Impact of Implantable Cardioverter Defibrillators on Survival of Patients with Centrifugal Left Ventricular Assist Devices WILLIAM LEE, M.B.B.S.,* ANDRE TAY, R.N.,* RAJESH N. SUBBIAH, PH.D.,* BRUCE D. WALKER, PH.D.,* DENNIS L. KUCHAR, M.D.,* KAVITHA MUTHIAH, M.B.B.S.,* PETER S. MACDONALD, PH.D.,* ANNE M. KEOGH, M.D.,* EUGENE KOTLYAR, M.D.,* ANDREW JABBOUR, PH.D.,* PHILIP SPRATT, M.B.B.S.,† PAUL C. JANSZ, PH.D.,† EMILY GRANGER, M.B.B.S.,† KUMUD DHITAL, PH.D.,† and CHRISTOPHER S. HAYWARD, M.D.* *From the Cardiology Department, St. Vincent’s Hospital, Sydney, Australia; and †Cardiothoracic Surgery Department, St. Vincent’s Hospital, Sydney, Australia Background: Both implantable cardioverter defibrillators (ICDs) and left ventricular assist devices (LVADs) have a positive impact on survival in the heart failure population. We sought to determine whether these positive effects on survival are additive or whether LVAD therapy supersedes ICD therapy. Method: We analyzed survival data of patients implanted with nonpulsatile LVADs between October 2004 and March 2013. Survival in patients with ICDs (n = 64) was compared to those without ICDs (n = 36). Patients exited the study at the time of heart transplantation or death. Results: A total of 100 patients underwent LVAD implantation during this time. Patients had a mean follow-up time of 364 ± 295 days. Death occurred in 15 (38%) patients in the no ICD group versus 18 (30%) in the ICD group. Univariate analysis demonstrated a marginal early survival benefit at up to 1 year post-LVAD implant in the ICD cohort; however, at time points greater than 1 year there was no statistically significant benefit in ICD therapy in LVAD patients (P = 0.56). Multivariate analysis did not show any significant predictor of survival. There were no patients who died of sudden cardiac death. There was no significant difference in the time to heart transplantation (443 days ± 251 no ICD vs 372 days ± 277 ICD, P = 0.37). Conclusion: The benefit of ICD therapy in the setting of continuous flow LVAD therapy is uncertain. Although prolonged ventricular arrhythmias (VAs) may potentially impact on patient survival, LVAD therapy is beneficial in prevention of sudden cardiac death due to VAs. (PACE 2015; 38:925–933) defibrillation – ICD, VT, congestive heart failure, heart transplantation
Introduction At present, cardiac transplantation is considered the “gold standard” in refractory endstage heart failure management.1 Over the last decade cardiac transplant rates have remained stable at approximately 3,800 transplants per year worldwide2 despite the growing heart failure
Conflict of Interest: Christopher S. Hayward has received consulting fees and a research grant, unrelated to the current study, from HeartWare, Inc. There are no other conflicts to declare. Address for reprints: Christopher Hayward, M.D., Department of Cardiology, St. Vincent’s Hospital, 438 Victoria St., Darlinghurst, NSW 2010, Australia. Fax: 61-2-8382 6881; e-mail: [email protected] Received September 3, 2014; revised February 20, 2015; accepted April 21, 2015. doi: 10.1111/pace.12654
population. Therefore, cardiac transplantation does not provide a sustainable long-term option for the growing heart failure population. Since the approval of left ventricular assist device (LVAD) destination therapy in 2010, up to 40% of new implants are designated as destination therapy LVADs.3 LVAD therapy is increasingly utilized as a definitive therapy for advanced heart failure, and may represent a sustainable solution for the growing heart failure population. As a result, the number of community LVAD patients is growing and the optimal balance between traditional heart failure therapy and mechanical circulatory support therapy is yet to be determined. Numerous multicenter trials have shown the survival benefit of implantable cardioverter defibrillator (ICD) therapy in the setting of heart failure.4–6 Current Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) guidelines suggest ICD
©2015 Wiley Periodicals, Inc. PACE, Vol. 38
August 2015
925
LEE, ET AL.
reactivation post-LVAD implant and primary ICD implantation post-LVAD implant.7 Despite this, recent observational studies show there is conflicting evidence as to the survival benefit of defibrillator therapy in the LVAD setting.8–11,16 Furthermore, ICD therapy may represent a doubleedged sword in the LVAD setting as numerous case studies12–14 report no loss of consciousness during ventricular arrhythmia (VA) and therefore patients may be subjected to painful conscious shocks with a consequent decrease in quality of life. The aim of this study was to examine the survival benefit of ICD therapy in the setting of LVAD therapy and the clinical impact of VAs in LVAD patients. Methods This single-center retrospective population study was conducted with the approval of the Human Research Ethics Committee of St. Vincent’s and Mater Health Service. Medical records of all adult patients who were implanted with a Ventrassist (Ventracor Limited, Brisbane, Australia) or Heartware MVAD (Heartware, Framingham, MA, USA) continuous flow LVAD at St. Vincent’s Hospital, Sydney, between October 2004 and March 2013 were examined. The study included both destination and bridge-totransplant patients as well as combined left and right ventricular assist devices (BiVADs). Patient data were collected beginning 6 months prior to LVAD implantation until the study end point was reached. This included: death, cardiac transplantation, or end of study (March 1, 2013), to ensure capture of the patients’ disease course from preimplant to death/cardiac transplantation. All outpatient LVAD patients were reviewed on a monthly basis at St. Vincent’s Hospital, Sydney, Australia. Outcomes Patient outcomes were analyzed according to ICD status. The primary outcome was allcause mortality. Date and primary cause of death were recorded. Patients were censored at the time of their cardiac transplantation or at the end of the study on March 1, 2013. Tachycardia therapies were deactivated at the time of LVAD implantation and then reactivated within 24 hours of LVAD implant. There was no other standard ICD programming protocol. A total of 11 patients had their tachycardia therapies turned off during the course of their follow-up. Survival by ICD status was analyzed by intention-to-treat analysis; a crossover survival analysis was also performed to account for patients who had their ICD therapies turned off during the course of their follow-up. To further evaluate the significance of deactivated
926
ICD therapies, a subgroup analysis of survival was performed excluding those patients who had deactivated ICD therapies at any time during their treatment course. A multivariate analysis was conducted using pre-LVAD implant covariates. This included: ICD status (ICD vs no ICD), heart failure duration (greater than 3 months vs less than 3 months), heart failure etiology (ischemic vs nonischemic), pre-LVAD mechanical support, age, creatinine, pre-LVAD left ventricular end-diastolic dimension, and pre-LVAD left ventricular ejection fraction. Secondary outcomes included: incidence of both VA and defibrillator shocks. Although patient follow-up was regular and ongoing, the follow-up was divided into four distinct clinical time periods for clinical relevance. This included: preimplant (6 months prior to LVAD implant), perioperative (less than 1 month post-LVAD implant), early (1– 3 months postimplant), where most patients have been discharged and were being treated in a clinicbased setting, and late (greater than 3 months postimplant), where the majority of patients were listed and awaiting cardiac transplantation. Due to the high incidence of asymptomatic VA, secondary outcomes were included only in those patients who had implantable devices with the ability to log and record VA events. All VA events recorded by the device were included in the study including nonsustained, sustained, and treated episodes. VA burden was described as number of VA episodes per month in each clinical time period: preimplant, perioperative, early, and late. Patients were categorized according to their VA burden as: none (0 VA events per month), low (1–200 VA events per month), or high (greater than 200 VA events per month). For survival analysis, patients were dichotomized according to VA burden. Treatment of sustained and recurrent VA episodes was at the discretion of the treating cardiologist. Patients with ICDs were interrogated for number of shocks received. Shock events were classified as appropriate or not appropriate. Shock burden was the cumulative number of shocks delivered in each clinical time period. Patients were categorized according to their shock burden as: none (0 shocks), low (less than or equal to five shocks), or high (greater than five shocks). For survival analysis, patients were dichotomized based on their first shock event; patients who did not receive any shocks during their follow-up period remained in the “no-shock” group. Statistical Analysis Patients’ baseline characteristics were statistically analyzed using Excel 2013 software
August 2015
PACE, Vol. 38
ICD AND LVAD INTERACTION
Table I. Baseline Demographics
n= Age Ischemic etiology Nonischemic Male Female Heart failure
Introduction At present, cardiac transplantation is considered the “gold standard” in refractory endstage heart failure management.1 Over the last decade cardiac transplant rates have remained stable at approximately 3,800 transplants per year worldwide2 despite the growing heart failure
Conflict of Interest: Christopher S. Hayward has received consulting fees and a research grant, unrelated to the current study, from HeartWare, Inc. There are no other conflicts to declare. Address for reprints: Christopher Hayward, M.D., Department of Cardiology, St. Vincent’s Hospital, 438 Victoria St., Darlinghurst, NSW 2010, Australia. Fax: 61-2-8382 6881; e-mail: [email protected] Received September 3, 2014; revised February 20, 2015; accepted April 21, 2015. doi: 10.1111/pace.12654
population. Therefore, cardiac transplantation does not provide a sustainable long-term option for the growing heart failure population. Since the approval of left ventricular assist device (LVAD) destination therapy in 2010, up to 40% of new implants are designated as destination therapy LVADs.3 LVAD therapy is increasingly utilized as a definitive therapy for advanced heart failure, and may represent a sustainable solution for the growing heart failure population. As a result, the number of community LVAD patients is growing and the optimal balance between traditional heart failure therapy and mechanical circulatory support therapy is yet to be determined. Numerous multicenter trials have shown the survival benefit of implantable cardioverter defibrillator (ICD) therapy in the setting of heart failure.4–6 Current Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) guidelines suggest ICD
©2015 Wiley Periodicals, Inc. PACE, Vol. 38
August 2015
925
LEE, ET AL.
reactivation post-LVAD implant and primary ICD implantation post-LVAD implant.7 Despite this, recent observational studies show there is conflicting evidence as to the survival benefit of defibrillator therapy in the LVAD setting.8–11,16 Furthermore, ICD therapy may represent a doubleedged sword in the LVAD setting as numerous case studies12–14 report no loss of consciousness during ventricular arrhythmia (VA) and therefore patients may be subjected to painful conscious shocks with a consequent decrease in quality of life. The aim of this study was to examine the survival benefit of ICD therapy in the setting of LVAD therapy and the clinical impact of VAs in LVAD patients. Methods This single-center retrospective population study was conducted with the approval of the Human Research Ethics Committee of St. Vincent’s and Mater Health Service. Medical records of all adult patients who were implanted with a Ventrassist (Ventracor Limited, Brisbane, Australia) or Heartware MVAD (Heartware, Framingham, MA, USA) continuous flow LVAD at St. Vincent’s Hospital, Sydney, between October 2004 and March 2013 were examined. The study included both destination and bridge-totransplant patients as well as combined left and right ventricular assist devices (BiVADs). Patient data were collected beginning 6 months prior to LVAD implantation until the study end point was reached. This included: death, cardiac transplantation, or end of study (March 1, 2013), to ensure capture of the patients’ disease course from preimplant to death/cardiac transplantation. All outpatient LVAD patients were reviewed on a monthly basis at St. Vincent’s Hospital, Sydney, Australia. Outcomes Patient outcomes were analyzed according to ICD status. The primary outcome was allcause mortality. Date and primary cause of death were recorded. Patients were censored at the time of their cardiac transplantation or at the end of the study on March 1, 2013. Tachycardia therapies were deactivated at the time of LVAD implantation and then reactivated within 24 hours of LVAD implant. There was no other standard ICD programming protocol. A total of 11 patients had their tachycardia therapies turned off during the course of their follow-up. Survival by ICD status was analyzed by intention-to-treat analysis; a crossover survival analysis was also performed to account for patients who had their ICD therapies turned off during the course of their follow-up. To further evaluate the significance of deactivated
926
ICD therapies, a subgroup analysis of survival was performed excluding those patients who had deactivated ICD therapies at any time during their treatment course. A multivariate analysis was conducted using pre-LVAD implant covariates. This included: ICD status (ICD vs no ICD), heart failure duration (greater than 3 months vs less than 3 months), heart failure etiology (ischemic vs nonischemic), pre-LVAD mechanical support, age, creatinine, pre-LVAD left ventricular end-diastolic dimension, and pre-LVAD left ventricular ejection fraction. Secondary outcomes included: incidence of both VA and defibrillator shocks. Although patient follow-up was regular and ongoing, the follow-up was divided into four distinct clinical time periods for clinical relevance. This included: preimplant (6 months prior to LVAD implant), perioperative (less than 1 month post-LVAD implant), early (1– 3 months postimplant), where most patients have been discharged and were being treated in a clinicbased setting, and late (greater than 3 months postimplant), where the majority of patients were listed and awaiting cardiac transplantation. Due to the high incidence of asymptomatic VA, secondary outcomes were included only in those patients who had implantable devices with the ability to log and record VA events. All VA events recorded by the device were included in the study including nonsustained, sustained, and treated episodes. VA burden was described as number of VA episodes per month in each clinical time period: preimplant, perioperative, early, and late. Patients were categorized according to their VA burden as: none (0 VA events per month), low (1–200 VA events per month), or high (greater than 200 VA events per month). For survival analysis, patients were dichotomized according to VA burden. Treatment of sustained and recurrent VA episodes was at the discretion of the treating cardiologist. Patients with ICDs were interrogated for number of shocks received. Shock events were classified as appropriate or not appropriate. Shock burden was the cumulative number of shocks delivered in each clinical time period. Patients were categorized according to their shock burden as: none (0 shocks), low (less than or equal to five shocks), or high (greater than five shocks). For survival analysis, patients were dichotomized based on their first shock event; patients who did not receive any shocks during their follow-up period remained in the “no-shock” group. Statistical Analysis Patients’ baseline characteristics were statistically analyzed using Excel 2013 software
August 2015
PACE, Vol. 38
ICD AND LVAD INTERACTION
Table I. Baseline Demographics
n= Age Ischemic etiology Nonischemic Male Female Heart failure
