Keywords

Case Presentation

A 68-year-old man with a single-chamber implantable cardioverter-defibrillator (ICD) presented because his ICD was nearing the end of battery life. The ICD was implanted 6 years ago for primary prevention of sudden cardiac death (SCD) . He has not had any appropriate ICD shocks. His left ventricular ejection fraction (EF) , which was 30% at the time of ICD implantation, has improved to 45% since then. Is the ICD generator replacement justified?

ICD Generator Replacement Statistics

Approximately 30,000 ICD generator replacement procedures are performed in the United States annually for end of battery life, constituting 28% of all ICD procedures [1,2,3]. The most common reason for ICD generator replacement is the device reaching elective replacement indicator (ERI) , an alert displayed by the ICD indicating that the battery may reach end of life in the next 3–6 months. It is recommended to replace the ICD generator within 3 months of reaching ERI. Other, less common reasons for replacing ICD generator are infection, upgrade to cardiac resynchronization therapy (CRT) , lead or generator malfunction, and advisory recalls for increased risk of failure of ICD components [2,3,4].

Approximately, 65–70% of primary prevention ICD recipients remain free of appropriate ICD therapy during the lifetime of their initial ICD generator [2, 5, 6]. While it is common practice to routinely replace ICDs that reach ERI, a number of factors may limit the potential benefit of ICD after generator replacement. Patients presenting for ICD generator replacement tend to be older and have more comorbidities than those having initial ICD implant, increasing their competing risk of death from non-cardiovascular causes [1, 5, 7]. In a propensity-matched analysis of the National Cardiovascular Data Registry, survival after ICD replacement was worse compared to initial implant, regardless of device type [2].

Furthermore, ICD generator replacement procedure may be associated with significant complications such as infection, hematoma, or lead damage, which may result in increased morbidity and mortality [2, 8,9,10]. Indeed, patients presenting for an ICD generator replacement due to ERI have a periprocedural major complication rate of 4–6% [1, 5, 8, 9, 11]. Those who have a concomitant lead replacement have a 6-month complication rate of up to 15% [9]. The highest risk is associated with the need to replace a left ventricular lead with complication rates ranging from 9% to 50% [9].

The risk of mortality after ICD generator replacement is close to 10% at 1 year and up to 50% at 5 years [5, 7, 12,13,14]. Factors associated with higher mortality include increased age, atrial fibrillation, heart failure, worsened ejection fraction, chronic lung disease, diabetes, renal dysfunction, and history of stroke [1, 15]. Excessive long-term mortality in these cases is a testament to the higher-risk status of these patients rather than the risks of the ICD generator replacement procedure.

Implantable cardioverter-defibrillator generator replacement procedures also have a significant economic burden to the US healthcare system [16, 17]. The approximate cost of a single-chamber ICD replacement was around $18,000 in 2005 but increased to nearly $23,000 by 2013 [7, 18]. Thus, roughly $700 million is spent for ICD generator replacement in the USA each year.

While there is a close audit of indications at initial ICD implantation, routine reassessment of ICD indications is not mandated when these patients present for ICD generator replacement [6]. Identifying the patients who are least likely to benefit from continued ICD therapy may significantly reduce medical expenses by avoiding unnecessary ICD generator replacement.

Frequency of EF Improvement in Patients with ICD

Left ventricular EF is the cornerstone of the criteria used in the decision process to recommend or decline ICD implantation for primary prevention of SCD [19]. Professional society practice guideline statements recommend ICD implantation in patients with EF ≤35% and mild to moderate heart failure symptoms while taking optimal medical therapy [20,21,22]. The EF cut-off is based on randomized controlled trials, in which patients assigned to ICD and medical therapy were more likely to survive compared to those assigned to medical therapy alone. However, patients presenting for ICD generator replacement have a left ventricular EF that is, on average, 4–5% higher than it had been at the time of the initial ICD implantation [1, 5]. As such, 25–40% of the patients who receive an ICD for primary prevention of SCD experience an improvement in their EF to the extent that they are no longer eligible for ICD therapy when they present for generator replacement. The proportion of patients with EF improvement has been consistent in cohorts that include ICD alone and those that also include CRT [14, 23].

Patients who experience an improvement in EF are younger, more likely to be women, more likely to be taking heart failure medications, and, most notably, more likely to have nonischemic cardiomyopathy [14, 23]. They also have less comorbidity, smaller left ventricular volume, and lower body mass index [24, 25]. Cardiomyopathy due to reversible causes such as tachycardia, myocarditis, pregnancy, hyperthyroidism, stress, pacing, or alcohol is more likely to improve after the offending etiology is treated or eliminated. Thus, 50% of the individuals with nonischemic cardiomyopathy assigned to ICD in the DEFINITE (Defibrillators in Nonischemic Cardiomyopathy Treatment Evaluation) trial experienced a significant (>5% absolute) improvement in EF [13]. In comparison, 20–25% of the patients with ischemic cardiomyopathy experience improvement in EF [26, 27]. Medical therapies and revascularization have been associated with improvement in EF in ischemic cardiomyopathy [28,29,30,31,32,33]. In addition to the factors associated with a higher likelihood of EF improvement, patients with a baseline EF in the range of 30–35% at the time of ICD implantation are more likely to be ineligible for ICD at the time of battery depletion [7, 26]. These data show that EF improvement is common after ICD implantation and 25–40% of the patients who qualified for ICD on the basis of a low EF will no longer be eligible for ICD implantation by the time they present for generator replacement.

Appropriate ICD Therapy After EF Improvement

With improvement in EF, the incidence of appropriate ICD therapy is reduced but not completely eliminated (Table 8.1) [5, 7, 12,13,14, 26, 34,35,36,37,38,39,40,41]. In recent cohort studies, improvement in EF was associated with a 70% reduction in the risk of appropriate ICD therapy, which ranged from 2.8% to 5% per year (Fig. 8.1) [5, 7, 14]. Conversely, two earlier studies had found a similar incidence of appropriate ICD shock among patients with improved or unchanged EF [13, 26]. The reason for the dissimilar results in these studies may have been aggressive ICD programming parameters, which have evolved over the years to reduce shocks delivered for arrhythmias that are likely to terminate spontaneously [42].

Table 8.1 Description of the studies assessing the outcomes of patients with and without improvement in left ventricular ejection fraction after defibrillator implantation
Full size table
Fig. 8.1
figure 1

Cumulative incidence of appropriate ICD therapy after generator replacement among patients with or without EF improvement. (Adopted with permission from Madhavan et al. [5])

Full size image

Absence of appropriate ICD therapy before generator replacement is not a sufficient reassurance for not having future ICD therapies. Approximately, 20% of patients who had an improvement in EF and no prior appropriate ICD therapies experience their first appropriate ICD therapy after generator replacement [26]. In a cohort study of such patients, the incidence of ICD therapy was 5% per year [5].

Patients with normalized EF constitute a special subgroup of EF improvement. These patients either have a reversible cardiomyopathy or can be classified as super responders to CRT. Super responders to CRT have a similar mortality risk to the general population [12, 40, 43]. Although the risk of appropriate ICD shocks decreases markedly in patients with normalized EF, a small risk remains. In a prospective cohort study by Zhang et al., only 1 of the 35 patients with normalized EF had an appropriate ICD shock during follow-up (1.7 shock/100 person-years) [14]. On the other hand, none of the 18 patients with EF >55% after CRT in a series by Manfredi et al. had appropriate ICD therapy [12].

Patients with CRT are more likely to experience improvement and normalization of LVEF compared to patients with ICD [14]. However, the association between changes in left ventricular EF and ICD therapy appears to be similar in ICD and CRT, suggesting that the improvement in EF itself, but not the means that caused the improvement, is responsible from the favorable results [14].

Collectively, these data show that 20–30% of patients with EF improvement are at risk of receiving appropriate ICD shock/therapy, but the risk appears to be lower as the EF approaches normal range. The persisting risk of arrhythmias, observed in some patients despite improvement in EF, may be partly explained by the presence of a fixed substrate for ventricular arrhythmias (e.g., fibrosis, myocardial scar, heterogeneous repolarization) that does not resolve even when EF improves [14, 44,45,46,47,48,49]. However, the other factors associated with a persisting risk of SCD in patients with improved EF are presently unknown.

Do ICDs Reduce Mortality After Improvement of EF?

Left ventricular EF is a major determinant of arrhythmic and non-arrhythmic mortality [50]. Thus, it should come as no surprise that improvement in EF among patients with ICD is associated with improved survival in comparison with unchanged EF in the great majority of the cohort studies to date (Table 8.1). However, because of a lack of a control group without ICD, these cohort studies cannot determine whether ICD improves the likelihood of survival in patients with an improvement in EF. In the absence of prospective randomized controlled trials, we assessed the efficacy of ICD in prolonging survival among patients with improved EF in a secondary analysis of the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) [41]. The SCD-HeFT was a randomized controlled trial of ICD, amiodarone, or placebo among 2511 patients with heart failure symptoms and EF ≤35% due to ischemic or nonischemic cardiomyopathy [51]. After a median 45.5 months of follow-up, the patients assigned to ICD had a lower likelihood of mortality than those assigned to placebo or amiodarone (22%, 29%, and 28%, respectively), resulting in a 23% reduction in the relative risk and a 7.2% reduction in the absolute risk of mortality in comparison with the placebo group. While not mandated by the study protocol, nearly 75% of the patients in SCD-HeFT had a repeated assessment of EF 1 year after enrollment. Of these, 30% assigned to ICD or placebo showed a significant improvement in EF where the mean EF increased from 27% to 45%. During a median follow-up of 30 months after the repeated EF measurement, all-cause mortality rate was lower in the ICD vs. placebo groups both in patients whose EF improved to levels >35% (2.6% vs. 4.5% per 100-person-year follow-up, respectively) and in those whose EF remained ≤35% (7.7% vs. 10.7% per 100-person-year follow-up, respectively) (Fig. 8.2). Compared with placebo, the adjusted hazard ratio for the effect of ICD on mortality was 0.64 (95% CI, 0.48–0.85) in patients with repeated EF ≤35% and 0.62 (95% CI, 0.29–1.30) in those with a repeated EF >35% (Table 8.2). There was no interaction between treatment assignment and repeated EF for predicting mortality, suggesting that the efficacy of ICD was similar in patients with improved or unchanged EF. Cumulatively, these results suggest that mortality is lower among patients with improved EF, but ICD remains effective in reducing all-cause mortality among these patients.

Fig. 8.2
figure 2

Incidence rate of all-cause mortality of patients assigned to ICD vs. placebo. Adjusted hazard ratios of all-cause mortality in the ICD vs the placebo groups were 0.64 (95% CI, 0.48–0.85) in patients with a repeated ejection fraction (EF) ≤35% and 0.62 (95% CI, 0.29–1.30) in those with an EF >35%. (Adopted with permission from Adabag et al. [41])

Full size image
Table 8.2 Incidence rates of all-cause mortality and SCD in the ICD and control groups in each EF category
Full size table

Proposed Algorithm

Patients who present for ICD generator replacement should be reevaluated for the appropriateness of continued ICD therapy (Fig. 8.3). The evaluation should first exclude any potential contraindications, such as advanced malignancy, that may have developed since the initial implant. A repeat echocardiogram to assess left ventricular function is prudent, if one has not been performed since the initial ICD implantation. A frank discussion to learn the patient’s values and wishes about continued ICD therapy is of utmost importance to help guide the decision and to clarify potential misconceptions.

Fig. 8.3
figure 3

Recommended algorithm for patients who present for ICD generator replacement

Full size image

We recommend replacement of the generator if the original indication for ICD was secondary prevention of SCD. The risk of appropriate ICD therapy is higher (10%/year versus 5%/year) if the ICD was implanted for secondary prevention of SCD [52].

We also recommend generator replacement if there was an appropriate ICD therapy (shock or antitachycardia pacing ) during the lifetime of the initial device. In addition to the host factors such as the rate and frequency of the ventricular tachycardia/ventricular fibrillation, the likelihood of appropriate ICD shocks also depends on the programmed tachycardia therapy parameters with a rise in the likelihood with more aggressive programming schemes. Indeed, it has been well documented that some ICD shocks for ventricular tachycardia/ventricular fibrillation are delivered prematurely for arrhythmias that would have terminated anyway. While we support utilization of newer ICD programming schemes to prevent inappropriate and appropriate—but unnecessary—shocks, we also recommend replacing the ICD generator in patients with prior appropriate ICD shock because of the increased risk of future shocks in these patients.

We also recommend generator replacement in patients with channelopathies/inheritable arrhythmogenic syndromes due to the continuation of risk. Similarly, patients whose EF remains ≤35% continue to be at SCD risk and should undergo generator replacement.

On the other hand, some patients who are no longer eligible for ICD due to improvement in EF deserve a fair discussion of whether the SCD risk warrants continuation of ICD therapy. Patients with nonischemic cardiomyopathy have a lower risk of SCD and may not benefit from ongoing ICD therapy if EF has improved [53, 54]. Patients with normalized EF (>55%) may also not benefit from continued ICD therapy [12, 14, 40, 43]. On the other hand, patients with a prior myocardial scar may continue to benefit from ICD even if their EF is better [49].

Older patients who have developed competing risks of death due to new comorbidities (e.g., renal failure) or those with frailty, disability, or cognitive dysfunction should have an opportunity to reevaluate continued ICD therapy with an extensive discussion of goals of care [55, 56]. In cases with difficulty in assessing risk additional markers such as inducibility of ventricular tachycardia, or magnetic resonance imaging to identify and quantify fibrosis may be useful [5, 57].

Management of Unreplaced ICDs

There is very limited data on whether or not to explant the ICD that has reached ERI and does not need replacement. Some device manufacturers recommend explanting the ICD to avoid any potential harm from erratic device behavior. The rationale for this recommendation comes from the concern that as the battery continues to deplete, the performance of the transistors (electrical switches) within the ICD may become unpredictable due to the lack of current supplied to these components, which control a number of functions including sensing, pacing, and shock delivery. However, other manufacturers note that at battery depletion, the ICD will revert to storage mode in which no functionality is present. Thus, although no data are available, it is very unlikely for an ICD at the end of life to be able to generate enough power to deliver a shock. Indeed, in our limited experience, patients prefer leaving the device in order to avoid the burden and stress of the explant procedure. In two anecdotal cases, we have left the device without any negative clinical consequences.

However, patients with CRT defibrillator constitute a special situation. Even if the decision is made not to replace a CRT defibrillator, the device, in most instances, should be replaced with a CRT pacemaker to continue synchronization of the left ventricle, particularly if the patient is a “responder” to CRT. Similarly, among patients with a pacing indication who do not wish continued ICD therapy, the ICD should be replaced with a pacemaker [58, 59].