury.
Venous injury is strongly associated with AHT. The mechanism of acceleration/deceleration associated or not with direct impact can lead to tearing of convexity bridging veins at the junction of the bridging vein and superior sagittal sinus. Additionally, rupture of the arachnoid membrane allows CSF to enter the subdural space, mixing with subdural blood producing a hematohygroma.
Hyperdense SDH is well shown at screening CT, although the detection of small-volume and iso- or hypodense collections may benefit from the increased contrast resolution of MRI.
SDH is generally thin and diffuse and is often unilateral or asymmetric; most commonly parafalcine in location and rarely causes significant mass effect. Fig 1, 2.
At CT, acute SDH can present with a variety of imaging appearances. Although, mixed-attenuation subdural haematomas are found with greater prevalence in AHT than in accidental head trauma, a descriptive approach of the density of the SDH at CT is recommended due to the number of variables involved in the appearance of SDHs on CT.
The appearance of a SDH can be affected by a number of factors, such as different time of resolution, haemoglobin levels, CSF or coexisting primary or secondary coagulopathy that can impair blood clotting.
Usually in the first three hours after trauma unclotted blood is isodense to the cortex. Fig 3. Hematohygroma can present as an acute iso- or hypodense subdural collection. Fig 4. Anemia (hemoglobin < 8–10 g/dL) can result in hypoattenuation of acute haemorrhage. Hyperdense SDH and bridging vein thrombosis are reliable indicators of an acute component of haemorrhage (up to 10 days after trauma), but the absence of these findings cannot be used to infer the opposite.
Mixed-density SDHs can result from acute-on-chronic haemorrhage but can also result from hyperacute bleeding, acute haemorrhage with sedimentation levels, and hematohygromas.
Therefore, follow-up MRI can be extremely useful and can give further information regarding the evolution of the extra-axial haemorrhages that may aid an estimation of the age of the lesions as well as providing prognostic information related to the resolution or persistence of damage to the brain itself.
The vertex should be scrutinized in all cases for bridging vein thrombosis, which will appear hyperdense at CT and hypointense on susceptibility-sensitive MRI sequences (T2* or SWI). Morphologically, vein rupture appears as a linear intravascular clot extending into more globular perivascular thrombus at the site of vessel rupture. This appearance has been termed the “tadpole” or “lollipop” sign.
Acute SDH either will rapidly resolve or will rapidly evolve into a subdural hygroma through the transudation of interstitial fluid or CSF; this subdural hygroma will undergo resorption in most cases.
A subdural hygroma may persist and form vascularized neomembranes along the outer margins and then the inner margins. Repeated bleeding, either spontaneous or with trivial trauma, from these fragile membranes can transform the collection into a chronic SDH. Chronic SDHs are differentiated from hygromas by their loculated configurations and evolving mass effect over time. Enhancing membranes are one of the few reliable indicators of nonacuity in a subdural collection.
b. Diffuse parenchymal injury
Although present in less than one-third of patients with suspected AHT, studies show that in these children hypoxic-ischemic injury (HII) mechanisms have worse clinical outcomes (Fig 5) and they can be a significant cocontributor to the mortality in AHT.
Hypoxia-ischemia, perhaps secondary to a combination of central apnea, vascular compromise or strangulation is likely the most prominent contributing factor to parenchymal damage in children with AHT.
MRI is more sensitive than CT in delineation of parenchymal injures. DWI dramatically increases the conspicuity of cytotoxic injury because the T2 prolongation of brain edema is subtle in the unmyelinated brain. Diffuse parenchymal injury is rarely noted as an isolated finding in AHT except in cases of asphyxiation or strangulation.
Most studies have shown a watershed pattern of diffusion restriction in most AHT cases; however, multiple patterns of AHT-related parenchymal injury can be seen. Kadom et al reported three main patterns of DWI abnormalities: (1) bilateral ischemia Fig 6, 7, 8, 9 (2) unilateral ischemia Fig 10, 11, and (3) contusion Fig 12.
The mechanism behind prominently asymmetric HII has not yet been clarified, it may result from transient unilateral vascular occlusion. Another theory called a “second impact syndrome” refers to an unilateral HII and ipsilateral SDH. Although of unknown cause, second impact syndrome has been hypothesized to result from disordered cerebral autoregulation and to possibly be directly attributable to the overlying SDH. These findings suggest that the HII seen in AHT is likely multifactorial. A thin SDH with disproportionate underlying mass effect can be seen with early unilateral HII and should prompt consideration of AHT and further evaluation with MRI.
c. Focal parenchymal injury
The focal parenchymal lesion may be due to injury mechanisms secondary to impulsive load or impact on AHT, it is less common and less characterized than diffuse parenchymal lesion.
Shear injury can result in distinctive parenchymal lacerations or contusions in children under 1 year of age. Focal haemorrhagic contusions may be present in patients of any age in the context of impact loading injuries, usually with a skull fracture superimposed and often with a coup-contrecoup pattern Fig 13.
Parenchymal lacerations, also known as contusional tears, contusional clefts or sliding contusions may appear as linear disruptions of the cerebral parenchyma predominantly involving the subcortical white matter or in the gray-white junction of the gyral crests, usually in the frontal lobes. In general, they are hemorrhagic with levels of sedimentation in both CT and MRI and may be associated with overlying subpial haemorrhage.
FLAIR images are increasingly sensitive for the detection of lesions, while the SWI is substantially more sensitive. It has been shown that microbleeds in the SWI correlate with a poor clinical outcome in patients with hypertension, especially when associated with an underlying diffuse parenchymal lesion.
d. Skull fractures
Skull fractures are not specific for AHT. Simple fractures and small associated extra-axial or subperiosteal hemorrhages are common findings in low-impact accidental cranial injuries and are commonly observed over short distances. However, multiple, complex, diastatic or growing fractures suggest a high-energy mechanism and are troubling without an appropriate history of trauma. Fig 14.
CT performed with bone algorithm reconstructed in a submillimeter thickness is the best modality for the evaluation of skull fractures in children. Multiplanar reconstructions and VR images are useful for identifying fractures. Figure 15.
e. Cervical injury
Infants have an increased risk of upper cervical spine injury. In children under 1 year of age, bone or ligamentous injuries in the CCJ prevail, and it is believed that they are due to a cranial movement similar to that of a whiplash with an impulsive load. An incidence of almost 10% in the context of positive findings in a skeletal survey, significantly associated intracranial injury.
MRI should be included in the diagnostic study when there is evidence of intracranial injury.
Types of injuries include extra-axial spinal hemorrhage; cord laceration, contusion and haemorrhage; avulsions of nerve roots; and ligamentous and soft- tissue injuries.