Does Combining Biomarkers and Brain Images Provide Improved Prognostic Predictive Performance for Out-Of-Hospital Cardiac Arrest Survivors before Target Temperature Management?

Seung Ha Son,Jin Hong Min,Jung Soo Park,Jin Woong Lee,In Ho Lee,Hong Joon Ahn,Sung Uk Cho,Seung Whan Kim,Se Kwang Oh,Yong Chul Cho,Changshin Kang,In Sool Yoo,Yeonho You,Byung Kook Lee,Won Joon Jeong,Seung Ryu,Chun Song Youn,Dong Hun Lee

doi:10.3390/jcm9030744

Abstract

We examined whether combining biomarkers measurements and brain images early after the return of spontaneous circulation improves prognostic performance compared with the use of either biomarkers or brain images for patients with cardiac arrest following target temperature management (TTM). This retrospective observational study involved comatose out-of-hospital cardiac arrest survivors. We analyzed neuron-specific enolase levels in serum (NSE) or cerebrospinal fluid (CSF), grey-to-white matter ratio by brain computed tomography, presence of high signal intensity (HSI) in diffusion-weighted imaging (DWI), and voxel-based apparent diffusion coefficient (ADC). Of the 58 patients, 33 (56.9%) had poor neurologic outcomes. CSF NSE levels showed better prognostic performance (area under the curve (AUC) 0.873, 95% confidence interval (CI) 0.749–0.950) than serum NSE levels (AUC 0.792, 95% CI 0.644–0.888). HSI in DWI showed the best prognostic performance (AUC 0.833, 95% CI 0.711–0.919). Combining CSF NSE levels and HSI in DWI had better prognostic performance (AUC 0.925, 95% CI 0.813–0.981) than each individual method, followed by the combination of serum NSE levels and HSI on DWI and that of CSF NSE levels and the percentage of voxels of ADC (AUC 0.901, 95% CI 0.792–0.965; AUC 0.849, 95% CI 0.717–0.935, respectively). Combining CSF/serum NSE levels and HSI in DWI before TTM improved the prognostic performance compared to either each individual method or other combinations.

Highlights

Despite recent advances in emergency medicine and resuscitation management such as target temperature management (TTM), only approximately 30% of cardiac arrest (CA) survivors are discharged with a good cerebral performance status [1].Current guidelines recommend determining the neurologic prognosis at 72 h post-CA; the withdrawal of life-sustaining treatment (WLST) earlier than 72 h has been reported to be common and was shown to increase the mortality rate among patients who survived CA [2,3]
We previously reported that neuron-specific enolase in cerebrospinal fluid (CSF neuron-specific enolase levels in serum (NSE)) predicts the prognosis of patients with CA [10] and we reported that results of brain computed tomography (CT) and magnetic resonance imaging (MRI) prior to TTM can be used to predict prognosis in CA survivors [8]
This was a retrospective analysis of prospectively collected data including adult comatose out-of-hospital cardiac arrest (OHCA) survivors treated with TTM at Chungnam National University Hospital (CNUH) in Daejeon, Korea, between May 2018 and August 2019

Summary

Introduction

Current guidelines recommend determining the neurologic prognosis at 72 h post-CA; the withdrawal of life-sustaining treatment (WLST) earlier than 72 h has been reported to be common and was shown to increase the mortality rate among patients who survived CA [2,3]. One study suggested that 26% of patients with an early WLST may have survived and that 64% of these patients may have had a functionally favorable outcome [3]. To avoid sub-optimal WLST, an early and accurate prognosis is necessary. Several prognostic methods have been evaluated for predicting neurologic outcome [2,4]. Sedatives have been shown to confuse outcome predictions of CA survivors during neurologic examination [5]. A somatosensory evoked potential recording requires appropriate skills and experience, and it has been reported that artefacts significantly affect the results [4]; brain imaging results and biomarkers levels do not exhibit such effects

Methods

Results

Discussion

Conclusion