Performance of neuron-specific enolase and neurofilament light chain in predicting death after hospital admittance from out-of-hospital cardiac arrest - A sub study of the IMICA trial. (2nd May 2022)

• Record Type: Journal Article
• Authors: Meyer, M; Wiberg, S; Grand, J; Obling, L; Isse, Y; Frikke-Schmidt, R; Kjaergaard, J; Hassager, C
• Is Part Of: European heart journal. Volume 11(2022)Supplement 1
• Publication Date: 2022-05-02
• Subjects: 616.1205
• DOI: 10.1093/ehjacc/zuac041.164
• Languages: English
• ISSNs: 2048-8726

Abstract

Abstract: Funding Acknowledgements: Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): The study was supported by funding from: The Danish Heart Foundation (Reference No. 19-R135-A9302-22125), and "Region Hovedstadens Forskningsfond til sundhedsforskning" (Capital Region Research Foundation, Denmark; Reference No. A6030). Background: Patients who remain comatose at hospital admission after resuscitation from out-of-hospital cardiac arrest have a high mortality risk. The majority of deaths can be attributed to anoxic brain injury and consequent withdrawal of life sustaining therapies after neuroprognostication. Accordingly, patients who do not regain consciousness after the initial days in the ICU are evaluated by a multimodal approach for neuroprognostication, which includes the level of neuron-specific enolase (NSE) obtained from a blood sample. At present NSE is the only clinically available marker for neuroprognostication, but Neurofilament Light chain (NFL) has also been evaluated post hoc in clinical studies and have shown great potential. Purpose: To establish the performance of both NSE and NFL in predicting death from all causes in out-of-hospital cardiac arrest patients who remain comatose following resuscitation. Methods: The IMICA trial was a single center randomized controlled trial in which comatose Out-of-hospital cardiac arrest patients were randomized to receive either a single one-hour infusion of tocilizumab of 8mg/kg (maximum 800mg) or placebo in addition to standard of care. NSE was measured at 48 and 72 hours as part of routine biochemistry and results were available for clinical use. NFL was measured from biobank samples at corresponding timepoints after end of trial using an NFL ELISA assay. The area under the receiver operator curves (AUROC) for prediction of all-cause mortality were determined for both NSE and NFL, and the difference in AUROC between the two was calculated. Results: Eighty patients were included in the IMICA trial. Overall mortality was 35% with no difference in frequency of mortality between the tocilizumab and placebo groups. At 48 and 72 hours 71 and 67 patients had available NSE and NFL values, respectively. There were no between group differences for tocilizumab versus placebo in NSE values at 48 or 72 hours (p=0.39 and p=0.22), nor for NFL values at 48 or 72 hours (p=0.75 and p=0.47), and groups were therefore pooled for further analysis. At 48 hours AUROC for NSE was 0.92 [0.85-0.99] (Figure 1), and for NFL 0.96 [0.91-1.00] (Figure 2), with a difference in AUROC of 0.04 [-0.03-0.10], p=0.315. At 72 hours AUROC for NSE was 0.91 [0.82-0.99] and for NFL 0.96 [0.92-1.00] (AUROC difference 0.05 [-0, 03-0, 14], p=0.198). Conclusions: In this single-center trial the prognostic performance of NFL at 48 hours was not significantly different from NSE with respect to predicting all-cause mortality. However, the AUROC of NFL was numerically greater than for NSE and a larger sample size may find NFL superior to NSE. We encourage future studies in evaluating NFL in a clinical setting.