Sudden Cardiac Death
Sudden cardiac death (SCD) remains a major health care problem despite the progressive decline in the incidence of coronary artery disease (CAD) over the past 50 years. Approximately 310,000 patients die each year from SCD, which accounts for half of all deaths in patients with CAD. Approximately 80% of cases are associated with underlying CAD, 10% to 15% with nonischemic cardiomyopathies, and 5% with primary electrical diseases. SCD is defined as sudden unexpected death from a cardiac cause within 1 hour after the onset of cardiac symptoms. Approximately 80% of cases occur at home, and 40% are unwitnessed. The overall survival rate to hospital discharge for out-of-hospital cardiac arrest in North America is about 8% despite the presence of advanced emergency response systems.
When the cardiac rhythm is assessed within 4 minutes of witnessed arrest, ventricular fibrillation (VF) is present in 95% of cases, and asystole is present in 5%. As the time from collapse to rhythm assessment increases to 12 minutes, VF is present in only 71% of cases, and asystole is present in 29%. Resuscitation outcome is thus critically dependent on time from arrest. Because no antiarrhythmic agent has been shown to prevent SCD in a meaningful way, and because survival is greatest when defibrillation is performed in the first 4 minutes of arrest, the implantable cardioverter-defibrillator (ICD) is a powerful therapeutic agent for prevention of sudden death. However,ICDs are expensive and carry long-term consequences in many patients. Identification of patients who will benefit most from an ICD and those who will not remains an important ongoing clinical challenge.
Although a patient profile at very high risk for sudden death can be defined, the majority of patients who manifest SCD do not fit this profile. Indeed, SCD is the first manifestation of heart disease in 50% of those who die suddenly. Thus, although patients previously diagnosed with CAD, heart failure, and left ventricular (LV) dysfunction have the highest annual risk of sudden death, in terms of absolute numbers, this patient group accounts for only a minority of all patients who will die from cardiac arrest ( Figure 22-1 ).
Implantable Cardioverter-Defibrillator Systems and Technology
Current ICD generator size is as small as 35 cc, compared with 10 cc for a pacemaker. ICD leads may be transvenous, epicardial, or subcutaneous ( Figure 22-2 ). Basic arrhythmia detection by the ICD requires a sensed heart rate greater than a programmable cut-off value for a prescribed duration or number of cardiac cycles. Additional analysis of tachycardia rate variation, electrogram morphology, and atrial/ventricular electrogram relationship may be applied to discriminate ventricular tachycardia (VT) from supraventricular tachycardia (SVT). Antiarrhythmic therapies include high-voltage shocks or antitachycardia pacing. All current ICDs provide bradycardia pacing. Patients with high defibrillation energy requirements may have a subcutaneous coil lead in place, and a completely subcutaneous ICD system is under development. Epicardial leads are now primarily limited to cases in which implantation of transvenous leads for resynchronization therapy fail. For patients with temporary contraindications to ICD implantation or for those with a temporary risk for ventricular arrhythmias, a wearable defibrillator vest is available for use over weeks or months (see Figure 22-2 ). ICDs are considered to be cost effective, with estimated costs of $40,000 to $60,000 per life-year saved.
Implantable Cardioverter-Defibrillators for Secondary Prevention of Sudden Death
The evidence that ICDs reduce mortality rates in those who have survived previous cardiac arrest derives from three randomized clinical trials ( Table 22-1 ). A significant mortality benefit with an ICD was demonstrated in the Antiarrhythmic Drug Versus Defibrillator (AVID) trial, but nonsignificant trends toward reduced mortality rates with an ICD were noted in the Cardiac Arrest Survival in Hamburg (CASH) and the Canadian Implantable Defibrillator Study (CIDS) trials. The latter two trials are believed to have been underpowered to detect a mortality benefit; however, both were found to have contributed more deaths in a meta-analysis of all three studies. This was due in part to longer follow-up in CASH and CIDS compared with AVID. This meta-analysis found a significant 25% relative reduction in mortality rate with an ICD compared with amiodarone therapy (hazard ratio [HR], 0.75; 95% confidence interval [CI], 0.64 to 0.87). The benefit was entirely due to a 50% relative reduction in sudden death (HR, 0.50; 95% CI, 0.34 to 0.62). The absolute reduction in all-cause mortality was 7% (95% CI, 5% to 10%). Fifteen patients needed to be treated with ICD to prevent one death.
Trial | Patients (N) | Follow-up | Patient Group | Trial Design | Primary Endpoint | Mortality Control * | Mortality ICD * | RR with ICD | NNT * |
---|---|---|---|---|---|---|---|---|---|
AVID | 1013 | 18 mo | Resuscitated from VF or sustained VT with syncope or sustained VT plus EF ≤40% and hemodynamic compromise | Randomized ICD vs. amiodarone | All-cause mortality | 24% | 15.8% | 0.73 ( P < .02) | 4.6 |
CIDS | 659 | 3 y | Resuscitated VF or VT or unmonitored syncope | Randomized amiodarone vs. ICD | All-cause mortality | 21% | 14.7% | 0.70 ( P = .142) | 11 |
CASH | 228 | 57 mo | Cardiac arrest secondary to VF or VT | Randomized to ICD vs. amiodarone vs. metoprolol | All-cause mortality | 44% | 36% | 0.61 ( P =.2) | 5.6 |
Implantable Cardioverter-Defibrillators for Primary Prevention of Sudden Cardiac Death
Clinical Trials
Based on success with ICD therapy for the secondary prevention of SCD, numerous trials have been performed for primary prevention in patients at high risk for fatal ventricular arrhythmias but who had not had an arrest ( Table 22-2 ). These trials have a common feature in that the major inclusion criterion was reduced LV ejection fraction (LVEF), which was proven in epidemiologic studies to be a powerful predictor of total mortality rate and SCD. Patients with infarct-related cardiomyopathy initially were targeted by ICD trials because of the known risk of SCD in this patient population. These trials include patients with a reduced LVEF who received ICD late after myocardial infarction (MI). All three of these studies found benefit with prophylactic ICD implantation, with HRs of 0.46 to 0.73 for total mortality. The evidence for benefit to primary prevention treatment is strongest for the cohort of patients with ischemic cardiomyopathy who received ICDs late after MI. Patients with reduced LVEF who received ICDs early (<40 days) after MI (Defibrillator in Acute Myocardial Infarction Trial [DINAMIT] and Insulin Resistance Intervention After Stroke [IRIS] Trial) or patients with a reduced LVEF who underwent surgical revascularization (coronary artery bypass graft [CABG] ) did not show benefit in controlled trials. In these trials, reduced rates of sudden death were offset by higher non-sudden death rates in ICD patients. In two very small studies, patients with a reduced LVEF and nonischemic cardiomyopathy showed no benefit to prophylactic ICD (Cardiomyopathy Trial [CAT] and Amiodarone Versus Implantable Cardioverter-Defibrillator Trial [AMIOVIRT]). The much larger Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) of nonischemic cardiomyopathy patients showed a strong trend toward improved total mortality rate in nonischemic cardiomyopathy patients (relative risk [RR], 0.65; 95% CI, 0.40 to 1.06; P = .08). The largest primary prevention trial to enroll ischemic and nonischemic cardiomyopathy patients, the Sudden Cardiac Death-Heart Failure Trial (SCD-HeFT) of 2521 patients, showed benefit to ICD implantation for all patients combined and for ischemic and nonischemic subgroups ( Figure 22-3 ). Although no single trial has shown a statistically significant reduction in total mortality rate (as a primary endpoint) for nonischemic cardiomyopathy patients treated with primary prevention ICDs, these devices are considered to benefit this patient cohort based on the overall weight of the results from these multiple studies.
Trial | Patients (N) | Follow-up | Patient Group | Trial Design | Primary Endpoint | RR (95% CI) with ICD | NNT |
---|---|---|---|---|---|---|---|
MADIT | 1232 | 20 mo | NYHA class I-III, LVEF ≤30% post MI | Randomized ICD vs. no ICD | All-cause mortality | 0.69 (0.51-0.93), P = .016; 0.74 after extended follow-up at 7.6 y | 8 at 20 mo |
AMIOVIRT | 103 | 2 y | Nonischemic cardiomyopathy and nonsustained VT | Randomized amiodarone vs. ICD | Total mortality | Stopped due to futility | |
CAT | 104 | 5.5 y | Dilated cardiomyopathy and LVEF ≤30% | Randomized ICD vs. no ICD | All-cause mortality | No difference between groups | |
DEFINITE | 458 | 29 mo | Nonischemic cardiomyopathy, LVEF <36%, history of heart failure and >10 PVCs/h or nonsustained VT | Randomized ICD vs. no ICD | All-cause mortality | 0.65 (0.40-1.06), P = .08 | 24 |
SCD-HeFT | 2521 | 45.5 mo | NYHA class II-III and EF ≤35%, ischemic or nonischemic heart disease | Randomized ICD vs. amiodarone vs. placebo | All-cause mortality | 0.77 (0.62-0.96), P = .007); absolute decrease in mortality of 7.2% | 14 |
DINAMIT | 674 | 30 mo | 6-40 days post MI with EF ≤35% and depressed heart rate variability or elevated resting heart rate | Randomized to ICD or no ICD <40 days after MI | All-cause mortality | 1.08 (0.76-1.55), P = .66 | |
IRIS | 898 | 37 mo | 5-31 days post MI with EF ≤40%, nonsustained VT, or heart rate >90 beats/min | Randomized to ICD vs. no ICD <31 days after MI | All-cause mortality | 1.04 (0.81-1.35), P = .78 | |
MADIT | 196 | 27 mo | NYHA class I-II, prior MI, LVEF ≤35%, nonsustained VT and inducible VT not suppressed with procainamide | Randomized to ICD vs. best medical therapy (amiodarone in 74%) | All-cause mortality | 0.46 (0.26-0.82), P = .009 | 2 |
MUSTT | 704 | 39 mo | Prior MI, LVEF ≤40%, and inducible sustained VT | Randomized medical therapy vs. EPS-guided therapy (58% received ICDs) | Cardiac arrest or arrhythmic death | 0.73 (0.53-0.99), P < .004 | 2.5 |
CABG-Patch | 900 | 2.7 y | LVEF ≤35% with abnormal SAECG and elective CABG | Randomized epicardial ICD or no ICD | All-cause mortality | 1.07 (0.81-1.42), P = .64 |
Risk Stratification for Sudden Cardiac Death
ICD therapy is expensive and is associated with long-term consequences. Ideal patient selection criteria would identify those at high risk for malignant arrhythmias and exclude those unlikely to benefit from ICDs because of low risk of SCD or competing nonarrhythmic death. All variables studied to date have limited positive predictive value for SCD. No other risk factor, other than reduced ejection fraction, is currently required for ICD implantation. The following risk stratification measures have received the greatest attention ( Table 22-3 ).
Variable | Critical Measure | PPV | NPV | Comment | Key Reference(s) |
---|---|---|---|---|---|
LVEF | LVEF <30%-40% | 10% | 75% | For VT after MI | |
Electrophysiologic study | Inducible MMVT >15 to 30 sec duration or requiring DCC or sustained PMVT with 1 or 2 extra stimuli | 11% | 95% | CAD patients, value of EPS in nonischemic cardiomyopathy patients limited 1-year follow-up | |
Nonsustained ventricular tachycardia | 10 PVCs/h | 19% | 94% | For VT post MI | |
Signal-averaged ECG | Filtered QRS >120 ms; RMS40 <20 uV or LAS40 >38 ms | 17% | 96% | From post-MI patients | |
T-Wave alternans | >1 min with V alt ≥1.9 uV at heart rate <110 beats/min | 9% | 95% | CAD patients for ICD implant | |
Baroreflex slope | <3 ms/mm Hg | 21% | 95% | Post MI | |
Heart rate variability | 22% | 91% | |||
Ventricular scar burden | Late myocardial enhancement presence or extant | Hazard ratio, 5.2 | Nonischemic cardiomyopathy |