Neuromarkers and neurological outcome in out-of-hospital cardiac arrest patients treated with therapeutic hypothermia–experience from the HAnnover COoling REgistry (HACORE)

Abstract

Background Neuron-specific enolase (NSE) and S-100b have been used to assess neurological damage following out-of-hospital cardiac arrest (OHCA). Cut-offs were derived from small normothermic cohorts. Whether similar cut-offs apply to patients treated with hypothermia remained undetermined. Methods We investigated 251 patients with OHCA treated with hypothermia but without routine prognostication. Neuromarkers were determined at day 3, neurological outcome was assessed after hospital discharge by cerebral performance category (CPC). Results Good neurological outcome (CPC≤2) was achieved in 41%. Elevated neuromarkers, older age and absence of ST-segment elevation after ROSC were associated with increased mortality. Poor neurological outcome in survivors was additionally associated with history of cerebrovascular events, sepsis and higher admission lactate. Mean NSE was 33μg/l [16–94] vs. 119μg/l [25–406]; p<0.001, for survivors vs. non-survivors, and 21μg/l [16–29] vs. 40μg/l [23–98], p<0.001 for good vs. poor neurological outcome. S-100b was 0.127μg/l [0.063–0.360] vs. 0.772μg/l [0.121–2.710], p<0.001 and 0.086μg/l [0.061–0.122] vs. 0.138μg/l [0.090–0.271], p = 0.009, respectively. For mortality, thresholds of 36μg/l for NSE and 0.128μg/l for S-100b could be determined; for poor neurological outcome 33μg/l (NSE) and 0.123μg/l (S-100b), respectively. Positive predictive value for NSE was 81% (74–88) and 79% (71–85) for S-100b. Conclusions Thresholds for NSE and S-100b predicting mortality and poor neurological outcome are similar in OHCA patients receiving therapeutic hypothermia as in those reported before the era of hypothermia. However, both biomarkers do not have enough specificity to predict mortality or poor neurological outcome on their own and should only be additively used in clinical decision making.