Combining brain computed tomography and serum neuron specific enolase improves the prognostic performance compared to either alone in comatose cardiac arrest survivors treated with therapeutic hypothermia

Aim of the study

We determined whether combining the grey-to-white matter ratio (GWR) on brain computed tomography (CT) and serum neuron specific enolase (NSE) improves the prognostic performance when compared to either alone in cardiac arrest patients treated with therapeutic hypothermia (TH).

Methods

We performed a retrospective study of a cohort of cardiac arrest patients treated with TH. The Hounsfield unit was measured in the caudate nucleus (CN), putamen (P), posterior limb of internal capsule (PIC) and corpus callosum (CC); GWR was calculated as CN/PIC and P/CC. The NSE value was obtained at 0, 24, and 48 h after restoration of spontaneous circulation (ROSC). We analysed the prognostic performance of GWR and NSE, singly and in combination, in predicting poor neurologic outcome (cerebral performance category 3–5).

Results

Of the 224 included patients, 82 showed good neurologic outcome at hospital discharge, while 142 showed poor neurologic outcome. The P/CC (area under receiver operating characteristics (AUROC) 0.864, sensitivity/specificity 52.9%/100%) showed better prognostic performance than did the CN/PIC (AUROC 0.721, sensitivity/specificity 19.8%/100%). The NSE value at 48 h after ROSC (AUROC 0.895, sensitivity/specificity 60.2%/100%) showed the highest prognostic value among the three NSE time points. Analysis of 119 patients undergoing both brain CT and NSE at 48 h indicated that combining P/CC and NSE improved the sensitivity (78.6%) compared to either alone (48.6%, 62.9%).

Conclusion

Combining brain CT and serum NSE improves the prognostic performance when compared to either alone in predicting poor neurologic outcome in cardiac arrest patients treated with TH.