Analysis of Catheter-Tip (8-mm) and Actual Tissue Temperatures Achieved During Radiofrequency Ablation at the Orifice of the Pulmonary Vein
Background— Many ablative approaches in or near the orifice of the pulmonary vein (PV) have demonstrated success in eliminating atrial fibrillation. Despite current practice, there are no data regarding the in vivo efficacy and safety of an 8-mm catheter tip for ablation at the PV orifice.
Methods and Results— Ten mongrel dogs were studied. Thermocouples were implanted in the atrial muscle of the PV orifice. Intracardiac echocardiography monitored catheter position, tip/tissue orientation, and microbubble formation. Ninety-four ablations were performed for 120 seconds. A temperature discrepancy >10°C between the catheter tip and tissue occurred during 47 (50%) of the ablations. Despite termination of energy delivery, the average tissue temperature remained within 1°C of the achieved steady state for 9 seconds. A temperature discrepancy >10°C was more common in the right superior PV, with oblique catheter positioning, when tissue temperatures were >60°C or 80°C, and with type 1 or type 2 microbubble formation. However, microbubbles were not present in 7 (13%, type 1) and 10 (40%, type 2) ablations with tissue temperatures >80°C. The maximum tissue temperature achieved with non–full-thickness lesions was 47.3±7.4°C vs 75.9±11.7°C (P<0.0001) for full-thickness lesions.
Conclusions— Marked discrepancies between catheter-tip and tissue temperatures occurred with higher temperatures, prolonged ablation times, and unfavorable catheter thermistor–tissue contact. Also, these data suggest a conservative approach to atrial ablation, because full-thickness lesions were obtained when tissue temperatures reached 50°C to 60°C and the tissue retained high heat levels despite termination of radiofrequency energy. Finally, microbubbles are inconsistent markers of tissue overheating.
Received March 31, 2004; revision received June 16, 2004; accepted June 21, 2004.
Predicated on pulmonary vein (PV) involvement in the pathophysiology of atrial fibrillation, many different ablative approaches at or near the orifice of the PV have demonstrated substantial therapeutic success. Nevertheless, a variety of complications, including PV stenosis, char and thrombus formation, and stroke and other embolic events remain a significant source of morbidity. The causes of these complications are incompletely understood, although several groups have suggested excess thermal injury or other components of the ablative strategy, such as repeated energy deliveries, as contributors.1–6 However, recent data suggest that the catheter-tip temperatures used to titrate energy delivery may be unreliable.7,8
These ongoing problems have provided the incentive for developing other means of grading energy titration, such as using the formation of microbubbles as a marker of tissue heating.4,5 Optimism for such an approach is premature without studies to correlate microbubble formation with actual tissue temperatures. Alternatively, 8-mm and irrigated-tip technologies have been developed to optimize ablation efficacy while minimizing complications.
Despite current clinical practice, few data exist regarding the in vivo efficacy and safety of using an 8-mm catheter tip for ablation at the PV orifice or in left atrial tissue. It is unclear whether the temperature recorded at the catheter tip accurately reflects tissue temperature at times or if an extensive “cooled-tip” effect is present near the PV orifice.9 Therefore, the purposes of this study were to (1) determine the catheter-tip versus tissue temperature profile during ablation with an 8-mm tipped catheter, (2) determine the clinical factors that may result in inaccurate catheter-tip reflection of actual tissue temperatures, and (3) evaluate the impact of repetitive burns at a single site on the dynamics of tissue heating.
The study was approved by the Mayo Foundation Institutional Animal Care and Use Committee. Ten mongrel dogs weighing between 30 and 40 kg were used. The animals were anesthetized with intravenous ketamine (10 mg/kg) and diazepam (0.5 mg/kg), intubated, and maintained on 1% to 3% isoflurane with positive-pressure ventilation. The hair at both groins, jugular regions, and chest wall was shaved for bilateral thoracotomies and ablation patch positioning. The body temperature was maintained at 37°C by water-flow heating pads and warmed intravenous solutions. The surface ECG, temperature, and blood pressure were recorded before catheterization and every 30 minutes from the time of catheterization.
By a percutaneous approach, 5 sheaths were positioned: an 8F into the left external jugular vein, a 12F into the right external jugular vein, an 8F into the right femoral vein and artery, and a 12F into the left femoral vein. A 6F coronary sinus catheter was introduced through the left external jugular vein and advanced to the distal coronary sinus for anatomic guidance and pacing.
After establishment of anesthesia, a small left-sided, posterolateral thoracotomy was performed. Careful dissection exposed the external aspects of each PV orifice. For each PV, 2 to 4 thermocouples were implanted in the atrial muscle orifice of the PV. A radiopaque marker was attached to the tip of the thermocouple for fluoroscopic localization. After completion of the left-sided implantation and ablation, the same procedure was followed to dissect out the right superior PV via a right thoracotomy.
The PV orifice tissue temperature was measured directly by embedded IT-21 copper/constantan T-type thermocouples (Physitemp Instrument Inc) with a time constant of 0.08 second. The thermocouples were 0.23 mm in diameter. The sensor signals were amplified and filtered by low-pass filters to minimize the power-source 60-Hz and radiofrequency (RF)-source 500-kHz interference. An analog-to-digital converter (National Instruments) sampled the data at 2.8 samples per channel when acquiring 4 simultaneous channels.
The accuracy of the temperature acquisition system was tested in vitro during RF ablation as previously reported.8 In brief, postablation temperature decay was examined and compared with that of water-heated tissue to examine the effects of the thermocouples acting as RF antennas. Sampling frequency was set at 2.8 samples per second. The temperature measurement system was calibrated before use against a digital thermometer, with an accuracy of 0.42°C.
A 10F, 64-element, 5.5- to 10.0-MHz, phased-array intracardiac echocardiography (ICE) catheter was introduced via the right external jugular vein and advanced to the level of the tricuspid annulus. ICE-guided transseptal catheterization to the left atrium was performed with an 8F Mullins sheath and a Brockenbrough needle, as previously described.10 The ICE catheter was advanced through the left femoral vein to the level of the tricuspid annulus to monitor the catheter positioning, tip/tissue orientation, the development of microbubbles, and/or tissue changes during ablation.
Catheterization and Ablation
A temperature-controlled, multisensor catheter with an 8-mm-tip electrode was inserted through the right femoral sheath. The catheter system referred to as the “tip” in this article consisted of 4 temperature sensors embedded within the distal catheter, with 1 at the most distal tip location, and 3 located 120° axially from each other in a more proximal position. Power output was limited by the highest of the 4 thermistor temperatures. The maximum temperature from the 4 thermistor recordings during a 1-second interval was considered the maximum catheter-tip temperature for the subsequent analysis.
Under guidance by fluoroscopy and ICE, the ablation catheter tip was placed endocardially, directly across from the epicardially implanted thermocouple. The catheter position and orientation were verified by fluoroscopy, ICE, and the radiopaque thermocouple marker.
Ablation energy was first titrated to achieve a steady-state temperature of 50°C at the catheter tip. Once this temperature was reached, the RF power was increased to increment the tip temperature by 5°C. When the steady-state temperature was achieved, the process of increasing the temperature was repeated over the course of 120 seconds. The rate of temperature increase was 5 to 15 seconds, depending on the time required to reach a steady-state catheter-tip temperature. Simultaneous data acquisition of catheter-tip and PV tissue temperatures were made, and power and impedance were recorded. Occurrence of microbubbles was noted and characterized as scattered microbubbles (type 1) and a brisk “shower” of dense microbubbles (type 2).4 The catheter temperature was increased throughout the ablation period, independent of microbubble formation.
At the end of the study, ventricular fibrillation was induced with high-rate burst pacing, and the animal was exsanguinated. The entire heart-lung preparation was removed with the pericardium intact. A gross examination was performed at this time, during which both endocardial and epicardial tissue surfaces of ablation lesions were reviewed. Evidence of device-related trauma to any cardiac structure was documented. Ablation lesions were characterized as full or partial thickness.
Continuous variables were reported as mean±SD, and comparisons between groups were based on a 2-sample t test (parametric) or the Wilcoxon rank-sum test (nonparametric). Categorical variables were summarized as percentages, and group comparisons were based on the χ2 test. Subsequent multivariate analyses were based on logistic-regression models in which a temperature discrepancy of >10°C and >20°C and the development of type 1and type 2 microbubbles were dependent variables. In each analysis, odds ratios (ORs) with 95% confidence intervals (CIs) were used to characterize the association of ablation characteristics (catheter orientation, PV, ablation number, and temperature >60°C and >80°C) with the likelihood of developing the dependent variable. A fit of the in vivo PV orifice temperature was examined by plotting this versus the catheter-tip temperature. The data correlation and trendline were assessed by the highest R2 result without evidence of lack of fit.
General Ablation Characteristics
A total of 94 ablations were performed at the superior and inferior PV orifices; the right superior PV in 38 (40%), left superior PV in 28 (30%), and left inferior PV in 28 (30%). The catheter position was oblique during 53 (56%) ablations, perpendicular in 26 (38%) ablations, and parallel in 15 (16%) ablations, as established by ICE and fluoroscopy. Details of the ablations are listed in Table 1. Examples showing the correlation between catheter-tip and tissue temperatures are shown in Figure 1 (A, minimal tissue/tip temperature discrepancy; B, marked [>20°C] tissue/tip temperature discrepancy). Figure 2 displays the actual tissue temperatures achieved per individual catheter-tip temperature recording. Figure 3 displays the average tissue and catheter-tip temperatures per 20-second intervals.
Impact of Ablation Number
The ablation attempts were examined at 30, 60, and 90 seconds, for 10 seconds, after a tissue steady-state temperature was reached (Table 2). Despite similar power, there was a trend toward increased tissue temperatures with more ablations at 30, 60, and 90 seconds.
The temperature discrepancy was >10°C during 18 (50%, ablation 1), 18 (56%, ablation 2), and 11 (42%, ablation 3) ablations. The discrepancy was >20°C during 9 (25%, ablation 1), 12 (38%, ablation 2), and 11 (42%, ablation 3) ablations (Figure 4A). Both the number of discrepancies >10°C and >20°C were reduced when assessing the first 60 seconds of energy delivery, although the incidence increased with ablation number (Figure 4B).
Predictors of Temperature Discrepancy
Although the overall average tissue and catheter-tip temperatures were similar, there were periods when there was a marked discrepancy. A temperature discrepancy >10°C was more common in the right superior PV, with an oblique catheter position, with tissue temperatures >60°C, and when type 1 or type 2 microbubbles were present (Table 3). In multivariate analysis, the ablation number and type 1 microbubbles were predictive of an increased risk of a >10°C tissue discrepancy (Figure 5A). Similarly, a marked temperature discrepancy of >20°C was more likely with ablation in the right superior PV, with an oblique catheter position, when tissue temperatures were >60°C, and when type 1 or type 2 microbubbles were present (Table 4). Multivariate predictors of a >20°C temperature discrepancy included ablation in the right superior PV, with a strong risk trend with type 2 microbubble appearance (Figure 5B).
Type 1 microbubbles occurred at 69.7±11.1°C versus type 2 microbubbles at 77.3±11.7°C (P=0.016). Type 1 microbubble formation occurred more frequently with ablation of the right superior PV and with tissue temperatures >60°C (Table 5). However, in 7 (13%) ablations with tissue temperatures in excess of 80°C, no type 1 microbubbles were present. Multivariate predictors of type 1 microbubble formation included ablation in the right superior PV (OR 4.25; 95% CI, 2.98 to 5.52; P=0.039) and temperatures >60°C (OR 5.64; 95% CI, 3.91 to 7.40; P=0.018). Similarly, type 2 microbubble formation occurred more frequently with ablation of the right superior PV, with temperatures >60°C, and when type 1 microbubbles were present (Table 6). In multivariate analysis, the appearance of type 1 microbubbles predicted the occurrence of type 2 microbubble development (OR 3.35; 95% CI, 2.23 to 4.47; P=0.067). However, in 10 (40%) ablations, type 1 microbubbles did not precede the formation of type 2 microbubbles. In addition, type 2 microbubbles were not present in 10 (40%) ablations when temperatures were >80°C.
Despite termination of energy delivery at 120 seconds, temperatures remained within 1°C of the achieved steady state for ≈9 seconds before decreasing (Figure 6). In comparison, tissue temperatures remained constant after 60-second ablations for only 6 seconds. The subsequent temperature loss was less dramatic after 60 seconds versus 120 seconds (maximum temperature decay for 60 seconds, 6 to 9 seconds: ablation 1, −0.53°C/s; ablation 2, −0.80°C/s; for 120 seconds, 10 to 13 seconds: ablation 1, −1.51°C/s; ablation 2, −1.82°C/s).
At autopsy, no device-related trauma was noticed. Full-thickness lesions were achieved in all but 3 (3.2%) of the PV orifices. A total of 9 energy deliveries were delivered to these PV orifices. Figure 7 contains a histogram of ablation lesion characteristics based on the maximum tissue temperature achieved. The maximum tissue temperature achieved during non–full-thickness ablation attempts was 47.3±7.4°C in comparison with a maximum temperature of 75.9±11.7°C (P<0.0001) achieved when full-thickness lesions were obtained.
This study has several important findings relevant to the clinical practice of RF ablation of atrial arrhythmias. First, the 8-mm catheter-tip temperatures were correlated, in general, with tissue temperatures during 60-second energy deliveries. Second, after completion of the ablation, the steady-state temperature achieved persisted. In addition, the subsequent temperature decay was dependent on the initial ablation time. Third, large tissue and catheter-tip temperature discrepancies were more common with higher temperatures, an oblique catheter orientation, and prolonged ablation times. Fourth, estimation of tissue heating with either type 1 or type 2 microbubble formation as a guide was inaccurate, even at very high tissue temperatures (>80°C).
Although catheter-tip temperatures have been traditionally used to guide RF energy delivery, very little is known about the accompanying in vivo tissue temperatures.11,12 Even less is known regarding tissue temperatures generated at the PV orifice. In general, in non-PV tissue, tissue temperatures are correlated with lesion size.13–15 With excessive temperatures, tissue charring, thrombus formation, and impedance rises occur, and the potential for collateral tissue injury is increased.13 These surface changes may provide the substrate for embolic phenomena. Furthermore, PVs subject to high temperatures (>60°C) lose structural integrity of the underlying collagen and elastin matrix, with resultant circumferential narrowing.3 These data underscore the need for accurate catheter-tip measurement of tissue temperatures.
Newer catheter designs incorporating 8-mm and irrigated-tip technologies have been developed to optimize efficacy while minimizing complications.14–22 With larger electrode sizes, the tip surface area exposed to the circulating blood increases, resulting in convective heat loss and tip cooling.22 The process is similar to cooling produced by saline-irrigated tips.15 Despite the use of these catheter designs in clinical practice, few data exist regarding the in vivo efficacy and safety. Furthermore, it is unclear whether temperatures recorded with an 8-mm tip accurately reflect tissue temperature at times or if an extensive “cooled-tip” effect is present near the PV orifice.
This study provides in vivo data generated at the PV orifice with use of an 8-mm tip and suggests a good correlation between catheter-tip and tissue temperatures during the first 60 seconds of ablation. A time-dependent augmentation in catheter-tip and tissue temperature discrepancy may reflect a thermal latency of the tissue in vivo. As previously reported and demonstrated in this study, although tip temperatures dropped instantaneously with termination of RF energy, tissue temperatures may continue to be elevated.15,23–25 Mechanistically, this is thought to result from 2 different heating processes. First, there is rapid resistive heating of a thin layer of tissue around the catheter tip during ablation. Second, there is a slower conductive heating of tissue at a distance.24 With time, the overall impact of conductive heating is augmented, resulting in higher tissue temperatures despite consistent local resistive heating, as evidenced by the higher initial and latter steady-state temperatures with subsequent ablations in the same area. Furthermore, conductive energy persists after termination of energy delivery, which is directly related to ablation duration. These data explain why shorter ablation attempts may really reflect longer heating periods.
These data suggest that catheter-tip orientation influences the accuracy of predicted tissue temperatures. In particular, the oblique position was most associated with large temperature discrepancies. This may, in part, reflect incomplete contact between the catheter-tip temperature sensor and the tissue. With this catheter design, the parallel/perpendicular positions allow a thermistor to be in direct contact with the tissue. If the temperature sensor is exposed to a mixed blood/tissue interface, as with the oblique orientation, convective heat loss into the circulating blood may result in lower registered temperatures at the catheter tip. In turn, the catheter will increase the power erroneously, resulting in higher tissue temperatures. As demonstrated previously, in high-blood-flow states, there is an increased tissue-to–catheter-tip temperature discrepancy in a tissue-bath model in comparison with low-blood-flow states.9 In the canine model, a perpendicular tip/tissue orientation in the right superior PV is difficult to achieve because of the angle created in the small left atrium. Furthermore, a parallel position is difficult to achieve in the right superior PV without entering the PV. Anatomic obstacles that prevent close contact of sensing thermistors to the tissue may, in part, account for the temperature discrepancies seen with the right superior PV in this study. These data reinforce the need to verify catheter position by ICE and fluoroscopy to ensure thermistor contact with the tissue.
This study suggests that a conservative approach to temperature titration be pursued. Marked catheter-tip and tissue temperature discrepancies were more likely to occur at higher temperatures. In addition, all lesions that reached a temperature of 60°C were transmural. Previous reports have demonstrated that irreversible electrophysiologic changes to myocytes occur early at 50°C.26 Because only 1 ablation >55°C did not result in a transmural lesion, these data suggest that the target tissue temperature at the PV orifice should be 55°C. Furthermore, tissue temperatures remain elevated even after termination of RF energy. These data suggest that shorter ablation times may achieve therapeutic outcomes while minimizing complications associated with overheating.
These data also demonstrate the fallibility of titration of energy to achieve microbubbles as a surrogate marker of tissue overheating. This finding is of important clinical relevance because of the increasing practice of titrating ablation energy to type 1 microbubble formation.4,5 Although, on average, microbubble formation was correlated with excessive temperatures, there were ablation attempts with temperatures >80°C without microbubbles. Furthermore, type 2 microbubbles did not always follow type 1 microbubble formation.
Data from an in vivo canine model may not be similar to human tissue characteristics and responses; however, this model represents a close approximation. Second, the thermocouples were implanted in a single area and may not fully represent 3-dimensional, conductive tissue heat transfer. Third, the data were generated from an 8-mm temperature-controlled catheter, and results may be different with other ablation tip and system designs. Fourth, the pathologic findings in this study must be viewed in the context of multiple regional ablation attempts. However, the interpretation of tissue temperatures in which nontransmural lesions occurred is more straightforward, thus allowing a general tissue temperature target. Fifth, these data were generated in left atrial tissue at the PV orifice. Linear ablation at other locations within the left atrium was not examined.
Current ablation approaches are limited because of the imprecise means of assessing actual in vivo tissue temperatures. Without accurate means to determine tissue temperatures, ablative complications that result from overheating will persist. This is a novel study revealing the actual in vivo tissue temperatures while using an 8-mm temperature-controlled catheter. In general, the catheter-tip temperatures were similar to those in tissue up to 60 seconds. Prolonged ablation times, tissue/tip contact, and catheter orientation influenced the accuracy of tissue temperature prediction. However, even after cessation of ablation, tissue temperatures remain elevated. To minimize excessive tissue heating, these findings should be considered when additional ablation attempts are made in a targeted area or when ablation delivery times are long. Additional studies of other catheter types in current ablative practice need to be performed to determine their in vivo accuracy to prevent complications associated with tissue overheating.
This study was supported by an unrestricted research grant from Boston Scientific/EPT Technologies.
Taylor GW, Kay GN, Zheng X, et al. Pathological effects of extensive radiofrequency energy applications in the pulmonary veins in dogs. Circulation. 2000; 101: 1736–1742.
Marrouche NF, Martin DO, Wazni O, et al. Phased-array intracardiac echocardiography monitoring during pulmonary vein isolation in patients with atrial fibrillation: impact on outcome and complications. Circulation. 2003; 107: 2710–2716.
Saad EB, Rossillo A, Saad CP, et al. Pulmonary vein stenosis after radiofrequency ablation of atrial fibrillation: functional characterization, evolution, and influence of the ablation strategy. Circulation. 2003; 108: 3102–3107.
Su WW, Johnson SB, Packer DL. Marked discrepancies between catheter-tip and actual tissue temperatures achieved during RF ablation of pulmonary veins. Circulation. 2002; 106: 3116. Abstract.
McRury ID, Panescu D, Mitchell MA, et al. Nonuniform heating during radiofrequency catheter ablation with long electrodes: monitoring the edge effect. Circulation. 1997; 96: 4057–4064.
Petersen HH, Chen X, Pietersen A, et al. Tissue temperatures and lesion size during irrigated tip catheter radiofrequency ablation: an in vitro comparison of temperature-controlled irrigated tip ablation, power-controlled irrigated tip ablation, and standard temperature-controlled ablation. Pacing Clin Electrophysiol. 2000; 23: 8–17.
Nakagawa H, Yamanashi WS, Pitha JV, et al. Comparison of in vivo tissue temperature profile and lesion geometry for radiofrequency ablation with a saline-irrigated electrode versus temperature control in a canine thigh muscle preparation. Circulation. 1995; 91: 2264–2273.
Nakagawa H, Wittkampf FH, Yamanashi WS, et al. Inverse relationship between electrode size and lesion size during radiofrequency ablation with active electrode cooling. Circulation. 1998; 98: 458–465.
Petersen HH, Chen X, Pietersen A, et al. Temperature-controlled irrigated tip radiofrequency catheter ablation: comparison of in vivo and in vitro lesion dimensions for standard catheter and irrigated tip catheter with minimal infusion rate. J Cardiovasc Electrophysiol. 1998; 9: 409–414.
Nath S, Whayne JG, Kaul S, et al. Effects of radiofrequency catheter ablation on regional myocardial blood flow: possible mechanism for late electrophysiological outcome. Circulation. 1994; 89: 2667–2672.
Wittkampf FH, Nakagawa H, Yamanashi WS, et al. Thermal latency in radiofrequency ablation. Circulation. 1996; 93: 1083–1086.
Nath S, Lynch C 3rd, Whayne JG, et al. Cellular electrophysiological effects of hyperthermia on isolated guinea pig papillary muscle: implications for catheter ablation. Circulation. 1993; 88: 1826–1831.