Abstracts

Rationale: Burst suppression pattern (BSP) on EEG  in post-cardiac arrest patients is traditionally associated with poor outcome. It is especially regarded as incompatible with recovery if it persists or arises spontaneously in the absence of anesthesia or cooling. However, recent models suggest that BSP may be a means of cellular preservation in a low energy environment. In this context, we reviewed a cohort of 73 post-cardiac arrest patients and characterized the evolution of spectral features of BSP in patients with favorable and unfavorable outcome. In two patients re-emergence of spontaneous BSP associated with complete neurological recovery motivating the comparison of outcomes across the group. Methods: Retrospective EEG data were collected from 73 patients who suffered cardiac arrest and underwent therapeutic hypothermia at Weill Cornell Medical Center between 2010-2013. Six patients with BSP EEG who did recover consciousness were randomly selected for further analysis. Age and sex matched patients with BSP EEG who did not recover consciousness were also selected for comparison. Power spectra taken from burst and suppression periods were generated for these patients. Spectral characteristics were then compared within patients over time and among patients to identify features that may differentiate between those with and without favorable outcomes. Results: BSP was seen within 72 hours of arrest in 45/73 (61%) patients. Comparable numbers of patients with and without BSP regained consciousness (31.1% and 35.7% respectively). We observed a prominent early theta feature (5-7 Hz) within ‘bursts’ in all selected patients who subsequently recovered consciousness. In patients who did not recover consciousness, this ‘burst-theta’ peak was either absent or present within 48 hours after arrest then lost. We also identified two unique cases where BSP re-emerged spontaneously in the absence of anesthetic sedatives after rewarming was complete. In both patients a prominent theta feature was seen in bursts during spontaneous BSP and importantly, both patients regained consciousness. Conclusions: Given these findings, we speculate that in some cases, the emergence of spontaneous BSP may serve as a neuro-protective mechanism during times of severe metabolic stress including anoxic coma after cardiac arrest. Current models suggest that BSP may prolong cell survival in such circumstances by activating the membrane potential in a discontinuous manner. We also speculate that the presence of a prominent, early, theta peak within bursts may indicate potentially viable neuronal networks necessary for eventual recovery of consciousness. These results support the idea that BSP may not always indicate a state of irreparable neuronal damage. Rather, that the prognostic implications of this brain state may vary based on various characteristics intrinsic to the signal itself including spectral content. Our work emphasizes the need for further investigation into the physiological significance of BSP in this patient population, to more clearly understand it’s potential significance in recovery. Funding: NIH NINDS K23 NS096222Leon Levy Foundation