Atrial volumes indexed to body surface area (AVI) are robust predictors of nonvalvular atrial fibrillation (AF) recurrence after direct current cardioversion (DCCV). The incremental value of atrial emptying fraction (EmF) compared with atrial volumes as a predictor for recurrent AF after DCCV has not been evaluated. We sought to compare the predictive ability of baseline left atrial (LA) EmF, right atrial (RA) EmF, LAVI, and RAVI for post-DCCV AF recurrence at 6 months. The first 95 patients enrolled in the AF Clinic Registry with adequate echocardiogram imaging constituted the study cohort. Each patient underwent echocardiogram within 6 months before cardioversion. Maximal LAVI and RAVI, LA EmF, and RA EmF were performed offline using 4-chamber single-plane Simpson’s method, averaged over 5 cycles. The mean age of the study cohort was 64 ± 12 years, and 67% were men. Only 28 patients (29%) who underwent DCCV remained in sinus rhythm at 6 months of follow-up. The remaining, 67 (71%) had reverted to AF or underwent ablation during the 6 months of follow-up. The overall performance for prediction of AF recurrence was greatest for RA EmF, area under the receiver operator characteristic curve (AUC): RA EmF 0.92, LA EmF 0.89, RAVI 0.76, and LAVI 0.63. RA and LA EmF AUCs were significantly higher than for LAVI or RAVI (max p = 0.02). In conclusion, although RAVI and LAVI are strong predictors of AF recurrence after DCCV, RA and LA EmF outperformed in this cohort.
Atrial fibrillation (AF) causes structural and electrical remodeling, including atrial enlargement and fibrosis. Both left atrial (LA) and right atrial (RA) enlargements are robust predictors of AF recurrence after direct current cardioversion (DCCV). Our group was the first to demonstrate that RA volume was superior to LA volume for prediction of AF recurrence after DCCV. However, atrial enlargement is likely a late finding of remodeling and atrial function may be an earlier indicator of AF-related changes. Several small studies have examined the predictive ability of LA function for recurrence of AF after pulmonary vein ablation. These studies show that impaired LA function is associated with AF recurrence after pulmonary vein ablation and that the predictive ability is superior to that of LA size. The relation of atrial function and recurrence of AF after DCCV has not been investigated. We sought to assess the utility of echocardiographic LA and RA emptying fraction (EmF) as a predictor of AF recurrence after DCCV.
Methods
The study cohort consisted of the first 95 patients with persistent AF who underwent DCCV and had adequate echocardiographic imaging since the launch of the Vancouver General Hospital Atrial Fibrillation Clinic. These patients met the following criteria: freedom from rheumatic or prosthetic valve disease, freedom from congenital heart disease, echocardiogram within 6 months before DCCV, and clinical follow-up for at least 6 months after DCCV. Subjects with echocardiogram images that were not available for repeat measurements were excluded. If there was more than one DCCV recorded for a specific subject, the subsequent cardioversions were excluded. This study was reviewed and approved by the University of British Columbia Clinical Review Ethics Board.
The electronic database included a standard set of parameters for all AF Clinic patients including: Congestive heart failure history, Hypertension, Age, Diabetes, Stroke/Transient ischemic attack (CHADS 2 score), gender, AF duration, antiarrhythmic drug use, pulmonary disease, dyslipidemia, and smoking status. Previous and subsequent procedures including repeat cardioversion, pulmonary vein ablation, and device implantation were documented. Clinical and electrocardiogram (ECG) follow-up was obtained for all patients at 1 week, 1 month, 3 months, 6 months, and with changes in clinical status.
Comprehensive transthoracic echocardiogram was performed in all patients before DCCV. Studies were performed with an iE33 (Philips Medical Systems, Andover, Massachusetts) equipped with an X3-1 or X5-1 transducer by experienced sonographers. Standard M-mode, Doppler imaging, and 2-dimensional cine loops of the parasternal long and short axis, and of the apical 2-, 3-, and 4-chamber views were obtained in each patient. All images and measurements were acquired from the standard views, according to the guidelines of the American Society of Echocardiography and were digitally stored for offline analysis.
Atrial volumes were measured offline by a reader blinded to the clinical data and DCCV outcomes. Maximum and minimum LA and RA volumes were measured and indexed to body surface area (LAVI, RAVI) using the 4-chamber single-plane Simpson’s method, which has been validated against magnetic resonance imaging. Maximum volume was obtained at end systole, on the frame just before mitral or tricuspid valve opening by tracing the inner border of the atrium taking care to avoid the area under the valve annulus, appendages, and pulmonary veins. Minimum volume was obtained at end diastole, on the frame of mitral or tricuspid valve closure. All volumes were averaged over 5 consecutive cardiac cycles. Left and RA EmF, a measure of reservoir function, was calculated as (maximum atrial volume − minimum atrial volume)/maximum atrial volume and expressed as a percentage for each side. Reservoir function is primarily determined by atrial compliance during ventricular systole and decreases with worsening diastolic dysfunction. All remaining measurements were obtained from the electronic echocardiography database.
Electronic medical records were available for review on all patients, and ECG confirmation was required for ascertainment of AF. At the 6-month time point, subjects were recorded as either being in sinus rhythm (SR) or experiencing AF recurrence. All but 2 subjects who underwent ablation during the 6-month period had recurrence of AF and were recorded as not in SR at 6 months.
Results are described as mean ± SD for continuous variables and percentage for categorical variables. Continuous variables are compared using a 2-sample t test or Mann–Whitney test if distributional assumptions are violated and categorical variables are compared using the Fisher’s exact test. Multivariable logistic regression analysis was used to obtain adjusted odds ratios (ORs) for SR at 6 months with p values for ORs obtained using a Wald test. Due to the large number of predictors relative to the sample size and number of subjects remaining in SR at 6 months, a penalized logistic regression model (Elastic Net) was used to obtain stable estimated ORs with p values obtained through resampling. Standard receiver operator characteristic curves were constructed, with estimated area under the receiver operator characteristic curve (AUC) compared through permutation tests and confidence intervals through bootstrapping. Optimal cutoffs for classification were determined using the Youden index. Accuracy is calculated as (TP +TN)/(P + N), where TP = true positives, TN = true negatives, P =positives, and N=negatives. To account for overfitting of regression models, optimism-adjusted AUCs were obtained through bootstrapping. All tests are 2 sided and compared with a 0.05 alpha level, as an exploratory analysis, there is no adjustment for multiple comparisons. Statistical analysis was conducted using R version 3.2 software and the following packages: glmnet, ROCR, and pROC.
Results
There were 156 elective DCCV completed in 136 subjects from December 15, 2010, to March 13, 2014. For the purposes of analysis, 41 subjects were excluded; 23 had echocardiograms who were older than 6 months at the time of DCCV, 17 subjects had echocardiograms who were not accessible for repeat measurement, and 1 subject had rheumatic valvular disease.
The remaining 95 patients (64 men and 31 women) who underwent DCCV were of mean age 64 ± 12 years; 46 patients were ≥65 years (48%). A history of systemic hypertension, heart failure, diabetes mellitus, and transient ischemic attack/stroke was identified in 60 (63%), 27 (28%), 14 (15%), and 5 (5%) patients, respectively. There was 1 patient with a permanent pacemaker before DCCV and another who underwent pacemaker implantation during the 6 months of follow-up for sinus node dysfunction. There were 9 patients who had undergone previous surgical atrial ablation (MAZE) procedures and/or pulmonary vein ablation and an additional 19 patients who had previous cardioversions. Antiarrhythmic use was common before DCCV, with 21 patients (22%) treated with either amiodarone or dronedarone, 14 patients (15%) treated with a class I antiarrhythmic (propafenone or flecainide), and 6 patients (6%) treated with sotalol. After cardioversion, 23 patients (24%) were treated with amiodarone or dronedarone, 11 patients (12%) with a class I antiarrhythmic (propafenone or flecainide), and 10 patients (11%) with sotalol. Antiarrhythmic use was not associated with reduced risk of AF recurrence (p = 0.65). Patients who failed to remain in SR were more likely to have been treated with amiodarone or dronedarone (36%) at any point, compared with class I antiarrhythmics (16%) or sotalol (10%).
At 6 months after DCCV, 28 patients (29%) remained in SR; 11 (12%) required ablation, 9 (9%) required at least one additional DCCV, and the remaining 47 had AF recurrence without further intervention. Of the 11 patients who underwent ablation, all but 2 patients had recurrent AF before the procedure. Of those ablated during follow-up, 8 (89%) were found to be in SR at 6 months after DCCV. These subjects were not considered as having a durable response to DCCV through 6 months for the purposes of the regression analyses.
In univariable analysis ( Table 1 ), subjects who remained in SR at 6 months were more likely to have been diabetic (p = 0.024); however, no other clinical characteristic was associated with durable response through 6 months. Atrial volumes and atrial EmF were significantly different between those who remained in SR and those who reverted to AF (all p <0.05).
| Variable | All Subjects | SR at 6 Months | Not in SR at 6 Months | P-Value | Ablated Within 6- Months of DCCV | Repeat DCCV Within 6-Months |
|---|---|---|---|---|---|---|
| N | 95 (100%) | 28 (29%) | 67 (71%) | 11 (12%) | 9 (9%) | |
| Age (years) | 63.5 ± 11.6 | 60.9 ± 11.2 | 64.5 ± 11.6 | 0.16 | 60.2 ± 11.6 | 62.4 ± 16.5 |
| Age (≥ 65 years) | 46 (48%) | 12 (43%) | 34 (51%) | 0.51 | 4 (36%) | 3 (33%) |
| Male | 64 (67%) | 19 (68%) | 45 (67%) | 1 | 9 (82%) | 5 (56%) |
| Current Smoker | 2 (2%) | 0 | 2 (3%) | 1 | 0 | 0 |
| Former Smoker | 39 (41%) | 14 (50%) | 25 (37%) | 0.26 | 2 (18%) | 5 (56%) |
| CHADS 2 score | 1.3 ± 0.9 | 1.3 ± 1 | 1.3 ± 0.9 | 0.92 | 1.3 ± 0.8 | 1.2 ± 1 |
| Congestive heart failure | 27 (28%) | 7 (25%) | 20 (30%) | 0.80 | 5 (45%) | 2 (22%) |
| Hypertension | 60 (63%) | 17 (61%) | 43 (64%) | 0.82 | 7 (64%) | 7 (78%) |
| Diabetes mellitus | 14 (15%) | 8 (29%) | 6 (9%) | 0.024 | 1 (9%) | 0 |
| Stroke/Transient ischemic attack | 5 (5%) | 2 (7%) | 3 (4%) | 0.63 | 0 | 0 |
| Atrial fibrillation duration (years) | 3.5 ± 4.9 | 3.2 ± 6.6 | 3.7 ± 4.1 | 0.45 | 5.1 ± 4.9 | 2.2 ± 2.6 |
| Antiarrhythmic use before or after cardioversion | 55 (58%) | 15 (54%) | 40 (60%) | 0.65 | 8 (73%) | 5 (56%) |
| Pulmonary Disease | 20 (21%) | 4 (14%) | 16 (24%) | 0.41 | 2 (18%) | 2 (22%) |
| Dyslipidemia | 40 (42%) | 11 (39%) | 29 (43%) | 0.82 | 6 (55%) | 3 (33%) |
| Mitral regurgitation severity ∗ | 0.9 ± 0.6 | 0.8 ± 0.5 | 1 ± 0.7 | 0.13 | 1 ± 1.1 | 1 ± 0.8 |
| Left ventricular ejection fraction (%) | 56.2 ± 10.3 | 56.2 ± 8 | 56.2 ± 11.1 | 0.98 | 55.2 ± 9.5 | 58.6 ± 12.1 |
| Left atrial diameter (mm) | 43.3 ± 5.8 | 42.1 ± 5.2 | 43.8 ± 6 | 0.18 | 46.6 ± 6.4 | 43.1 ± 4.7 |
| Left atrial volume index (mL/m 2 ) | 47.5 ± 11.7 | 44.2 ± 8.8 | 48.9 ± 12.5 | 0.041 | 52.5 ± 21.6 | 44.4 ± 10.2 |
| Right atrial volume index (mL/m 2 ) | 45.2 ± 12.4 | 39.2 ± 6.7 | 47.8 ± 13.4 | <0.001 | 44.9 ± 10.7 | 43.1 ± 13.1 |
| Left atrial emptying Fraction (%) | 43.1 ± 7.4 | 50.1 ± 4.4 | 40.2 ± 6.4 | <0.001 | 44.5 ± 7.6 | 39.9 ± 6.5 |
| Right atrial Emptying Fraction (%) | 41.5 ± 12.6 | 55.2 ± 6.2 | 35.8 ± 9.8 | <0.001 | 45 ± 11.8 | 38.9 ± 10.7 |
∗ Mitral regurgitation severity was quantified as: 1 = mild, 2 = moderate, 3 = moderate-severe, and 4 = severe.
The results of the multivariable logistic regression model are listed in Table 2 . This analysis identified predictors of durable SR through 6 months after DCCV, accounting for the influence of all other variables. Atrial EmF was a statistically significant echocardiographic predictor of durable SR in full multivariable regression and penalized multivariable regression ( Table 2 ). The penalized model provides stabilized OR estimates because the sample size was small relative to the number of predictors in the model. The penalized regression demonstrates that both LA and RA EmFs have substantial and independent predictive ability for durable SR, with EmF being protective (p = 0.031 and <0.001, respectively). The odds of durable SR are increased by 80% for every 5% increase in RA EmF and by 10% for every 5% increase in LA EmF.
| Variable | Full Model | Penalized Model | ||
|---|---|---|---|---|
| OR | P-Value | OR | P-Value | |
| Age (1 year) | 1.1 | 0.26 | 1.0 | 0.41 |
| Sex (ref: female) | 5.9 | 0.31 | 1.0 | 0.84 |
| Current or past Smoker | 3.0 | 0.49 | 1.6 | 0.25 |
| CHADS 2 (1 unit) | 0.3 | 0.43 | 1.0 | 0.92 |
| Hypertension | 0.4 | 0.67 | 1.0 | 0.71 |
| Stroke/Transient Ischemic attack | 126.4 | 0.62 | 1.0 | 0.70 |
| Diabetes | 2029.3 | 0.051 | 3.2 | 0.044 |
| AF Duration (1 year) | 0.8 | 0.089 | 1.0 | 0.45 |
| Antiarrhythmic use before or after DCCV | 0.4 | 0.52 | 1.0 | 0.65 |
| Pulmonary Disease ∗ | 8.4 | 0.20 | 1.0 | 0.83 |
| Dyslipidemia | 0.1 | 0.19 | 0.7 | 0.28 |
| Mitral regurgitation severity (per grade increase † ) | 0.1 | 0.12 | 0.7 | 0.24 |
| Left ventricular ejection fraction (10%) | 0.3 | 0.24 | 1.0 | 0.74 |
| Left atrial diameter (1 mm) | 0.7 | 0.034 | 0.6 | 0.13 |
| LAVI (5 mL/m 2 ) | 1.1 | 0.68 | 1.0 | 0.39 |
| RAVI (5 mL/m 2 ) | 0.8 | 0.57 | 1.0 | 0.48 |
| LA EmF (5%) | 0.3 | 0.27 | 1.1 | 0.031 |
| RA EmF (5%) | 15.7 | 0.015 | 1.8 | < 0.001 |
∗ Chronic obstructive pulmonary disease, obstructive sleep apnea, or pulmonary hypertension.
† Mitral regurgitation severity was quantified as: 1 = mild, 2 = moderate, 3 = moderate-severe, and 4 = severe.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
