Author + information
- Received July 31, 2017
- Revision received November 27, 2017
- Accepted November 30, 2017
- Published online February 2, 2018.
- Eran Leshem, MD, MHAa,
- Israel Zilberman, DVMb,
- Cory M. Tschabrunn, PhDa,
- Michael Barkagan, MDa,
- Fernando M. Contreras-Valdes, MDa,
- Assaf Govari, PhDb and
- Elad Anter, MDa,∗ ()
- aHarvard-Thorndike Electrophysiology Institute, Cardiovascular Division, Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts
- bBiosense Webster, Research and Development, Haifa, Israel
- ↵∗Address for correspondence:
Dr. Elad Anter, Harvard-Thorndike Electrophysiology Institute, Beth Israel Deaconess Medical Center, 185 Pilgrim Road, Baker 4, Boston, Massachusetts 02215.
Objectives This study sought to examine the biophysical properties of high-power and short-duration (HP-SD) radiofrequency ablation for pulmonary vein isolation.
Background Pulmonary vein isolation is the cornerstone of atrial fibrillation ablation. However, pulmonary vein reconnection is frequent and is often the result of catheter instability, tissue edema, and a reversible nontransmural injury. We postulated that HP-SD ablation increases lesion-to-lesion uniformity and transmurality.
Methods This study included 20 swine and a novel open-irrigated ablation catheter with a thermocouple system able to record temperature at the catheter-tissue interface (QDOT Micro Catheter). Step 1 compared 3 HP-SD ablation settings: 90 W/4 s, 90 W/6 s, and 70 W/8 s in a thigh muscle preparation. Ablation at 90 W/4 s was identified as the best compromise between lesion size and safety parameters, with no steam-pop or char. In step 2, a total of 174 single ablation applications were performed in the beating heart and resulted in 3 (1.7%) steam-pops, all occurring at catheter-tissue interface temperature ≥85°C. Additional 233 applications at 90 W/4 s and temperature limit of 65°C were applied without steam-pop. Step 3 compared the presence of gaps and lesion transmurality in atrial lines and pulmonary vein isolation between HP-SD (90 W/4 s, T ≤65°C) and standard (25 W/20 s) ablation.
Results HP-SD ablation resulted in 100% contiguous lines with all transmural lesions, whereas standard ablation had linear gaps in 25% and partial thickness lesions in 29%. Ablation with HP-SD produced wider lesions (6.02 ± 0.2 mm vs. 4.43 ± 1.0 mm; p = 0.003) at similar depth (3.58 ± 0.3 mm vs. 3.53 ± 0.6 mm; p = 0.81) and improved lesion-to-lesion uniformity with comparable safety end points.
Conclusions In a preclinical model, HP-SD ablation (90 W/4 s, T ≤65°C) produced an improved lesion-to-lesion uniformity, linear contiguity, and transmurality at a similar safety profile of conventional ablation.
The objective of pulmonary vein isolation (PVI) is to create a transmural, continuous, and permanent cellular damage. The current practice of radiofrequency (RF) ablation with irrigated catheters involves the delivery of moderate power (20 to 40 W) for a relatively long duration (20 to 40 s) at a contract force range of 10 to 20 g. At these conventional parameters, the incidence of pulmonary vein (PV) reconnection remains significant, occurring acutely and 3 months after PVI at a frequency of 22% and 15%, respectively (1,2). Although the mechanisms underlying PV reconnection are not entirely understood, incomplete ablation with partial-thickness and/or reversible injury may be a major contributor (3–5).
Ablation in striated muscle tissue preparation, where parameter and catheter stability are optimized, results in lesions of similar or greater depth than the human left atrium (6–9). This indirectly suggests that reversible or nontransmural injury may be related to other factors, such as catheter instability in a beating heart and tissue edema without permanent cellular damage, limiting the long-term efficacy of ablation.
RF ablation lesions result from thermal injury that occurs in 2 consecutive phases: resistive and conductive. During the resistive phase, electrical current delivered at the catheter-tissue interface leads to immediate heating of the superficial tissue layer (approximately 1 to 2 mm). This resistive heating phase creates a heat source that then extends passively to deeper tissue layers during the conductive phase. Conductive heating is time dependent and the result of the current applied and heat produced in the resistive phase (10). Irreversible myocardial tissue injury with cellular death occurs at temperature ≥50°C, whereas reversible tissue injury often occurs at lower tissue temperatures (Figure 1A).
A potential method to achieve uniform, transmural lesions during PVI is to modify the relationship between the resistive and conductive heating phases, such that the resistive heating phase is increased to deliver immediate heating to the full thickness of the human PV circumference, whereas the conductive heating phase is reduced to limit collateral tissue damage (Figure 1B) (11). To achieve this, a larger current must be delivered for a shorter duration.
However, the biophysical ablation properties of high-power and short-duration (HP-SD) ablation are not known. Specifically: 1) increased electrical current may result in char and/or steam-pop (tissue boiling) formation, and appropriate power parameters have not yet been determined; 2) ablation duration (and power over time) may be different compared with standard ablation, such that the safety range of ablation duration at higher power setting may be smaller; and 3) it is unclear whether HP-SD ablation can create irreversible, transmural heating in atrial tissue.
In this study, we examined the biophysical characteristics of HP-SD RF ablation in a thigh muscle preparation model and the beating heart. In addition, we compared single lesion dimensions, linear ablation continuity, and safety parameters between HP-SD and conventional RF ablation.
Animals and protocol
This prospective study included a total of 20 Yorkshire swine (55 to 70 kg). All experiments were performed under general anesthesia. The Beth Israel Deaconess Medical Center Institutional Animal Care and Use Committee approved the research protocol. Experiments were performed at 3 sites: 1) Beth Israel Deaconess Medical Center (Boston, Massachusetts); 2) Technion Institute of Technology (Haifa, Israel); and 3) CBSET, Inc. (Lexington, Massachusetts).
RF delivery system
To perform high-power ablation with adequate safety, it is imperative to record temperature at the catheter-tissue interface to avoid overheating that can result in char and/or steam-pop formation. We therefore used a novel ablation catheter that incorporates 6 thermocouples symmetrically embedded in the circumference of the tip electrode: 3 distal thermocouples positioned 75 μm from the tip and 3 proximal thermocouples positioned 3 mm proximally (Figure 2) (QDOT Micro Catheter, Biosense Webster, Irvine, California) (12). The symmetrical distribution of the proximal and distal thermocouples is optimized to record temperature at both perpendicular and parallel catheter orientations. Measurement of temperature at the catheter-tissue interface has been traditionally limited by the confounding effect of the cold irrigation fluid during ablation. To record accurate temperature at the catheter-tissue interface, an algorithm was developed, validated, and applied to scale the temperature recorded with these specific thermocouples and actual tissue temperature recorded using fluoroptic thermal probes embedded in tissue (data not shown). In addition, this catheter has an improved irrigation system that includes backward flow toward the proximal electrode, allowing increased irrigation during ablation in a parallel orientation. Ablation has been performed using a proprietary RF generator capable of delivering power up to 100 W with a rapid ramp-up time of ≤0.5 s that also provides real-time temperature feedback every 33 ms (nGEN RF Generator, Biosense Webster).
Experimental study design
This study was performed in 3 steps.
The purpose of this step was to evaluate multiple high-power, duration, and irrigation settings in a thigh muscle tissue preparation to identify an optimal setting that allows maximal power delivery at the shortest time possible, without char or steam-pop formation. Preliminary data (not shown) identified 3 potential power and duration settings that required further assessment: 90 W/4 s, 90 W/6 s, and 70 W/8 s with a maximal temperature limit of 65°C (power is automatically adjusted to limit temperature rise above 65°C) and at an irrigation rate of 8 ml/s starting 2 s before each ablation application. In 3 swine, bilateral thigh muscle preparations were performed as previously described (13). In brief, an incision was made over the thigh muscle, and the skin and connective tissue were elevated to create a chamber overlying the thigh muscle. Arterial blood was extracted (∼400 ml), and unfractionated heparin was added to maintain an activated clotting time of 350 to 400 s. The blood was then circulated at a rate of 200 ml/min and temperature of 37°C within the thigh preparation chamber. The ablation catheter was positioned perpendicular to the tissue at a constant pressure of 10 g against the muscle. After each ablation application, the chamber was evacuated from blood, and the underlying tissue and catheter were carefully examined for presence of char, coagulum, or tissue rupture indicative of tissue boiling. The catheter was then positioned at a new location, allowing adequate separation between ablation lesions. This protocol was repeated in the contralateral thigh muscle. Each lesion was referenced on a map to allow later identification and correlated to the RF ablation parameters stored on the Carto3 mapping system (Biosense Webster, Irvine, California). In each thigh muscle, 2 ablation settings were examined to avoid potential differences related to individual muscle preparation.
The purpose of this step was to evaluate the safety of HP-SD setting of choice as determined in step 1 in the beating heart. In 7 swine, HP-SD ablation applications were delivered in all 4 cardiac chambers at clinically relevant contact force range of 10 to 40 g, and at a variety of catheter angles, from perpendicular to parallel. To evaluate the effect of temperature monitoring, and specifically a maximal temperature limit, HP-SD ablation was performed without temperature limit in 3 animals, and with temperature limit of 65°C in 4 additional animals. The geometry of each chamber was defined using Carto3, and each ablation application was tagged on the map.
The purpose of this step was to examine the HP-SD setting of choice in formation of linear atrial lines, including PVI, and to compare these lines with lines made using standard ablation at 25 W for 20 s using standard catheter (Thermocool SmartTouch SF Catheter, Biosense Webster). In each of the 8 animals, 2 parallel right atrial lines were created (1 with HP-SD and 1 with standard ablation) from the superior to the inferior vena cava alternating between posterolateral and posteroseptal position. Each line consisted of 15 ± 3 applications placed 3 to 4 mm apart. In addition, the superior vena cava (SVC) was isolated with a circumferential line in 6 animals (3 HP-SD/3 standard). Following completion of the right atrial lines and SVC isolation, the left atrium was entered using the transspetal approach with a long steerable sheath (Agilis, St. Jude Medical, Saint Paul, Minnesota). Swine often have small left atrium and small PVs, limiting PVI to the right superior and/or the left inferior PV. We isolated the PVs using HP-SD and standard ablation, alternating between right superior and left inferior PVs. Electrical recording from the PV was performed before and after completion of the circumferential line. Entrance block was defined as elimination of conduction into the PV following ablation, whereas exit block was defined as the inability to capture the atrium when pacing the PV at multiple locations at 10 mA at 2 ms. The presence or absence of steam-pop was documented after each lesion. Steam-pop was defined as a sudden rise in the impedance associated with a temperature drop in the absence of significant catheter movement. The presence of char was assessed after the completion of each individual line by careful inspection of the catheter. Right phrenic nerve injury was assessed before and after each posterolateral line. Phrenic nerve injury was defined as loss of phrenic nerve capture (including transient) that occurred during or after ablation.
Analysis of ablation lesions
For steps 1 and 2, triphenyl tetrazolium chloride was infused 15 min before euthanasia. Because triphenyl tetrazolium chloride stains metabolically inactive tissue, it allows better identification of tissue necrosis, improving the accuracy of ablation lesion measurements. The heart and thigh muscles were harvested and fixed in formalin for ≥7 days. After fixation, ablation lesions were identified by indexed serial numbers, and cut through the center of each lesion. Tissue samples were digitally scanned using a high-resolution scanner and measured using electronic calipers at ×6 magnification. Measurements were performed using MatLab (MathWorks, Inc., Natick, Massachusetts) by an investigator blinded to the ablation parameters.
In step 3, following completion of linear lines and PVI, the fresh specimen were examined immediately following euthanasia and harvest of the heart. Presence or absence of visual gaps and transmurality were documented for each line. The specimens were then fixed in formalin, embedded in paraffin, and sectioned for histological analysis using Mason trichrome stain to allow accurate determination of the ablation lesion borders. For each ablation line, the presence or absence of contiguity and transmurality were determined using light microscopy (×10 to 40). In addition, the maximal lesion width and depth were measured. Histological analysis was performed by a pathologist blinded to the method of ablation and results compared independently for posterolateral and posteroseptal lines because of difference in the atrial wall thickness. In addition, in each animal, the right pleura adjacent to the posterolateral ablation line was examined for presence or absence of injury, including hemorrhage.
Biophysical ablation parameters
In all 3 steps, continuous temperature, impedance, and power data were recorded. We also examined the biophysical properties of HP-SD ablation using a thermal modeling of endocardial RF ablation based on the bioheat equation (14). This model allows calculation of current density and temperature distribution in the tissue under a spherical electrode. All models assumed tissue homogeneity, and that the blood perfusion rate is unaffected by the heating process.
Descriptive statistics are reported as mean ± SD for continuous variables. Comparison between HP-SD and standard ablation lesions was performed using the Wilcoxon rank sum test and the Fisher exact test, as appropriate. The different HP-SD protocols were compared using 1-way analysis of variance test with a Bonferroni correction for pairwise comparisons. A p < 0.05 was considered statistically significant. Statistical analyses were performed with Stata/MP version 14 (StataCorp, College Station, Texas).
Evaluation of HP-SD ablation settings in the thigh muscle preparation
In a thigh muscle preparation, 3 HP-SD ablation settings were evaluated: 90 W/4 s, 90 W/6 s, and 70 W/8 s with a similar maximal temperature cutoff limit of 65°C. A total of 84 lesions, 28 at each setting, were performed. Ablation lesion width was similar among 90 W/4 s, 90 W/6 s, and 70 W/8 s (10.36 ± 1.2, 10.57 ± 0.9, and 10.79 ± 1.0 mm, respectively; p = 0.36). Lesion depth was smallest with 90 W/4 s and largest with 70 W/8 s (3.62 ± 0.6, 4.01 ± 0.6, and 4.32 ± 0.6 mm; p < 0.001 for the total comparison). Pairwise comparison revealed significant differences in depth only between ablations at 90 W/4 s and ablations at 70 W/8 s (Figures 3A and 3B).
Maximal temperature was lowest at 90 W/4 s (59.3 ± 5°C vs. 64.3 ± 6°C [90 W/6 s]; p = 0.005; 65.2 ± 5°C [70 W/8 s]; p = 0.001). Impedance change was smallest at 90 W/4 s and largest at 70 W/8 s (28 ± 3, 29.1 ± 4, and 30.8 ± 4 Ω, respectively; p = 0.016). In a pairwise comparison, a significant difference was only present between ablation at 90 W/4 s and 70 W/8 s (p = 0.013) (Figures 3C and 3D).
The ablation data of HP-SD in thigh muscle preparation was compared with ablation using standard RF setting of 25 W for 20 s. Ablation in a thigh muscle preparation at 25 W/20 s resulted in lesion depth and width of 3.74 ± 0.6 and 6.3 ± 0.9 mm, respectively. In comparison with ablation at 25 W/20 s, HP-SD ablation resulted in wider lesions (p < 0.001 for the 3 comparisons) and similar depth (p = NS). The maximal temperature was higher with HP-SD ablation (p < 0.001 for all comparisons with the 3 HP-SD settings). The impedance change was larger with HP-SD ablation (p < 0.001 for all comparisons with the 3 HP-SD settings).
A total of 3 steam-pops occurred (3.6%) in the 84 ablation lesions, 2 at 90 W/6 s and 1 at 70 W/8 s. Char was present following 8 RF applications (9.5%): 5 at 90 W/6 s, and 3 occurrences at 70 W/8 s (none were associated with steam-pops). Ablation at 90 W/4 s did not result in either steam-pop of char formation (p = 0.013 for comparison to other settings). As such, the thigh muscle ablation data identified 90 W/4 s as the optimal compromise between lesion size and safety parameters. This setting was further evaluated in beating heart experiments.
Single ablation lesion in the beating heart
A total of 233 isolated ablation lesions at 90 W/4 s with a maximal temperature cutoff limit of 65°C were applied in the atrium (153 lesions) and ventricle (80 lesions) at a wide range of tissue contact force between 5 and 40 g, and at various catheter orientations to assess safety parameters and lesion dimensions. There were no occurrences of steam-pop or char. Atrial lesion depth was 3.39 ± 0.8 mm (range, 2.3 to 5.0 mm) with all lesions being transmural. Atrial lesion width was 7.0 ± 1.1 mm (range, 5.6 to 8.8 mm). Ablation lesions in the ventricles had similar dimensions to atrial lesions, however demonstrated a greater variance (depth, 3.82 ± 1.1 mm [range, 2.0 to 5.6 mm]; width, 6.92 ± 1.0 [range, 5.4 to 9.0 mm]). Ablation parameters including power, temperature, and change in impedance are shown in Figure 4.
To evaluate the safety value of temperature measurement and a cutoff limit at 65°C, 174 additional RF applications were performed in the atrium (114 lesions) and ventricle (60 lesions) at a similar ablation setting of 90 W/4 s, however without temperature cutoff (Online Figure 1). Of the 174 ablation applications, there were 3 (1.7%) occurrences of steam-pops: 2 in the right atrium and 1 in the right ventricle. In all 3 events, the maximal temperature recorded was ≥85°C (Online Table 1). Of note, these 3 events also coincided with fluctuations in tissue contact force, with contact force that exceeded 40 g.
In comparison with ablation at 90 W/4 s with temperature limit of 65°C, ablation without a temperature limit resulted in an increased maximal temperature and a higher average power delivered (Table 1).
Right atrial lines
In 8 swine, 2 parallel ablation lines were performed in the posterolateral and posteroseptal right atrium with HP-SD (90 W/4 s [T≤65°C]) and standard ablation (25 W/20 s). A total of 4 lines at each position were performed using each ablation strategy. Macroscopic examination of the gross pathological specimen showed consistent differences between lines made with HP-SD and standard ablation. Lines made with standard ablation were composed of distinct and separate lesions as compared with contiguous lines made with HP-SD. In addition, visual gaps were identified in 2 of 8 lines made with standard ablation, whereas no visual gaps were identified in lines made with HP-SD ablation. Figure 5 shows a gross pathological specimen comparing RA lines made with HP-SD and standard ablation. Online Video 1 shows an example of linear ablation (Figure 5) using HP-SD with real-time temperature feedback information.
Histopathologic analysis was performed to accurately determine linear contiguity, presence of gaps, and for measurement of lesion dimensions. Comparative analysis was performed separately for the posterolateral and the posteroseptal lines because of inherent differences in atrial wall thickness. In the posterolateral lines, HP-SD ablation resulted in all contiguous lines, composed of full-thickness lesions that extend from the endocardium to the epicardial fat. In comparison, lines made with standard ablation had lesion-to-lesion gaps and evidence of partial thickness injury (1 line with gaps between lesions, and 3 lines with partial thickness lesions). Figures 5D and 5E show a representative example of microscopic comparison between ablation lines made with HP-SD and standard settings demonstrating such a gap. Lesions width was ∼50% greater with HP-SD (6.16 ± 0.4 mm vs. 4.09 ± 0.8 mm; p < 0.001). Lesion depth was also greater with HP-SD ablation (1.81 ± 0.1 mm vs. 1.61 ± 0.3 mm; p = 0.036).
Posteroseptal lines made with HP-SD ablation were all contiguous and without gaps between lesions. In comparison, lines made with standard ablation had gaps within the line in 1 occasion and partial thickness in another. Lesion width was greater with HP-SD ablation (6.02 ± 0.2 mm vs. 4.43 ± 1.0 mm; p = 0.003); however, depth was similar between ablation methods (3.58 ± 0.3 mm vs. 3.53 ± 0.6 mm; p = 0.81). Standard ablation lesions had significantly higher variance in width and depth compared with HP-SD ablation (p < 0.05 for all comparisons). The comparative biophysical ablation parameters (temperature, power, and impedance) of these ablation methods are presented in Table 2.
There was no difference in the distance between the ablation applications between standard and HP-SD ablation. The interablation application distance with standard ablation was 3.97 ± 0.1 mm as compared with 4.11 ± 0.1 mm with HP-SD ablation (p = 0.40).
The architecture of the ablation lesions was different between HP-SD and standard ablation, such that with HP-SD ablation the maximal lesion width was measured on the endocardial surface, whereas with standard ablation maximal lesion width was subendocardial (Figure 6). The differences in the zone of maximal width may be partially explained by the relatively increased ratio between irrigation and power with standard ablation, resulting in more effective cooling of the catheter-tissue interface, a phenomenon referred as “subendocardial sparing.”
To further evaluate this observation, we used thermal modeling to compute the temperature distribution of endocardial RF ablation in the atrium with standard and HP-SD setting. These experiments demonstrated that the zone of maximal heating is indeed different between standard and HP-SD ablation, such that in HP-SD ablation the zone of maximal heating includes the endocardium while it is protected with standard ablation (Figure 7).
Three SVC isolations were performed with each ablation setting, and all resulted in electrical isolation with presence of bidirectional block. The number of ablation lesions in each circumferential lesion set was similar at 19 ± 3 versus 18 ± 3 (p = 0.68) for HP-SD and standard ablation, respectively. Evaluation of the gross pathological specimens showed consistent differences between lines made with HP-SD and standard ablation, similar to the right atrial lines. Lines made with standard ablation were composed of distinct and separate lesions as compared with contiguous lines made with HP-SD. In histopathologic analysis, SVC lines made with standard ablation demonstrated areas of full-thickness injury; however, also zones with gaps and partial-thickness injury (Figure 8).
Pulmonary vein isolation
A total of 5 PVs were ablated with standard ablation, and 4 PVs with HP-SD. Acute isolation (presence of entrance and exit block) was achieved in all cases after completion of the initial circular line. The comparative biophysical ablation parameters (temperature, power, and impedance) are presented in Table 3. HP-SD ablation resulted in higher power, average and maximal temperature, and larger impedance drop. Evaluation of the gross pathology was consistent with the findings described for right atrial lines. The standard ablation lesions had evidence of lesion-to-lesion gaps and partial tissue injury as compared with contiguous and full-thickness lines made with HP-SD ablation.
Safety parameters of HP-SD ablation
In all linear sets, including SVC isolation and PVI, there was no catheter char or cardiac perforation. A single lesion during PVI resulted in a steam-pop; however, without pericardial effusion. Examination of the tissue showed no evidence of tissue rupture. Review of this ablation application showed that tissue contact force was above the pre-determined force threshold at 42 g.
Temporary phrenic nerve injury occurred in 2 of the 4 posterolateral lines made with standard ablation and none with the HP-SD settings. Right lung hematoma and/or hemorrhage was also more prevalent with standard ablation, occurring in 2 of the 4 posterolateral lines compared with 1 of 4 posterolateral lines made with HP-SD ablation.
Durable PVI is the cornerstone of atrial fibrillation ablation. Nevertheless, it remains a challenge despite improved technologies allowing accurate catheter positioning, real-time contact force, and impedance feedback. This may be related to catheter instability in a constantly moving heart, and tissue edema limiting effective ablation. Although this can be partially ameliorated by highly skilled operators, it remains a general limitation of RF ablation for atrial fibrillation.
We postulate that delivering RF energy at high-power for a short-duration may overcome these challenges, limit the negative inherent effects of catheter instability resulting in tissue edema, and potentially improve lesion-to-lesion consistency. This approach modifies the relationship between resistive and conductive heating, allowing immediate heating and permanent tissue injury while limiting conductive heating and potential damage to neighboring structures. However, a prerequisite for the clinical use of this approach is the ability to consistently extend the resistive heating phase to a full-thickness PV circumference.
We examined the biophysical characteristics and safety profile of HP-SD RF ablation in a thigh muscle preparation and beating heart to evaluate this methodology for atrial ablation, including PVI. Ablation at 90 W/4 s with a temperature limit of 65°C is feasible, effective, and safe. Ablation at this setting resulted in contiguous ablation lines composed of transmural and irreversible lesions. Ablation at 90 W/4 s with a temperature limit of 65°C was superior to standard ablation at 25 W/20 s in the atrium. It reduced the frequency of linear gaps and nontransmural injury, increased lesion-to-lesion consistency, and exhibited a comparable safety profile to standard ablation. High-power RF ablation had a narrow efficacy to safety window, such that smaller changes in power and/or duration may result in char or boiling (steam-pop), and real-time temperature feedback is important for limiting power delivery and to prevent overheating.
Biophysical properties of HP-SD ablation
In this study, we modified the traditional relationship between resistive and conductive heating used for conventional RF ablation, delivering immediate and lethal heating to a similar tissue depth affected by conventional ablation in the atria. The results from computer thermal modeling were confirmed by experimental data in animals, including histological analysis.
It is important to note that HP-SD ablation is particularly suited for thin-walled structures, such as the atrium, and particularly the PV circumference. However, it may not be suitable for thicker tissue, either in the atria, such as the mitral annulus (i.e., for mitral lines), or for ablation in the ventricle. HP-SD ablation is largely based on immediate heat formation during the resistive phase, affecting a tissue depth of ≤3.5 to 4 mm using the energy settings evaluated in this study. Studies of human cadavers have shown that the thickness of the left atrium including the PV circumference is consistently ≤4 mm (6,15). In particular, the ridge separating the left PV from the left atrial appendage reaches its maximal thickness at its superior aspect, and on average measures 2.8 ± 1.1 mm (7). Another study measured the left atrium thickness using computed tomography in patients with persistent atrial fibrillation, and a presumably thicker and scarred atrium. The average atrial wall thickness was 1.89 ± 0.5 mm and never exceeded 3.5 mm (8). Thus, these cumulative data suggest that left atrial wall thickness is well within the ablation depth achieved in our study using HP-SD (90 W/4 s) ablation.
High-power RF energy from a small electrode produces high current that rapidly heats the tissue immediately under the catheter to an average temperature of 63.2°C ± 75°C (Table 2). In atrial tissue, this is often sufficient to heat the full wall thickness to a physiologically lethal temperature ≥50°C. Because the slope of temperature rise is very rapid, mechanisms to prevent tissue boiling must be used with this ablation strategy. In this study, we used a novel catheter design that is based on a Thermocool SmartTouch SF catheter; however, adding a sophisticated thermocouple system sensitive to detect small changes in temperature at any catheter orientation. The system also includes an algorithm to scale these recorded temperatures to actual tissue temperature.
HP-SD ablation at 90 W/4 s with a temperature limit of 65°C resulted in wider ablation lesions and lines, lesion depth similar to maximal depth achieved with standard ablation, and an overall improved lesion-to-lesion consistently. This was achieved with comparable safety parameter because of a closed loop system that provided real-time temperature measurements with a hard cutoff that forced automated adjustment of power. The importance of temperature cutoff was confirmed by an increased incidence of steam-pops with ablation at similar settings but without temperature limit. This highlights the importance of temperature measurement and automated power regulation in high-power ablation. In addition, even with a temperature limit cutoff, intermittent high force (≥40 g) resulted in a steam-pop. This narrow safety margin of high-power ablation reinforces the need for strict ablation parameters that are governed by an automated feedback system. Under these settings, HP-SD ablation may improve the results, consistency, and durability of PVI. Catheter stability is enhanced when the time span of ablation is greatly reduced (by a factor of 5, from 20 to 4 seconds). Additionally, it significantly reduces the RF ablation time to a total ablation time ≤7 min assuming ∼35 applications per ipsilateral pairs of pulmonary veins.
Several prior studies evaluated the utility of higher power and shorter duration ablation for PVI. Bhaskaran et al. (11) examined several combinations of high-power ablation from 50 to 80 W for 5-s duration with a temperature limit of 60°C in an in vitro and a sheep model using Tacticath catheter (Endosense SA, Geneva, Switzerland). They examined the safety and efficacy of these settings in comparison with ablation at 40 W for 30 s in the right atrium. They found that all of the HP-SD ablation settings resulted in 100% transmural lesions. In addition, all HP-SD ablation settings resulted in reduced collateral damage compared with ablation at 40 W for 30 s. Ablation at power ≥70 W were associated with increased incidence of steam-pop (9.5%; 2/21 occurrences) compared with ablation at power ≤60 W (0/32). Overall, the findings of this study are consistent with ours: 1) HP-SD duration ablation consistently results in full-thickness atrial ablation; 2) shorter ablation lesions seem to limit damage to neighboring structures; and 3) high-power ablation increased the risk for tissue boiling and steam-pops. In our study, the incidence of steam-pops was significantly smaller and similar to standard ablation, likely caused by the different catheter design, advanced thermocouple system, and a closed feedback loop governing power delivery. This again emphasizes the importance of an entire automated system designed to allow safe delivery of high-power ablation.
More recently, Liu et al. (16) reported an observational comparison between standard ablation (20 to 40 W/20 to 40 s) and high-power short-duration ablation (50 W/5 to 7 s) for patients undergoing first PVI. They found that high-power short-duration ablation was as effective as standard ablation with a similar safety profile.
As with any pre-clinical experiments, biophysical properties in animals may be different from human. This is particularly important given a smaller safety window with HP-SD ablation, and deserves special consideration in translational of this technology to human. In addition, this study used a novel technology (catheter and generator), limiting its immediate applicability to human. However, high-power ablation using this and other technologies is expected to soon be evaluated in human.
HP-SD ablation in the atrium results in wider lesions that are consistently transmural. This ablation paradigm modulates the relationship between resistive and conductive heating, allowing larger depth of immediate heating and reduced passive heating. It is particularly attractive for PVI, and can potentially improve its consistency and durability. Delivery of high-power ablation requires safety guards to prevent overheating. It requires new generation of catheters capable of accurate temperature measurement, and new RF generators capable of delivering high powers at short duration.
COMPETENCY IN MEDICAL KNOWLEDGE: PVI using standard radiofrequency ablation parameters is associated with frequent recovery of conduction because of gaps between lesions and nontransmural injury. The use of high-power and short-duration ablation has gained clinical popularity over the past few years to improve lesion-to-lesion consistency and wall transmurality. However, its biophysical properties, including efficacy and safety, have not been studied. This study examined multiple high-power ablation settings in a thigh muscle preparation and the beating heart to determine safety and efficacy of this approach for PVI, and compared it with standard ablation. High-power and short-duration ablation demonstrated improved efficacy and similar safety profile as compared with standard ablation.
TRANSLATIONAL OUTLOOK: This study examined the biophysical properties of ablation in animal models of muscle tissue preparation, and the beating heart. Ablation at high-power for short duration results in increased resistive heating and reduced conductive heating. This is particularly beneficial for ablation in the thin atrial walls. It demonstrated improved ablation efficacy at a similar safety profile. The findings of this study will serve as the basis for a clinical study evaluating the safety and efficacy of high-power and short-duration ablation for patients with atrial fibrillation.
Supported in part by an investigator-initiated grant from Biosense Webster and a National Institutes of Health grant (1R21HL127650-01). Dr. Leshem is a recipient of a National Institutes of Health training grant (5T32HL007374-37). Dr. Tschabrunn has received research grants from Biosense Webster. Dr. Anter has received research grants from Biosense Webster and Boston Scientific. Drs. Zilberman and Govari are employees of Biosense Webster. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- high power and short duration
- pulmonary vein
- pulmonary vein isolation
- superior vena cava
- Received July 31, 2017.
- Revision received November 27, 2017.
- Accepted November 30, 2017.
- 2018 American College of Cardiology Foundation
- Ullah W.,
- McLean A.,
- Tayebjee M.H.,
- et al.
- Bhaskaran A.,
- Chik W.,
- Pouliopoulos J.,
- et al.
- Leshem E.,
- Tschabrunn C.M.,
- Jang J.,
- et al.
- Leshem E.,
- Tschabrunn C.M.,
- Contreras-Valdes F.M.,
- Zilberman I.,
- Anter E.
- Sanchez-Quintana D.,
- Cabrera J.A.,
- Climent V.,
- Farre J.,
- Mendonca M.C.,
- Ho S.Y.
- Liu E.,
- Shaw E.,
- Belden W.,
- Thosani A.J.