John Mandrola


September 06, 2013

Cardiac resynchronization therapy (CRT) is special. It stands out as perhaps the most impactful procedure an electrophysiologist performs. The multilead pacing device can transform and extend the lives of selected patients who suffer from both impaired systolic function and electrical disease.

The degree of benefit in CRT responders cannot be understated. It's truly remarkable. Bearing witness to the CRT response gets your attention. This is why we went to medical school. Doctors want to extend this sort of benefit to as many patients as possible.

The unfortunate aspect of CRT is that it works only in CHF patients who also have delayed activation of the left ventricle. CRT is a purely electrical therapy. There's no magic. All it does is preexcite a delayed lateral LV wall. That's important because CRT devices are expensive, and like any invasive therapy, come with significant risks. Only patients likely to improve should be exposed to the risk of CRT.

The Echo-CRT trial , presented September 3, 2013 at the European Society of Cardiology (ESC) 2013 Congress and simultaneously published in the New England Journal of Medicine, contributes greatly to the understanding of patient selection for CRT devices. (See Steve Stiles' report.) The findings underscore the electrical nature of CRT therapy. Echo-RT warns us of the danger of letting our desire to do good decouple from basic physiology.

CHF affects millions. CRT is not a small intervention. To date, the biggest challenge for caregivers has been patient selection. I believe Echo-CRT ranks as the most important trial from ESC. Let's see if I can convince you.

Echo-CRT Trial Summary

Background Is Important

Previous CRT studies, which enrolled patients with QRS duration >120 or >130 ms, established benefit from biventricular pacing. Additional work confirmed the greatest benefit occurred in patients with QRS durations >150 ms and typical left bundle branch block (LBBB). Conflicting observational and small randomized trials were less clear for patients with shorter QRS durations—the majority of heart-failure patients. What's more, most cardiologists have seen patients with "modest" QRS durations respond to CRT.

Echocardiographic techniques, complicated as they are, held promise to identify mechanical dyssynchrony. CRT implanters have heard this before: the patient has a normal-duration QRS, but the echo shows marked dyssynchrony.

This was the question, then: Do CHF patients with mechanical dyssynchrony (determined by echo) but no electrical delay (as measured by the ECG) benefit from CRT?

Methodology Was Superb

Echo-CRT was an industry-sponsored (Biotronik) investigator-initiated prospective international randomized controlled trial. All patients had mechanical dyssynchrony by echo, QRS <130 ms, and an ICD indication. CRT-D devices were implanted in all patients. Blinded randomization to CRT-on (404 patients) vs CRT-off (405 patients) was performed after implantation. Programming in the CRT-off group was set to minimize RV pacing. The primary outcome was a composite of all-cause mortality or hospitalization.

Six Key Findings

1. Although entry criteria for the trial was a QRS duration <130 ms, the mean QRS duration of both groups was 105 ms.

2. The data safety monitoring board terminated the trial prematurely because of an increased death rate in the CRT group.

3. No differences were noted in the primary outcome.

4. More patients died in the CRT group (hazard ratio=1.8).

5. The higher death rate in the CRT group was driven by cardiovascular death.

6. More patients in the CRT group were hospitalized, due primarily to device-related issues.

These findings send clear and simple messages to all involved with treating patients with heart failure. My interpretation of Echo-CRT is as follows:

Do not implant CRT devices in patients with "narrow" QRS complexes.

The signal of increased death was strong. A hazard ratio of 1.8 translates to an almost doubling of the risk of death. This finding is unlikely to be a statistical anomaly, as it was driven by CV death. The risks of CRT in nonresponders are well-known and include: increased RV pacing, possible proarrhythmia from LV pacing, and the need for more device-related surgery. Patients who do not respond to CRT get none of the benefits but all the potential harms—an unfavorable ratio indeed.

Echo is not useful for assessing dyssynchrony in patients with narrow QRS complexes.

The problem isn't the echo techniques. It's more basic than that. CRT is only an electrical therapy. It preexcites the LV lateral wall. This works when the patient has delayed electrical activation there. I would refer readers to this excellent review article in the Journal of Cardiovascular Electrophysiology. Dr Samuel Asirvatham explains the concept of electropathy. He teaches us that the later the LV lateral wall is activated relative to the RV, the more the benefit of preexciting it is with an LV lead. That's why the chance for CRT increases with QRS duration, because—in most cases—a wide QRS means late activation of the lateral LV.

I wonder whether "resynchronization" was the right word for CRT. Perhaps it would have been better to call CRT something like . . . LVPED (LV preexcitation device).

Simple triumphs over complicated—CRT response best estimated with the old-fashioned ECG.

Isn't this refreshing? The ECG tells us a lot. Simple things like: in a right bundle branch block, the left ventricle is activated first; in LBBB, the LV lateral wall is last, and with a nonspecific ICD, all we can say is there's delayed conduction in either the His-Purkinje system or in ventricular muscle. And this: What does a normal QRS say? It says the wave front of activation activates the LV and RV simultaneously—just what we strive for with CRT in the first place.

CRT Benefit With Mild-Moderate QRS Prolongation Still Not Settled

I'd like to leave you with a useful lesson I learned from reading Dr Robert Myerburg's writings (here and here). He teaches us to make a distinction between trial entry criteria and the actual values of the cohort.

Consider how this applies to QRS duration: In COMPANION and CARE-HF, both powerful clinical trials that showed definitive CRT benefit, entry criteria required a QRS duration >120 ms (130 ms in CARE-HF). But the actual mean QRS duration of enrolled patients was 160 ms. A meta-analysis of CRT trials confirmed benefit at longer QRS durations and questioned it below 150 ms. CRT guideline recommendations incorporate study entry criteria, not the mean values of actual patients in the trial. I mention this distinction because patients enrolled in Echo-CRT had very narrow QRS complexes (105 ms). What to recommend in the common situation when a patient with a typical LBBB has a QRS duration straddling 130 ms is not entirely clear. The results of Echo-CRT might have been different had the actual QRS duration values been closer to 130 ms.


Echo-CRT resonates with me because it confirms basic cardiac physiology. Although the results are disappointing, findings that force caregivers to think about basic physiology are a good thing. Plus, in the practice of medicine, it's quite useful to know when not to do something.

The trial should not dampen enthusiasm for CRT. Rather, it should focus our attention to patient selection—and the value of the 12-lead ECG.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: