MADIT-CRT: Resynchronization therapy cuts heart-failure risk in patients with only mild disease

Wednesday, September 2, 2009

by Steve Stiles

Barcelona, Spain (updated) - The evidence base has another randomized trial supporting cardiac resynchronization therapy (CRT) in patients with systolic dysfunction, ventricular dyssynchrony by ECG, but only mild heart failure or LV systolic dysfunction without symptoms.

In such patients, for whom a primary-prevention implantable cardioverter defibrillator (ICD) was already indicated, the addition of resynchronization pacing (CRT-D) cut the risk of death or heart-failure events, the primary end point, by about a third over two and a half years in the Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy (MADIT-CRT) [1]. A 41% reduction in risk of heart-failure events accounted for most of the CRT-D benefit, which was independent of whether the cardiomyopathy was ischemic or nonischemic.

"This study provides evidence that preventive CRT-ICD therapy decreases the risk of heart-failure events in vulnerable patients with ischemic or nonischemic heart disease who have minimal heart-failure symptoms but a wide QRS complex," write the authors, led by Dr Arthur J Moss (University of Rochester Medical Center, NY).

The MADIT-CRT findings were published online today in the New England Journal of Medicine to coincide with their release at the European Society of Cardiology Congress 2009. The investigators had released the trial's primary results to the investment community in June, as heartwire reported at the time.

A prespecified subgroup analysis showed that the CRT-D benefit for the primary end point was driven by the 52% risk reduction (p=0.001) in patients with QRS duration of >150 ms vs shorter QRS. There was also a significant interaction by sex, with women showing greater benefit than men (p=0.01).

"The striking thing was that females did better [with CRT-D] irrespective of their QRS duration," Moss said when formally presenting the study at the meeting. "Men got a good result, but the women got twice as good a result."

The definition of heart-failure events in MADIT-CRT allowed for patients to have been treated for decompensation on an outpatient basis and so wasn't strictly limited to HF hospitalization, a much more widely used end point in CRT trials.

CRT was also associated with significant improvements in LV volumes and LVEF, indications of the reverse-remodeling effect long observed as a benefit of the device therapy, including in the REVERSE trial. The MADIT-CRT findings are in general consistent with REVERSE, which also entered patients with mild heart failure, specifically patients in NYHA 1 or 2 with a QRS duration >120 ms, an LVEF <40%, and echocardiographic evidence of LV dilatation, as previously reported by heartwire.

"I think it will very possibly change our indications in the future, to include also these patients," Dr Guenter Breithardt (University Hospital, Münster, Germany) said of MADIT-CRT as the discussant following Moss's presentation. He observed that patients in the trial had a lower LVEF and more dyssynchrony than those in REVERSE, but were otherwise very similar. "We of course have to be cautious about this conclusion, but both trials point in the same direction, that the greatest benefit is achieved in patients in class 2, and with longer QRS complexes."

But, he asked, "Do we have the same strong evidence for class 1 patients as we do for class 2? I don't think so." Class 2 patients were dominant in both trials, he said, so perhaps CRT should be aimed more at them.

Other views

The trial's eligibility criteria reflect an expanded population in terms of NYHA class and a tightening of criteria in terms of QRS duration, relative to class I indications for CRT in the current guidelines: LVEF <35 or less, QRS >120 ms or more, and sinus rhythm with NYHA class 3 or ambulatory class 4 heart failure on optimal medical therapy.

In an accompanying editorial [2], Dr Mariell Jessup (University of Pennsylvania, Philadelphia), points out that "it is not completely clear how the enrolled patients differ from those in earlier CRT trials, since no objective criteria were used to classify functional status at baseline," and functional status assessments thereafter weren't blinded.

"I think we don't understand the [MADIT-CRT] patient population very well," Jessup told heartwire, adding that it seems to be a weakness of the trial. She notes that the report says 10% of the patients had been in NYHA class 3 or 4 more than three months before study entry, as the protocol allowed. Therefore, not all patients in the trial had always been asymptomatic; some apparently had mild heart failure after having responded to medical therapy.

"Patients walk in the door, they have a wide QRS and a low LVEF, and what the guidelines tell us to do is put them on medical therapy and see if they get better. And if they get better and they're asymptomatic and they do well on their six-minute-walk test, we can put in an ICD, but we're not supposed to put in a CRT. This study suggests that you could, that you probably should," Jessup said.

"We have to understand which patients we should be doing this in, because it's an expensive therapy, and we probably need to start with the premise that the QRS must be really wide, not just a little wide. And we have to ask, is there a cheaper way to keep people out of the hospital than doing [CRT]?"

You're talking about a pretty significant healthcare expenditure that doesn't give you an absolute gain in each person who receives the technology.

Given the benefit primarily in patients with a QRS duration of >150 ms and cost-effectiveness concerns if CRT use is broadened based on MADIT-CRT, Jessup writes, it "appears prudent" that any expanded indications to include patients with milder heart failure be limited to those with such very wide QRS complexes and a history of "marked symptoms" that were under control with medical therapy.

As observed for heartwire by Dr Clyde W Yancy (Baylor University Medical Center, Dallas, TX), "a number of patients who undergo CRT don't derive a benefit. So you're talking about a pretty significant healthcare expenditure that doesn't give you an absolute gain in each person who receives the technology. So refining the candidacy and understanding more about the substrate we're targeting end up being pretty important."

Yancy, who is president of the American Heart Association, said he feels the MADIT-CRT results are most relevant in what they say about the current indications for CRT, "and I'm intrigued that there may be other patients who can benefit."

According to Dr Angelo Auricchio (Fondazione Cardiocentro Ticino, Lugano, Switzerland), MADIT-CRT represents a significant advance in heart-failure therapy. "The magnitude of the effect that we saw in this population is as big as what we had seen in past years with beta blockers, as big as what we saw with ACE inhibitors," he told heartwire.

Back then, no one really called for cost-effectiveness analyses before starting to use those drugs, he observed. But, "obviously we will come to a point where we will have to justify the use of devices in this very, very large group of patients. We will have to find the resources," he said.

Perhaps they will come from cost savings as CRT-D helps patients avoid heart-failure events. "This is a relatively healthy population, so it may take 10 years to see an effect on mortality, but now we see an immediate effect on hospitalization. This is really what is going to save money."

Both Auricchio and Dr Josep Brugada Terradellas (University of Barcelona, Spain) seem to agree on the probable impact of the study: not a broad expansion of the indications for CRT, but more physicians convinced of the value of CRT for existing indications. Both kinds of devices are underused in Europe, more so than in North America, they note.

"Every patient in MADIT-CRT has an indication for an ICD, but in real life many of them probably are not receiving one," Brugada said to heartwire. "Now probably they will. I expect an increase in the number of implantations [of both CRT-D and ICD-only devices] in patients who already have an indication for an ICD but are not implanted."

Dr Kenneth Dickstein (University of Bergen, Stavanger, Norway) foresees a huge expansion of the population getting CRT devices. "Now it's a whole new ball game with these mildly symptomatic patients, who obviously have a low mortality rate—think of all the patients who have ICDs who actually would now satisfy [CRT-D] criteria," he told heartwire.

Dickstein, who chairs the writing committee for the European CRT guidelines, observed that neither REVERSE nor MADIT-CRT was published when the last recommendations came out. "So now we'll have to update the guidelines."

Others have expressed caution about how tightly the findings should be embraced right now. "I think we need more data on this before [changing] the recommendations in the guidelines, because it's an invasive therapy," Prof Heinz Völler (Klinik am See, Rüdersdorf, Germany) said for heartwire. He observed that the MADIT-CRT report didn't alleviate all concern about the safety of CRT in its patient population. For example, it didn't say anything about the prevalence of appropriate and inappropriate ICD shocks. Also, he observed, the left-ventricular coronary-vein lead had to be repositioned within the first month in 4% of the CRT-D patients "for a variety of reasons," the report says.

How CRT made it

MADIT-CRT randomized 1820 patients at 110 centers in North America and Europe who were in NYHA class 1 or 2 and had an LVEF <30% and ventricular dyssynchrony as defined by an electrocardiographic QRS duration of at least 130 ms. Three patients received CRT-D devices for every two that got an ICD only. They were commercially available transvenous devices from Boston Scientific, the trial's sponsor. Optimal medical therapy was required in both groups.

The entry criteria specified that patients with ischemic cardiomyopathy could be in NYHA class 1 or 2 and that those with nonischemic disease had to be in NYHA class 2.

Over a follow-up averaging 2.4 years, the primary end point—death from any cause or nonfatal HF events—occurred in 17.2% of CRT-D patients and 25.3% of those getting an ICD (p=0.001). Heart-failure end points were identified by physicians who were aware of the patients' randomization group, but events were blindly adjudicated, according to Moss et al.

HR (95% CI) for the primary end point* and its components, CRT-D patients (n=1089) vs ICD-only patients (n=731)

End point
All patients
Patients with ischemic cardiomyopathy
Patients with nonischemic cardiomyopathy
Death or HF*
0.66 (0.52-0.84)a

0.67 (0.52-0.88)b
0.62 (0.44-0.89)c
HF only
0.59 (0.47-0.74)a
0.58 (0.44-0.78)a
0.59 (0.41-0.87)c
1.00 (0.69-1.44)
1.06 (0.68-1.64)
0.87 (0.44-1.70)

a. p<0.001

b. p=0.003

c. p=0.01

CRT-D was associated with significantly improved LV volumes in a subgroup of 746 CRT-D patients and 620 ICD patients who underwent two-dimensional echocardiography at baseline and at one year. LVEF also rose significantly higher with resynchronization.

Change in left ventricular volumes by echocardiography over one year

CRTD, n=746
ICD, n=620

LVEDV=left ventricular end diastolic volume. LVESV=left ventricular end-systolic volume.

p<0.001 for all differences between CRT-D and ICD

To download tables as slides, click on slide logo above

"It's an exciting study," Jessup said to heartwire, recalling her reaction when she learned of the results of the MADIT-2 trial. "I said, oh my God, that's a lot of devices. And that was my impression when I read this, that this is a lot of devices."

But MADIT-CRT should be far from the last word. Further trials should have narrower entry criteria. And "a really cool trial would be to say we're going to put [CRT devices] in a group of patients with a very wide QRS, 150 ms, and we're going to compare them to a nurse call center and see who stays out of the hospital more often for the least amount of money. That's the kind of question we have to ask now. Do we really need to be putting these expensive devices in, or is there another way to keep people out of the hospital?"

"We have a current patient population that is not reaping the full benefit of CRT, and if part of the reason is some trepidation regarding the real benefit, these data, along with REVERSE, really make the compelling argument that there are changes in ventricular size and function that are really quite striking," Yancy said.

"There's a totality of evidence accumulating that says that CRT is beneficial, and I'm hoping that totality of evidence will be especially effective for the patients who already have a well-established indication [for CRT]," he said. "I think for those physicians who have not accepted CRT, it's time."


  1. Moss AJ, Hall WJ, Cannom DS, et al. Cardiac-resynchronization therapy for the prevention of heart-failure events. N Engl J Med 2009; DOI: 10.1056.NEJMa0906431. Available at:
  2. Jessup M. MADIT-CRT—Breathtaking or time to catch our breath? N Engl J Med 2009; DOI:10.1056.NEJMe0907335. Available at:


Post a Comment