Background: Because of concerns about radiation exposure, some centers consider magnetic resonance imaging (MRIs) the preferred imaging modality for pediatric thoracic and/or lumbar compression fractures. The purpose of this study was to evaluate the sensitivity of computed tomography (CT) and MRI in diagnosing thoracolumbar compression fractures and the utility of MRI in their management.
Methods: Retrospective review identified 52 patients aged 0 to 18 years with 191 thoracic and/or lumbar compression fractures who had both CT and MRI during the initial trauma evaluation. The decision to perform CT and/or MRI was made by the attending pediatric spine surgeon. In all cases the CT scan was performed before the MRI. All imaging studies were reviewed by a board-certified pediatric radiologist and attending pediatric spine surgeon.
Results: Only 10 patients (19%) had a single-level injury. Of 42 with multiple compression fractures, 34 (81%) had fractures in contiguous levels, and 8 had noncontiguous injuries. Comparing CT and MRI, there was complete agreement in the number and distribution of fractures in 23 patients (44%). MRI identified additional levels of fracture in 15 patients (29%); 14 (27%) had fewer levels fractured on MRI than CT. Only one patient (2%) had fractures seen on MRI after a normal CT scan. Complete correlation between CT and MRI was seen in 59% (17/29) of patients aged 11 to 18 years, compared with 26% (6/23) of patients younger than 11.
Conclusions: In pediatric patients with mild thoracic or lumbar compression fracture(s), CT scan demonstrates a high sensitivity in determining the presence or absence of a fracture compared with MRI. Although some variability exists between the 2 modalities in the exact number of spinal levels involved, the definitive treatment and outcome were not changed by the addition of MRI. The information that may be obtained from an MRI must be weighed against the increased time and expense of the study, as well as the risks associated with sedation when necessary.
Level of evidence: Level II-diagnostic study.