medwireNews: Magnetic resonance imaging (MRI) scanning within 30 days of severe traumatic brain injury (TBI) may help to more accurately determine the long-term prognosis of children, US researchers believe.
Incorporating MRI findings into a conventional clinical model significantly improved the ability to predict a favorable or unfavorable outcome at 6 or 12 months compared with clinical markers alone, report Peter Ferrazzano (University of Wisconsin, Madison) and co-workers.
They assessed information from 233 children in the USA, UK, or Australia with severe TBI who were enrolled in the ADAPT study between February 2014 and September 2017. The children were aged up to 17.0 years (median 6.9 years), 57.5% were boys, and all had a Glasgow Coma Scale (GCS) score of 8 or less (median score 6).
The most common cause of TBI was impact (69.5%), followed by falls (15.5%) and acceleration/deceleration (9.0%), and MRIs were taken a median of 134.6 hours after TBI.
The scans were assessed for signs of contusion, ischemia, diffuse axonal injury, intracerebral hemorrhage, and brainstem injury, and the findings were collated with those of the Glasgow Outcome Scale–Extended for Pediatrics (GOSE-Peds) assessment of functional outcome at 3, 6, and 12 months, based on level of consciousness, activities of daily living, and return to school.
MRI revealed contusion in 79.8% of patients, ischemia in 35.2%, diffuse axonal injury in 72.1%, intracerebral hemorrhage in 16.3%, and brainstem injury in 39.1%.
In multivariable analysis of the clinical variables, both the motor component of the GCS (odds ratio [OR]=0.71) and pupil score (one or both fixed vs neither fixed, ORs=1.79 and 47.29, respectively) correlated significantly with the GOSE-Peds score, after adjusting for age, sex, and time of assessment.
When only the MRI measures were assessed, multivariable analysis showed a significant correlation between contusion volume, number of regions with ischemia, and presence of brainstem injury with GOSE-Peds score, with adjusted ORs of 1.19, 2.28, and 8.44, respectively.
And when both clinical and MRI measures were included in the analysis, there were significant correlations between contusion (OR=1.13), ischemia (OR=2.11), brainstem injury (OR=5.40), and pupil score (ORs=1.74 and 10.12 for one and two pupils fixed, respectively) with the GOSE-Peds score.
Moreover, when MRI measures were added to a clinical model for predicting GOSE-Peds score, there was a significant improvement in the ability to best explain variance in outcome with fewest predictors, with an Akaike information criterion score of 1763 versus 1795.
The area under the receiver operating characteristic curve (AUC) of the combined predictive model was greater than the clinical-only model at each timepoint for predicting a GOSE-Peds score of 3 or less (favorable outcome) versus 4–7 (unfavorable outcome).
Specifically, the MRI-only and combined model AUCs were significantly higher than the clinical-only model at 6 months (0.76 and 0.77 vs 0.67, respectively) and 12 months (0.71 and 0.69 vs 0.58), whereas there was no significant different between the MRI-only and combined models at these times.
“To our knowledge, this prognostic study is the first to establish the added value of clinical MRI findings in predicting long-term outcome after severe TBI in children,” Ferrazzano and co-workers say.
They suggest: “Inclusion of MRI scanning in future pediatric TBI clinical trials may allow for improved stratification of patients.
“Additional research is needed to define the optimal timing of MRI scanning after TBI, and to determine the evolution of MRI findings over the course of postinjury brain development.”
medwireNews is an independent medical news service provided by Springer Healthcare Ltd. © 2024 Springer Healthcare Ltd, part of the Springer Nature Group