757 Ambulatory blood pressure monitoring and 2D-echo in predicting non arrhythmic cardiac events after ICD implantation for MADIT II criteria. A new prognostic index

Abstract

Aim:We compared ambulatory blood pressure monitoring (ABPM) and 2D echo to predict cardiac death and heart failure (combined end-point) after implantation of internal cardiac defibrillator (ICD) in 135 consecutive patients, all fulfilling MADIT II criteria; 115 men and 20 females, mean age 66±7, 86 with dual chamber ICD, 20 with cardiac resinchronization therapy (CRT), 29 patients received single chamber ICD. Methods: ABPM was performed 2 weeks after ICD implantation on outpatients basis with Spacelabs 90207. Mostpatients were on medical therapy with betablockers (75%), ace-inhibitors (90%) and diuretics (90%) at the time of ABPM evaluation. ABPM variables considered were: mean 24hour Systolic Blood Pressure (m24hSBP), mean 24hour Diastolic Blood Pressure (m24hDBP), mean 24h Heart Rate (m24hHR), Pulse pressure (PP). 2D-echo was performed before ICD implantation. LVEF was calculated by Simpson rule. Age, Blood Creatinine (BC) and CRT were also considered in statistical analysis. Results: At 1 year follow-up there were 25 cardiac events (18%), 4 cardiac deaths (cardiogenic shock) and 21 hospitalizations for pulmonary oedema or congestive heart failure. Age (p= 0.007), m24hDBP (p=0.02) and BC (p=0.02) were significantly associated with events, but not LVEF (p = 0.6), m24hSBP (p=0.09), m24hHR (p=0.08), PP (p= 0.27) and CRT (p=0.96). Only age (Odds ratio 3.8, p=0.01) and m24hDBP (Odds Ratio 2.3 p=0.04) remained significantly associated with clinical events in multivariate analysis. A prognostic index (PI) was built from these variables (m24hDBP and age) according to the formula: 120− age + m24hDBP. PI was significantly lower in patients with clinical events (107± 9 vs 121± 14, p 108 remained event-free in contrast with 55% of patients with PI ≤108 (p <0.001 by Log-rank test). Multivariate Cox regression analysis (including as covariates PI, age, LVEF, m24hDBP) confirmed that only PI was significantly associated with events (Hazard Ratio 11.4 with 95% confidence interval 2.4-55 with p =0.003). Conclusion:The PI built bym24hDBP and age could be a simple method to stratify the hemodynamic risk in patients with uniformally depressed LVEF (≤ 30% by MADIT II criteria) and ICD. .