Diagnostic and prognostic utility of cardiac troponin in post-cardiac arrest care

Background: Cardiac troponin is routinely tested in the post-cardiac arrest setting, but its utility in identifying ischaemic aetiology and predicting left ventricular systolic dysfunction (LVSD) and survival is not known. Methods: In a retrospective single center registry, we identified 145 consecutive patients who had achieved return of spontaneous circulation after cardiac arrest and had undergone serial cardiac troponin T (cTnT) testing, echocardiogram, and expert adjudication of aetiology. Initial and peak cTnT were evaluated for assessing ischaemic aetiology, LVSD, and survival to discharge using area under the receiver operating characteristic curve (AUROC). Results: Mean age was 61 + 14 years and 71% were men. Of the 145 arrests, 19% had an ischaemic aetiology, 68% had LVSD post-arrest, and 55% survived to discharge. All patients had a positive initial cTnT at 0.01 ng/mL (clinical cut-off). Even at higher cut-offs of 10x, 100x and 1000x, initial cTnT performed poorly (AUROC 0.57, 0.56, and 0.56) and peak cTnT performed modestly (AUROC 0.55, 0.61, and 0.62) as diagnostic tests for ischaemic aetiology. Similarly, even at higher cut-offs, initial (AUROC 0.60, 0.62, 0.55) and peak (AUROC 0.57, 0.61, and 0.62) cTnT performed poorly to modestly at predicting LVSD. The test performed poorly for predicting survival to discharge (AUROC for all <0.6). Conclusions: At both current and several-fold higher thresholds, cTnT does not perform sufficiently well to guide clinical decision-making or predict patient outcomes. Routine post-cardiac arrest testing of cTnT should be reevaluated.