Coronary Artery Spasm During Pulsed Field vs Radiofrequency Catheter Ablation of the Mitral Isthmus

Key Points

Question

During catheter ablation of atrial fibrillation, does vasospasm of the left circumflex artery occur during adjacent ablation at the posterolateral mitral isthmus using either pulsed field or radiofrequency energy?

Findings

In this cohort study of 26 consecutive patients, left circumflex arterial spasm frequently occurred during pulsed field ablation at the mitral isthmus, near-universally when the ablation line was situated superiorly but not when situated inferiorly; most, but not all, instances of spasm were subclinical in nature. Spasm was not observed with radiofrequency ablation.

Meaning

In this study, during atrial fibrillation ablation, coronary vasospasm frequently occurred with adjacent mitral isthmus ablation, albeit usually subclinically.

This cohort study compares the incidence and severity of left circumflex arterial vasospasm between pulsed field ablation and radiofrequency ablation during adjacent ablation along the mitral isthmus.

Abstract

Importance

In treating atrial fibrillation, pulsed field ablation (PFA) is a novel energy modality with comparable efficacy to conventional thermal ablation, such as radiofrequency ablation (RFA), but with the benefit of some preferentiality to myocardial tissue ablation. Studies have demonstrated important safety advantages, including the absence of esophageal injury or pulmonary vein stenosis and only rare phrenic nerve injury. However, there is emerging evidence of coronary artery vasospasm provoked by PFA.

Objective

To compare the incidence and severity of left circumflex arterial vasospasm between PFA and RFA during adjacent ablation along the mitral isthmus.

Design, Setting, and Participants

This prospective cohort study enrolled consecutive adult patients receiving first-ever PFA or RFA of the mitral isthmus during catheter ablation of atrial fibrillation in 2022 with acute follow-up at a single referral European center.

Exposure

A posterolateral mitral isthmus line was created using either a multielectrode pentaspline PFA catheter (endocardial ablation) or a saline-irrigated RFA catheter. Simultaneous diagnostic coronary angiography was performed before, during, and after catheter ablation. Nitroglycerin was planned for spasm persisting beyond 20 minutes or for significant electrocardiographic changes.

Main Outcomes and Measures

The frequency and severity of left circumflex arterial vasospasm was assessed and monitored, as were time to remission and any need for nitroglycerin administration.

Results

Of 26 included patients, 19 (73%) were male, and the mean (SD) age was 65.5 (9.3) years. Patients underwent either PFA (n = 17) or RFA (n = 9) along the mitral isthmus. Coronary spasm was observed in 7 of 17 patients (41.2%) undergoing PFA: in 7 of 9 (77.8%) when the mitral isthmus ablation line was situated superiorly and in 0 of 8 when situated inferiorly. Conversely, coronary spasm did not occur in any of the 9 patients undergoing RFA. Of 5 patients in whom crossover PFA was performed after RFA failed to achieve conduction block, coronary spasm occurred in 3 (60%). Most instances of spasm (9 of 10 [90%]) were subclinical, with 2 (20%) requiring nitroglycerin administration. The median (range) time to resolution of spasm was 5 (5-25) minutes.

Conclusion and Relevance

When creating a mitral isthmus ablation line during catheter ablation of atrial fibrillation, adjacent left circumflex arterial vasospasm frequently occurred with PFA and not RFA but was typically subclinical.

Introduction

Pulsed field ablation (PFA) has engendered substantial interest, as it minimizes damage to periatrial tissue during atrial fibrillation (AF) ablation while retaining clinical effectivness.^{1,2,3,4,5,6,7} However, there is emerging evidence that PFA can cause subtotal coronary arterial spasm during cavotricuspid isthmus (CTI) ablation. Indeed, right coronary spasm occurred in 100% of patients undergoing PFA at the CTI using a multielectrode pentaspline catheter, albeit the spasm was subclinical—no ST-segment elevation, arrhythmias, or ventricular wall motion abnormalities—and attenuated by prophylactic administration of nitroglycerin.⁸ Furthermore, 2 patients undergoing PFA of the CTI developed symptomatic ST-segment elevation responsive to nitroglycerin, in one case degenerating to ventricular fibrillation requiring defibrillation.^6,9

While CTI ablation is typically not technically demanding, with the option of switching to conventional radiofrequency ablation (RFA), it is difficult to achieve bidirectional mitral isthmus conduction block with thermal ablation.¹⁰ We performed systematic coronary angiography during PFA at the posterolateral mitral isthmus to assess the frequency and severity of vasospasm of the adjacent left circumflex artery and to compare this response with RFA.

Methods

Study Participants

This was a prospective cohort study of consecutively enrolled patients undergoing AF ablation. This study was approved by the Ethical Committee at Homolka Hospital, Prague, and all patients provided written informed consent. Patients were planned for a first-ever AF ablation procedure and had characteristics suggesting a benefit to mitral isthmus ablation, ie, history of persistent AF or AF episodes exceeding 24 hours. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Ablation Protocol

As detailed in the eMethods in Supplement 1, under propofol sedation without endotracheal intubation, intracardiac ultrasonography–guided transeptal puncture was performed, followed by pulmonary vein isolation. PFA was performed using a commercially available pentaspline catheter with a biphasic bipolar 2 kV waveform (Farawave; Boston Scientific).^4,6,7 For point-by-point RFA, a saline-irrigated catheter (Thermocool; Biosense-Webster) was guided by electroanatomical mapping (CARTO; Biosense-Webster). pulmonary vein isolation was performed with PFA in 24 patients and RFA in 2 patients.

Mitral Isthmus Ablation

After left coronary artery cannulation with a diagnostic catheter, coronary angiography was performed during ablation along the posterolateral mitral isthmus with either the RFA or PFA catheter (flower configuration). PFA lesions, typically 2 applications per location, were placed either superiorly (1- to 2-o’clock position along the mitral annulus) or inferiorly (3- to 5-o’clock position), but all RFA lesions were placed superiorly (Figure, A and B; eFigure 1 in Supplement 1).

Figure. Left Circumflex Arterial Spasm During Pulsed Field Ablation (PFA) at the Posterolateral Mitral Isthmus.

By fluoroscopy, the PFA catheter is shown in a flower configuration positioned at the posterolateral mitral isthmus, either superiorly (A) or inferiorly (B), in 2 different patients (both right anterior oblique views). The black arrowheads indicate the left circumflex artery and the yellow arrowhead indicates spasm of the proximal left circumflex artery. C, In instances when coronary spasm was observed, the time to spontaneous relief of spasm is shown, aggregated by the intensity of spasm. The blue bars represent patients undergoing initial PFA, while the brown and orange bars represent patients undergoing crossover PFA, ie, those first undergoing radiofrequency ablation of the mitral isthmus without conduction block followed by PFA.

^aIn 1 patient, severe spasm did not spontaneously relieve by 20 minutes, so intravenous nitroglycerin was administered, prompting relief 5 minutes later.

^bIn 1 patient, severe spasm spontaneously resolved in 10 minutes, but because mitral isthmus block required additional pulsed field applications, intracoronary nitroglycerin was prophylactically administered before additional PFA.

Contrast angiography was performed during PFA and RFA mitral isthmus ablation lesions, typically 10 mL of contrast for initial and final injections and 2 to 4 mL for injections after individual ablation lesions.⁸ If vasospasm were observed, hemodynamic, electrocardiographic, and wall motion changes were monitored by intracardiac ultrasonography. Nitroglycerin, 1 mg, administration was planned if spasm did not self-resolve in 20 minutes or for significant ST-segment changes. Spasm was classified as mild (<50% luminal decrease), moderate (50% to 90%), or severe (>90%).⁸

Statistical Analysis

Descriptive data are reported as frequencies and percentages, medians with ranges, or means with SDs, as appropriate. Comparisons of spasm rates were assessed by Fisher exact tests. Significance was set at P < .05, and all P values were 2-tailed. Data management and statistical analyses were performed in Microsoft Excel version 2309 (Microsoft).

Results

Catheter Ablation Procedure

Of 26 included patients, 19 (73%) were male, and the mean (SD) age was 65.5 (9.3) years. (Table 1). The left atria were mildly dilated (mean [SD] size, 45.3 [8.1] mm), reflecting the predominance of persistent AF (18 [69%]).

Table 1. Baseline Patient Characteristics.

Characteristic	No. (%)
	Full patient cohort (N = 26)	PFA cohort (n = 17)	RFA cohort (n = 9)
Age, mean (SD), y	65.5 (9.3)	65.4 (10.6)	65.9 (6.4)
Sex
Female	7 (27)	5 (29)	2 (22)
Male	19 (73)	12 (71)	7 (78)
Body mass index, mean (SD)a	30.3 (7.8)	30.2 (4.9)	30.5 (4.8)
Type of atrial fibrillation
Paroxysmal	8 (31)	4 (24)	4 (44)
Persistent	18 (69)	13 (76)	5 (56)
Duration of atrial fibrillation, mean (SD), mo	39.9 (58.1)	51.0 (69.9)	18.9 (5.8)
CHA2DS2-VASc score, mean (SD)b	2.1 (1.4)	2.1 (1.5)	2.1 (1.3)
Hypertension	18 (73)	11 (65)	8 (89)
Diabetes	6 (23)	3 (18)	3 (33)
Dyslipidemia	15 (58)	9 (53)	6 (67)
Sleep apnea	1 (4)	1 (6)	0
Prior stroke or transient ischemic attack	2 (8)	2 (12)	0
Coronary artery disease	4 (15)	4 (24)	0
Congestive heart failure	3 (12)	1 (6)	2 (22)
Left ventricular ejection fraction, mean (SD), %	62.5 (3.2)	62.3 (3.5)	62.8 (2.6)
Left atrium dimension, mean (SD), mm	45.3 (8.1)	45.5 (9.7)	45.1 (3.9)
Medications
Warfarin	12 (46)	12 (71)	0
Nonwarfarin oral anticoagulant	14 (54)	5 (29)	9 (100)
Antiarrhythmic drugs
Class I	14 (54)	6 (35)	8 (89)
Class II	11 (42)	10 (59)	1 (11)
Class III	8 (31)	4 (18)	4 (44)
Class IV	0	0	0

Abbreviations: PFA, refers to pulsed field ablation; RFA, radiofrequency ablation.

^a Calculated as weight in kilograms divided by height in meters squared.

^b Calculated as congestive heart failure, hypertension, age 75 years and older, diabetes, stroke or transient ischemic attack, vascular disease, age 65 to 74 years, and sex category.

The mean (SD; range) skin-to-skin procedure time was 107.5 (32.6; 54-217) minutes, with a mean (SD; range) fluoroscopy time of 10.3 (3.8; 4.0-17.0) minutes. pulmonary vein isolation was successful in all pulmonary veins in 24 patients with PFA and 2 patients with RFA (eTable in Supplement 1).

Mitral isthmus ablation was performed with PFA or RFA in 17 and 9 consecutive patients, respectively (eFigure 2 in Supplement 1). PFA at the mitral isthmus achieved bidirectional conduction block in all 17 patients (100%), with a mean (SD) of 10.5 (3.9) pulsed field applications per patient. The mitral isthmus line was situated superiorly in 9 of 17 patients (53%), and inferiorly in 8 of 17 patients (47%) (eFigure 2 in Supplement 1).

All RFA lines were situated superiorly along the mitral isthmus. After a mean (SD) of 20.7 (10.4) lesions per patient, bidirectional block was achieved in 4 of 9 patients (44%). Block was not achieved in the remaining 5 patients (55%) despite RFA within the coronary sinus (CS) in 3 patients. Subsequently, crossover PFA was used and achieved block in all 5 patients.

Coronary Artery Effects

Coronary angiography at baseline revealed no significant coronary disease, defined as luminal stenosis or calcification of 70% or more. Patients did not receive nitroglycerin pretreatment.

Coronary Effects During PFA

During PFA of the mitral isthmus, left circumflex artery vasospasm occurred in 7 of 17 patients (41%). Interestingly, spasm occurred in 7 of 9 patients (77.8%) and 0 of 8 patients undergoing mitral isthmus ablation situated superiorly or inferiorly, respectively (P = .002; Table 2). The intensity of spasm was mild in 3 patients (43%), moderate in 2 (29%), and severe in 2 (29%) (Figure, C). While mild or moderate spasm spontaneously resolved, severe spasm prompted nitroglycerin infusion for either alleviation of spasm persisting beyond 20 minutes or for additional PFA at an unblocked mitral isthmus (Figure, C).

Table 2. Coronary Arterial Spasm Details.

Outcome	Patients, No./total No. (%)
Spasm incidence during initial PFA	7/17 (41.2)
Superior mitral isthmus line	7/9 (77.8)
Mild spasm	3/7 (42.9)
Moderate spasm	2/7 (28.6)
Severe spasm	2/7 (28.6)a
Inferior mitral isthmus line	0/8
Spasm incidence during RFA	0/9
Spasm incidence during crossover PFA	3/5 (60)
Mild spasm	2/3 (66.7)
Moderate spasm	0/3
Severe spasm	1/3 (33.3)
Time to resolution of spasm, median (range), min	5 (5-25)

Abbreviations: PFA, pulsed field ablation; RFA, radiofrequency ablation.

^a One of these 2 instances of severe spasm was associated with ST-segment elevation.

Coronary Effects During RFA and Crossover PFA

None of the 9 patients undergoing RFA developed coronary spasm. Among the 5 patients without conduction block subsequently receiving crossover PFA, vasospasm developed in 3 patients (Table 2). The spasm was mild in 2 patients and severe in 1 patient; spasm resolved spontaneously in 17 minutes in this patient.

Impact and Relief of Coronary Spasm

In total, 10 patients developed coronary spasm during initial (n = 7) or crossover (n = 3) PFA. There were no ventricular wall motion abnormalities or hemodynamic compromise. The median (range) time to spasm relief was 5 (5-25) minutes (Figure, C). There were 3 adverse events; none were related to vasospasm (eResults in Supplement 1).

Discussion

During mitral isthmus ablation, (1) left circumflex artery vasospasm occurred during PFA at the posterolateral mitral isthmus, but only with lesions situated superiorly; (2) RFA at this superior mitral isthmus location did not induce spasm; (3) during crossover PFA after failed RFA, spasm again occurred; and (4) spasm was severe in a minority of patients, with 1 instance of dynamic ST-segment changes without hemodynamic instability. Mitral isthmus ablation is often considered to fortify against perimitral flutters, but its utility is tempered by the technical difficulty in achieving durable bidirectional conduction block with RFA.¹⁰ This difficulty appears related to the epicardial cooling effect of the blood in the CS, the complex topography with associated annular movement, and the involvement of epicardial tissue sleeves. Additional strategies may be necessary for bidirectional mitral isthmus block, including ablation from within the CS itself, use of a CS balloon to displace blood during endocardial left atrial ablation, and infusion of alcohol into the vein of Marshall.¹⁰ This challenge is further compounded by the fact that incomplete linear ablation can be proarrhythmic by creating conduits of slow conduction, thereby potentiating macro-reentrant perimitral flutter.

The pentaspline PFA catheter studied herein was designed for pulmonary vein isolation, but after regulatory approval, it has been used for mitral isthmus ablation.⁶ Most cases of PFA-related spasm, even severe spasm, are subclinical or, when clinically evident, cause only transient ST-segment changes.⁸ Spasm was not observed during PFA inferiorly along the mitral isthmus. It is possible that at this location, the petals of the PFA catheter may not be oriented sufficiently parallel to the tissue for the energy field to affect the left circumflex artery. However, to our knowledge, there is 1 reported case of left circumflex artery spasm—manifesting as inferior ST-segment elevation requiring nitroglycerin—occurring during PFA inferiorly at the mitral isthmus.¹¹ Interestingly, spasm in this case was likely accentuated by the fact that the left circumflex supplied collaterals to the right coronary arterial territory, as this vessel had in-stent stenosis. Thus, while vasospasm is not frequent with PFA situated inferiorly, it certainly is possible.

While not studied here, we previously demonstrated that nitroglycerin pretreatment during PFA at the CTI eliminated severe vasospasm.⁸ This is consistent with spasm resulting from temporary membrane permeabilization of smooth muscle, local release of calcium or inflammatory mediators, or stimulation of ganglionated plexi causing autonomic imbalance.¹² Additional studies should assess whether routine prophylactic nitroglycerin pre-PFA is superior to expectant administration only when vasospasm occurs.

Significant (50% to 84%) left circumflex spasm was previously observed in 5 of 53 patients (9%) undergoing RFA of the mitral isthmus.¹³ In our study, RFA induced no discernable spasm, perhaps because of less frequent application of RF energy directly within the CS. This suggests that vasospasm during mitral isthmus ablation may be less related to the energy modality but rather a function of the energy field’s proximity to the vessel.

Limitations

This study has limitations. This study was nonrandomized; however, patients were treated consecutively—first PFA, then RFA. Furthermore, crossover PFA caused spasm in 3 of 5 patients after RFA induced no spasm. Also, left circumflex spasm only occurred during mitral isthmus ablation superiorly and not at all inferiorly, again highlighting the improbability of differential spasm being pure coincidence. Only 1 PFA catheter was studied, so these results may not apply to other catheters with different contact, maneuverability, and stability.

While the observed vasospasm was largely subclinical, overt clinical effects may be more common if there is preexisting coronary artery disease. Similarly, the long-term sequelae of spasm are unknown. In preclinical porcine studies, when PFA was applied immediately adjacent to coronary arteries, tunica media fibrosis and neointimal hyperplasia were observed 4 to 12 weeks later, albeit causing only minimal to mild stenosis.^14,15 Whether chronic coronary injury occurs late after PFA adjacent to coronary arteries remains to be determined.

Conclusions

When creating a mitral isthmus ablation line during catheter ablation of atrial fibrillation, unlike for RFA, PFA caused spasm of the adjacent left circumflex artery, albeit typically without clinical manifestations.

Supplement 1.

eMethods. Ablation Protocol

eResults. Additional Details

eTable. Procedural Characteristics

eFigure 1. PFA Catheter Positioning at the Postero-Lateral Mitral Isthmus

eFigure 2. Patient Flow for Mitral Isthmus Ablation

eReferences.

Supplement 2.

Data Sharing Statement