MADIT-CRT

Published

Moss AJ, et al. "Cardiac-resynchronization therapy for the prevention of heart-failure events". The New England Journal of Medicine. 2009. 361(14):1329-1338.
PubMed • Full text • PDF

Clinical Question

Among patients with HF with LVEF ≤30% and QRS duration ≥130 msec, does placement of an ICD with cardiac resynchronization therapy reduce mortality or HF events when compared to ICD placement alone?

Bottom Line

Among patients with HF with LVEF ≤30% and QRS duration ≥130 msec, placement of an ICD with cardiac resynchronization therapy reduces the rate of mortality or HF events when compared to ICD placement alone. This benefit was driven primarily by a reduction in HF events.

Major Points

The benefit of cardiac resynchronization therapy (CRT) for patients with LVEF ≤35%, cardiac dyssynchrony, and NYHA class III or IV symptoms was demonstrated by the 2002 MIRACLE and 2005 CARE-HF trials.The role of ICD plus CRT for patients with milder symptoms was less well established.

The 2009 Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy (MADIT-CRT) trial randomized 1,820 patients with LVEF ≤30%, QRS duration ≥130 msec, NYHA class I or II in a 2:3 ratio to ICD or ICD with CRT. With 2.4 years of follow up, the trial was stopped early because of an interim efficacy calculation. The ICD-CRT group had a lower rate of the primary outcome of first event of all-cause mortality or non-fatal HF event (17.2% vs. 25.3%; NNT 12). This outcome was driven primarily by a reduction in HF events (22.8% vs. 13.9%; NNT 11), which were essentially acute decompensated HF events. The ICD-CRT group had improvements in LVEF, LVEDV, and LVESV at 1 year. Subgroup analyses of the primary outcome revealed no benefit for those with QRS duration <150 msec. A post hoc analysis published in 2011 demonstrated efficacy of ICD-CRT in those with LBBB morphology; those with RBBB or other intraventricular morphologies did not experience a reduction in the primary outcome.

The 2010 RAFT study was similar in design and execution though its participants were men, had a greater burden of symptoms, and a higher rate of ischemic disease. A seven-year follow-up of patients with LBBB in MADIT-CRT published in 2014 showed an 11% absolute reduction in mortality between in patients with LBBB ICD-CRT.

Kirk and colleagues demonstrated that CRT therapy improves myofilament sensitivity to calcium in animal models, possibly through increasing phosphorylation by glycogen synthase kinase 3β (GSK-3β). This implicates GSK-3β as a potential pharmacologic target though trials demonstrating efficacy of GSK-3β agonists have yet to be completed.

Guidelines

AHA/ACCF Heart Failure Guidelines (2013, adapted)

• For patients with LVEF ≤35 on guideline-directed medical therapy:
  • CRT indicated if sinus rhythm, LBBB, QRS ≥150 msec, and NYHA class II (class I, level B) or NYHA class III or ambulatory class IV symptoms (class I, level A)
  • CRT useful if sinus rhythm, no LBBB, QRS ≥150 msec, with NYHA class III or ambulatory class IV symptoms (class IIa, level A)
  • CRT useful if sinus rhythm, LBBB, QRS 120-149 msec, with NYHA class II, III or ambulatory class IV (class IIa, level B)
  • CRT useful if AF if requiring V-pacing or meeting other CRT criteria and rate will allow near 100% V-pacing with CRT either by AV-nodal ablation or rate-controlling medications (class IIa, level B)
  • CRT useful if undergoing placement/replacement of a device and anticipated need for >40% of ventricular pacing (class IIa, level C)
• CRT not recommended for NYHA class I or II symptoms, no LBBB, and QRS <150 msec (class III, level B)
• CRT not indicated if comorbid conditions are limiting survival to <1 year (class III, level C)

These guidelines do not reflect the 7 year follow-up data published in 2014.

Design

• Multicenter, randomized, open label, controlled trial
• N=1,820
  • ICD (n=731)
  • ICD-CRT (n=1,089)
• Setting: 110 centers in the US (88), Canada (2), and Europe (20)
• Enrollment: 2004-2008
• Average follow-up: 2.4 years (stopped June 2009 because of efficacy endpoint)
• Analysis: Intention-to-treat
• Primary outcome: First event of all-cause mortality or non-fatal HF event

Population

Inclusion Criteria

• Age ≥21 years with HFrEF (LVEF ≤30%) defined by one of the following:
  • ICM with NYHA class I or II symptoms
  • NICM with NYHA class II symptoms
• Sinus rhythm
• QRS ≥130 msec
• Indication for ICD therapy

Exclusion Criteria

• Other indication for CRT
• PPM, ICD, or CRT in place already
• NICM and NYHA class I
• NYHA class III or IV in prior 3 months
• CABG or PCI in prior 3 months or CAD on angiogram likely requiring CABG or PCI in the near future
• Elevated cardiac enzymes in the prior 3 months
• 2nd or 3rd degree heart block
• Irreversible brain damage from preexisting disease
• Women planning to become pregnant or women already pregnant
• Reversible cause of NICM (acute viral myocarditis or alcoholic cardiomyopathy with discontinuation of alcohol intake)
• Chronic AF in prior month
• Other chronic illness likely to limit survival during the trial (including CKD with creatinine >3 mg/dL or BUN >70 mg/dL)
• Participation in other trials
• Unwillingness or unable to cooperate with the protocol
• Live too far from study site to have frequent follow-up or planning to move during the trial
• Lack of consent

Baseline Characteristics

From the ICD group.

• Demographics: Age 64 years, male 75.6%
  • Race: White 90.7%, black 7.7%, other 1.5%
• HF specifics:
  • ICM: NYHA class I 15.5%, NYHA class II 39.4%
  • NICM: NYHA class II 45.1%
  • NYHA class III or IV >3 mo before enrollment: 10.0%
• Heart disease risk factors: Treatment for HTN 63.2%, AF >1 mo before enrollment 12.6%, DM 30.6%, smoking 12.8%, BMI ≥30 kg/m², CABG 28.5%
• Baseline data: BP 121/71 mmHg, creatinine 1.2 mg/dL, BUN ≥26 mg/dL 24.5%, LBBB 71.3%, RBBB 12.6%, QRS ≥150 msec 65.1%, LVEF 24%, 6MWT 363 m
  • Cardiac: LVEDV 251 mL, LVESV 179 mL
• Baseline medications: Aldosterone antagonist 30.9%, amiodarone 7.0%, ACE-inhibitor 77.0%, ARB 20.2%, beta-blocker 93.2%, class I antiarrhythmic 0.4%, dig 24.2%, diuretic 72.9%, statin 67.2%

Interventions

• Randomized in a 2:3 ratio, stratified by clinical center and ischemic status, to a group:
  • ICD - Placement of an ICD-only device set to VVI for single-chamber units and DDI for dual chamber units, lower rates of 40 BPM, and hysteresis off
  • ICD-CRT - Placement of a ICD-CRT device set to DDD mode, lower rates of 40 BPM, and hysteresis off
• Optimal HF medical management

Outcomes

Presented as ICD vs. ICD-CRT.

Primary Outcome

First event of all-cause mortality or non-fatal HF event

Non-fatal HF event defined as HF signs or symptoms responsive to IV decongestive therapy as an outpatient or augmented decongestive therapy with oral or parenteral medications as an inpatient.

25.3% vs. 17.2% (HR 0.66; 95% CI 0.52-0.84; P=0.001; NNT 12)

ICM, NYHA class I or II: 29.2% vs. 20.4% (HR 0.67; 95% CI 0.52-0.88; P=0.003)

NICM, NYHA class II: 20.6% vs. 13.2% (HR 0.62; 95% CI 0.44-0.89; P=0.01)

Secondary Outcomes

Heart failure event

22.8% vs. 13.9% (HR 0.59; 95% CI 0.47-0.74; P<0.001; NNT 11)

ICM, NYHA class I or II: 26.2% vs. 16.1% (HR 0.58; 95% CI 0.44-0.78; P<0.001)

NICM, NYHA class II: 18.8% vs. 11.2% (HR 0.59; 95% CI 0.41-0.87; P=0.01)

All-cause mortality

7.3% vs. 6.8% (HR 1.00; 95% CI 0.69-1.44; P=0.99)

ICM, NYHA class I or II: 8.7% vs. 8.9% (HR 1.06; 95% CI 0.68-1.64; P=0.80)

NICM, NYHA class II: 5.5% vs. 4.3% (HR 0.87; 95% CI 0.44-1.70; P=0.68)

Additional Analyses

No device placement

2.6% vs. 1.0%

Placement of the other group's assigned device

7.5% vs. 12.4%

Device removal

0.7% vs. 1.3%

Change in echocardiographic measurements at 1 year

LVEF: +3% vs. +11% (P<0.001)

LVEDV: -15 vs. -52 mL (P<0.001)

LVESV: -18 vs. -57 mL (P<0.001)

Subgroup Analysis

For the primary outcome There was no difference by age, NYHA class, LVEF ≤25% or >25%, LVEDV ≤240 mL or >240 mL, or LVESV ≤170 mL or >170 mL.

Sex

Male: HR 0.76; 95% CI 0.59-0.97

Female: HR 0.37; 95% CI 0.22-0.61

P-value for interaction 0.01

The authors note that women had benefit from ICD-CRT independent of duration of QRS

QRS duration

<150 msec: HR 1.06; 95% CI 0.74-1.52

≥150 msec: HR 0.48; 95% CI 0.37-0.64

P-value for interaction 0.001

Adverse Events

Mortality related to device insertion hospitalization

0 vs. 1 death (from PE)

In 30 days after device insertion

Pneumothorax: 0.8% vs. 1.7%

Infection: 0.7% vs. 1.1%

Evacuation of pocket hematoma: 2.5% vs. 3.3%

Serious device-related adverse events

5.2 vs. 4.5/100 device-months

ICD-CRT group-only complications

Coronary venous dissection and pericardial effusion: 0.5%

LV coronary vein repositioning in first 30 days: 4.0%

Criticisms

• Open-label design may have influenced the diagnosis of heart failure complications
• Unclear clinical significance of the primary endpoint
• Minimal reduction in the rate of heart failure events given the cost of the intervention
• Relatively high rate of unsuccessful lead placement, device placement, and device removal
• Unclear why the remodeling benefit did not result in mortality benefit (which it subsequently did for patients with LBBB on 7 year follow-up)

Funding

• Grant from Boston Scientific, the manufacturer of the ICD and ICD-CRT devices
• Authors with multiple financial conflicts