Prognostic yield of echocardiographic measures of cardiac structure and function in HFrEF patients with vs without an ICD or CRT-D

Abstract Background Selected echocardiographic measures of cardiac structure and function are known to be associated with mortality in patients with Heart Failure with reduced Left Ventricular (LV) Ejection Fraction (HFrEF). Whether the predictive ability of measures of cardiac structure and function remain after implantation of an Implantable Cardioverter Defibrillator (ICD) or Cardiac Resynchronization Therapy-Defibrillator (CRT-D) is unknown among patients with HFrEF. Purpose To compare the predictive ability of echocardiographic measures of cardiac structure and function in HFrEF patients with or without an ICD/CRT-D. Methods Echocardiographic examinations were performed between 2003 and 2013 and retrieved from 2168 patients with HFrEF. Patients that had an ICD or CRT-D implanted during follow-up were identified through registries. The endpoint was all-cause mortality and follow-up was 100% to date. Images were analyzed offline in accordance with current guidelines. Associations were assessed by Cox multivariable regression and Cox univariable interaction models. Results After a median of 169 days, 677 patients had an ICD or CRT-D implanted. In total, 68% vs 80% of patients with and without an ICD, respectively, were men (p<0.001) and the median age were 70 vs 71 years (p=0.192). There was a significant difference in resting heart rate (median 74 vs 69, p<0.001) and frequency of ischemic heart disease (64% vs 81%, p<0.001) between the groups. Patients with an ICD had a higher mortality rate (69 per 1000 people/year vs 92 per 1000 people/year, p<0.001) and died significantly faster (median 9.3 vs 8.2 years, p<0.001). Patients with an ICD had higher Left Atrial (LA) End Systolic Volume (59.2 vs 67.3, p<0.001), LA Volume Index (28.5 vs 33.9, p<0.001), LV End Diastolic Volume (137.7 vs 153.7, p<0.001), and LV Mass Index (114.5 vs 124.5, p<0.001), as well as lower LA strain (9.8 vs 6.3, p<0.001) and TAPSE (2.0 vs 1.7, p<0.001). In patients without an ICD, Right Ventricular strain was associated with mortality (HR = 1.05 per 1% decrease mm/m2 [1.005, 1.11] when adjusting for covariates whereas in patients with an ICD, LA Volume Index (HR 1.14 per 1 increase [1.03, 1.25], and LV Ejection Fraction (HR 0.96 per 1% increase [0.92, 0.99]) were associated with mortality when adjusting for covariates. ICD/CRT-D implantation significantly modified the prognostic yield of LV End Diastolic Volume (HR per 1 ml increase: 0.99 without vs 1.01 with an ICD/CRT-D, p=0.005) and LV Mass Index (HR per 1 g/m2 increase: 1.00 without and 1.01 with an ICD/CRT-D, p=0.031). Conclusion Different echocardiographic measures were stronger prognosticators in HFrEF patients with and without an ICD or CRT-D, respectively, and ICD implantation modified the prognostic yield of LV End Diastolic Volume and LV Mass Index in HFrEF patients.Table 1Forest plot