Atrial Fibrillation in CRT With a Defibrillator
The population of this retrospective, multicenter, observational, follow-up study consisted of 543 patients that were consecutively assigned to CRT-D in Erasmus Medical Centre, Rotterdam, the Netherlands and Medical Centre Alkmaar, Alkmaar, the Netherlands, between February 2005 and March 2011. Therapy assignment was based on the ESC guidelines for cardiac pacing and cardiac resynchronization therapy. All patients suffered from chronic heart failure for at least 3 months and were treated with optimal medication. Follow-up on mortality lasted until December 2011.
Baseline characteristics of all patients were collected by reviewing hospital records and included demographics, medical history, medication use, cardiovascular risk factors, and pacing and electrocardiographic characteristics.
Intraventricular conduction disorders (IVCDs) were defined according to the AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram. Left bundle branch block (LBBB) criteria are QRS duration ≥120 milliseconds; QS or rS in lead V1; wide R (frequently notched) waves in leads I, aVL, V5, or V6; and absent q waves in leads V5 and V6. Left anterior fascicular blocks (frontal plane axis deviation between −45° and −90°; qR pattern in aVL; and R-peak time in lead aVL of 45 milliseconds or more) with QRS duration ≥120 milliseconds were also counted as LBBB. Right bundle branch block (RBBB) required: QRS duration ≥120 milliseconds; rsr', rsR', rSR', or qR in leads V1 or V2; and occasionally, a wide and notched R wave and wide S waves in leads I, V5, and V6.
The placement procedure was similar in the 2 participating medical centers.
First, the right ventricular ICD-lead was positioned, mostly in the right ventricular apex using a cephalic cut down technique. Next, the LV pacing lead was inserted through the coronary sinus and positioned in the venous system, preferably in the (postero-)lateral vein. Finally, the atrial lead was positioned in the right atrial appendage. If transvenous LV pacing lead placement failed, an epicardial procedure was scheduled and performed.
During in-hospital and remote ICD follow-up, as part of usual care, ICD printouts were obtained every 3 months to determine the number and type of arrhythmias and the number of appropriate and inappropriate shocks. In Alkmaar, the majority of devices had a 3-zone configuration. The first zone was a monitor zone only, which was set to 160 ±10 bpm, the VT zone was set to 190 ±12 bpm and the VF zone was set to >209 ±15 bpm. In Rotterdam, a 2-zone configuration was programmed in 83% of the patients, and 16% had a 3-zone configuration. A monitoring zone was activated in the majority of patients with a 3-zone configuration. The mean VT detection rate was 171 ±11 bpm; the mean VF detection rate was set to 216 ±15 bpm. In the Medtronic Inc. (Minneapolis, MN, USA) devices, the number of intervals to detect was set to 18/24 episodes in all zones. In the Boston Scientific Inc. (Natick, MA, USA) devices, the number of intervals to detect was set to 8/10 intervals, with a duration of 8 seconds in the VT zone and 5 seconds in the VF zone. For all patients, ICD programming was intended to avoid inappropriate therapy by activating the available discriminators, e.g., dual-chamber algorithms, onset, stability, and morphology. For each patient, programming was tailored according to the clinical presentation.
The aim of this study was to assess risk factors for all-cause mortality, appropriate and inappropriate shocks. A shock was considered appropriate if it was delivered on ventricular fibrillation or ventricular tachycardia. ICD interrogations were performed 10 days, 3, 6, 9, and 12 months after implantation and then also every 6 months. Every arrhythmia in the ICD memory was interpreted by 2 experienced physicians per center and declared as either appropriate or inappropriate.
Normality of distribution was determined by Kolmogorov–Smirnov's test. Continuous data were expressed as mean ± SD, if normally distributed, otherwise by median and interquartile range (IQR). Continuous data were analyzed with Student's t-test or Mann–Whitney U-test when appropriate. Categorical data were compared by using the χ test or Fisher's exact test when appropriate. Multivariable Cox proportional hazards regression was used to analyze the effect of patient baseline characteristics on the outcome measures, e.g., mortality, appropriate, and inappropriate shocks. A backward elimination procedure was used and variables were withdrawn from the model if the corresponding P value was ≥0.1 until all factors had a P value <0.1. All available baseline characteristics were used for the backward elimination procedure, except for the variable "CRT-P upgraded to CRT-D," due to the small number of patients in whom this occurred.
The preventive cardioverter/defibrillator function of the device is always indicated in patients who have CRT-D as secondary prevention for life-threatening arrhythmias. For this reason, we repeated our analysis on appropriate and inappropriate shocks solely on the group of patients who had received CRT-D as primary prevention.
The influence of appropriate and inappropriate shocks on mortality was investigated by Cox regression analysis. For this purpose, appropriate and inappropriate shocks were treated as 2 separate time-dependent covariates, which allowed us to take into account the occurrence of multiple shocks that occurred consecutively in time.
We investigated the performance of the combination of independent risk factors, resulting from the multivariable analyses described above. For this purpose a C-index was calculated using the area under the receiver operating characteristic (ROC) curve.
SPSS 14.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis.
Methods
Patients and Study Design
The population of this retrospective, multicenter, observational, follow-up study consisted of 543 patients that were consecutively assigned to CRT-D in Erasmus Medical Centre, Rotterdam, the Netherlands and Medical Centre Alkmaar, Alkmaar, the Netherlands, between February 2005 and March 2011. Therapy assignment was based on the ESC guidelines for cardiac pacing and cardiac resynchronization therapy. All patients suffered from chronic heart failure for at least 3 months and were treated with optimal medication. Follow-up on mortality lasted until December 2011.
Baseline Data Collection
Baseline characteristics of all patients were collected by reviewing hospital records and included demographics, medical history, medication use, cardiovascular risk factors, and pacing and electrocardiographic characteristics.
Intraventricular conduction disorders (IVCDs) were defined according to the AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram. Left bundle branch block (LBBB) criteria are QRS duration ≥120 milliseconds; QS or rS in lead V1; wide R (frequently notched) waves in leads I, aVL, V5, or V6; and absent q waves in leads V5 and V6. Left anterior fascicular blocks (frontal plane axis deviation between −45° and −90°; qR pattern in aVL; and R-peak time in lead aVL of 45 milliseconds or more) with QRS duration ≥120 milliseconds were also counted as LBBB. Right bundle branch block (RBBB) required: QRS duration ≥120 milliseconds; rsr', rsR', rSR', or qR in leads V1 or V2; and occasionally, a wide and notched R wave and wide S waves in leads I, V5, and V6.
Placement Procedure and ICD Follow-Up
The placement procedure was similar in the 2 participating medical centers.
First, the right ventricular ICD-lead was positioned, mostly in the right ventricular apex using a cephalic cut down technique. Next, the LV pacing lead was inserted through the coronary sinus and positioned in the venous system, preferably in the (postero-)lateral vein. Finally, the atrial lead was positioned in the right atrial appendage. If transvenous LV pacing lead placement failed, an epicardial procedure was scheduled and performed.
During in-hospital and remote ICD follow-up, as part of usual care, ICD printouts were obtained every 3 months to determine the number and type of arrhythmias and the number of appropriate and inappropriate shocks. In Alkmaar, the majority of devices had a 3-zone configuration. The first zone was a monitor zone only, which was set to 160 ±10 bpm, the VT zone was set to 190 ±12 bpm and the VF zone was set to >209 ±15 bpm. In Rotterdam, a 2-zone configuration was programmed in 83% of the patients, and 16% had a 3-zone configuration. A monitoring zone was activated in the majority of patients with a 3-zone configuration. The mean VT detection rate was 171 ±11 bpm; the mean VF detection rate was set to 216 ±15 bpm. In the Medtronic Inc. (Minneapolis, MN, USA) devices, the number of intervals to detect was set to 18/24 episodes in all zones. In the Boston Scientific Inc. (Natick, MA, USA) devices, the number of intervals to detect was set to 8/10 intervals, with a duration of 8 seconds in the VT zone and 5 seconds in the VF zone. For all patients, ICD programming was intended to avoid inappropriate therapy by activating the available discriminators, e.g., dual-chamber algorithms, onset, stability, and morphology. For each patient, programming was tailored according to the clinical presentation.
Definitions
The aim of this study was to assess risk factors for all-cause mortality, appropriate and inappropriate shocks. A shock was considered appropriate if it was delivered on ventricular fibrillation or ventricular tachycardia. ICD interrogations were performed 10 days, 3, 6, 9, and 12 months after implantation and then also every 6 months. Every arrhythmia in the ICD memory was interpreted by 2 experienced physicians per center and declared as either appropriate or inappropriate.
Statistical Analysis
Normality of distribution was determined by Kolmogorov–Smirnov's test. Continuous data were expressed as mean ± SD, if normally distributed, otherwise by median and interquartile range (IQR). Continuous data were analyzed with Student's t-test or Mann–Whitney U-test when appropriate. Categorical data were compared by using the χ test or Fisher's exact test when appropriate. Multivariable Cox proportional hazards regression was used to analyze the effect of patient baseline characteristics on the outcome measures, e.g., mortality, appropriate, and inappropriate shocks. A backward elimination procedure was used and variables were withdrawn from the model if the corresponding P value was ≥0.1 until all factors had a P value <0.1. All available baseline characteristics were used for the backward elimination procedure, except for the variable "CRT-P upgraded to CRT-D," due to the small number of patients in whom this occurred.
The preventive cardioverter/defibrillator function of the device is always indicated in patients who have CRT-D as secondary prevention for life-threatening arrhythmias. For this reason, we repeated our analysis on appropriate and inappropriate shocks solely on the group of patients who had received CRT-D as primary prevention.
The influence of appropriate and inappropriate shocks on mortality was investigated by Cox regression analysis. For this purpose, appropriate and inappropriate shocks were treated as 2 separate time-dependent covariates, which allowed us to take into account the occurrence of multiple shocks that occurred consecutively in time.
We investigated the performance of the combination of independent risk factors, resulting from the multivariable analyses described above. For this purpose a C-index was calculated using the area under the receiver operating characteristic (ROC) curve.
SPSS 14.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis.
SHARE
