Markers of neutrophil mediated inflammation associate with disturbed continuous electroencephalogram after out of hospital cardiac arrest

Background: Achieving an acceptable neurological outcome in cardiac arrest survivors remains challenging. Ischaemia-reperfusion injury induces inflammation, which may cause secondary neurological damage. We studied the association of ICU admission levels of inflammatory biomarkers with disturbed 48-hour continuous electroencephalogram (cEEG), and the association of the daily levels of these markers up to 72 hours with poor 6-month neurological outcome. Methods: This is an observational, post-hoc sub-study of the COMACARE trial. We measured serum concentrations of procalcitonin (PCT), high-sensitivity C-reactive protein (hsCRP), osteopontin (OPN), myeloperoxidase (MPO), resistin, and proprotein convertase subtilisin/kexin type 9 (PCSK9) in 112 unconscious, mechanically ventilated ICU-treated adult OHCA survivors with initial shockable rhythm. We used grading of 48-hour cEEG monitoring as a measure for the severity of the early neurological disturbance. We defined 6-month Cerebral Performance Category (CPC) 1-2 as good and CPC 3-5 as poor long-term neurological outcome. We compared the prognostic value of biomarkers for 6-month neurological outcome to neurofilament light (NFL) measured at 48 hours. Results: Higher OPN (P = 0.03), MPO (P < 0.01) and resistin (P = 0.01) concentrations at ICU admission were associated with poor grade 48-hour cEEG. Higher levels of ICU admission OPN [OR 3.18; 95% CI 1.25 - 8.11 per ln(ng/mL)] and MPO (OR 2.34; 95% CI 1.30 - 4.21) were independently associated with poor 48-hour cEEG in a multivariable logistic regression model. Poor 6-month neurological outcome was more common in the poor cEEG group (63% vs. 19% P < 0.001, respectively). We found a significant fixed effect of poor 6-month neurological outcome on concentrations of PCT (F = 7.7, P < 0.01), hsCRP (F = 4.0, P < 0.05), and OPN (F = 5.6, P < 0.05) measured daily from ICU admission to 72 hours. However, the biomarkers did not have independent predictive value for poor 6-month outcome in a multivariable logistic regression model with 48-hour NFL. Conclusion: Elevated ICU admission levels of OPN and MPO predicted disturbances in cEEG during the subsequent 48 hours after cardiac arrest. Thus, they may provide early information about the risk of secondary neurological damage. However, the studied inflammatory markers had little value for long term prognostication compared to 48-hour NFL. Editorial Comment: These findings show that new biomarkers and continuous EEG are disturbed following successful resuscitation after cardiac arrest at the ICU. However, these parameters alone do not add to the prediction of long-term outcomes. Thus, established markers and especially time and neurological assessments remain hallmarks in the prediction of outcome in post-cardiac arrest patients.