Reed Miller

June 19, 2012

June 18, 2012 (Nice, France) — The value of cardiac resynchronization therapy (CRT) for patients with a QRS between 120 and 150 ms remains a major source of controversy among the experts [1].

As reported by heartwire , the Heart Failure Society of America's latest guidelines on heart failure changed the recommended QRS threshold for CRT from >120 ms to >150 ms, in the absence of right bundle-branch block (RBBB). These guidelines state that CRT for otherwise-eligible patients with QRS durations >120 ms to <150 ms is merely "to be considered." Then last month, the European Society of Cardiology Heart Failure Association released guidelines giving the strongest recommendation for CRT in patients with NYHA class 2 heart failure, left-bundle-branch block (LBBB), LVEF <30%, and a QRS duration >130 ms. These guidelines state CRT should be considered for patients with non-LBBB morphology and a QRS duration >150 ms.

Here at Cardiostim 2012, Dr Johannes Holzmeister (University of Zurich, Switzerland) said that he was "shocked" that the guidelines did not address the non-LBBB heart-failure patients with a QRS duration between 120 and 150 ms. "These new guidelines, which are totally based on subgroup analysis, exclude the majority of heart-failure patients from CRT." He argued that physicians need to "more diligently tailor" device therapy to this subpopulation, "but at the moment, with the current guidelines, we're just excluding them and giving the physicians no guidance and no comment about these patients in the guidelines. Nothing.

These new guidelines, which are totally based on subgroup analysis, exclude the majority of heart-failure patients from CRT.

"Unfortunately, the new guidelines will have huge implications for patients and physicians, because insurances in Europe will pick up these guidelines and use this and just look at the QRS, because it's very easy to measure QRS and bundle-branch block," Holzmeister argued.

The trials that inform these guidelines were not powered to show whether CRT benefits this group of patients, he argues. For example, 90% of patients in CARE-HF, "which sets the stage for CRT overall," had LBBB, "so you can't come to a conclusion about patients with other conduction abnormalities from CARE-HF," he said. In MADIT-CRT, patients with a QRS <150 did not respond to CRT as well as those with a higher QRS, on average, but the confidence interval for that group was very wide, he pointed out, so that trial should not be used as evidence that an individual patient with a relatively narrow QRS will not benefit from CRT.

As reported by heartwire , a recent meta-analysis by Dr Ilke Sipahi (University Hospitals and Case Medical Center, Cleveland, OH) suggested that CRT does not generally help patients with a QRS duration <150 ms, but Holzmeister argued that the trials included in this meta-analysis did not include enough non-LBBB patients to draw definitive conclusions about this group.

Will EchoCRT Answer More Questions?

Holzmeister argued that mechanical dyssynchrony defines which patient groups benefit from CRT, and therefore echocardiographic measures may provide a better guide to CRT therapy than QRS morphology. "There are different levels of mechanical dyssynchrony. In the narrower-QRS patients, you have dyssynchrony within the left heart, and as QRS widens, intraventricular dyssynchrony adds up, and then you also have dyssynchrony on the atrial ventricular level, so it's more complex than what is reflected in the QRS complex alone," he explained. He estimates that about half of patients with a QRS <130 ms have mechanical dyssynchrony.

Holzmeister is one of the lead investigators for the EchoCRT trial, sponsored by Biotronik, which is comparing CRT with controls (CRT functions on or off) in patients with heart failure due to left ventricular systolic dysfunction with a narrow QRS width (<130 ms) and echocardiographic evidence of ventricular dyssynchrony. So far the trial has enrolled about 800 of the planned 2300 patients.

"A patient has a very high chance of mechanical dyssynchrony if their QRS is above 160 ms, but the majority of patients are [below 150 ms], and we need new data on them," he said. "Because if we proceed like this now, we exclude 70% of HF patients from CRT. . . . With EchoCRT, we'll have a good answer on how to take care of these patients."

Will We Ever Find This Hoped-for Key to Mechanical Dyssynchrony?

The pros and cons of CRT for patients with QRS duration <150 ms was the topic of a staged debate at Cardiostim between Dr Kamil Sedlácek (Institute for Clinical and Experimental Medicine, Prague, Czech Republic) and Dr Cecilia Linde (Karolinksa University Hospital, Stockholm, Sweden).

Sedlácek argued that "there is no reproducible and widely applicable method beyond QRS for CRT patient selection that corresponds to or signifies electrical abnormality amenable to CRT. Currently patients with progressively narrower QRS have less chance to benefit and more chance to derive harm, so we remain very conservative and cautious when applying CRT to patients with a QRS less than 150 ms."

Linde disagrees with the most recent changes to the guidelines on CRT with respect to the QRS thresholds. "For those with intermediately prolonged QRS, 120 to 150 ms, the new recommendation is that it may be considered, which really means it should perhaps not be done, and that causes terrible confusion," she said. "That means that the big part of the heart-failure patients who were before implantable no longer have the right to receive such treatment, and I think that's terribly wrong."

Sedlácek acknowledged that "there are a substantial proportion of patients with a broad QRS with no mechanical dyssynchrony and patients with a narrow QRS may still have signs of dyssynchrony." But despite all of the wealth of imaging modalities now available, none of the proposed measures of dyssynchrony have been able to beat QRS for identifying CRT responders, he said. Single-center studies support the application of echo indicators of ventricular dyssynchrony, including pulsed-Doppler and tissue-Doppler imaging, M-mode echocardiography, and myocardial strain-rate imaging. However, the PROSPECT study found that--so far at least--none of these methods is sufficiently sensitive and specific to recommend for widespread clinical practice, Sedlácek pointed out.

We've been struggling with dyssynchrony for years . . . and we came up with basically zero.

"We've been struggling with dyssynchrony for years, and we dedicated the most echocardiographers to dyssynchrony assessment, and we've come up with basically zero," he said. "There are very powerful predictors of response in CRT, and if you put them in multivariate analysis, you get very nice models of response prediction, but obviously they don't work in individual patients."

Sedlácek does not have much hope that EchoCRT will solve this puzzle, because the results may not be generalizable beyond the academic medical centers participating in the trial. "To me, if EchoCRT is positive, it would mean a disaster, and even more to the echo groups in this work, because then the whole world will have to send their echos to Pittsburgh for validation," he quipped. "If EchoCRT is positive, they will probably need to do another trial, which will be multicenter and again assess if people are able to assess dyssynchrony, not only in experienced centers, but in the real world."

He also pointed out that the REVERSE and MIRACLE studies showed that some patients who do not respond to CRT actually suffer worsening heart failure, "so inappropriate application of this therapy may actually minimize benefit and cause harm."

Linde agreed with Sedlácek that imaging dyssynchrony may be difficult at most centers, but EchoCRT will be tremendously important if it confirms that narrow-QRS patients may also have dyssynchrony that is potentially treatable with CRT, she said.

She stressed that QRS is a continuous variable, so "even though the wider the QRS the greater the benefit, it doesn't mean patients with intermediately prolonged QRS do not derive a benefit." The thresholds set by the guidelines reflect the inclusion criteria of the trials rather than discrete, identifiable patient groups, she said. "It's clear that the greater benefit is derived in the groups whose QRS surpasses 150 ms in all of these trials, but they are not based on individual cases."

Even though the wider the QRS the greater the benefit, it doesn't mean patients with intermediately prolonged QRS do not derive a benefit.

Linde apparently agrees with Holzmeister that the guidelines appear to be giving too much credence to the meta-analyses. The meta-analyses performed so far "are really meta-analyses of results of major trials but not case-to-case results," she said. "But really the best way to do this is perhaps not to throw the results from all the major trials into one meta-analysis but to do a proper meta-analysis where you have individual results." Linde indicated that results from this type of case-by-case meta-analysis will soon be available. "That will be much better evidence of what is really true."

After the debate, Sedlácek acknowledged that "we actually don't know what happens to those narrow-QRS patients. We'll have to do new trials to really know. I absolutely agree that always denying that therapy to these patients [just based on the current evidence] would not be correct." He said that, at his center, "we don't draw this dichotomy. We probabilistically assess the likelihood of response in those patients . . . and the more symptoms the patient has, the more likely our patients are to get CRT."

Linde added that "the biggest problem in CRT is underimplementation. There is a vast population that does not get therapy, even among those with QRS above 150 ms."