Evaluation of time course and predicting factors of progression of paroxysmal or persistent atrial fibrillation to permanent atrial fibrillation.

BACKGROUND: To evaluate time course and predictors of progression of paroxysmal or persistent atrial fibrillation (AF) to permanent AF. METHODS AND RESULTS: We included 460 patients referred for paroxysmal (n = 337) or persistent (n = 123) AF between 1994 and 2012. Mean follow-up was 13.2 ± 6.5 years. AF progression rate was 3.7% per year, 19.7% at 5 years, and 38.1% at 10 years. Lone AF was diagnosed in 217 patients (47%). Predictors of permanent AF were: age, persistent AF, left atrial (LA) size, left ventricular-fractional shortening (LV-FS), lack of antiarrhythmic (AA) drugs, VVI pacing (P < 0.001 for all), and valvular disease (P < 0.02). Independent predictors were age (P < 0.001), persistent AF (P < 0.001), LA diameter (P < 0.005), lack of AA drugs (P < 0.005), and VVI pacing (P < 0.01). When adjusted at means of covariates, persistent AF and age >75 years remained highly significant (P < 0.01). LA dimension >50 mm was highly significant at univariate model (P < 0.001) but to a lesser extent when adjusted (P < 0.05). In patients with paroxysmal AF-with age <75 years-on AA drugs, progression rate to permanent AF was 6.5% at 5 years and 23.7% at 10 years. Among four predictors (age, LA size, LV-FS, and VVI pacing), only age (P < 0.01) and LA size (P < 0.005) remained independently significant, but LA size was not significant when adjusted. CONCLUSIONS: Progression to permanent AF is a slow process. Aging, LA size, VVI pacing, lack of AA therapy, and a persistent form of AF independently increased the progression to permanent AF.

Bigeminy pacing: a new protocol to unmask atrial vulnerability.

INTRODUCTION: One of the most exciting developments in our understanding of atrial fibrillation (AF) mechanisms has been the recognition that "AF begets AF" in a process termed atrial remodeling. Little information is available about the events that mediate short-term remodeling. In a bigeminy atrial pacing protocol that produces a continuous extrasystole-postextrasystole cycle length, we sought to evaluate the electrophysiologic consequences of irregular atrial pacing. METHODS AND RESULTS: This study included 22 consecutive patients with documented paroxysmal AF and 10 control subjects. After evaluating the effective refractory period (ERP) and functional refractory period (FRP), bigeminy atrial pacing was performed for 5 minutes. The S1-S2 coupling interval during bigeminy pacing was programmed to a mean value of 275 +/- 45 msec, i.e., 45 msec longer than the basic ERP measured at 100 beats/min. During bigeminy pacing, AF that lasted longer than 1 minute occurred in 12 AF patients and in none of the control subjects (group I). Short salvos of AF occurred in 5 patients and 3 controls (group II). No arrhythmia occurred in 5 patients and 7 controls (group III). Sensitivity, specificity, and negative and positive predictive values of sustained AF induced by bigeminy pacing were 54%, 100%, 50%, and 100%, respectively. No differences were observed between different pacing rates during bigeminy, the premature coupling interval S1-S2, or the conduction parameters S2-A2 and A2. Group I had the shortest basic ERP (222 +/- 38 msec) and group III the longest ERP (242 +/- 21 msec, P < 0.05); group II was intermediate. Atrial ERPs and FRPs measured immediately after termination of 5 minutes of bigeminy pacing were shorter than during baseline. The degree of shortening was similar in AF patients and in controls. The locoregional conduction delay A2 did not change after the bigeminy protocol. CONCLUSION: This study demonstrates that atrial bigeminy pacing highly increases atrial vulnerability. This protocol appears interesting because its sensitivity and specificity are higher than those of the conventional extrastimulation test. This makes it attractive for routine diagnosis of undocumented paroxysmal AF. Because it may induce atrial arrhythmias independently of the classic mechanisms of wavelength shortening, this study emphasizes the need for new modalities in the prevention of atrial arrhythmias.

P wave duration and morphology predict atrial fibrillation recurrence in patients with sinus node dysfunction and atrial-based pacemaker.

P wave duration and morphology have never been systematically evaluated as markers of AF in patients with a conventional indication to pacing. This study correlated sinus P wave duration and morphology and the incidence of AF in patients with sinus node dysfunction (SND), previous history of AF before implant, and atrial-based pacemaker. Included were 140 patients (86 men, 54 women; mean age 71.8 +/- 10.4 years) with recurrent paroxysmal AF and who received a DDD (128 patients) or AAI (12 patients) pacemaker for SND. Forty-nine patients had structural heart disease. Sinus P wave duration and morphology was evaluated in leads II, III. Twenty-two patients had an abnormal P wave morphology, diphasic (+/-) in 5 and notched (+/+) in 17. The basic pacemaker rate was programmed between 60 and 70 beats/min. Rate responsive function was activated in 65 patients. During a follow-up of 27.6 +/- 17.8 months, AF was documented in 87 patients. Forty-four patients developed permanent AF, following at least one episode of paroxysmal AF in 26 cases. Statistical analysis used Cox model regression. Univariate predictors of AF (P < 0.10) were drugs (mean: 2 +/- 1.4) and DC shock before pacing (16/140 patients), P wave duration (mean 112.5 +/- 24.6 ms), basic pacemaker rate (mean 68 +/- 5 beats/min), and drugs in the follow-up (mean 1.2 +/- 0.94). Multivariate analysis showed that P wave duration (b = 0.013, s.e. = 0.004; P = 0.003), and drugs before pacing (b = 0.2; s.e. = 0.08; P < 0.01) resulted in a significant independent predictor of AF. Actuarial incidence of patients free of AF at 30 months was 35%: 56% in patients with a P wave < 120 ms, and 13% in those with P wave > or = 120 ms (P < 0.01 by Score test). Univariate predictors of permanent AF were drugs and DC shock before pacing, left atrial size (mean 39 +/- 6 mm), P wave duration, abnormal P wave morphology (22/140 patients), and drugs in the follow-up. Multivariate analysis showed that P wave morphology was the most important predictor of permanent AF (b = -0.56, s.e. = 0.2; P = 0.008). Incidence of patients free of permanent AF at 30 months was 69%: 74% in patients with normal P wave, compared to 28% in the case of abnormal P wave morphology (P < 0.01). P wave duration and morphology are good markers of postpacing AF recurrence in patients with SND and an atrial-based pacemaker. This observation suggests that intra- and interatrial conduction disturbances be extensively evaluated before implantation, and the indication for atrial resynchronization procedures be reevaluated.