Atrial fibrillation (AF) affects about 3% of the population, making it the most common arrythmia encountered by healthcare professionals [1]. Among patients undergoing cardiac surgery, preoperative AF reaches an incidence of about 10% [2], with a higher incidence among patients presenting for mitral valve replacement (30–50%) and a lower incidence in patients undergoing coronary artery bypass graft (CABG) (5%) [3, 4].

AF is characterized by loss of atrial contraction and rapid, irregular ventricular contraction, leading to progressive cardiac dysfunction. Many patients experience significant symptoms such as palpitations, dyspnea, and weakness, leading to reduced quality of life (QoL) [5]. Beyond symptoms, AF is an independent risk factor all-cause mortality, stroke, heart failure and hospitalization compared to normal sinus rhythm [6, 7], and if present at the time of surgery, it negatively affects 30 day outcome and survival [8].

Surgical ablation (SA) is a well-established treatment option for patients with AF undergoing cardiac surgery. Current ESC/EACTS guidelines recommend concomitant AF ablation for all patients with a history of AF if the heart team believes added rate control may be beneficial [9, 10]. The aim of SA is to eliminate AF by using surgical lesions to block aberrant electrical conduction which therefore inhibits the generation and propagation of macro re-entrant circuits in the atria [11, 12].

Technical aspects include a wide range of energy including “cut and sew”, radio-frequency ablation (RFA) and cryoablation (CA). Similarly, current procedures include simple pulmonary vein isolation (PVI), left atrial (LA) ablation and the biatrial Cox-Maze IV procedure (CM-IV) [13, 14, 15].

On follow-up SA has repeatedly been shown to promote a return to NSR [16]. However, it is not clear if there is a global benefit of imposing NSR, and whether this translates to an improvement in patient important outcomes such as mortality, stroke and QoL.

The aim of this study is to review randomized controlled trials, propensity score matched studies, cohort studies and metanalysis to assess the benefits and risks associated with concomitant SA in patients undergoing cardiac surgery.

Literature presents many randomized control trials (RCTs) [5, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33], propensity score matched studies [34, 35] and cohort studies [3, 36, 37, 38, 39]. Tables 1,2 Ref. [16, 17, 18, 19, 20, 21, 22, 23, 34, 35, 36] show a summary of the RCTs and propensity matched studies included in the review. A large variety of lesion sets and energy sources were used to perform the ablation and there were also significant differences in the type of AF that was allowed according to specific inclusion criteria. Most studies included persistent AF (PerAF), long standing persistent AF (LsPerAF) or permanent AF.

A total of 5 metanalysis have been published in literature about concomitant treatment of AF during cardiac surgery (Table 3, [41, 42, 43, 44, 45]). The milestone Cochrane systematic review published by Huffman et al. [41] in 2016 summarized all the significant literature, concluding that concomitant AF surgery significantly halves the risk of recurrent arrhythmias (RR 2.04, 95% CI 1.63–2.55), while increasing the risk of permanent pacemaker implantation (RR 1.69, 95% CI 1.12–2.54), but with no differences in all-cause mortality (RR 1.14, 95% CI 0.81–1.59) and early mortality (RR 1.25, 95% CI 0.71–2.20). Starting from the late 1980’s when first RCTs were reported, this paper included 34 reports of 22 RCTs with about 1900 patients. Notably, authors reported a high risk of bias across at least one domain, and although they included only RCTs by study design, the risk of detection bias, small-study bias and reporting bias negatively affected the quality of the evidence, which were considered to be qualitatively poor. Most trials (77%) had less than 200 patients and were performed in a single center. Surgical technique also was very variable, with cut-and-sew technique, microwave ablation, cryoablation and radiofrequency ablation equally distributed among studies, thus increasing the inter-study variability and confounding the interpretation of results. At the publication of this review, 8 studies were still under recruitment and therefore results were not included but most studies were halted prematurely. Authors concluded that adequately powered RCTs and future secondary analysis would overcome those limitations, but this remains the most comprehensive metanalysis of RCTs on this topic.

Wang et al. in 2018 [42] confirmed the previous findings, but found a higher risk of mortality and morbidity for AF surgery. Cappabianca et al. in 2019 [43] summarized studies comparing biatrial vs left atrial ablation, concluding that BA is superior at costs of increased risks of bleeding and pacemaker implantation. However, those results were partially confirmed by Guo et al. in 2021 [44] as they found a higher risk of early mortality with BA ablation. Maesen et al. in 2021 [45] published the only metanalysis focused on quality of life, concluding that QoL improves after both stand-alone and concomitant arrhythmia surgery, as this improvement can be attributed to both the cardiac procedure itself and to the arrhythmia surgery, confirming that the main factor determining quality of life is rhythm at long term follow up.

In the present review, ablation appeared to be very effective at converting AF into NSR, however there was significant heterogeneity between studies in the tools used to assess freedom from AF. 18% of the trials used ECG alone at follow up to establish presence of AF, only using Holter monitoring if the patient claimed to be symptomatic. We were concerned that this may have caused over-estimation of freedom from AF. In one study, asymptomatic AF occurred 12 times more frequently than symptomatic AF when followed-up by Holter monitoring [49]. Also, the DISCERN-AF study showed that increased intensity of monitoring resulted in increased AF detection, therefore studies are less likely to achieve freedom from AF if a more intensive method, such as an implantable loop recorder (ILR), is used to assess rhythm [50]. As a result of this, we believe over estimation of freedom from AF to be a limitation of the present review and suggest that future trials contributing to the ablation argument should aim for a more homogenous, intensive monitoring technique to make results more reliable.

Although studies have claimed that SA can be performed concomitantly to cardiac surgery with no increased risk to the patient [39] others have shown an almost three fold higher risk of permanent pacemaker placement associated with SA [16]. This phenomenon is believed to be due to underlying sinus node dysfunction, only revealed when the aberrant macro-reentrant circuits associated with AF are terminated following successful ablation [51, 52]. Increased risk of pacemaker implantation has been observed especially with bi-atrial Cox Maze-IV lesions rather than LA or PVI [41]. This evidence was considered robust enough for the ESC/EATCS to downgrade their recommendation for concomitant ablation from I to IIA [10]. Although adequate statistical sub-group analysis has not yet been performed and is needed to confirm our findings, it initially appears that we have corroborated this view. It should however be added that this risk of pacemaker implantation is relatively low, and in our opinion, it should not prevent clinicians from performing ablation on patients if the heart team believe it to be in their best interest.

An important patient-centered outcome that is frequently overlooked in studies is health related quality of life (HRQoL), again due to the inference that return to NSR is directly linked with increased QoL [53]. In the present review, of the 26 studies included, only 3 reported on HRQoL using approved assessment tools such as the SF-36. As shown by the RACE study, AF can commonly cause significant symptoms and reduce QoL [47]. However, in the context of concomitant surgical ablation, the main cardiac disease requiring intervention might have a crucial role in limiting HRQoL preoperatively. Therefore, it can be difficult to ascertain if the improvement in the QoL is due to treatment of AF or the concomitant disease.

Chernyavskiy et al. [5] found that in the immediate post-operative period, HRQoL was significantly improved in all domains in both the control and the ablation groups. However, at 1 year of follow up, the increases in the control group had reduced and had ceased to be significant whereas in the ablation group they remained significant. Furthermore, at this time as recorded by ILR, only 44.1% of the control group where in stable NSR, whereas in the PVI group 80% were in NSR and in the CM-IV group the figure was 86%. Therefore, it is likely that the reduction in QoL in the control group was due to a persistence or relapse into AF in the follow up period which was not observed in the ablation groups.

Our opinion is therefore in agreement with Chernyavskiy et al. [5], and we suggest that improvement in QoL following cardiac surgery with concomitant surgical ablation is attributed to not only resolution of the underlying disease but also return to NSR. Berkowitsch suggested that this may be due to a placebo effect whereby fewer symptoms are reported because the patients believe their treatment has been successful [54]. Also maybe being told that their heart-rate is beating in a “normal” rhythm may increase their psychological well-being by reducing their anxiety [55]. We therefore believe that treatment should be comprehensive and focused on not only the elimination of angina symptoms but also the clinical manifestations of AF. More evidence is needed however in the form of large RCTs investigating HRQoL over long periods to support this argument further.

Surgical ablation is the most effective procedure to treat AF and is a unique opportunity for most patients undergoing cardiac surgery affected by AF to return to sinus rhythm with no added mortality risk and a potential improvement in quality of life. There is however a low risk of potential pacemaker implantation and renal failure which must be weighed against the benefits before selection of treatment. It is also not clear how long-term outcomes are affected due to underpowered RCTs and this review would urge these studies to report on their long-term outcomes in the future to add to this argument.

There is significant heterogeneity between the studies in terms of their inclusion and exclusion criteria. This may question the legitimacy of looking at the results as a collective, however we believe that the comparison is valid since it reflects the heterogeneity in operating procedure observed between surgeons and institutions. Adequate evaluation of SA in specific surgical settings, i.e., patients undergoing isolated mitral valve repair or combined mitral and tricuspid valve repair, is warranted to define the optimal treatment of those patients. Also, homogeneous or comparable technical details (lesion set and energy) are vital to evaluate outcomes. This will add considerable weight to the argument of whether the benefits of surgical ablation extend further than simple maintenance of NSR.