Background
Implantable cardioverter defibrillator (ICD) implantation is contraindicated in those with <1 year life expectancy.
Objective
To develop a risk prediction score for 1-year mortality in patients with primary prevention ICDs and to determine the incremental improvement in discrimination when incorporating serum-based biomarkers to traditional clinical variables.
Methods
We analyzed data from the PROSE-ICD study, a large prospective observational study of patients undergoing primary prevention ICD implantation who were extensively phenotyped for clinical and serum markers. We identified variables predicting 1-year mortality and synthesized them into a comprehensive risk scoring construct using backward selection.
Results
Among 1,189 patients deemed by their treating physicians as having reasonable 1 year life expectancy, 62 patients died within 1 year of ICD implantation. The risk score, comprised of 6 clinical factors (age ≥75 years, New York Heart Association class III/IV, atrial fibrillation, eGFR <30mL/min/1.73m2, diabetes, and use of diuretics), had good discrimination (AUC=0.77) for 1-year mortality. Addition of 3 biomarkers (TNF-αRII, pro-BNP, and cTnT) further improved model discrimination to 0.82. Patients with 0-1, 2-3, 4-6, or 7-9 risk factors had 1-year mortality rates of 0.8%, 2.7%, 16.1% and 46.2%, respectively.
Conclusions
Individuals with more co-morbidities and elevations of specific serum biomarkers were at increased risk for all-cause mortality despite being deemed as having reasonable 1 year life expectancy. A simple risk score comprised of readily available clinical data and serum biomarkers may better identify patients at high risk of early mortality and improve patient selection and counseling for primary prevention ICD therapy.