The incidence of sudden cardiac death depends on the clinical characteristics of the studied population. While in the general population, SCD incidence is low (0.1-0.2%), it reaches 20-30% in high-risk subgroups, such as survivors of SCD, who are at a particular high-risk for recurrent events. However, when speaking in absolute numbers, SCD mostly occurs in patients who have not yet suffered arrhythmia or SCD. This has led to the investigation of ICD benefits in such subpopulations and to the publication of several major clinical trials in the last twelve years. A broadening of the indications for ICD implantation in the guidelines issued by the main scientific societies ensued.
1- Evidence from major clinical trials
The first major clinical trial on SCD primary-prevention was the MADIT trial (1996). This trial enrolled patients with prior myocardial infarction (three weeks or more before enrollment), non-sustained ventricular tachycardia, left ventricular dysfunction (LVEF≤35%), New York Heart Association (NYHA) functional class I-III and inducible sustained VT which was still inducible following procainamide infusion (1). One-hundred ninety six patients were enrolled (n=196), and were randomly assigned to receive an ICD (n=95) or conventional medical therapy including antiarrhythmic drugs (n=101). During an average follow-up of 27 months, ICD therapy was superior to medical therapy in terms of survival (HR of overall mortality 0.46, p=0.009). Nevertheless, it should be noted that the use of beta-blockers was low and it was twice as frequent in the ICD group compared to the control group (31% vs 14%).
CABG-PATCH Trial (1997)
(2) was published one year later. This trial assessed the benefit of ICD prophylactic implantation at the time of elective coronary bypass graft surgery. Patients were eligible if they had coronary heart disease scheduled for elective coronary bypass surgery, ventricular dysfunction (LVEF≤35%), and abnormalities on signal-averaged electrocardiogram. Four-hundred and fourty-six patients were randomised to the ICD group and 454 to the control group. At follow-up there were no significant differences in overall mortality (HR 1.07, p=0.64) and more postoperative infections were reported in the defibrillator group (30% vs 14%, p<0.05). On the other hand, more patients developed myocardial infarction (MI) at long-term follow-up in the control group (0.5% vs 4.2%, p<0.05). Interestingly, mortality rate (18%) in both groups was lower than that found in previous trials, such as MADIT (32%) and AVID (24%) trials. These latter differences could be related to several factors such as that CABG-PATCH patients may have had lower arrhythmic risk: abnormalities in signal-averaged electrocardiogram seemed to be a less strong risk indicator than the presence of unsustained ventricular tachycardias (in MADIT trial) or an arrhythmic event (in AVID trial). Other potential factors to explain such a low mortality could be related to all the patients undergoing complete revascularization and a subsequent decrease of ischemically triggered arrhythmias and antiarrhythmic drugs compared to the MADIT and AVID patients (3).
MUSTT trial (1999)
The MUSTT trial results (4) supported those found in the MADIT trial and led to changing the level of evidence of this indication in the ACC/AHA/NASPE 2002 Guidelines (5).
However, it should be noted that this was not a defibrillator trial but a trial of risk stratification by invasive electrophysiologic evaluation. The primary objective of the trial was to test if antiarrhythmic therapy guided by electrophysiologic evaluation decreased mortality compared to non-antiarrhythmic conventional empirical therapy. Inclusion criteria were ischemic heart disease with LVEF≤40% and asymptomatic non-sustained ventricular tachycardia. Patients in whom sustained ventricular tachyarrhythmia was induced were randomly assigned to receive either antiarrhythmic therapy with either drugs or ICD or non-antiarrhythmic therapy. At five-year follow-up arrhythmic mortality was significantly lower in the electrophsyiologically guided group than in the control group (25% vs 32%, RR 0.73, p=0.04). Interestingly, the lower arrhythmic mortality in the electrophysiologically guided therapy group was largely attributable to the use of ICD. Both arrhythmic and overall mortality was lower in the ICD subgroup than in those patients who were treated with guided antiarrhythmic therapy and who did not receive a defibrillator (9% vs 37% y 24% vs 51% respectively, p<0.001). Arrhythmic and overall mortality was similar among patients who were treated with antiarrhythmic drugs and those who received no therapy. Nevertheless, it should be noted that this was not a randomised comparison of ICD versus conventional therapy.
MADIT II (2002)
The third major trial was the MADIT II trial (6) . It was published in 2002, and showed a beneficial effect of ICD in two-year overall survival in patients who had had MI more than one month before enrollment, LVEF<30%, and functional class I-III. Patients were randomly assigned in a 3:2 ratio to receive ICD (742 patients) or conventional therapy (490 patients). Betablockers and angiotensin-converting-enzyme inhibitors were used in around 70% of the patients in both groups. Overall mortality was 14.2% in the ICD group and 19.8% in the conventional-therapy group (HR=0.69, p=0.016) at twenty month follow-up. This mortality difference is the smallest seen among all ICD trials with differences statistically in favour of the ICD group. It is remarkable that the two survival curves began to diverge after nine months and there was a slight increase in hospitalizations due to new or worsened heart failure in the ICD group. This latter finding could be explained by different factors: a longer survival of the ICD patients, the negative effect of backup ventricular pacing on hemodinamics, and ventricular function impairment due to the delivery of electrical shocks by the device.
DINAMIT trial (2004)
The DINAMIT trial (7) is the only SCD primary prevention trial during the acute phase of MI:. Patients were enrolled if they had recently had MI (6-40 days), functional class I-III, reduced left ventricular function (LVEF≤35%, mean LVEF 28%), and impaired cardiac autonomic function (manifested as depressed heart-rate variability or an elevated average 24-hour heart rate on Holter monitoring). They were randomized in a 1:1 ratio to receive an ICD (332 patients) or conventional therapy (342 patients). There were no significant differences in overall mortality (HR 1.08, p=0.66) between both groups. There was less arrhythmic mortality in the ICD group (HR 0.42, p=0.009), but this effect was balanced by an increase of non-arrhythmic mortality (HR 1.75, p=0.02) without significant mortality related to the surgical act of ICD implantation. Secondary analysis by DINAMIT investigators showed that this increase of non-arrhythmic mortality was confined to the subgroup of patients who had received shocks from the ICD. This suggested that ICD prevented SCD by terminating ventricular tachyarrhythmias in some of these patients who later will die due to pump failure without a net effect on overall survival. This is also supported by other observation such as the lack of ICD benefits during first months after implantation in the MADIT II trial.
The SCD-HeFT trial was published in 2005 and is the largest ICD study to date. It enrolled 2551 patients, all of them with left ventricular dysfunction (LVEF≤35%), functional class II/III (70 %/30%, respectively) and stable congestive heart failure due to ischemic (52%) and non-ischemic causes (48%). Patients were randomly assigned to conventional therapy plus placebo, conventional therapy plus amiodarone or conventional therapy plus single-lead ICD.
The device was programmed with an only zone of electrical shocks without antitachycardia pacing. Risk of death was similar in the amiodarone and placebo groups (HR 1.06, p=0.53) but significantly lower in the ICD group (HR 0.77, p=0.007). Subgroups analysis showed the greatest ICD benefit in class II patients (46% relative reduction and absolute reduction of 12% at five years) and no benefit in class III patients. Nevertheless, this evidence is not considered enough to exclude class III patients from ICD implantation because it was found in a subgroups analysis (8).
COMPANION trial (2004)
Finally, the COMPANION trial (2004), also supports the benefit of ICD in patients with ischemic heart disease. 1520 patients with advanced heart failure (class III or IV) due to ischemic or non-ischemic causes, LVEF≤35% and QRS complex≥120 ms were randomized in a 1:2:2 ratio to receive optimal conventional therapy alone or in combination with cardiac-resynchronization therapy with either a pacemaker or an ICD. The cardiac-resynchronization therapy group had significantly less mortality and hospitalizations than the other two groups. However, subanalysis of this population showed that only patients with ICD have significant reduction of overall mortality (36%, p=0.003), (9).
2 - Present practice guidelines recommendations from Scientific societies
The results of the previously mentioned clinical trials have been incorporated on the guidelines of practice of the main scientific societies, which have increased their indications for ICD implantation.
|ACC/AHA/ESC 2006 Guidelines for management of patients with ventricular arrhythmias and the prevention of sudden cardiac death (10)||ACC/AHA/HRS 2008 Guidelines for Device-Based Therapy of Cardiac Rhythm Abnormalities (11)|
Level of evidence A
|LVEF≤ 30-40% due to prior MI (≥40 days before) and NYHA II-II(*)|| - LVEF<35% due to prior MI (≥40 days post-MI) and NYHA II-III. (SCF-HeFT)
- LVEF<30% due to prior MI (≥40 days post-MI) and NYHA I (MADIT II)
Level of evidence B
|Non-sustained VT due to prior MI, LVEF<40%, and inducible VF/ VT at electrophysiological study (MADIT/MUSTT)|
Level of evidence B
|LVEF≤30-35% due to prior MI ( ≥40 days post-MI) and NYHA class I (**)|
(*) Supported in MADIT II and SCD-HeFT trials but LVEF was increased up to 40%
(**) Supported in MADIT II but LVEF was increased up to 35%
This broadening of indications for ICD implantation has resulted in an increase of cost which led to controversy about cost-effectiveness of this therapy in certain indications. Some national guidelines, such as the British guidelines from the NICE (The National Institute for Health and Clinical Excellence) only contemplates the ICD implantation in MADIT patients and in those MADIT II patients who in addition show a QRS complex width≥120 ms. The experts stated that a QRS complex width≥120 ms confers an additional risk for SCD which justify ICD implantation from results a cost-effectiveness prospective (12).
3 - Further patient selection
A post-hoc analysis of the MADIT II trial showed that not all patients have the same benefit from ICD implantation (13). A risk score was developed with the factors that were associated with a higher risk: functional class greater than II, atrial fibrillation, BUN >26 mg/dL, QRS>120 ms and age>70 years. An extremely high-risk profile was also identified in patients with BUN≥50 mg/dl and/or serum creatinine≥2.5 mg/dl. Differences in mortality between ICD and conventional therapy was more evident among patients with 1 or 2 risk factors, who accounted more than one half of the patient population (mortality at two years was 10% and 15% respectively in the ICD group vs 22% and 27% in the conventional therapy group). Mortality reduction in this group was around 60% (13). In contrast, ICD was not associated with mortality differences in those patients without risk factors, in those with more than 2 risk factors, and in the very high-risk subgroup.
Mortality reduction in patients with severe left ventricular dysfunction due to MI by ICD implantation is well supported by the results of major clinical trials. This has led to broader indications in guidelines, and therefore, to an increment in costs. Further risk stratification is warranted to reduce the health care cost of this therapy and to offer it to those patients who would benefit most.