OUP user menu

Time-dependent benefit of preventive cardiac resynchronization therapy after myocardial infarction

Alon Barsheshet, Arthur J. Moss, Michael Eldar, David T. Huang, W. Jackson Hall, Helmut U. Klein, Scott McNitt, Jonathan S. Steinberg, David J. Wilber, Wojciech Zareba, Ilan Goldenberg
DOI: http://dx.doi.org/10.1093/eurheartj/ehq392 1614-1621 First published online: 12 November 2010

Abstract

Aims Cardiac remodelling is a progressive process after myocardial infarction (MI). However, currently there are no data regarding the effect of elapsed time from MI on the benefit of cardiac resynchronization therapy with defibrillator (CRT-D). The present study was designed to evaluate the relationship between elapsed time from MI and the benefit of preventive CRT-D therapy in patients with ischaemic cardiomyopathy (ICM).

Methods and results The risk of heart failure (HF) or death as a function of elapsed time from MI to enrolment, by treatment with CRT-D vs. implantable cardioverter defibrillator (ICD)-only therapy, was assessed among 704 ICM patients with a documented MI enrolled in MADIT-CRT, and separately in a subset of ICM patients without a documented prior MI (n = 237). In ICD patients, the adjusted risk of HF or death increased by 4% (P = 0.01) for each year elapsed from MI. Multivariate analysis demonstrated that patients with remote MI [categorized at the median value (≥8 years)] derived a significantly greater benefit from CRT-D [HR = 0.42 (P < 0.001)] than those with a more recent MI [HR = 1.26 (P = 0.35); P-value for interaction <0.001]. Consistently, the benefit of CRT-D was directly related to increasing quartiles of elapsed time from MI [Q1 (<3 years): HR = 1.67; P = 0.20, Q2 (3–8 years): HR = 1.12; P = 0.71, Q3 (8–15 years): HR = 0.47; P = 0.02, and Q4 (≥15 years): HR = 0.38; P = 0.001]. The ICM subgroup with no documented MI also derived enhanced benefit from CRT-D (HR = 0.43; P = 0.003).

Conclusion In patients with ischaemic cardiomyopathy, the risk of HF or death and the magnitude of CRT-D benefit are directly related to elapsed time from MI.

  • Heart failure
  • Cardiac resynchronization therapy
  • Myocardial infarction

See page 1580 for the editorial comment on this article (doi:10.1093/eurheartj/ehr016)

Introduction

Left ventricle (LV) adverse remodelling after myocardial infarction (MI) has been shown to be a chronic progressive process that continues for years after the initial insult.13 Adverse remodelling and gradual dilation leading to increased LV volumes after MI are major predictors of poor outcomes among heart failure (HF) patients.4 Accordingly, several therapies were found to slow the process of adverse LV remodelling, including blockers of the renin–angiotensin–aldosterone and adrenergic systems.1,2,57 In addition, a progressive adverse remodelling has been described among patients with ischaemic cardiomyopathy (ICM) without a documented prior MI.810

In MADIT-CRT, cardiac resynchronization therapy with a defibrillator (CRT-D) reduced the risk of HF or death in mildly symptomatic ICM and non-ICM patients when compared with implantable cardioverter defibrillator (ICD) treated patients. The study also showed that CRT-D therapy was associated with reverse remodelling and significant improvements in LV volumes in this population.11

We hypothesized that in patients with ICM: (i) the clinical benefit from CRT-D therapy would be directly related to the length of time elapsed from MI to device implantation; and (ii) patients with ICM but without a prior MI would also benefit from CRT-D therapy.

Methods

Study population

The design and results of the MADIT-CRT study have been reported previously.11 Briefly, 1820 patients who had ICM [NYHA class I or II (n = 1046)] or non-ICM [NYHA class II (n = 774)], an ejection fraction of ≤0.30, and prolonged intraventricular conduction with QRS ≥ 130 ms were randomized to receive CRT-D or ICD therapy in a 3:2 ratio. Screened patients were excluded from enrolment if they had existing indication for CRT, NYHA class III/IV in the past 90 days before enrolment, implanted pacemaker, coronary artery bypass graft surgery, percutaneous coronary intervention, or MI within the past 90 days before enrolment. The primary endpoint was all-cause mortality or HF event, whichever occurred first.

Of the 1046 patients with ICM enrolled in MADIT-CRT, 105 patients were excluded due to insufficient data regarding the occurrence of prior MI or the date in which the event occurred. Thus, the final study population comprised 941 ICM patients, of whom 704 (75%) had documented prior MI and 237 (25%) had documented coronary artery disease without documented MI.

Definitions and outcome measures

For the diagnosis of ICM, study patients had to have one or more of the following: (i) a documented prior MI (Q-wave or enzyme-positive); (ii) a history of a coronary revascularization procedure [one or more prior coronary artery bypass graft surgeries or percutaneous coronary interventions (balloon and/or stent angioplasty)]; (iii) documented significant coronary artery disease at coronary angiography and history of angina pectoris or other coronary-related symptoms or signs such that the enrolling physician assessed the patient as having ICM. Data regarding ICM, the number of prior MI events, and the date of last MI were collected. For the present analysis, time from MI was defined as the time in years from the date of the most recent MI to enrolment (i.e. device implantation), and was assessed both as a continuous measure (per year increment) and as a categorical variable (dichotomized at the median value and into approximate quartiles).

The primary endpoint of the current study was the occurrence of a HF event or death, whichever occurred first. Patients were categorized to have a HF event when they had symptoms and/or signs consistent with congestive HF and (i) received decongestive therapy (IV diuretics, IV nesiritide, IV inotropes) in an out-patient setting, or (ii) received an augmented HF regimen with oral or intravenous medications during an in-hospital stay.

Two-dimensional echocardiography was performed at baseline and at the 1-year follow-up as previously reported11 to assess changes in the LV volumes, left atrial volume, and LV ejection fraction. Volumes were measured according to Simpson's method, and the ejection fraction was calculated using a standardized protocol.12 Paired echocardiographic data were available for 392 CRT-D and 312 ICD patients.

Statistical analysis

Characteristics of patients categorized by history of MI and quartiles of time from MI were compared with one-way ANOVA or Chi-square tests, and P for trend was calculated using regression analysis or Mantel–Haenszel test, as appropriate. The probability of HF or death by time from MI and by treatment group was graphically displayed according to the method of Kaplan and Meier, with comparison of cumulative events by the log-rank test. Multivariate analysis was carried out using Cox proportional hazards regression modelling. Pre-specified covariates in the multivariate model included: age at enrolment (assessed as a continuous measure), gender, NYHA functional class, EF < 25%, QRS ≥ 150 ms, diabetes mellitus, and reduced estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m2. The risk factor thresholds for categorization into lower and higher risk subsets were pre-specified with the use of clinical and laboratory accepted criteria and those that were employed in the primary MADIT-CRT analysis. The effect of elapsed time from MI on outcome among ICD-only patients and the benefit of CRT-D when compared with ICD therapy by elapsed time from MI were assessed by including a treatment-by-time-from-MI interaction term to the multivariate models. A significant interaction implies that the adjusted HR for CRT-D vs. ICD effect among patients who had remote MI is significantly different from the adjusted HR for CRT-D vs. ICD effect among patients who had recent MI. The subset of study patients who did not have a documented prior MI was evaluated by including a no-MI subcategory in the multivariate models that assessed time from MI as a categorical variable. In order to confirm the consistency of the results, time from MI was categorized in several ways (divided by quartiles, median, 2-year intervals or by log scale) and a further adjustment was made for left bundle branch block in addition to the pre-specified covariates that included QRS ≥ 150 ms or for baseline LV volumes instead of baseline EF. Left ventricular volume changes [(1 year volume − baseline volume)/baseline volume] were compared between patients with documented MI and patients without documented prior MI using a two-tailed t-test. Analyses were conducted with SAS software (version 9.2, SAS institute, Cary, NC, USA).

Results

Patient characteristics

Study patients exhibited a wide distribution of time-periods from MI to enrolment (Figure 1) ranging from 3 months to 38 years (median: 8 years; interquartile range: 3–15 years). The baseline characteristics of study patients by time from MI and of patients without a documented prior MI are shown in Table 1. Patients in the higher time-from-MI quartiles were older and exhibited larger LV volumes when compared with patients in the lower quartiles, but had a lower frequency of diabetes mellitus. The ICM subgroup without a documented prior MI had a more advanced NYHA class, higher systolic blood pressure, wider QRS duration, and higher frequency of left bundle branch block when compared with MI patients. Notably, baseline LV volumes among ICM patients without a documented prior MI were similar to those among patients in the higher time-from-MI quartiles.

View this table:
Table 1

Patient characteristics by quartiles of time from myocardial infarction and no myocardial infarction

MI patients by quartiles of time since MIAll study patients
Q1 (<3 years), n = 176Q2 (3–8 years), n = 160Q3 (8–15 years), n = 190Q4 (≥15 years), n = 178Documented MI, n = 704No documented MI, n = 237
CRT-D (%)67.260.457.457.4*60.460.8
Age (years)65.0 ± 9.865.5 ± 10.166.3 ± 8.869.9 ± 7.5**66.7 ± 9.367.8 ± 9.2
Female sex (%)14.916.49.69.1*12.416.9
Diabetes (%)43.731.630.929.5**33.735.0
Hypertension (%)74.766.766.862.5*67.774.2
Smoking (%)17.411.413.412.113.510.9
NYHA functional class II (%)74.772.368.671.071.781.4††
Systolic blood pressure (mmHg)122.7 ± 17.7122.4 ± 19.1121.1 ± 17.3122.6 ± 17.1122.2 ± 17.7125.9 ± 17.5††
eGFR(mL/min/1.73 m2)67.2 ± 19.568.2 ± 20.665.5 ± 20.664.9 ± 17.166.3 ± 19.668.1 ± 18.4
eGFR < 60 mL/min/1.73 m2 (%)34.936.740.140.038.235.0
QRS (ms)151.8 ± 15.9153.2 ± 18.3151.5 ± 17.6154.9 ± 18.8152.9 ± 17.8158.9 ± 20.0††
QRS ≥ 150 ms (%)54.054.747.958.053.765.0††
LBBB (%)56.948.450.353.752.472.2††
LVEF%24.8 ± 4.925.0 ± 4.824.3 ± 4.924.3 ± 4.824.6 ± 4.924.1 ± 5.2
LVEF < 25% (%)28.728.933.538.6*32.735.9
LVESV/BSA (mL/m2)82.9 ± 17.184.2 ± 17.488.1 ± 22.386.1 ± 18.2*85.5 ± 19.088.5 ± 22.6
LVEDV/BSA (mL/m2)116.4 ± 20.8118.2 ± 21.7122.9 ± 28.0121.6 ± 23.1*119.9 ± 23.8123.9 ± 27.7
ACEI/ARB (%)92.591.896.393.293.696.2
Beta-blockers (%)92.594.388.392.091.891.6
Diuretics (%)74.772.373.472.273.473.8
  • Values are percent of patients or mean ± SD.

  • ACEI/ARB, angiotensin converting enzyme inhibitor or angiotensin receptor blocker; BSA, body surface area; eGFR, estimated glomerular filtration rate; LBBB, left bundle branch block; LVEF, left ventricle ejection fraction; LVEDV, left ventricle end-diastolic volume; LVESV, left ventricle end-systolic volume; MI, myocardial infarction; NYHA, New York Heart Association.

  • *P < 0.1 and **P < 0.01 for trend among quartiles of time after MI.

  • P < 0.1 and ††P < 0.01 for comparison between documented and no documented MI.

Figure 1

Distribution of time periods from myocardial infarction to enrolment. Median (interquartile range): 8 (3–15) years.

There were 237 patients with no documented MI; 180 had a history of coronary revascularization (CABG or PCI) and 57 patients (24%) had documented significant coronary artery disease at coronary angiography and coronary-related symptoms or signs such that the enrolling physician assessed the patient as having ICM (but did not undergo revascularization).

Out of the 941 patients included in the current study, 225 patients (23.9%) experienced the primary endpoint of HF or death, 82 patients (8.7%) died, 172 patients (18.3%) were hospitalized due to HF, and 17 patients (1.8%) experienced an HF event treated in an outpatient setting.

Effect of elapsed time from myocardial infarction on outcome among patients allocated to implantable cardioverter defibrillator-only therapy

Patients allocated to ICD-only therapy exhibited significantly higher rates of HF or death with increasing time from MI (Figure 2). Similarly, multivariate analysis demonstrated a 63% (P = 0.058) and >three-fold increase (P = 0.006) in the risk for HF or death among ICD-allocated patients who experienced prior MI above the median and in the upper quartile time-from-MI periods, respectively, when compared with ICD-allocated patients in the respective lower time-from-MI categories (Table 2). The subgroup of ICM patients without documented prior MI displayed similar event rates as patients in upper time-from-MI subgroup (Figure 2), and consistently, had a similar adjusted risk of HF or death as the upper quartile time-from-MI subgroup (Table 2).

View this table:
Table 2

Multivariate analysis: risk of heart failure or death in patients with implantable cardioverter defibrillator only by elapsed time from myocardial infarction and among ischaemic cardiomyopathy patients without a documented prior myocardial infarction

Hazard ratio95% CIP-value
Time from MI as a continuous measure
 Per 1-year increment1.041.01–1.070.013
Time from MI dichotomized at median
 ≥8 years vs. <8 years1.630.98–2.700.058
Time from MI categorized by quartile
 Q1Reference
 Q22.120.91–4.950.082
 Q32.080.91–4.790.084
 Q43.101.38–6.940.006
Ischaemic patients: no documented MI
 No MI vs. MI1.310.83–2.060.251
 No MI vs. MI < 8 years1.650.94–2.900.081
 No MI vs. Q1 MI2.791.20–6.490.017
  • Adjusted for age at enrolment, gender, NYHA functional class, EF < 25%, QRS ≥ 150 ms, diabetes mellitus, and eGFR < 60 mL/min/1.73 m2.

Figure 2

Two-year Kaplan–Meier estimates of heart failure/death events in the two treatment groups by quartiles of time since myocardial infarction. Q1 denotes <3 years since MI; Q2, 3–8 years; Q3, 8–15 years; Q4 ≥ 15 years.

Relation between elapsed time from myocardial infarction and the benefit of cardiac resynchronization therapy with defibrillator therapy

Treatment with CRT-D was associated with a significantly lower probability of HF or death at 2 years of follow-up (16%) when compared with ICD-only therapy (25%; P = 0.002) among all 941 ICM patients in the study. Subgroup analysis demonstrated that 2-year event rate reduction with CRT-D when compared with ICD-only therapy was pronounced among patients who experienced more remote MI [10 vs. 26%, respectively; P < 0.001 (Figure 3B)] and among ICM patients without a documented prior MI [16 vs. 31%, respectively; P = 0.002 (Figure 3C)], whereas among patients with a more recent MI, event rates in the CRT-D and ICD-only groups were similar [22 vs. 17%, respectively; P = 0.188 (Figure 3A)]. Similarly, quartile analysis showed a direct relationship between event-rate reduction associated with CRT-D therapy and elapsed time from MI (Figure 2).

Figure 3

Kaplan–Meier estimates of probability of death/heart failure event by treatment group among patients enrolled (A) less than median time (8 years) since myocardial infarction, (B) more than median time (8 years) since MI, or (C) no documented MI.

Multivariate analysis showed that CRT-D therapy was associated with an overall 39% reduction (P = 0.001) in the risk of HF or death among ICM patients (Table 3A). The benefit of CRT-D therapy was evident among ICM patients who experienced prior MI [32% risk-reduction (P = 0.025)] and pronounced in the ICM subgroup of patients who did not have a documented prior MI [57% risk-reduction (P = 0.003)]. Similar to the Kaplan–Meier survival analysis, multivariate analysis showed that the benefit of CRT-D therapy was directly related to elapsed time from MI. Thus, among patients who experienced MI ≥8 years prior to enrolment, CRT-D therapy was associated with a pronounced 58% (P < 0.001) reduction in the risk of HF or death, whereas among patients with a more recent MI, CRT-D therapy benefit was significantly attenuated [HR = 1.26 (P = 0.345); P-value for interaction (i.e. P-value for the difference between the hazard ratios) <0.001 (Table 3B)]. Consistently, quartile analysis (Table 3C) showed a pronounced benefit with CRT-D therapy among the upper time-from-MI quartile (Q4 HR = 0.38, P = 0.001 and Q3 HR = 0.47, P = 0.023) and no significant benefit in the lower quartiles (Q2 HR = 1.12 and Q1 HR = 1.67, P for both = ns). Similar results were found when time from MI was categorized differently (e.g. by log scale or by 2-year increment, data not shown) or when further adjustment was made for left bundle branch block in addition to the pre-specified covariates that included QRS ≥ 150 ms.

View this table:
Table 3

Multivariate analysis: cardiac resynchronization therapy with defibrillator vs. implantable cardioverter defibrillator-only risk of heart failure or death by elapsed time from myocardial infarction and among ischaemic cardiomyopathy patients without a known prior myocardial infarction

Hazard ratio95% CIP-value
(A) All ischaemic patients0.610.46–0.820.001
 Documented MI0.680.49–0.970.025
 No documented MI0.430.26–0.730.003
(B) Time from MI dichotomized at median
 <8 years1.260.78–2.030.345
 ≥8 years*0.420.27–0.65<0.001
(C) Time from MI categorized by quartile
 Q11.670.76–3.690.203
 Q21.120.61–2.070.711
 Q30.470.24–0.900.023
 Q40.380.21–0.690.001
  • Adjusted for age at enrolment, gender, NYHA functional class, EF < 25%. QRS ≥ 150 ms, diabetes mellitus, and eGFR < 60 mL/min/1.73 m2.

  • *P for interaction (with MI < 8 years) <0.001.

  • P for interaction (with Q1) <0.01.

Consistent with the results showing that CRT-D therapy was associated with a pronounced 57% reduction in the risk of HF or death among ICM patients without a documented prior MI, resynchronization therapy was associated in this subgroup with significant reverse remodelling effects, including improvements in LV ejection fraction, LV end-systolic, LV-end diastolic and left atrial volumes, which were significantly greater than in the subgroup of ICM patients with documented prior MI (Table 4).

View this table:
Table 4

Cardiac resynchronization therapy with defibrillator effects on reverse remodelling at 1 year in the myocardial infarction subgroup and no documented myocardial infarction subgroup

Ischaemic cardiomyopathy patientsP-value
Documented MI (n = 288)No documented MI (n = 104)
▵LVEF (%)9.4 ± 5.011.2 ± 4.90.002
▵LVEDV/baseline volume (%)−17.3 ± 10.2−19.9 ± 9.60.027
▵LVESV/baseline volume (%)−27.7 ± 14.7−32.0 ± 13.20.009
▵Left atrial volume/baseline volume (%)−24.3 ± 13.2−27.0 ± 11.90.054
  • Data are presented as mean ± SD.

  • ▵LVEF = 1 year LVEF (%) − baseline LVEF (%).

  • ▵Volume/baseline volume = (1 year volume − baseline volume)/ baseline volume.

  • LVEF, left ventricle ejection fraction; LVEDV, left ventricle end-diastolic volume; LVESV, left ventricle end-systolic volume.

Discussion

The present study is the first to assess the time-dependent effect of elapsed time from MI on the benefit of preventive CRT-D therapy among ICM patients with no or only mild symptoms of HF. We have shown that: (i) among ICM patients who do not receive preventive CRT-D therapy there is a direct correlation between elapsed time from MI and the risk of major HF events or death; (ii) ICM patients with a history of prior MI derive a significant benefit from preventive CRT-D therapy. However, the benefit of CRT-D in this population is directly related to elapsed time from MI, and is significantly higher among patients with more remote MI; and (iii) ICM patients without a known prior MI derive a pronounced benefit from preventive CRT-D therapy, that is similar to the benefit shown among ICM patients with remote MI.

Adverse remodelling after myocardial infarction

The present study demonstrates that among patients who do not receive preventive CRT-D therapy, the risk of HF or death increases as a function of time from MI despite optimal medical therapy. There are several possible mechanisms for these findings. The occurrence of a large transmural infarction is followed by a series of events referred to as ventricular adverse remodelling and characterized by progressive dilatation, hypertrophy, change in ventricular geometry, and reduced ejection fraction.4,13 The increase in systolic and diastolic wall stress activates cell surface mechanoreceptors that affect intracellular signalling for myocyte hypertrophy and for modulation of the extracellular matrix.14,15 The renin–angiotensin–aldosterone system and the sympathetic nervous system compound the alterations in haemodynamic that occur and play an important role in the progressive remodelling process.4 Gaudron et al.16 demonstrated that following the acute phase after MI, patients can be divided into three prognostic groups: those exhibiting no dilatation, limited dilatation, or progressive LV dilatation. In the later group, LV dilatation is compensatory at first, then progresses to non-compensatory dilatation, and finally results in severe global LV dysfunction. It appears that the adverse remodelling process in patients with progressive dilatation continuous for years after MI.13 The magnitude of LV dilatation after MI is an important determinant of risk for HF, ventricular arrhythmias, and death.4,17 Consistently, the present study also showed that LV- end-systolic and end-diastolic volumes were related to elapsed time from MI (Table 1), further supporting the mechanism relating to the ongoing process of adverse remodelling many years after MI.

Time dependence of the benefit of myocardial infarction therapy on elapsed time from myocardial infarction

Our findings suggest that the benefit of CRT-D therapy among ICM patients with a history of MI is directly related to the time elapsed from the last MI to device implantation, and is pronounced among higher-risk patients who experienced more remote MI. Notably, a recent study by Wilber et al.18 also observed a time-dependent effect of elapsed time from MI on defibrillator efficacy. The study consistently showed enhanced ICD efficacy among higher-risk patients who experienced a more remote MI. The findings from the two studies suggest that progressive adverse mechanical and electrical remodelling associated with elapsed time from MI can be used to identify higher-risk ICM patients who derive enhanced benefit from combined device therapy.

We observed no significant benefit conferred by CRT-D among patients who were enrolled within the first 8 years after MI [HR = 1.26 (0.78–2.03); P = 0.345]. This finding could be partially explained by the fact that during the first 8 years after MI, there was a low risk for HF/death among ICD patients (20% event rate at 2.5 years of follow-up, Figure 3A). In contrast, after 8 years, there was a higher risk for HF/death among ICD patients (31% event rate at 2.5 years of follow-up, Figure 3B) and accordingly a significant CRT-D benefit.

The greater CRT-D benefit observed among patients who experienced remote MI (≥8 years) when compared with a more recent MI (<8 years) was driven mainly by HF events necessitating hospitalization. It should also be noted that the time-dependent effect of elapsed time from MI on CRT-D benefit remained significant after adjustment for multiple variables including baseline EF or alternatively baseline LV volumes implying that several mechanisms are involved beyond time-dependent LV remodelling.

Implantable cardioverter defibrillator and cardiac resynchronization therapy with defibrillator effects in ischaemic cardiomyopathy patients without a documented prior myocardial infarction

Atkinson and Virmani9 described that 26% of consecutive necropsies with ICM had severe coronary artery disease without MI, demonstrating biventricular dilatation with interstitial fibrosis but without pathological evidence of acute or healed MI. Several investigators tried to elucidate the pathogenesis of ICM without a documented prior MI and suggested that repetitive stunning or hibernation causing regional dysfunction might play a direct role in ischaemic LV remodelling and development of ICM if coronary blood flow is not restored.10,19,20 Notably, consistent with the findings from the present study, it was shown that ICM patients without a documented prior MI had significantly worse prognosis compared with ICM patients post-MI.8,21

Currently, there are no data regarding the effect of device therapy in the distinct subset of ICM patients without a documented prior MI. We demonstrated that ICM patients without a documented prior MI who do not receive preventive CRT-D therapy are at an increased risk for HF or death, whereas treatment with CRT-D in this high-risk ICM subset is associated with enhanced clinical and echocardiographic benefit. It is conceivable that ICM patients without documented prior MI respond better to CRT when compared with post-MI patients with myocardial scars because they are more likely to have hibernating myocardium, representing a myocardial contractile reserve which is a strong predictor of LV reverse remodelling and clinical response.22,23 Our findings are consistent with this mechanism and demonstrate that the reverse remodelling effects of CRT-D in MADIT-CRT were significantly greater in the subgroup of ICM patients without a documented prior MI when compared with those with prior MI.

Limitations

A history of prior MI and elapsed time from MI to enrolment were not randomization factors in MADIT-CRT. Accordingly, there were significant differences in the clinical characteristics of the post-MI and no-MI subsets in the present study. Our findings, however, persisted after multivariate adjustments for multiple risk factors and echocardiographic parameters, further suggesting that there is a time-dependent effect of elapsed time from MI on CRT-D efficacy. There were insufficient data regarding the occurrence of prior MI or the date in which MI occurred in 105 patients with ICM. This patient subset showed similar clinical characteristics and derived a similar magnitude of benefit from CRT-D as the ICM subgroup with documented MI, suggesting that the missing data did not contribute a bias to the results regarding the time-dependent effect of MI on CRT-D benefit. The treating physician reported whether a patient had a history of MI (enzyme positive or Q wave MI), the number of MIs, and the date of most recent MI. We did not have a core lab analysing ECG in order to confirm history of MI. In addition, data on the no documented MI subgroup should be interpreted with caution as prior MI might not be absolutely excluded by the physician's clinical judgment.

Summary

Preventive cardiac resynchronization therapy is associated with a significant clinical benefit in mildly symptomatic (NYHA functional class I-II) ICM patients with wide QRS, who are being treated with all currently recommended modalities, including beta-blockers, renin–angiotensin–aldosteron system inhibitors, and ICD. Our findings have major clinical implications among ICM candidates for CRT-D, as they help identify which patients are at greater risk for HF events or death without preventive CRT-D therapy, and accordingly are more likely to derive enhanced benefit from this mode of preventive therapeutic modality. These high-risk subsets include patients with a remote MI with progressive adverse remodelling and those without a documented MI. In contrast, our findings suggest that the benefit of CRT-D therapy is significantly attenuated among lower-risk patients with more recent MI, in whom further risk stratification should be carried out prior to device implantation.

These findings suggest that clinical data regarding MI history should be incorporated during risk assessment among ICM patients who are candidate for preventive CRT-D therapy.

Funding

The MADIT-CRT study was supported by a research grant from Boston Scientific, St Paul, Minnesota to the University of Rochester School of Medicine and Dentistry.

Conflict of interest: A.J.M. reports receiving grant support from Boston Scientific and lecture fees from Boston Scientific, Medtronic, and St Jude Medical; J.S.S. lecture and consultant fees and grant support from Boston Scientific and lecture fees from Medtronic, St Jude Medical, and Biotronik; D.J.W., lecture fees and grant support from Boston Scientific and lecture and consulting fees and grant support from Medtronic and St Jude; and W.Z., grant support and lecture fees from Boston Scientific, and grant support from Medtronic. No other potential conflict of interest relevant to this article was reported.

Acknowledgement

This research was carried out while Dr Alon Barsheshet was a Mirowski-Moss Career Development Awardee at the University of Rochester Medical Center, Rochester, NY.

References

View Abstract