The Cardiac Arrest Study Hamburg

Abstract & Commentary

Synopsis: Implantable cardioverter defibrillator therapy results in a modest reduction in all-cause mortality and a larger reduction in sudden death mortality compared to the mortality rate seen in drug-treated groups. The reductions are most apparent in the first five years of therapy and become less prominent over time.

Source: Kuck KH, et al, for the CASH investigators. Circulation 2000;102:748-754.

The cardiac arrest study hamburg (cash) initially randomized survivors of cardiac arrest in four treatment groups: amiodarone, metoprolol, propafenone, and implantable cardioverter defibrillators (ICD). The primary end point for the trial was all-cause mortality with sudden cardiac death and recurrent cardiac arrest being secondary end points.

A total of 349 patients were entered into the trial between 1987 and 1996. Assignment to propafenone was discontinued in 1992 after a preliminary analysis showed a higher mortality in the treatment group. The final study included 99 ICD patients, 92 amiodarone patients, and 97 metoprolol patients. Their mean age was 58 ± 11 years, 80% were male, 73% had coronary artery disease, and 10% had no structural heart disease. The index arrhythmia was ventricular fibrillation (VF) in 84% and ventricular tachycardia (VT) in 16%. The average left ventricular ejection fraction was 46%. Both epicardial (55 patients) and endocardial (44 patients) ICD systems were used. Daily maintenance doses of amiodarone and metoprolol were 225 ± 75 and 85 ± 73 mg, respectively. During a mean follow-up of 57 ± 34 months, the crude mortality rates were 36% in the ICD group and 44% in the combined amiodarone/metoprolol group. This difference was not statistically significant (P = 0.081; hazard ratio 0.77). Although the death rate was decreased in the ICD group by more than 25% in each of the first four years of follow-up, less differences were seen at later time points. The secondary analysis of sudden death survival did show a significant advantage in the ICD group (13% vs 33%; P = 0.005). Interestingly, there was no difference between the amiodarone and metoprolol subgroups in either total or sudden death mortality.

There were five deaths associated with ICD implantation and 23% of the ICD patients experienced one or more nonfatal complications. No serious toxicity was noted in either the amiodarone or the metoprolol groups.

Kuck and associates conclude that ICD therapy results in a modest reduction in all-cause mortality and a larger reduction in sudden death mortality compared to the mortality rate seen in drug-treated groups. The reductions are most apparent in the first five years of therapy and become less prominent over time.

Comment by John P. DiMarco, MD, PhD

CASH is the third of the large, randomized, clinical trials that compared ICD use to drug therapy in patients resuscitated from sustained ventricular arrhythmias. CASH differs from the two previously reported North American trials (AVID and CIDS) in that only cardiac arrest survivors could be enrolled and patients with hypotensive VT were excluded. This resulted in a study group that was younger and included more patients without structural heart disease than the group in the other trials. Unlike those studies, CASH did not show a statistically significant improvement in survival in the ICD group. However, this study was underpowered to detect a clinically meaningful difference, so its results must be considered in the context of the other trials.

Despite the fact that this study cannot stand solely on its own, CASH does provide some interesting insights. First, it is clear that a group of out-of-hospital cardiac arrest survivors is different from a group comprising patients with both cardiac arrest and sustained VT. The latter group will have more recurrent arrhythmias, a higher total mortality rate and lower ejection fractions. When planning the sample size for any study, these factors must be considered. Second, it is interesting that amiodarone and metoprolol were not different. Amiodarone is certainly more useful in sustained VT patients but the effects of beta-adrenergic blockade may be of more importance in preventing death and VF. Finally, it must be noted that the ICD group had a relatively high rate of both fatal and nonfatal device complications. Presumably, most of these occurred with the epicardial ICD systems that were used early in the trial. Without these early deaths, it seems likely that the advantages of ICD therapy would have reached statistical significance.