*Review* **Ablation Modalities for Therapeutic Intervention in Arrhythmia-Related Cardiovascular Disease: Focus on Electroporation**

**Shauna McBride <sup>1</sup> , Sahar Avazzadeh <sup>1</sup> , Antony M. Wheatley <sup>1</sup> , Barry O'Brien <sup>2</sup> , Ken Coffey <sup>2</sup> , Adnan Elahi 3,4 , Martin O'Halloran <sup>3</sup> and Leo R. Quinlan 1,5,\***


**Abstract:** Targeted cellular ablation is being increasingly used in the treatment of arrhythmias and structural heart disease. Catheter-based ablation for atrial fibrillation (AF) is considered a safe and effective approach for patients who are medication refractory. Electroporation (EPo) employs electrical energy to disrupt cell membranes which has a minimally thermal effect. The nanopores that arise from EPo can be temporary or permanent. Reversible electroporation is transitory in nature and cell viability is maintained, whereas irreversible electroporation causes permanent pore formation, leading to loss of cellular homeostasis and cell death. Several studies report that EPo displays a degree of specificity in terms of the lethal threshold required to induce cell death in different tissues. However, significantly more research is required to scope the profile of EPo thresholds for specific cell types within complex tissues. Irreversible electroporation (IRE) as an ablative approach appears to overcome the significant negative effects associated with thermal based techniques, particularly collateral damage to surrounding structures. With further fine-tuning of parameters and longer and larger clinical trials, EPo may lead the way of adapting a safer and efficient ablation modality for the treatment of persistent AF.

**Keywords:** electroporation; pulsed field ablation; cardiac; heart; arrhythmia; atrial fibrillation

#### **1. Introduction**

The Centers for Disease Control and Prevention in the USA reports that 1 in every 4 deaths in the United States is related to general cardiovascular disease, with an estimated 12.1 million people predicted to develop arrhythmias such as atrial fibrillation (AF) by 2030 [1]. In recent years there has been a rapid growth in the technology base and clinical appetite for targeted ablative procedures for arrhythmias, with some reports showing procedures to be effective, with quick procedural timelines, minimal associated risks and rapid recovery times [2,3]. Catheter-based ablation for AF is considered a safe and effective approach for patients who are refractory to medication. The cornerstone of catheter-based approaches to date is pulmonary vein isolation (PVI) but, increasingly, additional sites beyond the pulmonary veins are now being targeted [4]. In this review we report on the available data exploring energy-based ablative technologies, highlight the differing

**Citation:** McBride, S.; Avazzadeh, S.; Wheatley, A.M.; O'Brien, B.; Coffey, K.; Elahi, A.; O'Halloran, M.; Quinlan, L.R. Ablation Modalities for Therapeutic Intervention in Arrhythmia-Related Cardiovascular Disease: Focus on Electroporation. *J. Clin. Med.* **2021**, *10*, 2657. https:// doi.org/10.3390/jcm10122657

Academic Editor: Charles Guenancia

Received: 25 May 2021 Accepted: 14 June 2021 Published: 16 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

modalities that have been developed with a particular focus on anti-arrhythmic therapies. This review also considers the factors involved in achieving successful ablation of cardiac tissue and the evidence from in vitro and in vivo preclinical work which has informed clinical studies using EPo approaches.

#### **2. Current Ablation Approaches for Treating Arrhythmia**

Several relatively simple non-invasive ablative procedures have been developed to date, such as alcohol septal ablation, which involves the injection of ethanol into the septal coronary artery to target portions of the septal wall [5]. This minimally invasive ablation method has been extensively employed as a treatment for structural related heart defects such as hypertrophic cardiomyopathy, targeting the attenuation of outflow tract obstruction [2,6]. Alcohol septal ablation is often applied when previous lower intensity therapies have failed [5]. Stereotactic radioablation is another non-invasive modality under development. While not currently used in clinical practice to the best of our knowledge, a number of animal-based feasibility studies with stereotactic radioablation have been performed and reviewed elsewhere [7,8].

Typically, more invasive ablation techniques require entry into the body cavity to access targeted areas of the myocardium (Figure 1). These techniques up to more recently generally involved the use of thermal energy and either induced hyper- or hypo-thermal injury at the target site [9]. Hyperthermal approaches are most commonly based on the application of radiofrequency (RF) or laser energy. Hypothermal approaches, termed cryoablation, are commonly achieved by passing cooled, thermally conductive, fluids through hollow probes at the target site.

approaches is often made in relation to the target area and patient's disease substrate **Figure 1.** Access to the heart for invasive ablation purposes. This can be achieved via an internal endocardial approach (**A**) via the femoral vasculature (Table 2). Ablation catheter access can also be gained from an external epicardial (B) method. The extremities of the heart are reached by this technique. Access via an epicardial approach can be achieved through ports in the intercostal spaces (**1B**), a sub-xiphoid puncture (**2B**) or via open heart surgery (**3B**). The choice made between the two approaches is often made in relation to the target area and patient's disease substrate [10].

#### *2.1. Hyperthermal Techniques*

Hyperthermal approaches can use various energy sources including the use of ultrasound, lasers, radiofrequency technology applied via electrode catheter, or hot balloon ablation systems [11,12]. Focused ultrasound causes the destruction of a target area due to a thermal heating effect, while remaining minimally invasive [11,13,14]. Several studies have highlighted the challenges associated with using high-intensity ultrasound for cardiac ablation [15–17]. Due to the incidence of oesophageal fistula and subsequent fatalities, ultrasound as a modality needs considerably more development involving lower-intensity energy and better targeting if it is to be more widely adopted [15,18]. Similar techniques

are employed using lasers and have often been applied to target tumours in a variety of locations in the body [11]. Laser-based approaches employ slow heating, with low-power lasers delivered through optical fibres to induce protein denaturation. However, areas targeted by hyperthermal techniques are often difficult to control, with blood circulation proving problematic, acting to dissipate the temperature field. Some of the initial negative associations with laser or ultrasound methods are being overcome as devices become smaller and more user friendly [11]. More recent developments in commercial laser ablation systems and low-intensity ultrasound systems are beginning to compete with RF technologies in PVI applications [19].

By far the most common hyperthermal approach is based on RF technology. This modality has been in use since the 1990s and has surpassed all other energy delivery methods in popularity for use in cardiac ablation [20]. Ablation with RF relies on thermal energy from high frequency sinusoidal waves (500–750 kHz), to induce controlled damage or region-specific necrosis of heart tissue [11,20–25]. Temperatures of ≥50 ◦C induce tissue necrosis; however, temperatures approaching 100 ◦C can cause a coagulum of denatured proteins and plasma to form on the catheter tip, impeding current delivery [20] (Figure 2). RF-induced lesions typically have well-defined borders and their shape can adopt a monopolar egg shape or a bipolar round-brick shape [25,26]. RF lesion shape is dependent on a number of factors including catheter tip diameter, inter-tip spacing, tissue contact, temperature and duration of energy pulse delivered [25–27]. RF is a clinically significant technique as an ablation treatment for atrial fibrillation (AF), with success rates generally ranging anywhere between 45% and 80% [4,28,29]. – – . Temperatures of ≥50 –

**Figure 2.** Effects of thermal and non-thermal energies. Diagram highlights the differing outcomes exhibited post-ablation between radiofrequency (**A**), cryoablation (**B**) and electroporation (EPo) (**C**) modalities. Radiofrequency and cryoablation can induce necrosis upon application which is followed by scarring with the intention to break arrhythmic circuits. Meanwhile, EPo increases membrane permeability which can lead to apoptotic or necrotic cell death, ultimately resulting in scarring.

#### *2.2. Hypothermal or Cryoablation Techniques*

–

Cryoablation has proven to be a clinically effective and a safe ablation method for use on cardiac tissue and has been studied since the early years of interventional cardiology [30]. Cryoablation uses hypothermal energy to induce ablation by freezing (≤−40 ◦C) [21,31] (Figure 2). Lesion shape with cryoablation is sharper, more homogeneous and less thrombogenic than lesions resulting with RF [32]. Cryoablation has been employed as a treatment particularly for arrhythmias such as AF. A study by Bárta et al. yielded success rates of approximately 90% immediately post-ablation and 48.5% of patients were free from AF at 12-month follow-up [30,33,34]. A comparison study by Kim et al. showed atrial

contractility recovery rates of 32.2% and 48.8% following RF or cryoablation treatment at 12-month follow-up, respectively [35]. Similar data were reported in the FIRE and ICE clinical trial with comparable numbers of patients requiring repeat ablation procedures to sufficiently isolate pulmonary veins (PVs) and terminate arrhythmias, highlighting the success, efficacy and challenges associated with both procedures [36].

#### *2.3. Challenges with Current Ablative Approaches*

While there have been many positive reports particularly for RF ablation and cryoablation, their efficacies rely on the precise positioning of catheters, adequate catheter-to-tissue contact and the energy level applied to the target area [37]. Collateral damage of surrounding areas is common during the application of thermal energy, including cardiac tamponade, thromboembolism, PV stenosis, phrenic nerve and oesophageal injury or fistula, and mitral valve trauma [21,38]. Overheating of the ablation site with RF energy has resulted in 'steam pop' and can cause myocardial perforation or cardiac tamponade [39–41]. Similarly, blood flow can cause a heat sink effect during RF and cryoablation procedures, preventing uniform tissue heating or lesion formation [42]. RF has also been shown to cause coagulation within coronary vessels, induce initial hyperplasia and instigate the shrinkage of collagen fibres within the coronary arterial walls [43].

Over the last decade, the interest and demand for a more controllable and safer alternative ablation technique has been growing. The advances in electroporation (EPo) and its refinement as pulsed electric field (PEF) technology, or pulsed-field ablation (PFA), has expanded to such a degree that it can now be considered a cutting-edge, nominally thermal ablation approach. The capacity to customise parameters for further enhanced application in humans may be a turning point in the treatment of specific targeted CVDs, improving procedure management and outcomes for both clinician and patient.

#### **3. Electroporation as an Ablative Approach**

Catheter-based electroporation (EPo) using monophasic pulses was first employed with cardiac tissue in the 1980s but it was found to be associated with negative side effects such as the induction of an electrically isolating "vapor globe" resulting in a spark (arcing), followed by an explosion and damaging pressure waves [44–46]. Serious complications such as barotrauma and a pro-arrhythmic effect saw voltage-based energy systems superseded by RF ablation [46,47]. However, Ahsan et al. demonstrated that the cautious use of electroporation at lower energies could successfully avoid arcing and produce sufficient therapeutic lesions [48]. Modern voltage-based systems typically employ pulsed electric fields (PEFs) [49,50]. Ablation based on EPo is growing in popularity as an alternative to thermal ablation and causes a biophysical phenomenon to arise following the application of PEF [2,51]. These electric fields induce irreparable pore formation in cell membranes [3]. As a result, so-called PFA is considered minimally thermal and creates more predictable and controllable lesions, with minimal interaction with blood flow.

Since 2005, both irreversible (IRE) and reversible (RE) EPo has received considerable attention as a method of disrupting cell membranes for drug delivery or inducing selective cell death, respectively [11]. Both IRE and RE have the potential to be tissue-specific in terms of lethal or effective thresholds, with extracellular and endothelial structures commonly remaining intact following exposure to electric fields [52,53]. The permanent opening of nanopores in cell membranes activates intracellular molecular pathways, increases ionic and molecular transport, resulting in an overall disruption of the cell membrane and intracellular homeostasis [11,21,54]. Exposure to sufficiently large field strength results in IRE, and permanent damage and cell death ensues due to localized rearrangement within membrane structures, while supporting structures remain unscathed [9,55–58]. RE, in contrast, only transiently opens membrane pores, maintaining cell viability, and is commonly employed in the targeted delivery of drugs and nucleotides [11].

The extent and targeting of ablation with IRE can be controlled at least to some degree by changing parameters such as pulse amplitude, frequency, duration of the application

and pulse number [2,59]. The lethal thresholds for many cell types have been reported based on these parameters; however, many contradictory data exist as it is still an active area of ongoing research. On the face of it, short exposures and microsecond EPo impulses can be used for biomedical applications aimed at drug delivery and gene transfer, while more prolonged impulses are related to cellular injury and ablation by IRE [55,60,61]. The shape of the applied pulse is an under-explored, and in many cases a poorly documented, parameter that has not received the same degree of experimental testing as amplitude, frequency and others (Tables 1 and 2). Using a lung cell line, Kotnik et al. demonstrated that of the parameters used to describe pulse shape, the major factor determining electropermeabilization was the amount of time the pulse amplitude exceeded a certain threshold value [62]. They suggest that any differences observed between various pulse shapes may in fact be reflecting the difference in time the pulse is above the critical threshold for that cell type. Meanwhile, Stankevic et al. reported that it is the pulse shape and total energy input that contribute to the efficiency of IRE [63]. Sano et al. (2017) reported that asymmetric waveforms have significantly lower IRE thresholds compared to equivalent symmetrical waveforms, at least for neuroblastoma cells in vitro [64]. Both symmetrical and asymmetrical biphasic pulses have proven effective in IRE cardiac ablation procedures in both animals and a small number of pilot human trials [45,65–68]. Overall, asymmetric waveforms appear to produce more effective pore opening than symmetric pulses, possibly due to the different amplitudes of their phases. We recommend that all elements of pulse profile need to be reported, according to a set of recommended guidelines, as the extent that pulse shape contributes towards the safety and efficacy for AF treatment with IRE is unclear [69]. Overall, this is an area that requires substantial and more fundamental research before it can become part of standard clinical application [67].

More recently, the field has focused on pulse timing issues [70]. With nanosecond-PEFs in particular, this has been shown to improve the controllability of pore size. Short duration nsPEFs have been shown to minimise the electrophoretic effects associated with cell membrane transport [70,71]. When compared with longer pulse durations, shorter durations are reported to limit solute movement, overall reducing the osmotic imbalance and improving cell targeting with PEF exposure. nsPEF stimuli are too short to induce capacitive charging and instead aim to influence displacement currents over conduction currents [70]. Elementally, every cell behaves independently, deeming intercellular electric connections ineffective on membrane charging [72]. However, the mechanism by which such short stimuli can influence pore opening is still not fully understood and is the subject of ongoing research [70].

#### *3.1. Pre-Clinical Evaluation of IRE, towards Optimization of Parameters for Clinical Use*

Ex vivo studies were a milestone in the adaptation of IRE for in vivo applications as early work by Krassowska et al. emphasized the formation of pores in tissue exposed to EPo in a 2D model of cardiac tissue [73]. Selective pore formation can prevent excessively high transmembrane potentials, limiting damage in surrounding tissues. As work progressed, studies investigated the therapeutic thresholds and biophysical effects of EPo at a cellular level [3]. This was achieved by altering some therapeutic variables (pulse duration, pulse frequency, amplitude) and comparing the induction of injury on tissue through lactate dehydrogenase activity and the integrity of cell membranes.

Experiments involving the murine atrial cardiac cell line HL-1, cultured as adherent monolayers, showed that the damage was proportional to the number of IRE pulses and field strength applied [3] (Table 2). When compared to previous work from the same group looking at the human prostate cancer line LNCaP, data suggested that cardiac cells were more suspectable to IRE at higher field strengths greater than 1000 V/cm. Kaminska et al. showed that pulse intensities above 375 V/cm were destructive in the immature rat H9C2 myoblast cell line [74] (Table 2). A scan of the potential range of field strengths that might induce cell death is required and would be enhanced by the addition of threshold data on neuronal, cardiomyocyte and fat tissue found in the heart. Very recently, Hunter et al. showed that cardiac cells are more susceptible to electroporation damage than cortical

neurons and oesophageal smooth muscle cells [75]. However, there are very few reports of this nature examining the different IRE thresholds of cardiac cells relative to other appropriate cardiac–neuronal model systems.

In animal studies, the application of IRE to cardiac tissue for the treatment of arrhythmias was found sufficient to block aberrant conductive pathways and reduce conduction and propagation of disruptive electrical signals [76]. In a study by Zager et al., it was shown that longer pulse durations (100 µs versus 70 µs) and increased pulse number (20 versus 10) were associated with a larger volume of damage in the ventricular myocardium in a rat model [2]. In a porcine model, the controlled delivery of electrical pulses both monophasic and biphasic, over a few microseconds or nanoseconds, has been shown to create tissue injury while avoiding negative effects [45].

In terms of pulse polarity, studies have found that biphasic pulses show better efficacy than monophasic stimuli in penetrating epicardial fat and overcomes the impedance by fatty cells during ablation of cardiac tissue [43,60]. Similarly, while both monophasic and biphasic pulses have proven efficient at producing feasible ablation outcomes, biphasic waveforms have been shown to create more durable lesions than monophasic applications [77]. This may be owing to biphasic pulses significantly altering electric field bias, reducing ion charging and prolonged post-ablation depolarizations [70]. Due to a "cancellation effect", higher amplitudes are often required to achieve ablation when using biphasic shocks at the nanosecond level [60]. Ablation success is seen to be influenced by the time between successive pulses. Nanosecond pulses can induce a uniform activation of the myocardium by forming a consistent electric field distribution [70]. This has been found to reduce the risk of new wave-fronts arising that could reinitiate arrhythmia and fibrillation [70]. Studies have also demonstrated that patients receiving monophasic pulses commonly require general anaesthesia and neuromuscular paralytics during procedures [78]. In comparison, patients that received biphasic pulses or high frequency energy were able to have the procedure under conscious sedation due to minimal resulting skeletal muscle activation [60,67,78] (Figure 3). It has been proposed that direct current (DC) monophasic energy be replaced by short alternating current (AC) biphasic energy to target larger areas, as it appears to reduce capture of nearby excitable tissues, thus reducing muscle spasms and acute pain during ablation [79].

#### *3.2. Controlled Lesion Formation with IRE In Vivo*

Human studies using comparably greater pulse durations and frequencies, ranging from microseconds to milliseconds (Table 2), show that the ablative effect and lesion area depends on the electric resistivity of the tissue, presence of cell membranes and the applied electric field [47]. Short electric field pulses cause rapid lesion formation, which is favorable to procedural work-flow [68]. However, the rapid nature of the delivery of IRE provides little, if any, opportunity for clinicians to change position of catheters or the profile of the energy delivered during the active phase of energy delivery.

While clinical outcome reports for IRE are limited, they are growing, and success has been noted in early clinical trials. Reddy et al. and Loh et al. have in a series of papers shown the safety and efficacy of IRE in the clinic [68,80,81]. Firstly, the authors highlighted the safety of the IRE procedure by successful acute pulmonary vein isolation (PVI) in 100% of patients [68,78,81]. This was a turning point in IRE ablation as it highlighted the potential of IRE and its capability to replace current thermal techniques in PVI procedures. Freedom from AF was later recorded in 94.4 ± 3.2% of patients by Reddy et al., in a recent trial using a either a combined RF/IRE (pulsed-field) approach or IRE as a standalone ablation procedure [80]. Similarly, 100% of PVs in patients with symptomatic paroxysmal or persistent AF were successfully isolated by Loh et al., using IRE alone [81].

Compared to thermal approaches, IRE appears to be less reliant on specific anatomical catheter positioning or catheter–tissue contact to produce adequate lesions, however this has not been examined systemically and requires more evidential data [51]. Successful ablations have been demonstrated even when delivering less precise, more widespread energy,

suggesting that tissue vulnerability and lesion formation depends on tissue susceptibility and tissue type, facilitating cell-specific targeting with controlled parameter selection [51]. Studies have shown that IRE lesions occur locally in regions directly associated with electrodes [82,83]. Regions surrounding the electrodes are exposed to lower electric fields which induce RE, thus cells in this region recover and revert to normal function. Whether IRE lesions are transmural or not varies with the increasing thickness of the myocardium, requiring lesions to be wider to ensure adequate penetration [84]. IRE-induced lesions of the myocardium can be observed at a cellular level and are differentiated from unaffected tissue by a sharp border, similar to those induced by RF and cryoablation [9]. Appropriate transmural lesions are necessary to ensure the isolation of targeted regions to prevent disease relapse, thus avoiding the need for repeated procedures [9].

**Figure 3.** Structure of monophasic and biphasic pulses and their effect on muscle during ablation. (**A**) Monophasic pulses have been shown to induce excessive skeletal muscle spasm in patients thus requiring the use of general anaesthesia and paralytics. (**B**) Muscle activation induced by biphasic is minimal and therefore requires sedation.

– It has been suggested that IRE parameters can be fine-tuned to achieve different lesion profiles [78]. The data to date, suggest that lesion geometry is significantly influenced by a number of pulse parameters and electrode spacing, with lesion size and depth corresponding mainly to the magnitude of field strength delivered [40,72,85] (Table 2). Early studies in which IRE was applied to porcine tissue noted that no charring or tissue disruption was visible upon gross inspection, and a clear demarcation line was evident around electroporated regions at the cellular level [9]. Further histological inspection demonstrated that while avoiding a significant local temperature change, successful electrical isolation was evident in the atria and was accompanied by transmural destruction. In addition, these studies demonstrated that lower field strengths can create sufficient lesions on PVs, while higher energies result in tissue shrinkage of the ventricle [53,86]. In porcine models, IRE-induced lesions were found to be similar to those formed with RF energy, however microscopic analysis revealed that IRE lesions have reduced epicardial fat-associated inflammation and fewer intralesional sequestered cardiomyocytes [45]. Furthermore, fibrotic regions formed during remodeling at the site of energy delivery were more homogenous in those areas exposed to IRE than to RF energy [45]. As expected, both modalities were linked to neointimal thickening on the undisrupted endocardium, fibrosis of intralesional nerves, and absence of endocardial thrombus formation. Similar histological results were recorded in canine studies upon the targeting of epicardial ganglia with IRE, highlighting preservation of cardiomyocyte architecture, minimal inflammatory response and fibrosis [76]. Studies have also shown that blood has a higher conductivity than tissues, which may affect IRE lesion depth of procedures done using an endocardial approach; however, unlike thermal

techniques, this interaction with blood does not cause coagulation [87]. While substantial information has been collected from animal models, evidence of lesion geometry in humans cannot be assessed to such degrees. Instead, lesions are observed from a gross, clinical perspective and their electrical conduction is monitored during procedures. The inspection of the effect of IRE on vasculature is commonly achieved via imaging techniques.

#### *3.3. Advantages and Disadvantages of EP*

Preclinical and clinical data overall support the efficacy and safety of IRE with its capacity to limit injury to surrounding structures [51,68,73,74]. Due to the importance of protecting the coronary vasculature, several studies have investigated the short-term (3 weeks) and longer-term (3 months) effects of IRE on blood vessels (Table 2). Du Pre et al. investigated the effect of IRE when applied directly to the coronary vasculature in a porcine model. They observed that the coronary vessels remained free from clinically relevant damage, regardless of the lesion proximity to the vessel [43]. This further supports the targeting benefits of IRE which is not influenced by arterial blood flow, even when applied directly to vessels [43]. Long-term follow-up studies of porcine models by Neven et al. also highlighted the safety of IRE when targeted at the coronary vasculature [44]. They demonstrated that even deep lesions had no effect on luminal coronary artery diameter in their long-term study, proving IRE to be a safe ablation strategy for use on or in close approximation to the coronary vasculature.

The minimal damage IRE induces to vasculature makes it an attractive ablation modality for further development [77]. The major drawback with current thermal techniques is the induction of vessel stenosis, particularly of the PVs. The mechanism behind PV stenosis is believed to be a combination of intimal hyperplasia, medial thickening and smooth muscle activation, which is often followed by scar retraction and vein narrowing [88]. Comparative studies have drawn interesting comparisons between the use of RF energy and IRE for ablation of PVs. Early investigations were conducted on porcine models showing the effect of the given energies on inducing PV stenosis [89] (Table 2). A study revealed the significant damage caused by RF on PV tissue with observations of necrotic myocardium, large amounts of scar tissue surrounding the myocardial sleeve and intimal and elastic lamina proliferation. In contrast, only minor intimal proliferation was noted on tissue targeted with IRE. An initial decrease in PV diameter was noted in IRE experiments, however later studies revealed an overall increase in diameter [89]. It was evident that no PV stenosis arose due to IRE-exposed subjects at 3-month follow-up, while those who underwent RF ablation exhibited stenosis immediately that persisted for the follow-up survival period [89]. In recent years, the effects of IRE and RF were investigated in humans and research showed that the incidence of stenosis and narrowing of PV diameter following PFA was virtually eliminated (0%), while patients who received RF energy saw a 12.0% reduction in diameter and 32.5% incidence of stenosis at 3-month follow-up [90].

Some early preclinical work on porcine PVs by Wittkampf et al. reported that lower field strengths can create feasible and safe lesions on PV ostia, without evidence of collateral damage to surrounding structures [53,91] (Table 2). Preservation of nerve tissue was observed with no significant damage to the phrenic nerve post-IRE procedure [92]. With the oesophagus lying near the heart, the avoidance of trauma here is also a major concern of electrophysiologists. Similar findings were reported in a canine model, with subjects showing no signs of oesophageal injury [93]. In a porcine model, Neven et al. reported no disruption to oesophagus architecture, even with purposeful targeting of the adventitia [94]. This highlights the feasibility of IRE applications for cardiac tissue. In the first human trials IMPULSE and PEFCAT, incorporating the FARAPULSE endocardial ablation system, combined analysis by Reddy et al. showed the tissue-specific nature of IRE [78]. There were no reports of oesophageal or phrenic nerve injury in patients receiving the therapy, demonstrating that IRE possesses a major safety advantage over thermal techniques as an ablation modality for cardiac tissue [78].


**Table 1.** Comparison of clinical IRE studies on cardiac tissue.

Another major difference between IRE and thermal catheter-based procedures is the time taken to perform the procedure. From a practical perspective, IRE procedures require significantly less time, energy and number of applications of energy in comparison to those of RF or cryoablation [3,59,72,78,93]. In the FIRE and ICE trial, studies by Reddy et al., on the isolation of PVs, highlighted a notable difference in mean total procedure time for IRE (92.2 min) compared to RF and cryoablation procedure times, which required 141 and 124 min, respectively [78]. Similarly, left atrial catheter-dwell time was much lower in IRE procedures (34 min) in contrast to RF and cryoablation (109 and 92 min, respectively) [78]. Similar results have been recorded in AF treatments with cryoablation yielding significantly shorter procedure times (ranging between 73.5 ± 16 min to 192.9 ± 44 min), in comparison to RF energy techniques (from 118.5 ± 15 min to 283.7 ± 78.0 min), with IRE procedures yielding even less time overall (from 22.6 ± 8.3 min to 92.2 ± 27.4 min) [78,80,95]. Shorter procedure times also incorporate less fluoroscopy duration, reducing a patient's radiation exposure [96]. While procedure time is not a crucial factor in ablation, shorter duration can enhance productivity by reducing healthcare costs overall. Therefore, IRE has a clear advantage over current procedures for efficacy, safety and reduced procedure times.

While IRE overcomes many complications associated with thermal ablation techniques, it does pose some similar risks such as thrombosis, haemorrhage and infection, however these are common to all procedures employing similar access techniques [50]. Specific to IRE there is an associated risk of electrolysis when untuned current is passed through body fluids with dissolved electrolytes, instigating gas formation [97,98]. One study reported that different current polarity may decrease gas bubble formation as a side-effect of IRE, highlighting that a reduced number of gas bubbles are released when using anodal IRE, compared to RF or cathodal IRE [98]. Gaseous microemboli could result in myocardial damage and in some instances with symptomatic cerebral ischemic events due to the obstruction of capillaries [98]. However, the risk of microemboli appears very low with preclinical canine reports by Neven et al. detailing that no treatment-related cerebral events occurred due to gas formation during IRE procedures [99]. In addition, MRI imaging and histopathology confirmed the absence of cerebral emboli, supporting the safety of this procedure [99].

The most immediate effects of electric field application to the myocardium are electrophysiological, leading to possible changes in ECG, such as in the ST segment and T-wave, or an overall decrease in resting cell membrane potential [50]. In some instances, the use of PEFs in non-cardiac applications has also been involved in the evolution of lethal and non-lethal cardiac arrhythmias in animal studies [51]. To minimise the risk of induced arrhythmia, it is presumed that ECG synchronisation integrated with pulse delivery is critical to ensure the energy is applied only during the absolute refractory period of the heart to avoid a critical increase in cell permeability [66,82,100]. A recent clinical trial by Loh et al. incorporated ECG synchronisation for the delivery of monophasic IRE pulses [81]. During this study, no peri-procedural complications were recorded. Likewise, in a study by Reddy et al. no adverse effects were reported when using biphasic pulses without synchronization to depolarization of the atria or ventricles [80]. Thus, the absolute requirement for synchronization is unclear. The use of IRE requires meticulous monitoring of blood pressure and electrolytes, as instances of induced mild hypertension and epileptic-like seizures have been reported with the use of high voltages and general anaesthesia during IRE procedures [81,90,101]. Cases of such intraoperative complications could jeopardize patient safety, therefore a clear understanding and rapid, appropriate management by the clinician is paramount. Nevertheless, it has been concluded that an application of irreversible PEFs directly to cardiac tissue both endocardially and epicardially in preclinical studies is safe when timed with the R-wave during sinus rhythm, and in early clinical studies regardless of timing [51,66,80–82,90,100,101]. However, as PEF is a novel technique, safety boundaries and significant safety data remain scarce for human application and further investigation is required.


#### **Table 2.**Comparison of preclinical IRE studies on cardiac tissue.


**Table2.***Cont.*


**Table2.***Cont.*


**Table2.***Cont.*


45

#### **4. Conclusions**

IRE has seen its stock rise substantially as a therapeutic intervention in recent decades and there has been much interest in its safety and feasibility for use on cardiac tissue. While significant advances have been made based on animal studies, particularly involving porcine and canine models, and preliminary parameters have been developed for use in humans (Table 1), much optimisation remains to be achieved. Further testing and finetuning are required to adapt and potentially individualise these parameters for specific patients or patient groups, while ensuring precise delivery of energy to achieve efficient EP ablation. There is significant room for the development of more complex representative in vitro model systems that incorporate both functional and histological outcomes, that are multi-cellular and more easily translatable. This will facilitate rapid development of pulse parameters and potentially catheter design by looking at the catheter not just to deliver energy, but to also provide feedback on target site and success of the ablation.

Similarly, while there are substantial preclinical data for IRE from animal models, the number of clinical trials is limited. Studies completed to date include small cohorts of approximately eighty patients with varying follow-up times of 3, 4 and 12 months [78,80,90]. Therefore, not only larger, multicentre trials are required to analyse the effects of IRE but also long-term evaluation of the permanence of the ablation.

Lesions are difficult to investigate in human studies, thus, most information is to be acquired regarding the true depth and volume of lesions is collected from animal studies. Follow-up times of preclinical trials generally exceed no longer than 3 or 4 months (Table 2). Similarly, long-term studies would challenge the durability of lesions in humans and examine any relapse to the electrical or structural induced CVD originally treated by IRE. Another limitation to current IRE trials is the lack of consistency between experiments. Some studies are limited to one energy magnitude, while others either use smaller or greater magnitudes on different sized animals (Table 2). While there are few published clinical trials related to the use of IRE on cardiac tissue, preclinical studies provide a promising baseline representation of its use. IRE bypasses many of the complications and drawbacks of the more commonly used thermal ablation modalities. With further improvements and refinement of parameter specifics, IRE may prove to be the gold standard for ablative CVD therapy.

**Author Contributions:** Conceptualization, S.M., S.A. and L.R.Q. Writing—Original draft preparation, S.M. and S.A.; Writing—review and editing, S.A., A.M.W., B.O., K.C., A.E., M.O. and L.R.Q. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Enterprise Ireland Disruptive technology (DTIF) grant number [DT20180123].

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Riyaz A. Kaba 1,2,\*, Aziz Momin 1,2 and John Camm <sup>1</sup>**


**\*** Correspondence: rkaba@sgul.ac.uk; Tel.: +44-208-725-4571

**Abstract:** Atrial fibrillation (AF) is a global disease with rapidly rising incidence and prevalence. It is associated with a higher risk of stroke, dementia, cognitive decline, sudden and cardiovascular death, heart failure and impairment in quality of life. The disease is a major burden on the healthcare system. Paroxysmal AF is typically managed with medications or endocardial catheter ablation to good effect. However, a large proportion of patients with AF have persistent or long-standing persistent AF, which are more complex forms of the condition and thus more difficult to treat. This is in part due to the progressive electro-anatomical changes that occur with AF persistence and the spread of arrhythmogenic triggers and substrates outside of the pulmonary veins. The posterior wall of the left atrium is a common site for these changes and has become a target of ablation strategies to treat these more resistant forms of AF. In this review, we discuss the role of the posterior left atrial wall in persistent and long-standing persistent AF, the limitations of current endocardial-focused treatment strategies, and future perspectives on hybrid epicardial–endocardial approaches to posterior wall isolation or ablation.

**Keywords:** persistent atrial fibrillation; posterior wall; hybrid ablation; convergent ablation

#### **1. Introduction**

Atrial fibrillation (AF) is the most commonly diagnosed sustained cardiac dysrhythmia and is characterised by rapid and irregular activation of the atria. It is associated with an increased risk of ischemic stroke, heart failure and mortality and can have a substantial impact on quality of life. Atrial fibrillation can be paroxysmal, lasting 7 days or less with or without intervention, or be continuous beyond 7 days (persistent, PersAF) or beyond 12 months (long-standing persistent, LSPersAF) [1]. Permanent AF is the term used for long-standing persistent AF when any attempt to restore sinus rhythm has been abandoned or has proved impossible. As each episode of AF continues, progressive electroanatomical remodelling occurs that may serve to perpetuate and sustain AF, known as 'AF begets AF' [2]. Therefore, it is not surprising that treatment strategies vary in effectiveness depending on the extent and duration of AF.

Overall, optimal AF management should include a holistic, comprehensive, multidisciplinary approach that collectively considers modifiable risk factors, stroke prevention, and patient- and symptom-focused rate and rhythm control [3] (Figure 1). Using this approach, known as the AF Better Care (ABC) pathway, AF is managed with lifestyle modifications to address risk factors such as obesity and hypertension, and medical therapy which can include anticoagulation for stroke prevention as well as rate and rhythm control drugs depending on the patient and symptoms [1,3]. When antiarrhythmic drugs fail or are intolerable, ablation is recommended. This typically takes form as standalone endocardial catheter ablation or as surgical ablation if performed concomitantly with a primary cardiac surgical procedure. In both cases, pulmonary vein isolation (PVI) is paramount, although

**Citation:** Kaba, R.A.; Momin, A.; Camm, J. Persistent Atrial Fibrillation: The Role of Left Atrial Posterior Wall Isolation and Ablation Strategies. *J. Clin. Med.* **2021**, *10*, 3129. https:// doi.org/10.3390/jcm10143129

Academic Editor: Charles Guenancia

Received: 20 June 2021 Accepted: 13 July 2021 Published: 15 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

other regions often emerge as potential substrates in PersAF [4]. One of these regions, arguably the most influential after the pulmonary veins (PVs), is the posterior wall of the left atrium, which is known to generate AF triggers and is subject to electrical and structural changes that occur with the persistence of AF. However, this region is where endocardial catheter ablation is more limited in its capacity to comprehensively address the AF substrate owing to the elevated risk of collateral damage to adjacent structures such as the oesophagus. This review aims to discuss the published literature on the role of the left atrial posterior wall in PersAF and LSPersAF, outline practical limitations of endocardial catheter ablation to safely and durably isolate the posterior wall and describe the rationale for a hybrid epicardial–endocardial ablation strategy for silencing the PVI and posterior wall.

**Figure 1.** Risk factors for perpetuation of AF. Modifiable risk factors are highlighted separately. LSPersAF, long-standing persistent AF; LA, left atrium; PWD, p-wave duration; LVH, left ventricular hypertrophy; BMI, body mass index; OSA, obstructive sleep apnoea; DM, diabetes mellitus.

#### **2. Burden of Atrial Fibrillation**

Atrial fibrillation was estimated to affect more than 43 million people worldwide in 2016, a figure that continues to rise every year, with observed increases during the last few decades in associated disability and mortality [3]. Atrial fibrillation increases the risk of stroke around five-fold [5], more so with multiple co-existing risk factors, and is also associated with increased mortality [6] even within the first few months of diagnosis [7]. Atrial fibrillation can also overlap with heart failure in that it can exacerbate existing heart failure or lead to tachycardia-induced cardiomyopathy in patients with chronic, poorly managed AF. Therefore, the impact of AF on the global healthcare system is significant. In addition, AF is associated with a decreased quality of life, which can be attributed to the burden of symptoms, as well as the complex interplay with other patient comorbidities commonly associated with AF [8]. In effect, treating the syndrome with AF is not only aimed at reducing the risks of stroke and cardiac death but also decreasing AF burden and, consequently, AF symptoms to improve patient quality of life.

#### **3. Treatment of Paroxysmal Atrial Fibrillation**

In a seminal 1998 paper, Häissaguerre and colleagues identified the PVs as the primary sites of arrhythmogenicity in paroxysmal AF and that these AF triggers could be destroyed or isolated with radiofrequency ablation [9]. Favourable success rates have been

demonstrated for endocardial catheter ablation focused on the PVs for the treatment of drug-refractory paroxysmal AF [10], further supported by advancements in catheter-based radiofrequency, cryoballoon and other technologies [11,12].

#### **4. Paroxysmal vs. Persistent Atrial Fibrillation: Differences in Treatment Outcomes**

The consistent clinical success of endocardial catheter ablation in paroxysmal AF is not paralleled in persistent and long-standing persistent forms of AF. The discrepancy between paroxysmal and non-paroxysmal AF outcomes is well-evidenced by long-term results of endocardial radiofrequency ablation in these subgroups. With a median follow-up of 4.8 years after circumferential PVI, Ouyang et al. reported 46.6% of 161 patients with paroxysmal AF were free from atrial arrhythmia recurrence after a single procedure, and this success rose to 79.5% with multiple procedures (median 1: range 1–3) [13]. However, in long-standing PersAF, with a median follow-up of 4.7 years, the same investigators reported 20.3% of 202 patients were free from arrhythmia recurrence after PVI with additional CFAE/linear ablation [14]. After multiple procedures (median 2, range 1–5), 45% of patients with LSPersAF were in sinus rhythm. A comprehensive meta-analysis of persistent and long-standing PersAF treatment outcomes reported similarly disappointing results with much shorter follow-up times [15].

One explanation for the suboptimal effectiveness of PVI in non-paroxysmal AF is that areas outside of the PVs can drive and act as substrates as AF continues [16]. It has been well-documented that AF triggers are present outside of the PVs [17,18]. While extra-PV triggers may be present in paroxysmal AF, the majority of triggers are located in and around the PVs (Figure 2); this may, at least in part, explain why PV isolation alone is more effective in treating this type of AF [19]. However, as AF becomes persistent, there is a shift towards extra-PV triggers for atrial tachyarrhythmias and, given the progressive electrophysiological and structural changes that occur with the persistence of AF, these extra-PV regions may be appropriate substrates for ablation in PersAF and LSPersAF. Having said that, what and how to ablate in PersAF and LSPersAF is still unclear. Data from the STAR-AF II trial appeared to show that additional endocardial ablation utilising CFAEs or certain linear lesions (roof and mitral lines) adjunctive to PVI did not improve clinical outcomes over PVI alone in PersAF [20]; although, dedicated posterior wall ablation was not specifically tested in this study.

**Figure 2.** Triggers and substrates for PAF vs. PersAF. In PAF, the majority of these are located within and around the PVs, whereas in PersAF there are many more non-PV locations, especially in the posterior wall (between the four PVs and below the lower PVs). LA, left atrium; LAA, left atrial appendage; RA, right atrium; RAA, right atrial appendage; SVC, superior vena cava; IVC, inferior vena cava.

The left atrial posterior wall has been shown to house the highest proportion of non-PV triggers. Lin et al. reported 38% of non-PV ectopic beats emanated from the posterior

wall [17]. Additionally, with continued PersAF, the left atrial posterior wall is the most common non-PV site to contain AF re-entrant drivers [16]. In the next section, we review the unique arrhythmogenic properties of the posterior wall that underscore the rationale for its role in PersAF and LSPersAF.

#### **5. The Posterior Wall of the Left Atrium in Non-Paroxysmal Atrial Fibrillation**

#### *5.1. Intrinsic Features*

The left atrial posterior wall has several inherent anatomic and electrophysiological properties that are conducive to arrhythmogenicity. When these factors are combined with the structural changes that develop with more prolonged episodes of AF (see next section on 'Effects of prolonged atrial fibrillation on posterior wall'), the posterior wall then emerges as one of the key regions in the pathophysiology of PersAF. The posterior wall is derived embryonically from the same tissue as the pulmonary veins [21]. Between approximately 6–8 weeks of gestation, the common pulmonary vein, lined with mediastinal myocardium distinct from the primary myocardium that lines systemic venous structures, bifurcates and becomes incorporated into the left atrial wall [22]. Of note, the mediastinal myocardium is composed of fast-conducting cells compared to the slower conducting cells of the primary myocardium. Given the shared tissue origin with the pulmonary veins, it is not surprising then that the posterior left atrial wall is also a site of AF triggers and plays a role in sustaining PersAF.

Myocytes within the left atrial posterior wall have unique electrophysiological properties that may be intrinsically suited to initiate or sustain AF. These cells are characterised by having larger late sodium currents and smaller potassium currents [23]. The intracellular calcium transient and content within the sarcoplasmic reticulum are high. In effect, the cells of the posterior wall have (i) a low resting membrane potential; (ii) short action potential duration; (iii) the shortest refractory period of any cell in the heart. Taken together, these cellular characteristics make the posterior wall prone to misfiring.

Other structural aspects of the posterior wall can contribute to AF initiation and facilitate re-entry. The myocardial fibres in the left atrial posterior wall, particularly near the junction with the pulmonary veins, have a heterogenous orientation with respect to each other [24]. As a consequence, non-uniform anisotropy can occur in which conduction velocity and depolarisation differ between adjacent tissues, including the transition between the epicardial and endocardial layers. Subsequently this can lead to delayed conduction, unidirectional block and, thus, local re-entry.

The autonomic nervous system is a key player in the initiation and sustainment of AF. The posterior wall of the left atrium has the highest density of autonomic neurons in the heart [25]. Ganglionated plexi are groups of autonomic neurons embedded in epicardial fat pads, and some of the ganglionated plexi are located at the posterior left atrium, near the pulmonary veins. Ganglionated plexi are thought to contribute to AF and at times are adjunctive targets in ablation procedures.

#### *5.2. Effects of Prolonged Atrial Fibrillation on Posterior Wall*

As described above, there are intrinsic functional and anatomical characteristics of the left atrial posterior wall that make it prone to the initiation and maintenance of AF. Once AF occurs and persists over time, progressive changes in the left atrium then serve to propagate and further sustain AF. As such, the left atrial posterior wall is acknowledged as a key AF substrate in persistent forms of the disease. The development of fibrosis is thought to be a contributing factor to the propagation and persistence of AF. Fibrosis can develop due to other cardiac abnormalities or health conditions that are coincident with AF, as well as aging. Fibroblasts comprise 50–70% of cardiac cells [25], and their function is to compose and dynamically maintain the heart's scaffold [26]. These fibroblasts can differentiate into myofibroblasts under various pathologic conditions, including inflammation and mechanical overload. Myofibroblasts, in turn, produce, turn over and deposit collagen and other extracellular matrix components, which lead to the hardening and scarring of

cardiac tissue. This fibrotic tissue can slow conduction, serve as a unidirectional block and contribute to macro re-entry [27]. Cochet et al. demonstrated through MRI delayed enhancement that fibrosis tends to develop on the posterior left atrium [28]. This may be in part due to chronic, increased stress in the regions adjacent to the pericardial reflections that anchor the posterior heart to the chest wall [29]. Additionally, increased pressure and dilation of the left atrium due to prolonged AF leads to stretching, followed by inflammation and leading to fibrosis [30].

The accumulation of epicardial fat on the posterior wall can also contribute to AF in two ways. Firstly, adipose tissue produces inflammatory signals that support remodelling and fibrosis [31]. Secondly, animal studies have suggested infiltration of epicardial adipose tissue into the myocardium may create tissue disorganisation that can serve as a substrate for aberrant conduction [32]. Areas of abnormal conduction in the posterior left atrium have been shown to be associated with adjacent epicardial adipose tissue in obese patients with AF [33].

#### *5.3. Difficulties with Endocardial Ablation of Posterior Wall*

Given the evidence for the posterior wall as an AF substrate, both in triggering and sustaining AF, the posterior wall has been explored as a target of radiofrequency and cryoablation to improve clinical outcomes in AF, particularly PersAF and LSPersAF. This is evident from the Cox-Maze IV surgical ablation lesion set, which isolates the posterior wall of the left atrium with epicardial ablation lines on the right and left pulmonary vein antrum followed by roof and floor ablations anchored to the left atriotomy [34]. However, Cox-Maze IV is typically performed concomitantly with open cardiac surgeries, limiting its reach to patients who do not need or want an open procedure.

Endocardial catheter isolation of the left atrial posterior wall has been studied with both radiofrequency and cryothermal ablation (Table 1). The majority of these studies included only patients with PersAF and LSPersAF, which is in line with current guideline recommendations when considering posterior wall isolation in conjunction with PVI [1]. Meta-analyses of a few randomised and observational comparison studies have suggested an overall benefit of endocardial posterior wall ablation compared to pulmonary vein isolation alone in PersAF [35,36], but results of the individual studies, including the randomised clinical trials [37–39], are mixed (Table 1). This may, in part, be due to the lack of a standardised approach to posterior wall isolation, which is evidenced by the various lesion sets used in published studies. These approaches to posterior wall isolation include a single ring around the PVs and posterior left atrium [39], linear lesions (left atrial roof and posterior-inferior) to create the so-called posterior 'box' lesion [37,38,40,41], or extensive point-by-point radiofrequency [42] or segmental cryoballoon ablation [43,44] to debulk the posterior wall. Adjunctive lesions also vary among these studies.


**Table 1.** Summary of select studies evaluating addition of posterior wall isolation to pulmonary vein isolation.


**Table1.***Cont*.

1 Percentages depicted in Kaplan–Meier curve in Aryana et al. 2018 as noted in Della Rocca et al. 2020; AAD: antiarrhythmic drugs; AF: atrial fibrillation; CTI: cavotriscupid isthmus; LSPersAF: long-standing persistent AF; MTI: mitral isthmus; PAF: paroxysmal AF; PersAF: persistent AF; PVAI: pulmonary vein antrum isolation; PVI: pulmonary vein isolation; PW: posterior wall; RF: radiofrequency; SVC: superior vena cava.

In addition to a lack of standardised posterior wall ablation strategy, other practical challenges limit the extent to which endocardial posterior wall isolation can be achieved and thus may contribute to varied clinical outcomes. One major concern with endocardial catheter ablation of the left atrial posterior wall is potential collateral damage. The tissue of the posterior wall is thin, particularly at the superior aspect, in part to accommodate the stress of limited cardiac motion at the pericardial reflections [45]. It has been shown using post-mortem hearts that the posterior wall tissue is generally thinner in patients with AF, with an overall mean thickness of ≤3 mm [45]. Endocardial catheters apply ablative energy away from the heart towards the pericardium, therefore there are risks of cardiac perforation and tamponade as well as thermal injury to the oesophagus and other adjacent structures. Atrio-oesophageal fistula is the most devastating consequence of oesophageal thermal injury. While the documented incidence is low (<0.1%) with endocardial posterior wall ablation, the potential risk remains, and the consequences can be fatal [46]. Oesophageal temperature monitoring during ablation may be used as an alert for thermal injury; however, there are well recognised limitations such as the temperature can continue to rise after ablation is stopped and the probe may cause oesophageal damage by thermal effect. Consequently, despite the use of this device, atrio-oesophageal fistula can still develop [47], limiting the widespread use of such an approach for monitoring. Indeed, a recent randomised study demonstrated a similar rate of endoscopically-detected oesophageal lesions following endocardial catheter ablation with and without the use of an oesophageal temperature probe [48]. Additionally, aborting ablation due to an unexpected rise in temperature may result in incomplete ablation lines and gaps. Reducing the power and/or duration of energy delivery during ablation on the posterior wall is normally undertaken to reduce the risk of collateral damage, but this also reduces the efficacy of lesion formation. Taken together, active mitigation of thermal injury is important, yet it may also contribute to incomplete isolation of the posterior wall and varied clinical outcomes.

Reported rates of acute and continued isolation of the posterior wall using endocardial catheter ablation suggest there is difficulty in creating transmural and durable lesions. A meta-analysis of endocardial posterior wall isolation found an acute procedural success rate of 78% (95% CI, 59.4–94.4%) with results from box, single ring and debulking techniques combined [35]. The same meta-analysis also reported a substantial rate of posterior wall reconnections observed at repeat electrophysiology procedures for arrhythmia recurrence after endocardial catheter ablation: the pooled rate of posterior wall reconnection was 63.1% (95% CI, 42.5–82.4%) [35]. Markman et al. assessed chronic posterior wall isolation at repeat ablation after a single procedure of PVI and posterior wall ablation. They found a 40% rate of posterior wall reconnections in patients who experienced arrhythmia recurrence, with most reconnections at the atrial roof and most recurrences classified as atrial flutter in patients with failed posterior wall isolation [49]. Bai et al. reported 37.5% of patients had posterior wall reconnections three months after a single endocardial posterior wall debulking procedure [42]. In fact, four of the studies comparing PVI to PVI with posterior wall isolation discussed herein suggest suboptimal durability of posterior wall isolation using endocardial catheter ablation (Table 2).

Evidence of endocardial–epicardial dissociation in atrial fibrillation may also limit the effectiveness of endocardial posterior wall isolation, especially when considered in the context of suboptimal transmurality. Endocardial–epicardial dissociation, as evidenced by asynchronous activation of the epicardial and endocardial surfaces, was initially demonstrated in animal [50] and computational models [51]. More recently, real-time mapping has shown there may be up to 50–55% asynchronous activation between the epicardial and endocardial surfaces in patients with AF [52,53]. One contributing factor to endocardial– epicardial dissociation in AF may be the presence of fibrosis in the epicardial layer, which was first suggested by animal studies [54] and recently supported by computational modelling with validation in a small number of patients [55]. The cumulative evidence for endocardial–epicardial dissociation suggests that endocardial-only mapping and abla1

tion may be insufficient to adequately address conduction abnormalities on both cardiac surfaces in AF.


**Table 2.** Posterior wall (PW) connection rates in studies comparing pulmonary vein isolation (PVI) to PVI + PW isolation.

Includes pulmonary vein and PW reconnections; PVI: pulmonary vein isolation; PW: posterior wall; RF: radiofrequency.

#### *5.4. Hybrid Epicardial–Endocardial Approach to Address Posterior Wall Silencing*

In effect, there are three main challenges in the treatment of PersAF and LSPersAF: (i) limited candidates for concomitant surgical ablation; (ii) limited effectiveness of catheter ablation in non-paroxysmal AF; (iii) challenges with endocardial catheter ablation focused on the left atrial posterior wall, which is a source of AF triggers and a substrate. These issues prompted the development of hybrid epicardial–endocardial approaches to ablation. Hybrid approaches combine minimally invasive epicardial ablation by a cardiothoracic surgeon and endocardial ablation by an electrophysiologist to complete a transmural lesion set that effectively isolates the pulmonary veins and left atrial posterior wall.

There are two general strategies for hybrid epicardial–endocardial ablation. The primary difference is the surgical epicardial ablation technique, including epicardial access, ablation tools and posterior wall lesion set. Hybrid ablation can be achieved with totally thoracoscopic (TT) epicardial ablation followed by endocardial ablation. In hybrid TT ablation, surgical access to the pericardium is achieved thoracoscopically and the epicardial lesion set is focused on PVI and creating a box lesion set across the posterior wall. Endocardial ablation is performed by an electrophysiologist to complete PVI and address gaps. The first report of this approach was published in 2011 [56]. Recent retrospective studies have reported mid-term (2–3 year) outcomes ranging from 67–79% arrhythmia-free survival off AADs in patients with PersAF and LSPersAF [57–59]. Safety and effectiveness of hybrid TT ablation are being evaluated in two randomised clinical trials (NCT02441738, NCT02695277) and one single-arm trial (NCT02393885).

In the other hybrid epicardial–endocardial approach, commonly referred to as the hybrid Convergent procedure, the surgeon uses a single, small subxiphoid incision to gain access to the pericardial space without the use of additional ports. It was initially proposed in 2009 [60] and the ablation set has evolved over time. In early studies, an ex-Maze lesion set was performed through a transabdominal, transdiaphragmatic approach [60]. A box lesion set then became the preferred method to isolate the posterior wall. Since 2012, epicardial posterior wall homogenization has been achieved with 2–3 rows of linear lesions spanning between the pulmonary veins [61], which is another distinction from the TT lesion set. Beginning in 2016, the pericardial space has been accessed via the subxiphoid incision [62], eliminating the need to divide the central tendon of the diaphragm. Endocardial mapping and ablation are subsequently performed by the electrophysiologist on the same day, sequential day, or several weeks later, with the goal of ensuring PVI and addressing any gaps following the epicardial procedure. Further, since there is recovery of electrical conduction following epicardial ablation, it remains important to undertake both components of the hybrid technique to achieve long-lasting, widespread transmurality [63]. Observational clinical outcomes from contemporary analyses have suggested favourable outcomes with this technique [64–68], which were recently corroborated by the results of the multi-centre, randomised controlled CONVERGE trial [69]. The trial compared hybrid Convergent ablation with endocardial catheter ablation in PersAF and LSPersAF and met its primary safety and effectiveness endpoints. Twelve-month freedom from atrial arrhyth-

mias without new/increased doses of AADs was 67.7% with hybrid Convergent ablation compared to 50.0% with catheter ablation (*p* = 0.036). Significantly better effectiveness off AADs (53.5% vs. 32.0%, *p* = 0.013) and irrespective of AADs (76.8% vs. 60.0%, *p* = 0.033) were also achieved with hybrid Convergent ablation. The 30-day major adverse event rate with the hybrid Convergent procedure was 7.8% (vs 0.0% in the catheter arm, *p* = 0.0525), primarily relating to inflammatory pericardial effusions. Of note, no cardiac perforations, deaths or atrio-oesophageal fistulas occurred.

One important aspect of both hybrid ablation strategies is a collaborative, heart team approach to patient management in order to optimise clinical outcomes and safely mitigate risks [70].

#### **6. Future Directions**

Isolation of the left atrial posterior wall with a combination of epicardial and endocardial ablation to increase the likelihood of durable, transmural lesion has shown promising results in observational studies during the last decade as well as in a randomised controlled trial. More recently, these studies have described outcomes using a subxiphoid approach to reach the posterior left atrium, and additional studies dedicated to this approach will be important. Concomitant application of the AtriClip®, the most widely employed left atrial appendage exclusion device, is gaining popularity [4] and future studies should assess the precise impact on AF outcomes of including this technique. Another endpoint of interest for a hybrid approach is evaluating the length of stay for comparison with other minimally invasive surgical ablation approaches. For example, in our experience, we have seen rapid recovery after hybrid Convergent ablation, with a median length stay of 1 day, in contrast to recovery times for patients who undergo totally thoracoscopic Maze procedures, who typically require several days prior to discharge.

#### **7. Conclusions**

The left atrial posterior wall is likely an important driver and substrate as AF progresses and, as such, its isolation has been explored during AF ablation procedures to improve clinical outcomes in PersAF and LSPersAF. Surgical-only approaches to isolate the posterior wall are limited by invasiveness and patient eligibility for a concomitant procedure. Endocardial ablation alone to isolate the posterior wall has yielded mixed results in PersAF and LSPersAF. Electrophysiological differences between the endocardium and epicardium may not be safely addressable with an endocardial approach alone. The combination of the two concepts into a hybrid electrophysiological–surgical collaboration, such as in the Convergent procedure, may help to optimise lesion durability and transmurality to effectively isolate the posterior wall.

**Author Contributions:** All authors have contributed to the preparation of this paper, R.A.K., A.M. and J.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Acknowledgments:** The authors wish to thank Yashasvi Awasthi and Kristen Plasseraud for their kind assistance in the preparation of this article.

**Conflicts of Interest:** R. A. Kaba is a consultant for Daiichi Sankyo, Bayer, Atricure and Biotronik. A. Momin is a consultant for Atricure. A. J. Camm has received personal fees from Abbott, Boston Scientific, Medtronic and Atricure.

#### **References**

