T-wave alternans, heart rate turbulence and ventricular extrasystoles for the prediction of appropriate implantable cardioverter-defibrillator discharges

Abstract

Purpose: Predictors for survival or Implantable Cardioverter-Defibrillators (ICD) shocks are necessary for improved patient selection and optimized therapeutic strategies. We prospectively studied the prognostic value of Premature Ventricular Contractions (PVC, normalized per 24 hours), non-sustained VT (nsVT), Heart Rate Variability (HRV), Heart Rate Turbulence (HRT) and deceleration capacity from Holter monitoring and exercise T-Wave Alternans (TWA, Cambridge heart method). Methods: Consecutive patients (pts) with sinus rhythm undergoing ICD implantation between 2008 and 2012 at our institution for approved indications were prospectively enrolled, tests were performed within four weeks after ICD implantation. Follow-Up (FU) was scheduled every 6 months (m). Deaths of any cause or appropriate (ap) ICD shocks were defined as endpoints. Data are reported as mean±SD. Risk markers for ap shocks and death were identified using Cox regression analysis. Results: A total of 282 pts [age 65±13 yrs, 75% male, NYHA 2.5±0.7, QRS 123±31 ms, Left Ventricular Ejection Fraction (LVEF) 33±13%, 36% secondary prophylaxis, 24% history of myocardial infarction, 36% dual chamber and 24% biventricular devices] were enrolled. During 24±13 m FU, 34 (12%) pts received ap shocks and 38 (14%) pts died [5 (13%) of the latter received ap shock]. When univariate predictors of [A] ap shock, [B] death and [C] combined endpoint of both were entered in a multivariate cox regression model (using dichotomization by the median for continuous variables), the following variables remained after backward elimination: [A] PVC HR 3.2 (1.23-7.99, p=0.017), HRT category HR 2.32 (0.78-6.84, p=0.129), [B] estimated glomerular filtration rate HR 2.29 (1.09-4.81, p=0.029), Low Frequency/High Frequency quotient (LF/HF) from HRV HR 2.69 (1.25-5.78, p=0.011), LVEF HR 2.90 (1.20-7.04, p=0.019) and [C] PVC HR 2.02 (1.10-3.69, p=0.023), LF/HF from HRV HR 2.01 (1.11-3.65, p=0.022) and HRT category HR 2.24 (1.04-4.84, p=0.039) Specifically, exercise TWA or nsVT did not predict all-cause mortality or ap shocks. Conclusions: The number of PVC independently predicts ap shocks in this ICD population, additionally, HRV and HRT are predictive for a combined endpoint, whereas TWA is not predictive.