Implantable Cardioverter Defibrillator. Description

Transcription

1 Subject: Implantable Cardioverter Defibrillator Page: 1 of 41 Last Review Status/Date: December 2015 Implantable Cardioverter Defibrillator Description The automatic implantable cardioverter defibrillator (ICD) is a device designed to monitor a patient s heart rate, recognize ventricular fibrillation (VF) or ventricular tachycardia (VT), and deliver an electric shock to terminate these arrhythmias to reduce the risk of sudden death. Background Automatic implantable cardioverter defibrillators (ICD) monitor a patient s heart rate, recognize ventricular fibrillation (VF) or ventricular tachycardia (VT), and deliver an electric shock to terminate these arrhythmias to reduce the risk of sudden death. Indications for ICD implantation can be broadly subdivided into 1) secondary prevention, i.e., their use in patients who have experienced a potentially life-threatening episode of ventricular tachyarrhythmia (near sudden cardiac death); and 2) primary prevention; i.e., their use in patients who are considered at high risk for sudden cardiac death but who have not yet experienced life-threatening VT or VF. The standard implantable cardioverter defibrillator (ICD) involves placement of a generator in the subcutaneous tissue of the chest wall. Transvenous leads are attached to the generator and threaded intravenously into the endocardium. The leads sense and transmit information on cardiac rhythm to the generator, which analyzes the rhythm information and produces an electrical shock when a malignant arrhythmia is recognized. A totally subcutaneous ICD (S-ICD ) has also been developed. This device does not employ transvenous leads and thus avoids the need for venous access and complications associated with the venous leads. Rather, the S-ICD uses a subcutaneous electrode that is implanted adjacent to the left sternum. The electrodes sense the cardiac rhythm and deliver countershocks through the subcutaneous tissue of the chest wall. Several automatic ICDs are approved by FDA through the premarket application (PMA) approval process. FDA-labeled indications generally include patients who have experienced life-threatening VT associated with cardiac arrest or VT associated with hemodynamic compromise and resistance to pharmacologic treatment. Devices manufactured by Guidant are approved by FDA for use in patients at high risk of sudden cardiac death due to ventricular arrhythmias and who have experienced at least 1 of the following: an episode of cardiac arrest (manifested by the loss of consciousness) due to a ventricular tachyarrhythmia; recurrent, poorly tolerated sustained ventricular tachycardia (VT); or a

2 Subject: Implantable Cardioverter Defibrillator Page: 2 of 41 prior myocardial infarction (MI), left ventricular ejection fraction of less than or equal to 35%, and a documented episode of nonsustained VT, with an inducible ventricular tachyarrhythmia. On July 18, 2002, FDA expanded the approved indications for the Guidant ICD devices to include the prophylactic use of Guidant ICDs for cardiac patients who have had a previous heart attack and have an ejection fraction that is 30% or less. This expanded indication is based on the results of the second Multicenter Automatic Defibrillator Implantation Trial (MADIT II trial), which is discussed here. Medtronic devices are approved to provide ventricular antitachycardia pacing and ventricular defibrillation for automated treatment of life-threatening ventricular arrhythmias. Other devices have approval language similar to that of Medtronic. Regulatory Status Transvenous ICDs The Food and Drug Administration (FDA) has approved a large number of ICDs through the PMA process (FDA product code: LWS). A 2014 review of FDA approvals of cardiac implantable devices reported that between 1979 and 2012, FDA approved 19 ICDs (7 pulse generators, 3 leads, 9 combined systems) through new PMA applications. (1) Many originally-approved ICDs have undergone multiple supplemental applications. A summary of some currently-available ICDs is provided in Table 1 (not an exhaustive list). Table 1: Implantable Cardioverter Defibrillator With FDA Approval Device Manufacturer Original PMA Approval Date Ellipse/Fortify Assura Family (originally: Cadence Tiered Therapy Defibrillation System) Current Plus ICD (originally: Cadence Tiered Therapy Defibrillation System) Dynagen, Inogen, Origen, and Teligen Family (originally: Ventak, Vitality, Cofient family) Evera Family (originally: Virtuosos/ Entrust/Maximo/ Intrisic/ Marquis family) Subcutaneous Implantable Defibrillator System St. Jude Medical Inc. (St. Paul, MN) St. Jude Medical Inc. (St. Paul, MN) Boston Scientific Inc. (Marlborough, MA) Medtronic Inc. (Minneapolis, MN) Cameron Health Inc. (San Clemente, CA); acquired by Boston Scientific Inc. Jul 1993 Jul 1993 Jan 1998 Dec 1998 Nov 2012 Type Transvenous Transvenous Transvenous Transvenous Subcutaneous

3 Subject: Implantable Cardioverter Defibrillator Page: 3 of 41 Subcutaneous ICDs In September 2012, the FDA approved the Subcutaneous Implantable Defibrillator (S-ICD ) System (Cameron Health, Inc, San Clemente, CA; acquired by Boston Scientific, Inc,. Marlborough, MA), through the PMA process for the treatment of life-threatening ventricular tachyarrhythmias in patients who do not have symptomatic bradycardia, incessant ventricular tachycardia, or spontaneous, frequently recurring ventricular tachycardia that is reliably terminated with anti-tachycardia pacing. In March 2015, the Emblem S-ICD (Boston Scientific, Inc., Marlborough, MA), which is smaller and longer-lasting than the original S-ICD, was cleared for marketing through a PMA supplement. NOTE: ICDs may be combined with other pacing devices, such as pacemakers for atrial fibrillation, or biventricular pacemakers designed to treat heart failure. This evidence review addresses ICDs alone, when used solely to treat patients at risk for ventricular arrhythmias. Related Policies Biventricular Pacemakers (Cardiac Resynchronization Therapy) for the Treatment of Heart Failure Policy *This policy statement applies to clinical review performed for pre-service (Prior Approval, Precertification, Advanced Benefit Determination, etc.) and/or post-service claims. Adults The use of the automatic implantable cardioverter defibrillator (ICD) may be considered medically necessary in adults who meet the following criteria: Primary Prevention ischemic cardiomyopathy with New York Heart Association (NYHA) functional Class II or Class III symptoms, a history of myocardial infarction at least 40 days before ICD treatment, and left ventricular ejection fraction of 35% or less; or ischemic cardiomyopathy with NYHA functional Class I symptoms, a history of myocardial infarction at least 40 days before ICD treatment, and left ventricular ejection fraction of 30% or less; or nonischemic dilated cardiomyopathy and left ventricular ejection fraction of 35% or less, after reversible causes have been excluded, and the response to optimal medical therapy has been adequately determined; or hypertrophic cardiomyopathy (HCM) with 1 or more major risk factors for sudden cardiac death (history of premature HCM-related sudden death in 1 or more first-degree relatives younger than 50 years; left ventricular hypertrophy greater than 30 mm; 1 or more runs of nonsustained ventricular tachycardia at heart rates of 120 beats per minute or greater on 24-hour Holter

4 Subject: Implantable Cardioverter Defibrillator Page: 4 of 41 monitoring; prior unexplained syncope inconsistent with neurocardiogenic origin) and judged to be at high risk for sudden cardiac death by a physician experienced in the care of patients with HCM diagnosis of any one of the following cardiac ion channelopathies and considered to be at high risk for sudden cardiac death (see Policy Guidelines ): o congenital long QT syndrome; OR o Brugada syndrome; OR o short QT syndrome; OR o catecholaminergic polymorphic ventricular tachycardia. Secondary Prevention patients with a history of a life-threatening clinical event associated with ventricular arrhythmic events such as sustained ventricular tachyarrhythmia after reversible causes (e.g., acute ischemia) have been excluded. The use of the ICD is considered investigational in primary prevention patients who: have had an acute myocardial infarction (i.e., less than 40 days before ICD treatment); have New York Heart Association (NYHA) Class IV congestive heart failure (unless patient is eligible to receive a combination cardiac resynchronization therapy ICD device); have had a cardiac revascularization procedure in past 3 months (coronary artery bypass graft [CABG] or percutaneous transluminal coronary angioplasty [PTCA]) or are candidates for a cardiac revascularization procedure; or have noncardiac disease that would be associated with life expectancy less than 1 year. Pediatrics The use of the ICD may be considered medically necessary in children who meet any of the following criteria: survivors of cardiac arrest, after reversible causes have been excluded; symptomatic, sustained ventricular tachycardia in association with congenital heart disease in patients who have undergone hemodynamic and electrophysiologic evaluation; or congenital heart disease with recurrent syncope of undetermined origin in the presence of either ventricular dysfunction or inducible ventricular arrhythmias. hypertrophic cardiomyopathy (HCM) with 1 or more major risk factors for sudden cardiac death (history of premature HCM-related sudden death in 1 or more first-degree relatives younger than 50 years; massive left ventricular hypertrophy based on age-specific norms; prior unexplained syncope inconsistent with neurocardiogenic origin) and judged to be at high risk for sudden cardiac death by a physician experienced in the care of patients with HCM. diagnosis of any one of the following cardiac ion channelopathies and considered to be

5 Subject: Implantable Cardioverter Defibrillator Page: 5 of 41 at high risk for sudden cardiac death (see Policy Guidelines ): o congenital long QT syndrome; OR o Brugada syndrome; OR o short QT syndrome; OR o catecholaminergic polymorphic ventricular tachycardia. The use of the ICD is considered investigational for all other indications in pediatric patients. Subcutaneous ICD The use of a subcutaneous ICD may be considered medically necessary for adults or children who have an indication for ICD implantation for primary or secondary prevention for any of the above reasons and meet all of the following criteria: Have a contraindication to a transvenous ICD due to one or more of the following: (1) lack of adequate vascular access; (2) compelling reason to preserve existing vascular access (ie, need for chronic dialysis; younger patient with anticipated long-term need for ICD therapy); or (3) history of need for explantation of a transvenous ICD due to a complication, with ongoing need for ICD therapy. Have no indication for antibradycardia pacing; AND Do not have ventricular arrhythmias that are known or anticipated to respond to antitachycardia pacing. The use of a subcutaneous ICD is considered investigational in adult and pediatric patients. for individuals who do not meet the criteria outlined above. Policy Guidelines This policy addressed the use of ICD devices as stand-alone interventions, not as combination devices to treat heart failure (i.e., cardiac resynchronization devices) or in combination with pacemakers. Unless specified, the policy statements and policy rationale are referring to transvenous ICDs. Indications for pediatric ICD use are based on American College of Cardiology/American Heart Association/Heart Rhythm Society (ACC/AHA/HRS) guidelines published in 2008, which acknowledged the lack of primary research in this field on pediatric patients (see Rationale section). These are derived from nonrandomized studies, extrapolation from adult clinical trials, and expert consensus. Criteria for ICD Implantation in Patients with Cardiac Ion Channelopathies Individuals with cardiac ion channelopathies may have a history of a life-threatening clinical event associated with ventricular arrhythmic events such as sustained ventricular tachyarrhythmia, after reversible causes, in which case they should be considered for ICD implantation for secondary prevention, even if they do not meet criteria for primary prevention.

6 Subject: Implantable Cardioverter Defibrillator Page: 6 of 41 Criteria for ICD implantation in patients with cardiac ion channelopathies are derived from results of clinical input, a 2013 consensus statement from the HRS, European Heart Rhythm Association (EHRA), and the Asia-Pacific Heart Rhythm Society on the diagnosis and management of patients with inherited primary arrhythmia syndromes (Priori et al, 2013), 2013 guidelines from the ACC, AHA, HRS, the American Association of Thoracic Surgeons, and the Society of Thoracic Surgeons on device-based therapy of cardiac rhythm abnormalities (Tracy et al, 2013), and a report from the HRS/EHRA s Second Consensus Conference on Brugada syndrome (Antzelevitch et al, 2005). Indications for consideration for ICD implantation for each cardiac ion channelopathy are as follows: Long QT syndrome: o Patients with a diagnosis of LQTS who are survivors of cardiac arrest. o Patients with a diagnosis of LQTS who experience recurrent syncopal events while on betablocker therapy. Brugada syndrome: o Patients with a diagnosis of BrS who are survivors of cardiac arrest. o Patients with a diagnosis of BrS who have documented spontaneous sustained ventricular tachycardia (VT) with or without syncope. o Patients with a spontaneous diagnostic type 1 ECG who have a history of syncope, seizure, or nocturnal agonal respiration judged to be likely caused by ventricular arrhythmias (after noncardiac causes have been ruled out). o Patients with a diagnosis of BrS who develop ventricular fibrillation (VF) during programmed electrical stimulation. Catecholaminergic polymorphic ventricular tachycardia: o Patients with a diagnosis of CPVT who are survivors of cardiac arrest. o Patients with a diagnosis of CPVT who experience recurrent syncope or polymorphic/bidirectional VT despite optimal medical management, and/or left cardiac sympathetic denervation. Short QT syndrome: o Patients with a diagnosis of SQTS who are survivors of cardiac arrest. o Patients with a diagnosis of SQTS who are symptomatic and have documented spontaneous VT with or without syncope. o Patients with a diagnosis of SQTS or are asymptomatic or symptomatic and have a family history of sudden cardiac death. NOTE: For congenital LQTS, patients may have one or more clinical or historical findings other than those outlined above that may, alone or in combination, put them at higher risk for sudden cardiac death. These may include patients with a family history of sudden cardiac death due to LQTS, infants with a diagnosis of LQTS with functional 2:1 atrioventricular block, patients with a diagnosis of LQTS in conjunction with a diagnosis of Jervell and Lange-Nielsen syndrome or Timothy syndrome, and patients a diagnosis of LQTS with profound QT prolongation (>550 msec). These factors should be evaluated on an individualized basis by a clinician with expertise in LQTS in considering the need for an ICD implantation.

7 Subject: Implantable Cardioverter Defibrillator Page: 7 of 41 Rationale Transvenous ICDs for Primary Prevention in Adults Transvenous implantable cardioverter defibrillators have been evaluated for primary prevention in a number of populations considered at high risk of sudden cardiac death, including those with ischemic cardiomyopathy, nonischemic dilated cardiomyopathy, and hypertrophic cardiomyopathy. There is a large body of evidence, including a number of randomized clinical trials (RCTs) and systematic reviews of these trials addressing the role of ICDs for primary prevention, and identifying specific populations who may benefit. Overview and Summary of TEC Assessments Automatic implantable cardiac defibrillators (ICDs) were first used in survivors of near sudden cardiac death. There has been ongoing interest in using ICDs as primary preventive therapy in patients with risk factors for sudden cardiac death. Several BCBSA TEC Assessments have addressed the use of ICDs for primary prevention of sudden cardiac death. The first ICD TEC Assessment, published in 2002, focused on the Multicenter Automatic Defibrillator Implantation Trials (known as MADIT I and MADIT II) that compared the use of an ICD with conventional therapy among patients with coronary artery disease with a history of myocardial infarction (MI) and a reduced ejection fraction. (2)The key difference in the 2 trials was the patient selection criteria. In the MADIT I trial, patients were required to have a left-ventricular ejection fraction (LVEF) of less than 35% but also ventricular tachyarrhythmia, as evidenced on an electrophysiologic study. In the subsequent, MADIT II, trial, patients were required to have a lower ejection fraction, less than 30%, but no electrophysiologic study was required. Therefore, the patient selection criteria of the MADIT II trial could potentially identify a much larger number of candidates for ICD implantation. The 2002 TEC Assessment concluded, For patients who have coronary artery disease with prior MI and reduced LVEF and who are similar to those selected in MADIT I and MADIT II, the available evidence demonstrates an improvement in overall mortality associated with ICD treatment compared with conventional therapy. In October 2004, TEC reassessed ICDs for primary prevention of sudden cardiac death. (3) The 2004 TEC Assessment focused on the results of the 5 randomized clinical trials (RCTs) included in the 2002 Assessment (including the Multicenter Unsustained Tachycardia Trial [MUSTT], MADIT I, MADIT II, Coronary Artery Bypass Graft [CABG] Patch Trial, and the Cardiomyopathy Trial [CAT]) and 5 additional RCTs: 1. Defibrillator in Acute Myocardial Infarction Trial (DINAMIT); 2. Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) 3. Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION); 4. Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation (DEFINITE); and 5. Amiodarone versus Implantable Defibrillator Randomized Trial (AMIOVIRT).

8 Subject: Implantable Cardioverter Defibrillator Page: 8 of 41 The 2004 TEC Assessment made the following conclusions about the use of ICD devices. The use of ICD devices meets the TEC criteria in the prevention of sudden death from VT in patients who have: Symptomatic (defined as the presence of dyspnea on exertion, angina, palpitations, or fatigue) ischemic dilated cardiomyopathy with a history of MI at least 40 days before ICD treatment and LVEF of 35% or less; or Symptomatic (defined as the presence of dyspnea on exertion, angina, palpitations, or fatigue) nonischemic dilated cardiomyopathy for more than 9 months duration and LVEF of 35% or less. The use of ICD devices does not meet the TEC criteria in the prevention of sudden death from ventricular tachyarrhythmia in patients who: have had an acute MI (i.e., less than 40 days before ICD treatment); have New York Heart Association (NYHA) class IV congestive heart failure (unless patient is eligible to receive a combination cardiac resynchronization therapy ICD device); have had cardiac revascularization procedure in past 3 months ( CABG or percutaneous transluminal coronary angioplasty [PTCA]) or are candidates for a cardiac revascularization procedure; or have noncardiac disease that would be associated with life expectancy less than 1 year. The 2004 TEC Assessment based its conclusions on the following indication-specific evidence. Patients Who Have Prior MI and Reduced LVEF. The 2002 Assessment concluded that the evidence was sufficient to demonstrate that ICD therapy improves net health outcome for primary prevention in patients with prior MI and reduced LVEF. Both new studies (SCD-HeFT and COMPANION) and the reanalysis of MUSTT findings provide additional supportive evidence of improved outcomes in patients with prior MI. The hazard ratio (HR) for all-cause mortality in the ischemic subgroup of SCD-HeFT was 0.79 (95% confidence interval [CI] 0.60 to 1.04), which is close to that observed in MADIT II (HR: 0.69, 95% CI: 0.51 to 0.93), and these findings provide additional supportive evidence that ICD therapy reduces mortality. There may be slight but not statistically significantly increased rates of adverse effects associated with ICD therapy; however, serious device-related events are not common. On balance, the significant reductions in mortality associated with ICD therapy outweigh the harms associated with ICD therapy in comparison to conventional treatment. Thus, the available evidence again demonstrates that ICD therapy improves health outcomes in patients with coronary artery disease and prior MI and reduced LVEF. Patients Who Have Acute MI and Reduced LVEF. The evidence reviewed in the 2004 TEC Assessment was insufficient to permit conclusions regarding the effect of ICD therapy as primary prevention on net health outcome for t patients with acute MI and reduced LVEF. Patients Who Have No Prior MI and Reduced LVEF (e.g., Nonischemic Dilated Cardiomyopathy, NIDCM). Results from subjects with nonischemic dilated cardiomyopathy included in SCD-HeFT and

9 Subject: Implantable Cardioverter Defibrillator Page: 9 of 41 DEFINITE suggest a mortality benefit from ICD therapy, although statistical significance that was not achieved in these studies was likely related to insufficient power. A meta-analysis (4) of 5 trials including nonischemic subjects reports a statistically significant reduction in mortality associated with ICD therapy. Furthermore, when the body of evidence for ICD therapy in both ischemic and nonischemic populations is considered together, the preponderance of evidence suggests that ICD therapy improves health outcomes compared with medical management alone with a relative risk reduction in all-cause mortality between 21% and 35%. While the risk of adverse events is not wellreported in studies of patients without prior MI, it seems reasonable to expect similar low rates of device-related adverse events as seen in studies of patients with prior MI. Subsequent Evidence and Guidelines for the Use of ICDs in Adults Relevant evidence and most current guidelines identified through Medline published following the 2004 TEC Assessment through October 2015 relates to the following subjects: Identification of predictors of better/worse outcomes after ICD placement. Use of ICD after acute MI: Reports of BEST-ICD (Beta-blocker Strategy + ICD), and IRIS trials Use of ICD in nonischemic dilated cardiomyopathy, with focus on implantation timing Use of ICD in hypertrophic cardiomyopathy In 2013, the Agency for Healthcare Research and Quality (AHRQ) Technology Assessment Program published a technology assessment on the evidence of ICDs for primary prevention of sudden cardiac death. (5) The review was structured around 3 questions: Key Question 1 examined the clinical effectiveness of the ICD versus no ICD, ICD with antitachycardia pacing (ATP) versus ICD alone, or ICD with CRT versus ICD alone, and differences among subgroups. Key Question 2 examined early and late adverse events and inappropriate shocks after ICD implantation, and differences among subgroups. Key Question 3 examined eligibility criteria and evaluation methods for patients included in comparative studies and the risk of SCD. The review included 14 studies comparing ICD with no ICD, 3 studies comparing combined ICD/CRT (CRT-D) with ICD, and 59 articles contributing data on adverse events after ICD implantation. Focusing on the evidence reported in the AHRQ report related to the efficacy of ICDs alone, a metaanalysis of 7 RCTs comparing ICD with control was associated with a summary hazard ratio (HR) of 0.69 (95% CI 0.60 to 0.79) for death (favoring ICD treatment). A meta-analysis of 5 studies comparing ICD with control was associated with a summary HR of 0.37 (95% CI 0.26 to 0.52) for reducing SCD. There was low-strength evidence that failed to show a consistent effect of ICDs on quality of life. One subgroup evaluated was time since MI. The authors conclude, An indirect comparison of ISIS and DINAMIT (which included patients with recent MIs, within 31 or 40 days) versus the remaining trials, suggests that patients with recent MIs may have no reduction in all-cause mortality (HR 1.05 [95% CI 0.86, 1.30]) than patients with more distant or no prior MIs (HR 0.69 [95% CI 0.60, 0.79]). By metaregression, the difference between IRIS and DINAMIT and the other seven RCTs is statistically significant (P = 0.012).

10 Subject: Implantable Cardioverter Defibrillator Page: 10 of 41 ICDs for Primary Prevention in Adults Post Myocardial Infarction BEST-ICD trial. The BEST-ICD (Beta-blocker Strategy + ICD) trial randomized 143 patients 5 30 days after acute MI to evaluate whether electrophysiology studies were useful to guide ICD placement and improve outcomes in patients at high risk of sudden death. (6) Entry criteria included an LVEF less than or equal to 35% along with 1 or more non-invasive risk factors (e.g., premature ventricular contractions, heart rate variability, SAECG [signal-averaged electrocardiography]-positive) and be given maximal tolerated beta-blocker (metoprolol) therapy. The authors concluded that using electrophysiology studies to guide ICD placement within 5 30 days after MI did not significantly improve outcomes and survival. This is consistent with the conclusions that ICD placement after early MI does not improve outcomes. The authors also noted that the study screened over 15,000 patients but ended after randomizing only 12% of the targeted study population, largely because there were far fewer patients with LVEF less than 35% than expected based on experience reported in the literature. IRIS Trial. The Immediate Risk Stratification Improves Survival (IRIS) trial evaluated ICD implantation early after myocardial infarction. (7) Eligible patients were required to have an LVEF 40% or less and either: 1) a heart rate 90 or more beats per minute on initial ECG, or 2) nonsustained ventricular tachycardia during Holter monitoring, or both. From 92 centers and 62,944 patients post-mi, 898 were randomized 5 to 31 days following the MI to ICD implantation or medical therapy. Seventy-seven percent had experienced ST elevation MI, of whom 72% underwent percutaneous transluminal coronary angioplasty (PTCA). During a mean 37-month follow-up, overall mortality was similar in the two arms (ICD vs. medical therapy, HR 1.04; 95% CI: 0.81 to 1.35). However, the risk of sudden cardiac death was lower following ICD (HR 0.52; 95% CI: 0.35 to 0.78), but non-sudden cardiac death risk was greater (HR 1.8; 95% CI: 1.0 to 3.2). These results are consistent with guidelines and previous trials. ICDs for Primary Prevention in Adults with High-Risk Hypertrophic Cardiomyopathy (HCM). Maron and colleagues (8) reported appropriate ICD discharge rates (terminating either ventricular tachycardia or fibrillation) from an international registry of HCM patients enrolled at 42 referral and nonreferral institutions. Between 1986 and 2003, ICDs were implanted in 506 patients with HCM 383 for primary prevention and 123 for secondary prevention. The mean age of patients was 42 years (standard deviation [SD]: 17), and 28% were 30 years of age or younger; 36% were female; mean follow-up was 3.7 years (SD: 2.8). Criteria considered in the study placing patients at high risk and, therefore, candidates for primary prevention included: 1) history of premature HCM-related sudden death in 1 or more first-degree relatives younger than 50 years of age; 2) left-ventricular hypertrophy greater than 30 mm; 3) one or more runs of nonsustained ventricular tachycardia at heart rates of 120 beats per minute or greater on 24-hour Holter monitoring; and 4) prior unexplained syncope inconsistent with neurocardiogenic origin. Abnormal exercise blood pressure was not reported. In the primary prevention group, appropriate discharges occurred at an annual rate of 3.6% (95% CI: 2.7% to 4.8%), in the secondary prevention group 10.6% (95% CI: 7.9% to 13.9%); respective 5-year cumulative probabilities of first appropriate discharge were 17% and 39%. If each appropriate discharge was life-saving, 5-year numbers needed to benefit (NNTBs) could be as low as 5.9 and 2.6 for primary and secondary prevention, respectively, when considering only the first appropriate discharge.

11 Subject: Implantable Cardioverter Defibrillator Page: 11 of 41 However, when analyzed in nonischemic dilated cardiomyopathy (NIDCM), Ellenbogen and colleagues 9) concluded that approximately one half of arrhythmias terminated by appropriate ICD discharges are not life-threatening. The NNTBs calculated in the Maron et al study, therefore, represent lower bounds or greatest potential benefit, and the true benefit is likely less (only 6.3% of primary prevention patients had more than one appropriate discharge). Adverse events rates included one or more inappropriate discharges (27%); infections (3.8%); hemorrhage or thrombosis (1.6%); lead fractures, dislodgement, and oversensing (6.7%). While the number of risk factors present was not associated with cumulative probability to first appropriate discharge for primary prevention, patient selection for ICD implantation was performed by experienced clinicians. These results, obtained outside the setting of a clinical trial, apply under such conditions. In 2015, Magnussen et al reported outcomes for 321 patients with HCM treated with an ICD enrolled in a Swedish registry. (10) Over a mean 5.4 years of follow-up, appropriate ICD discharges in response to ventricular tachycardia or fibrillation occurred in 77 patients (24%), corresponding to an annual rate of appropriate discharges of 5.3%. At least 1 inappropriate shock occurred in 46 patients (14.3%), corresponding to an annualized event rate of 3.0%. Ninety-two patients (28.7%) required at least 1 surgical intervention for an ICD-related complication, with a total of 150 ICD-related reinterventions. Most reinterventions (n=105, 70%) were related to lead dysfunction. ICDs for Primary Prevention in Adults with Nonischemic Dilated Cardiomyopathy. For patients with nonischemic cardiomyopathy, the optimal timing of ICD implantation remains uncertain. A substantial percent of patients diagnosed with nonischemic cardiomyopathy (NICM) will improve following initial diagnosis, even when a reversible cause of NICM cannot be identified. Given the current available evidence, it is not possible to predict which patients with idiopathic NICM will improve, nor is it possible to accurately estimate the time course for improvement. The specification of a 9-month waiting period prior to ICD implantation arises from the selection criteria of the CAT trial, (11) which restricted enrollment to patients with onset of NICM within 9 months. While the results of this trial did not show a benefit for patients with recent onset of NICM, the trial was stopped early due to an unexpectedly low rate of events and was thus underpowered to detect a difference in mortality between groups. Kadish and Subacius (12) performed a post-hoc analysis of the DEFINITE trial data to examine whether the time from diagnosis of nonischemic dilated cardiomyopathy (NIDCM) was associated with the magnitude of benefit from ICD implantation. Survival benefit was found only for those diagnosed less than 9 months prior to implantation (n=216); no benefit was apparent when NIDCM was diagnosed greater than 9 months prior (n=242). However, there was a significant discrepancy between arms in the time from diagnosis to randomization standard therapy patients were randomized a median of 20 months after diagnosis, while those in the ICD arm had a median of 8 months. The trial was neither designed nor powered to examine a time effect, and the analyses conflict with findings of the smaller (n=104) Cardiomyopathy (CAT) trial (11) reviewed in the 2002 TEC Assessment. Further evidence is necessary to define when in the natural history of the disease ICD implantation is appropriate. The Definite trial enrolled NICM patients without regard to time since onset, and a post-hoc analysis revealed that the benefit was found mainly in patients with onset of NICM for less than 9 months. Neither of these pieces of evidence represents strong data to support a specific time interval prior to implanting an ICD in patients with NICM.

12 Subject: Implantable Cardioverter Defibrillator Page: 12 of 41 Zecchin et al. (13) performed a cohort study on 503 consecutive patients diagnosed with idiopathic NICM to determine the extent to which indications for an ICD evolved over the several months following an initial NICM diagnosis. At initial diagnosis, 245 met Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) criteria for an ICD, based on an ejection fraction less than 35% and Class II-III heart failure, and 258 did not meet criteria for an ICD. At a mean follow-up of 5.4 months during which patients were treated with angiotensin-converting enzyme inhibitors and beta blockers, there were consistent improvements in ejection fraction and symptoms, such that less than one-third of evaluable patients (31%) still had indications for ICD. Of patients who initially did not have an indication for an ICD, a total of 10% developed indications for an ICD at follow-up. This study highlights the fact that a decision for ICD implantation should not be made prior to optimal treatment and stabilization of patients with newly diagnosed NICM, since the indications for ICD are not stable over time and will change in a substantial numbers of patients following treatment. A prospective registry sponsored by the National Heart, Lung and Blood Institute (NHLBI) enrolled 373 patients with recent-onset NICM, and compared mortality in patients receiving an early ICD with those receiving the device at a later time.(14) Forty-three patients received an ICD within 1 month of diagnosis, with a 1-year survival for this group of 97%. Three-hundred thirty patients received an ICD between 1 and 6 months, with a 1-year survival of 98%. Seventy-three patients received an ICD at a time period longer than 6 months, with a 1-year survival. Survival at 2 and 3 years was also similar between groups, with no significant differences. Some experts consider patients with recently diagnosed NICM and either sustained VT or unexplained syncope to be candidates for earlier ICD implantation due to their higher risk of lethal arrhythmias. However, evidence on this specific population is lacking, and the natural history of patients in this category is not well-characterized. The most recent ACC/AHA guidelines (15,16) do not specifically address the optimal waiting period prior to implantation of an ICD for patients with newly diagnosed Section Summary: ICD for Primary Prevention for in Adults A large body of RCTs has addressed the effectiveness of transvenous ICD implantation for primary prevention in patients at high risk of sudden cardiac death due to ischemic cardiomyopathy and nonischemic dilated cardiomyopathy. Evidence from several RCTs demonstrates improvements in outcomes with ICD treatment for patients with symptomatic heart failure due to ischemic or nonischemic cardiomyopathy with LVEF of 35% or less. The notable exception is that data from several RCTs, including the BEST-ICD and IRIS trials, and subanalyses from earlier RCTs, is that outcomes with ICD therapy do not appear to be improved for patients who are treated with an ICD within 40 days of an acute MI. Less evidence is available for the use of ICDs for primary prevention in patients with hypertrophic cardiomyopathy. In several cohort studies, the annual rate of appropriate ICD discharge ranged from 3.6% to 5.3%. Given the long-term high risk of patients with hypertrophic cardiomyopathy for SCD risk, with the assumption that appropriate shocks are life-saving, these rates are considered adequate evidence for the use of SCDs in patients with hypertrophic cardiomyopathy.

13 Subject: Implantable Cardioverter Defibrillator Page: 13 of 41 ICDs in Patients with Hereditary Arrhythmia Syndromes ICDs have been used for both primary and secondary prevention in patients with a number of hereditary disorders that predispose to ventricular arrhythmias and sudden cardiac death, including long QT syndrome (LQTS), Brugada syndrome (BrS), short QT syndrome (SQTS), and catecholaminergic polymorphic ventricular tachycardia (CPVT). Some of these conditions are extremely rare, but the use of ICDs has been described in small cohorts of patients with BrS, LQTS, and SQTS. Congenital Long QT Syndrome In 2010, Horner et al reported on outcomes for 51 patients with genetically confirmed LQTS treated with an ICD from 2000 to 2010 who were included in a single-center retrospective analysis of 459 patients with genetically confirmed LQTS. (17) Of the patients treated with ICDs, 43 (84%) received the device as primary prevention. Twelve patients (24%) received appropriate ventricular fibrillation or torsades de pointes-terminated ICD shocks. Factors associated with appropriate shocks included secondary prevention indications (P=0.008), QT corrected (QTc) duration greater than 500 ms (P=0.0008), non- LQT3 genotype (P=0.02), documented syncope (P=0.05), documented torsades de pointes (P=0.003), and a negative sudden family death history (P=0.0001). Inappropriate shocks were delivered in 15 patients (29%). Patients with the LQT3 genotype had only received inappropriate shocks. Brugada Syndrome Conte et al described outcomes for a cohort of 176 patients with spontaneous or drug-induced Brugada type 1 electrocardiographic findings who received an ICD at a single institution and were followed for at least 6 months. (18) Before ICD implantation, 14.2% of subjects had a history of aborted sudden cardiacdeath (SCD) due to sustained spontaneous ventricular arrhythmias, 59.7% had at least 1 episode of syncope, and 25.1% were asymptomatic. Over a mean follow up of 83.8 months, 30 patients (17%) had spontaneous sustained ventricular arrhythmias detected. Sustained ventricular arrhythmias were terminated by ICD shocks or antitachycardia pacing in 28 patients (15.9%) and 2 patients (1.1%), respectively. However, 33 patients (18.7%) experienced inappropriate shocks. Eight patients (4.5%) died during follow up, 3 of whom died of cardiac causes. Dores et al reported results of a Portuguese registry that included 55 patients with Brugada syndrome, 36 of whom were treated with ICDs for either primary or secondary prevention.(19) Before ICD implantation, 52.8% of subjects were asymptomatic, 30.6% had a history of syncope with suspected arrhythmic cause, and 16.7% had a history of aborted SCD. Over a mean follow up of 74 months, 7 patients experienced appropriate shocks, corresponding to an incidence of 19.4% and an annual event rate of 2.8%. In multivariable analysis, predictors of appropriate shocks were a history of aborted SCD (HR 7.87, 95% CI 1.27 to 49.6, P=0.027) and nonsustained ventricular tachycardia during follow up (HR 6.73, 95% CI 1.27 to 35.7, P=0.025). In data from a US cohort of 33 patients with Brugada syndrome treated with ICDs, Steven et al reported that 2/3 patients with a prior history of aborted SCD received appropriate shocks over a mean 7.9 years of follow up, while none of the 30 patients without a history of aborted SCD had an arrhythmia detected. (20) In a smaller registry that included 25 patients with Brugada syndrome treated with ICDs, over an

14 Subject: Implantable Cardioverter Defibrillator Page: 14 of 41 average follow up of 41.2 months, appropriate shocks were delivered in 3 patients, all of whom had prior cardiac arrest. (21) Catecholaminergic Polymorphic Ventricular Tachycardia Roses-Noguer et al reported results of a small retrospective study of 13 patients with CPVT who received an ICD. (22) The indication for ICD therapy was syncope despite maximal beta-blocker therapy in 6 patients (46%) and aborted SCD in 7 patients (54%). Over a median follow-up of 4.0 years, 10 patients (77%) received a median 4 shocks. For 96 shocks, 87 ECGs were available for review; of those, 63 (72%) were appropriate and 24 (28%) were inappropriate. Among appropriate shocks, 20 (32%) were effective in restoring sinus rhythm. Section Summary: ICD for Patients with Hereditary Arrhythmia Syndromes The evidence related to the use of ICDs in patients with hereditary arrhythmia syndromes includes primarily single-center cohort studies or registries of patients with LQTS, Brugada syndrome, and CPVT that report on appropriate shock rates. Patient populations typically include a mix of those requiring ICD implantation for primary prevention and those requiring it for secondary prevention. The limited available data for ICDs for LQTS and CPVT report high rates of appropriate shocks. For Brugada syndrome, more data are available and suggest that rates appropriate shocks are similarly high. Studies comparing outcomes between patients treated and untreated with ICDs are not available. However, given the relatively small patient populations and the high risk of cardiac arrhythmias, clinical trials are unlikely. Use of Automatic Transvenous ICDs in the Pediatric Population There is limited direct scientific evidence on the efficacy of ICDs in the pediatric population. The majority of published studies in this area are retrospective analyses of small case series. A review of some of the representative publications of this type is summarized below. The largest published series was a combined series of pediatric patients and patients with congenital heart disease from 4 clinical centers. (23) The median age of this population was 16 years, although some adults were included up to the age of 54 years. A total of 443 patients were included. The most common diagnoses were tetralogy of Fallot and hypertrophic cardiomyopathy. ICD implantation was performed for primary prevention in 52% of patients and for secondary prevention in 48%. Over a 2- year period of follow-up, appropriate shocks occurred in 26% of patients and inappropriate shocks occurred in 21%. Silka et al. (24) compiled a database of 125 pediatric patients treated with an ICD, through query of the manufacturers of commercially available devices. Indications for ICD placement were survivors of cardiac arrest in 95 patients (76%), drug-refractory ventricular tachycardia in 13 patients (10%), and syncope with heart disease plus inducible ventricular tachycardia in 13 patients (10%). During a mean follow-up of 31 +/- 23 months, 73 patients (59%) received at least one appropriate shock and 25 patients (20%) received at least one inappropriate shock. The actuarial rates of sudden-death-free survival were 97% at one year, 95% at 2 years, and 90% at 5 years.

15 Subject: Implantable Cardioverter Defibrillator Page: 15 of 41 Alexander et al. (25) reported on 90 ICD procedures in 76 young patients with a mean age of 16 years (range: 1 30). Indications for placement were 27 patients (36%) with cardiac arrest or sustained ventricular tachycardia, 40 patients (53%) with syncope, 17 patients (22%) with palpitations, 40 patients (53%) with spontaneous ventricular arrhythmias, and 36 patients (47%) with inducible ventricular tachycardia. Numerous patients had more than one indication for ICD in this study. Over a median of 2 years follow-up, 28% of patients received an appropriate shock, and 25% of patients received an inappropriate shock. Lewandowski et al. (226) reported on long-term follow-up of 63 patients between the ages of 6-21 years who were treated with an ICD device. After a 10-year follow-up, there were 13 (21%) patients with surgical infections. Fourteen patients (22%) experienced at least one appropriate shock and 17 patients (27%) had at least one inappropriate shock. Serious psychological sequelae developed in 27 patients (43%). Section Summary: ICDs in Pediatric Patients The available evidence for the use of ICDs in pediatric patients is limited and consists primarily of small case series that include patients with mixed indications for ICD placement. Overall, these studies report relatively high rates of appropriate shocks, but also high rates of inappropriate shocks. Pediatric patients may also be eligible ICD implantation if they have hereditary arrhythmias syndromes (see section, ICDs in Patients with Hereditary Arrhythmia Syndromes. ) Adverse Events Associated with Transvenous ICDs Perrson et al published a systematic review and meta-analysis of adverse events (AEs) following ICD implantation.(27) The authors included data from 35 cohort studies, reported in 53 articles. In-hospital serious AE rates ranged from 1.2% to 1.4%, most frequently pneumothorax (0.4%-0.5%) and cardiac arrest (0.3%). Post hospitalization complication rates were variable: device-related complications occurred in 0.1% to 6.4%; lead-related complications in 0.1% to 3.9%; infection in 0.2% to 3.7%; thrombosis in 0.2% to 2.9%; and inappropriate shock in 3% to 21%. In another systematic review and meta-analysis of AEs following ICD implantation, Ezzat et al compared rates of AEs reported in clinical trials of ICDs with those reported in the US National Cardiovascular Data Registry.(28) The review included 18 RCTs with a total of 6796 patients. In pooled analysis, the overall AE rate was 9.1% (95% CI 6.4 to 12.6%). Rates of access-related complications, lead-related complications, generator-related complications, and infection were 2.1% (91% CI 1.3 to 3.3%), 5.8% (95% CI 3.3 to 9.8%), 2.7% (95% CI 1.3 to 5.7%), and 1.5% (95% CI 0.8 to 2.6%), respectively. Complication rates in the RCTs were higher than those in the US registry, which reports only in-hospital complications (9.1% in the RCTs vs 3.08%, P<0.01). The overall complication rate was similar to that reported by Kirkfelt et al, in a population-based cohort study including all Danish patients who underwent a cardiac implantable electronic device procedure from (562/5918 patients [9.5%] with at least 1 complication).(29) In 2011, van Rees et al reported results of a systematic review of implantation-related complications in RCTs of ICDs and cardiac resynchronization therapy (CRT) devices. (30) The review included 18 trials and 3 subgroup analyses. Twelve trials assessed ICDs, 4 of which used both thoracotomy and nonthoracotomy ICDs (n=951) and 8 of which used nonthoracotomy ICDs (n=3,828). For

16 Subject: Implantable Cardioverter Defibrillator Page: 16 of 41 nonthoractomy ICD implantations, the rate of in-hospital and 30-day mortality was 0.2% and 0.6%, respectively, and pneumothorax was reported in 0.9% of cases. For thoracotomy ICD implantations, the average in-hospital mortality rate was 2.7%. For nonthoracotomy ICD implantations, the overall leaddislodgement rate was 1.8%. In a large retrospective single-center study including 1043 transvenous lead extraction procedures in 985 patients, Maytin et al reported a cumulative mortality rate of 2.1%, 4.2%, 8.4%, and 46.8% at 30 days, 3 months, 1 year, and 10 years post-procedure, respectively. (31) Most lead extractions were due to infection (50%) or lead malfunction (30%). Of the 21 patients with an initial extraction due to infection, 10 required another lead extraction procedure for infection. Lead extraction due to infection was associated with a significantly increased mortality risk. However, it is unclear whether this mortality risk is related to the ICD lead extraction or underlying patient morbidity. Lead Failure The failure of ICD leads in several specific ICD devices has lead the U.S. Food and Drug Administration (FDA) to require St. Jude Medical to conduct 3-year postmarket surveillance studies to address concerns related to premature insulation failure and to address important questions related to follow-up of affected patients.(32) A 2010 report found that 57 deaths and 48 serious cardiovascular injuries associated with device-assisted ICD or pacemaker lead extraction were reported to the FDA s Manufacturers and User Defined Experience (MAUDE) database.(33) Birnie et al reported clinical predictors of lead failure among 3169 Sprint Fidelis leads implanted from 2003 to 2007 at 11 centers selected from the 23 centers participating in the Canadian Heart Rhythm Society Device Committee. (34) A total of 251 lead failures occurred, corresponding to a lead failure rate at 5 years of 16.8%. Factors associated with higher rates of failure included female sex (HR 1.51, 95% CI 1.14 to 2.04, P=0.005), axillary vein access (HR 1.94, 95% CI 1.23 to 3.04), and subclavian vein access (HR 1.63, 95% CI 1.08 to 2.46). In a previous study from 3 centers reporting on predictors of Fidelis lead failures, compared with Quattro lead failures, Hauser et al reported a failure rate for the Fidelis lead of 2.81% per year (vs 0.42%/year for Quattro leads, P<0.0001). (35) In an earlier study from 12 Canadian centers, Gould et al reported outcomes from ICDs replacements due to ICD advisories from 2004 to 2005 which included 451 replacements (of 2635 advisory ICD devices).(36) Over 355 days of follow up, 41 complications (9.1%) occurred, including 27 (5.9%) requirements for reoperation and 2 deaths. In another multicenter study, Eckstein et al reported rates of lead failure among 1317 consecutive patients with an ICD implanted at 3 European centers from 1993 to (37) The endpoint of lead failure, defined if a case required surgical revision to correct the lead-related problem, occurred in 38 patients. Lead malfunction resulted in inappropriate ICD therapies in 29 of the failures (76%), while the remaining lead malfunctions were detected during routine follow-up. Over a median follow up of 3.1 years, among the 38 patients with lead malfunction, the cumulative incidences of lead failure recurrence were 4.4%, 14.1%, and 19.8% at years 2, 3, and 4, respectively. A study reporting on another single-center European registry which included 990 patients who underwent first implantation