Calling death's bluff: implantable defibrillators can save thousands of lives--but which ones?

Some hearts seem to be beating normally one moment, but then quiver for an instant and fall still forever. In these cases, the heart's electrical circuitry goes haywire, and its contractions accelerate until the organ flails uselessly and shuts down. Sudden cardiac death awaits nearly half a million people nationwide in the next 12 months. Once a person's heart stops, there is usually little that any hospital can do.

However, doctors do have an earlier chance to intervene. By implanting a specialized device into a person's chest, they can equip the heart to recover instantly when death comes knocking. When the electrical leads of these so-called implantable cardioverter-defibrillators, or ICDs, detect an abnormal heart rhythm, the attached pager-size generator delivers a shock that restores the normal cadence.

Use of these devices began in the 1980s and has increased exponentially in recent years. This year, U.S. doctors will surgically implant some 200,000 defibrillators at a cost of about $30,000 to $50,000 per operation.

Yet most of the devices will never see action. About four-fifths of defibrillators sit quietly in people's chests until their batteries conk out. Replacing a battery, which is necessary every 5 years or so, requires another round of surgery.

Defibrillators also can be cantankerous. They sometimes mistake mild arrhythmia, or rhythm disruption, for a deadly spasm and shock the heart unnecessarily, causing pain but no permanent harm. And while some people die because a device fails to fire when it should--a problem that spawned massive recalls of certain models last year--many more perish because no one recognized that a defibrillator could have saved their lives.

Doctors would like to improve their ability to distinguish people who truly need defibrillators from those who don't. The current method uses a crude measure: the heart's mechanical efficiency at pumping blood. That value, called ejection fraction, has no direct relationship with the heart's electrical circuitry.

Recent studies suggest that measures of the heart's electrical function and its responsiveness to the nervous system may reveal additional information about the heart's vulnerability to arrhythmia. Medical researchers recently concluded the first systematic test in which they weighed one such promising factor--an electrical characteristic called T-wave alternans--for deciding who should get a defibrillator. They plan to announce their results at a medical meeting this November.

If such trials prove that T-wave alternans and other new measures can supplement ejection fraction in clinical evaluation of patients, doctors may save more lives even though they implant fewer devices than they do now.

HIGH PRICE AT THE PUMP With each contraction of the heart, the organ's left ventricle squeezes blood into the body's arteries. A healthy chamber pumps out half to three-quarters of its contents, while a defective heart might expel 30 percent or less. This ejection fraction, expressed on a 100-point scale, can be gauged painlessly by performing a sonogram of the chest.

An insult such as a myocardial infarction--commonly called a heart attack--causes muscle damage that can instantly trigger fatal arrhythmia, or it can permanently reduce ejection fraction. The same muscle damage can also interfere with the conduction of electrical impulses through the heart, thereby increasing the chance that a subsequent life-threatening arrhythmia will occur.

"Ejection fraction has been studied extensively and is dearly associated closely with sudden death" says cardiac electrophysiologist David S. Rosenbanm of Case Western Reserve University's MetroHealth Campus in Cleveland.

The form of arrhythmia that usually triggers cardiac arrest is called ventricular fibrillation. In that event, uncoordinated muscle contractions cause the ventricles to flutter ineffectually. The malfunction develops when electrical impulses that control the pace and rhythm of the ventricles' contractions become irregular.

"The lethal event is electrical--a catastrophic disturbance in rhythm," says Richard Verrier, a cardiovascular electrophysiologist at Beth Israel Deaconess Medical Center in Boston.

That's precisely the problem that defibrillators are designed to solve. People with an ejection fraction lower than 30 have been shown to benefit from defibrillators, so physicians currently use that value as a threshold when deciding who should get a device. With that cut-off and other criteria, such as previous ventricular fibrillation or advanced heart failure, about 1.6 million people in the United States are eligible, according to Minneapolis-based device maker Medtronic.

Yet ejection fraction reveals limited information about the heart's electrical integrity. "You can have a lethal disturbance without having any [preexisting physical] damage to the heart," Verrier notes. Or, the organ can have damage that never leads to an arrhythmia.

"You're using a measure of mechanical abnormality to predict an electrical event, so you can't expect it to be overwhelmingly accurate," says Rosenbaum. "Doctors have to implant 17 defibrillators to save one person. The benefit is real, but it's modest and has to be weighed against the risk."

Hazards of the surgery include infections, clots, and internal bleeding. And when devices fire unnecessarily, patients "get a flurry of shocks," says cardiologist J. Thomas Bigger Jr. of Columbia University's New York--Presbyterian Hospital.

PROBING THE CIRCUITRY In 1994, researchers led by Rosenbaum determined that hearts that go into arrest tend to have a history of beat-to-beat fluctuations in the size of the T wave. That wave is a small blip on the familiar trace of an electrocardiogram (EKG), and it reflects the electrical recharging of the heart between beats.

Today, doctors examine T waves by attaching electrodes to a patient's chest as he or she exercises on a treadmill or stationary bicycle. As their heart rate rises, some people develop an alternating pattern, or alternans, in the amplitude of the T wave. Such a propensity to develop T-wave alternans is considered a harbinger of the problem that defibrillators protect against.

Many people meet the current criteria for defibrillator implantation but never get a device. Sometimes, doctors conclude that a device is probably unnecessary; in other cases, patients can't be convinced that they need it. Broader use of T-wave alternans testing might combat both those obstacles by giving doctors and patients additional information, Bigger says.

Ten years after Rosenbaum's initial report on T-wave alternans, Bigger and his colleagues published a report on 177 people whose low ejection fractions and other aspects of their health status qualified them to receive defibrillators. Among the two-thirds who had either T-wave alternans or an ambiguous T-wave pattern, 17.8 percent died within the 2-year study period. By contrast, the 2-year death rate was only 3.8 percent among the group with a normal T-wave pattern.

In follow-up research, the investigators analyzed T-wave patterns from 549 participants, including those from the first study and others who appeared to be at slightly lower risk of sudden cardiac death. Overall, only 69 of these volunteers had received a defibrillator.

Afterward, "people who had the normal test had a very low incidence of death or sustained arrhythmias," says Bigger. Within 2 years among those 189 people, one person with a normal T-wave pattern died from an arrhythmia and one died from an unrelated cancer. Two other volunteers in this group survived serious arrhythmias.

But among the 360 volunteers with atypical T-wave alternans, there were 38 deaths and 9 nonfatal arrhythmias within 2 years, Bigger's team reported in the Jan. 17 Journal of the American College of Cardiology.

Notably, in the second analysis, the researchers included some volunteers whose ejection fractions were as high as 40--people not currently considered defibrillator candidates.

When T-wave alternans results are normal, comments Rosenbaum, "we're doing the patient a big favor by not implanting a defibrillator." Using the new measure to inform clinical decisions, he and other researchers argue, could prevent thousands of needless procedures in patients who are currently advised to get the devices.

But electrophysiologist Rachel Lampert of the Yale University School of Medicine cautions, "If you're going to narrow that group, you want to be pretty sure that anyone you're excluding is not going to die suddenly." Only a clinical trial rigorously comparing criteria for implants could provide such assurance.

Rosenbaum says that he and his colleagues have just completed the first clinical trial to use T-wave alternans to guide implantation of defibrillators. He won't yet reveal the results but says that he will present them this November at a meeting of the American Heart Association. Defibrillator manufacturer St. Jude Medical of Saint Paul, Minn., supported the study.

Such studies "will certainly shed important light on the T-wave alternans question" says Michael R. Gold of the Medical University of South Carolina in Charleston. However, he adds, those studies are already somewhat outdated because certain procedures that were standard when the studies began, such as an invasive procedure called electrophysiology testing, have since fallen out of favor.

Some doctors are incorporating T-wave analysis into clinical practice even as they continue to collect experimental data. In a collaborative effort, heart specialists in Cincinnati and Ann Arbor, Mich., evaluated T-wave alternans in 768 consecutive patients whose ejection fractions had fallen, after a heart attack, to 35 or less.

Each doctor in the study decided on a case-by-base basis whether to recommend a defibrillator. Collectively, the investigators implanted the devices in 30 percent of the 254 volunteers who had normal T-wave patterns and in 62 percent of the 514 people with T-wave alternans. The doctors then kept track of the patients for an average of 18 months.

In the May 2 Journal of the American College of Cardiology, Paul S. Chan of the Veterans Affairs Ann Arbor Healthcare System and his colleagues reported that, other factors being equal, patients with atypical T-wave alternans were 2.3 times as likely to die from arrhythmia as those with normal electrical activity were.

Bigger proposes a two-step method for evaluating patients who might need defibrillators. First, anyone with an ejection fraction of 40 or lower should be tested for T-wave alternans. Making the ejection fraction threshold higher than the current standard of 30 would identify more people who are at risk of arrhythmia, he says.

Then, patients with normal T-wave patterns--presumably about one-third of those who get tested for T-wave alternans--would be discouraged from getting a device. The latter step, says Bigger, would "pull out those who are at such low risk that they'd have very little if any benefit from an ICD."

"T-wave alternans has real merit for keeping people who don't need devices from getting them and convincing people who need them to get them," he says.

NERVOUS ENERGY As promising as the test for T-wave alternans is, it's unlikely to be the final word in predicting cardiac arrest or in assigning defibrillators to patients. Researchers anticipate that adding new tests to cardiac care will make for even better decisions.

"I think T-wave alternans has proven itself," says Verrier. But he adds, "the field is going [toward] what we call the multi-parameter approach."

In that vein, doctors might analyze each patient's ejection fraction, T-wave pattern, and other factors to calculate "a risk-assessment score for sudden cardiac death," he says. Similar scoring systems exist for heart attack risk and other health threats.

Several of the most promising parameters reflect what electrophysiologist Georg Schmidt calls the "software of the heart," or the organ's responses to the nervous system.

Like T-wave alternans, these parameters can be gleaned from EKG data. Each is calculated by a complex algorithm that analyzes the time intervals between many consecutive heartbeats.

These parameters offer a window into the baroreflex, a biological response to short-lived perturbations in blood pressure. When blood pressure momentarily falls in a healthy person, which is a common and inconsequential occurrence, part of the nervous system instantaneously commands the heart to accelerate as compensation. Over several seconds, the heart rate returns to normal.

In some people, "baroreflex doesn't work very well. A poor baroreflex is a very strong indicator of risk for cardiac mortality," says Schmidt of Technical University Munich in Germany. He and other researchers have studied this dangerous bug in the heart's software by using several measures.

In a 2003 study, for example, Schmidt and his collaborators measured baroreflex over 24 hours by using a marker called heart-rate turbulence. They found that an abnormal baroreflex was linked to death nearly six times as often as was a normal value. An ejection fraction below 30 percent, by comparison, multiplied risk by a factor of only 4.5.

All 1,455 volunteers had had heart attacks before participating in the study, but most didn't qualify for defibrillator implantation under current guidelines. That suggests that the new measure could be used to save lives in people whom current methods don't identify as endangered, Schmidt says.

While the heart's acceleration after a blood pressure drop is one facet of heart-rate turbulence, "the deceleration pattern seems to be more meaningful," Schmidt says.

In subsequent research in 1,256 other volunteers, Schmidt's group confirmed that what it calls deceleration capacity is a powerful predictor of whether a person will die within 2 years. The results appeared in the May 20 Lancet.

Doctors might weigh intermediate deceleration-capacity scores along with other relevant factors, including ejection fraction and T-wave alternans, Schmidt says. He argues that people with severely reduced deceleration capacity need a defibrillator, even if their ejection fraction appears normal. But people with normal deceleration capacity and an ejection fraction greater than 30 percent--which would include two-thirds of the study volunteers--need not receive a defibrillator.

"Deceleration capacity may be an important sudden-death-risk stratifier that will complement T-wave alternans," Verrier comments.

A third measure of nervous system function, called heart-rate variability, is also being examined. For 3 years, Eric Rashba of the University of Maryland at Baltimore and other investigators studied 274 volunteers with low ejection fractions but no past heart attacks. High variability hallmarks a heart that, in spite of its mechanical problems, has relatively little risk of going into arrest, the researchers reported in the March Heart Rhythm.

"It makes sense to combine the information from all these tests,' Schmidt says. "What we need for all these new risk predictors is studies," he says. "Large trials. Intervention trials."

The vanguard of that data wave, says Bigger, may come in November with the results of the pending trial of T-wave alternans as a predictor of heart problems.