There Should Never Be Another Antiarrhythmics Versus Implantable Defibrillator (AVID) Trial

Abstract

T Antiarrhythmics versus Implantable Defibrillator (AVID) trial was halted on April 7, 1997 following demonstration of 38% improved survival at 1 year in patients randomized to defibrillators.1 Some have already hailed this as a landmark study. This is entirely understandable, because those who have supported this study need to justify its existence and execution. A rationale can always be constructed to justify a scientific endeavor, even if its logic or ethics are questionable. We also agree that AVID will be known as a landmark study, but for reasons quite different than those proffered by its proponents. We— and others—have been unequivocal in our criticisms of this study for many years.2–9 Our concerns, detailed previously, center not only on the flawed scientific method, but most importantly, the ethics of conducting such a study in the light of monumental evidence (supporting implantable cardioverter defibrillator [ICD] benefit compared with the best available antiarrhythmic drugs) that preceded it. The basic question asked in AVID was whether, in patients with a history of sustained, life-threatening ventricular tachyarrhythmias, the ICD prolongs life and whether the undisputed efficacy of ICDs in terminating ventricular tachycardia and ventricular fibrillation translates into an overall survival benefit. Further, it pits the ICD against 2 antiarrhythmic drugs shown to be most effective for therapy of lethal ventricular arrhythmias, amiodarone and sotalol. The underlying question of the study was whether the acknowledged superiority of ICD therapy in protecting patients at risk of sudden arrhythmic death would extend to overall survival, from all causes of death. Our objections arose from the fact that—before and without the study—there existed more than ample evidence to answer this question. Multiple series have demonstrated far better sudden death and overall survival for patients treated with ICDs10–17 compared with ostensibly similar patients receiving antiarrhythmic drugs.18–23 In addition, there were 2 carefully performed matched case control studies, in which ICD patients were closely matched on all critical baseline variables (age, sex, presenting arrhythmia, left ventricular ejection fraction, etc.) to patients receiving amiodarone in the first study and d,l sotalol in the second.24,25 The results in terms of survival and recurrences of ventricular tachycardia were unanimously and significantly in favor of ICD therapy in both studies. Probably the most persuasive evidence that drugs were not going to provide the same protection as the ICD comes from the large prospective randomized studies on amiodarone and sotalol,26–28 and concurrently, 2 prospective, randomized trials of ICDs versus conventional medical therapy.29,30 Using the implantable Holter function of modern ICDs, there are now thousands of examples of sustained, dangerously high-rate ventricular tachycardias terminated by the device. If you are the physician whose patient is still with you many years after such episodes, the question of whether that patient’s life has been prolonged becomes nonsensical! The amassed evidence from so many sources led 2 of the highly respected arrhythmia research groups in the world, both fully cognizant of—but manifestly unwilling to wait for proof coming from—AVID to issue editorials explicitly stating, ‘‘defibrillators are superior to antiarrhythmic drugs . . . .’’31,32 The persistent response of the AVID proponents has been that there was not as yet any evidence from prospective, randomized trials (even after 2 such trials had been completed,29,30) showing that the ICDs protection against sudden death translated to meaningful extension of life.33–37 We and others have long argued3,5–8 that this question has no real meaning; in the right population, the ICD is unequaled in doing what it was meant to do—convert potentially life-ending ventricular tachycardia or ventricular fibrillation—allowing a patient to live life until the nonarrhythmic cardiac or other illness results in the patient’s death. We have also argued that the right question of such trials involving ICDs is not extension of life—for the moment a patient’s sudden death is terminated—the patient has already had an extension of life. So, why should there never be another AVID? The main lesson that needs to be learned and adhered to is that substantial, unidirectional, and conclusive evidence should not be ignored simply due to lack of data from prospective randomized trials. Penicillin, pacemakers, and multiple other obviously beneficial therapies have existed for years, unquestioned in their efficacy, without ever having been subjected to prospective randomized trial evidence. We are not a priori objecting to randomized clinical trials. Such trials are desirable and mandatory when the question is truly open and no strong evidence exists to resolve an important clinical or scientific dilemma. This is not the case in the ICD effectiveness issue. The Cardiac From the Division of Electrophysiology and Pacing, Department of Medicine, University of Louisville, Louisville, Kentucky, and Guidant Europe, Zaventem, Belgium. Manuscript received and accepted May 7, 1997. Address for reprints: Igor Singer, MBBS, Electrophysiology and Pacing, University of Louisville, 530 S. Jackson Street, Louisville, Kentucky 40292.