This Month in Psychopharmacology

Novel MRI Screening Is Sensitive to Intracranial Hemorrhage for Well-Appearing Infants at Risk for Abusive Head Trauma

Abusive head trauma (AHT) in infants, if misdiagnosed, can result in serious medical consequences, even death. In a 1999 study, 31% of infants diagnosed with AHT had been evaluated by a physician after the injury occurred, but were originally misdiagnosed as having viral gastroenteritis and viral syndrome. Of the five deaths in the study, four were considered to be preventable if the initial diagnosis of AHT had been made. A neuroimaging technique that is sensitive enough to identify subdural hemorrhage in this population is important because it is present in more than 90% of the cases of AHT in infants. Computerized tomography (CT) has been the gold-standard for assessing infants with potential AHT because the scan can be completed quickly and does not require sedation; however, there is increasing concern about the radiation exposure linked to cancer, particularly in young children. In severe cases, assessment of brain damage outweighs the risk of radiation exposure as a result of CT. However, in well-appearing infants who present with non-specific symptoms—such as vomiting, fussiness, bruising, and lethargy—there is reluctance to subject the infant to a CT scan.

Conventional magnetic resonance imaging (MRI) is challenging for this population due to the need for sedation, which elicits concerns about the effects of anesthesia on the developing brain. In contrast, rapid-sequence MRI is a fast T2-weighted sequence that eliminates the costs and risks associated with sedation of infants, along with concerns about radiation exposure. In a recent study by Flom et al. a novel screening MRI protocol was designed and tested to determine if it could identify intracranial hemorrhage in well-appearing infants at risk for AHT, thus reducing the need for exposure to CT. The derivation cohort consisted of 9 children with AHT and the validation cohort was comprised of 78 children with AHT and age-matched controls with normal brain imaging. All children were between 30 and 364 days old. The diagnosis of AHT was made by the hospital-based Child Protection Team that includes child abuse pediatricians. All infants were well-appearing and presented with seven possible symptoms: fussiness and poor feeding, fussiness and/or vomiting, possible seizure activity, apparent life-threatening event, head swelling, macrocephaly, or bruising.

For the validation cohort, the sensitivity of the screening MRI for intracranial hemorrhage was 100%, while specificity was 83%. The purpose of this study was not to eliminate the need for CT, but to reduce use when not needed, thus decreasing radiation exposure. The novel MRI technique is suggested as an appropriate screening tool for well-appearing infants, so that brain injury is not missed, and AHT is not misdiagnosed. If there is evidence of intracranial hemorrhage observed in the screening MRI, then CT is recommended to determine the extent of brain injury in these patients. Importantly, the findings from this study offer a novel screening technique for well-appearing infants who may be at risk for AHT, a population that is often missed because CTs are not routinely conducted for well-appearing infants. Of the 53 infants in the study with normal CT, 44 had a normal screening MRI and would not have needed to undergo CT. This would have resulted in an 83% decrease in head CT use.

This study presented a highly specific and sensitive screening MRI technique that does not require sedation for infants, and has the potential to 1) accurately diagnosis well-appearing infants who are at risk for AHT, and 2) reduce the need for infants to be exposed to radiation by decreasing the need for CT scans in this population. Although additional research is needed, this screening tool has the potential to accurately diagnosis well-appearing infants who are at risk for AHT, and reduce the need for infants to be exposed to radiation by decreasing the need for CT scans in this population.

>> Flom L et al. Pediatr Radiol 2016;46(4):219-26.