Response to programmed ventricular stimulation and clinical outcome in cardiac arrest survivors receiving randomised assignment to implantable cardioverter defibrillator or antiarrhythmic drug therapy.

BACKGROUND

Using a prospective design which randomly assigned implantable cardioverter defibrillator (ICD) versus antiarrhythmic drugs (AADs), we investigated the usefulness of programmed ventricular stimulation (PVS) for prediction of outcome and therapy effectiveness in cardiac arrest (CA) survivors.

METHODS AND RESULTS

We performed baseline PVS in 285 survivors of CA enrolled in the Cardiac Arrest Study Hamburg (CASH) and randomised to ICDs or AADs. Sustained ventricular arrhythmia (VA) was induced in 134 (47.0%) patients. We compared the outcomes of different subgroups based on response to baseline PVS and randomly assigned therapy. Patients were followed for a median of 55 months. The raw death rate was greater among inducible (51.3% [95% CI: 44.9-58.3%]) than non-inducible patients (28.8% [CI: 23.4-36.1%, p = 0.0003]). When challenged in a multivariate model, inducibility still had an independent power for predicting all-cause death (hazard ratio (HR), 1.5 [95% CI, 1.1-2.3], p = 0.041), but not sudden death (SD) (HR, 1.2 [95% CI, 0.7-3.6], p = 0.162). Subgroup analysis showed that, when compared to AADs, assignment to ICDs was associated with a lower risk of all-cause death (HR, 0.4 [95% CI, 0.1-0.9], p = 0.044) in patients with EF < or =0.35 and non-inducible arrhythmias, but not in other patient subgroups.

CONCLUSIONS

In CA survivors, inducibility at baseline PVS is independently associated with an increased risk of all-cause death, but not SD. In addition, response to PVS may help to identify subgroups of patients who could most benefit from ICD.