09 Dec Who Benefits from an Implantable Defibrillator?

MedicalResearch.com Interview with:

Dr. Wayne Levy

Wayne C. Levy, MD
Division of Cardiology
University of Washington
Seattle, Washington

MedicalResearch.com: What is the background for this study?

Response: There is uncertainty how effective ICDs are outside of clinical trials in real world patients who are often older with more comorbidities. The recent DANISH ICD only reiterates provider and patient concerns regarding the effectiveness of an ICD, that may be life saving, but does not improve heart failure symptoms or reduce hospitalizations. Many patients die without ever having an ICD shock, so the ICD was not necessary for the patient.

We postulated that the effectiveness of an ICD is not driven by the absolute risk of sudden death (event rate/year) but rather the proportion of all deaths that are due to sudden death vs non sudden death. If a patient has a 3% annual risk of sudden death and this is decreased to 1% with an ICD (a 67% reduction in sudden death), the patient benefit will be much greater if the non sudden death rate is 1%/year than if it is 12%/year. In the first scenario the absolute mortality is decreased from 4% to 2% (a 50% reduction) whereas in the second patient, the mortality would be decreased from 15% to 13% (a 13% benefit).

We developed the Seattle Proportional Risk Model (SPRM) using 10 clinical variables that had a differential impact on the mode of death, sudden vs. non sudden, in ~10,000 patients with ~2,500 deaths. Sudden death was more common in younger patients, male, without diabetes mellitus, NYHA 1 or 2 vs. 3 or 4, lower EF, SBP closer to 140, normal sodium and creatinine, higher BMI, and digoxin use.

MedicalResearch.com: What are the main findings?

Response: We prospectively applied the Seattle Proportional Risk Model to the HF ACTION randomized trial of exercise training where all patients had an EF ≤35% and 99% were NYHA 2 or 3. ICDs were present prior to death in only 62% of the patients. The predicted proportion of sudden death of all deaths was 56% and the observed proportion was 59%, supporting adequate calibration of the model. The SPRM was predictive of sudden vs. non sudden death (p=0.002) in patients without an ICD. Overall the ICD benefit on all cause mortality was 37% and varied with the SPRM (interaction p=0.001), with a greater benefit in patients with a higher predicted proportion of sudden death. There was no ICD benefit with the predicted proportion of sudden death was <32%.

MedicalResearch.com: What should readers take away from your report?

Response: Older, female, NYHA 3 patients with comorbidities like diabetes, CKD and likely other comorbidities, derive less benefit from a primary prevention ICD.

MedicalResearch.com: What recommendations do you have for future research as a result of this study?

Response: We have applied this model to ~88,000 primary prevention ICD patients and ~10,000 control patients without an ICD with similar results (AHA 2015 – http://circ.ahajournals.org/content/134/Suppl_1/A12090). The SPRM is a prospectively validated model that can be used to identify patients with greater, lesser, and potentially no benefit from a primary prevention ICD with observational ICD cohorts. Validation within randomized ICD trials like SCD-HeFT and MADIT 2 are ongoing.

MedicalResearch.com: Is there anything else you would like to add?

Response: This model may have the greatest utility at the time of ICD generator replacement when patients have not received ICD therapy. These patients would be anticipated to be older, with more comorbidities and likely derive less benefit than when a primary prevention ICD is implanted.

An interactive model like the SPRM can provide information to patients and providers, to make better informed decisions regarding the potential benefit of a primary prevention ICD at initial implantation and potentially at the time of generator replacements.

MedicalResearch.com: Thank you for your contribution to the MedicalResearch.com community.

Citation:

Does the ICD benefit vary with the estimated proportional risk of sudden death in heart failure patients?

Does the Implantable Cardioverter-Defibrillator Benefit Vary With the Estimated Proportional Risk of Sudden Death in Heart Failure Patients?
Wayne C. Levy, MD,  Yanhong Li, MD, MS,  Shelby D. Reed, PhD,  Michael R. Zile, MD, Ramin Shadman, MD, Todd Dardas, MD, David J. Whellan, MD, Kevin A. Schulman, MD,  Stephen J. Ellis, PhD,  Matthew Neilson, PhD, Christopher M. O’Connor, MD, for the HF-ACTION Investigators
JACC: Clinical Electrophysiology  Available online 23 November 2016

Note: Content is Not intended as medical advice. Please consult your health care provider regarding your specific medical condition and questions.

More Medical Research Interviews on MedicalResearch.com

Last Updated on December 9, 2016 by Marie Benz MD FAAD