ICD Effectiveness: Do Patients Appreciate Their Limits? The Debate Continues CME

News Au= thor: Steve Stiles
CME Author: Désirée= Lie, MD, MSEd
Disclosures

From <= span style=3D'font-size:11.0pt;font-family:Arial;color:black'>Heartwire — a professional news service of WebMD

September 23, 2008 —= ; The benefits of implantable cardioverter-defibrillator= (ICD) therapy are overstated and their limitations are downplayed in clinical pra= ctice, considering limitations of the technology and of the clinical trials on whi= ch the ICD guidelines are based, according to a "state-of-the-art paper&q= uot; in the latest Journal of the American College of Cardiology [1]. The article also questions whether I= CD candidates are told enough about the treatment's limitations to make an informed decision about receiving a device.

"ICD therapy has cle= arly been shown to be effective in aborting sudden arrhythmic death. However, ... how much this capability, which modestly prolongs life, outweighs potential adverse effects on morbidity, quality of life, and the mode of death is less clear," write the authors, led by = Dr Roderick Tung (University of Califo= rnia, Los Angeles). "It is ethically imperative that we are honest with the data, so that we can be honest with our patients."

Appearing with the group's report is a counterpoint from Dr = Andrew E Epstein (University of Alabama at Birmingham) [2], who chaired the writing committee for the 2008 guidelines on "device-based therapy of cardiac rhythm abnormalities" sponsored by the Heart Rhythm Society, American College of Cardiology= , and American Heart Association [3].<= /p>

Until there are clinical trial data to guide improvements in how patients are selected for ICD thera= py and in its effect on quality of life, "we are left with the results of clinical trials that in the aggregate show improved survival in a broad selection of patients with left ventricular dysfunction and either demonstr= ated or anticipated risk for arrhythmic death," he writes. "Although we all freely admit that there are problems with ICDs, the weight of evidence supports their use for the indications listed [in the 2008 guidelines] in appropriate patients."

Both papers were published online September 22, 2008 and are slated for the journal's September 30 iss= ue.

The 2008 guidelines, rele= ased in May of this year, actually address some of the reservations Tung et al express about ICDs and their use. They include, for example, that the previ= ous set of guidelines from 2006 held a left ventricular ejection fraction (LVEF= ) of 40% as the threshold at which devices are recommended despite scant clinical trial support for their use when LVEFs exceed 35%. The more recent guidelin= es, in contrast, recommend device therapy specifically for patients who are consistent with the entry criteria of the various trials.=

"I think the benefit= s of ICDs are important, and there are many therapies that have equal or less benefit that we give to patients," Dr William H Maisel (Beth Israel Deaconess Medical Center, Boston, MA) said to heart<= /strong>wire after reading the two arti= cles.

"To me, the elephant= in the room is the large number of patients with strong indications for primary-prevention ICDs who don't get them," he said. "I see pati= ents literally every week who have indications for defibrillator therapy but don't get one, and then come into the hospital wi= th cardiac arrest."

"= A work in progress"

ICDs started out as a secondary prevention therapy for patients clearly at increased risk for sud= den death, observed Tung (who acknowledged the review was written before releas= e of the 2008 guidelines). Their indications later broadened to encompass patien= ts of progressively lower risk, such that today the overwhelming majority of primary-prevention recipients ultimately never need them, he observed for <= strong>heartwire. But even for them, he said, the devices are oft= en oversold.

"They make the most sense for someone who is a survivor of cardiac arrest, but even those patie= nts in the secondary prevention trials who received a defibrillator lived an average of two to four months longer than those who didn't get a defibrilla= tor. And those are the patients for whom you'd think the devices would be a no-b= rainer." Even if two to four more months is statistically significant, he said, whet= her it's clinically meaningful is open to question.

His group's article isn't saying ICDs don't provide a benefit, "we're just saying there are prob= lems with it, that we need to refine our selection of patients, and that they ar= e a work in progress," Tung said. "It's not about being anti-defibrillator. It's about considering what the most appropriate defibrillator use is. We're trying to acknowledge the fact that in the real= world, devices are fallible, that defibrillator implantation has up-front risks, a= nd that receiving shocks, whether appropriate or inappropriate, has a signific= ant impact on quality of life."

Other points made by the group:

The population implanted with ICDs today is broader — they are older, for example — than the patient groups entered into the clinical trials.
Those trials are fraught with issues that make them hard to interpret, such as differences in use of beta-blockers and antiarrhyth= mic drug therapy between actively treated patients and controls that could favor one group over the other.
The survival advantage with ICDs is overrated: of all the trial= s, the oft-cited Multicenter Automatic Defibrillator Implantation Trial (MADIT-2) showed the smallest gains with ICDs, 5.6 percentage points, while there was a five percentage point rise in heart-failure hospitalizations.
ICDs "reprogram" the mode of death from sudden death = to heart failure progression, which is "substantially more morbid.&q= uot;
A large proportion of ICD recipients receive inappropriate shoc= ks as well as shocks for ventricular arrhythmias that would have been nonsustained. Some evidence suggests that shocks may damage the myocar= dium and that ICD lead insertions themselves might be proarrhythmic. Also, device changeouts due to safety advisories inherently increase risk.

ICD cost-effectiveness analyses routinely represent best-case scenarios that "exaggerate the clinical benefit and underestimate the adverse effects, skewing the analyse= s in favor of defibrillators," according to Tung.

"With patients we of= ten refer to defibrillator therapy as a safety net, or like taking out an insur= ance policy," he said. "But insurance policies don't have the potential for causing harm. Up to 80% of patients will never receive any benefit from= the defibrillator, yet they still have to incur all the risks. We're not saying they're more harmful than good, but they need to know that there are some potential risks."

To express ICD benefits in terms of average increase in survival time is "misleading," Maisel said. "Many patients who get a primary prevention defibrillator live longer, and some patients might live considerably longer. To me it's less a= bout the average patient and more about emphasizing that one out of every six patients might live two or three years longer. Some patients aren't going to benefit, but some are going to benefit a lot."

"= When you're talking population medicine..."<= /o:p>

Epstein told heartwire that he "couldn't agree more that our first responsibility as physicians is to do the right thing for patients, and that includes identifying those who have the greatest expectation for benefit and trying to offer therapies that pose the least possible risk. So in that reg= ard, we have no disagreement whatsoever."

It's important to "g= et back to the bedside" and communicate with patients about both the poss= ible benefits and risks of ICD therapy, he said, noting that device therapy will= be recommended or not recommended on a case by case basis. "But when you'= re talking population medicine rather than individuals, you have to go by what= the clinical trials show."

In his counterpoint, Epst= ein addresses all or most of the concerns stated by Tung et al, often by citing= the 2008 guidelines but sometimes by offering a different slant on the same information:

Are we to ignore the results of the MADIT 2 study because the degree of benefit is not more than we would like? For better or worse,= the results of the MADIT 2 study are the best we have. <= /li>
To equate a 5% increase in hospitalization as counterbalancing = a 5% decrease in mortality ignores the point that if patients live longer a= nd quality of life can be preserved, the 'trade' for hospitalization will= be accepted by many patients.
The guidelines have long noted that "ICD cost-effectiveness would be greatest in patients at high risk of arrhythmic death and at = low risk for other causes of death," and "that cost-effectiveness would be improved by lowering the cost of the device and improving reliability and longevity."

Although ICDs malfunction, have proarrhythmic effects, and negatively affect quality of life, Epstein's article concedes, "they do not negate the r= esults of studies that in the aggregate show benefit."

Both the review from Tung= et al and his counterpoint conclude that "further work needs to be done," Epstein said. "And that further work is to refine identification of people who will benefit and who will not." But curre= nt guidelines, he added, have to go by the evidence we have now.

According to Maisel, the = two reports show that physicians are taking stock of ICDs and their value to patients, "and while its not explicitly stated, I think it comes in the wake the Medtronic Sprint Fidelis [ICD lead] recall and other device recalls" over the last three years or so. Also, he observed, no major trial that could potentially expand ICD use has been reported since Sudden Cardiac Death Heart Failure Trial (SCD-HeFT).

"The devices had in = many ways been the poster child for modern medicine," Maisel said, "and now the pendulum has swung in the other direction because of the device performance issues. I think it's critical to get past that if the therapy is going to save as many lives as it can."

Coauth= or Dr. Mark E Josephson (Beth Israel Deaconess Medical Center) has disclosed receiving educational grants and honoraria from Medtronic. Dr. Maisel has been a consultant to the US Food and Drug Administration, is a member of the Medicare Coverage Advisory Committee, and has disclosed no relevant financial relationships.

Sources

Tung R, Zimetbaum P, Josephson ME. A critical appraisal= of implantable cardioverter-defibrillator therapy for the prevention of sudden cardiac death. J Am Coll Cardiol. 2008;52:1111-1121. doi:10.1016/j.jacc.2008.05.058.
Epstein AE. Benefits of the implantable cardioverter-defibrilla= tor. J Am Coll Cardiol= . 2008;52:1122-1127. doi:10.= 1016/j.jacc.2008.06.035.
Epstein AE, Dimarco JP, Ellenbogen KA, et al. Heart Rhythm 2008;5:e1-62. published online before print May 21, 2008.= PMID: 18534360.

The complete contents of H= eartwire, a professional news service of WebMD, can be found at www.theheart.org<= /span>, a Web site for cardiovascular healthcare professionals.

Clinical Context

In the past 15 years, the annual insertion of ICDs has increased 15-fold. The recent 2008 practice guidelines on device-based therapy for the implantation of ICDs and pacemak= ers are considered the current evidence in making recommendations, with the hel= p of independent experts. The guidelines include the identification of patients = who are most likely to benefit from ICD implantation.

This is a critical apprai= sal of evidence that led to the 2008 practice guideline recommendations with a = rebuttal from 1 author of the guidelines. The critical appraisal questions whether t= he benefits have been overrated, the risks underestimated, and whether cost-effectiveness has been overestimated at the current cost of $30,000 per device.

Study Highlights=

Overestimation of benefit and underestimation of risk

The authors noted that 3 of 4 trials showing a mortality benef= it for ICDs also had disproportionate rates of beta-blocker treatment in= the ICD group as an argument that benefits have been overrated.
The comment that clinical benefit was not observed in patients with ejection fractions greater than 35% and less than 20%.
They observed that the unadjusted improvement in mean survival= was only 0.21 years (2.6 months), with a number needed to treat of 11 per life saved in 1 trial, with the cost of 4% absolute increase in the r= isk for hospitalization for patients with ICD implants (60% vs 56%; <= span style=3D'font-family:Arial'>P =3D .04) and that the cost = for the prolongation in life was $85,522.
They argued that smaller trials for secondary prevention using ICDs failed to demonstrate significant benefits and that similar ineq= uity in beta-blocker use was observed in the ICD groups.
They cited a meta-analysis showing only 4.4 months of prolonga= tion in life during 6 years of follow-up for ICD implantation, with very modest benefits across 3 trials.
They criticized another study for high rates of beta-blockade = in the ICD group.
They noted a high incidence of death in 1 study within 30 days= of device implantation with high morbidity form sternal and wound infections.
In the postinfarct setting, they noted unclear data bout the timing of device placement after myocardial infarction and questionab= le benefit for patients with New York Heart Association Class III heart failure.
They noted no benefit for patients with nonischemic cardiomyopathy.
25% of patients with ICDs receive inappropriate shock; such treatment is cited as being harmful, with a reduction in quality of l= ife and higher rates of depression and anxiety.
Inappropriate shocks were also cited as being responsible for potentially lethal effects, with an hazard= ratio of 1.97 in 1 study.
Local lead effects and mechanical irritation and late fibrosis= may be potential mechanisms for ventricular tachycardia.
Hardwire malfunction remained a risk for manufacturing defects, with the potential for lethal effects.

Rebuttal to opinion

The rebuttal reiterates the use of best evidence to develop the 2008 guidelines for device implantation.
Most of the evidence was based on trials that showed improved survival in a broad selection of patients with left ventricular dysfunction.
The guidelines already recommend that patients being considered for ICD implantation receive optimal therapy including appropriate us= e of beta-blockers.
The appropriate use of beta-blockers in trials showing a benef= it for ICDs does not invalidate their beneficial effects.
The guidelines already recommend adequate evaluation of heart failure status and plan for ongoing care in the consideration of ICD implantation.
The guidelines do recognize that ICD implantation was not a substitute for resuscitation for cardiac arrest and note that ICDs can have proarrhythmic effects and negative effects on quality of life. <= o:p>
The rebuttal notes that patient selection was a key to best outcomes of ICD implantation and that the weight of current evidence supported the use of ICDs in carefully selected patients. =
It recommends that future research be directed at recognizing patients who are unlikely to benefit from ICD therapy.

Pearls for Practice

The evidence on which the current ICD implantation practice guidelines are based may overestimate the benefits and underestimate t= he risks.
However, there remains good evidence to consider ICD implantati= on for selected patients with the appropriate criteria who receive optimal medical treatment.