Defibrillation Therapy

Implantable Cardioverter-Defibrillators

Cardiac defibrillation was first developed by Dr. Naum Gurvich, a Russian physician who in 1939 performed the first successful defibrillation in animals. Subsequent development of cardiac defibrillation was advanced by Dr. Claude Beck, a cardiothoracic surgeon from Cleveland who researched cardiac defibrillation in animal models and subsequently in 1946 successfully used defibrillation during cardiac arrest of a young boy undergoing surgery for pectus excavatum. This intracardiac use of defibrillator was followed by development of external defibrillation as described by Zoll in 1956.

In the 1960s and 1970s, Dr. Mieczyslaw (Michel) Mirowski developed an idea that the same device that defibrillates externally could defibrillate internally using a system of generator and lead similar to implantable cardiac pacemakers that were introduced in the 1960s. This led to collaborative work of key inventors: Dr. Michel Mirowski, Dr. Morton Mower, and Dr. Stephen Heilman, who patented the concept of automatic implantable cardioverter-defibrillators with first implantations occurring in 1980 in the United States and in the early 1980s in Europe. The advance of this concept and technology has been very successful in the last 35 years, which has led to widespread use of these devices with several hundred thousand implanted every year worldwide.

Early Clinical Trials with Implantable Cardioverter-Defibrillators in Cardiac Arrest Survivors

SCD survivors were the first group of patients in which ICDs were tested. It was well known that survivors of cardiac arrests have a very high risk of recurrent cardiac arrest and sudden death and early ICD studies evaluated the usefulness of ICD in preventing cardiac arrest in such patients.

The Antiarrhythmics Versus Implantable Defibrillators (AVID) trial was the largest of these trials enrolling 1016 patients, 81% of them with ischemic etiology, who were survivors of cardiac arrest, had sustained VT with syncope, or presence of sustained VT in the presence of ejection fraction (EF) <40%. Patients treated with antiarrhythmic drugs (mostly amiodarone and only 3% with sotalol) had a 24% 2-year mortality, whereas ICD-treated patients had a 16% mortality (hazard ratio [HR] = 0.73). ICD reduced overall mortality by 39% (1 year), 27% (2 years), and 31% (3 years) ( P < 0.02) and reduced arrhythmic death by 56% ( P = 0.0002). There was a 38% relative reduction of death in ICD patients compared with patients treated with class III antiarrhythmic drugs.

The Canadian Implantable Defibrillator Study (CIDS) was a randomized, multicenter, clinical trial comparing ICD therapy ( n = 328) versus amiodarone ( n = 331) in patients with prior cardiac arrest or hemodynamically unstable VT. The study also enrolled patients with sustained VT >150 beats per minute (bpm) causing presyncope or angina with EF ≤35%. Overall mortality was reduced by 20% ( P = 0.142) and did not reach significance, but arrhythmic death was reduced by 33%. The benefit was greatest among patients with EF <35% and advanced New York Heart Association (NYHA) class.

The CASH (Cardiac Arrest Study Hamburg) evaluated the effectiveness of ICD therapy ( n = 99) versus metoprolol ( n = 97), amiodarone ( n = 92), and propafenone ( n = 58) in sudden cardiac arrest survivors. ICD therapy reduced overall mortality by 23% compared with amiodarone and metoprolol ( P = 0.081). Propafenone was discontinued early in the study due to excess mortality (presumably ventricular proarrhythmia). ICD reduced arrhythmic death by 58% ( P = 0.005). Because all three studies enrolled similar type of patients with prior arrhythmic events, the meta-analysis of these three studies was performed with 934 patients receiving ICDs and 932 receiving amiodarone. It demonstrated that ICD therapy was associated with a 28% reduction in total mortality (HR = 0.72; P = 0.0006) and with 50% reduction of arrhythmic mortality (HR = 0.50; P < 0.0001). Importantly, this meta-analysis also showed that there was no significant benefit from ICD in 643 patients with EF >35%, whereas a benefit from ICD therapy was observed in the remaining 1189 patients with EF ≤35%. The above evidence established class I indications for ICD therapy in secondary prevention of mortality in patients with prior arrhythmic events. Table 16-1 summarizes data from the above secondary prevention trials and also puts them in perspective of key results of primary prevention ICD trials. Secondary prevention ICD trials demonstrated that ICD is superior to antiarrhythmic drugs in reducing the risk of mortality in patients with cardiac arrest or documented tachyarrhythmias. However, cardiac arrest survivors and patients with documented VT/VF constitute only a small fraction of patients at risk of SCD.

Clinical Trials in Primary Prevention of Sudden Cardiac Death

Extensive risk stratification research conducted mostly in postinfarction patients in the 1980s demonstrated that left ventricular (LV) dysfunction measured by low EF is the most powerful predictor of mortality and SCD. In addition, presence of frequent ventricular premature beats, nonsustained VT on Holter, as well as inducibility of ventricular tachyarrhythmias in electrophysiologic studies were found to be useful predictors of mortality and SCD. Based on observations by Wilber et al, the MADIT (Multicenter Automatic Defibrillator Implantation Trial) was designed to determine whether patients with EF ≤35%, nonsustained VT, and inducible and nonsuppressible sustained VT induced during electrophysiologic study will benefit from ICD in comparison with a conventional no ICD treatment arm. The trial enrolled 196 patients randomized to ICD versus optimal medical therapy (with 74% receiving empirically administered amiodarone). After a mean 27-month follow-up, there was a significant 54% reduction in mortality with ICD therapy when compared with the non-ICD arm that showed 32% 2-year mortality (HR = 0.46; P = 0.009). About half of enrolled patients received transthoracic ICD devices, and the remaining half re eived transvenous ICD devices that became available during the trial.

A few years later, in 2000, the Multicenter Unsustained Tachycardia Trial (MUSTT) results were published. The study had enrolled 704 postinfarction patients with EF ≤40%, nonsustained VT, and inducible sustained VT. Patients with induced sustained VT were randomized to no antiarrhythmic therapy or to electrophysiologic-guided (EP-guided) therapy. The EP-guided therapy showed substantially lower mortality, and this benefit was entirely due to empiric ICD therapy (not randomized). Non-ICD patients showed 55% mortality, which was reduced to 24% by ICD (HR = 0.49; P < 0.001). Figure 16-1 shows the superimposition of cumulative survival curves in patients treated and not treated with ICD in MADIT and MUSTT trials with a remarkable consistency in the findings. These studies demonstrated the prophylactic use of ICDs in postinfarction patients with depressed EF, nonsustained VT, and inducible sustained VT significantly reduces the risk of SCD and total mortality.

As discussed earlier, low EF especially below 30% identifies a very high risk of mortality (over 20% mortality at 2 years); therefore additional risk stratifiers might not be needed. MADIT II was designed to evaluate the survival benefit of prophylactic ICD therapy in patients with a prior myocardial infarction and LVEF ≤30%, without eligibility requirement for arrhythmia induction using invasive electrophysiologic testing. Patients at least 1 month after myocardial infarction with EF ≤0.30 were randomized to receive either an ICD or conventional medical therapy. The primary end point was all-cause mortality. The MADIT II enrolled 1232 patients, of whom 490 were randomized to conventional therapy and 742 to ICD therapy. The mean EF of studied patients was 23% ± 5% with two thirds of patients in NYHA class ≥II, indicating that MADIT II population represented postinfarction patients with advanced LV dysfunction and symptoms of heart failure in most of them. During a mean 20-month follow-up there were 97 (19.8%) deaths in the conventional arm and 105 (14.2%) deaths in the ICD arm. Figure 16-2 shows significantly improved survival in patients randomized to ICD therapy when compared with patients randomized to conventional therapy with HR = 0.69 ( P = 0.016). The ICD therapy resulted in a 31% reduction in the risk of all-cause mortality and 67% reduction in SCD achieved in addition to optimal ICD therapy.

The SCD-HeFT was a primary prevention trial randomizing 2521 patients with EF ≤35% and ischemic or nonischemic cardiomyopathy NYHA class II or III to amiodarone, ICD, or placebo therapy. There was no benefit from amiodarone therapy, whereas ICD therapy contributed to a significant 23% reduction in mortality ( Fig. 16-3 ) when compared with placebo (HR = 0.77; P = 0.007). The main benefit was observed in patients with EF ≤30%.

The Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) trial was primarily focused on the effects of cardiac resynchronization therapy (CRT) in patients with advanced heart failure and QRS ≥120 msec, but one of the arms of the trial received device combining ICD with cardiac resynchronization pacemaker therapy. The ICD-resynchronization therapy in ischemic patients (55% of study population) was associated with 27% reduction in mortality ( P = 0.082).

ICD therapy in nonischemic cardiomyopathy patients was evaluated in the Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) trial, which enrolled 458 patients with nonischemic dilated cardiomyopathy, EF ≤35%, and premature ventricular complexes or nonsustained VT who were randomized to ICD or no ICD. During a mean 29-month follow-up there were 28 deaths in the ICD group, as compared with 40 in the no ICD group (HR = 0.65; P = 0.08). There were 17 sudden deaths from arrhythmia: 3 in the ICD group, as compared with 14 in the standard therapy group (HR = 0.20; P = 0.006).

The above primary prevention trials, including MADIT, MADIT II, SCD-HeFT, DEFINITE, and COMPANION, showed very consistent results in somewhat diverse groups of patients who were not in acute or postacute ischemic state. However, three trials that evaluated ICD therapy in early ischemia period failed to show benefit from ICD when compared with conventional therapy. The CABG Patch Trial enrolled 900 postinfarction patients with EF ≤35% and abnormal signal-averaged electrocardiogram (ECG). During a 32-month follow-up there was no significant difference in mortality between no ICD and ICD arms (24% vs. 27%, respectively). The Defibrillator in Acute Myocardial Infarction Trial (DINAMIT) enrolled 674 patients 6 to 40 days after acute myocardial infarction with EF ≤35% and signs of impaired cardiac autonomic modulation (standard deviation of NN intervals [SDNN] from heart rate variability ≤70 msec or elevated heart rate ≥80 bpm). A total mortality was 6.9% in the control group versus 7.5% in ICD patients ( P = 0.66). ICD therapy reduced the cumulative risk of arrhythmic death by 58% (95% CI, 0.22-0.83) compared with control. This reduced risk of arrhythmic death in the ICD arm was offset by an increase in nonarrhythmic death compared with control (6.1% vs. 3.5%, P =0.016). IRIS trial enrolled 898 patients 5 to 31 days after a myocardial infarction with EF ≤40%, heart rate of 90 bpm, or nonsustained VT. During a mean follow-up of 37 months, 233 patients died and there was no difference in mortality between patients randomized to ICD or no ICD. SCD was reduced by 45% in ICD patients (HR = 0.55; P = 0.049), but non-SCD was increased in these patients (HR = 1.92; P = 0.001).

These studies enrolled patients with recent ischemic events when the myocardium undergoes early stages of remodeling and healing supported by revascularization therapies. EF might recover in such patients in the first few months after myocardial infarction and implanting ICD at this early stage might not be effective in reducing mortality. These patients might benefit from temporary protection with WCD, which was found to be effective in reducing mortality in early postinfarction patients. This discrepancy in the beneficial WCD effects versus no effect of ICD in DINAMIT and IRIS studies is puzzling. Both DINAMIT and IRIS trials subselect patients with low heart rate variability or impaired autonomic control of the heart. This subselection might have been counterproductive because impaired autonomic tone in patients with low EF predisposed to pump failure death instead of arrhythmic death. The ongoing VEST trial will help to answer this question.

Current Implantable Cardioverter-Defibrillator Indication Guidelines

The clinical trials described above contributed to establishing the 2008 ACC/AHA/HRS guidelines, updated in 2012, regarding optimal use of ICD and CRT therapy. Box 16-1 shows current recommendations for ICD implantation with class I indications supported by level of evidence A or B coming from randomized clinical trials. These recommendations seem to be followed in clinical practice; however, there are significant disparities in utilization of ICDs. Only about 20% to 30% of eligible patients receive devices for primary prevention of mortality despite benefits clearly demonstrated in clinical trials and subsequent real-practice cohorts evaluated by the NCDR ICD Registry investigators. The beneficial effects of CRT, described below, should further encourage physicians to consider devices in much broader patient population.

The 2012 recommendations also include ICD indications in rare conditions and these guidelines are in class IIa and class IIb with the level of evidence C coming from observational studies, not randomized clinical trials. In 2013, the HRS/EHRA/APHRS Expert Consensus Statement on the Diagnosis and Management of patients with Inherited Primary Arrhythmia Syndromes was published and Box 16-2 provides details regarding specific syndromes and current recommendations regarding ICD treatment in these conditions.