REVERSE, MADIT-CRT Make Case for Extending CRT to Mild Heart Failure: Should Guidelines Change?

October 01, 2009

October 1, 2009 (Washington, DC) — The literature now has another major randomized trial suggesting that cardiac resynchronization therapy (CRT) benefits patients with mild or asymptomatic heart failure, clinically and in terms of reverse remodeling, much as it benefits patients with more severe heart failure. Although the guidelines recommend CRT for patients in NYHA functional class 3-4, a prospectively planned 24-month analysis from the Resynchronization Reverses Remodeling in Systolic Left Ventricular Dysfunction (REVERSE) trial [1] mirrors the results of the recently published, much larger MADIT-CRT trial [2] in showing that NYHA 1-2 patients can benefit as well.   The guidelines should now be changed to recommend CRT for such patients with milder heart failure, REVERSE principal investigator Dr Cecilia Linde (Karolinska University Hospital, Stockholm, Sweden), said to heartwire .

"The results of MADIT-CRT and REVERSE at 24 months are remarkably consistent," agreed Dr Claude Daubert (Hôpital Pontchaillou-CHU, Rennes, France), another lead REVERSE investigator. "There is thus clear evidence today that CRT may modify disease progression in mildly symptomatic patients already on optimal medical treatment. I personally believe that time has come to update the guidelines on clinical indications for CRT," he told heartwire in an email.

The trial's primary outcomes based on 610 patients, all recipients of CRT-defibrillator (CRT-D) devices who were then randomized at 73 North American and European centers, have already been published [3]: no significant difference at one year in rate of clinical worsening, as gauged by a "clinical composite response" end point, between patients whose devices had the CRT function programmed "on" (n=419) vs "off" (n=191). On the other hand, CRT was associated with improved LV end-systolic and end-diastolic volumes, which were associated with significantly higher LV ejection fractions.

We should probably drop the NYHA criteria, at least when it comes to choosing interventions to alter patients' prognosis.

As previously reported by heartwire , enrollment criteria had included NYHA functional class 1–2 with a QRS duration >120 ms, an LVEF <40%, and a LV end-diastolic dimension of at least 55 mm. According to current guidelines, CRT is indicated in patients with NYHA class 3–4 heart failure, LVEF <35%, and the same degree of QRS prolongation despite optimal medications.

In the prespecified two-year analysis of the trial's 262 patients randomized at European centers, those assigned to "CRT-on" improved echocardiographically but also with a significant reduction in the clinical composite end point: 19% vs 34% (p<0.001). That end point was defined as clinical worsening indicated by death, heart-failure hospitalization, crossover to the other treatment group, worsened NYHA class, and patient global assessment.

CRT-on patients also showed a 62% reduction in risk of heart-failure hospitalization or death (p=0.003), apparently driven by the 61% reduction in risk of HF hospitalization by itself (p=0.01). Their reverse-remodeling benefits were similar to those seen in the larger REVERSE trial; most of the effect occurred in the first six months.

The European-cohort results had been reported previously at the American College of Cardiology 2009 Scientific Sessions and covered then by heartwire . They're now published online September 30, 2009 in the Journal of the American College of Cardiology with Daubert as first author.

As the invited commenter after the MADIT-CRT trial was presented at the recent Heart Failure Society of America 2009 Scientific Meeting, Prof John GF Cleland (University of Hull, Kingston-upon-Hull, UK) said its finding of a CRT clinical benefit in NYHA class 1–2 patients "is consistent in size and direction" with those of REVERSE and earlier trials such as MIRACLE and CARE-HF.

"I think together they should change the guidelines. We should probably drop the NYHA criteria, at least when it comes to choosing interventions to alter patients' prognosis," said Cleland, who led the CARE-HF trial.

"I think that symptoms are not a useful guide to the prognostic benefits of CRT, and patients should no longer be excluded from having CRT just because medical therapy has controlled symptoms. But perhaps we can go one step further," he said. In REVERSE and possibly in MADIT-CRT, the clinical benefits of CRT were more pronounced in sicker patients. In CARE-HF, he noted, patients with higher natriuretic peptide levels had the worst outcomes and showed a significant benefit from CRT. "But the absolute reduction in mortality was similar, while the relative reduction was greater in patients with less-elevated [natriuretic peptides]."

So, Cleland said, it may be that CRT "can modify outcomes in terms of years" in patients with less severe heart failure, "but perhaps only in terms of months" in patients with more advanced disease.

"I am absolutely convinced that the guidelines ought to change," Linde told heartwire . There remains some concern, however, that even among patients like those who were eligible for REVERSE or MADIT-CRT, not all benefit the same from CRT. For example, those with a QRS duration <150 ms seem to benefit less.

Another reservation I have is that the NYHA class 1 patients are few in both in REVERSE and MADIT CRT, so that the results from those patients might not be representative of the entire group.

"Another reservation I have is that the NYHA class 1 patients are few in both in REVERSE and MADIT CRT, so that the results from those patients might not be representative of the entire group," Linde said. "In those patients, I think the evidence for a benefit is less strong."

She emphasized that NYHA class 1 doesn't include patients with LV systolic dysfunction who have never had symptoms; little is known about CRT in that group. REVERSE limited enrollment to patients with mild symptoms and those who were previously symptomatic but became asymptomatic on medical therapy.

"I believe it is too early to recommend CRT use in NYHA class-1 patients," Daubert agreed. He also said the guidelines may have to account for the greater degree of QRS-interval lengthening, compared with the >120 ms stated for NYHA 3–4, which seems necessary for successful CRT in patients with mild heart failure.

Or, Linde observed, any guidelines-writing committee reconsidering CRT indications may simply stick to the trials' entry criteria and not consider subgroup analyses.

Still, "If you want to generalize our findings to a wider population, then I think these questions remain to be answered," Linde said. A pooled analysis of REVERSE and MADIT-CRT might help, she proposed. "At least, it might be worthwhile to do, because after their results, which are so congruent and point in the same direction, it would be very hard to have a placebo-controlled trial in mild heart failure with 'CRT-off' as one of the randomization arms."

In an editorial [4] accompanying the REVERSE 24-month report, Dr Derek V Exner (University of Calgary, AB) writes that the results "are interesting" but also that "they should not be considered definitive."

He also echoed other experts in expressing doubts about MADIT-CRT, in particular about the proportion of heart-failure events in the primary end point reduced by CRT that actually consisted of heart-failure hospitalizations. Also, the population's baseline six-minute-walk distances "suggest that many of these patients would have been categorized as NYHA functional class 3 in past trials."

Exner concludes, "It is tempting to recommend [CRT] beyond present guidelines. However, it is premature to recommend CRT as a routine intervention to patients with asymptomatic LV dysfunction or those with mildly symptomatic heart failure today."

REVERSE was sponsored by Medtronic. MADIT-CRT was sponsored by Boston Scientific. Linde reports receiving honoraria for speaking and consulting fees from Medtronic and St Jude Medical and research grants from Medtronic; disclosures for the other authors are listed in the paper. Cleland reports receiving research grants from AstraZeneca, Medtronic, and Philips and honoraria for speaking from AstraZeneca and Medtronic and being a consultant or serving on an advisory board for Medtronic. Exner reports receiving honoraria and research support from Medtronic and St Jude Medical, honoraria from Boston Scientific, and research support from Sorin/ELA.