First-in-human trial of atrial fibrillation ablation using real-time tissue optical assessment to predict pulsed field lesion durability

Abstract

Aims

Loss of bipolar electrograms immediately after pulsed field ablation (PFA) makes lesion durability assessment challenging.

Objective

The aim of this trial (NCT 06700226) was to evaluate a novel ablation system that can optically predict lesion durability by detecting structural changes in the tissue during ablation.

Methods and results

Patients with paroxysmal atrial fibrillation underwent pulmonary vein isolation (PVI) using PFA (AblaView®, MedLumics). Using polarization-sensitive optical coherence reflectometry (PS-OCR), reflective characteristics of myocardial tissue and visualization of real-time contrast between healthy tissue and ablated tissue using a drop in tissue birefringence (BiR) was assessed. Wide antral PVI was performed using single point irrigated PFA (unipolar, 1800V, 3 trains, 21 s). Remapping was performed at 3 months. Primary efficacy outcome was the ability of PS-OCR to predict lesion durability at 3-month remapping. Serious adverse events were recorded. Ten patients were included. In total, 38/40 PVs could be isolated with the system. The mean drop of BiR was 17.3 ± 11.5%. Dragging across the ablation lines showed a persistent drop in BiR. During the remap procedures (8/10 patients ablated only with PFA), 12 PVs (37.5%) were found to be electrically reconnected. The mean loss of BiR during all PFA for durable lesions was 20.9%, while only 10.1% BiR loss was observed during the index ablation for reconnected areas (P < 0.001). None of the points with ≥17% loss of birefringence was found to be reconnected.

Conclusion

This first-in-human study supports the use of real-time drop in tissue BiR for lesion assessment and durability during PFA delivery, and its procedural safety.

Graphical Abstract

Open in new tabDownload slide

Introduction

Catheter ablation is the cornerstone treatment for patients with symptomatic atrial fibrillation (AF).¹ For decades, it has been performed using thermal energies, but pulsed field ablation (PFA) is rapidly replacing thermal technology with very promising safety, efficiency, and efficacy.^2–4 The initial excellent results of PFA, showing a very high rate of durable pulmonary vein isolation (PVI), were tempered by the results of real-life multicentre registries.^5–7 Indeed, in patients undergoing a redo procedure for symptomatic AF recurrence after an initial PFA ablation using the pentaspline catheter, PVI durability was only 71–73% on a per-pulmonary vein (PV) basis, and 38–43% on a per-patient basis.^8–10 Such a high reconnection rate is in part explained by the lesion depth typically achieved by current bipolar PFA systems, restricted to 2–4 mm.^4,11,12 Additionally, the disappearance of bipolar signals in tissue is not a reliable indicator of lesion formation with PFA, as even subtherapeutic pulses can lead to signal attenuation due to cellular stunning.^4,12,13 Consequently, confirming the creation of transmural PFA lesions remains challenging, forcing operators to rely on an empiric number of applications in the hope of achieving complete lesion formation.¹⁴

Direct tissue visualization through optical imaging could aid in assessing the depth and durability of lesions created during ablation by leveraging polarization sensitive optical coherence reflectometry (PS-OCR), which assesses the birefringent properties of tissue.¹⁵ PS-OCR has demonstrated its ability to predict both lesion depth and durability in the context of radiofrequency and PFA ablation in animal models.^16–18 However, its ability to monitor lesion formation in humans has never been evaluated.

Thus, this first-in-human study aimed to assess (1) whether real-time optical tissue assessment could predict lesion durability in patients referred for paroxysmal AF ablation using PFA and (2) the safety of the system.

Methods

Trial design

The objective of this prospective, nonrandomized, single-arm trial was to evaluate the safety and efficacy of a novel PFA system integrating a real-time optical assessment of ablation lesions using PS-OCR (AblaView® Unipolar PFA System, MedLumics, Spain) for the treatment of patients with paroxysmal AF (NCT06700226). The steering committee was responsible for design, execution, and study oversight. The sponsor was not involved in adjudicating safety and efficacy events. Data analysis was performed by the sponsor, and data interpretation was provided by the principal investigators. All the authors approved data analyses and interpretation, article contents. The Ethics Committee of the Ministry of Health of the Republic of Uzbekistan approved the study on 5 September 2024, which was performed in accordance with the Declaration of Helsinki—Ethical Principles for Medical Research Involving Human Participants.

Patient selection

Patients with symptomatic paroxysmal AF eligible for AF ablation (first-line or failure/ineligibility of class I or III antiarrhythmic drugs) were treated using the AblaView system. Inclusion and exclusion criteria are listed in Supplemental Materials. All patients provided written informed consent. All patients were included, and all procedures were performed in Ezgu Niyat Hospital, Tashkent, Uzbekistan.

Ablation procedure

Before the ablation, patients were on continuous, uninterrupted oral anticoagulation therapy for at least 3 weeks. Antiarrhythmic drugs were not required to be stopped before ablation, except amiodarone, which was stopped ahead of the procedure. Procedures were performed under general anaesthesia, with the administration of paralytic agents per operator discretion. Venous access was obtained via the femoral vein. A 6F decapolar catheter (Inquiry™, Abbott, Irvine, CA, USA) was placed in the coronary sinus. A single transseptal puncture was performed under fluoroscopic and pressure guidance with use of peri-procedural transesophageal echocardiography if needed. Heparin was infused before transseptal puncture and heparin was administered to keep the activated clotting time at 300–350 s with monitoring every 20–30 min. Electroanatomical maps were created using a 3D mapping system (EnSite Velocity NavX® System, Abbott, Minneapolis, USA) and a multielectrode catheter with a fixed interelectrode spacing (Advisor HD Grid Mapping Catheter Sensor Enabled, Abbott). The ablation strategy was a wide antral circular ablation lesion set around each pair of PVs at least 1 cm from the PV ostia. Carinal lines were also performed if first-pass isolation was not achieved with circular lesions alone.

AblaView® unipolar PFA system

Patients underwent PV isolation (PVI) using the AblaView® Unipolar PFA System integrating real-time optical assessment of ablation lesions. The AblaView® system was developed to measure the birefringent properties of the tissue and is made up of 3 components as shown in Figure 1:

1. AblaView® Unipolar PFA Catheter, hereafter referred to as the ‘Catheter’

2. AblaView® Unipolar PFA Console, hereafter referred to as the ‘Console’ and its ‘Software’

3. AblaView® Unipolar PFA Generator, hereafter referred to as the ‘Generator’.

The Catheter is unipolar, uni-directional, radiopaque, and has an open irrigated-tip. The tip electrode is 8Fr with 4 mm between the proximal and distal electrodes and has 7 optical viewports (6 radial beams at 60° symmetrically distributed around the cap and 1 looking forward beam). The focal catheter was manoeuvred through an 8.5 Fr steerable sheath (Agilis™ NXT; Abbott).

Light emitted from the catheter measures the reflective characteristics of myocardial tissue providing an ultrasound-like structural tissue representation (A scan), which can be used to confirm catheter contact with the tissue. The A scan algorithm evaluates the contact by assessing the distance between the catheter and tissue with an axial resolution of ∼50 µm. The A scan refreshes at a rate of ∼50 ms and acts as a measure of contact. An example of stable contact and one of the unstable contact encountered during the procedure are shown in Figure 2.

The catheter also performs real-time visualization of contrast between healthy and ablated tissue using birefringence (BiR) based on PS-OCR. As tissue is ablated, the natural BiR of the myocardium drops. Pulsed field ablation was delivered through the catheter by the generator (unipolar, 1800V, 3 trains, 21 s per lesion). Stable contact was confirmed using the A-scan. During PFA delivery, real-time drop in BiR was presented to the user in terms of percentage of loss with a target of 18.5% drop, which was based on preclinical porcine data showing durable lesions previously published.¹⁹ However, we did not have any prior human target data.

Therefore, the endpoint of the index procedure was a clinical endpoint of an absence of PV potentials on the multielectrode catheter in each PV ostia (entrance block) at least 30 min after the completion of each set of veins. The completeness of the ablated lines was further confirmed by dragging the catheter along the ablation lines (from the PV ostium into the atrial chamber) to confirm a persisting BiR drop and a lack of gaps across the lines.

Study follow-up

Patients were brought back 3 months after the index procedure for PV remapping to assess durable PV isolation, presence of gaps, and correlation to tissue BiR. The identification of the gaps was based on methodology previously published.²⁰ If reconnection was found, re-isolation could be performed at the discretion of the local physicians with standard, commercially available ablation technology. During the 3 months, patients were monitored with bi-weekly and symptomatic transtelephonic monitoring (AliveCor® Kardia 6L, Mountain View, CA, USA). Follow-up visits were performed at 1-, 2- and 3-months post-ablation including physical examination, 12-lead ECG, and quality-of-life questionnaires [Atrial Fibrillation Effect on Quality-of-Life (AFEQT) and European Quality of Life-5 Dimensions (EQ-5D) questionnaires]. Additionally, 48-h-Holter monitoring was performed at 2 and 3 months. Any documented atrial arrhythmia >30 s was considered a recurrence.

Endpoints

The primary efficacy outcome was the ability of PS-OCR to predict PFA lesion durability and continuity at the 3-month remapping procedure utilizing the AblaView® PFA catheter. The primary safety outcomes were serious procedure- and device-related adverse events. Secondary endpoints included procedural, fluoroscopy and left atrial dwell time, total amount of PFA energy delivered, number of applications delivered, quantification of changes in BiR during ablation, acute PVI, recurrence of arrhythmias controlled by Holter 48 h and by bi-weekly TTM and triggered by symptoms, and quality of life evaluated by AFEQT (Abbott) and EQ-5D (EuroQol Research Foundation, 2021) questionnaires.

Statistical analysis

Normally distributed variables were expressed as means ± SD and compared using Student’s t-test. Non-normally distributed variables were expressed as median and interquartile ranges and compared using Mann–Whitney U test. Categorical variables were expressed as counts and percentages and were compared using the χ² test or exact Fisher test when needed. A P value <0.05 was considered statistically significant. The analysis was performed with the SPSS statistical package, version 11.0 (SPSS Inc., Chicago, IL, USA).

Results

Study population

A total of 10 patients (80% women) were included in the study. Patients were 55 ± 11 years old, with a mean body mass index of 29 ± 4 kg/m² and a CHA₂DS₂VASc score of 2.4 ± 0.8. Eight (80%) patients failed prior antiarrhythmic drug therapy (n = 5 class I/III, n = 3 class II/IV). Left ventricular ejection fraction was 58 ± 3%, and left atrial diameter was 41 ± 2 mm. Full details are described in Table 1.

Procedural characteristics

During AF ablation, a total of 65 ± 24 PFA applications were delivered to obtain complete PVI, representing 652 ± 181 Joules of energy delivered. The increase in temperature measured on the catheter tip thermocouples was 2.9 ± 1.4°C. A total of 38/40 PVs could be isolated using the study catheter. Two PVs could not be isolated in one patient due to generator lock-out due to EP laboratory power supply instability. These PVs were completed with radiofrequency. The mean number of beams in contact with the tissue was 3 ± 1, and the overall drop of birefringence was 17.3 ± 11.5%. Left atrial dwell time was 147 ± 53 min, improving progressively from 180 to 70 min, and total fluoroscopy time 20 ± 10 min. The number of ablation lesions applied decreased by >15 from the first patients to the last.

Only the first two patients required neuromuscular blockage during ablation (Table 2). It was decided to remove the neuromuscular blockers from the second patient onwards due to the absence of muscular contractions detected in the patients and symptom-free recovery. Only one complication was observed acutely unrelated to the Investigational System, which was an embolization of a peeled-off wire fragment during vascular access which required percutaneous retrieval. Free haemoglobin remained stable (0.3 ± 0.2 g/L pre-ablation vs. 0.4 ± 0.1 g/L post-ablation, P = NS), haptoglobin decreased by 26% (125 ± 98 pre-ablation vs. 89 ± 54, P = NS), and no haemolysis was detected on blood samples or on urine samples. All the patients were discharged in sinus rhythm.

Follow-up and remapping data

Patients remained free of complications during the 1-month follow-up after ablation. During follow-up, only one patient (10%) experienced symptomatic AF recurrences, while durable sinus rhythm maintenance was observed in the remaining patients through trans-telephonic monitoring and 48-h-Holter ECGs.

At 3 months, one patient (who had only two out of four PVs isolated with the investigational catheter initially) withdrew consent for remapping. Among the remaining 9 remap procedures, we considered the 8 patients ablated only with PFA, 12 PVs (37.5%) were found to be electrically reconnected. On a per-vein basis, the overall PVI durability rate was 62.5%. The mean loss of BiR during the PFA for durable lesions performed on the 9 patients was 20.9%, while only 10.1% BiR loss was observed during the index ablation for reconnected areas (P < 0.001). Violin plots showing the percentage of birefringence loss are depicted in Figure 3A. Of note, none of the points with ≥17% loss of birefringence was found to be reconnected (specificity: 1.0; sensitivity 0.6; positive predictive value: 1.0; negative predictive value: 0). Electrogram abatement did not differ significantly between durable and reconnected PFA lesions (0.5 ± 0.7 mV vs. 0.4 ± 0.4 mV; Figure 3B). Examples showing the BiR loss for each point of wide antral circumferential ablation are depicted in Figures 4 and 5.

There was no significant statistical difference between the number of ablations performed between groups with all veins isolated and those with some of the veins reconnected, being 65 ± 5 vs. 65 ± 8, respectively.

Safety

No patients demonstrated any symptoms of esophageal or phrenic injury during follow-up. No SAEs, strokes, or deaths related to the PFA system occurred in any patient out to 3 months post-procedure. Full details are provided in the Supplemental Material. Although the generator locked-out because of power supply instability in two patients, this was not considered a serious adverse event since the generator appropriately did not deliver PFA with an unstable power supply.

Quality of life

All patients reported an improvement in their quality of life based on specific questionnaires conducted during follow-up visits. The results of the AFEQT test are shown in the Figure 6.

Discussion

Main results

A novel PFA system utilizing real-time assessment of tissue BiR used in a first-in-human pilot study achieved acute intraprocedural PVI leaving most patients free of AF. There were no serious adverse events attributed to the PFA system during the 3-month follow-up. Importantly, the pilot study demonstrates the ability of the Ablaview® System to differentiate durable from non-durable PFA lesions through loss of BiR, demonstrating a statistically significant difference between reconnected and non-reconnected lesions. A BiR drop of ≥17% seems to differentiate durable from non-durable PFA based on 3-month remapping. This is substantially better than the loss of bipolar electrogram amplitude, which can occur with even sub-therapeutic doses of PFA.^11,21,22 Indeed, in this study, electrogram amplitude abatement did not discriminate between durable and reconnected lesions.

Although we targeted a drop in BiR of at least 18.5% for each lesion during PVI, this was not the endpoint of the ablation procedure. The endpoint of the ablation was not achieving a minimum BiR drop at each lesion, but rather confirmation of PVI after a standard 30-minute wait after each set of veins. The study was specifically designed this way since we could not be sure what the ideal target of BiR drop would be in humans. So, the endpoint was kept as clinical as possible—the standard of PVI. The advantage of the design is that we had a ‘real-world’ data set of complete and incomplete points by which a human target could be determined. The disadvantage of our choice of design is that we probably had more reconnections than if we had targeted a specific BiR drop per ablation delivered. Despite the reconnections seen, 80% of the patients did not have recurrent AF during the first three months. As mentioned earlier, ‘real-world’ results with other PFA systems have shown reconnection rates of 70% per vein and 40% per patient.^2–4 Recent publications have also demonstrated variability in PV reconnection rates of 38–90% with pentaspline PFA undergoing repeat ablation.^23,24

Other than the fact that BiR was not the clinical endpoint of the study, there are some additional explanations why an ideal BiR drop was not achieved at every point. First, the 18.5% target was determined by using highly controlled applications in a porcine preclinical model at the right atrial-superior vena cava junction. Navigating through the human left atrium is clearly more challenging, and we were using a system for the first time in humans, so we may not have been able to get ideal contact and delivery at each point despite guidance from the A scan in the system. Our data showed there was clearly a learning curve for the 10 patients with procedural time and number of ablations dropping with each ablation due to better understanding of the optical guidance. The optimal inter-lesion distance was also not well known. If the catheter is moved too little, the neighbouring tissue BiR may already be low from the prior lesion, preventing much further drop. However, if the catheter is moved to a point of completely normal BiR, there may be a good drop but a small gap between lesions. Further analysis and study of these data will allow us to define optimal inter-lesion distances. However, it seems that a lesions with >17% drop regardless of baseline BiR predicted durable isolation, based on the analysis presented in this paper.

What our data does show is that there is a discriminatory power based on BiR drop that predicts a durable PFA lesion. This addresses an important need since it is well known that standard bipolar electrograms are abated substantially even with subtherapeutic doses of PFA.^11,21,22 Even in this study, we found no clear correlation between bipolar electrogram abatement and durable PVI lesions at 3 months. The reason for this is due to myocardial stunning with application of any PFA, which creates a mixture of reversible and irreversible electroporation.^11,21,22 Although manufacturers of PFA systems provide guidelines as to how many lesions should be applied, the lack of ‘real-time’ assessment is a major limitation. Waiting time and adenosine infusion have also never been shown to be additive to assessment of PFA durability. Recovery of reversibly electroporated tissue can take hours.^21,22,25,26 Some promising data have been presented on acute changes in unipolar electrograms during effective PFA, but this data have yet to show any conclusive clinical use.^11,21,26 Contact force or use of a contact force ‘index’ has also been proposed as a surrogate for completeness of PFA lesions.^27,28 While the importance of contact for complete PFA lesions is a given, the additive utility of contact force beyond 5–10 g is less clear.²⁹ The use of a ‘PFA index’, like the ablation index used for radiofrequency, may offer more than just empiric numbers of applications, but it remains an indirect surrogate for durable lesions. In contrast, measuring loss of tissue BiR provides a direct, real-time evaluation of the tissue and the critical drop in BiR necessary to get a durable lesion is based on actual changes in the tissue structure.

One of the key points of the technology presented in this study is that the changes in BiR are maintained over time. Within the duration of the initial ablation procedure, our dragging tests globally showed that BiR remained low over the ablated lines. We demonstrated persisting drop in myocardial BiR in our preclinical study, but this is the first time we have shown this in humans.

Finally, it should be mentioned that the catheter did successfully deliver a safe and clinically effective PFA pulse. The pulse had previously been assessed in preclinical models¹⁹ but had yet to be tested in humans. When appropriate BiR drop was attained, the lesions were durable. Furthermore, the lesions were delivered with an absence of oesophageal damage, phrenic damage, or coronary vasospasm. There was also no need for gated delivery based on the protocol design. Tests of haemolysis also showed no significant change. Although we successfully delivered a safe and effective PFA recipe, the optical technology can be adapted to any size/shape of catheter and therefore does not need to be associated with this specific catheter/generator system.

Limitations

In addition to the limitations acknowledged above, it must be emphasized that this is a first-in-human, single-centre trial with a very small number of patients. Extrapolation of these data to clinical practice would require further clinical investigation. However, the promising results show that detection of BiR change by PS-OCR can address an important need in the field of PFA. We also included two operators in this study, which may have introduced some procedural variability which could have affected the 3-month outcome. We were also very strict in our assessment of reconnections using methods previously published.²⁰ Voltage map settings were set at an upper limit of 0.5 mV and lower limit of 0.2 mV. However, we did not rely solely on the voltage map to identify gaps and also used the appearance of any near-field electrogram on the mapping catheter to define reconnection. Two operators also separately had to agree upon whether a vein was durably isolated or not. Finally, because of an unanticipated problem in catheter supply we were not able to measure the tissue BiR at the repeat 3-month remapping procedure to see if tissue BiR was still low after 3 months. However, our preclinical data showed that drops in tissue BiR were persistent even weeks after ablation.

Conclusion

In this first-in-human pilot trial, PFA using a novel ablation system guided by optical assessment of tissue BiR achieved acute PVI with no serious adverse events attributed to the PFA technology. Lesions which demonstrated >17% drop of tissue BiR predicted durable lesions during a 3-month remapping procedure.

Supplementary material

Supplementary material is available at Europace online.

Funding

Clinical Investigation sponsored and funding by MedLumics S.L.

Data availability

All data are incorporated into the article and its online Supplementary material.

References

Author notes