DAVID's shadow: MADIT II substudy supports theory that ventricular pacing in ICD patients promotes a

Laurent Castellucci

May 22, 2003

Thu, 22 May 2003 22:15:00

Washington, DC - Despite the vast improvement in mortality for those implanted with a device, the ICD population of the Multicenter Autonomic Defibrillator Implantation Trial II (MADIT II) had a higher rate of hospitalization for heart failure than those without ICDs. MADIT II investigators have found that this initially surprising result may be explicable in light of the findings from the Dual Chamber and VVI Implantable Defibrillator (DAVID) trial: that ventricular pacing in patients who do not need it worsens heart failure by creating dyssynchrony in the heart. Looking back at the ICD cohort of MADIT II in a substudy, investigators presented evidence at the late-breaking trials of the NASPE 2003 Scientific Sessions that the same effect seen in DAVID was at work in MADIT II.

"In a substantial number of patients you can just avoid ventricular pacing," said Dr Jonathan S Steinberg (St Luke's-Roosevelt Hospital Center, New York), who presented the substudy. "If it's not indicated, don't do it."

Of the original 742 patients randomized to the ICD arm of MADIT II, Steinberg's group took the 567 (76%) patients for whom data from a final interrogation of the device were available. Patients in this group had a mean age of 64 years, were primarily male, had previous heart failure symptoms (class 2-4), and a mean EF of 23.4%. Just under half of them (46%) had a dual-chamber ICDs implanted.

Investigators segregated the patients into two groups based on the cumulative time they were paced: low pacing meant they were paced for 0% to 50% of the time (198 patients) during the trial, and high pacing meant they were paced 51% to 100% of the time (369 patients). Steinberg pointed out that the 50% marker was something of an arbitrary cutoff, since the distribution was quite bimodal, with the majority of patients grouped either around the 90%-to-100% mark or the 0%-to-10% mark.

In general, the highly paced patients were sicker to start with: they had higher NYHA classes, lower ejection fractions, and more left bundle branch block, and they were less likely to have a normal PR interval.

The investigators looked at three outcomes: new or worsened heart failure (119 events, 21%), heart failure or death (130 events, 23%), and appropriate ICD therapy (142 events, 25%). For each of these outcomes, the highly paced group had a higher rate of occurrence than the less-paced group. High pacing was an independent predictor of these events, and under multivariate analysis, being highly paced resulted in a significant increase in the risk of heart failure or death and a strong trend to more ICD therapy.

Predictive value of ventricular pacing


HR (95% CI)


Heart failure

2.1 (1.3-3.4)

Heart failure or death

1.9 (1.2-3.0)

ICD therapy

1.5 (1.0-2.2)


1.7 (0.8-3.9)
To download table as a slide, click on slide logo below

Substudy limitations

Steinberg stressed repeatedly that the analysis is severely limited by being a substudy. MADIT II had no prespecified ICD pacing program. Not all of the enrolled ICD patients were analyzed for this study, because there was no final interrogation of the ICD available, often due to death. Because it was not a prospective, randomized trial, the patient populations were not identical, with the highly paced patients being significantly sicker than those who were paced less often.

"It is just a substudy and so doesn't really prove anything," Steinberg told heartwire , "The implicit conclusion is that this is like DAVID. Sort of a pre-echo of what we eventually saw there." Steinberg went on to say that it does provide a plausible explanation for the rise in heart failure hospitalizations seen in MADIT II.

Steinberg said that he's convinced pacing should be avoided in ICD patients who don't have indications for it. "There will be patients you can't avoid it in," he told heartwire , "such as those already with severe LV dysfunction, and then you have to put it in. The real challenge will be in those patients who sort of teeter on the edge. Then you have to wonder if dual-chamber pacing would help."

Steinberg couldn't answer whether or not biventricular pacing would help in these patients, "There are no data on that and it needs to be tested."

Dr Bruce L Wilkoff (Cleveland Clinic Foundation, OH), lead author of the DAVID trial, told heartwire he agreed that there needs to be a study examining the possible effectiveness of biventricular pacing in these sorts of patients and to find out whether implanting these devices can actually slow the progression of heart failure.

But even before those data are in, Wilkoff suggests there would be benefit in changing how physicians program patients. "We learned everything about how to program pacing in secondary-prevention patients, and now we keep expanding out to all kinds of primary-prevention indications, and it isn't the same patient population. We need to program these people differently," Wilkoff said.

Wilkoff told heartwire there are thousands of pacing combinations, and it is too much to think that every doctor is programming patients according to the best evidence-based medicine information. "It's too complicated," Wilkoff said, "We're not even using the experience we do have."

Related links

1. [HeartWire > News; Mar 19, 2002]

2. [HeartWire > News; Dec 19, 2002]


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.