Abstracts

Rationale: Infantile spasms (IS) is an epilepsy syndrome characterized by clinical spasms with hypsarrhythmia and developmental delay or regression. Brain MRI is a standard part of the initial work-up in newly diagnosed infantile epilepsy or IS and shows etiologically relevant findings in about 50% of patients. However, in the vast majority of patients with Down syndrome (DS) and IS, the etiology of IS is thought to be secondary to molecular or microstructural changes at the synapse, rather than large structural anomalies. Neuroimaging is done in many patients with IS and DS, which may be an unnecessary expense and risk due to anesthesia. To determine if clinically useful etiologic information is provided by neuroimaging in the DS population with IS we performed a detailed analysis of the neuroimaging findings. Methods: The clinical research tool I2B2 was used to identify patients with IS and DS cared for at Boston Children’s Hospital (BCH) between 2001 and 2016. Retrospective chart review confirmed the diagnosis of IS and DS, identified timing, indication and type of imaging done, and determined any significant imaging findings that directly impacted medical decision-making. All available images were independently reviewed by a pediatric neuro-radiologist, who was blinded to the prior MRI reports. We classified the MRI as normal, abnormal with structural abnormalities that are commonly reported with IS, or abnormal with incidental findings. Results: We identified 69 patients of which 51 patients (65% male, 35% female) had both IS and DS and sufficient documentation in the BCH system. Of these patients, 38 (75%) had brain MRIs in the BCH imaging system, and 4 more patients had reported normal MRIs at outside institutions (total 42, or 82% of patients with brain MRIs). Of these 42 patients, 33 patients (79%) had imaging primarily because of the diagnosis of IS. Of the 38 scans in the BCH system, 16 patients (42%) had normal imaging. Nineteen (50%) patients had incidental imaging findings; the most common findings were mild hippocampal dysmorphisms (n = 6, 16%) and/or prominent ventricles or sylvian fissures (n =9, 24%). Two patients had imaging confirmation of previously known diagnoses (hypoxic-ischemic injury and in utero stroke). One patient had evidence of prior intra-ventricular hemorrhage, not clinically suspected, without any subsequent hydrocephalus, as well as abnormal T2 signal which prompted a metabolic work-up that was ultimately normal, with no change made to the treatment plan. Conclusions: In patients with IS and DS, MRI is not likely to demonstrate additional structural abnormalities or provide further etiological diagnoses that would alter treatment decisions. Given the higher risk of anesthesia in the DS population and the low yield of imaging, routine brain MRIs may not be indicated in the work-up of IS in the DS population, except when there is a high index of suspicion for a secondary lesion. Funding: Fred Lovejoy House-staff Research and Education Fund