Early NSE measurements at the ICU bear prognostic value in patients with and without CPR

Abstract Funding Acknowledgements Type of funding sources: None. Background Neuron-specific enolase (NSE) represents a diagnostic tool for identifying poor neurological outcomes in patients following CPR. As part of multimodal prognostication, NSE values above 60 µg/L after 48h and/or 72h were associated with a worse neurologic outcome (European Resuscitation Council Guidelines 2021). Despite the prognostic value of NSE 48h to 72h after CPR, measurements at admission to the ICU remain the topic of discussion as divergent findings are present in the literature (primarily in small collectives). Purpose Within the proposed study, we aimed to illuminate the prognostic value of NSE measurements on the day of admission to the ICU. Furthermore, we conducted a comparison of patients with and without CPR prior to ICU admission. Due to hemolysis sensitivity, we adjusted NSE values according to a previously published protocol to reveal potential changes in predictive potential through the occurrence of hemolysis. Methods This was a prospective, observational cohort study. 232 consecutive patients admitted to a cardiac ICU were screened over the course of one year. Patients with active malignancies or traumatic brain injuries were excluded. Furthermore, samples with hemolysis index >30 units were excluded to guarantee stable measurements. NSE values were measured on the day of admission to the ICU in a total of 202 patients (154 without previous CPR, 48 with CPR) using specific ELISA kits. According to an earlier publication, hemolysis was considered to account for an artificial increase of 0.29µg/L per unit (range 5-30 units) and was subsequently subtracted. Results Median age was 66.96 (IQR: 55.7-76.9), and 41.1% of the included patients were female. The total 30-day-mortality was calculated to be 27.2%, and 48 patients (23.8%) were admitted following CPR. Interestingly NSE at admission was not higher in patients after CPR (3.23 ± 3.88 µg/L vs 5.43 ± 5.78 µg/L, p = 0.072). However, NSE measurements on the day of admission were significantly elevated in 30-day none-survivors (3.23 ± 3.84 µg/L vs 6.91 ± 6.55 µg/L, p < 0.001). The same findings were made when excluding all patients after CPR (30-day survivors: 3.24 ± 3.99 µg/L vs 30-day none-survivors: 6.15 ± 6.17µg/L, p = 0.001). No statistically significant differences in survival were found when stratifying for the primary cause of admission. Comparing hemolysis-corrected NSE values to raw data showed a slight increase in positive predictive value for 30-day mortality in the total cohort (ROC-AUC: 0.732 vs 0.721) as well as in patients without prior CPR (ROC-AUC: 0.722 vs 0.703). Conclusion We provided evidence for the prognostic value of early NSE measurements at the ICU. In both the total cohort as well as a sub-cohort without patients admitted after CPR, 30-day none-survivors displayed markedly elevated levels of NSE. Moreover, the positive predictive value of these measurements was slightly increased when correcting for hemolysis >5 units.