Relevance of guideline-based ICD indications to clinical practice

Abstract

The implantable cardioverter-defibrillator (ICD) has established itself as life-saving therapy in patients at risk for sudden cardiac death. Remarkable technological advances have made ICDs easier and safer to implant, with improved therapeutic and diagnostic functions and reduced morbidity.

Guidelines on ICD indications have been proposed by American and European scientific societies since a number of years, based upon trials and expert opinion. In the context of variable economic and political constraints, it is questionable whether these guidelines may be applied to all settings. This review discusses the guideline-based indications, critically examines their applicability to clinical practice, and discusses alternatives to ICD therapy.

1. Introduction

Implantable cardioverter defibrillators (ICDs) were approved for general market release in the United States in 1985, but only for patients who had survived prior cardiac death (SCD). Over the last two decades, the indications have evolved from secondary prevention of sudden death to also include primary prevention in selected patients. The most recent international guidelines that include indications for ICDs are the 2006 ACC/AHA/ESC guidelines on ventricular arrhythmias and sudden death,¹ the 2012 update of the 2008 ACC/AHA/HRS device guidelines,² and the 2012 ESC guidelines on heart failure.³ The main ICD indications are summarized in Table 1 and their rationale are discussed below.

Table 1.

Summary of the main ICD indications of the latest international guidelines.

Patient characteristics	ACC/AHA/ESC ventricular arrhythmia and sudden death 20061	ACCF/AHA/HRS device-based therapy 20122	ESC heart failure 20123
Cardiac arrest survivors, unstable sustained VT	Class I, LOE A	Class I, LOE A	Class I, LOE A
Structural heart disease, spontaneous sustained VT (stable or unstable)	–	Class I, LOE B	–
Post MI (≥40 days)
NYHA II, III LVEF ≤35%:	Class I, LOE A (LVEF ≤30–40%)	Class I, LOE A	Class I, LOE A
NYHA I LVEF ≤30%:	Class IIa, LOE B (LVEF ≤30–35%)	Class I, LOE A	–
LVEF ≤40%, NSVT, EPS+	Class I, LOE A	Class I, LOE B
NICM LVEF ≤35%
NYHA II, III	Class I, LOE B (LVEF ≤30–35%)	Class I, LOE B	Class I, LOE B
NYHA I	Class IIb, LOE C (LVEF ≤30–35%)	Class IIb, LOE C	–

Shaded gray boxes: secondary prevention; white boxes: primary prevention.

Abbreviations: EPS+: induction of sustained ventricular tachycardia or ventricular fibrillation at electrophysiological study; LOE: level of evidence; MI: myocardial infarction; NICM: non-ischemic cardiomyopathy; VT: ventricular tachycardia; VF: ventricular fibrillation.

2. Secondary prevention

Secondary prevention refers to prevention of SCD in patients who have survived a prior sudden cardiac arrest or sustained ventricular tachycardia (VT).² Early data on ICDs were drawn primarily from uncontrolled series of refractory patients who had failed antiarrhythmic drugs. In the 1990's, three randomized controlled trials of secondary prevention of sudden death were conducted: AVID,⁴ CASH⁵ and CIDS.⁶ All three trials enrolled patients who had survived SCD without a reversible cause. The trials randomized an ICD to antiarrhythmic therapy (including amiodarone, sotalol or propafenone). A meta-analysis of these trials⁷ including 1866 patients showed that ICD therapy significantly reduced the risk of total mortality by 28% (p = 0.0006) and that of sudden death by 50% (p < 0.0001). Based upon the positive results of these trials, the guidelines state that:

• ICD implantation is a class I level of evidence A indication in survivors of sudden death due to ventricular fibrillation (VF) or hemodynamically unstable sustained ventricular tachycardia (VT), after a completely reversible cause has been excluded.

Secondary prevention of sudden death (without a reversible cause for arrhythmia) is seldom disputed as being an indication for an ICD. However, subgroup analyses (with all their caveats) indicate that not all patients benefit from this therapy. In the meta-analysis of AVID/CIDS/CASH, patients aged >75 years did not benefit from the ICD.⁸ In AVID, the largest of all the secondary prevention trials, patients with a left ventricular ejection fraction (LVEF) of >34% (comprising approximately 40% of the study population), also did not derive any survival benefit from ICD therapy.⁹ In the setting of limited resources, these subgroup analyses call for careful evaluation in order to prescribe therapy to those patients who are most likely to benefit, even in the setting of secondary prevention.

The AVID registry¹⁰ showed that patients who presented with sustained VT without serious symptoms or hemodynamic compromise (and who were not implanted with an ICD), also had a high mortality. The presenting arrhythmias may be a marker for a substrate capable of producing a more malignant arrhythmia. Therefore, current guidelines state that:

• ICD therapy is a class I indication in patients with structural heart disease and spontaneous sustained VT, whether hemodynamically stable or unstable.

ICD implantation is however not indicated in certain VTs in patients with structurally normal hearts, such as infundibular or fascicular VTs who generally have a good prognosis and in whom catheter ablation may be curative.²

In patients with structural heart disease, an electrophysiological study (EPS)-guided approach with antiarrhythmic therapy has been evaluated. A study¹¹ included 84 patients with ischemic heart disease who presented with sustained VT or VF and in whom arrhythmias were inducible at EPS; half of the patients were no longer inducible after amiodarone loading. The non-inducible patients were then maintained on amiodarone, and an ICD was implanted in the others. After a mean follow-up of 5 years, total mortality and sudden cardiac death were significantly higher in the patients only on amiodarone therapy, indicating that this strategy does not offer optimal protection. Amiodarone may nevertheless be considered (class IIb, level of evidence C) in patients with LV dysfunction due to prior MI with an ICD indication, in patients who cannot (e.g. for financial reasons) or refuse to have an ICD implanted.¹

Another alternative to ICD implantation that may be considered is catheter ablation of ventricular arrhythmias, a domain in which increasing experience is being gained. In a 2009 EHRA/HRS expert consensus statement,¹² catheter ablation was proposed as an adjunct to ICD therapy but not as an alternative to it. It is stated that “results of catheter ablation of VT should have little influence on the indications for ICD implantation”, as even if VTs are rendered non-inducible by ablation, the recurrence rate may remain substantial. This was observed in the multicentre VTACH study,^13,14 in which 107 ICD patients with ischemic heart disease, reduced (≤50%) LVEF and a history of stable VT were randomized 1:1 to catheter ablation versus conventional therapy. Even though VT/VF recurrence was significantly reduced in patients randomized to ablation, ventricular arrhythmias still recurred in half of them.

A stepwise approach in patients with ischemic heart disease presenting with sustained VT has been published.¹⁵ In this report, 45 patients with a mean LVEF of 40% underwent radiofrequency ablation of VT, the endpoint of which was non-inducibility, and which was acutely successful in 89% of patients. The EPS was repeated at 2–3 months (without anti-arrhythmic drugs) and in case of re-inducibility, and new ablation procedure was performed. Patients in whom monomorphic VT of <270 bpm remained inducible despite radiofrequency ablation were implanted with an ICD. Thus only 19/45 (42%) of patients had ICD implantation. After a median follow-up of 45 months, there were no significant differences in mortality between the two groups (p = 0.11), with a trend in better survival in patients without an ICD. Only 1 patient without an ICD died suddenly (and may therefore have benefited from ICD implantation). These very promising results however need to be interpreted with caution, namely that the procedures were performed in an expert center, using state-of-the art techniques and technology (such as electro-anatomical mapping). The results therefore may not be extrapolated to centers with less experience/infrastructure, but nevertheless offer a future perspective to a cheaper alternative to ICDs.

3. Primary prevention of sudden death

This indication is more a subject of debate than secondary prevention, especially as trials have gradually included patients with less stringent selection criteria. A summary of the ICD indications stated in different guidelines is shown in Table 1.

3.1. Patients with ischemic heart disease

Three studies have randomized patients with ischemic heart disease and reduced LVEF to conventional therapy versus an ICD for primary prevention of sudden death: MADIT,¹⁶ MUSTT,¹⁷ and MADIT II.¹⁸ The first two studies included patients with an LVEF of ≤35% and ≤40% respectively, with documented non-sustained VT, and inducible VT by EPS that was not suppressible by anti-arrhythmic drugs. MADIT II was the largest study (1232 patients) and did not require documentation of arrhythmias or inducible VT for inclusion, but the LVEF cutoff was lower at ≤30%. All three studies showed a reduction in mortality in ICD patients. Based upon the results of MADIT and MUSTT, the guidelines state that:

• Patients with ischemic heart disease and LVEF ≤40%, non-sustained VT and inducible VT or VF by an EPS have a class I indication for an ICD.

Based upon MADIT II, the guidelines state that:

• Patients with an LVEF ≤30% and ischemic heart disease have a class I indication for an ICD (without documenting any ventricular arrhythmias, and independently of heart failure status; cutoff of LVEF is ≤35% in patients with NYHA II-III heart failure).

The guidelines also state that life expectancy should be >1 year, and that the interval between myocardial infarction and ICD implantation should be >40 days, and that in case of heart failure, that optimal medical therapy should have been followed for ≥3 months. This was due to two studies^19,20 that showed not survival benefit if ICDs were implanted early after myocardial infarction. This was attributed to the possibility that ICDs, by shocking VF in these patients, merely transform sudden death to eventual death from pump failure, without significantly prolonging life.

A subgroup analysis of the MADIT study showed that “the sickest patients benefit the most”, as efficacy of the ICD was limited to those with an LVEF <26%.²¹ In MADIT-II, subgroup analysis also showed that only patients with an LVEF <25% had a statistically significant improvement in survival conferred by the ICD, although there was a trend in improvement in patients with an LVEF >25%. Inducibility of VT by an EPS is sometimes used for risk-stratification in order to aim for a target population who would be most likely to benefit from an ICD. In the MADIT II study however, patients without VT inducibility still had a 25% incidence of ventricular arrhythmias at follow-up that was not statistically different from those patients who were inducible.²² This implies that an EPS has limited value in a MADIT-II type patient population.

It is unknown whether ablation of inducible VT in primary prophylaxis patients can avoid ICD implantation. In a small series²³ of 38 primary prevention ICD patients with structural heart disease, 18 of whom had undergone catheter ablation of inducible VT, only 5% of these patients received appropriate shock compared to 45% without catheter ablation (p = 0.018) after a mean follow-up of 50 months. These preliminary data however do not allow us to conclude that VT ablation can be used as first line therapy to avoid ICD implantation.

3.2. Patients with non-ischemic cardiomyopathy

Data supporting prophylactic ICD implantation in patients with non-ischemic cardiomyopathy are less strong as compared to ischemic heart disease. Three studies (CAT,²⁴ AMIOVIRT²⁵ and DEFINITE²⁵) evaluated prophylactic ICD implantation in a total of approximately 660 patients with LVEF ≤30–35% with or without documented non-sustained VT. All showed a non-significant trend in reduced mortality in the ICD arms. SCD-HeFT²⁶ was a large trial that evaluated prophylactic ICD implantation in patients with NYHA II-III heart failure and LVEF ≤35% and that randomized 2521 patients to conventional therapy vs. amiodarone vs. an ICD. The trial showed a survival benefit with ICD therapy, and approximately half of these patients had non-ischemic cardiomyopathy without any significant difference in outcome compared to those with ischemic heart disease. These two subgroups analyzed separately only showed a trend in improved mortality, as was also the case in patients with an LVEF of 30–35%. Current guidelines state that:

• Patients with non-ischemic dilated cardiomyopathy who have an LVEF <35% and who are in NYHA functional Class II or III have a class I (level of evidence B) indication for prophylactic ICD implantation. Patients who are in NYHA functional class I have a class IIb indication (level of evidence C).

The other question that arises in patients with heart failure is whether cardiac resynchronization therapy (CRT) is indicated. The European guidelines have recently been updated,²⁷ but are outside the scope of this review.

As can be appreciated from Figs. 1 and 2, that efficacy of ICDs in primary prevention trials is comparable, if not higher, than for that of secondary prevention.

Fig. 1.

Comparison of results of primary (left) and secondary (right) prevention ICD trials. ARR = absolute risk reduction.

Fig. 2.

Number of patients needed to treat (NNT) with an ICD to save 1 life in the primary (left) and secondary prevention trials (right).

4. Other considerations

4.1. Individual patient decision-making

The guidelines discussed above should serve as an outline to help in decision-making. Indications for an ICD should be considered individually in each patient taking into account factors such as patient age, comorbidities (which will mitigate effectiveness of ICD therapy²⁸), psychological profile, quality of life, access to healthcare (mandatory for ICD follow-up) etc.

4.2. Device follow-up

A crucial but sometimes neglected aspect of ICD therapy is optimal programming and follow-up. Programming of arrhythmia detection and therapy settings has been shown to significantly affect mortality,²⁹ as well as morbidity in terms of inappropriate shocks.^29–31 Troubleshooting of ICD follow-up (e.g. interpretation of stored episodes) may be complex and requires proper training of qualified personnel. National societies should define standards to ensure that patients receive optimal care. For example, the Working Group of Pacing and Electrophysiology of the Swiss Society of Cardiology stipulates that cardiologists who perform pacemaker or ICD follow-up should have passed the EHRA accreditation exam in cardiac pacing.³² Educational programs are available to assist physicians in obtaining certification (www.escardio.org/communities/EHRA/education/).

4.3. Economic aspects

ICD therapy is expensive, although its cost-effectiveness is acceptable according to Western standards.³³ Nevertheless, healthcare cost constraints are being faced worldwide, and rational clinical decisions need to be made. Device manufacturers are facing increasing pressure on prices, although a wider prescription of ICD therapy with greater volumes may help in reducing unit cost. Simpler models (“shock boxes”) may also be a cheaper solution than premium models, especially in patients with a primary prevention indication. Adequate programming of these simpler devices needs nevertheless to be performed to avoid inappropriate/unnecessary shocks. Re-use of explanted and resterilised ICDs with at least 3 years projected battery longevity has been shown to be safe,³⁴ and may be an alternative if a new device is simply not an option. Finally, political instances need to be made aware of the importance of providing adequate funding for this life-saving therapy.

5. Conclusions

Randomized controlled trials have established ICD therapy for secondary prevention of sudden death as well as for primary prevention in selected patients with structural heart disease. Anti-arrhythmic therapy does not provide optimal protection against ventricular arrhythmias in this setting, even when arrhythmias are no longer inducible by EPS. Catheter ablation of VT is a promising adjunct to ICD therapy in order to reduce ICD shocks but does not replace device therapy. Despite these facts, economic constraints in some countries oblige physicians to prescribe anti-arrhythmic therapy or catheter ablation if ICD therapy cannot be afforded.

Conflicts of interest

All authors have none to declare.

Financial support

H.B. is funded in part by a research grant from the La Tour Fund for Cardiovascular Research.

Disclosures

N.A-J. has no conflicts of interest to report. H.B. has received speaker fees and fellowship/research support from Biotronik, Boston Scientific, Medtronic, Sorin and St-Jude Medical.