Abstracts

Rationale: Infantile spasms (IS) typically present in association with hypsarhythmia (HYPS) and developmental regression. However, infantile spasms can present without HYPS 20% of the time but the significance of this finding as it relates to etiology, treatment or outcomes remains unclear. Methods: The multicenter National Infantile Spasms Consortium prospective database was used to compare IS patients less than 2 years of age with and without HYPS in terms of age, sex, prior developmental delay, etiology, first-line treatment choice and response to first-line treatment. Hypsarhythmia (HYPS) was defined as an EEG with multifocal spikes, poor background organization and an amplitude over 200 uV peak-to-peak. Patients without hypsarhythmia had abnormal EEGs but did not meet criteria for HYPS. Chi-square test was used to evaluate statistical significance. We fit a multivariable logistic regression models to predict response to therapy including interactions between HYPS and age. Results: Of 396 patients, 326 (82%) had hypsarhythmia and 70 (18%) did not. Overall, there was no statistically significant difference in gender, average age (7.08 months), or chance of preexisting developmental delay between the two groups. However, patients < 3 months of age at onset of infantile spasms were less likely to have HYPS compared to patients older than three months (65% and 84.9% respectively, P = 0.007). EEG background at diagnosis was significantly associated with etiology (P=0.0013). Specifically, neurocutaneous disorders were less likely to have HYPS compared to all other etiologies combined (Genetic/ metabolic, Neonatal injury, stroke, and unknown, (58.3% vs 83%, P=0.0019)). Infants without HYPS were less likely to receive standard treatment (e.g. ACTH, prednisolone, or vigabatrin), than those with HYPS (71% vs 87 %, p < 0.0005). EEG background, however, was not a determinant of response to medication (resolution of spasms at 2 weeks and no relapse for 3 months). Response was driven by treatment choice. Of the patients without HYPS none (0 of 20 patients) responded when treated with non-standard therapy compared to 48% (24 of 50 patients) of those treated with standard therapy (P = 0.0001). In a logistic regression model, only standard treatment and older age of onset predicted a response to the first medication. Conclusions: Patients who were less than 3 months old or had a neurocutaneous disorder were less likely to have hypsarhythmia on EEG at the time of diagnosis. Patients without hypsarrhythmia were more likely to be prescribed non-standard therapies compared to patients with hypsarhythmia at diagnosis. Receipt of a first line therapy (ACTH, Prednisolone or Vigabatrin) and older age of onset, not the EEG is the prime predictor of successful resolution of infantile spasms. Funding: Pediatric Epilepsy Research Foundation