Aim of the study: Extracorporeal cardiopulmonary resuscitation (ECPR) is an evolving technique to improve cardiopulmonary resuscitation (CPR) outcomes. Identifying a readily available tool helpful for predicting patient's outcome is warranted. The aim of the study was to evaluate the capability of cranial near-infrared spectroscopy (cNIRS) to identify non-survivors or patients with unfavorable neurologic outcome prior to cannulation for ECPR to avoid futile cannulations. Methods: Retrospective analysis (2015-2021) of 97 patients requiring ECPR due to cardiac arrest with prior cNIRS measurement, which was performed immediately after ECPR team arrived on scene. Lowest possible regional cerebral oxygen saturation (rSO 2 ) is 15%. Results: Mortality was 72.1% (70/97). Survivors showed in 88.9% (24/27) good neurological outcome (Cerebral Performance Category (CPC) 1 + 2). rSO2 = 15% (11/97) prior to cannulation was only found in non-survivors. Among survivors, initial rSO 2 was not associated with neurological outcome. Non-shockable initial rhythm was associated with higher mortality (44/50). In survivors, time to ECPR was shorter (p = 0.006), and initial lactate was significantly lower, whereas initial pH and hemoglobin levels were higher (p = 0.001). Survivors and those with favorable neurological outcome showed lower maximal NSE levels in the first 72 hours (p < 0.001; p = 0.041).
Conclusion:In our patient cohort, rSO 2 = 15% immediately prior to cannulation for ECPR did not result in any survivors, thus might be a marker for futile cannulation in ECPR. Higher rSO 2 values were not associated with favorable neurologic outcome. Lower initial lactate and lower maximal NSE within the first 72 h after arrest were associated with favorable outcome.