In the modern era of catheter-based arrhythmogenic substrate suppression, thermal modalities have long held the evidence-based guideline recommendations for both atrial [1] and ventricular [2] abnormalities. As such, these temperature-based technologies predominate in clinical electrophysiology labs and operating rooms across the world. The use of cryoballoon ablation for atrial fibrillation (AF) management was found to be non-inferior to radiofrequency (RF) ablation, with a comparable, if not improved, procedural workflow, safety profile, and complication rate [3]. Nonetheless, multiple serious but infrequent adverse events have marked these thermal techniques, including esophageal perforation [4], phrenic nerve injury or palsy [5, 6], pulmonary vein stenosis [7], and increased risk of thromboembolic events [8]. In addition, the durability of thermal ablation for AF has been found to wane rather quickly for single procedures, with approximately 60% of patients back in AF after approximately two years [9].

Irreversible electroporation (IRE), the technique underlying pulsed field ablation (PFA), has broad clinical applications beyond cardiology, particularly within oncology, with DNA/RNA translocation into cells for anti-tumor effects [10], delivering chemotherapeutics into tumor cells [11], and notably ablating solid tumors near nervous structures without permanent damage [12, 13]. Current data on the clinical use of catheter-based PFA for arrhythmia management, though limited, support PFA as a method with increased myocardial tissue specificity, with the main advantage of minimal ancillary damage, and non-inferior short-term success rates [14, 15].

PFA is being considered as an alternative therapy option in the armamentarium of rhythm specialists that could tune to precisely target arrhythmogenic foci in either the atria or ventricles [16] with minimal collateral damage to surrounding structures such as the nervous tissue (example: phrenic nerve) or esophagus, as can be observed with thermal ablation techniques. Despite these exciting applications, there are still safety concerns with PFA procedures, specifically regarding potential neurocardiac [17] disruptions. High-risk mediastinal structures in the cardiac nervous system include efferent axons from the stellate ganglion and sympathetic ganglia, which facilitate sympathetic tone, cardiac-specific axons from the tenth cranial nerve (vagus), which facilitate parasympathetic outputs, as well as cervical and vertebral ganglia axons from the superficial and/or deep cardiac plexi [18, 19, 20]. Equally complex in distribution and number are the afferent neurons responsible for returning information to the central nervous system (example: parasympathetic nodose ganglia), as well as visceral interneurons that contribute to the intrinsic cardiac nervous system. These structures form a highly complex and likely functionally redundant network that allow for precise control of cardiac function via subconscious cortico-brainstem reflexes in conjunction with autonomic outputs. Disruption of this neurocardiovascular system via neuron lysis/necrosis or delayed apoptosis has been documented secondary to thermal cardiac ablation or surgical procedures, however preservation may be feasible with a nonthermal ablation technique that affords sufficient specificity to cardiomyocytes. This succinct review aims to highlight neurological outcomes in preclinical and clinical cardiac ablation procedures, with an emphasis on the epicardial efferent axons and ganglia.

Early in vitro studies determined that clusters of ganglia derived from the intrinsic cardiac autonomic system (ANS), known as ganglionated plexi (GP), were critically involved in the initiation and sustenance of reentrant tachyarrhythmias [21, 22]. Interest has been growing regarding the IRE parameters needed to selectively ablate cardiomyocytes while preserving neuronal function, to potentially reduce adverse ablation events. The IRE threshold of 375 V/cm has been described as the minimum for irreversibly ablating cardiomyocytes [23], though specifically in reference to immature rodent myoblasts, which may not be representative of mature human cardiomyocytes. However, the exact threshold for selective GP ablation remained elusive.

In more recent studies, it was found that increasing the total number of pulses, regardless of field threshold at either 1000 V/cm or 1250 V/cm produced a greater proportion of delayed (24 hours post-electroporation) cardiomyocyte death [24]. Interestingly, at various field strengths ranging from 12.5 to 1250 V/cm, neuronal cell lines were found to be similarly susceptible to cell death when compared to cardiomyocytes [24]. When adjusted for biphasic pulses with longer intra-pulse intervals (increased interphase delay), the extent of cell death, for both cardiac and neural lines, was increased due to a minimization of any cancellation effect [25]. Though these in vitro data suggest GPs could be targeted with cardiac IRE protocols, other studies have suggested that phrenic nerves require exposure to much higher field strengths to facilitate sustained cell death [26, 27]. This may be partly due to neurons having cell bodies proximal to the site of ablation, inherently having larger cell body diameters, and being myelinated, providing electrical insulation.

In the context of GP ablation isolation, GPs are inextricably tied to the central nervous system (CNS) and ANS, which likely potentiate pathologic cardiac fibrosis and tachyarrhythmia susceptibility. This is illustrated in patients who have undergone cardiac allotransplantation and exhibit a low permanent AF incidence, possibly due to cardiac denervation secondary to mediastinal manipulation.

Preclinical investigation with canine models demonstrated that PFA to epicardial GPs effectively reduced AF inducibility post-ablation with no signs of collateral nervous damage [28]. IRE performed on sciatic nerves in adult rats, yielded a pattern of post-IRE myelin disintegration, followed by Wallerian degeneration and finally regeneration of axons with no quantifiable functional deficits seven weeks post-procedure, compared to sham control rodents [29]. Interestingly, the field strength utilized in this rodent study approximated 3800 V/cm, 10-fold above the in vitro cardiomyocyte threshold previously described [23]. IRE performed on rabbit sciatic nerve implicated in a solid tumor revealed similar findings in which the nerve axons disintegrated, and then later had signs of nerve repair with improved functional improvement in the rabbits’ ability to stand with near-normal toe-spreading reflexes [30].

These findings have been replicated in additional rabbit studies [31], as well as large animal models [32, 33]. Identified porcine studies utilized a gradation of IRE field strengths ranging from 1000 to 2500 V/cm (Table 1, Ref. [29, 30, 32, 34, 35, 36]). Under histopathology, the external architecture of the axons was frequently preserved post-IRE, whereas RF produced extensive axonal, extracellular matrix, and collagen denaturation not conducive to regeneration. Although lesions created with IRE did exhibit histopathological signs of nerve damage, the preserved external architecture allowed the regeneration of the axons [32].

Though the number of clinical studies aimed at describing neuronal functional deficits secondary to PFA are increasing [37], not one study designed to detect damage specific to the neurocardiovascular system could be identified. Only one study appropriately powered to provide molecular insights, in addition to companion physiologic data, into the non-cardiac neurologic effects of cardiac PFA was found.

In a small (n = 18) single-arm prospective observational study [34], patients suffering from symptomatic paroxysmal AF underwent pulmonary vein isolation via PFA to assess therapeutic response and residual cardiac nerve function. Using serologic nerve injury biomarkers (neurofilament light chain, glial fibrillary acidic protein, brain-type fatty acid binding protein, S100$\beta$) as surrogates of structural neurologic damage, there were no significant differences pre-versus immediately post-ablation or 24 hours post-ablation. Further, the calculated heart rate variability exemplified high specificity to cardiac tissues without signs of collateral nerve damage in the neighboring structures. This data is corroborated by the lack of phrenic nerve paresis or palsy during PFA pulsing with no primary neurological events after short-term follow up.

In a larger (n = 91) prospective cohort study, S100$\beta$ was found to increase (p 0.001) in the blood twenty-one minutes after pulmonary vein isolation with PFA, however the increase was less than that observed twenty minutes after pulmonary vein isolation with cryoballoon ablation [38]. When S100$\beta$ was normalized to the concomitant high-sensitivity troponin-I, PFA produced a ratio approximately four times smaller than cryoballoon’s ratio, suggesting less relative neurotoxcitiy.

Increases in heart rate post-thermal pulmonary vein isolation for AF have been documented up to 2 months post-ablation in patients [39]. This thermal ablation-heart rate finding was confirmed in two repeat clinical studies [38, 40], and is contrasted with no significant change in heart rate over 1-month post-PFA [34] or 3 months post-PFA [40]. It was postulated that this effect was due to retained GP and/or a lesser degree of cytotoxicity to the local neurocardiac system. Interestingly, in a single study based out of Germany, approximately half of the PFA pulmonary vein isolation patients exhibited transient bradycardia [41], however this was without any serologic evidence of increased neurocardiac damage relative to a thermal study arm.

Catheter-based pulsed field ablation is a promising tool for invasive electrophysiologists to manage arrhythmogenic substrates. With comparable procedural skill requirements, a strong safety profile, and decent antiarrhythmic effect durability, one of the few remaining challenges for widespread PFA adoption is possible neurologic consequences. This review emphasized the effects of PFA on the neurocardiovascular system, but due to the limited data, also included data from somatic and cranial nerve studies as well. The data reveal that PFA is capable of affecting neuronal structures and inducing damage, however the preservation of perineural structures allows for long-term regeneration with minimal innervation deficits. These data are not supported by appropriately rigorous clinical studies that incorporate molecular parameters in addition to physiologic data and histopathology, given that no such studies exist to date.

DC, AA, and IRC all met the International Committee of Medical Journal Editors (ICMJE)-defined requirement for authorship via substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work. In addition, DC contributed to the present review article via manuscript and table drafting, as well as revisions. AA contributed to the present review article via manuscript drafting. IRC contributed to the present review article via manuscript and table drafting, as well as revisions. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

Not applicable.

This research received no external funding.

The authors declare no conflict of interest.