Reduced left ventricular ejection fraction (LVEF) is the best available predictor of sudden cardiac death (SCD) in survivors of myocardial infarction (MI). Nearly 50% of all coronary artery disease-related death is SCD. Use of the implantable cardioverter defibrillator (ICD) reduced mortality by 31% over two years in the Multicenter Automatic Defibrillator Implantation Trial II (MADIT-II) of patients with LVEF ⩽30% and previous MI.1 However, the average time from MI to enrolment was 81 months and no survival benefit of ICD was found among patients who were enrolled 6 months or earlier after revascularisation in this trial.2 It is also suggested that another reason for the higher rate of event in the ICD trials is enrolment bias, resulting in the recruitment of patients at very high risk.3 Real-world effectiveness of ICD is not clear in all patients meeting the MADIT-II criteria, and this result may not be generalised.

Early reperfusion reduces short-term and long-term mortality in patients with acute MI.4 Primary percutaneous coronary intervention (PCI) is better than thrombolytic therapy, affording a significant reduction in short-term and long-term outcomes of death, non-fatal reinfarction and stroke.5 The aim of this study was to determine the incidence of SCD according to LVEF in survivors of MI in the primary PCI era.

METHODS

In all, 4122 consecutive patients with acute MI who were discharged alive were enrolled immediately in a Japanese multicentre observational prospective registered cohort study, the Heart Institute of Japan Acute Myocardial Infarction-II (HIJAMI-II) study. In participating medical centres, approximately 80% of patients with acute MI underwent early reperfusion therapy.6

Acute MI was defined as MI meeting two or more of the following criteria: (1) typical chest pain, (2) an increase in creatine kinase to twice the normal level or more and (3) a persisting elevation/decrease in the ST-segment or a change in T waves in two or more contiguous leads or development of a new abnormal Q wave or left bundle branch block on the electrocardiograms. An ST-elevation MI was defined as a persistent elevation in the ST segment (>0.1 mV at J points) in two or more contiguous leads, and non-ST-elevation MI was defined as a change in the T waves in two or more contiguous leads or new development of left bundle branch block without ST-segment elevation.

Written informed consent was obtained from all patients. The protocol was approved by the institutional review boards of the participating medical centres.

Follow-up

After discharge, patients were seen in the outpatient clinic of the participating medical centre at 6-monthly intervals. Information about deceased patients was obtained from family members, their general practitioners, the hospitals to which they had been admitted and the resident registers. All causes of death were obtained from medical records, death certificates or pathological reports.

The primary end-point was SCD, and a secondary end-point was death from any cause. Sudden death was defined as unexpected, endogenous death within 24 hours after last having been seen alive, not related to a specific cause of circulatory failure.

Determination of LVEF

LVEF was calculated from the results of ventriculography, echocardiography or radionuclide ventriculography, performed before hospital discharge. All LVEF determinations were done by at least two independent investigators blinded to the patients’ data.

Statistical analysis

Data are presented as mean (SD) and are medians with interquartile ranges or frequencies. One-way analysis of variance (ANOVA) was used to compare groups with respect to normally distributed continuous variables, and the Kruskal-Wallis test was used for other variables. The χ² test test was used to compare nominally scaled variables. Cumulative probabilities of SCD and death from any cause were estimated with the Kaplan-Meier method from the enrolment, and by means of comparison of cumulative events based on LVEF with a log-rank test. To evaluate the influence of LVEF with respect to subsequent death events, the Cox proportional hazards models were used. The proportional hazards assumption was confirmed by the log (− log survival function). Two-tailed p values of <0.05 were considered to indicate a statistically significant difference.

RESULTS

Of 4122 infarct survivors, 77.8% received PCI and 3.7% received coronary artery bypass graft (CABG) surgery during their hospitalisation. During an average follow-up of 4.1 years, overall mortality was 13.1% and SCD was 1.2%. Patient distribution was 82.9% with LVEF >40%, 12.3% with LVEF ⩽40% and >30%, and 4.8% with LVEF ⩽30%.

Baseline characteristics of patients are shown in table 1. Body mass index, systolic and diastolic blood pressure, serum triglyceride and estimated glomerular filtration rate were lower, and heart rate, blood HbA_1c, serum C-reactive protein, and serum creatinine were higher in patients with LVEF ⩽30%. Incidences of complicated diabetes mellitus, dialysis, prior MI, prior PCI and prior CABG were higher in this group of patients. As regards medications at discharge, angiotensin-converting enzyme inhibitor, β-blocker and aspirin were less frequent, but nitrate and warfarin were more frequent in this group.

Kaplan-Meier curves for SCD-free status in the three groups are shown in figure 1. There was significantly higher incidence with lower LVEF . Patients with LVEF ⩽30% and LVEF ⩽40% and >30% were at increased risk for SCD (HR 5.99, 95% CI 2.73 to 13.14, p<0.001, HR 3.37, 95% CI 1.74 to 6.50, p<0.001, respectively) compared to patients with LVEF >40%.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Kaplan-Meier curve for 5-year sudden cardiac death in survivors of myocardial infarction with LVEF >40%, LVEF ⩽40% and >30%, and LVEF ⩽30%. LVEF, left ventricular ejection fraction.

Kaplan-Meier curves for death from any cause in the three groups are shown in figure 2. There was also a significantly higher incidence with lower LVEF . Patients with LVEF ⩽30% and LVEF ⩽40% and >30% were at increased risk for mortality (HR 3.85, 95% CI 2.96 to 5.00, p<0.001, HR 2.06, 95% CI 1.66 to 2.57, p<0.001, respectively) compared to patients with LVEF >40%. Causes of death are shown in table 2. There was a higher incidence of death due to heart failure and non-cardiovascular death, as well as SCD, with lower LVEF .

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Kaplan-Meier curve for 5-year mortality in survivors of myocardial infarction with LVEF >40%, LVEF ⩽40% and >30%, and LVEF ⩽30%. LVEF, left ventricular ejection fraction.

DISCUSSION

Our study showed that the frequency of patients with LVEF ⩽40% was 17.1%, and 4.8% had LVEF ⩽30% (meeting the MADIT-II criteria) among all registered MI patients. Reports from the national registry of MI in the United States showed that patients with low LVEF, of ⩽40%, accounted for approximately 23% of all MI patients, among whom more than half did not receive reperfusion.7 Results from the databases of the two largest multicentre, randomised AMI trials of primary PCI, the Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications (CADILLAC) and the Stent-Primary Angioplasty in Myocardial Infarction (Stent-PAMI) trials in the United States and Europe, showed that the frequency of LVEF <40% was 20.5% in 2982 patients with acute MI treated with primary PCI.8 A Japanese registered observational study in 2211 patients with acute MI undergoing PCI found that the frequency of patients with LVEF ⩽40% was 11.1%.9 Furthermore, a recent registered observational study (Zwolle Myocardial Infarction Study) showed that the frequency of patients with LVEF ⩽30% was 13.4% in 2544 consecutive patients with acute ST-elevation MI who survived ⩾30 days after primary PCI in The Netherlands.10 Although we cannot simply compare these results, because patients’ characteristics or inclusion/exclusion criteria were not the same, the frequency of surviving Japanese MI patients with LVEF ⩽30% seems to be lower than that in Western populations.

There was a tendency towards a higher incidence of SCD and total death in patients with lower LVEF. LVEF remains the single most powerful predictor of survival in the primary PCI era.8 Kaplan-Meier estimates of SCD in our patients with LVEF ⩽30% were 2.9%, 5.1% and 5.1% at 1, 3 and 5 years, respectively. These values of SCD probability are considerably lower than those for MI patients with LVEF ⩽30% in previous studies: MADIT-II (conventional arm: 12.1% at 2 years)11 and TRAndolapril Cardiac Evaluation (TRACE) study (15.5% at 3 years).12 A large-scale trial in 14 609 patients who had acute MI (between 0.5 and 10 days previously), the Valsartan in Acute Myocardial Infarction Trial (VALIANT), showed Kaplan-Meier estimates of SCD or cardiac arrest with resuscitation in 3852 patients with LVEF ⩽30% were approximately 8% at 1 year and 10% at 2 years. However, only 30% of them received reperfusion therapy.13 In the primary PCI era, the Zwolle Myocardial Infarction Study reported only eight SCD (2.3%) within 1 year among 342 surviving MI patients after primary PCI with LVEF ⩽30%.10 Tanno and colleagues reported only two SCD among 90 patients (77% PCI and 10% CABG) with prior MI and LVEF ⩽30% during the mean follow-up period of 37 months in a Japanese single centre retrospective study.14 These results are comparable with ours, and emphasise the importance of achieving optimal revascularisation to reduce SCD.

Primary PCI in acute MI is now central to optimal ST-elevation MI treatment, reducing infarct size, minimising myocardial damage, preserving LV function and decreasing morbidity and mortality.15 However, the reason for a reduction in SCD incidence in MI patients in the primary PCI era is not clear. Several factors are thought to contribute directly or indirectly to this beneficial effect. First, myocardial salvage by early revascularisation after acute MI can prevent the LV remodelling process and may inhibit the development of arrhythmia. Second, early revascularisation may contribute to an improvement in LV function and a reduction of infarct size late after acute MI, and this improvement will be associated with better survival (lower SCD).16 17 Third, revascularisation improves coronary circulation and reduces ischaemic events, which may inhibit triggering of arrhythmia and prevent activation of neurohormonal systems.

Our study demonstrated that overall mortality in patients with LVEF ⩽30% was similar to or higher than the levels found in MADIT-II, Tanno’s report and the Zwolle Myocardial Infarction Study.1 10 14 Our subjects had higher mean age than the patients in those studies, and this might be expected to result in higher mortality, including cardiovascular and non-cardiovascular deaths. Our patients with LVEF ⩽30% had significantly higher rates of complicated diabetes mellitus, renal insufficiency, dialysis, prior MI, prior revascularisation and warfarin use, and were older. Therefore, this group included more patients who had high risk of stroke, peripheral embolisation, heart failure and bleeding, or who had more severe clinical conditions compared with other groups. Furthermore, the proportions of patients using β-blocker, ACE inhibitors/angiotensin II receptor blockers and statin, which are effective in reducing the incidence of post-MI death, were lower in our study. These factors might have had some influence on the cardiovascular death rate in our patients with LVEF ⩽30%. Data from the placebo arms of four randomised trials showed that all deaths in MI patients with LVEF ⩽10% were non-arrhythmic, although LVEF significantly predicted mortality. This result suggested that consideration of LVEF alone might be inappropriate in selecting patients for ICD therapy.18 Notwithstanding the results of ICD trials, the optimal stratification of patients at risk for SCD is still open to debate.19

We could not clarify from our study which patients should receive ICD, and at what time following MI or revascularisation. The Defibrillator in Acute Myocardial Infarction Trial (DINAMIT) failed to show a survival benefit of early prophylactic ICD implantation in patients who had acute MI and low LVEF, although ICD therapy reduced arrhythmic death.20 The Coronary Artery Bypass Graft (CABG) Patch Trial and MADIT-II substudy showed no survival benefit of ICD therapy when implanted early after revascularisation in patients with ischaemic LV dysfunction.2 21 However, the MADIT-II substudy suggested a significant survival benefit in patients receiving ICD implantation more than 6 months after revascularisation.2 Regarding LV remodelling, patients with acute MI show an improvement in LV function and a reduction in infarct size within 6 months after revascularisation.17 To select patients for ICD therapy for primary prevention of SCD, it may be useful to assess LVEF at more than 6 months after acute MI or revascularisation and to add risk stratification to low LVEF. Further investigation is needed.

There were some limitations in this study. First, the number of subjects, especially those with LVEF ⩽30%, was relatively small. Therefore, subgroup analysis was not feasible. Second, longitudinal measurement of LVEF was not performed to evaluate the improvement in LV function after MI. Third, treatment bias existed because this was an observational study.

CONCLUSION

Although LVEF is a predictor of increased risk for SCD, there is a low incidence of SCD in survivors of MI in the primary PCI era. It appears that use of the ICD makes only a minor contribution to decreasing the overall mortality in MI survivors with low LVEF.