NSE and S-100B are not sufficiently predictive of neurologic outcome after therapeutic hypothermia for cardiac arrest.

INTRODUCTION: Prognostication of cardiac arrest survivors is challenging since therapeutic hypothermia (TH) has been introduced. We evaluated serum biomarkers and motor response.
METHODS: This was a retrospective data analysis including patients in the years 2007-2012. Blood was drawn and a neurological examination was performed on admission and every morning. Outcomes were evaluated 6 months after discharge and dichotomized into good (cerebral performance category (CPC)=1 or 2) and poor (CPC=3, 4 or 5).
RESULTS: 123 patients (79.7% male, 63±14 years) received TH; 50% had a good outcome. On admission, S-100B (P=0.004) was significantly associated with the outcome, as well as neuron-specific enolase (NSE; P=0.020) and S-100B (P=0.004) on day 1 after admission. NSE on day 2, NSE progression from day 1 to 2 and motor response on day 3 also predicted the outcome (all P<0.001). NSE>33μgl(-1) only predicted a poor outcome with a specificity of 76%. An absent motor response on day 3 was the most sensitive marker (94%). NSE>41.1μgl(-1) combined with S-100B>0.461μgl(-1) on day 1 was the most specific marker (96%).
CONCLUSION: Although NSE and S-100B levels are associated with the outcome, the use of previously described cut-off values was insufficiently predictive of neurologic outcome. Caution should be exercised in the use of these tests to provide neuroprognostication.