Author + information
- Received April 3, 2017
- Revision received August 28, 2017
- Accepted September 6, 2017
- Published online October 16, 2017.
- Saurabh Kumar, BSc(Med)/MBBS, PhDa,b,
- Jorge Romero, MDa,c,
- William G. Stevenson, MDa,
- Lori Foley, BSa,
- Ryan Caulfield, BSa,
- Akira Fujii, MDa,
- Shinichi Tanigawa, MDa,
- Laurence M. Epstein, MDa,
- Bruce A. Koplan, MDa,
- Usha B. Tedrow, MDa,
- Roy M. John, MDa,d and
- Gregory F. Michaud, MDa,d,∗ ()
- aCardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts
- bCardiovascular Division, Westmead Hospital, University of Sydney, Westmead, New South Wales, Australia
- cArrhythmia Services, Department of Medicine (Cardiology), Montefiore-Einstein Center for Heart and Vascular Care, Bronx, New York
- dArrhythmia and Electrophysiology Program, Department of Medicine, Vanderbilt University Medical Center, Nashville, Tennessee
- ↵∗Address for correspondence:
Dr. Gregory F. Michaud, Arrhythmia and Electrophysiology Program, Department of Medicine, Vanderbilt University Medical Center, 1211 Medical Center Drive, Nashville, Tennessee 37232.
Objectives The authors sought to investigate the effect of low irrigation flow rate on lesion characteristics and ablation outcomes in a clinicopathological study.
Background Irrigated ablation produces deeper lesions compared with nonirrigated ablation, which may not be desirable in the thin-walled posterior left atrium (LA), where collateral esophageal injury is possible.
Methods Lesions were placed on the smooth posterior right atrium in 20 swine and posterior LA in 60 patients at a maximum power of 20 to 25 W with either: 1) power-controlled ablation at an irrigation flow rate of 17 ml/min (high-flow group 10 swine; n = 40) or 2) temperature-controlled ablation at an irrigation flow rate of 2 ml/min (low-flow group 10 swine; n = 20). Safety and efficacy was also compared in 326 patients undergoing AF ablation using high-flow (n = 160) or low-flow settings (n = 166) for posterior LA ablation.
Results Low-flow, compared with high-flow, lesions in swine had a higher incidence of lesions with: impedance fall ≥10 Ω, loss of pace capture, electrograms characteristic of transmural lesions, and visible lesions on anatomic inspection (p < 0.05 for all). Low-flow lesions had a maximal diameter at the endocardial surface, whereas high-flow lesions had a maximal diameter at the epicardial surface. In humans, impedance, pace capture, and transmurality data also strongly favored low-flow lesions. There was no difference in acute pulmonary vein isolation, complications, or 12-month arrhythmia-free survival between the groups.
Conclusions Low-flow irrigated ablation provides favorable lesion characteristics for posterior LA ablation without increasing the risk of adverse events.
Radiofrequency (RF) catheter ablation resistively heats tissue to create a thermal injury, which is permanent when tissue temperatures exceed 50°C to 60°C (1). In lesions created with a nonirrigated catheter lying on the endocardium, tissue temperature is highest just beneath the endocardial surface; however, increased power application is limited by creation of coagulum formation over the electrode when temperature exceeds 70°C to 80°C (2). Open irrigated catheters that are in widespread use, use saline to cool the catheter tip–tissue interface, allowing higher power delivery for a longer duration without coagulum formation; this increases lesion size, thereby affording the creation of deeper, transmural lesions (3). Open irrigation also changes lesion geometry such that the maximal lesion width is typically found 4 mm below the endocardial surface, corresponding to the depth of hottest tissue temperatures, and a smaller area of endocardial surface is exposed to tissue temperatures necessary for irreversible tissue destruction, resulting in relative endocardial sparing (3–6). Producing deep lesions with irrigated ablation may not be optimal in thin-walled tissue such as the posterior left atrium (LA) that lies in close proximity to structures vulnerable to collateral injury, such as the esophagus. In the posterior LA, one possible scenario is that irrigated ablation may potentially expose the esophagus to much higher temperatures in order to achieve a sufficient endocardial lesion compared with nonirrigated lesions (Online Figure 1). Esophageal heating is routinely seen with temperature monitoring during posterior LA ablation, frequently limiting ablation. However, the concern with nonirrigated ablation is the risk of coagulum formation and emboli.
We hypothesized that short-duration, low-flow, irrigated radiofrequency ablation of the posterior LA would allow creation of transmural lesions with maximum diameter at the endocardial surface, without compromising safety or efficacy. To evaluate this, we first compared: 1) lesion biophysical parameters and pathology with low-flow (2 ml/min) versus usual, high-flow rates (17 ml/min) during irrigated catheter ablation in 20 swine randomized to either low-flow or high-flow ablation; 2) biophysical parameters in 60 patients undergoing de novo catheter ablation for atrial fibrillation (AF) who had posterior LA ablation was performed in the first 30 consecutive patients with usual high-flow rates (comparator), and the next 30 consecutive patients with low-flow rates (comparator); and 3) reviewed the safety and efficacy of this approach in a total of 326 consecutive patients undergoing AF ablation with the low-flow settings (first 166 patients) compared with patients undergoing ablation with high-flow settings (next 160 patients) for posterior LA ablation. The human component of the study was not randomized.
Animal experimental protocol
The Institutional Ethical Committee for animal research at the Brigham and Women’s Hospital approved the experimental protocol. Studies were performed in 20 juvenile female swine between 30 and 35 kg under anesthesia and mechanical ventilation. The experimental protocol is detailed in the Online Methods and in Online Figure 2A. Briefly, a 3-dimensional mapping system (CARTO 3, Biosense Webster, Diamond Bar, California) was used with right atrial (RA) geometry created using a multielectrode catheter (Pentarray, Biosense Webster), intracardiac echo (ICE), and lesions were made with a 3.5-mm-tip open-irrigated contact force (CF)-sensing catheter (Thermocool SMART-TOUCH, Biosense Webster) delivered via a steerable sheath. Lesions were delivered in the smooth posterior RA at sites that had stable electrogram (EGM) morphology, demonstrable pace capture at an output of 10-mA and 2-ms pulse width with unipolar and bipolar pacing before ablation. The superior, mid-, and low posterior RA, were chosen as the 3 minimum pre-defined sites for RF ablation. A minimum of 15-mm distance was maintained between lesions with a maximum of 6 lesions placed in the RA (Online Figure 2A).
The following characteristics were required for ablation targets: 1) average contact force 18 to 22 g with a stable CF profile and the force vector directed toward the target tissue; 2) stable contact with no sliding or lateral movement detected by ICE; and 3) RF duration of 10 to 15 s.
Each animal was randomized to receive all of its lesions with either: 1) high-flow settings: irrigation flow rate 17 ml/min, power-controlled ablation (20 W, cutoff temperature 50°C); or 2) low-flow settings: irrigation 2 ml/min, temperature-controlled ablation (maximum power 20 W, maximum tip temperature 60°C).
For each ablation site, the following were recorded: unipolar and bipolar EGM amplitude and morphology, pace capture (10 mA/2-ms output) immediately before and after ablation, average contact force, force-time integral (FTI) (force ∙ ablation duration) (7), total energy delivered, average and maximum catheter tip temperature, and impedance decrease at the end of ablation.
Assessment of biophysical parameters and pathology was made with the investigator blinded to group assignment. Post-procedure analysis included: percentage of lesions that met a minimum 10-Ω impedance decrease, percentage reduction in peak-to-peak unipolar and bipolar EGM amplitude, and percentage of lesions meeting unipolar and bipolar EGM-based criteria for transmurality, as described by Otomo et al. (8). Briefly, lesions were considered to have met criteria for transmurality if the negative deflection was eliminated on unipolar EGMs. On bipolar EGMs, lesions were considered transmural lesions if they exhibited elimination of a positive deflection or ≥75% attenuation of the R wave at sites exhibiting QRS pattern pre-ablation or complete elimination of the R′ wave at sites exhibiting RSR′ pattern pre-ablation (8). Details of animal sacrifice and lesion volume assessment are given in the Online Methods.
Human AF ablation
Patients undergoing percutaneous catheter ablation of paroxysmal or persistent AF between September 2014 and September 2016 were included. All patients provided written informed consent for the procedure. The Brigham and Women’s Hospital Subject Protection Committee approved the data analysis.
The approach to AF ablation is described in the Online Methods. Briefly, all procedures were performed under general anesthesia with use of a 3-dimensional mapping system and a Thermocool SMART-TOUCH CF-sensing ablation catheter (Biosense Webster). The esophageal course was identified and tagged by ICE, and then merged onto the LA map (Online Figures 2B and 2C). The lesion set consisted of 2 wide circumferential antral ablation lines around the ipsilateral pulmonary vein (PV) pairs with the endpoint of inexcitability to pacing at 10 mA, 2-ms pulse width along the ablation line, demonstrable entrance and exit block, and lack of PV reconnection after a minimum waiting time of 30 min with or without a rapid bolus of 12 to 18 mg of adenosine according to operator preference (9,10). Lesions were applied in a point-by-point, contiguous fashion with lesion markers set at 2 mm in diameter. LA ablation was performed with a power of 30 W with an irrigation rate of 17 ml/min for 30 to 60 s for anterior sites. At the posterior LA, power was limited to 20 to 25 W for 10 to 30 s. CF targets were a minimum of 10 g; for posterior LA ablation, a CF range of 16 to 22 g was sought to maintain consistency with the swine experiments.
Human AF ablation: Assessment of biophysical parameters
Two approaches to posterior LA ablation were compared. The first consecutive group of patients had an irrigation flow rate of 17 ml/min (high-flow group) during posterior LA ablation. In the next consecutive group of patients, we attempted to avoid deep RF lesions in the posterior LA, potentially near the esophagus. In the second consecutive group of patients, rather than using a second 4-mm solid-tip nonirrigated electrode catheter when ablating over the posterior LA, the irrigated catheter was used with the irrigation rate maintained at the baseline rate of 2 ml/min and 20 to 25 W of power applied for 10 to 20 s according to the impedance fall. In both groups of patients, RF was terminated immediately if the esophageal temperature increased by >0.1°C. Further lesions in that area were avoided until esophageal temperature returned to baseline. Power settings were not changed during RF delivery. There were no crossovers from the high-flow to the low-flow settings, or vice versa, in any patient during posterior LA ablation. Ablation settings in both arms were identical to that described in the swine experiments, with the exception that up to 25 W was permitted. For nonposterior LA ablation, power was set to 30 to 35 W and performed with the usual, high-flow settings.
As previously published, given that chronic PV conduction recovery was strongly associated with poor impedance fall during the index procedure (11), RF applications were predominantly guided by impedance decrease (Online Figures 2B and 2C). RF applications were aborted if a pre-specified decrease <5 Ω did not occur in the first 10 s (Online Figure 2B) (considered low-quality RF lesions) (9,10). In such cases, the steerable sheath and ablation catheter were repositioned in order to improve catheter contact. If impedance continued to fall >5 Ω in the first 10 s, ablation was continued for a total duration of 30 to 60 s for the anterior wall, roof, and septum, and up to 10 to 30 s for the posterior wall (considered high-quality lesions) (Online Figure 2B) (9,10). Sites achieving a ≥10-Ω fall were marked (red tags for irrigated, blue tags for low-flow irrigation); sites with an intermediate impedance decrease (5 to 9 Ω) were also marked (pink or white tags; considered intermediate-quality lesions) (Online Figure 2B).
After anatomic completion of the encircling ablation line and adenosine testing, pacing was performed during sinus rhythm to evaluate pace capture along the line encircling the ipsilateral veins. If atrial tissue was still excitable to pacing, additional ablation lesions were delivered until loss of pace capture was achieved at that location (12). No additional RF energy was applied to posterior wall sites adjacent to the esophagus if significant esophageal temperature rises had occurred in that location.
For the purpose of the comparing biophysical parameters of low-flow versus high-flow irrigation settings in humans, only lesions delivered to the posterior LA were included. Furthermore, only patients undergoing their first-time AF ablation (de novo ablation) were included in analysis of biophysical parameters. The following biophysical parameters were evaluated blinded to group assignment: average CF, FTI, impedance decrease, percentage of lesions meeting a minimum 10-Ω impedance decrease, loss of bipolar pace capture (at output of 10 mA and 2 ms), and the percentage of lesions meeting bipolar EGM-based criteria for transmurality (8).
Human AF Ablation: Assessment of Safety and Efficacy
For comparison of safety and efficacy, patients who underwent posterior LA ablation using the high-flow settings (first 160 consecutive patients, including the 40 patients who had biophysical data were analyzed) were compared with the next 166 consecutive patients who underwent posterior LA ablation using the low-flow settings (including the 20 patients who had biophysical parameters were assessed). Endpoints evaluated were acute PV isolation, acute major complications (thromboembolism, tamponade, symptomatic esophageal injury, or atrioesophageal fistula), and recurrence of atrial tachyarrhythmia in follow-up. Our approach to post-ablation follow-up has been previously described (9,10).
The Statistical Package for Social Sciences (IBM, Armonk, New York) was used for analysis. Continuous variables were expressed as mean ± SD if normally distributed; median and interquartile range 25% to 75% (Q25 to Q75) were used if the data were clearly skewed. Where normal distribution was not present, log transformation of the raw values was performed to meet the assumption of homogeneity of variance. In the animal study, to account for multiple observations in the same animal, and between different RA regions, and the resulting correlation between animals and region, we applied linear mixed models to compare the differences in biophysical and pathological lesion parameters. Similarly, when comparing biophysical parameters in low- vs. high-flow groups, linear mixed models were used to account for multiple lesions in the same patient. To test for associations between categorical variables, chi-square tests or the Fisher exact test was used. Survival free of atrial tachyarrhythmia in follow-up was estimated using the Kaplan-Meier procedure. A 2-tailed p value < 0.05 was considered statistically significant. Graphs were constructed using either SPSS, Prism, version 5.0 d (GraphPad Software, La Jolla, California) or Microsoft Excel (Microsoft, Redmond, Washington).
Seventy lesions were delivered to the smooth posterior RA (21, 23, and 26 in the high, mid-, and low posterior RA, respectively). A median of 3 lesions (Q25-Q75: 3 to 4, range 2 to 6) per swine were delivered. Thirty-six lesions were low flow, and 34 were delivered with the high-flow ablation settings. Average CF, RF duration, FTI, and total energy delivered were similar; average and maximum tip temperatures were higher in low-flow compared with high-flow lesions (Table 1). There were no instances of impedance rise, steam pops, or thrombus visible on the catheter tip on ICE or with visual inspection of the catheter in either group.
Impedance decrease was significantly greater for low-flow compared with high-flow lesions, as were percentage of lesions with a ≥10-Ω impedance decrease (Figures 1A and 1B). The percentage of lesions with loss of unipolar and bipolar pace capture after ablation was significantly higher in low-flow compared with high-flow lesions (Figure 1C). Degree of unipolar and bipolar EGM amplitude reduction after ablation was similar (Figure 1D). A significantly higher percentage of low-flow lesions met EGM-based unipolar and bipolar criteria for lesion transmurality (Figure 1E).
Mean tissue thickness in the swine posterior RA was 1.9 ± 0.8 mm (range 0.6 to 3.5 mm), overall being thinner in swine randomized to low-flow lesions (Table 2). On anatomic inspection of the endocardial surface, 27 of 36 low-flow lesions (75%) were visible, whereas only 14 of 34 high-flow lesions (41%) were visible after staining (p = 0.007). There was no significant difference in biophysical parameters in low-flow lesions visible on pathology versus nonvisible lesions (Online Table 1). Among high-flow lesions, nonvisible lesions had a similar impedance decrease (10 ± 8 Ω vs. 9 ± 6 Ω; p = 0.66); however, tip temperature was lower (35 ± 2°C vs. 36 ± 1°C; p = 0.01), RF duration was shorter (11 ± 2 vs. 13 ± 2 s; p = 0.006), FTI was smaller (231 ± 42 g ∙ s vs. 286 ± 59; p = 0.003); and had lower energy delivery compared with visible high-flow lesions (215 ± 43 J vs. 260 ± 46 J; p = 0.006). Higher tip temperature (48 ± 3°C vs. 35 ± 2°C; p < 0.001) and greater impedance decrease (20 ± 11 vs. 10 ± 8 Ω; p = 0.002) were the most significant parameters that were different between visible low-flow lesions and nonvisible high-flow lesions, respectively (Online Table 1).
Figure 2 and Table 2 summarize the dimensions of visible lesions using the low-flow versus high-flow setting (figure not to scale for dimensions) with examples shown in Online Figure 3. No thrombus was noted on any visible lesion on pathology. Among the visible lesions, low-flow compared with high-flow lesions had: similar lesion volume, similar endocardial surface diameter, similar lesion depth corrected for tissue thickness, and similar percentage of transmural lesions, but a smaller maximum diameter. Low-flow lesions had the maximum diameter at or 0.04 mm below the endocardial surface; this was displaced 1.5 mm below the endocardial surface in high-flow lesions (Table 2, Figure 2). A greater proportion of low-flow lesions had a maximal diameter at the endocardial, rather than the epicardial surface (Table 2, Online Figure 3). By contrast, a greater proportion of high-flow lesions had maximal diameter located at the epicardial, rather than the endocardial surface (Table 2, Online Figure 3). When epicardial extension was evident, the maximal epicardial lesion diameter was larger in high-flow compared with low-flow lesions (Table 2, Online Figure 3).
Human data: Biophysical parameters
The baseline characteristics and procedural data of the 60 patients (40 high-flow and 20 nonirrigated) who underwent a de novo AF ablation who had biophysical parameters were evaluated are shown in Table 3. Notably, the total number of lesions delivered on the posterior LA wall was similar between high-flow and low-flow patients. However, the total ablation time and energy delivered to the posterior LA was significantly lower in patients receiving low-flow versus high-flow posterior LA ablation (Table 3).
Analysis of force, impedance, and RF duration was performed on 1,345 high-flow lesions in 40 patients (median 25 lesions/patient, Q25 to Q75: 21 to 39) and 558 low-flow irrigation lesions in 20 patients (median 32 lesions/patient, Q25 to Q75: 25 to 41). There were no instances of impedance rise, steam pops, or thrombus noted on the catheter during the procedure.
Assessment of EGM-based transmurality was possible in 33 patients (17 low-flow, 16 high-flow; and assessment of post-ablation pace capture was possible in 40 patients (19 low-flow, 21 high-flow). EGM analysis was not possible in 27 patients due to the presence of AF during ablation (n = 15) or failure to synchronize the ablation point on the electroanatomic mapping system with the EGM recording system (n = 12). Analysis of pacing data was not possible in 20 patients because it was not performed (n = 10) or there was failure to synchronize the ablation point on the mapping and recording system (n = 10).
Average CF was identical between low-flow and high-flow lesions (Figure 3A). High-flow lesions, compared with low-flow lesions, had a longer RF duration in an attempt to achieve a minimum impedance fall of 10 Ω (Figure 3B), and consequently had a larger FTI (Figure 3C). Low-flow lesions had lower average power and higher tip temperatures compared with high-flow lesions (Figure 3D). Low-flow, compared with high-flow lesions had a significantly larger mean impedance decrease (Figure 3E) and a greater proportion of lesions that reached an impedance decrease of ≥1 0Ω (Figure 3F), despite shorter RF duration and a smaller FTI. Representative examples of cases are shown in Online Figures 2B and 2C.
A greater percentage of lesions were transmural on bipolar EGM analysis in low-flow compared with high-flow lesions (Figure 3G). Loss of bipolar pace capture was also more common in low-flow compared with high-flow lesions (Figure 3H). Representative examples of EGM-based assessment of transmurality in high-flow versus low-flow lesions is shown in Online Figure 4.
Human data: Safety and efficacy
None of the 326 patients who underwent ablation using either the low-flow or high-flow settings experienced thromboembolic events, symptomatic esophageal injury, or atrioesophageal fistula. Of the 326 patients included, 250 had a de novo AF ablation and 76 had a redo-AF ablation (baseline clinical and procedural characteristics are shown in Online Table 2). Successful PV isolation was achieved in all 250 de novo AF ablations. Freedom from atrial tachyarrhythmia in patients undergoing de novo AF ablation was similar in patients with paroxysmal AF (at 12 months using the Kaplan-Meier procedure, low-flow 75 ± 11% vs. high-flow 76 ± 9%; p = 0.99) or persistent AF (at 12 months, using the Kaplan-Meier procedure, low flow 33 ± 18% vs. high flow 54 ± 11%; p = 0.5). The atrial arrhythmia-free survival after the last ablation procedure was similar when pooling the de-novo and redo-AF ablation patients.
This study conveys clinically relevant preliminary observations on the rationale, safety, and efficacy of lowering the irrigation flow rate on lesion formation in thin atrial tissue from swine experiments and a clinical study. The salient findings were:
1. In thin atrial tissue (0.6 to 3.5 mm), low-flow irrigation (2 ml/min) produces lesions that have their greatest diameter at or slightly below the endocardial surface and greater epicardial sparing, and this can be done without producing char. The effect is reflected in a greater impedance decrease and higher maximum tip temperatures consistent with greater endocardial surface heating. By contrast, standard irrigated ablation (high flow 17 ml/min) caused endocardial sparing, with a maximal lesion diameter 1.5 mm below the endocardial surface, while also producing a more frequent epicardial lesion. This effect is likely attributed to endocardial surface cooling resulting in endocardial surface sparing whilst paradoxically increasing epicardial surface diameter, which may not be desirable in thin atrial tissue in close proximity to structures vulnerable to collateral injury such as the esophagus.
2. In humans, lowering the irrigation flow rate, compared with high-flow rates, was more likely to produce lesions with EGM characteristics of transmural lesions and an increase of pacing threshold to >10 mA despite a shorter RF time and less energy delivery to the posterior LA.
3. Low-flow, compared with high-flow, irrigation settings for posterior LA ablation were similar in safety and efficacy when compared in a nonrandomized, retrospective sample of 326 patients undergoing AF ablation. A randomized study is needed to compare the safety and efficacy of both approaches.
Irrigated catheters produce deeper, transmural lesions and minimize the risk of thrombus formation during high-power, long-duration RF applications. With nonirrigated ablation, lesion width and depth increases linearly as a function of electrode–tissue interface temperature up to 90°C; however, coagulum formation on the catheter tip occurs with temperatures exceeding 80°C, and a sudden rise in impedance can occur, frequently limiting the duration of RF delivery and thus the extent of energy delivery to tissue. Tissue temperatures are highest just beneath the endocardial surface, and small, superficial lesions may result if power or ablation duration is limited to avoid high catheter tip temperatures (3).
Irrigation allows cooling of the tip–tissue interface, avoiding coagulum formation such that high power can be delivered for a longer duration. It also changes the surface geometry of the lesion due to the fact that a smaller endocardial surface area is exposed to lethal tissue temperatures (>50°C to 60°C) (6), and that peak tissue temperature is displaced deeper into the tissue (3). This results in the maximum lesion diameter typically being 4 mm below the endocardial surface (“gum drop”-shaped lesion), with the potential for endocardial sparing, particularly if the lesion duration is shortened, such as is frequently the case when the catheter tip opposes the esophagus (3).
Although creation of deeper lesions is desirable in thick tissue, its geometrical profile may not be favorable in thin walled tissue, particularly the posterior LA, which lies in close proximity to the esophagus. A hypothetical example is described in Online Figure 1 where irrigated ablation could potentially expose the esophagus to much higher temperatures to achieve the same endocardial lesion dimensions as a nonirrigated lesion. A nonirrigated lesion would create maximum temperature within 1 mm of the endocardial surface with a decreasing (not increasing) thermal gradient in the direction of the epicardium and esophagus, in which case, the maximum diameter of the lesion is at the endocardial surface. It is plausible that nonirrigated lesion delivery may allow one to achieve a thin transmural lesion more quickly with relative sparing of the esophagus. However, the perceived thromboembolism risk and potentially inefficient energy delivery with nonirrigated ablation makes it less readily acceptable based on the current recommendations for irrigation with contact force-sensing catheters. Low-flow irrigation has not previously been studied for this purpose.
In this study, we examined whether a reduction of the irrigation flow rate from a standard of 17 ml/min to 2 ml/min would favorably alter lesion geometry and potentially create thin, transmural lesions with less energy delivery outside the heart wall. We used a swine model of ablation over thin-walled smooth posterior RA that has similar tissue thickness (1.9 ± 0.8 mm, range 0.6 to 3.5 mm) to the human posterior LA (2.2 ± 0.9 mm, range 0.9 to 7.4 mm) (13).
The most striking finding was that high-flow lesions, compared with low-flow lesions, were less frequently seen on the endocardium during pathology despite similar ablation settings and notable impedance fall during ablation. A likely reason for this is that irrigation produced a greater extent of endocardial sparing from surface cooling that reduced endocardial lesion formation. A longer lesion delivery may have eventually overcome endocardial sparing, but the lesion would also continue to grow in the deeper within tissue with the potential for collateral extension (e.g., esophagus). High-flow lesions were similar in volume, with maximum diameter 1.5 mm below the endocardial surface and a larger diameter on the epicardial, rather than the endocardial, surface. The maximum lesion dimension being 1.5 mm below the endocardial surface suggests greater potential for collateral injury of tissue directly underneath. By contrast, low-flow lesions had a maximal lesion diameter at or slightly below the endocardial surface and smaller diameter at the epicardial surface without compromising lesion transmurality or lesion volume. Indiscriminate application of the same irrigation settings throughout thick- and thin-walled atrial tissue may also help explain limitations of lesion efficacy in contemporary AF ablation.
As noted in swine and human studies, low-flow, compared with standard, irrigation also produced a more rapid impedance decrease with shorter RF applications and one-half the total posterior LA wall ablation time, while achieving a higher incidence of electrical inexcitability with pacing compared with higher-flow irrigation. Importantly, there was no evidence of abrupt rise in impedance or steam pops; coagulum formation was noted with low-flow ablation at 20 to 25 W of the limited durations studied (10 to 15 s). Greater tip temperatures seen with low-flow lesions seem to be responsible for the larger impedance decreases, which may partially explain why impedance may not be a good predictor of overall lesion size (14), but rather it may be a better predictor of endocardial surface heating.
Observations from the swine experiments were supported by analysis of biophysical parameters in 60 patients who underwent posterior LA ablation with standard or low-flow irrigation. Low-flow, compared with standard, irrigation was associated with a more rapid and greater impedance decrease, despite shorter RF applications, and a higher incidence of transmural lesions and electrical inexcitability with pacing. Indeed, total ablation time and total posterior wall energy delivered were significantly lower in the low-flow, compared with the high-flow, group. No adverse thromboembolic complications or esophageal injury occurred in 166 consecutive patients undergoing low-flow posterior LA ablation. Efficacy was not compromised in the short term or in follow-up, with similar rates of acute PV isolation and freedom from atrial tachyarrhythmias in patients treated with low-flow posterior LA lesions (n = 166), compared with patients treated with standard irrigation (n = 160), during posterior LA ablation.
These preliminary data suggest that low-flow irrigation has the potential for safer and more efficient RF energy delivery to the posterior LA by creating more superficial endocardial lesions compared with deeper lesions with endocardial sparing created with standard irrigation settings. The optimal combination of power, duration, contact force, and irrigation remains to be determined.
The data serve as a preliminary, but important, set of observations that provide impetus for a randomized clinical trial of low-flow versus standard irrigation settings. Although esophageal temperature was monitored, we did not systematically record this information for analysis and cannot prove that low-flow lesions reduce esophageal exposure to higher tissue temperature. Routine esophageal endoscopy was not performed to assess potential injury. Although no cerebrovascular or other thromboembolic events were noted, cranial imaging was not performed. We used surrogate markers for lesion completeness, namely, impedance fall, inexcitability with pacing, and EGM-based assessment of transmurality in the human study. These surrogate markers have been extensively studied and correlate with lesion completeness (8,15,16). The findings were impressive, supporting effective lesions creation, and PV isolation was achieved, but magnetic resonance imaging-based assessment of lesion completeness was not used. The optimal contact force, duration of ablation, and power and temperature settings required for low-flow ablation need to be elucidated in titration experiments. Surface lesions could potentially be produced with solid-tip electrode catheters. Rather than changing catheters, RF was applied with during the low irrigation rate used for mapping, as has been our common practice when smaller ablation lesions are desired, as when AV nodal re-entry is encountered in patients undergoing ablation for atrial fibrillation. Low-flow irrigation may still have benefit in reducing thrombus formation and findings might differ for non-irrigated catheters or catheters with other irrigation characteristics.
Low-flow irrigation for the posterior LA creates superficial lesions without comprising safety or efficacy of transmurality, when compared with standard irrigation settings, which produce larger epicardial lesions with endocardial sparing and thus potential for extension to collateral structures. Whether low-flow irrigation can improve safety and efficacy of post-LA ablation compared with high-flow irrigation warrants study in a randomized trial.
COMPETENCY IN MEDICAL KNOWLEDGE: Low-flow (2 ml/min), compared with high-flow (17 ml/min), irrigated ablation of thin atrial tissue results in more-superficial atrial lesions while affording relative epicardial sparing. This may translate to reduced collateral esophageal injury during posterior left atrial ablation.
TRANSLATIONAL OUTLOOK: A randomized trial comparing the safety and efficacy of low-flow, compared with high-flow, posterior wall irrigation is needed. Further work defining the optimal contact force, power, and duration of ablation to achieve transmural lesions in the posterior left atrium while avoiding collateral esophageal injury is needed.
The authors wish to thank Dr. John Ashton, PhD, Timothy Campbell, and Kathryn Finnerty from Biosense Webster Inc., for their assistance with the study.
Dr. Kumar is a recipient of the Neil Hamilton Fairley Overseas Research scholarship cofunded by the National Health and Medical Research Council and the National Heart Foundation of Australia and the Bushell Travelling Fellowship funded by the Royal Australasian College of Physicians. The animal experiment in this study was funded by a study grant from Biosense Webster Inc. Dr. Stevenson is coholder of a patent for needle ablation that is consigned to Brigham and Women’s Hospital; and his spouse receives research support from St. Jude Medical. Dr. Epstein has received consulting fees/honoraria from Abbott, Boston Scientific, Medtronic, and Spectranetics. Dr. Tedrow has received consulting fees/honoraria from Boston Scientific and St. Jude Medical; has received research funding from Biosense Webster and St. Jude Medical; and is on the faculty of the fellows program for St. Jude Medical, Boston Scientific, Biosense Webster, and Medtronic. Dr. John has received consulting fees/honoraria from St. Jude Medical. Dr. Michaud has received consulting fees/honoraria from Boston Scientific, Medtronic, and Abbott/St. Jude Medical; and has received research funding from Boston Scientific and Biosense Webster. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Kumar and Romero contributed equally to this work.
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
- atrial fibrillation
- contact force
- force-time integral
- intracardiac echo
- left atrial/atrium
- pulmonary vein
- interquartile range
- right atrial/atrium
- Received April 3, 2017.
- Revision received August 28, 2017.
- Accepted September 6, 2017.
- 2017 The Authors
- Wood M.,
- Goldberg S.,
- Lau M.,
- et al.
- Haines D.E.,
- Verow A.F.
- Nakagawa H.,
- Yamanashi W.S.,
- Pitha J.V.,
- et al.
- Petersen H.H.,
- Chen X.,
- Pietersen A.,
- Svendsen J.H.,
- Haunso S.
- Nath S.,
- Lynch C. 3rd.,
- Whayne J.G.,
- Haines D.E.
- Chinitz J.S.,
- Kapur S.,
- Barbhaiya C.,
- et al.
- Steven D.,
- Sultan A.,
- Reddy V.,
- et al.
- Lemola K.,
- Sneider M.,
- Desjardins B.,
- et al.
- Yokoyama K.,
- Nakagawa H.,
- Shah D.C.,
- et al.