Contact force sensing in ablation of ventricular arrhythmias using a 56-hole open-irrigation catheter: a propensity-matched analysis

Purpose The effect of adding contact force (CF) sensing to 56-hole tip irrigation in ventricular arrhythmia (VA) ablation has not been previously studied. We aimed to compare outcomes with and without CF sensing in VA ablation using a 56-hole radiofrequency (RF) catheter. Methods A total of 164 patients who underwent first-time VA ablation using Thermocool SmartTouch Surround Flow (TC-STSF) catheter (Biosense-Webster, Diamond Bar, CA, USA) were propensity-matched in a 1:1 fashion to 164 patients who had first-time ablation using Thermocool Surround Flow (TC-SF) catheter. Patients were matched for age, gender, cardiac aetiology, ejection fraction and approach. Acute success, complications and long-term follow-up were compared. Results There was no difference between procedures utilising either TC-SF or TC-STSF in acute success (TC-SF: 134/164 (82%), TC-STSF: 141/164 (86%), p = 0.3), complications (TC-SF: 11/164 (6.7%), TC-STSF: 11/164 (6.7%), p = 1.0) or VA-free survival (TC-SF: mean arrhythmia-free survival time = 5.9 years, 95% CI = 5.4–6.4, TC-STSF: mean = 3.2 years, 95% CI = 3–3.5, log-rank p = 0.74). Fluoroscopy time was longer in normal hearts with TC-SF (19 min, IQR: 14–30) than TC-STSF (14 min, IQR: 8–25; p = 0.04). Conclusion Both TC-SF and TC-STSF catheters are safe and effective in treating VAs. The use of CF sensing catheters did not improve safety or acute and long-term outcomes, but reduced fluoroscopy time in normal heart VA.


Introduction
Catheter ablation is an important treatment for ventricular arrhythmias (VA) [1], but long-term outcomes remain suboptimal. Measuring the contact force (CF) between the radiofrequency (RF) ablation catheter and myocardium aims to improve mapping accuracy, energy delivery and procedure safety, by optimising catheter-tissue contact. In comparison to the large volume of data available on measuring CF in AF ablation, much less work has focused on its real-world usefulness in VA. This study compares acute success, complication rate and long-term arrhythmia-free survival in catheter ablation procedures for VA using the 56-hole RF ablation Thermocool Surround Flow (TC-SF) catheter versus the Thermocool SmartTouch Surround Flow (TC-STSF) catheter (both Biosense Webster, Diamond Bar, CA, USA). The latter combines the ability to monitor the magnitude and direction of catheter-tissue contact with entire tip irrigation.

Patient population
Procedure-and outcome-related data were collected for the 164 patients undergoing first-time ablation for VA using TC-STSF catheter between 2016 and 2018 at St George's University Hospital, London, UK (91 with structurally normal heart, 38 with ischemic cardiomyopathy (ICM) and 35 with non-ischemic cardiomyopathy (NICM)). From a database of 264 first-time VA ablation procedures performed in the same institution using TC-SF catheter, one procedure was propensity-matched against each TC-STSF procedure (Fig. 1). The study was approved by the local institutional research ethics committee.
There were 209 procedures for ventricular ectopy (VE) and 119 for ventricular tachycardia (VT). The data comprised baseline patient demographics, left ventricular ejection fraction (LVEF), presence and indication for intracardiac devices, number of prior ablations and underlying cardiac aetiology. All patients underwent a transthoracic echocardiogram (TTE) or cardiac magnetic resonance imaging (cMRI), or both, preprocedurally. Coronary angiography was performed in all cases of structural heart disease and was discretional in the case of normal heart VAs.
The indication for VA ablation was categorised as symptomatic despite medical therapy, presumed VA-induced cardiomyopathy unresponsive to medical therapy, recurrent implantable cardioverter defibrillator (ICD) shocks or VT storm. Antiarrhythmic drug therapy was discontinued for 5 halflives, except for patients presenting with VT storm or those on Amiodarone. Scar-dependent VT ablation was performed under general anaesthesia or conscious sedation. Sub-xiphoid epicardial approach was undertaken at the discretion of the operator and was based on clinical aetiology, pre-procedural imaging or prior procedural mapping.

Definitions
VA was defined as either VT or VE. Normal heart VA was defined as the absence of structural abnormality on TTE or cMRI. Acute procedural success was defined as noninducibility of all VT(s) and the complete elimination of scar-related abnormal electrograms in VT cases and as > 80% reduction in clinical VE frequency despite isoprenaline challenge, during a 30-min wait period in VE cases. VA recurrence was defined as sustained monomorphic VT (> 30 s) or VT requiring ICD therapy (anti-tachycardia pacing or shock) for patients undergoing VT ablation and as < 80% reduction of the burden of clinical ectopy on 24-h Holter monitoring in patients undergoing VE ablation.

Ablation strategy
Dual left ventricular (LV) access was employed in scardependent VT cases, using both the trans-septal and the retrograde aortic approaches. Anticoagulation during LV procedures was maintained with unfractionated Heparin and titrated to achieve an activated clotting time of 300-400 s.
Patients 2016. Both catheters have a 3.5 mm tip with 56 pores in the distal electrode through which normal saline was infused at a rate of 2 ml/min during mapping. While using the TC-STSF catheter, the average and maximum CF between the catheter tip and the myocardium were measured and expressed in grams (g), aiming for CF of 10-40 g. Mapping was supplemented by multi-electrode mapping catheters, which were more commonly used in the latter years of the study. Voltage mapping was performed during sinus rhythm or right ventricular (RV) pacing at the outset. Programmed stimulation was performed with up to 3 extrastimuli at 2 cycle lengths. Ablation was performed in the power control mode, with temperature cut-off of 40°C, maximum power of 40 W (rarely up to 50 W on the interventricular septum), flow rate of 17 ml/min, duration of RF application of 60 s at each point, with lower energy and shorter duration at aortic sinus of Valsalva and para-Hisian locations. Ablation was terminated immediately if there was an impedance rise or an audible steam pop.
For VT cases, ablation of all clinical VTs was attempted. Abnormal scar-related electrograms were targeted with focal ablation in scar with endocardial voltage < 1.5 mV, aiming at the elimination of near-field electrograms, impedance drop of 10 Ω and loss of capture with moderate output pacing [2].
For VE cases, ablation was performed at the earliest site of local activation and/or the site of best pacemap (> 90%) match to the clinical VE.

Follow-up
All major intra-and post-procedural complications were recorded up to 30 days. Antithrombotic therapy was recommended for 1-3 months after LV endocardial procedures with either direct oral anticoagulants or warfarin with a goal international normalised ratio of 2-3. Long-term follow-up data were obtained from outpatient clinic, ambulatory electrocardiographic recordings and device interrogation reports. Follow-up for at least 1 year was available for 317 of 328 (96.6%) procedures, with a median follow-up duration of 5.4 years (interquartile range (IQR) = 4.1-6.3 years) for TC-SF and 2.3 years (IQR = 1.6-3.4 years) for TC-STSF.

Statistical analysis
A propensity score was calculated for all eligible patients undergoing first-time RF ablation for VAs through binary logistic regression with ablation catheter (TC-STSF or TC-SF) as the binary outcome and baseline variables were used as covariates for estimating the propensity score. Propensity matching was performed in a 1:1 fashion using the nearest neighbour approach with a two-decimal calliper for age, sex, aetiology, LVEF and approach (endocardial or epicardial). If no match could be found, then the TC-STSF subject was removed from the analysis. Univariate analyses of dichotomous, categorical and continuous variables were performed to determine similarities or differences between TC-SF and TC-STSF in normal heart, ICM and NICM groups. Categorical data were compared using the Chi-square test. The distribution of continuous variables was assessed for normality with Shapiro-Wilk test. Student's t test was used to compare normally distributed data and Mann-Whitney U test to compare non-parametric data between TC-SF and TC-STSF groups. Kaplan-Meier time event analysis was used to asses VA recurrence in each group, comparing TC-SF vs TC-STSF and significance determined by the log-rank test. Similarly, ICD shock recurrence and overall survival were compared in patients with ICM and NICM. Receiver operating characteristic (ROC) analysis was performed to find the average CF correlating with acute procedure success. Statistical analysis was performed using SPSS statistical software, version 26 (IBM Corp., Chicago, IL, USA).

Patient details
Baseline characteristics are presented in Table 1. Procedures using either TC-SF or TC-STSF were well matched in patients' gender distribution, age, distribution of cardiac aetiology and LVEF. There was no difference in the proportion of patients with implanted cardiac devices, indication for device implantation or indication for ablation.
Pre-procedure cMRI was performed in 57/164 (35%) patients in the TC-SF group and 43/164 (26%) patients in the TC-STSF group (p = 0.09), with no difference in demographic characteristics between the patients in whom cMRI was performed in both groups. There was no difference between the two groups regarding the size of scar in patients with ICM, as measured by late gadolinium enhancement, TC-SF: median = 6 segments, IQR = 4-7; TC-STSF: median = 8 segments, IQR = 5-9, p = 0.12.

Procedure characteristics
Procedure characteristics are shown in Table 2. There was no difference in the type of arrhythmia targeted, the approach used, procedure duration, duration of RF energy application or maximum power applied.
In patients presenting with VT, there was no difference in the number of VTs induced between both study groups (TC-SF: median = 1, IQR = 1-2; TC-STSF: median = 1, IQR = 1-2; p = 0.7). The clinical VT as well as any induced sustained VT was targeted for ablation.
When examining patients undergoing VE and VT ablation separately (Table 3)   There were no other differences in procedure duration, duration of RF energy application or maximum power applied.
An average CF greater than 5 g during RF delivery could not be reached in 4 of 164 patients (2.4%); 2 cases involved ablation in dilated RV outflow tracts, one involved the anterior mitral annulus and one involved basal LV ablation for LV summit VE.
Likewise, there was no difference in complications between TC-SF (11 of 164, 6.7%) and TC-STSF (11 of 164, 6.7%, p = 1.0) catheter groups, and this applied to VE and VT ablations (Table 3) as well as normal hearts, ICM and NICM (Table 4). There was no association between maximum CF and incidence of tamponade, but the numbers were too small to demonstrate a difference. There were two deaths within 30 days of the procedure; both patients had advanced heart failure with severely depressed LVEF (< 20%).
Tamponade requiring drainage occurred in 9 (2.7%) procedures, 8 of which resolved with pericardiocentesis and one associated with TC-STSF use required surgical repair. Of the four tamponades occurring with TC-SF, two occurred with ablation in the right ventricular outflow tract (RVOT), one with ablation in the right coronary cusp and one occurred 1.5 months later due to pericarditis following an epicardial ablation. Of the five tamponades occurring with TC-STSF, two occurred with ablation in the left ventricle, one with ablation in the aorto-mitral continuity, one with ablation in the RVOT and one with ablation in the RV free wall.

Long-term follow-up
In patients who had undergone VT ablation, there was no difference in the utilisation of antiarrhythmic medications during long-term follow-up (Table 5).

Redo of TC-SF procedures with TC-STSF
Of the 164 patients in the TC-SF catheter group, 7 patients subsequently underwent repeat ablation using the TC-STSF catheter for recurrent VA. Reasons for failure in the index procedure were adjudicated by 2 senior electrophysiologists. In two cases, recurrence of VT occurred after endocardial ablation of NICM requiring an endocardial/epicardial approach. Difficulties with mapping were identified in three cases and non-inducibility in one case. Recurrence due to mid-myocardial location of substrate occurred in one case. In two cases, the use of CF was considered to aid the success of the subsequent procedure for VEs mapped to the superior RV inflow in one and the aorto-mitral continuity in another.

Discussion
This study is the largest propensity-matched series assessing the role of CF in the ablation of VAs to date and the first to investigate whether measuring CF improves on the outcomes achieved by entire tip irrigation. The main finding from this study is that the addition of CF sensing to a 56-hole irrigation RF ablation catheter was not associated with improved acute procedural success, complications or long-term outcomes. Prior studies in sheep models and small human series suggested that measuring CF improved substrate characterisation and lesion formation, especially in scar zones [3,4]. CF achieved in different regions of the left ventricle has been found to depend on the approach used (whether transseptal or transaortic) [5]. The consistent use of both approaches to the left ventricle for patients with scar-dependent VA at our center might have lessened the effect of measuring CF. Although the size of our study is insufficient to make firm conclusions, we postulate that the use of CF may aid ablation in specific areas within the ventricles. Areas such as the inflow portion of the right ventricle, the aortomitral continuity, the subaortic vestibule (infero-septal recess), particularly when the ascending aorta is horizontally situated or tortuosity of the descending aorta is encountered, may benefit from the use of CF. In these situations, the manual feedback afforded to the operator is reduced and often requires significant catheter torque to manoeuvre the catheter. Optimising cathetertissue contact may also be of greater value with deeper intramyocardial foci. Additionally, CF sensing catheters enable the operator to visualise the direction of force which may be advantageous during epicardial ablation. Our propensity- The failure of CF sensing to offer increased efficacy in our series may be explained by the fact that there is no clear guidance on how to act on CF values in different ventricular sites of varying wall thickness. The use of surrogate markers of effective RF lesion formation, such as initial impedance fall, shown to be a correlate of good CF [6], and electrical unexcitability of ablated tissue, may have limited the relative utility of CF sensing. Other variables which influence final lesion formation are not accounted for by the CF value displayed, such as the duration of RF energy application, catheter tip angulation as well as the spatio-temporal variation in contact caused by cardiac and respiratory motion [7,8]. These might in the future be encompassed by a derived, weighted parameter combining power, time and CF, as recent reports suggest [9]. Due to the heterogeneity in tissue thickness and composition in both ventricles as well as in diverse cardiac pathologies, region-and tissue-specific parameters will be required to maximise effective lesion formation without compromising safety.
Our data shows both the TC-SF and TC-STSF catheters to be equally safe in treating VAs. This is reassuring as prior concerns were expressed regarding the safety of the TC-SF design due to a higher reported frequency of cardiac perforation as well as atrio-esophageal fistula in left atrial ablation procedures. This was corroborated by increased steam pops in canine models and in vitro simulations [10,11], prompting the manufacturer to issue a safety notice for the TC-SF catheter in 2014. The safety advantage gained by adding CF sensing to preexisting platforms may be offset by an increase in catheter stiffness, as CF-sensing catheters were associated with higher reported rates of atrio-esophageal fistula as adverse events in left atrial procedures [12]. This is concordant with our observation of a higher number of adverse events related to catheter manipulation with TC-STSF, including cardiac perforation and coronary injury due to catheter manipulation in the aortic root. This, although not statistically significant, would warrant caution while using TC-STSF in more vulnerable locations, e.g. RVOT and aortic root. CF is monitored by operators during ablation, but abrupt catheter movements during manipulation can lead to unanticipated and potentially injurious rises in CF.
Compared with the TC-SF catheter, the use of TC-STSF was associated with reduced fluoroscopy time, mainly in normal heart VA ablation. A similar observation was noted in a previous comparison of both catheters, which mainly comprised atrial ablation procedures [13]. This may indicate that operator confidence in catheter tip positioning and contact is less reliant on fluoroscopy when CF data is available especially in the outflow tracts. An alternative explanation might be that the TC-STSF cases were more recent and there has been a steady decline in fluoroscopy exposure over the past 10 years because of other factors, e.g. improvement in mapping systems. However, our finding that fluoroscopy times did not decrease with structural heart disease in the same epoch would argue against this.
Concordant with our findings, a previous observational study investigating the utility of CF in VAs using six-hole irrigation systems failed to show an improvement in procedure outcome or safety profile [14]. The findings are also analogous to the literature available on CF sensing in atrial ablation procedures. While non-randomised studies showed lower CF values to be associated with pulmonary vein reconnection [15,16], randomised controlled trials failed to demonstrate that the use of CF improves intermediate to long-term outcome [17].

Limitations
This was a retrospective single-center study. Specific parameter settings were left to operator discretion. However, casespecific ablation endpoints were the same for both catheters.
The large number of cases and the diversity of etiologies provide real-world experience of the use of both catheters.
The catheters were used mostly sequentially. Due to the lack of propensity-matched cases and as only first-time ablations were included, the study comprised a small proportion of cases using the epicardial approach and no surgical ablation cases. Image integration with pre-procedure cMRI was not available; it would have been useful to compare the area of scar as measured by cMRI and by voltage mapping both with and without CF data. As the data was collected The modest number of patients undergoing VT ablation may have been insufficient to find a small potential difference in outcomes due to the use of contact force in this subset.

Conclusion
Both TC-SF and TC-STSF catheters are safe and effective in treating VA. The addition of CF sensing to a 56-hole openirrigation RF catheter did not confer additional benefits in terms of safety, acute success or long-term outcomes in ablating VA, but reduced fluoroscopy time in normal heart VA.