Author + information
- Received April 7, 2010
- Revision received June 28, 2010
- Accepted August 10, 2010
- Published online January 11, 2011.
- Massimo Santini, MD⁎,⁎ (, )
- Maurizio Gasparini, MD†,
- Maurizio Landolina, MD‡,
- Maurizio Lunati, MD§,
- Alessandro Proclemer, MD∥,
- Luigi Padeletti, MD¶,
- Domenico Catanzariti, MD#,
- Giulio Molon, MD⁎⁎,
- Giovanni Luca Botto, MD††,
- Laura La Rocca, MS‡‡,
- Andrea Grammatico, PhD‡‡,
- Giuseppe Boriani, MD§§,
- cardiological centers participating in the ClinicalService Project
- ↵⁎Reprint requests and correspondence:
Dr. Massimo Santini, Institute of Cardiology, San Filippo Neri Hospital, Via Giovanni Martinotti, 20, Rome 00135, Italy
Objectives The purpose of this analysis was to evaluate the correlation between atrial tachycardia (AT) or atrial fibrillation (AF) and clinical outcomes in heart failure (HF) patients implanted with a cardiac resynchronization therapy defibrillator (CRT-D).
Background In HF patients, AT and AF have high prevalence and are associated with compromised hemodynamic function.
Methods Forty-four Italian cardiological centers followed up 1,193 patients who received a CRT-D according to current guidelines for advanced HF, New York Heart Association functional class ≥II, left ventricular ejection fraction ≤35%, and QRS complex ≥120 ms. All patients were in sinus rhythm at implant.
Results During a median follow-up period of 13 months, AT/AF >10 min occurred in 361 of 1,193 (30%) patients. The composite end point (deaths or HF hospitalizations) occurred in 174 of 1,193 (14.6%). Multivariate time-dependent Cox regression analyses showed that composite end point risk was higher among patients with device-detected AT/AF (hazard ratio [HR]: 2.16, p = 0.032), New York Heart Association functional class III or IV compared with II (HR: 2.09, p = 0.002), and absence of beta-blockers (HR: 1.36, p = 0.036). Furthermore, the composite end point risk was inversely associated with left ventricular ejection fraction (HR: 1.04, p = 0.045), increasing by a factor of 4% for each 1% decrease in left ventricular ejection fraction.
Conclusions In HF patients with CRT-D, device-detected AT/AF is associated with a worse prognosis. Continuous device diagnostics monitoring and Web-based alerts may inform the physician of AT/AF occurrences and identify patients at risk of cardiac deterioration or patients with suboptimal rate or rhythm control. (Italian ClinicalService Project; NCT01007474)
Among chronic heart failure (HF) patients, atrial tachycardia (AT) or atrial fibrillation (AF) are common comorbidities and are associated with a worse prognosis (1–3). For patients with HF, systolic dysfunction and cardiac dyssynchrony, cardiac resynchronization therapy (CRT), through biventricular pacing, reduces the risk of death, reduces complications, and improves symptoms and quality of life (4,5). Benefits of CRT have also been demonstrated in patients with AT/AF (6,7).
Only a few studies have evaluated the incidence and clinical consequences of AT/AF in HF patients with CRT pacemakers (8) or CRT implantable cardioverter-defibrillators (CRT-D) (9). Device diagnostics allows a continuous monitoring of cardiac arrhythmias and an accurate evaluation of AT/AF occurrence and duration. The aim of our research was to characterize patients' AT/AF profiles, in terms of maximum AT/AF duration, and to evaluate the correlation between AT/AF and clinical outcomes, such as survival or HF hospitalizations, in a large population of patients with CRT-D.
Patients were included consecutively by 44 cardiological centers that participate in the Italian ClinicalService Project, a national medical care project aiming to evaluate and improve the use of implantable cardiac devices in clinical practice. Each patient signed an informed consent approved by each site's institutional review board.
All patients received a CRT-D, according to current guidelines (10,11), namely, advanced HF (New York Heart Association [NYHA] functional class II, III, or IV), depressed left ventricular function (left ventricular ejection fraction [LVEF] ≤35%), and wide QRS complex (≥120 ms). A history of AT/AF was defined as ≥1 AT/AF episode, documented by electrocardiogram or Holter monitor, in the period preceding implantation. Patients with permanent AT/AF or treated by pulmonary vein AT/AF ablation were excluded from the analysis.
Follow-up and end points
In-hospital follow-up visits were scheduled according to each center's clinical practice. Patients' clinical management, such as treatment of AT/AF episodes, was performed according to expert cardiologists' discretion.
Information about clinical outcomes such as hospitalizations and deaths were collected during scheduled or unscheduled hospital visits, or by phone calls for patients who missed programmed visits.
The incidence and duration of AT/AF was derived by device data, which comprise the total time spent by the patient in AT/AF for each day of the follow-up period. The AT/AF detection and its sensitivity and specificity, in the CRT-D used, have been previously described (12). Patients were considered to have experienced AT/AF episodes if the device detected a cumulative AT/AF duration >10 min in a day, a cut-off duration recognized as appropriate to discard false detections (8). As a function of the duration of the longest AT/AF period, observed during the follow-up period, each patient was classified according to 5 AT/AF profiles, whose length was >10 min, >6 h, >24 h, >7 days, and >6 months; the first 3 cut-offs characterize different forms of paroxysmal AT/AF, whereas the last 2 mimic persistent and permanent AT/AF, respectively.
Biventricular pacing percentage (BIVP%) was estimated as the number of paced biventricular beats divided by the number of paced or sensed ventricular beats over given time periods; in particular, BIVP% was estimated both during the whole lifetime of the device and during AT/AF periods.
The pacing mode was atrial-synchronous ventricular pacing. The automatic mode-switch feature was enabled in all patients to switch the pacing mode to a nonatrial tracking mode during AT/AF. The detection and therapy programming of the defibrillator was left to the clinical practice of each center.
Descriptive statistics were reported as mean and standard deviation for normally distributed continuous variables, or median with 25th and 75th percentiles in the case of skewed distribution. Categorical variables were reported as percentages. Comparisons of categorical variables were performed by means of the Fisher exact test or chi-square, as appropriate.
Survival analysis was performed by the Kaplan-Meier method, and the log-rank test was applied to evaluate differences between survival trends.
Hazard ratios (HRs) and their 95% confidence intervals (CIs) were computed by means of time-dependent Cox regression models, where AT/AF during follow-up was considered as a time-dependent covariate and baseline predictors as fixed covariates. After checking for collinearity, we included in the multivariate Cox models any baseline variable with p < 0.05 on univariate analysis, and age and sex regardless of their p values. For statistical analysis, Stata/SE version 11.0 for Windows (StataCorp LP, College Station, Texas) were used.
The baseline characteristics of the 1,193 patients are shown in Table 1.
AT/AF type and incidence during follow-up
The median follow-up period was 13 months (25th and 75th percentile: 8 and 20 months). AT/AF lasting longer than 10 min occurred in 361 of 1,193 (30%) patients. Figure 1 shows the number of patient with AT/AF occurrence, as a function of their AT/AF profile. Among 882 patients with no previous history of AT/AF, 178 (20%) had a new-onset AT/AF.
The incidences of the composite end point and its components—deaths and heart failure hospitalizations—are summarized in Table 2, for various patient subgroups identified as a function of AT/AF occurrence, duration, and type.
Table 3 shows results of multivariate time-dependent Cox regression analyses performed to evaluate predictors of death, HF hospitalizations and the composite end point. In the whole population, the median value of the lifetime BIVP% was 98% (25th and 75th percentiles: 95% and 99%). In patients with AT/AF, it was 95% (91% and 99%) during the whole follow-up period, and more particularly, 98% (95% and 99%) during sinus rhythm and 71% (51% and 92%) during AT/AF (p < 0.01 vs. sinus rhythm period). In patients with AT/AF occurrence, the condition of nonoptimal CRT, namely, lifetime BIVP% <95%, was found to be an independent predictor of clinical outcomes, associated with a higher risk of the composite end point (HR: 2.0, 95% CI: 1.00 to 4.05, p = 0.05), of death (HR: 6.6, 95% CI: 1.70 to 25.63, p = 0.006), and of HF hospitalization (HR: 3.14, 95% CI: 1.36 to 7.25, p = 0.007).
The Kaplan-Meier survival analyses showed higher incidences of both the composite end point and HF hospitalizations among patients with AT/AF during follow-up when compared to patients without AT/AF, as shown in Figure 2.
Real-life registries have an important role in identifying novel risk factors. The present analysis adds to the previous literature in that it shows the impact of device-detected AT/AF on adverse clinical outcomes, associating this impact with specific AT/AF profiles and with suboptimal CRT.
Impact of AT/AF on clinical outcomes
Device-detected AT/AF was significantly correlated with the composite end point (death or HF hospitalizations) at multivariate time-dependent Cox regression analysis, at Kaplan-Meier survival analysis, and by simple proportion analysis test. The explanation for these findings may be that AT/AF causes loss of atrial systole, irregular ventricular rhythm, and tachycardia-related contractile dysfunction. Moreover, in CRT-D recipients, AT/AF may worsen a patient's health by dramatically reducing biventricular pacing; in fact, BIVP% achieved 98% during sinus rhythm and only 71% during AT/AF. A BIVP% <95% was significantly associated with both HF hospitalizations and death.
Summary of knowledge from previous studies
The importance of BIVP% has been shown by Gasparini et al. (6,7) in CRT patients with permanent AT/AF by comparing atrioventricular node ablation versus pharmacological rate control, apparently achieving adequate biventricular pacing. These authors showed that atrioventricular node ablation, ensuring 100% effective CRT, was associated with significant improvement in left ventricular function and volumes, exercise capacity, and mortality.
Our results indicating that AT/AF occurrence during follow-up is an independent predictor of adverse clinical events contribute and extend the discussion of whether AT/AF is independently associated with a worse outcome (1–3,6–9). Benjamin et al. (1) showed that patients in who AT/AF developed during 40 years of follow-up of the Framingham Heart Study cohort were associated with a 1.9-fold mortality risk. More recently, the CARE-HF (Cardiac Resynchronization in Heart Failure) trial (8) showed that patients who had AT/AF presented with higher rates of deaths or unplanned cardiovascular hospitalizations. Evaluating this correlation, the CARE-HF study (8) authors only considered AT/AF as documented by electrocardiography, whereas our findings extend the correlation with death or HF hospitalizations to device-detected AT/AF.
Borleffs et al. (9) studied CRT-D patients with no AT/AF history and showed that new-onset AT/AF was associated with a higher incidence of cardiac hospitalizations. Our data confirm and extend this finding since we found that any AT/AF occurrence, new-onset or recurrence, correlated with a higher incidence of HF hospitalizations.
Clinical relevance of AT/AF duration assessment
Patients with persistent or permanent AT/AF had higher incidences of the composite end point or HF hospitalization when compared with patients in sinus rhythm. This finding may be explained by the negative impact of a prolonged CRT loss due to AT/AF-induced high spontaneous ventricular rate.
Figure 1, showing that many new-onset AT/AF episodes have long duration, suggests that this population is at high risk of embolic complications and may benefit from continuous AT/AF monitoring.
Our results may not be extended to a general, nondevice, HF population and, in particular, not to HF patients without CRT.
For HF patients treated with CRT-D, device-detected AT/AF was correlated with worse prognosis in terms of deaths or HF hospitalizations. In particular, this association appeared mainly mediated by the subgroup of patients with persistent AT/AF, because of the negative impact of a prolonged CRT loss caused by AT/AF-induced high ventricular rate.
These correlations suggest that AT/AF continuous monitoring may be a useful tool for identifying patients at risk of cardiac deterioration or patients with suboptimal rate or rhythm control. AT/AF occurrences may trigger re-evaluation of current treatment and, if appropriate, pharmacological or electrical interventions to control ventricular rate, such as pharmacological blockade or ablation of the atrioventricular node conduction.
Our findings suggest the importance of utilizing continuous device monitoring and Web-based care alerts based on AT/AF duration or on ventricular rate or BIVP% during AT/AF.
Dr. Santini has received research grants from Medtronic and St. Jude. Dr. Proclemer has received research grants from Medtronic. Dr. Padeletti has received research grants from Medtronic, Boston Scientific, St. Jude, and Sorin SpA. Ms. LaRocca and Dr. Grammatico are employees of Medtronic Italia, an affiliate of Medtronic Inc. All other authors have reported that they have no relationships to disclose.
- Abbreviations and Acronyms
- atrial fibrillation
- atrial tachycardia
- biventricular pacing percentage
- confidence interval
- cardiac resynchronization therapy
- cardiac resynchronization therapy defibrillator
- heart failure
- hazard ratio
- left ventricular ejection fraction
- New York Heart Association
- Received April 7, 2010.
- Revision received June 28, 2010.
- Accepted August 10, 2010.
- American College of Cardiology Foundation
- Benjamin E.J.,
- Wolf P.A.,
- D'Agostino R.B.,
- Silbershatz H.,
- Kannel W.B.,
- Levy D.
- Wang T.J.,
- Larson M.G.,
- Levy D.,
- et al.
- Young J.B.,
- Abraham W.T.,
- Smith A.L.,
- et al.,
- Multicenter InSync ICD Randomized Clinical Evaluation (MIRACLE ICD) Trial Investigators
- Gasparini M.,
- Auricchio A.,
- Regoli F.,
- et al.
- Gasparini M.,
- Auricchio A.,
- Metra M.,
- et al.
- Hoppe U.C.,
- Casares J.M.,
- Eiskjær H.,
- et al.
- Vardas P.E.,
- Auricchio A.,
- Blanc J.J.,
- et al.
- Zipes D.P.,
- Camm A.J.,
- Borggrefe M.,
- et al.