NATURE Vol 465 20 May 2010 parasites (Toxoplasma, Leishmania and active compounds — and an earlier, partial y These reports1,2 offer tremendous opportunities trypano somes) and on replicating human cel described set8 identified in a high-throughput to develop the next generation of antimalarial lines, and found that most of the compounds screen against P. falciparum — should be a first drugs. They also sound a call for the academic
Doi:10.1016/j.amjcard.2005.02.02Implantable Cardioverter-Defibrillator Efficacy in
Patients With Heart Failure and Left Ventricular
Dysfunction (from the MADIT II Population)
Wojciech Zareba, MD, PhD, Katarzyna Piotrowicz, MD, Scott McNitt, MS, and Arthur J. Moss, MD, for the MADIT II Investigators The Multicenter Automatic Defibrillator Implantation
The Multicenter Automatic Defibrillator Trial II
Trial II demonstrated a significant 31% reduction in
(MADIT II) demonstrated that postinfarction pa- the risk of mortality in postinfarction patients with low
tients with greatly reduced left ventricular function ejection fraction (EF <30%). Recently, results from the
benefit from prophylactic therapy with implantable Sudden Death in Heart Failure Trial indicated that a
cardioverter-defibrillators There was a 31% subgroup of patients with New York Heart Associa-
reduction in the risk of all-cause mortality in patientsrandomized to an ICD compared with patients receiv- tion (NYHA) class III heart failure had less benefit
ing conventional therapy with high usage of ␤ block- from an implantable cardioverter-defibrillator (ICD)
ers, angiotensin-converting enzyme inhibitors, and sta- than patients with less advanced heart failure. This
tins. Recently, findings from the Sudden Death in study evaluates the association between NYHA class,
Heart Failure Trial (SCD-HeFT) were presented, and EF, and blood urea nitrogen (BUN) levels as measures
ICD therapy effectively reduced mortality in patients of heart failure and left ventricular dysfunction, and
with ischemic and nonischemic The ICD benefit in reducing mortality as well as the asso-
magnitude in the mortality risk reduction was 23%, ciation of these parameters with ICD therapy for ven-
similar to the 31% reduction observed in MADIT II.
tricular tachyarrhythmias. NYHA class I was identi-
However, SCD-HeFT patients with New York Heart fied in 442 patients (36%), class II in 425 (35%), and
Association (NYHA) class III heart failure did not class III in 350 patients (29%). EF
receive significant benefit from ICD therapy (hazard <20% was present
ratio 1.16; 95% confidence interval 0.84 to 1.61), in 472 patients (38%), EF of 21% to 25% in 359
whereas the benefit was evident in patients in NYHA patients (29%), and EF of 26% to 30% in 401 patients
class II. Previously, data from the the Antiarrhythmics (33%). BUN <25 mg/dl was present in 850 patients
Versus Implantable and trials, (70%) and >25 mg/dl in 368 patients (30%). Patients
which included patients with a higher ejection fraction with higher NYHA class and BUN had higher mortal-
(EF) than those in MADIT II and SCD-HeFT, indi- ity (34%) and a higher risk of arrhythmic events
cated that patients with more advanced left ventricular (33% to 35%) than patients in lower functional groups
dysfunction benefitted more from ICDs than those (16% to 20%). EF did not differentiate the risk. There
with higher EFs. Therefore, we conducted a detailed was no evidence for significant interactions between
secondary analysis of the MADIT II population to mortality, ICD therapy, and tested parameters. In
determine the efficacy of ICD therapy in high-risk conclusion, patients with more advanced heart failure
subgroups defined by NYHA functional class, EF, andblood urea nitrogen (BUN) levels.
have a higher risk of mortality and arrhythmic events
than patients with less severe disease. However,
Details of the MADIT II have been previously there is no significant difference in the benefit from
Briefly, MADIT II enrolled patients who ICD therapy among the above subgroups of patients
had a myocardial infarction within 1 month and left in the Multicenter Automatic Defibrillator Implanta-
ventricular EF ⱕ30%. Patients who had undergone tion Trial. 䊚2005 by Excerpta Medica Inc.
recent revascularization procedures, those in NYHA (Am J Cardiol 2005;95:1487–1491)
class IV at enrollment, and patients with major co-morbidities were excluded. Patients were randomizedto ICD or conventional therapy (3:2 ratio), with highusage of ␤ blockers, angiotensin-converting enzymeinhibitors, and statins in both groups.
Patients had data collected with regard to left ven- From the Heart Research Follow-up Program, Cardiology Unit, Depart- tricular EF, NYHA class, and BUN levels at enroll- ment of Medicine, University of Rochester Medical Center, Rochester,New York. This report was supported by a research grant from ment. EF, reflecting myocardial damage and the re- Guidant Corporation, St. Paul, Minnesota, to the University of Roch- modeling process after myocardial infarction, was ester School of Medicine and Dentistry. Dr. Zareba's address is: Heart divided into 3 subcategories: ⱕ20%, 21% to 25%, and Research Follow-up Program, Box 653, University of Rochester Medical 26% to 30%. The NYHA class, reflecting functional Center, Rochester, New York 14642-8653. E-mail: wojciech_zareba@ response to left ventricular decompensation, was cat- urmc.rochester.edu. Manuscript received December 28, 2004; re-vised manuscript received and accepted February 9, 2005.
egorized (based on direct examination of patients by 2005 by Excerpta Medica Inc. All rights reserved.
0002-9149/05/$–see front matter The American Journal of Cardiology Vol. 95 June 15, 2005 TABLE 1 Two-year Cardiac Event Rates in Studied Subgroups of MADIT II Patients
Mortality in Conventional Arm Cardiac Events in ICD Arm The p values are by log-rank statistics comparing 2-year event rates from Kaplan-Meier curves.
SCD ⫽ sudden cardiac death, classified as described by Greenberg et FIGURE 2. Cumulative probability of appropriate ICD therapy in
FIGURE 1. Cumulative probability of survival in MADIT II patients
MADIT II patients randomized to ICD therapy by NYHA class (A ),
randomized to conventional therapy by NYHA class (A), EF (B), and
EF (B), and BUN levels (C). The p value was estimated by log-rank
BUN levels (C). The p value was estimated by log-rank statistics.
statistics. VF ⴝ ventricular fibrillation; VT ⴝ ventricular tachycardia.
1488 THE AMERICAN JOURNAL OF CARDIOLOGY姞
FIGURE 3. Cumulative probability of survival in MADIT II patients randomized to conventional versus ICD therapy: patients in (A)
NYHA class, (B) NYHA class II, and (C) NYHA class III; and patients with (D) EF of 26% to 30%, (E) EF of 21% to 25%, and (F) EF of
<21%. HR ⴝ hazard ratio.
trained study nurses) into 3 subgroups: class I, II, and randomized to ICD treatment. Antitachycardia pac- III, separately. BUN levels were dichotomized at ⱕ25 ing was programmed on at enrollment in 41% of and ⬎25 mg/dl based on our previous ICD patients.
indicating that this cutoff is highly predictive of car- Kaplan-Meier curves with log-rank statistics were diac events in postinfarction patients.
used to analyze time to end point events. Cox propor- The analysis was focused on ICD efficacy, i.e., tional-hazards analyses were used to determine hazard comparison of survival in patients randomized to ICD ratios for risk of mortality and arrhythmic events in therapy versus conventional treatment in each evalu- respective subgroups with interaction terms tested for ated subgroup of We also evaluated the their significance. A p value ⬍0.05 was considered association between the aforementioned selected variables and mortality and sudden cardiac death in The MADIT II population consisted of 1,232 pa- conventionally treated patients. Sudden cardiac death tients (1,040 men and 192 women; median age 65 was based on modified Hinkle-Thaler criteria used years). Patients had documented myocardial infarction by an independent review The association on average 5 years before enrollment. Their mean EF between these clinical variables defining high-risk was 23% (median 25%). Half of the patients had subgroups and appropriate ICD therapy (antitachy- undergone coronary artery bypass grafting. Patients cardia pacing or ICD shocks) for ventricular tachy- were treated pharmacologically with ␤ blockers cardia or ventricular fibrillation (documented by interrogated ICDs) was also analyzed in patients (70%), and statins (66%). The 3:2 randomization lead NYHA classes I and II. The risk of TABLE 2 Cox Proportional Hazards Ratios for the Appropriate Implantable
ventricular tachycardia or ventricular Cardioverter-defibrillator (ICD) Therapy in Studied Subgroups of MADIT IIPatients fibrillation, which required appropri-ate ICD therapy, followed a similar Appropriate ICD Therapy* pattern although patients in NYHA class II had a 2-year risk ofarrhythmic events closer to that of patients in NYHA class III compared with patients in NYHA class I Antitachycardia pacing was administered as therapy, converting arrhythmia without a need for ICD shock in 12% of patients in NYHA class I, in 13% patients in class II, and in 15% patients in class III. Cox proportional-hazards analysis dem- *Appropriate ICD therapy defined as antitachycardia pacing or ICD shock to terminate ventricular onstrated a 40% higher risk for ap- tachycardia or ventricular fibrillation.
propriate ICD therapy in patients ⫽ confidence interval; HR ⫽ hazard ratio.
with NYHA class II than class I heartfailure, and an 80% higher risk in TABLE 3 Cox Proportional Hazards Ratios for the Implantable Cardioverter-
patients with NYHA class III than defibrillator (ICD) Effectiveness (ICD vs conventional therapy) in Studied class I heart failure Ele- Subgroups of MADIT II Patients vated BUN also indicated a 40% in- Mortality—ICD vs Conventional creased risk for appropriate ICD ther- apy compared with patients with BUN ⱕ25 mg/dl.
The effectiveness of ICD therapy in studied subgroups was evaluated by comparing mortality between both arms of the study. In all sub- groups, patients treated with ICDs had better survival than those taking conventional therapy and and There was no evi- dence of a significant p value for interaction between mortality, ther- Abbreviations as in apy, and examined subgroups, afinding indicating that patients in dif-ferent groups by NYHA class, EF, to 490 patients being randomized to conventional ther- and BUN levels had similar degrees of mortality re- apy and 742 to ICD therapy.
duction with an ICD.
Subgroups of patients by NYHA class consisted of Additionally, we analyzed ICD effectiveness in 442 patients (36%) in class I, 425 (35%) in class II, patients having none, 1, 2, or 3 studied risk factors and 350 (29%) in class III. There were 472 patients dichotomized as NYHA ⬎II versus ⱕII, EF ⱖ25% (38%) with EF ⱕ20%, 359 (29%) with EF 21% to versus ⬍25%, and BUN ⱖ25 versus ⬍25 mg/dl.
25%, and 401 (33%) with EF 26% to 30%. Subgroups There were 382 patients with none of these factors of patients by BUN levels consisted of 850 patients with a hazard ratio of 0.69, 452 patients with 1 factor (70%) with BUN ⱕ25 mg/dl and 368 (30%) with with a hazard ratio of 0.67, 274 patients with 2 factors BUN ⬎25 mg/dl.
and a hazard ratio of 0.63, and 95 patients with all 3 shows the 2-year cumulative probability factors present and a hazard ratio of 0.67. There were (from Kaplan-Meier curves) of cardiac events in pa- no significant differences among these hazard ratios.
tients randomized to conventional and ICD therapy, In this study, we analyzed 3 parameters reflecting NYHA class III had 2 times higher mortality than signs and symptoms of more advanced heart failure: patients in NYHA class I or II Elevated EF, NYHA class, and BUN levels. They appear to BUN levels identified a similarly high (34%) risk represent somewhat different features of decompen- subset of patients Although there was a sated hearts, but they did not correlate well with each trend toward increasing mortality with decreasing EF other (only 95 patients had all 3 parameters present).
the separation was less significant than EF is a reflection of a damaged and remodeled heart NYHA class and BUN levels.
after myocardial infarction. The NYHA class provides Sudden cardiac death was observed 2 times more insight into functional consequences of myocardial often in patients in NYHA class III than in patients in and circulatory changes. BUN is a measure of renal 1490 THE AMERICAN JOURNAL OF CARDIOLOGY姞
failure than in patients with lower classes. Because ourfindings from MADIT II also demonstrate that patientsin NYHA class III benefit at least to the same extent aspatients with less expressed heart failure, the questionneeds to be asked: what is the reason for the discrepancybetween the findings from those studies? Patient popu-lations appear very similar, but SCD-HeFT includedpatients with EFs 31% to 35%. This difference wasunlikely to contribute to the observe differences in ICDeffectiveness by NYHA class. However, there is a dif-ference in the ICD therapy administered in SCD-HeFTand MADIT II trials. In the SCD-HeFT, ICD therapyconsisted of ICD shocks only without antitachycardiapacing enabled, whereas in MADIT II, both antitachy-cardia pacing and ICD shocks were used. Antitachycar-dia pacing successfully terminated most documentedtachyarrhythmia episodes (58%) in MADIT II withoutthe need for ICD shocks in patients who had antitachy-cardia pacing Similar findings were reportedby the Antiarrhythmics Versus Implantable Defibrilla-tors It is conceivable that patients withNYHA class III heart failure in the SCD-HeFT hadventricular tachycardia or fibrillation episodes alreadylong and advanced enough that ICD therapy had lessereffectiveness in converting rhythm. However, in MADIT FIGURE 4. Cumulative probability of survival in MADIT II patients
II, tachyarrhythmias in patients in NYHA class III were randomized to conventional versus ICD therapy by BUN levels.
treated in the earlier stage before full "warm-up" and Abbreviation as in
development of nonconvertible ventricular fibrillation.
Fifteen percent of patients in NYHA class III in the function that correlates well with a degree of decom- MADIT II had antitachycardia pacing therapy as the pensation in heart only modality terminating arrhythmia episodes.
Patients with more advanced disease measured by NYHA class and BUN levels had a 2 times highermortality than those in the respective comparison 1. Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS, Daubert
groups. These expected associations were paralleled JP, Higgins SL, Brown MW, Andrews ML. Prophylactic implantation of adefibrillator in patients with myocardial infarction and reduced ejection by a substantially increased risk of arrhythmic events fraction. N Engl J Med 2002;346:877– 883.
in these more advanced groups, indicating that they 2. Bardy GH, Lee KL, Mark DB, Poole JE, Packer DL, Boineau R, Domanski M,
should benefit from ICD therapy. In addition, our Troutman C, Anderson J, Johnson G, et al. Amiodarone or an implantablecardioverter-defibrillator for congestive heart failure. N Engl J Med 2005;352: analysis confirmed that the benefit in patients with more advanced disease was not significantly different 3. The Antiarrhythmics versus Implantable Defibrillators (AVID) Investigators.
from the benefit observed in patients with less ad- A comparison of antiarrhythmic-drug therapy with implantable defibrillators inpatients resuscitated from near-fatal ventricular arrhythmias. N Engl J Med vanced disease: a reduction in the risk of mortality in the order of 28% to 35% regardless of differences in 4. Moss AJ, Hall WJ, Cannom DS, Daubert JP, Higgins SL, Klein H, Levine JH,
the baseline mortality risk between subgroups. EF is a Saksena S, Waldo AL, Wilber D, Brown MW, Heo M. Improved survival with animplanted defibrillator in patients with coronary disease at high risk for ventric- confirmed predictor of total and sudden death, but it ular arrhythmia. Multicenter Automatic Defibrillator Implantation Trial Investi- was of lesser prognostic value among patients in the gators. N Engl J Med 1996;335:1933–1940.
MADIT II. The MADIT II entry criterion of EF 5. Moss AJ, Bigger JT Jr, Odoroff CL. Postinfarction risk stratification. Prog
Cardiovasc Dis 1987;29:389 – 412.
30% (mean 23%) limited the spectrum of patients 6. Greenberg H, Case RB, Moss AJ, Brown MW, Carroll ER, Andrews ML.
with left ventricular dysfunction. The limited accuracy of Analysis of mortality events in the Multicenter Automatic Defibrillator Implan- EF calculation and inadequacy of EF to reflect regional tation Trial (MADIT-II). J Am Coll Cardiol 2004;43:1459 –1465.
7. Aronson D, Mittleman MA, Burger AJ. Elevated blood urea nitrogen level as
wall motion abnormalities may be additional factors con- a predictor of mortality in patients admitted for decompensated heart failure. Am J tributing to a decreased predictive value of EF within the Med 2004;116:466 – 473.
MADIT II population.
8. Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation (DEF-
INITE) Investigators. Prophylactic defibrillator implantation in patients with
The SCD-HeFT enrolled patients with ischemic and nonischemic dilated cardiomyopathy. N Engl J Med 2004;350:2151–2158.
nonischemic cardiomyopathy, but reported findings re- 9. Moss AJ, Greenberg H, Case RB, Zareba W, Hall WJ, Brown MW, Daubert
garding lack of benefit from ICDs are relevant for both JP, McNitt S, Andrews ML, Elkin AD. Long-term clinical course of patients aftertermination of ventricular tachyarrhythmia by an implanted defibrillator. Circu- subsets. Defibrillators in DEFibrillators In Non-Ischemic lation 2004;110:3760 –3765.
cardiomyopathy Treatment Evaluation showed that 10. Klein RC, Raitt MH, Wilkoff BL, Beckman KJ, Coromilas J, Wyse DG,
in nonischemic dilated cardiomyopathy, benefit from Friedman PL, Martins JB, Epstein AE, Hallstrom AP, et al. Analysis of implant-able cardioverter defibrillator therapy in the Antiarrhythmics Versus Implantable ICD was greater in patients with NYHA class III heart Defibrillators (AVID) Trial. J Cardiovasc Electrophysiol 2003;14:940 –948.
Diario Oficial de las Comunidades Europeas (Actos cuya publicación es una condición para su aplicabilidad) REGLAMENTO (CE) No 993/2001 DE LA COMISIÓN de 4 de mayo de 2001 que modifica el Reglamento (CEE) no 2454/93 por el que se fijan determinadas disposiciones de aplicación del Reglamento (CEE) no 2913/92 del Consejo por el que se aprueba el Código aduanero (Texto pertinente a efectos del EEE)