cardiomyopathy : a multicentre experience implantation for non-ischaemic dilated improvement after primary prevention ICD Incidence of left ventricular function

Background The SCD-HeFT study demonstrated a benefit of primary prevention implantable cardioverter-defibrillator (ICD) implantation in patients with non-ischaemic dilated cardiomyopathy (NIDCM). However, NIDCM may improve spontaneously, even after waiting 6e9 months on optimal medical treatment. Objective To assess the incidence of left ventricular (LV) function improvement in patients receiving primary prevention ICDs for NIDCM. Methods All patients with NIDCM receiving primary prevention ICDs (non-cardiac resynchronisation therapy) from 2005 to the present at our institutions were retrospectively studied. All patients had NIDCM confirmed by a lack of significant stenoses on coronary angiography, a lack of significant valvular abnormalities on echo, and LV dysfunction with ejection fraction (EF) <35%. All patients had to have had a diagnosis of NIDCM for at least 9 months and be receiving optimal medical treatment for at least 3 months before implant according to the guidelines. All patients had at least New York Heart Association (NYHA) II symptoms. Baseline and follow-up EF was documented by quantitative echo and/or multi-gated acquisition scan. Results 332 patients were identified by a database search. Patients were aged 67611 years, 75% of them were male, NYHA 2.360.7, with EF 25613%, and LV diastolic diameter 61610 mm. Time from initial NIDCM diagnosis to implant was 1166 months and duration of medical treatment before implant was 865 months. Treatment at the time of implant included ACE inhibitors or ARBs (85%), b blockers (77%), spironolactone (53%), loop diuretic (63%) and digoxin (50%). Repeat EF assessment was available in 309/332 (93%) 866 months after implant. EF improved to >35% in 37/309 (12%) patients. Patients who improved had a shorter time from diagnosis to implant (963 vs 1365 months respectively, p1⁄40.03). No other significant predictors were identified for patients with improved EF. Conclusions In spite of following guidelines for implantation of primary prevention ICDs in patients with NIDCM, a substantial number of patients (12%) experience improvement in LV function to levels above those recommended for ICD implant. A shorter time from diagnosis to implant may predict post-implant improvement. INTRODUCTION The benefit of implantable cardioverter-defibrillator (ICD) treatment for primary prevention of sudden cardiac death has been long established in patients with ischaemic cardiomyopathy. Data demonstrating a similar benefit in patients with nonischaemic dilated cardiomyopathy (NIDCM) has been more recent. 3 Based on studies such as SCDHeFT and DEFINITE, the most recent joint AHA/ ACC/HRS guidelines for device-based treatment now categorises as a class I indication the implantation of an ICD in patients with NIDCM with ejection fraction (EF) #35% who are New York Heart Association (NYHA) functional class II or III. Although the guidelines suggest that it would be best to offer ICDs to patients who have “clinical profiles as similar to those included in the trials as possible,” they do not specify a minimum time between diagnosis of NIDCM and device implantation, nor is there a recommendation on the length of optimal medical treatment before implant. Yet, it is well known that some patients with NIDCM may experience spontaneous improvement in their cardiac function over weeks, but sometimes over months. Furthermore, improved medical treatment may also reverse the cardiomyopathy over time. Even patients with severe, end-stage cardiomyopathy on mechanical ventricular support may resolve to the point of not requiring such support with medical treatment alone. This is different from the ischaemic cardiomyopathy population where previous scar due to infarction is unlikely to regain contractile function. Unfortunately, data on the incidence and predictors of such left ventricular (LV) improvement in NIDCM are limited. The incidence of improvement has important implications for ICD treatment in this population since patients who experience substantial improvements in EF may no longer be at sufficient risk of sudden cardiac death towarrant an ICD. Thus, the purpose of this study was to assess in multiple centres the incidence of LV function improvement in patients receiving primary prevention ICDs for NIDCM and to determine if there are any predictors of such improvement in this population. PATIENTS AND METHODS Patient population Patient data were obtained from four different participating institutions. Consecutive patients Heart Rhythm Program, Southlake Regional Health Centre, Ontario, Canada; Department of Cardiology, University of Kansas Medical Center, Kansas City, Missouri, USA; Department of Cardiology, MetroHealth Campus, Case Western Reserve University, Cleveland, Ohio, USA; Texas Cardiac Arrhythmia Institute, St David’s Medical Center, Austin, Texas, USA Correspondence to Atul Verma, Southlake Heart Rhythm Program, 105e712 Davis Drive, Newmarket, Ontario, Canada L3Y 8C3; atul.verma@utoronto.ca Accepted 20 October 2009 Published Online First 1 November 2009 510 Heart 2010;96:510e515. doi:10.1136/hrt.2009.178061 Heart failure and cardiomyopathy group.bmj.com on April 11, 2010 Published by heart.bmj.com Downloaded from