The use of implantable devices to treat arrhythmias started with the implantable pacemaker back in the late 1950s. The risks and costs of implantation limited the indications for the procedure initially. A similar pattern was seen with the initial implantable cardioverter-defibrillators (ICDs) in the early 1980s. In the late 1990s, technology of both these devices focused on the management of systolic heart failure through cardiac resynchronization therapy (CRT). Currently, there are implantable devices being evaluated for optimizing the management of heart failure by frequently monitoring intracardiac pressures and surrogates for volume status. As the technology has evolved, the indications for these devices have expanded. There has been concern that too many devices are being implanted without solid clinical indication. A mandated registry for Medicare patients receiving ICDs is now in place to evaluate appropriate use of this technology. In the future, this registry will likely be expanded to include pacemakers. How the data in this registry are used will impact the future indications for device management. We will review the history and future of device management for arrhythmias and heart failure.
The first implantable pacemakers appeared the 1950s/60s. These preliminary devices were limited due to battery life. The first transvenous leads were developed in the mid-1960s. Initial indications were restricted to symptomatic complete heart block. The first American Heart Association (AHA) and American College of Cardiology (ACC)-sponsored guidelines for pacemakers were published in 1984. Indications for pacemakers have markedly expanded in the last 40 years. The most common indication for a permanent pacemaker is symptomatic bradyarrhythmias, including sinus node dysfunction and various degrees of heart block. Other indications include management of atrial fibrillation due to the coexistence of sinus node dysfunction and atrial fibrillation, the development of bradycardia as a common side effect of pharmacological interventions aimed at rate or rhythm control, and the need for pacing after atrioventricular node ablation. The application of atrial pacing for suppression and therapy of atrial tachyarrhythmias will be expanded.
The concept of an ICD came to light in the late 1960s, but it was not until 1980 that it became a reality for patients.1 The ICD was approved by the US Food and Drug Administration (FDA) in 1985 based only on observational data showing survival benefit in those receiving an ICD in the initial studies of feasibility. Indications were very restrictive: only patients who had survived two cardiac arrests were eligible.
Three landmark clinical trials were initiated from 1987 to 1993 to evaluate the mortality benefit of ICDs in survivors of cardiac arrest compared with treatment with antiarrhythmic medications. The US study Antiarrhythmics versus implantable defibrillators (AVID),2 like the studies conducted in Canada and Germany, showed a survival benefit in the ICD group of the study. All three trials had results in 1997 showing superiority of the ICD compared with antiarrhythmic medications in survivors of cardiac arrest.
Around the same time, in 1996, the results of the Multicenter automatic defibrillator trial (MADIT)3 reported the first evidence from a randomized controlled trial that ICDs improved survival among individuals at risk without a history of cardiac arrest. In this small study of fewer than 200 patients, patients with a history of myocardial infarction (MI), left ventricular ejection fraction (LVEF) ≤35 %, non-sustained ventricular tachycardia (VT) and inducible sustained VT/ventricular fibrillation (VF) by electrophysiology study (EPS) showed a significant survival advantage when receiving an ICD compared with those receiving conventional medical therapy. In fact, this study was stopped prematurely due to the survival benefit seen early in the ICD group (15 deaths compared with 39 deaths in the control group).
As a result of these secondary and primary prevention trials, the ICD indications expanded in 1999 to include survivors of a single cardiac arrest from VT/VF not due to an acute MI or reversible cause, and those with sustained ventricular tachyarrhythmias (even those induced in the electrophysiology laboratory). Also, individuals with documented inherited conditions such as hypertrophic cardiomyopathy or long QT syndrome at high risk of ventricular arrhythmias started to be considered for ICDs.
The marked survival benefit with ICDs in the MADIT trial led to larger hallmark trials evaluating the utility of ICDs in primary prevention of sudden cardiac death. The sequel to the first trial, MADIT II,4 evaluated individuals with MI and ejection fraction ≤30 %. This trial was stopped early, just like the MADIT trial, due to marked survival benefit in the ICD group, with a 31 % relative risk reduction in all-cause mortality in two years. The largest ICD trial, of more than 2,500 patients, was the Sudden cardiac death in heart failure trial (SCD HeFT),5 evaluating patients with New York Heart Association (NYHA) functional class II and III heart failure and LVEF ≤35 % from either ischemic or non-ischemic etiology on stable medical therapy. This trial showed a 23 % reduction in all-cause mortality. The consequences of these findings included expanding the indications to far more of the population at risk. Figure 1 shows an example of how ICDs can save lives.
With this evidence, the Centers for Medicare and Medicaid Services (CMS) did approve the expanded indications for primary prevention of sudden cardiac death, with the caveat that patients be followed for further confirmatory evidence of benefit—a policy that was termed Coverage with Evidence Development (CED). The CED included mandatory enrollment of patients in a national database for evaluation. In 2005, at the time of the initial approval of the current primary prevention indications, the data were entered in a quality de-identified database. However, within the same year, the Heart Rhythm Society and the ACC Foundation recommended an improved data collection tool. By the second quarter of 2006, mandatory reporting of data for Medicare primary prevention patients was introduced in version 1 of the National Cardiovascular Data Registry (NCDR) ICD Registry. This registry included 159 data points per case. It was also recommended, but not mandatory, that all ICD recipients be enrolled. As of early 2010, more than 77 % of the 1,500+ sites were enrolling all ICD implant patients.6 The accuracy of this data collection has not been validated. Certainly, the process is time-consuming without easy tools for data entry. Who enters the data and how the data are entered differ from institution to institution and the accuracy is likely to be variable.
Nonetheless, the data collected from this endeavor are being evaluated for the purpose of gaining further evidence of benefit or harm from these implanted devices. A recent study analyzing data from the NCDR ICD Registry suggests a high percentage of inappropriate use of ICDs among participating centers.7 This raises the question of whether the guidelines are being followed and whether evidence-based medicine is being practiced correctly in regard to ICDs. However, these conclusions do not consider that the accuracy of the registry may be variable or take into account the fact that clinical situations rarely fit the strict entry criteria designed for controlled clinical trials. The clinician is more often than not challenged by clinical scenarios that deviate from practice guidelines and is forced to make decisions based on best clinical judgment—the inherent nature of the practice and art of medicine. Generator replacements after partial left ventricular (LV) systolic function recovery; ICD implants after incomplete revascularization; children and adults with congenital heart disease and severe systemic or non-systemic systolic dysfunction, with or without high-risk features such as syncope or non-sustained VT; and complete heart block with newly diagnosed cardiomyopathy and definitive pacing indication: these are just a few examples of patients who could be deemed inappropriate by NCDR standards, not clearly addressed in published guidelines but indicated from the clinician’s perspective, and only supported by limited evidence in observational cohort evaluations or expert opinion (see Figure 2).
In the last year, there has been an expansion of the NCDR ICD Registry data requirements to more than 250 data points. With this expansion comes the increased work and effort for data entry, but hopefully such data collection will help answer some of the questions and clinical scenarios listed above to optimize the future management of these patients. However, great care will need to be taken in entering these data, as the results from the registry will be as good as the data entered.
Cardiac Resynchronization Therapy
The recognition of left bundle branch block and QRS width as markers of bad prognosis in heart failure led to the understanding of interventricular and intraventricular dyssynchrony.8–14 In the late 1990s, much investigation was focused on CRT in heart failure, resulting in the initial randomized controlled trial of CRT pacemakers, Multicenter insync randomized clinical evaluation (MIRACLE),15 followed by a larger trial of CRT pacemakers and defibrillators, the Comparison of medical therapy, pacing, and defibrillation in heart failure (COMPANION) trial.16 Both of these trials showed an improvement in quality of life in patients with systolic heart failure, ejection fraction less than 35 %, wide QRS, and NYHA functional class III and IV. The COMPANION trial, which included ICD, also showed a survival benefit. These, and many other studies involving CRT, led to a change in the therapeutic approach to heart failure. Beyond the benefits of the ICD, selected patients with electrocardiographic markers of dyssynchrony manifesting as a wide QRS, and moderately to severely symptomatic heart failure, are now known to benefit from a device intervention that decreases hospital admissions and reduces mortality. In addition, studies have shown that CRT can decrease filling pressures, increase cardiac output and ejection fraction, reduce ventricular size, and prevent ventricular arrhythmias.17–20 The indication for CRT defibrillators (CRT-Ds) and CRT pacemakers (CRT-Ps) was rapidly included in the professional organization guidelines and approved by the CMS and private insurances in the US. Figures 3 and 4 show some of the benefits of CRT demonstrated on electrocardiograms and by echocardiography, respectively.
Attempts to expand the indications for CRT have faced the challenges inherent to managing clinical heart failure. The Left ventricular-based cardiac stimulation post AV nodal ablation evaluation study (PAVE)21 conferred the indication for CRT to patients with difficult-to-control atrial fibrillation, heart failure symptoms, and LVEF ≤45 % undergoing atrioventricular nodal ablation. It showed an improvement in the six-minute walk distance among the CRT recipients.
Initial reports of benefit among patients with heart failure, narrow QRS, and echocardiographic evidence of dyssynchrony22–24 could not be confirmed in a larger clinical trial.25 Even more discouraging, the study of Predictors of response to CRT (PROSPECT),26 designed to identify the most appropriate echocardiographic assessment of ventricular dyssynchrony, showed poor performance of the echocardiographic evaluation and failed to change the standard for screening, which remains the QRS width.
Despite a multiplicity of clinical and paraclinical parameters available for the evaluation of heart failure, the definition of response to therapy has been a complex puzzle. Mortality, hospital admissions, and NYHA functional class are measures of dramatic changes in the patient’s condition; lack of progression of heart failure, reduction in LV size, pulmonary pressures, brain natriuretic peptide levels, and improvement in six-minute walk are variables with relatively less prognostic value and are more difficult to interpret.27 Patients undergoing atrioventricular nodal ablation and pacing as treatment of difficult-to-control atrial fibrillation have been shown to benefit from biventricular pacing when compared with right ventricular pacing alone, as have patients with other indications for pacing, regardless of their LVEF.28,29
Similarly, discrete benefit in LV diameter and reduction in hospital readmissions was clearly established for patients with ejection fractions <40 and 35 %, wide QRS, and NYHA class II heart failure symptoms, in the Resynchronization reverses remodeling in systolic left ventricular dysfunction (REVERSE)30,31 and MADIT CRT32 trials. These promising results in patients with less symptomatic heart failure have resulted in FDA approval for CRT ICDs in NYHA class II patients. However, there has not yet been an update of the current guidelines, nor in CMS reimbursement. The pediatric population, adults with congenital heart disease, and valvular heart disease patients represent examples of populations in which large clinical trials will not be feasible to demonstrate definitive clinical benefit using traditionally accepted endpoints, and among whom CRT-P or CRT-D may offer long-term benefit. Room for clinical judgment, individualized patient care, and more inclusive registries are needed and will probably result in expanded indications for CRT. At the same time, while factors limiting the benefits of CRT have been described—including the presence of extensive scarring in the anterolateral wall of the left ventricle, right bundle branch block, suboptimal anatomical placement of the LV lead, and atrial fibrillation33–36—our ability to predict clinical response in an individual patient is very limited.
Pacing Therapies for Atrial Fibrillation
Coexistence of sinus node dysfunction and atrial fibrillation could be explained by the same disease process in the atria or the dispersion of repolarization in atrial tissue induced by sinus node dysfunction, which facilitates the occurrence of atrial premature complexes and multiple wavelets of re-entry.37,38 Atrial pacing to prevent bradycardia and atrial pauses has been considered as a strategy to prevent atrial fibrillation and atrial flutter. Further elaboration of this and other basic electrophysiology concepts led to the development of complex algorithms of atrial fibrillation suppression through atrial pacing.39–42 Pacing at a faster rate than baseline to prevent spontaneous electrical atrial activity, pacing faster in response to atrial premature complexes or after an episode of atrial fibrillation, anti-tachycardia pacing in the atrium, and even automated atrial defibrillation have been evaluated as therapeutic pacing interventions for atrial fibrillation and atrial flutter (see Figure 5).43–46 Unfortunately, these pacing techniques have not eliminated atrial arrhythmias. These atrial interventions have demonstrated a modest reduction in atrial fibrillation frequency and duration.47,48 Today, selected pacemakers and defibrillators are available with atrial fibrillation suppression algorithms and anti-tachycardia pacing as a therapeutic intervention, but atrial defibrillators are not commercially available in the US.
Paradoxically, less pacing, rather than more, in the right ventricle has been proven to result in less incidence of atrial fibrillation and shifted the focus from atrioventricular synchrony to LV synchrony.49 Algorithms designed to minimize ventricular pacing at the expense of prolonging the atrioventricular delay to non-physiologic intervals have become standard of care among pacemaker recipients without complete heart block. The Dual chamber and VVI implantable defibrillator (DAVID) II trial50 brought to an end the popular clinical suspicion among implanters that dual-chamber ICDs were preferred to single-chamber ICDs, due to atrial pacing for prevention of atrial fibrillation or for sustained VT discrimination. At present, an atrial lead is not indicated for a defibrillator unless there is an indication for atrial pacing, which nevertheless frequently happens to be the case among ICD recipients with coexisting atrial fibrillation. Minimizing ventricular pacing in patients with ICDs who will require up-titration of beta-blockers or antiarrhythmic drugs requires the implantation of an atrial lead.
The main indication for atrial pacing among patients with atrial fibrillation remains the treatment of sinus pauses after conversion or sinus bradycardia resulting as a side effect from rate- or rhythm-control pharmacological agents. Pacing is also indicated when atrioventricular nodal ablation is considered the intervention of choice for rate control. This definitive rate control intervention is frequently used in symptomatic patients who have failed to respond, or do not wish to undergo, pulmonary vein isolation or pharmacological therapy.
Other Heart Failure Management Devices
There are currently several implantable devices under investigation that may improve heart failure management in the future. Many of these investigational devices use pressure or volume surrogates for assessment of heart failure status, and these measurements then direct heart failure management. The rationale for such devices is that early detection of volume overload or weight gain and early intervention with medical changes—such as increasing diuretics at home—can prevent acute or chronic heart failure exacerbations and heart failure hospitalization. These monitoring devices have been used as part of an ICD system such as thoracic impedance measurement in ICDs.51 Other monitors incorporate daily weight and blood pressure measurements that are followed and managed remotely by heart failure teams.52 There have also been stand-alone devices measuring pulmonary or left atrial pressure changes that can be early markers of heart failure decompensation.53,54 The preliminary data regarding such devices for heart failure management show promise, but these require further testing before they would be routinely considered for contemporary management of heart failure.
Starting from the initial mid-century pacemakers, implantable devices have been used to aid in the management of arrhythmias and heart failure for more than 40 years. The first 30 years showed a marked expansion in the use of pacemakers. In the last 10 years, we have seen improved survival and quality of life with ICDs and CRT-Ds. With these results from large clinical trials comes some degree of hesitation and criticism. Challenges from the scientific, financial, and regulatory standpoints are various. The NCDR ICD registry has the potential to generate the data needed to further evaluate and potentially validate the benefit of these devices in the future. The increased longevity of the general population and improved outcomes in the treatment of cardiovascular diseases have also been determinant factors in the pressing need for newer technologies aimed at optimizing the management of heart failure and cardiac arrhythmias. In addition to the existing therapies, other implantable devices are under investigation to assist in the optimal management of the chronic heart failure patient. The demand for implantable cardiac devices is likely to continue to increase over time, perhaps not as fast as initially expected, but at a slow, progressive rate.