A 33-year-old man is referred to you for implantable cardioverter-defibrillator (ICD) therapy for primary prevention of sudden death. He was healthy until one year ago, when he was diagnosed with a nonischemic dilated cardiomyopathy (NIDCM), and has been on guideline-directed medical therapy since then, using a maximally-tolerated beta-blocker and angiotensin-converting enzyme (ACE) inhibitor. His left ventricular ejection fraction (EF) is 29% with moderate left ventricular (LV) dilation, but he has not shown much improvement over the past year. He has no coronary artery disease. His EKG shows sinus rhythm with left ventricular hypertrophy and a QRS duration of 110 msec. He tries to be active, but has class II heart failure symptoms and gets short of breath after walking one block. He has no indication for bradycardia pacing or cardiac resynchronization pacing therapy (CRT). What would you recommend?
If this patient lived in the United States, he would almost certainly be offered an ICD. There are at least four reasons for this recommendation. Patients with a DCM have a greater risk of dying from a cardiac arrest compared to the normal population, implantable defibrillators are the most effective therapy in preventing someone from dying from a sustained ventricular arrhythmia, there are data from large clinical trials that support their use, and ICD therapy is a class I recommendation in the most recent ACC/AHA/HRS device guidelines. But what are the data? The largest randomized trial that relates to this patient is the SCD-HeFT trial, which showed that ICD therapy was associated with a 33% reduction in mortality over five years in over 2,500 patients with NYHA class II or III systolic heart failure.1 However, it is true that the reduction in mortality was not statistically significant in the subgroup of patients with a NIDCM (P=0.06). It is also true that unlike patients with coronary disease, the death rate in patients with a NIDCM seems to be lower, and it has been harder to demonstrate a mortality benefit in patients with a NIDCM using ICD therapy. Nonetheless, a meta-analysis by Desai et al in 20042 came to the conclusion that ICD therapy does reduce mortality in patients with a NIDCM, and a large CRT trial showed benefit of CRT defibrillators in patients with advanced heart failure; thus, ICD therapy has become the standard of care in the United States for patients with a NIDCM and heart failure.
A major randomized controlled trial of ICD use in patients with a NIDCM was just published in the New England Journal of Medicine, coinciding with its presentation by the DANISH study group at the 2016 European Society of Cardiology meeting.3 The results were strikingly negative, with no reduction in overall mortality with ICD therapy. It is not clear why the results were negative, but a careful review of the study raises more questions than answers. One potential reason for failure to demonstrate a mortality benefit is that the overall mortality rate in all patients was low because medical therapy for systolic heart failure in general, and specifically in the patients enrolled in this trial, has greatly improved since earlier ICD studies were done. So it may have been underpowered in the modern era to show a mortality benefit. It is also important to note that ICD therapy resulted in a 50% reduction in sudden cardiac death. In addition, it is possible that the low event rates resulted in longer than initially planned follow-up time, which erased the initial benefit of ICD therapy. For example, the survival curves appear to be maximally separated at five years with a possible benefit from ICD therapy, but at eight years, that benefit disappears.
Perhaps the most critical data from the DANISH study are not shown in the primary publication, presumably because they pertain to a subset analysis. Nevertheless, in a negative study such as this one, which goes against the grain in many ways, it is critical to look carefully for subgroups who might benefit from the therapy rather than entirely pan the therapy. There is a figure in the supplemental file (Figure S1A) that tells an important story related to age. It shows a 36% reduction in overall mortality in the group in the lowest two age tertiles (age <68 years) with a P-value of 0.011. The curves separate at five years in a very similar fashion as the five-year survival curves in the nonischemic patients in SCD-HeFT. This is critical information, and it is logical. Defibrillator therapy is up against competing causes of death, so one might not expect the same benefit of ICD therapy in elderly patients, who are much more likely to develop other life-threatening diseases. The rate of non-cardiovascular death was high at 31%. The age of the patients in the DANISH trial was three to four years older compared to SCD-HeFT.
The 33-year-old patient with a NIDCM presented at the beginning of this editorial was one of the first recipients of a totally subcutaneous defibrillator as part of the U.S. IDE trial over five years ago. He recently underwent generator change for battery depletion and is doing well. The DANISH study should not be interpreted as a negative study — it should also not be used by physicians (or by administrators in a universal health care system) to withhold potentially lifesaving ICD therapy for all patients with a NIDCM. Instead, the results support the use of ICDs in our younger patients who have a NIDCM, and allow us to have a more informed conversation with our elderly patients who have a NIDCM and are considering ICD therapy for primary prevention.
- Bardy GH, Lee KL, Mark DB, et al. Amiodarone or an implantable cardioverter-defibrillator for congestive heart failure. N Engl J Med. 2005;352:225-237.
- Desai AS, Fang JC, Maisel WH, Baughman KL. Implantable defibrillators for the prevention of mortality in patients with nonischemic cardiomyopathy a meta-analysis of randomized controlled trials. JAMA. 2004;292:2874-2879.
- Kober L, Thune JJ, Nielsen JC, et al. Defibrillator implantation in patients with nonischemic systolic heart failure. N Engl J Med. 2016 Aug 27. [Epub ahead of print] DOI:10.1056/NEJMoa1608029.