Pulsed field ablation (PFA) is supposed to be a non-thermal energy technique that uses high-voltage electrical fields of micro- or nanosecond duration to induce irreversible electroporation that destabilises cell membranes and results in cellular necrosis. Before its adoption in cardiology, PFA was used in oncology and pain management.1 Its application in cardiac electrophysiology initially focused on AF, with ongoing research exploring broader indications.
Electrical field ablation was pioneered in the 1980s when Scheinman et al. delivered a full defibrillator shock through a catheter in the heart.2 The investigators achieved heart block, but the accompanying heat and barotrauma sidelined direct current ablation and paved the way for radiofrequency (RF) ablation. The key technological advance for today’s PFA is that a ‘large charge’ can be broken into a series of multiple applications of very brief duration (microseconds or nanoseconds) that are controlled by a generator. The combination of pulses, their duration, and the specifics of catheter design are unique to each system and determine treatment efficacy and safety.
Basic Principles of Electroporation
When electrical fields are applied to the cardiac tissue, cellular membrane integrity and cellular homeostasis are disrupted. Included in these changes is the redistribution of extracellular and intracellular charged species that are important to cellular function. In addition, the pulsed electrical fields locally alter tissue pH, generate reactive oxygen species, stimulate mitochondrial cytochrome c release and activate caspase enzymes, which are among some of the mechanisms leading to cell death.3 If the loss of homeostasis is of sufficient magnitude and/or duration, the cell cannot restore normal function and dies through either immediate necrotic processes or various programmed cell death pathways. These processes occur over days to weeks, leading eventually to replacement fibrosis over a period of 2–4 weeks. The mechanism of cell death with PFA is predominantly apoptosis but includes necroptosis and pyroptosis (Figure 1 ).4
Unlike thermal ablation, PFA does not permanently disrupt the local tissue extracellular matrix or vascular supply, leaving the tissue capable of some continued function and potentially repopulation with new progenitor cells.5 For example, electroporation can disrupt both oesophageal smooth muscle and nerve axons, but in both cases repopulation with progenitor muscle or myelin cells can repair these tissues over days to weeks. Therefore, the risk of collateral damage to non-cardiac tissues such as the oesophagus, phrenic nerve and pulmonary veins (PVs) is potentially minor. Conversely, cardiac tissue is terminally differentiated and is incapable of such regeneration. The ability of PFA to differentially affect cardiac versus other collateral tissues was one of the chief reasons that it was pursued as an ablation option.
Designing a Pulsed Field Ablation System
Designing an optimal PFA system requires a balancing act between achieving adequate lesion depth for therapeutic efficacy and minimising the unwanted effects of delivering electrical impulses within the heart. Chief among these unwanted effects are generalised musculoskeletal stimulation (causing pain and patient movement), pain sensation (which can be separate from muscle stimulation) and unwanted thermal effects. Figure 2 shows how each of the various elements of pulse design can affect the various outcomes of therapeutic versus adverse outcomes.
Although monophasic pulses are more efficient in their ability to kill cardiac tissue, they cause more musculoskeletal stimulation and gaseous bubbles and increase the risk of arrhythmogenicity, so most current systems use biphasic pulses.3
Voltage is obviously important to achieving cardiac cell death, but increasing voltage can again cause musculoskeletal stimulation and thermal effects.3 The duration of exposure is also important for cell death. Breaking up a large voltage into small-duration pulses with the same voltage amplitude but delivered in several bursts can achieve the same magnitude of cell death while avoiding muscle stimulation and thermal effects. Pulse width, the number of pulses and the number of trains all determine the duration of cell exposure to the voltage. Most systems use pulse widths measured in the low microsecond (1–10 μs) range. Even small changes in pulse width (1 μs) can create major changes in lesion depth and muscle stimulation.3,6 Some systems use pulse widths of several hundred nanoseconds. A reduction in pulse width to the nanosecond range may help further reduce muscular stimulation, but requires an increase in voltage to keep the efficacy of the pulses similar to those of microsecond pulses (Figure 3 ).
Despite the ‘non-thermal’ nature of cell death in PFA, all PFA systems do create some incidental heat.3 As long as adequate pauses are programmed between the positive and negative phases of the pulse or between bursts of many pulses, then there is adequate time for the tissue to cool and avoid any thermal stacking. Electrode design is also important because electrical fields tend to concentrate on the edges of electrodes and can cause high current density resulting in heat. If electrodes are appropriately rounded and/or insulated, then some of this edge-effect heating can be mitigated.7
Thus, optimal PFA system design involves a constant compromise between treatment efficacy and reductions in unwanted muscular stimulation, heating, microbubble formation and other adverse effects. This results in a variety of different systems that have been developed or are currently in development (Tables 1 and 2 ).8–16
Safety of Pulsed Field Ablation
As noted above, the ability of PFA to differentially affect cardiac versus other collateral tissues was one of its chief purported benefits. Although there are extensive preclinical data supporting the safety of PFA concerning the oesophagus, phrenic nerve and pulmonary venous tissue, the question remains as to whether this has been borne out in clinical reality.17–19
Phrenic Nerve
Although transient phrenic nerve capture may occur during PFA applications, permanent phrenic nerve palsy remains an uncommon complication.20 Studies using porcine models to evaluate the effects of PFA on the phrenic nerve all observed that transient stunning of the nerve could occur in close proximity, but this typically recovered within minutes to hours.19 In the MANIFEST-17K registry, only transient phrenic palsies have been reported (0.06%), with no permanent phrenic palsy, although a prior case report of a permanent phrenic palsy was published.20,21
Oesophagus
Preclinical evidence indicates that oesophageal tissue exhibits notable resistance to PFA-induced damage. One preclinical study did note some oesophageal lesions acutely after PFA application, but all of these healed within 2 weeks, likely due to the preservation of the extracellular matrix and the ability for progenitor cells to repopulate.22 Adipose tissue, with its insulating properties, also likely plays a protective role, because even thin layers of fat separating the oesophagus from the left atrium may mitigate potential harm.3 However, some significant temperature rises have been recorded in the oesophagus during human ablation procedures using the pentaspline catheter, although temperature rises are almost undetectable with other systems.23,24 However, none of the European or US Food and Drug Administration approval studies have shown any evidence of oesophageal injury, and there were none documented in the MANIFEST-17K registry.8,10,24-28
Pulmonary Veins
Preclinical data show that PFA does not cause PV stenosis even when delivered deep in the veins.17 Most of the PFA approval studies to date have included subsets of data evaluating serial CT or MRI imaging to look for evidence of PV stenosis.8,10,24,25 None of the major clinical studies has demonstrated any evidence of PV stenosis in their subset data.8,10,24–27
New Problems Introduced by Pulsed Field Ablation
Based on the clinical trials, the safety profile of PFA is extremely promising. For example, the PULSED AF trial reported a total complication rate in both the paroxysmal AF and persistent AF cohorts of only 0.7%, one of the lowest complication rates ever reported in an AF ablation trial. However, as PFA use has increased, some unexpected complications have been reported, including haemolysis, coronary spasm, microbubble formation and unexpected thermal effects.
Haemolysis
In the MANIFEST-17K registry, which encompasses over 17,000 patients worldwide, there were isolated reports of haemolysis-related renal failure necessitating temporary dialysis.28 In that cohort, 5 (0.03%) individuals required dialysis after PFA.28 In addition, Venier et al. described two cases of acute kidney injury after PFA that led to extended hospitalisation.29 Although all PFA causes some haemolysis, the degree of haemolysis is dependent on the amount of electrode surface area exposed to the bloodstream, which is related to specific catheter size and design. Based on the available literature, the amount of haemolysis varies substantially between catheters; for example, the pentaspline and variable circular design catheters cause more haemolysis than a gold electrode array or spline balloon catheter.30–32 For catheters most prone to causing high degrees of haemolysis, the number of pulsed-field applications was identified as a significant risk factor for haemolysis and the development of renal injury.30,33
It has been suggested that renal injury from haemolysis can be mitigated by the infusion of 2 l of fluid after the procedure.34 However, this is a considerable fluid load, akin to volumes that were administered in the era of open irrigation catheters, and this does not prevent subclinical renal injury. Therefore, the solution lies in better system design to minimise haemolysis below the severe threshold.
Another potential cause of renal dysfunction could be myolysis, particularly in systems that involve greater musculoskeletal activation. Myoglobin release, although less studied, may contribute to kidney toxicity because circulating myoglobin has a similar nephrotoxic effect as free hemoglobin.35 Although most myoglobin release after ablation is presumably cardiac in origin, the rise seen in clinical studies is in excess of what would be expected from limited ablation in the left atrium. In fact, the peak levels would be akin to a very large anterior MI, which suggests that some of the contribution is coming from skeletal muscle activation.36 Even if myoglobin release is not extensive, the combination of free plasma haemoglobin and myoglobin is particularly toxic.
There are other, less understood consequences of haemolysis. Haemolysis triggers a consumption of nitric oxide, which, in turn, leads to vasoconstriction and platelet activation.35,37 Reports of delayed vasospasm could potentially be triggered by the consumption of nitric oxide, which can take hours to replete.38 Furthermore, platelet activation can lead to coagulopathy, which is not reduced by intravenous heparin and oral anticoagulants.37
Coronary Vasospasm
Coronary vasospasm, particularly in regions adjacent to coronary arteries (e.g. the cavotricuspid and mitral isthmus), has also been reported.39,40 In some instances, this has led to life-threatening ventricular arrhythmias. Pre-emptive administration of high doses of intravenous nitroglycerin has been suggested to mitigate this risk, although the doses required are 2–3 mg IV bolus followed by 2 mg every 2–5 minutes.41 These doses are so large that significant hypotension requiring vasopressor support almost always ensues.42 This makes such an approach quite impractical. Encouragingly, the rate of coronary vasospasm in the MANIFEST-17K study was only 0.14% (25 cases among 17,642 patients).28 A substudy on coronary spasm from the MANIFEST-17K trial identified two distinct types of coronary spasm.43 The first type was a generalised spasm, which occurred during PVI with PFA applied distally from the coronary arteries, which affected five patients. This type of spasm was observed after an average of five PFA applications at all PVs. Notable acute complications included VF (8.7%), transient high-degree atrioventricular block (8.7%) and severe coronary stenosis requiring angioplasty and stenting.43 The second type of spasm was proximity-induced spasm, observed during cavotricuspid isthmus or mitral annulus ablations in 18 patients. Over a mean follow-up period of 8.3 months, no further adverse events were reported, with no recurrence of angina or need for additional coronary interventions in any of the groups.43 It is hypothesised that generalised spasm may occur in patients already prone to spasm, such as those with Prinzmetal angina. It is unclear whether nitroglycerin pretreatment will actually prevent all coronary vasospasm or only prevent subclinical spasm. Long-term sequelae of vasospasm may be the development of neointimal hyperplasia causing long-term mild coronary stenosis.44,45 This has also been reported with thermal ablation, and it is unclear whether this poses a long-term risk.46 Ultimately, the best way to avoid vasospasm is to use RF in these high-risk areas, which is becoming much more feasible with dual-energy catheters.
Microbubbles and Thermal Profile
An additional concern is the formation of microbubbles during PFA delivery, which may play a role in the pathogenesis of silent cerebral embolism, which occurs in 9–15% of patients undergoing PFA.10,12,28,47 Three mechanisms of bubble formation have been identified: degassing due to decreased gas solubility at elevated temperatures; boiling because of thermal effects; and electrochemical reactions involving water electrolysis at the electrodes.8 The most likely contributing factor of bubble formation with biphasic high-frequency pulses is the thermal effect.48 There are large differences in microbubble formation seen with the various PFA systems approved or currently pending approval. Although silent cerebral emboli in and of themselves may not cause clinical stroke, the combination of a high thermal profile and coagulopathy stimulated by haemolysis can lead to clinical stroke. The variable loop catheter, for example, was recalled because of visible char formation on the catheter and a three- to fourfold increased risk of clinical stroke.49 The presence of char indicates that the thermal profile of the catheter is unacceptably high.
Clinical Outcomes of Pulsed Field Ablation for AF
There have been numerous single-arm approval trials that have assessed the efficacy of PFA for the treatment of AF.8,10,24–27,50 In these trials, rates of freedom from AF at 1 year have ranged from 70% to 75%, which is similar to rates reported in previous thermal ablation trials.51–54 However, there have only been a very limited number of randomised trials that have compared PFA to historical thermal ablation, either with RF or cryoablation. The key randomised clinical trials that have evaluated the efficacy and safety of PFA for AF are detailed below.
ADVENT (FARAPULSE Catheter)
The ADVENT trial was the first randomised controlled trial comparing pulsed field ablation (FARAPULSE, Boston Scientific) with conventional thermal ablation (either contact-force RF or cryoballoon) in patients with paroxysmal AF.8 In that multicentre prospective non-inferiority study, PFA demonstrated non-inferiority in both safety and efficacy endpoints. The recurrence rate for the PFA arm at 1 year was 73.3%, compared with 71.3% in the thermal arms.8 PFA was associated with significantly shorter procedure and left atrial dwell times, and less mild PV narrowing. However, silent cerebral emboli were observed in approximately 10% of patients undergoing PFA, compared with no patients in the thermal ablation group; however, the numbers were too small for statistical analysis.8
SPHERE Per-AF Trial (Sphere-9 Catheter)
The SPHERE Per-AF trial was a multicentre randomised non-inferiority trial evaluating a dual-energy (RF+PFA) lattice-tip catheter (Sphere-9; Affera-Medtronic) versus conventional contact-force RF ablation in patients with persistent AF.27 In all, 420 patients were randomised 1 : 1. At 1 year, the endpoint was achieved in 73.5% of the investigational arm, compared with 65.2% in the control group.27 Although superiority was not met, the Sphere-9 arm showed significantly shorter ablation and procedure times, and safety outcomes were comparable between groups.27 Importantly, because the investigational catheter delivered both RF and PFA, the SPHERE Per-AF trial does not represent an exclusively direct comparison between PFA and RF; however, the trial stated that the majority of ablations performed with the Sphere-9 system were PFA, although the exact numbers were not detailed.27
SINGLE SHOT CHAMPION Trial
The most recent SINGLE SHOT CHAMPION trial was a randomised open-label non-inferiority trial comparing the FARAPULSE PFA with cryoballoon ablation (Arctic Front; Medtronic) in 210 patients with paroxysmal AF.55 All patients received implantable loop recorders to monitor arrhythmia recurrence. The mean procedure duration was shorter with PFA (55 versus 73 minutes), with a similar fluoroscopy time.55 PFA resulted in higher post-procedural troponin levels (1,920 versus 1,114 ng/l). From day 91 to day 365, the first recurrence rate of atrial tachyarrhythmias was numerically lower in the PFA group, with a between-group difference of −13.6% (95% CI [−26.9%, −0.3%]), suggesting a trend towards superiority.55 However, secondary endpoints, such as overall AF burden, were identical in both groups and, in fact, the number of repeat ablations was higher in the PFA arm.55
Developing Pulsed Field Ablation Further for Applications Beyond Pulmonary Vein Isolation
Dual-energy Catheters
The safety of PFA in the treatment of atrial flutter, whether typical, targeting the cavotricuspid isthmus, or atypical, involving the mitral isthmus, remains uncertain due to the risk of vasospasm. Conversely, RF ablation has proven to be highly effective and generally safe for these types of arrhythmias. To harness the advantages of both energy sources, catheters and systems have been designed to enable toggling between PFA and RF.
In 2023, the Sphere catheter, a lattice-tip catheter compatible with both RF and PFA generators, was clinically evaluated, enabling operators to switch between RF and PFA as needed.11 PFA was universally used for posterior left atrium applications within the PV isolation lesion set, whereas anterior left atrium applications were performed using either RF or PFA, depending on operator preference. In addition to AF ablation, all spontaneously occurring atrial flutters (24 typical cavotricuspid isthmus-dependent and 25 atypical) were mapped and successfully ablated without requiring a catheter exchange to address the newly identified arrhythmia.11 In addition, two other catheters that have been historically used for RF ablation of AF, namely the SmartTouch (Biosense Webster) and TactiFlex (Abbott), are being evaluated for dual-energy applications switching between RF and PFA. The SmartFire trial has already been published, demonstrating favourable clinical outcomes and the FOCALFLEX trial is currently ongoing (NCT06271967).14
Sequential Application of Radiofrequency and Pulsed Field Ablation
As novel PFA technologies continue to emerge, various catheter–generator combinations will enable operators to seamlessly switch between RF and PFA energy, as described above.11,14 Because both modalities alter the electrical and thermal properties of tissue, their sequential application at the same site may have a synergistic effect on lesion depth.56 This phenomenon has been demonstrated in cancer treatment, where both RF applied before PFA and PFA applied before RF altered tissue electrical conductivity due to intercellular oedema and reduced impedance.57,58 In both cases, the combination of these energy sources resulted in deeper and wider lesions.
A preclinical cardiac study investigated the depth of sequential RF and PFA application, demonstrating greater lesion depth when both energies were applied sequentially (mean [±SD] lesion depth 4.9 ± 0.8 mm for RF alone, 3.5 ± 0.6 mm for PFA alone, 6.2 ± 1.8 mm for RF + PFA and 5.7 ± 1.3 mm for PFA+RF; p<0.0001 for both versus RF or PFA alone).59 Histological analysis of the combined lesions revealed central thermal necrosis surrounded by a large haemorrhagic and transitional PF zone (Figure 4 ).59
Safety of Pulsed Field Ablation
As noted above, the ability of PFA to differentially affect cardiac versus other collateral tissues was one of its chief purported benefits. Although there are extensive preclinical data supporting the safety of PFA concerning the oesophagus, phrenic nerve and pulmonary venous tissue, the question remains as to whether this has been borne out in clinical reality.17–19
Phrenic Nerve
Although transient phrenic nerve capture may occur during PFA applications, permanent phrenic nerve palsy remains an uncommon complication.20 Studies using porcine models to evaluate the effects of PFA on the phrenic nerve all observed that transient stunning of the nerve could occur in close proximity, but this typically recovered within minutes to hours.19 In the MANIFEST-17K registry, only transient phrenic palsies have been reported (0.06%), with no permanent phrenic palsy, although a prior case report of a permanent phrenic palsy was published.20,21
Oesophagus
Preclinical evidence indicates that oesophageal tissue exhibits notable resistance to PFA-induced damage. One preclinical study did note some oesophageal lesions acutely after PFA application, but all of these healed within 2 weeks, likely due to the preservation of the extracellular matrix and the ability for progenitor cells to repopulate.22 Adipose tissue, with its insulating properties, also likely plays a protective role, because even thin layers of fat separating the oesophagus from the left atrium may mitigate potential harm.3 However, some significant temperature rises have been recorded in the oesophagus during human ablation procedures using the pentaspline catheter, although temperature rises are almost undetectable with other systems.23,24 However, none of the European or US Food and Drug Administration approval studies have shown any evidence of oesophageal injury, and there were none documented in the MANIFEST-17K registry.8,10,24-28
Pulmonary Veins
Preclinical data show that PFA does not cause PV stenosis even when delivered deep in the veins.17 Most of the PFA approval studies to date have included subsets of data evaluating serial CT or MRI imaging to look for evidence of PV stenosis.8,10,24,25 None of the major clinical studies has demonstrated any evidence of PV stenosis in their subset data.8,10,24–27
New Problems Introduced by Pulsed Field Ablation
Based on the clinical trials, the safety profile of PFA is extremely promising. For example, the PULSED AF trial reported a total complication rate in both the paroxysmal AF and persistent AF cohorts of only 0.7%, one of the lowest complication rates ever reported in an AF ablation trial. However, as PFA use has increased, some unexpected complications have been reported, including haemolysis, coronary spasm, microbubble formation and unexpected thermal effects.
Haemolysis
In the MANIFEST-17K registry, which encompasses over 17,000 patients worldwide, there were isolated reports of haemolysis-related renal failure necessitating temporary dialysis.28 In that cohort, 5 (0.03%) individuals required dialysis after PFA.28 In addition, Venier et al. described two cases of acute kidney injury after PFA that led to extended hospitalisation.29 Although all PFA causes some haemolysis, the degree of haemolysis is dependent on the amount of electrode surface area exposed to the bloodstream, which is related to specific catheter size and design. Based on the available literature, the amount of haemolysis varies substantially between catheters; for example, the pentaspline and variable circular design catheters cause more haemolysis than a gold electrode array or spline balloon catheter.30–32 For catheters most prone to causing high degrees of haemolysis, the number of pulsed-field applications was identified as a significant risk factor for haemolysis and the development of renal injury.30,33
It has been suggested that renal injury from haemolysis can be mitigated by the infusion of 2 l of fluid after the procedure.34 However, this is a considerable fluid load, akin to volumes that were administered in the era of open irrigation catheters, and this does not prevent subclinical renal injury. Therefore, the solution lies in better system design to minimise haemolysis below the severe threshold.
Another potential cause of renal dysfunction could be myolysis, particularly in systems that involve greater musculoskeletal activation. Myoglobin release, although less studied, may contribute to kidney toxicity because circulating myoglobin has a similar nephrotoxic effect as free hemoglobin.35 Although most myoglobin release after ablation is presumably cardiac in origin, the rise seen in clinical studies is in excess of what would be expected from limited ablation in the left atrium. In fact, the peak levels would be akin to a very large anterior MI, which suggests that some of the contribution is coming from skeletal muscle activation.36 Even if myoglobin release is not extensive, the combination of free plasma haemoglobin and myoglobin is particularly toxic.
There are other, less understood consequences of haemolysis. Haemolysis triggers a consumption of nitric oxide, which, in turn, leads to vasoconstriction and platelet activation.35,37 Reports of delayed vasospasm could potentially be triggered by the consumption of nitric oxide, which can take hours to replete.38 Furthermore, platelet activation can lead to coagulopathy, which is not reduced by intravenous heparin and oral anticoagulants.37
Coronary Vasospasm
Coronary vasospasm, particularly in regions adjacent to coronary arteries (e.g. the cavotricuspid and mitral isthmus), has also been reported.39,40 In some instances, this has led to life-threatening ventricular arrhythmias. Pre-emptive administration of high doses of intravenous nitroglycerin has been suggested to mitigate this risk, although the doses required are 2–3 mg IV bolus followed by 2 mg every 2–5 minutes.41 These doses are so large that significant hypotension requiring vasopressor support almost always ensues.42 This makes such an approach quite impractical. Encouragingly, the rate of coronary vasospasm in the MANIFEST-17K study was only 0.14% (25 cases among 17,642 patients).28 A substudy on coronary spasm from the MANIFEST-17K trial identified two distinct types of coronary spasm.43 The first type was a generalised spasm, which occurred during PVI with PFA applied distally from the coronary arteries, which affected five patients. This type of spasm was observed after an average of five PFA applications at all PVs. Notable acute complications included VF (8.7%), transient high-degree atrioventricular block (8.7%) and severe coronary stenosis requiring angioplasty and stenting.43 The second type of spasm was proximity-induced spasm, observed during cavotricuspid isthmus or mitral annulus ablations in 18 patients. Over a mean follow-up period of 8.3 months, no further adverse events were reported, with no recurrence of angina or need for additional coronary interventions in any of the groups.43 It is hypothesised that generalised spasm may occur in patients already prone to spasm, such as those with Prinzmetal angina. It is unclear whether nitroglycerin pretreatment will actually prevent all coronary vasospasm or only prevent subclinical spasm. Long-term sequelae of vasospasm may be the development of neointimal hyperplasia causing long-term mild coronary stenosis.44,45 This has also been reported with thermal ablation, and it is unclear whether this poses a long-term risk.46 Ultimately, the best way to avoid vasospasm is to use RF in these high-risk areas, which is becoming much more feasible with dual-energy catheters.
Microbubbles and Thermal Profile
An additional concern is the formation of microbubbles during PFA delivery, which may play a role in the pathogenesis of silent cerebral embolism, which occurs in 9–15% of patients undergoing PFA.10,12,28,47 Three mechanisms of bubble formation have been identified: degassing due to decreased gas solubility at elevated temperatures; boiling because of thermal effects; and electrochemical reactions involving water electrolysis at the electrodes.8 The most likely contributing factor of bubble formation with biphasic high-frequency pulses is the thermal effect.48 There are large differences in microbubble formation seen with the various PFA systems approved or currently pending approval. Although silent cerebral emboli in and of themselves may not cause clinical stroke, the combination of a high thermal profile and coagulopathy stimulated by haemolysis can lead to clinical stroke. The variable loop catheter, for example, was recalled because of visible char formation on the catheter and a three- to fourfold increased risk of clinical stroke.49 The presence of char indicates that the thermal profile of the catheter is unacceptably high.
Clinical Outcomes of Pulsed Field Ablation for AF
There have been numerous single-arm approval trials that have assessed the efficacy of PFA for the treatment of AF.8,10,24–27,50 In these trials, rates of freedom from AF at 1 year have ranged from 70% to 75%, which is similar to rates reported in previous thermal ablation trials.51–54 However, there have only been a very limited number of randomised trials that have compared PFA to historical thermal ablation, either with RF or cryoablation. The key randomised clinical trials that have evaluated the efficacy and safety of PFA for AF are detailed below.
ADVENT (FARAPULSE Catheter)
The ADVENT trial was the first randomised controlled trial comparing pulsed field ablation (FARAPULSE, Boston Scientific) with conventional thermal ablation (either contact-force RF or cryoballoon) in patients with paroxysmal AF.8 In that multicentre prospective non-inferiority study, PFA demonstrated non-inferiority in both safety and efficacy endpoints. The recurrence rate for the PFA arm at 1 year was 73.3%, compared with 71.3% in the thermal arms.8 PFA was associated with significantly shorter procedure and left atrial dwell times, and less mild PV narrowing. However, silent cerebral emboli were observed in approximately 10% of patients undergoing PFA, compared with no patients in the thermal ablation group; however, the numbers were too small for statistical analysis.8
SPHERE Per-AF Trial (Sphere-9 Catheter)
The SPHERE Per-AF trial was a multicentre randomised non-inferiority trial evaluating a dual-energy (RF+PFA) lattice-tip catheter (Sphere-9; Affera-Medtronic) versus conventional contact-force RF ablation in patients with persistent AF.27 In all, 420 patients were randomised 1 : 1. At 1 year, the endpoint was achieved in 73.5% of the investigational arm, compared with 65.2% in the control group.27 Although superiority was not met, the Sphere-9 arm showed significantly shorter ablation and procedure times, and safety outcomes were comparable between groups.27 Importantly, because the investigational catheter delivered both RF and PFA, the SPHERE Per-AF trial does not represent an exclusively direct comparison between PFA and RF; however, the trial stated that the majority of ablations performed with the Sphere-9 system were PFA, although the exact numbers were not detailed.27
SINGLE SHOT CHAMPION Trial
The most recent SINGLE SHOT CHAMPION trial was a randomised open-label non-inferiority trial comparing the FARAPULSE PFA with cryoballoon ablation (Arctic Front; Medtronic) in 210 patients with paroxysmal AF.55 All patients received implantable loop recorders to monitor arrhythmia recurrence. The mean procedure duration was shorter with PFA (55 versus 73 minutes), with a similar fluoroscopy time.55 PFA resulted in higher post-procedural troponin levels (1,920 versus 1,114 ng/l). From day 91 to day 365, the first recurrence rate of atrial tachyarrhythmias was numerically lower in the PFA group, with a between-group difference of −13.6% (95% CI [−26.9%, −0.3%]), suggesting a trend towards superiority.55 However, secondary endpoints, such as overall AF burden, were identical in both groups and, in fact, the number of repeat ablations was higher in the PFA arm.55
Developing Pulsed Field Ablation Further for Applications Beyond Pulmonary Vein Isolation
Dual-energy Catheters
The safety of PFA in the treatment of atrial flutter, whether typical, targeting the cavotricuspid isthmus, or atypical, involving the mitral isthmus, remains uncertain due to the risk of vasospasm. Conversely, RF ablation has proven to be highly effective and generally safe for these types of arrhythmias. To harness the advantages of both energy sources, catheters and systems have been designed to enable toggling between PFA and RF.
In 2023, the Sphere catheter, a lattice-tip catheter compatible with both RF and PFA generators, was clinically evaluated, enabling operators to switch between RF and PFA as needed.11 PFA was universally used for posterior left atrium applications within the PV isolation lesion set, whereas anterior left atrium applications were performed using either RF or PFA, depending on operator preference. In addition to AF ablation, all spontaneously occurring atrial flutters (24 typical cavotricuspid isthmus-dependent and 25 atypical) were mapped and successfully ablated without requiring a catheter exchange to address the newly identified arrhythmia.11 In addition, two other catheters that have been historically used for RF ablation of AF, namely the SmartTouch (Biosense Webster) and TactiFlex (Abbott), are being evaluated for dual-energy applications switching between RF and PFA. The SmartFire trial has already been published, demonstrating favourable clinical outcomes and the FOCALFLEX trial is currently ongoing (NCT06271967).14
Sequential Application of Radiofrequency and Pulsed Field Ablation
As novel PFA technologies continue to emerge, various catheter–generator combinations will enable operators to seamlessly switch between RF and PFA energy, as described above.11,14 Because both modalities alter the electrical and thermal properties of tissue, their sequential application at the same site may have a synergistic effect on lesion depth.56 This phenomenon has been demonstrated in cancer treatment, where both RF applied before PFA and PFA applied before RF altered tissue electrical conductivity due to intercellular oedema and reduced impedance.57,58 In both cases, the combination of these energy sources resulted in deeper and wider lesions.
A preclinical cardiac study investigated the depth of sequential RF and PFA application, demonstrating greater lesion depth when both energies were applied sequentially (mean [±SD] lesion depth 4.9 ± 0.8 mm for RF alone, 3.5 ± 0.6 mm for PFA alone, 6.2 ± 1.8 mm for RF + PFA and 5.7 ± 1.3 mm for PFA+RF; p<0.0001 for both versus RF or PFA alone).59 Histological analysis of the combined lesions revealed central thermal necrosis surrounded by a large haemorrhagic and transitional PF zone (Figure 4 ).59
Two studies have demonstrated the feasibility of a hybrid approach (RF+PFA or PFA+RF) using a lattice-tip dual-energy catheter.60,61 The first study included four patients, whereas the other enrolled 18 patients with structural heart disease and ventricular tachycardia (VT) refractory to prior RF ablation. In the first study, the use of PFA first or RF first was at the operator’s discretion, whereas in the second study, an approach was used in which RF was applied initially, followed by PFA to consolidate the lesions.60,61 Epicardial ablation was performed in four patients, with coronary spasm observed in three through coronary angiography (and in only one patient via ECG changes). This spasm was successfully resolved with intracoronary nitroglycerin infusion. The lesion depth, potentially measured by MRI, has not been evaluated, and long-term follow-up data have not been reported.
More Effective Pulsed Field Ablation
Current PFA systems have been optimised for atrial ablation with tissue depths of 3–5 mm. However, if we need to address substrates in thicker tissues, such as the ventricle, then we need to push the limits of tissue depth to 10 mm or beyond. The simplest way to do this is to increase the voltage, but because most PFA systems are currently at the thermal threshold, this will only further increase the thermal profile of the energy. Perhaps in the ventricle this may not be an issue because we are not as close to sensitive structures such as the oesophagus. Furthermore, traditional RF energy with new catheter forms (like a spherical design) can create 10-mm-deep lesions without the use of PFA.62 If we do deeper lesions with PFA, then various strategies need to be considered.
One strategy is to simply repeat applications of PFA. Yavin et al. investigated the impact of single versus multiple PFA applications in an in vivo porcine model using a bipolar configuration.62 In that study, mean (±SD) lesion depth increased significantly from 5.6 ± 1.4 to 8.8 ± 0.7 mm (p<0.001) at 24 hours after the procedure.62 More importantly, these differences persisted over time, with lesion depth at 24 days measuring 3.9 ± 0.9 mm after a single application, compared with 7.3 ± 0.8 mm with four applications.62 However, there is a plateau effect because the lesion cannot go deeper than that predicted by the electric field strength and properties of the cardiac tissue.63 Therefore, this is unlikely to be a long-term solution (Figure 5 ).62,64–68
Another strategy is to change the return electrode for the catheter. Instead of delivering from a catheter to a return patch on the skin, we could consider delivering from the tip of a catheter to a multipolar return catheter in the coronary sinus or inferior vena cava. There is evidence that this could help drive deeper lesions in the ventricle.69 Bipolar ablation involving the use of two catheters can increase current density across the tissue volume enclosed between the electrodes, thereby enhancing the energy effect.70 Thus far, the feasibility of bipolar electroporation ablation has only been assessed in a single preclinical study, which investigated the use of two lattice-tip catheters applied to the septal and free wall of the left ventricle. Lesion analysis performed after 2 days revealed a mean (±SD) depth of 14.3 ± 4.7 mm and, unlike RF ablation, this approach preserved structural integrity.71
Finally, duty cycling of various electrode configurations on a catheter can help drive tissue depth. On the nine-electrode gold array, the electrodes are activated on every second electrode, so that only the odd and then the even electrodes are activated at any given time. This ended up driving increased tissue depth for the same amount of voltage.18 With a novel single-point catheter, there have been reports of very large ventricular lesions being formed in excess of 11 mm.72 This catheter also involves duty cycling of various unipolar and bipolar arrangements to create a very large field. The voltages are also much higher, but offset somewhat by using alternating microsecond and nanosecond pulse widths to mitigate muscular stimulation. This does generate a high distal electrode temperature, but it is buried from exposure to the tissue and irrigated to mitigate this effect.
Early Stages of Ventricular Applications of Pulsed Field Ablation
In 2019, the first study investigating the use of PFA in the ventricle was published in an animal model demonstrating that endocardial focal PFA application in the ventricle is both safe (preserving arterioles and nerves while not inducing arrhythmias) and feasible (producing homogeneous fibrosis with a mean [±SD] depth of 6.5±1.7 mm). Subsequent preclinical studies have confirmed these findings and highlighted additional advantages of PFA over RF ablation.62,64-68
When RF is applied to scar tissue, the heterogeneity of the substrate leads to variable effects, where parameters such as power, duration and contact force have a relatively predictable effect on lesion formation. Histological analysis of RF lesions in scar tissue has revealed irregular tissue injury and an unpredictable effect on surviving cardiomyocytes.73 In contrast, studies assessing electroporation in scar tissue suggest that this novel energy modality can generate homogeneous lesions by penetrating viable myocytes interspersed within the collagen and adipose matrix.74–76
There have been published clinical cases in which PFA has been used to treat idiopathic ventricular arrhythmia with potentially deeper substrates, such as premature ventricular complexes (PVCs) originating from the summit or refractory intramural/epicardial PVCs originating from the mitral annulus region.77-79
Peichl et al. reported a series of 44 patients with ventricular arrhythmias, including PVCs (48%) and scar-related VT in patients with structural heart disease (52%), using 3.5-mm irrigated-tip commercially available catheters connected to the CENTAURI system.80 Regarding PVCs, the acute success rate and sustained suppression at 3 months were 81%, with the advantage of safely delivering PFA in the great cardiac vein in 11 patients without observing vasospasm.81,82 However, for patients with VT and structural heart disease, although the acute non-inducibility rate was favourable (83%), a high recurrence rate (48%) was observed at 116±75 days.83 This suggests that non-inducibility at the end of the procedure may be a less reliable predictor of long-term ablation success compared with RF ablation, raising questions about whether PFA can achieve sufficient myocardial penetration for effective long-term lesion formation.84,85
More recent registries published in 2025, including patients with VT and PVC, such as the first US series with the pentaspline catheter and the European AVAAR Registry using a dual-energy lattice-tip focal catheter, among others, have reported comparable results in terms of safety and efficacy.83,86,87
Assessing Lesion Completeness With Pulsed Field Ablation: An Unresolved Issue
Whether PFA is applied to the atrium or ventricle, one of the key unresolved issues is how to determine whether adequate energy has been delivered or not. After PFA, the amplitude of the local bipolar electrogram decreases significantly compared with after RF ablation, making the electrograms unreliable for assessing electrophysiological endpoints.88 As a result, we are left with instructions provided by the manufacturers stating that if a minimum number of lesions is applied following a particular workflow, we can hope for good results. However, such instructions have proven unreliable. Although initial remapping studies suggested that PFA caused durable lesions in more than 95% of patients (Figure 6 ), the recent MANIFEST REDO study showed that, in the real world, reconnection of PVs occurs in at least 30–40% of patients and is comparable to thermal ablation.13,89–92,93 A recent prospective remapping study comparing PFA and RF in patients with persistent AF showed comparable rates of durable PV isolation at 7 months between energies.94
These findings underscore the need for reliable intraprocedural markers of irreversible electroporation. Unipolar electrogram morphology may offer a promising tool to differentiate truly transmural, irreversibly electroporated tissue from areas of reversible or incomplete ablation. Amorós-Figueras et al. analysed electrogram changes in an animal model 60 minutes after ablation in the epicardial surface and correlated these changes with histological samples.95 Immediately after ablation, a clear change in unipolar electrogram characteristics was observed, specifically a significant decrease in the amplitude of the RS wave, a large elevation in the ST segment and a notable decrease in their maximal slope (dV/dT).95 Interestingly, distinct dynamics between the reversible and irreversible areas were observed, with changes measured after 30 minutes able to predict which areas would progress to stable lesions and which would remain reversible. However, after 30 minutes, electrogram changes became less pronounced.95
That study was further expanded by research conducted in a porcine model by Stublar et al., which decomposed unipolar signals into high- and low-frequency components.96 Changes such as ST elevation are low-frequency changes. This approach enabled the identification of significant differences between the applied PFA dose and electrogram characteristics as early as 3–5 minutes after ablation.96 Specifically, a peak-to-peak reduction in high-frequency electrograms and an increase in low-frequency electrograms were observed within this time frame as the number of pulses and electric field intensity increased.96 These findings suggest that lesion magnitude could potentially be assessed through unipolar electrograms as an alternative to the ‘bipolar blindness’ observed after electroporation.
Other studies have attempted to identify different markers of irreversible lesions (Figure 7 ).69,95–98
The High Frequency Dielectric Sensing Lesion Assessment System (HFDS-LAS) was introduced in 2022.98 The HFDS-LAS is a prototype system consisting of a PFA focal ablation catheter with an electrode redesigned as an antenna sensor. This system is capable of intraoperatively measuring the high-frequency reflection impedance electrical properties of the tissue in contact with the electrode, thus enabling assessment of durable lesion formation. The HFDS-LAS has been successfully tested in both in vitro and in vivo models, confirming the catheter’s ability to generate transmural lesions and the system’s capacity to predict irreversibility.98 Nevertheless, one potential limitation is that, to date, it is system-dependent and cannot be applied to commercially available catheters.
Optical coherence reflectometry, a system already known for its ability to predict lesion formation following RF application, was tested as a lesion marker for PFA.69,99 It was demonstrated that not only can PFA induce acute changes in tissue optical characteristics but that the loss of tissue optical birefringence can also be quantified in real time during the procedure. Specifically, an acute reduction in birefringence of ≥20% was shown to predict chronic lesion formation with a sensitivity of 96% and a specificity of 83%.69 The practical advantage of this system lies in its potential for easy integration into any PFA system. Initial human data suggest that drops in birefringence can predict durable PV isolation lesions.100
Conclusion
PFA is reshaping the field of catheter ablation for AF. Its uniquely favourable safety profile stands out as a true paradigm shift, offering a reproducible, cardiac-preferential approach that minimises the risk of damage to extracardiac structures. Procedural efficiency and promising acute outcomes further support its role as a frontline therapy for AF ablation.
However, major challenges remain. The durability of lesions in real-world settings is inconsistent and dependent on the system used and protocol parameters. Moreover, although the initial data on ventricular applications are encouraging, these remain exploratory and require dedicated trials to define their value and limitations.
In summary, PFA has introduced a transformative advance in the safety of cardiac ablation, but its long-term efficacy and broader applicability still need to be validated. As the field evolves, ongoing technological refinement and well-designed clinical studies will be critical to establish its definitive place across arrhythmia subtypes.
Clinical Perspective
- PFA is a novel ablation technology that causes cardiac tissue cell death via non-thermally mediated irreversible electroporation.
- An understanding of the basic science underlying electroporation is essential for electrophysiologists to make informed decisions.
- Clinical evidence suggests that pulsed field ablation is more efficient and likely safer than thermal ablation, but evidence is lacking to suggest that efficacy is definitively better.
- Novel problems associated with pulsed field ablation must be understood and addressed by advancing the science of pulse and catheter design.