Primary prevention implantable cardioverter–defibrillator treatment: how to identify patients most likely to benefit?

Keywords::

• ICD
• implantable cardioverter–defibrillator
• primary prevention
• sudden cardiac death
• ventricular arrhythmia

Sudden cardiac death (SCD) is defined as death from an unexpected circulatory arrest, mostly due to a cardiac arrhythmia in patients with coronary artery disease, occurring within an hour of the onset of symptoms [1]. Approximately 50% of all deaths in patients with ischemic heart disease are unexpected, occurring shortly after onset of symptoms. Concomitantly with the increasing number of patients with ischemic heart disease – especially in the Western world – annual worldwide mortality rates due to SCD have risen to an estimated 7 million patients [2–4].

An effective treatment to prevent arrhythmic death is implantation of an implantable cardioverter–defibrillator (ICD). Large randomized trials have proven the beneficial effect of ICD therapy in patients at high risk for SCD. The first trials demonstrated benefit in patients who survived a life-threatening ventricular arrhythmia (secondary prevention [5–7]). However, since the chance of surviving such an event is low (estimated at 6%), focus shifted to the identification of patients at high risk, prior to a first life-threatening event (primary prevention [8]). A clear example here of is the MADIT II trial, which demonstrated that postinfarct patients with a left ventricular ejection fraction (LVEF) of less than 30% had a significantly better survival if they underwent ICD implantation compared with conventional medical therapy only [9]. In addition, the SCD-HeFT trial demonstrated that both ischemic and non-ischemic symptomatic heart failure patients with an LVEF ≤35% had an improved survival if treated with a defibrillator [10].

Following the inclusion of primary prevention ICD treatment in the international guidelines, the implanted population changed from survivors of ventricular arrhythmias to patients with a low LVEF [11,12]. As a consequence, the total number of worldwide ICD implantations significantly increased to 275,000 in 2008 [13].

Clinical trials: identification of primary prevention ICD patients most likely to benefit

Although the large randomized trials clearly demonstrated the positive effect of ICD treatment on total mortality, these positive results are not observed in all patients currently indicated. A clear example is the analysis of mortality data from the MADIT II trial, including patients with ischemic heart disease and a reduced LVEF without prior ventricular arrhythmias [14]. Benefit from ICD treatment was not observed in patients (5%) with severe renal failure (defined as blood urea nitrogen ≥50 mg/dl or serum creatinine ≥2.5 mg/dl). After exclusion of these patients, 17 prespecified potential risk factors were assessed for their predictive value for all-cause mortality in the non-ICD arm of the trial. This resulted in the following five risk factors: age >70 years, New York Heart Association (NYHA) functional class >II, blood urea nitrogen >26 mg/dl, atrial fibrillation and QRS duration >120 ms. All patients were grouped by the number of risk factors (0, 1, 2 or ≥3), and the effect of ICD treatment was assessed per group. The results showed a beneficial effect in patients with one or two risk factors (52% of patients) but no effect in patients without risk factors or with three or more risk factors (43% of patients). This implies that the patient who benefits most from ICD treatment should be at high enough risk for SCD (e.g., at least one risk factor) but not at too high a risk for mortality from other causes (e.g., three or more risk factors). Additionally, the results show that in 48% of MADIT II patients, no benefit could be observed. An analysis of the SCD-HeFT trial showed similar results in a population with poor LVEF and symptomatic heart failure (NYHA functional class II or III) due to ischemic or nonischemic heart disease [15]. A total of 2483 patients were included and stratified into five different risk groups according to their predicted mortality calculated from their baseline variables. In the highest-risk patients (20% of patients included), a 4-year mortality of 50% was observed and ICD implantation had no beneficial effect on survival. These outcomes again suggest that patients with a high risk for mortality do not benefit from primary prevention ICD treatment.

Routine clinical practice: identification of primary prevention ICD patients most likely to benefit

More recently, efforts have been made to construct a risk model on an ICD-treated population in a real-world population outside the setting of a clinical trial. Initially, risk models were constructed to identify patients at high risk for all-cause mortality, implying that these patients may have less ICD benefit. The analysis was performed with data from a 1036 (68% ischemic) primary prevention patients registry with a mean follow-up of 873 ± 677 days. The risk score consisted of simple baseline variables such as age, LVEF and renal clearance that could stratify ischemic and nonischemic patients in low (6 years mortality of <5%), intermediate and high risk (6 years mortality of >45%) for mortality and therewith create an individual patient-tailored estimation on mortality risk that could aid clinicians in daily practice [16]. Although, according to the large studies mentioned earlier, high mortality risk can be expected to point out patients at low benefit, one could take it to a higher analytical level by trying to identify death prior to ICD discharge as the ideal endpoint for nonbenefit. At our center, we developed a tool to identify patients with ischemic heart disease who, although currently indicated for ICD treatment, will have a high risk of dying prior to actually receiving a potentially life-saving ICD shock [17]. In this study, a total of 900 primary prevention ICD patients with ischemic heart disease were followed up for 669 days (IQR 363–1322 days). During follow-up, 150 (17%) patients died, of whom 114 (76%) patients did not receive appropriate device therapy and therefore had no clear benefit from ICD treatment. Accordingly, five independent predictors of death without appropriate ICD therapy were selected – NYHA ≥III, age ≥75, diabetes mellitus, LVEF ≤25% and a history of smoking – and a risk score model was included for death without appropriate ICD therapy (nonbenefit). The score was named the FADES score (acronym for functional class, age, diabetes, ejection fraction, smoking) and the individual patient risk score cutoffs were determined for a population at low, intermediate and high risk of death without prior appropriate ICD therapy. Five-year cumulative incidence for death without prior appropriate therapy (nonbenefit) was 10% in low-risk patients, 17% in intermediate-risk patients and 41% in high-risk patients. These results demonstrate that especially in the high-risk patient group, which comprises 23% of the total primary prevention ICD population, a significant number of patients had no benefit from ICD therapy. However, it is important to realize that patients classified as high risk for death without prior appropriate ICD therapy do not per se receive no appropriate ICD therapy at all. Paradoxically, factors as advanced age, depressed ejection fraction and smoking are also identified as predictors of SCD or appropriate ICD therapy [14,15,18]. Nevertheless, following potentially life-saving ICD therapy, life expectancy in those high-risk patients remains short: within 5 years following ICD therapy, 61% of the high-risk patients died. Consequently, the FADES score may offer additional inputs to improve patient care.

Potential of ICD treatment

A meta-analysis of all randomized clinical trials on primary prevention ICD treatment reported a number needed to treat of 13 (e.g., 13 ICDs to prevent one death [19]). Taking into consideration that, within the currently indicated population, it may be relatively easy to identify patients who do not benefit from ICD therapy, one can imagine the high potential of ICDs, if better allocated than currently according to the guidelines-directed clinical practice. Consequently, this will improve the cost–effectiveness of ICD therapy even more and optimize the utilization of limited financial resources and trained personnel, which is so evidently needed [20].

Conclusions

ICD therapy in primary prevention patients is effective and may save many patients from dying suddenly. However, the challenge will be to develop and implement criteria allowing better identification of high-risk patients and to limit the number of implants in patients who will not benefit. This is important not only from a cost perspective but also because ICD therapy is not harmless. In other words, inappropriate shocks, infections and device or lead malfunction are serious issues.

Financial & competing interests disclosure

The Department of Cardiology, Leiden University Medical Centre, Leiden, The Netherlands received unrestricted grants from: Biotronik (Berlin, Germany); Boston Scientific (Natick, MA, USA); GE Healthcare (Buckinghamshire, UK); Medtronic (MN, USA); and St Jude Medical (MN, USA). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.