OUP user menu

So you implanted an implantable cardioverter defibrillator, now what to do with it?

Lara Dabiri Abkenari , Luc Jordaens
DOI: http://dx.doi.org/10.1093/europace/eus099 921-922 First published online: 31 May 2012

This editorial refers to ‘Long-term follow-up on high-rate cut-off programming for implantable cardioverter defibrillators in primary prevention patients with left ventricular systolic dysfunction’ by N. Clementy et al., on page 968

It is well established that patients with a low ejection fraction, be it ischaemic cardiomyopathy or not, benefit from implantable cardioverter defibrillator (ICD) therapy.1,2 Since the hallmark studies, this has been shown repeatedly in real-world studies and in a meta-analysis.3,4 However, this therapy comes with implant-related complications and the psychological strain of shocks, both appropriate and inappropriate.5 Recently, the relation between shocks and mortality has been investigated. The initial question of who to implant with an ICD is closely followed by many others. Single-chamber or dual-chamber pacing (and sensing)? How many zones to programme? Which cut-off rate (and duration to intervention) should be used? What is the benefit or disadvantage of anti-tachycardia pacing (ATP), and how should this be programmed? Finally, which detection intervals and parameters are optimal to treat all important arrhythmias, while avoiding unnecessary and inappropriate shocks. Most of these questions have never been studied systematically.

Ventricular pacing can have detrimental effects in patients with left ventricular dysfunction and should be avoided. It is clear now that pacing should be atrial, or at 40 b.p.m. in the ventricle.6

The effects of zone programming have never been studied as such. In MADIT (Multicenter Automatic Defibrillator Implantation Trial) II, programming was left to physician discretion, and both single- and double-chamber devices were used.1 In SCD-HeFT (Sudden Cardiac Death in Heart Failure Trial) only single-chamber devices were used with a cut-off rate of 187 b.p.m.2 MADIT II allowed ATP whereas SCD-HeFT only allowed shock therapy. Programming one or two zones affects both detection and intervention with shocks or ATP. For ATP, there have been observational, and later randomized or cohort studies. Small differences in zone programming have a serious impact on further events, as we have recently shown in a study comparing events in two centres with different programming practices. The centre using two zones, with ATP, had significantly more arrhythmias treated by an ICD, than the ‘brave’ centre which limited therapy to one zone with a rather high rate cut-off.7 Fortunately, this did not result in a higher mortality in patients with the zone using ATP, but the series was too small to detect such outcome.

Earlier studies of ATP have been performed in secondary prevention patients, when ventricular tachycardia is more likely to occur than in patients with a prophylactic indication.8 Later randomized trials were composed of mixed cohorts with both primary and secondary prevention indications, while PREPARE (Primary pREvention PARameters Evaluation) included only primary prevention, and used subsets from MIRACLE ICD (Multicenter InSync Randomized Clinical Evaluation) and EMPIRIC (Comparison of Empiric to Physician-Tailored Programming of Implantable Cardioverter-Defibrillators) for controls.9,10

In this issue of the Journal, Clementy et al.11 conclude that programming a very high cut-off (> 220 b.p.m.), a shock only zone, in a primary prevention population, consisting of ischaemic and non-ischaemic cardiomyopathy, can be safely performed. They report a lower rate of appropriate and inappropriate ICD discharges, in comparison with recent reports. In this prospective registry, four different ICD brands with nominal settings for the arrhythmia detection were used. During follow-up, 11.2% received appropriate shocks, and 6.6% received inappropriate shocks (these were related to supraventricular tachycardia and noise or oversensing). Only 12% of patients had ventricular tachycardia episodes in the monitoring zone (170–220 b.p.m.) and only 1.9% had symptoms, but no deaths occurred in this group. When a symptomatic event (syncope or pre-syncope) was observed in the monitoring zone, programming was modified. The key outcome of the study was adverse events related to specific programming including death, heart failure, hospitalization, and symptoms. No unexpected results were observed with regard to the mortality rate and the mode of death.

The results of this study are reassuring for those who do not programme ATP, and who tend to use an ICD as a ‘shock box’. The PainFREE Rx (Pacing Fast Ventricular Tachycardia Reduces Shock Therapies) II study demonstrated that shocks could be avoided by ATP, but 15% of patients had inappropriate therapy (with a mean of four shocks per patient in the primary prevention group).9 Anti-tachycardia pacing intervenes fast, and can effectively terminate an arrhythmia before the shock can be delivered, but it may also deliver therapy to an episode which could have otherwise been self-terminating. Time before therapy is crucial. In the PREPARE trial, which examined a primary prevention population (with historic controls), prolonged detection from 18/24 to 30/40 intervals (about 2–3 s) and the number of treated episodes was reduced by 50%.10 The shortest delay to intervention was still 9.8 s, with a cut-off rate of 182 b.p.m. The drawback of a long intervention time would be syncope; however, this was not observed in the present study. All these considerations did not address the likelihood of acceleration, leading to an appropriate, and sometimes an inappropriate shock.8 Inappropriate ATP is reported in up to 9% of patients.12 Prolonged ATPs can lead to a delay in shocks and longer periods of discomfort.12 Still, the results of studies as PainFREE and PROVE (Programming Ventricular Tachycardia Therapy in Patients with a Primary Prevention ICD Indication), which used very modest ATP (only two bursts), remain encouraging for ATP.8,9,13

The lessons are that we need a long delay before therapy, and that we should not be too aggressive, with the words of the authors of PainFREE in mind: that ‘the greatest predictor of ATP success was non-sustained ventricular tachycardia,’ as its presence doubles the number of ATP interventions.8

The findings reported by Clementy et al. suggest that using an ICD more as a ‘shock box’ could be a safe option to decrease inappropriate and unnecessary shocks which decrease the quality of life. Unnecessary ICD interventions could lead to earlier battery depletion and thus add extra morbidity due to more box changes. It may not be completely true that ‘shocks kill’, but shocks hurt.14 In MADIT II and SCD-HeFT, inappropriate shocks were not associated with increased mortality. A pooled analysis of four ATP trials supported the hypothesis that ATP reduced shocks and lowered mortality.15 In this study, ventricular arrhythmia burden was higher in patients who received shocks than in patients who were treated with ATP.

In the present study, mortality was 17% over 40 months, which was difficult to compare with mortality rates reported in other trials because of a shorter follow-up in these studies, but it is in line with recently published survival rates for heart failure patients with an ICD.14 In order to gain the maximum benefit and to use ICDs appropriately, studies like this, and furthermore, studies which are randomized and controlled are necessary.

With all these considerations it seems that once an ICD is implanted, the decision-making really begins. To summarize, with this study, we can believe that for primary prevention ICDs, a single zone with a high cut-off rate can be programmed. This is safe and is associated with fewer inappropriate interventions. Anti-tachycardia pacing can potentially be withheld until a patient has proven to benefit from this therapy, and this constitutes only a small minority. We have now the opportunity to implant a subcutaneous defibrillator, which cannot pace or deliver ATP.16 In our initial clinical experience this did not seem to be a disadvantage.17 Who would benefit most from such an implant is not clear. The findings in this study could be helpful in better defining who would profit from such a simple device, which is (until now) not associated with the lead problems that make ICD therapy in general less attractive.

Conflict of interest: none declared.


  • The opinions expressed in this article are not necessarily those of the Editors of Europace or of the European Society of Cardiology.