Obese patients with atrial fibrillation (AF) are frequently treated with AF ablation. We sought to examine whether a body mass index (BMI) threshold exists beyond which the odds of experiencing a complication from AF ablation increases. All patients enrolled in the Vanderbilt AF Registry who underwent catheter-based AF ablation from May 1999 to February 2012 were included. Major complications were recorded. Morbid obesity was defined as a BMI >40 kg/m 2 and examined in multivariable analysis. A total of 35 complications (6.8%) occurred in 512 ablations. Morbidly obese patients experienced a greater rate of complications (6 of 42, 14.3%) than the nonmorbidly obese (29 of 470, 6.2%; p = 0.046). Using a discrete BMI cutoff, the odds of complications increased 3.1-fold in those with morbid obesity (odds ratio [OR] 3.1, 95% confidence interval [CI] 1.1 to 8.4, p = 0.03) and 2.1-fold for female gender (OR 2.1, 95% CI 1.04 to 4.38, p = 0.04). With BMI as a continuous variable, the odds of complications increased by 5% per 1 unit increase in BMI (OR 1.05, 95% CI 1.0 to 1.11, p = 0.05), and the increase for female gender was 2.2-fold (OR 2.2, 95% CI 1.1 to 4.6, p = 0.03). In conclusion, morbid obesity represents a BMI threshold above which the odds of complications with AF ablation increase significantly. The increase in complications appears to be driven primarily by events in women, suggesting that morbidly obese women are a special population when considering AF ablation.
Many studies have examined the predictors for procedural complications with atrial fibrillation (AF) ablation, but only a few have included obesity or body mass index (BMI) in the analysis. Previous studies have not found an association between complications from AF ablation and obesity, which might have been because of the nonlinear effect of BMI that has been recognized in a variety of other outcomes. In studies examining the rate of complications in patients undergoing percutaneous coronary intervention (PCI), the rate of complications was found to paradoxically decrease in patients with mild to moderate obesity before increasing in patients with morbid obesity. Morbid obesity was considered to represent the BMI threshold in which the rate of complications significantly increased in patients undergoing PCI. In the present study, we sought to examine the hypothesis that morbid obesity represents a threshold beyond which the risk of procedural complications with AF ablation significantly increases.
Methods
We included 445 patients with symptomatic AF who were prospectively enrolled in the Vanderbilt AF Registry, a clinical and genetic database, and underwent catheter-based ablation for AF from May 1999 to March 2012. Ablations included de novo and repeat procedures. Surgical and combined hybrid catheter plus surgical ablation procedures were excluded. The patient characteristics and procedural details were entered into a central database. Paroxysmal AF was defined as episodes lasting <7 days and spontaneously terminating. Nonparoxsymal AF was defined as AF episodes lasting >7 days and/or requiring termination with pharmacologic or electrical cardioversion. The number of lifetime direct current cardioversions before ablation was recorded according to patient report. The BMI was calculated as the weight in kilograms divided by the height in meters squared. Using current recommendations, obesity was defined as a BMI of 30 to 40 kg/m 2 and morbid obesity as a BMI >40 kg/m 2 .
Major complications were defined as those resulting in permanent injury, death, intervention, or requiring or prolonging hospitalization. Pulmonary vein (PV) stenoses were included if they required venoplasty. Acute lung injury was defined as pneumonia or fluid-overload requiring or prolonging ventilator support.
All patients received general anesthesia during ablation. Vascular access was obtained from the right and/or left femoral veins with or without the right internal jugular veins, according to operator preference for coronary sinus cannulation. Vascular access was obtained using anatomic and fluoroscopic landmarks. All ablations were performed using biplane fluoroscopy. Left atrial access was obtained using transseptal puncture under anteroposterior and left anterior oblique fluoroscopic views with assistance from intracardiac echocardiography.
The standard periprocedural anticoagulation strategies used varied throughout the study period. In cases in which the preprocedure therapeutic international normalized ratio for ≥1 month was not documented, a transesophageal echocardiogram was performed before ablation to document the absence of left atrial thrombus. Before 2009, warfarin was discontinued 5 days before the procedure with the use of a low-molecular-weight heparin bridge. Warfarin was restarted the morning of the procedure, and therapeutic anticoagulation with heparin was performed with transseptal puncture and maintenance of the activated clotting time at 300 to 350 seconds throughout the procedure. Heparin or low-molecular-weight heparin was continued after the procedure until a therapeutic international normalized ratio was achieved. Since 2009, catheter ablation has been performed without interruption of warfarin and with a therapeutic international normalized ratio goal of 2 to 2.5 on the day of the procedure. Since 2011, patients who received anticoagulation with dabigatran held the anticoagulant for 12 to 48 hours before the procedure according to the individual patient risk factors for bleeding (withheld for >24 hours if patient age >75 years, creatinine clearance <30 ml/min, or aspirin or clopidogrel use) with resumption 4 to 24 hours after removal of the sheath and hemostasis. Protamine was variably administered at completion of the procedure to facilitate sheath removal in patients with a therapeutic international normalized ratio. Anticoagulation with warfarin or dabigatran was continued for ≥3 months after ablation.
Ablations were performed according to standard techniques, contemporary for the year of ablation. The standard ablation procedure throughout the study period consisted of antral or segmental PV isolation (PVI) confirmed by entrance block, with additional linear ablation and ablation of non-PV foci, according to operator discretion. From 1999 to 2003, the primary ablative approach consisted of segmental PVI with use of a multipolar circular mapping catheter. Ablation was performed with an 8F, 5-mm tip, temperature-controlled radiofrequency catheter at the venoatrial junction. Radiofrequency energy was applied with a temperature of 55°C to 60°C and a maximum power of 50 W for 30 to 45 seconds at each site. Since 2003, a wide-area circumferential antral ablation has been performed with use of a 3-dimensional electroanatomic mapping system and placement of contiguous lesions 5 to 15 mm from the PV ostia, with testing for entrance block. As of 2009, an 8F, 3.5-mm tip, open irrigated-tip, power-controlled ablation catheter was used. For circumferential ablation, a maximum power of 25 W was applied on the posterior wall and 30 to 35 W on the anterior wall and roof. Entrance and exit block were tested using high output pacing with isoproterenol and/or adenosine administration, directing additional lesions as necessary to achieve complete bidirectional electrical PVI of all PVs. Since 2011, PVI has also been performed in select cases using cryoablation (Arctic Front, Medtronic, Minneapolis, Minnesota).
Continuous variables are expressed as the mean ± SD. The categorical variables are presented as percentages and frequencies. Patients were categorized by their BMI, and the characteristics are displayed according to BMI category. Univariate analysis was performed using a chi-square test for nominal variables and the Mann-Whitney U test for continuous variables. Multivariable analysis was performed using a binary logistic regression model to predict the occurrence of complications. The covariates chosen for inclusion in the final multivariable logistic regression analysis included age, gender, and coronary artery disease. They were prespecified and selected because of previous studies that demonstrated them to be significant predictors of complications on multivariable analysis. Multivariable analysis was performed to examine morbid obesity as defined by the BMI and dichotomized at 40 kg/m 2 (model 1) and BMI as a continuous variable (model 2). Two-sided p values <0.05 were considered statistically significant.
Results
The complete baseline patient characteristics and procedural details are listed according to BMI in Table 1 . A total of 564 ablations were performed, of which 512 met the eligibility criteria (52 were excluded because the procedure was hybrid or surgical ablation). The mean BMI for the overall cohort was 31 ± 6 kg/m 2 . Of the 512 patients, 42 (8%) were morbidly obese. Of these 42 patients, 26 (62%) were men 16 (38%) were women. Morbidly obese patients were no more likely than the overall cohort to be taking clopidogrel (2.4% vs 2.5%) or aspirin (41% vs 45%) or to be undergoing a redo procedure (14% vs 22%).
| Variable | BMI (kg/m 2 ) | ||
|---|---|---|---|
| <30 (n = 260) | 30–39 (n = 210) | >40 (n = 42) | |
| Age (yrs) | 61 ± 10 | 57 ± 9 | 56 ± 10 |
| Men | 181 (70%) | 159 (76%) | 26 (62%) |
| Body mass index (kg/m 2 ) | 27 ± 2.4 | 34 ± 2.6 | 44 ± 5 |
| Duration of atrial fibrillation (months) | 78 ± 82 | 86 ± 74 | 71 ± 49 |
| Paroxysmal atrial fibrillation | 133 (51%) | 88 (42%) | 17 (40%) |
| Direct current cardioversions | 1.3 ± 1.8 | 1.4 ± 2.0 | 1.7 ± 3.2 |
| Coronary artery disease | 35 (14%) | 46 (22%) | 15 (36%) |
| Obstructive sleep apnea | 36 (14%) | 69 (33%) | 24 (57%) |
| Clopidogrel | 3 (1%) | 9 (4%) | 1 (2%) |
| Aspirin | 117 (45%) | 97 (46%) | 17 (40%) |
| History of amiodarone use | 84 (32%) | 83 (40%) | 18 (43%) |
| Left atrial size ∗ (mm) | 37 ± 7 | 41 ± 7 | 46 ± 9 |
| Left ventricular ejection fraction ∗ (%) | 63 ± 11 | 62 ± 12 | 60 ± 13 |
| Redo ablation | 54 (21%) | 54 (26%) | 6 (14%) |
| Cavotricuspid isthmus ablation | 50 (19%) | 46 (22%) | 11 (26%) |
| Roof line | 90 (35%) | 78 (37%) | 23 (55%) |
| Mitral valve line | 56 (22%) | 43 (20%) | 8 (19%) |
| Complex fractionated atrial electrogram ablation | 18 (7%) | 28 (13%) | 8 (19%) |
| Cryoballoon | 5 (2%) | 2 (1%) | 0 (0%) |
| Procedure time (min) | 280 ± 78 | 293 ± 78 | 310 ± 78 |
A total of 35 complications occurred, for an overall rate of 6.8% (35 of 512; Figure 1 ). Six complications were observed in patients classified as morbidly obese, for a complication rate of 14.3% (6 of 42; Figure 2 ). Of the 16 morbidly obese women, 4 (25%) experienced a complication compared to 2 (7.7%) of the 26 men ( Figure 3 ). Of the 6 complications within the morbidly obese cohort, 4 were major vascular complications, 1 was an acute lung injury, and 1 was a PV stenosis requiring angioplasty.
Per the standard protocol at our institution, all patients were admitted for observation overnight after AF ablation. Of the 35 patients who experienced a complication, 26 (74%) required ≥1 extra day of inpatient hospitalization related to the complication. The median hospitalization was 4 days (interquartile range 2 to 8, with an absolute range of 1 to 23). One death occurred 5 days after the procedure in a morbidly obese woman in whom a femoral vein dissection was discovered at the sheath exchange. The patient was conservatively treated and demonstrated no hematoma or hematocrit decrease during 48 hours of inpatient observation. A complete autopsy was unable to demonstrate the cause of death.
On univariate analysis, the patients who experienced a complication were more likely to be women (49% vs 27%, p = 0.01), require a longer procedure time (322 ± 94 vs 285 ± 76 minutes, p = 0.05), and have a greater left ventricular ejection fraction on magnetic resonance imaging (68 ± 10% vs 62 ± 12%, p = 0.01). The complete results of the univariate analysis are presented in Table 2 . The year of ablation and physician did not increase the rate of procedural complications ( Figure 4 ).
| Variable | Complication | p Value | |
|---|---|---|---|
| Yes (n = 35) | No (n = 477) | ||
| Age (yrs) | 61 ± 12 | 58 ± 11 | 0.11 |
| Men | 18 (51%) | 348 (73%) | 0.01 |
| Body mass index (kg/m 2 ) | 32 ± 8 | 31 ± 6 | 0.56 |
| Duration of atrial fibrillation (months) | 71 ± 67 | 85 ± 78 | 0.28 |
| Paroxysmal atrial fibrillation | 16 (46%) | 222 (47%) | 0.99 |
| Direct current cardioversions | 1 ± 2 | 2 ± 2 | 0.48 |
| Coronary artery disease | 4 (11%) | 92 (19%) | 0.36 |
| Obstructive sleep apnea | 7 (20%) | 122 (26%) | 0.46 |
| Clopidogrel | 1 (3%) | 12 (3%) | 0.90 |
| Aspirin | 11 (31%) | 220 (46%) | 0.09 |
| History of amiodarone use | 15 (43%) | 170 (36%) | 0.39 |
| Left atrial size ∗ (mm) | 39 ± 8 | 39 ± 8 | 0.77 |
| Left ventricular ejection fraction ∗ (%) | 68 ± 10 | 62 ± 12 | 0.01 |
| Redo ablation | 5 (14%) | 109 (23%) | 0.24 |
| Cavotricuspid isthmus ablation | 8 (23%) | 99 (21%) | 0.77 |
| Roof line | 15 (43%) | 176 (37%) | 0.48 |
| Mitral valve line | 7 (20%) | 100 (21%) | 0.89 |
| Complex fractionated atrial electrogram ablation | 3 (9%) | 51 (11%) | 0.69 |
| Cryoballoon | 0 (0%) | 7 (2%) | 0.47 |
| Procedure time (min) | 322 ± 94 | 285 ± 76 | 0.05 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
