EBM 10/09/09 Lisa Rose-Jones MD

Cardiac-Resynchronization Therapy for the Prevention of Heart-Failure Events

Authors: Arthur J. Moss MD, W. Jackson Hall PHD, David S. Cannon MD, et al

*referred to as MADIT-CRT Study~(Multicenter Automatic Defibrillator Implantation Trial- Cardiac Resynchronization Therapy)

Background: It is estimated that around half of allpatients with heart failure suffer from sudden cardiac death secondary to ventricular arrhythmias. There are approximately 4-5 million Americans with heart failure with the AHA estimating about 550,000 new diagnosis eachyear. Since 1996 when the MADIT study arrived on the scene in the NEJM we have known that prophylactic implantable cardioverter defibrillator (ICD) therapy improves survival in high risk patients. In 2002, MADIT II was published confirming again this hypothesis and showed that heart attack survivors w/ EF <30% w/o any other risk stratification (ie: EP testing as had been conducted in MADIT) had a mortality benefit. ICDs have been recommended as secondary prevention in Clinical Practice Guidelines by the ACC/AHA since 2005 for those heart failure patients with ICM greater than 30 days out from MI and EF<30% or NICM with EF<35%. MIRACLE trial in JAMA in 2003led to current recommendations for combination ICD therapy+cardiac resynchronization therapy in those patientsw/ EF 35%NYHA class 3-4 symptoms of heart failure who are still symptomatic despite optimal medical therapyand have demonstratable ventricular dyssynchrony as evidenced by QRS duration 130 ms. The thought behind CRT's benefit is that it synchronizes muscle contraction of both the left and right ventricle simultaneously and thus improved efficiency. It differs from standard pacemaker in that an extra lead is attached in left ventricle. This study, MADIT-CRT, was designed to evaluate whether CRT could also help patients with milder heart failure who have signs of dyssynchrony. The implicationsif it did reveala positive outcome would be potential cost savings from slowing progression of heart failure or reducing hospitalizations.

Clinical Q: Does combined ICD-CRT therapy reduce the risk of mortality and heart failure events in patients with Mild cardiac symptoms (NYHA Class 2 ICM or NICM or NYHA Class 1 ICM), LV dysfunction (EF<30%) and wide QRS (>/= 130 ms) as compared to ICD therapy alone?

Study Type: Multi-centered, randomized, open label Interventional Trial

Inclusion Criteria:

~ Age 21 yo or older, Both genders eligible

~Ischemic Heart Disease defined as:

  • NYHA Class I or II for the past 3 calendar months prior to and at time of enrollment
  • 1 or more clinically documented (Q wave or + enzyme) prior MI, but not w/in 2 calendar months of enrollment; and/or
  • 1 or more prior CABG surgeries or PCI (POBA, Stent angioplasty) but not w/in 3 calendar months of enrollment

OR

Non-ischemic Heart Disease defined as:

  • dilated cardiomyopathy characterized by a low ejection fraction and increased ventricular volume, w/ ventricular compliance that is normal or increased
  • NYHA Class II for the past 3 calendar months prior to, and at the time of enrollment

~All of the following had to be met:

  • stable optimal pharmacologic therapy
  • EF of < or = to 0.30 by angiographic, radionuclide, or Echo methods w/in 1 yr prior to enrollment and re-confirmed by ECHO w/in 14 days prior to randomization
  • Resting QRS > or = 130 ms
  • Sinus rhythm by ECG (including RBBB and 1st deg heart block w/ PR<250 ms)

Exclusion Criteria:

*Existing indication for CRT

*implanted pacemaker

*existing ICD or CRT device

*NYHA Class 1 w/ NICM

*NYHA Class III, IV in the past 3 calendar months

*CABG or PCI w/in past 3 calendar months

*Enzyme positive MI w/in the past 3 calendar months prior to enrollment

*Candidates for revascularization who likely to undergo such in the foreseeable future

*2nd or 3rd degree heart block

*Irreversible brain damage from preexisting cerebral disease

*Females who were pregnant or planned to become so, those of childbearing age w/ neg preg test

*reversible NICM such discontinuation of alcohol in alcohol-induced heart dz

*chronic A fib

*presence of any disease other than cardiac associated w/ reduced likelihood of survival for duration of trial (Ie: Cr>3.0, liver failure, cancer)

*participants in other trials

*living distance from clinic was such that follow-up would be difficult

Study Design:

-Patients randomly assigned 3:2 ratio to receive either CRT w/ ICD or only ICD, stratified according to clinical center and ischemic status w/ use of algorithm to ensure balance

-Patient seen in clinical follow-up at 1 month after randomization and then at 3 month intervals until termination of trial

-Commerically available Boston Scientific devices were used

Outcomes:

Primary end point was combined endpoint of all-cause mortality or nonfatal heart failure events

*Heart failure events was diagnosis made by physicians aware of study assignments, req'd signs and symptoms consistent w/ CHF that was responsive to decongestive therapy. Adjudication of end points was carried out by indep mortality committee & by a HF committee according to prespecified criteria & who was unaware of randomization

Statistical Analysis:

-Intention to Treat Analysis, followed 1820 patients over 4.5 year follow-up of which 1089 received ICD+CRT and 731 ICD alone

- group sequential design w/ power 95% to detect hazard ratio of 0.75 at a two-sided significance level of 0.05. Prespecified terms for stopping boundaries (based on log-rank statistic) for an independent monitoring board to terminate the trial early if data supported favor of CRT-ICD, favor or ICD only, or no significant difference. On 6/22/09 after 9th interim analysis, trial was stopped because end points meet superiority boundary (Figure 1)

-Primary analysis endpoint was based on statistical log-rank test stratified according to study center & ischemic status

Results:

-of 1089 patients assigned to CRT, 11 did not receive a device; of 731 patients assigned to ICD only, 19 did not receive a device. There were 173 crossovers thatoccurred: 91 assigned to ICD only received CRT; 82 assigned to CRT-ICD received ICD only 2/2 technical difficulties. 14 devices removed in CRT arm, 5 in ICD only. 44 participants in the CRT group and 55 in the ICD only group withdrew or were lost to follow-up

-demographic characteristics and severity of disease were similar b/w 2 groups (see TABLE 1)

-In non-stratified Analysis:

Primary endpoint occurred in 187 (17.2%: 36 deaths, 151 HF events) of CRT-ICD groupvs. 185 (25.3%: 18 deaths, 167 HF events) in ICD only (p=0.001). The calculated hazard ratio of 0.66 indicates a 34% reduction in the risk of death or nonfatal HF in those receiving CRT in addition to lone ICD. (Similar findings between ischemic and non-ischemics) This difference was driven entirely by a dramatic 41% reduction in heart-failure symptoms:

Primary Endpoint (p=0.001)

  • ICD: 25.3%
  • ICD+CRT: 17.2%

Heart Failure Only (p<0.001)

  • ICD: 22.8%
  • ICD +CRT: 13.9%

Mortality (p=0.99)

  • ICD: 8.7%
  • ICD + CRT: 8.9%

-Kaplan-Meier estimates of event-free outcomes in 2 study groups shown in Fig 2

-There were 7 prespecified subgroup analysis (see Figure 3) which identified a greater benefit in women than in men & in pts w/ QRS duration of 150 msec or more

Are the Results Valid?

  • Was the assignment of patients to treatment randomized? Yes
  • Was randomization concealed? Yes
  • Was follow-up complete? Yes
  • Were patients analyzed in the groups to which they were randomized? Yes (Intention to treat)
  • Were patients, health workers, and study personnel “blind” to treatment? No
  • Were the groups similar at the start of the trial? Yes
  • Aside from the experimental intervention, were the groups treated equally? Yes

Limitations

*Did investigators knowledge of study-group assignments led to lower frequency of dx heart failure in the experimental arm?

*More HF in ICD only group 2/2 pacemaker induction? Probably not since demand rate was 40 bpm

*Composite Endpoint

*Group Sequential Design -> precise way to determine when clinical trial reaches statistical significance. Trial was stopped early because of perceived benefit. Ethical? We unfortunately can generate overestimated treatment effects from this and need to consider ensuring that there is a large number of outcome events that accrue before stopping

* Subgroup Analysis:This can be very misleading because they are prone to over-interpretation. The problem is that by chance you can be lead to apparently significant differences between sub-groups, and these are really only helpful in very large trials which show very large overall differences in the treatment and control groups.