Bariatric Surgery and Atrial Fibrillation: Does the End Justify Themeans?

Kadhim Kadhim; Melissa E. Middeldorp; Jeroen M. Hendriks; Dennis H. Lau; Prashanthan Sanders


Europace. 2019;21(10):1454-1456. 

The epidemiological link between obesity and atrial fibrillation (AF) is well established, with the upsurge in prevalence of both conditions reaching epidemic proportions.[1] Obesity is independently predictive of AF development and is associated with increased incidence of cardiovascular risk factors including diabetes, hypertension, vascular disease, and sleep-disordered breathing. This in turn negatively influences AF-related outcomes by elevating the risk of stroke and reducing the efficacy of AF treatments. Obesity results in progressive atrial remodelling characterized by increased pericardial fat, myocardial fat infiltration, and atrial fibrosis which results in areas of low voltage, fractionation, and conduction abnormalities.[2] A recent meta-analysis highlighted the impact of obesity on recurrence of AF after ablation with every 5 U increase in body mass index (BMI) being associated with a 13% greater risk of recurrence of AF following ablation,[1] highlighting the crucial role of targeting weight adjunctively with ablation.

Studies have demonstrated that weight loss, particularly as part of a comprehensive risk factor management program, can significantly improve catheter-based and pharmacological treatments for AF.[3] However, to date the studies where weight loss was achieved, resulting in subsequent AF reduction have done so in a cohort with a BMI of ~33 kg/m2.[3,4] Whether such weight loss can be achieved and result in improvement in ablation outcomes in those who are morbidly obese is not known. A recent study by Mohanty et al.[5] focused on a morbidly obese cohort with persistent AF undergoing ablation. Despite a significant weight reduction of 25.9 kg [interquartile range −19.1 to −56.7] in the intervention group, the authors demonstrated no change in AF burden, symptom control, or the need for ablation. Although this study was underpowered, these patients were not only morbidly obese but also had long-standing persistent AF, raising the possibility that there may be a threshold of arrhythmia burden or degree of obesity that may preclude any arrhythmic benefit from weight loss.

In this context, the study by Donnellan et al.[6] in this issue of the Journal is timely and highly relevant. The authors aimed to investigate the impact of bariatric surgery performed prior to catheter ablation on AF recurrence. Out of 239 patients with morbid obesity who underwent catheter ablation for AF, 51 patients had undergone bariatric surgery prior to the ablation procedure. The remaining 188 patients were enrolled in weight management clinics and formed the control arm of this observational study. The mean BMI of the entire study cohort at the time of AF ablation was 41.1 ± 5.3 kg/m2, being higher for patients undergoing bariatric surgery compared to controls. Atrial fibrillation ablation was performed using radiofrequency catheter ablation in almost all patients, with only three patients receiving cryoablation therapy. Over a mean follow-up period of 3 years, 80% of patients who underwent bariatric surgery had no recurrence of AF compared to 39% of the controls (P < 0.001). This was mirrored in a reduced need for repeat AF ablation procedures in the bariatric surgery group (12% vs. 41%, respectively).

In the Swedish Obese Subjects (SOS) study, 2000 AF-naïve individuals who underwent bariatric surgery were matched to a similarly sized cohort of controls. Although not a pre-specified endpoint, the SOS study observed a reduction in AF incidence by 29% in the bariatric surgery arm.[7] Recently, Lynch et al.[8] reproduced these observations in a large cohort of patients also using propensity-score matching, and noted a reduced incidence of AF in obese patients undergoing bariatric surgery compared to those managed medically (0.8% vs. 2.9%, P = 0.0001). While the benefits of weight loss in patients with AF are increasingly well established[3,4] (Figure 1), bariatric surgery has been postulated to have unfavourable effects. In a self-controlled case series study using population-based data from three US states, Shimada et al.[9] reported a paradoxical increase in AF-related hospitalizations or Emergency Department visits for at least 2 years after bariatric surgery.

Figure 1.

Proposed mechanisms of weight loss and favourable outcomes in atrial

Mechanistically, Donnellan et al. observed significant improvements in glycaemic and blood pressure control with bariatric surgery, potentially promoting atrial reverse remodelling and enhancing the likelihood of maintaining sinus rhythm, even prior to catheter ablation. Consequently, 22% of bariatric surgery patients were re-classified as paroxysmal from persistent AF while awaiting catheter ablation, a finding that echoes the recently published REVERSE-AF study.[4] Interestingly, in patients who reduced their BMI by >10%, 33% experienced AF recurrence, compared to 78% in those with increased BMI prior to ablation. With regards to glycaemic control, only 12% of patients whose HbA1C improved by at least 10% pre-ablation experienced AF recurrence, compared to 86% of those whose HbA1C increased. These parallel findings further support the notion of favourable atrial reverse remodelling with tighter control of modifiable AF risk factors.

The results of this single centre, retrospective observational study by Donnellan et al. should be interpreted with caution. The non-randomization to bariatric surgery could have introduced some selection bias. This is highlighted by the fact that propensity-score based analysis failed to find a match for 68% of the patients in the bariatric surgery group—an issue that a larger sample size could have potentially been able to resolve. Additionally, there were not enough patients to be able to establish if the type of bariatric surgery could impact on outcome. Weight loss achieved with bariatric surgery can be dramatic and Donnellan et al. observed about 23% reduction in BMI post-bariatric surgery and prior to AF ablation, but interestingly, there was no significant weight loss in patients enrolled in weight management clinics. While difficulty in achieving weight loss in this cohort may be a genuine reflection of a clinically challenging cohort to manage, weight loss while awaiting bariatric surgery has been reported to range from 8% to 12%, and is in fact a prerequisite for bariatric surgery in some centres.[10] This raises concerns regarding the intensity of interventions in those clinics, and whether other modifiable AF risk factors could have been more aggressively controlled, further contributing to the poor outcomes in the non-bariatric surgery group independent of the failure to lose weight. Further, Donnellan et al. did not report on the incidence of procedure-related complications resulting from either bariatric surgery or ablation. In a study by Winkle et al.,[11] they observed nearly doubled the complication rate in patients with a BMI ≥40 kg/m2 undergoing AF ablation compared to those with BMI <40 kg/m2 (3.5% vs. 6.7%; P = 0.023).

Donnellan et al. ought to be commended on their work. In addition to their study being the first to demonstrate an improvement in rhythm-related outcomes post-bariatric surgery and AF ablation, the mere fact that they were willing to offer ablation for as many morbidly obese patients is remarkable. Within the electrophysiology community, there remains an understandable trepidation regarding offering catheter ablation therapy to morbidly obese patients, given the unfavourable risk-to-gain ratio from offering such procedures. Catheter ablation for AF, despite technical and safety advances, remains an intervention with potentially serious complications that should not be offered lightly. Bariatric surgery is a highly effective treatment to achieve weight loss with resultant improvement in cardiometabolic profile but is not without risk.[10] The authors concluded that weight loss surgery should be offered to all morbidly obese patients with AF; however, we feel that more evidence is required to provide further compelling data for this invasive method of AF management, including longer-term follow-up data. We do agree that this study provides some promise for a group of patients with otherwise limited management options. But prospective, randomized controlled trials are required, both to establish the safety profile and to elucidate the true effect of bariatric surgery in terms of rhythm control for AF patients, prior to calls for radical shifts in management. In the meantime, a structured, patient-centred approach with goal-directed risk factor modification is a safe and effective tool that can be used, potentially in combination to bariatric surgery in the morbidly obese, to achieve sustainable weight loss and improve rhythm-related and quality of life outcomes for AF patients.