Introduction: Reliably predicting neurological outcomes following cardiac arrest can guide care. Neuron specific enolase (NSE), a neuronal enzyme released following anoxic brain injury, offers a quantitative marker of brain injury when measured serially. However, because NSE is also present in red blood cells and platelets, elevations may occur from non-neurological sources in critically ill patients. In this study, we measured NSE levels over the first five days post-arrest in 180 patients at Boston Medical Center and Yale New Haven Hospital.
Methods: We assessed how daily NSE levels and five-day trends following arrest differed between patients with good and poor neurological outcomes. We also determined the highest NSE values in patients with good outcomes each day. Lastly, we examined how NSE trends varied by demographic and arrest-related factors: sex, ethnicity, race, age, arrest location, whether the arrest was witnessed, and whether bystander CPR was performed. Good neurological outcomes were defined as Cerebral Performance Category (CPC) score of 0 to 2 and modified Rankin scale (mRS) score of 0 to 3, indicating patients who can meaningfully interact with their environment and function independently, in addition to ambulating on their own, respectively. Discharge outcomes included brain death, withdrawal of life-sustaining therapy, death from a non-neurologic cause, and survival to discharge.
Results: Some patients had high NSE values on day 1 or 2 but improved if levels declined over time. Among those with good outcomes, peak NSE values steadily decreased after day two post-arrest. In contrast, patients who did not survive had persistently high values. Our findings suggest that the trend in NSE values over the first five days following arrest provide more prognostic value than isolated measurements, with some patients showing early elevations that declined over time and ultimately achieved a good neurological recovery.
Conclusions: NSE remains a valuable quantitative biomarker post-cardiac arrest, with persistently elevated values over serial testing associated with a poor outcome, whereas diminishing values associated with potential for good outcome. Serial testing is more informative than relying on any single value on any specific day post-arrest.
