The influence of hemolysis upon admission on the performance of neuron-specific enolase in predicting death after resuscitated out-of-hospital cardiac arrest – a substudy of the IMICA trial

Abstract

Abstract Background The predominant cause of mortality among out-of-hospital cardiac arrest (OHCA) patients admitted to the intensive care unit in a comatose state is hypoxic-ischemic brain injury and consequent withdrawal of life sustaining therapies. Current guideline for post-resuscitation care suggests a multimodal approach for neuroprognostication including measuring a blood-born biomarker of neuronal injury termed neuron-specific enolase (NSE). However, NSE is also present in erythrocytes and therefore hemolysis – which frequently occurs immediately after resuscitation – must be assessed by measuring free-hemoglobin (free-hgb), and samples discarded if the threshold index is surpassed to minimize false positive results. In addition, the plasma half-life of NSE is much longer than free-hgb, and hemolysis occurring prior to the sampling point of NSE measurement, may cause elevated NSE levels due to destruction of erythrocytes rather than neurons, although free-hgb is no longer detectable in plasma. Purpose To investigate the relationship between hemolysis at hospital admission and NSE measured at 48 hours, and to determine the influence of hemolysis on the clinical performance of NSE in predicting all-cause mortality in comatose OHCA patients. Methods Free-hgb was measured at admission (0h) to quantify hemolysis. NSE was measured at 48 hours after admission. In each NSE sample hemolysis index was measured and mathematically corrected according to standard laboratory procedure. The relationship between free-hgb at 0h and NSE at 48h was investigated by Pearson's correlation analysis. The area under the receiver operating characteristic curves (AUROC) for prediction of all-cause mortality at 30 days were determined for both NSE and free-hgb, and the difference in AUROC between the two was calculated. Further, the positive and negative predictive values (PPV and NPV respectively) of NSE with a cut-off value at >60 μg/l (as recommended by guidelines) in predicting all-cause mortality were calculated. Results Admission free-hgb and NSE at 48 hours were available for 64 consecutive patients surviving beyond 48 hours. The 30-day mortality in this population was 33%. There was no correlation between free-hgb at 0h and NSE at 48h (r=0.13, p=0.30; Figure 1). The AUROC for NSE was 0.92 (Figure 2), and 0.59 for free-hgb at 0h. The PPV for NSE>60 μg/l was 1.0 (1.0–1.0) and NPV 0.84 (0.74–0.94). Conclusions The performance of NSE at 48 hours for predicting all-cause mortality among out-of-hospital cardiac arrest patients was not influenced by the degree of hemolysis at the time of admission, and NSE at 48 hours had both high positive and negative predictive values irrespectively of early hemolysis. Funding Acknowledgement Type of funding sources: None.