Effects of Contact Force on Lesion Size During Pulsed Field Catheter Ablation: Histochemical Characterization of Ventricular Lesion Boundaries

A 7Fr 3.5-mm irrigated-tip electrode ablation catheter with an optical CF sensor (TactiCath SE, Abbott Inc, Plymouth, MN) was used for this study.

A PFA generator system (CENTAURI, Galvanize Therapeutics Inc, Redwood City, CA) was employed to deliver a biphasic, monopolar electric current between the ablation electrode and a skin patch. PFA pulse bursts were delivered in synchronization with the QRS complexes to avoid the vulnerable period of the ventricles.

The experimental protocol was approved by the Committee on the Use and Care of Animals at T3-Labs, Atlanta, GA. Five Yorkshire swine weighing 57 to 74 kg were anesthetized with isoflurane and ventilated mechanically. No neuromuscular paralytic agents were administered. A 6Fr multielectrode catheter was inserted into the right jugular vein and positioned into the distal coronary sinus. An 8.5-Fr ultrasound catheter (AcuNav, Acuson, Mountain View, CA) was inserted into the left femoral vein and advanced into the right atrium to be used for intracardiac echocardiography (ICE). Heparin (10 000 IU) was administered intravenously with additional doses, as necessary to maintain the activated clotting time ≥300 seconds. Transeptal puncture was performed under ICE and fluoroscopic guidance. After a transeptal sheath (Agilis, Abbott) was advanced into the left atrium and left ventricle (LV), an anatomic shell of the LV chamber was first created using a multielectrode mapping catheter (Advisor HD-Grid, Abbott) with the EnSite Precision Mapping System (Abbott). The mapping catheter was then switched to the ablation catheter (TactiCath SE, Abbott). The ablation catheter was initially positioned centrally within the left atrium, without endocardial contact (confirmed by ICE), to calibrate the CF-sensor to 0 g (baseline noncontact value) using a CF monitoring system (TactiSys Quartz, Abbott). The ablation catheter was then advanced into the LV for ablation. After LV ablation was completed, the catheter was withdrawn into the right atrium and then advanced into the right ventricle (RV) for ablation, as described below.

Biphasic PFA current was delivered between the ablation electrode and the skin patch positioned on the posterior chest during constant atrial pacing via the coronary sinus catheter at a rate of 90 beats/min. In all 5 swine, PFA current was delivered in synchronization with the QRS complexes at 28 Amp in the RV, whereas higher PFA current at 35 Amp was delivered in the LV with saline irrigation at a rate of 4 mL/min (Figure 1).

During PFA, delivered pulse voltage and current were recorded. The quotient between delivered voltage and current was used to evaluate the impedance between the ablation electrode and a skin patch during PFA.

In each swine, PFA was performed with 4 different levels of CF: (1) low (4 g ≤average CF <15 g); (2) moderate (15 g ≤average CF <30 g); (3) high CF (30 g ≤average CF≤55 g); and (4) no electrode-myocardium contact ≈2 mm away from the endocardium, confirmed by the ICE and CF monitoring (Figures 1 through 3). The presence or absence of microbubbles during PFA current delivery was monitored by ICE. These ablation sites were stored on the EnSite mapping system and were sufficiently far apart to accurately identify during lesion assessment (Figure 2).

Following completion of PFA, a minimum dwell period of 2 hours was initiated to allow the PFA lesions to mature. Fifteen minutes before euthanasia, 250 mL of 1% triphenyl tetrazolium chloride (TTC) was intravenously administered.¹⁸ TTC stains the mitochondrial enzyme succinate dehydrogenase of living cells a red color, distinguishing viable (red) and nonviable (pale) tissue.^19–21 The hearts were excised and fixed in 10% formalin. The hearts were then sectioned at 2 to 3 mm slices perpendicular to the intraventricular groove from apex to the base. All tissue sections were scanned at high resolution (1200 dpi resolution) where each of the lesions was identified using the 3-dimensional electroanatomical maps. For each lesion, maximal depth and diameter were measured using ImageJ²² by investigators blinded to the ablation parameters.

Histological examination of the ablation lesions was performed using hematoxylin and eosin staining and Masson trichrome staining.²³

COX (cytochrome c oxidase) and SDH (succinic dehydrogenase) dual staining (COX/SDH, VitroView Biotech, Rockville, MD) was performed to examine the intensity of mitochondrial activity to delineate reversible and irreversible lesion boundaries in one additional swine (body weight, 64.5 kg). After 2 hours, an ablation lesion section was frozen in the optimal cutting temperature compound and sectioned at 10 µm following the standard procedures to obtain COX/SDH staining.²⁴

Statistical analyses were performed using StatView (Version 5.0, SAS Institute Inc, Cary, NC). Continuous variables are presented as mean±SD for normally distributed variables. Median with range is also presented for non-normally distributed variables. Lesion size was compared between the 3 levels of CF using ANOVA analysis with Fisher PLSD. Comparison of lesion size between low and high PFA intensity was undertaken with ANOVA analysis. Categorical variables are described as absolute values and percentage. The comparison between categorical variables was performed with the χ² test or the Fisher exact test, as indicated. The relationship between average CF and lesion depth and diameter and the relationship between impedance decrease and lesion depth and diameter were assessed using a nonlinear regression model and a random effects model. We compared a nonlinear regression model and a random effects model using a Hausman Test. A probability value of <0.05 was considered to be statistically significant.

After slicing the TTC-stained ventricular myocardium, all of the 80 ablation lesions (32 RV lesions and 48 LV lesions) produced with electrode-tissue contact were clearly identified macroscopically. In contrast, there were no detectable lesions in any of the 15 ablations (8 RV and 7 LV sites) without electrode contact (≈2 mm away from the endocardium).

The ablation lesions were well demarcated with TTC staining and demonstrated 3 distinct zones: (1) a dark central zone; (2) a pale zone; and (3) a hyperstained red zone in 58 of the 80 lesions (72%, Figure 4A). Although the dark central zone and the pale zone were clearly demarcated in all 80 lesions, the outer margin of the hyperstained red zone was not very evident in the remaining 22 lesions (28%) due to overstained normal (nonablated) myocardium by TTC. Microscopic histological examination using hematoxylin and eosin staining of these 3 zones of the ablation lesions demonstrated the following: (1) destruction of myocyte architecture with contraction band necrosis and nuclear pyknosis as well as hemorrhage (red blood cell) and increased interstitial space (edema); (2) relatively preserved myocyte architecture with nuclear pyknosis; and (3) no apparent histological changes (unaffected normal myocardium), respectively. There was no evidence of coagulation necrosis in any of the ablation lesions (Figure 4A).

Microscopic histological examination with Masson trichrome staining demonstrated a central dark blue zone surrounded by a light blue color zone. These 2 zones corresponded to the dark central zone and the pale zone with TTC staining (Figure 4B).

COX/SDH staining of the ablation lesion showed a nonstained central zone surrounded by a hyperstained zone (Figure 4C).

Seventy-one of the 80 lesions (89%) produced with electrode-tissue contact were nontransmural. Lesion depth and diameter were measured by combining the dark central zone and the pale zone with TTC staining as irreversible lesion boundaries (Figure 4A). Lesion size measurements were excluded in the remaining 9 transmural lesions, all of which were observed in the RV (2 lesions with low CF, 4 lesions with moderate CF, and 3 lesions with high CF) because these values would be artificially low.

At the constant PFA current applications (28 Amp applications in the RV and 35 Amp applications in the LV), the lesion depth (combined with the dark central zone and the pale zone) was the smallest with low CF, followed by moderate CF and the greatest with high CF for in both RV and LV (RV, 3.9±0.7 mm, 4.9±0.6 mm, and 5.7±0.8 mm, respectively, P<0.01; and LV, 4.0±1.0 mm, 5.6±1.5 mm, and 6.8±1.0 mm, respectively, P<0.01; Figure 5A and 5B). The depth of the dark central zone was also the smallest with low CF, followed by moderate CF, and the greatest with high CF (RV, 1.8±1.3 mm, 2.5±1.4 mm, and 3.3±0.9 mm, respectively, P<0.05; and LV, 2.6±1.2 mm, 3.5±1.9 mm, and 5.0±1.6 mm, respectively, P<0.01; Figure 5A and 5B). Compared with the lower PFA current applications in the RV, the lesion depth was significantly greater with higher PFA current applications in the LV with high CF (P<0.05), and there was a trend of greater lesion depth with higher PFA current applications for moderate and low CF.

The lesion diameter (combined with the dark central zone and the pale zone) was significantly greater with high CF compared with low CF in the RV (P<0.01), and there was a trend of greater diameter with high CF, compared with low CF ablation in the LV (RV: low CF, 4.7±1.5 mm; moderate CF, 4.7±1.5 mm; and high CF, 5.7±1.7 mm; and LV: low CF, 4.9±1.1 mm; moderate CF, 5.4±0.9 mm; and high CF, 5.9±1.6 mm; Figure 5C and 5D). The diameter of the dark central zone was significantly greater diameter with high CF compared with low CF in the RV (P<0.05), and there was a trend of greater diameter with high CF, compared with low CF ablation in the LV (RV: low CF, 1.7±1.0 mm; moderate CF, 2.6±1.0 mm; and high CF, 3.1±0.8 mm; and LV: 2.5±1.0, 3.1±1.1, and 3.7±1.1 mm; P<0.05; Figure 5C and 5D). There was a trend of greater lesion diameter with higher PFA current applications in the LV, compared with lower PFA current applications in the RV.

For the ratio of diameter to depth (diameter divided by depth), there was a trend of greater ratio with low CF, compared with moderate and high CF in the RV (low CF: ratio of diameter/depth, 1.24±0.43; moderate CF: 0.97±0.35; and high CF: 1.01±0.37, upper in Figure 5E). The ratio was significantly greater with low CF, compared with high CF in the LV (P<0.01, low CF: ratio of diameter/depth, 1.29±0.37; moderate CF: 1.06±0.37; and high CF: 0.88±0.30, lower in Figure 5E).

For the 49 nontransmural lesions with the well-demarcated 3 distinct zones (dark central zone, pale zone, and hyperstained red zone), the depth including all 3 zones was also smallest with low CF, followed by moderate CF and the greatest with high CF for in both RV and LV (RV: 4.6±1.0 mm, 6.4±0.8 mm, and 7.2±0.9 mm, respectively, P<0.01 and LV: 6.2±1.3 mm, 7.2±2.2 mm, and 8.5±1.4 mm, respectively, P<0.01; Figure 6A and 6B). For the diameter (combined with all 3 zones), there was no significant difference between 3 levels of CF in the RV and LV (RV: low CF, 7.8±1.2 mm; moderate CF, 7.5±1.6 mm; and high CF, 9.6±2.6 mm; and LV: low CF, 8.8±2.3 mm; moderate CF, 9.3±1.6 mm; and high CF, 9.3±1.5 mm; Figure 6C and 6D).

For all 71 nontransmural lesions with electrode-tissue contact and the constant PFA current applications (28 Amp applications in the 23 RV sites and 35 Amp applications in the 48 LV sites), the lesion depth (combined with the dark central zone and the pale zone) increased significantly with increasing CF in the RV and LV (R=0.729 and R=0.745; P<0.0001, respectively, Figure 7A and 7B; Table S1). Although the relationship between the lesion diameter and CF in the RV was not significant (NS, R=0.091, P=NS; Figure 7C; Table S1), the lesion diameter significantly increased with increasing CF in the LV (R=0.354, P=0.0137; Figure 7D; Table S1).

The degree of impedance decrease during PFA was very small in both RV and LV (RV: average 2.1±1.0 Ohms and LV: average 2.5±1.5 Ohms). There was no significant relation between lesion depth and impedance decrease in the RV and a weak relationship in the LV (R=0.280, P=NS and R=0.301, P=0.0379, respectively; Figure 8A and 8B; Table S1). There was no significant relation between lesion diameter and impedance decrease in both RV and LV (R=0.118 and R=0.172, P=NS, respectively; Figure 8C and 8D; Table S1).

It has been reported that some types of PFA delivery to myocytes result in cell membrane permeabilization attributed to the formation of nanometric pores in the cell membrane.^{11,12,25–27} Depending upon the PFA intensity applied, the effect of PFA delivery may be transient and associated with the repair of cell membranes and cellular recovery after electric field exposure (reversible injury).^{11,12,25–27} Greater PFA applications lead to irreversible cell death by apoptosis (due to permanent permeabilization leading to cell lysis or disruption of cell homeostasis, including damage of mitochondria and DNA) as well as necrosis (destruction of myocyte architecture), resulting in irreversible injury.^11,12,26,27

In the present study, the acute ablation lesions 2 hours after PFA current applications were well demarcated using TTC staining and demonstrated 3 distinct zones: (1) dark central zone; (2) pale zone, and (3) hyperstained red zone (Figure 4A). The microscopic histological examination using hematoxylin and eosin staining of the dark central zone exhibited destruction of myocyte architecture with contraction band necrosis and nuclear pyknosis, consistent with an acute necrosis zone. The presence of contract band necrosis suggests significantly increased Ca⁺ influx into the myocyte associated with cell membrane rupture, which has been shown to occur in the setting of reperfusion injury in acute myocardial infarction.²⁸ There are red blood cell infiltrations and increased interstitial space, indicating hemorrhage and edema resulting from increased microvascular permeability by PFA.²⁹ Importantly, there is no evidence of coagulation necrosis resulting from thermal injury, which is commonly observed in radiofrequency ablation lesions,^21,30,31 indicating that there is no or little tissue heating associated with PFA current delivery using this ablation system.

Microscopic examination (hematoxylin and eosin staining) of the pale zone with TTC staining exhibited relatively preserved myocyte architecture with nuclear pyknosis. No TTC staining indicates the damage of the mitochondria (no activity of succinate dehydrogenase) and nuclear pyknosis indicates the damage of DNA.”¹¹ These histochemical observations are consistent with an apoptotic pathway.¹¹

The most outer zone of the ablation lesions with TTC staining exhibited the hyperstained red zone. Microscopic examination of this zone showed no apparent morphological changes. Positive staining of TTC indicates preserved mitochondrial activity, suggesting a reversibly affected area.^11,12 Jeong et al utilized TTC staining to delineate reversible and irreversible PFA lesion boundaries in a potato model. Similar to our findings in the present study, they found that: (1) a TTC-unstained pale area delineated the irreversible lesion boundary; and (2) TTC-hyperstained deep red area was formed surrounding the irreversible lesion boundary, indicating the reversibly affected area.²⁰ The TTC-hyperstained deep red area was located peripherally at lower PFA voltage delivery sites (100–200 V/cm), while the TTC-unstained area was located centrally at higher PFA voltage delivery sites (>250 V/cm).²⁰ Based on the mechanism of TTC staining, TTC is consumed by succinate dehydrogenase in the mitochondria, resulting in a red color staining of the cell. Therefore, the color intensity of TTC staining could be related to degrees of the mitochondrial activity. Several studies have shown that PFA delivery results in ATP leakage and preserved or enhanced mitochondrial activity compensates the ATP depletion over 8 to 24 hours after PFA applications,^32,33 consistent with reversible injury. In contrast, histochemical examination of radiofrequency ablation lesions demonstrates coagulation necrosis (ie, thermal injury) surrounded by a thin rim of contraction band necrosis, but no hyperstained red zone.^1,7,18,21 The relatively large areas of hyperstained red zone (presumably reversible injury) with PFA suggest that there are larger areas of stunned myocardium or transient conduction block zone with PFA.¹⁸

COX and SDH dual staining (COX/SDH) have been developed for quantitative analysis of mitochondrial activity, staining cells with normal mitochondrial function in light brown.²⁴ Consistent with TTC staining, COX/SDH staining of the ablation lesion showed no staining of the central zone surrounded by a hyperstained dark brown zone, delineating irreversible and reversible lesion boundaries, respectively (Figure 4C).

Masson trichrome stain can delineate collagen (scar, stained in blue) from normal myocardium (stained in red). Recently, this staining method has also been utilized to detect acute myocardial infarction/ischemic border.²³ Microscopic examination with Masson trichrome staining in the present study demonstrated a central dark blue zone surrounded by a light blue color zone, corresponding to the dark central zone and the pale zone with TTC staining, identifying the irreversible lesion boundary (Figure 4B).¹⁸

To our knowledge, this is the first study to systematically examine the relationship between CF and lesion size using a biphasic monopolar PFA system in a closed chest beating heart model. Using a constant PFA current intensity, lesion depth increased significantly with increasing CF from 2.9 to 7.3 mm in the RV and from 2.5 to 8.4 mm in the LV (CF, 4–55 g, R=0.729 in the RV and R=0.745 in the LV, P<0.0001, respectively, Figures 5 through 7). The slope of nonlinear correlation equation was different between low and high PFA intensity applications (slope, 1.019 versus slope, 1.922), suggesting different degrees of CF contribution in lesion formation (Figure 7A and 7B). Importantly, there were no detectable lesions resulting from PFA without electrode contact (2 mm away from the endocardium), indicating the requirement of electrode-tissue contact for effective lesion formation.

The correlation between the lesion diameter and CF was significant but somewhat lower in the LV (R=0.354) and not significant in the RV (Figure 7C and 7D). Changes in electrode orientation (perpendicular versus parallel) might result in greater variation of lesion diameter with similar CF. It has been demonstrated that a greater pulsed electric field distribution and lesion formation are achieved when the electrodes are positioned parallel to the muscle fiber orientation (longitudinally), compared with perpendicular to the muscle fiber.³⁴ The dependency on the muscle fiber orientation may also result in greater variation of lesion diameter.

PFA lesion size depends on the applied electric field, which is determined by voltage/current, pulse duration, frequency, and polarity.^11–13 These experimental studies on PFA have been primarily performed with stable electrode-tissue contact, including needle electrodes inserted into the tissue.³¹ Recent studies have shown that, using a biphasic, bipolar PFA system in an isolated perfused heart preparation, epicardial shallow lesions are created without electrode-tissue contact (2 mm electrode offset from the epicardium). However, lesion size significantly increased with electrode-tissue contact and increased CF in these studies.^35,36 Epicardial PFA using an isolated perfused heart preparation would be quite different from clinical settings where endocardial ablation is performed with cardiac contraction, respiratory motions, and blood flow. The isolated heart preparation without cardiac contraction may be associated with a more focused field of PFA, leading to some lesion formation without electrode contact.

The effects of CF on lesion size in PFA may be less compared with radiofrequency ablation. However, the dependency of CF on radiofrequency lesion size is variable based on radiofrequency power and application times.³⁷ Therefore, scientifically fair comparison of the effects of CF on lesion size between PFA and radiofrequency ablation would be challenging and the relative importance of CF to lesion size in PFA compared with radiofrequency is still undetermined.

We also found that the mean diameter/depth ratio ranged from 0.88 to 1.29, showing similar dimensions between depth and diameter with PFA (Figure 5E). In contrast, much greater ratios of diameter/depth are associated with radiofrequency ablation (ratio, 1.61–2.22).^21,37 While PFA is a nonthermal ablative modality, radiofrequency ablation lesions are primarily produced by tissue heating (thermal injury). Myocardial fibers run circumferentially in the middle layer of the ventricle and thermal lesions would extend more along the myocardial fiber orientation than crossing the myocardial fiber, resulting in significantly greater diameter than depth for radiofrequency ablation.^38,39

An impedance decrease during radiofrequency ablation usually ranges between 5 and 50 ohms (average ≈25 ohms), which is associated with tissue heating (ie, lower impedance with higher tissue temperature). Several studies have shown significant relationships between impedance decrease and radiofrequency lesion size.^7,8,40 During PFA, tissue impedance may be expected to decrease due to disruption of myocyte cell membrane.⁴¹ However, in our study, there was only a very limited impedance decrease (only <6 ohms decrease), suggesting that no or little tissue temperature increase was induced by the PFA current applications. There was no appreciable correlation between the small impedance decrease measured during PFA and lesion size.

Histochemical examination of PFA lesions demonstrated a relatively large hyperstained zone with TTC and COX/SDH staining, indicating potentially reversible injury. The large reversible zone may be a marker of large areas of transient conduction block. Zyl et al⁴² reported that bipolar PFA applications at the basal interventricular septum resulted in AV block, which recovered 2 to 61 hours after ablation, suggesting a large transient conduction block zone. Several experimental studies also have shown that high voltage applications on myocytes induce transient or permanent depression of excitation, resulting in reduction of resting membrane potential, action potential amplitude, and upstroke (dV/dt).^43–45 Clinical studies on pulmonary vein isolation using multielectrode PFA systems have shown quick elimination of pulmonary vein potentials with a few PFA applications. However, additional PFA applications with different ablation catheter placements and orientation are recommended after complete elimination of pulmonary vein potentials to increase the likelihood of durable lesion formation and avoid the phenomenon of transient conduction block with subsequent reconnection.^46–49 Considering the PFA lesion shape (similar dimensions between diameter and depth), interlesion distance should be carefully adjusted during point-by-point PFA for linear lesion formation.

Additionally, given that no lesions were formed with no contact, stable electrode-tissue contact should be ascertained before a first PFA application to obtain durable lesion formation. Previous studies have shown that radiofrequency lesion depth can be described accurately by a logarithmic function of CF, radiofrequency power, and application time (force-power-time index formula, ablation index).⁵⁰ A new formula incorporating CF and PFA current intensity is desired for accurate prediction of PFA lesion size.

One limitation of this study is that histochemical examination was performed only 2 hours after PFA. It was not an objective of this study to evaluate the temporal evolution of PFA lesions on a longer timescale.

We hypothesized that the TTC or COX/SDH hyperstained zone indicates a preserved or enhanced mitochondrial activity, suggesting a reversible (transient conduction block) area. Further studies are required to examine the relationship between cellular electrophysiology and these histochemical observations.^43–45 TTC staining intensity may not be uniform in some animals, because the outer margin of the TTC-hyperstained red zone was not very evident in 22/80 (28%) lesions due to the overstained normal (nonablated) myocardium in specific animals.

Another limitation of this study is that only a biphasic, monopolar ablation system was tested. Effects of CF on PFA lesion size may be different using different ablation systems, such as multielectrode bipolar PFA or voltage-controlled treatments.

Finally, we note that PFA was performed in the LV and RV and not in the atrium, which allowed accurate measurement of lesion depth (ie, nontransmural lesions were intended).

Table S1