MRI shows traumatic brain lesions not shown on CT

MRI indicated brain contusions and axonal injury independently associated with poorer three-month outcome after mild traumatic brain injury (MTBI), according to a study published online Dec. 7, 2012, in Annals of Neurology.

Although many of the approximately 1 million patients who suffer MTBI in the U.S. have a good prognosis, a subset of patients develops persistent dysfunction. However, there is a lack of clinical, laboratory and imaging biomarkers for MTBI. In addition, researchers have recognized that existing classification schemes for TBI may be limited and obscure differences among diverse patient populations.

Esther L. Yuh, MD, PhD, from the Brain and Spinal Injury Center in San Francisco, and colleagues designed a prospective study to assess the relevance of early CT and MRI on three-month MTBI outcomes, controlling for known factors such as age, prior head injury, educational background and employment status associated with poorer outcomes.

The researchers enrolled 135 patients presenting to the emergency departments of three Level 1 trauma centers. Patients underwent head CT on admission and brain MRI 12 days after injury. Both imaging studies were analyzed using TBI common data elements (TBI-CDEs). The eight-point Extended Glasgow Outcome Scale (GOS-E) at three months served as the primary outcome measure.

Yuh and colleagues reported that MRI identified many more acute traumatic intracranial lesions than CT. A total of 27 percent of participants had abnormal CT results. However, among the 98 patients without CT evidence of skull fracture or acute intracranial injury, 28 percent had abnormal MRI results. These findings included 23 patients with hemorrhagic axonal injury, three patients with brain contusions and four patients with extra-axial hematomas.

The researchers devised three multivariate models focused on features predictive of three-month GOS-E. The most comprehensive model incorporated clinical, demographic/socioeconomic, CT and MRI features and was highly statistically significant for the variability in three-month GOS-E. It accounted for 20.6 to 21.9 percent of the variability, according to Yuh et al. Presence of one or more brain contusions was the strongest predictor of poor outcome.

Yuh and colleagues emphasized that this study brings progress toward evidence-based classification of injury severity. However, they noted it may not be cost-effective to refer all MTBI patients for routine MRI. Developments in MRI technology, including less costly head-only systems, may help to address this challenge, according to the researchers.