(1407) Validation of Neuron-Specific Enolase in Cardiac Arrest Patients

Introduction: Neuron-specific enolase (NSE) is a widely used prognosticator in cardiac arrest survivors, despite the disadvantage that it might differ depending on the analysis method or equipment. To validate NSE cut-off values, we analyzed the prognostic performance of different NSE cut-off values using single-center data in a setting with the limited withdrawal of life-sustaining therapy.

Methods: We analyzed the data of adult CA survivors who underwent targeted temperature management (TTM) between January 2014 and December 2020. We measured the NSE level 48 hr and 72 hr after CA. We performed receiver operating characteristics (ROC) and used the reference value of NSE ( < 17 μg/L), the Youden index indicated NSE, and the guidelines-suggested NSE value ( < 60 μg/L) as a threshold. The primary outcome was 6-month neurological outcomes with Cerebral Performance Category (CPC), dichotomized into good (CPC 1 or 2) or poor (CPC 3–5).

Results: Of the 522 included patients, 355 (68.0%) patients had poor neurological outcomes. The area under ROC (AUC) of NSE at 48 hr was 0.887 (95% confidence intervals [CIs], 0.856–0.913) with a Youden index value of 35.6 μg/L. A false positive rate (FPR) of < 2% was observed (54.1 μg/L). Analyses with specific cutoff values of 17 μg/L, 35.6 μg/L, and 60 μg/L showed a sensitivity of 86.8%, 72.4%, and 58.0% and a specificity of 66.5%, 93.4%, and 98.8%, respectively. The AUC of NSE at 72 hr was 0.892 (95% CIs, 0.854–0.923) with a Youden index value of 30.4 μg/L. Analyses with specific cutoff values of 17 μg/L, 30.4 μg/L, and 60 μg/L showed a sensitivity of 86.5%, 76.6%, and 60.8% with specificities of 72.2%, 90.4%, and 98.3%, respectively. An FPR of < 2% was observed (53.6 μg/L).

Conclusions: The cut-off value of 60 μg/L was appropriate to predict poor neurological outcomes with an FPR < 2% in patients with cardiac arrest who underwent TTM, irrespective of NSE at 48 hr or at 72 hr.