Abstracts

RETURN TO ABSTRACT LISTING


E1890. Predictors of Positive MRI in the Diagnostic Approach to Pediatric Epilepsy
Authors
  1. Kyle Robbins; University of Texas Southwestern Medical Center
  2. Patrick Arraj; University of Texas Southwestern Medical Center
  3. Cory Pfeifer; University of Texas Southwestern Medical Center
Objective:
Seizure imaging is a major component of pediatric neuroradiology. In facilities where 3T MRI is available, it is often requested to assess for epileptogenic foci provided that there are no contraindications for 3T imaging. Evaluation for focal cortical dysplasia may be dependent on the experience of the radiologist given the subtlety of the abnormality. This study looks at 1,373 consecutive MRI exams in children ordered with seizure as the indication to assess the effect of age, static field strength, and radiologist on the outcome of the MRI.

Materials and Methods:
This study was performed at a large children's hospital with 8 MRI scanners in the system, 3 with a static field strength of 3T and 5 with a static field strength of 1.5T. 1,373 consecutive MRI examinations of the brain ordered for seizure evaluation and interpreted by 6 pediatric neuroradiologists were reviewed. The exam was considered positive if it demonstrated a structural abnormality, cortical dysplasia, heterotopia, neoplasm, or signal abnormality that could be linked to the presentation of seizure. In cases where there was ambiguity as to whether the exam was positive, the study was reviewed by a radiologist with fellowship training in pediatric radiology and pediatric neuroradiology who made the determination. Logistic regression analysis was performed with a positive MRI as the dependent variable. Independent variables tested were age, sex, interpreting radiologist, use of general anesthesia, and static field strength (1.5T vs. 3T). Statistical analyses were performed using Stata.

Results:
The average age of the patient scanned was 6.45 years, and 738 patients were male (54%). The experience range of the interpreting radiologist ranged from 1 to 29 years. 1,073 studies (73%) were performed under general anesthesia. There were 303 positive studies for a positive rate of 22%. When assessed independently, age (p = 0.057), sex (p = 0.526), use of general anesthesia (p = 0.640), and interpreting radiologist (p = 0.872) were not significant predictors of positivity. Static field strength of 3T was a significant predictor of positivity (p < 0.001). Among 28 studies that revealed a focal cortical dysplasia for a first-time MRI, 10 were performed at a static field strength of 3T. There were 12 studies in which an MRI performed at 3T followed an exam performed at 1.5T. None of these studies demonstrated a lesion not seen at 1.5T, however, 1 exam was able to exclude a possible abnormality visualized at 1.5T.

Conclusion:
MRI uncovers a seizure focus in 22% of studies. The value of 3T MRI is demonstrated via logistic regression in this large study assessing MRI performed to evaluate pediatric patients with epilepsy. Despite the large range of experience between 6 pediatric neuroradiologists interpreting MRI studies ordered for seizure, the identity of the radiologist did not predict a positive scan. The rate of children receiving general anesthesia in this exam type is high.