Randomized trial comparing robotic to manual ablation for atrial fibrillation Waqas Ullah, MBBS, Ailsa McLean, RN, Ross J. Hunter, MBBS, PhD, Victoria Baker, RN, Laura Richmond, RN, Emily J. Cantor, MBBS, Mehul B. Dhinoja, MBBS, Simon Sporton, MBBS, MD, Mark J. Earley, MBBS, MD, Richard J. Schilling, MBBS, MD From the Cardiovascular Biomedical Research Unit, St Bartholomew’s Hospital, Barts Health NHS Trust, London, United Kingdom. BACKGROUND Catheter ablation of atrial fibrillation (AF) is a physically demanding procedure for the operator, involving radiation exposure, and has limited success rates. Remote robotic navigation (RRN) may offer benefit to the procedure, though only 1 previous small randomized trial has assessed this. OBJECTIVE This study aimed to investigate the impact of RRN on 1-year single-procedure success rates. METHODS RRN was compared to manual ablation in a randomized control trial setting by using an intention-to-treat analysis. RESULTS A total of 157 patients underwent ablation (116/157 (74%) persistent AF; 67/116 (58%) of these long-standing persistent AF). There were no significant differences between the RRN and manual groups with respect to 1-year single-procedure success rates (19/78 (24%) and 26/78 (33%), respectively; P ¼ .29), acute wide area circumferential ablation reconnection rates, complication rates, or procedure times. On multivariable analysis, fluoroscopy times were significantly shorter in the RRN group. The number of catheter displacements during ablation was lower in the RRN group, as was subjectively assessed operator fatigue. The crossover rate from RRN
Introduction Catheter ablation has become an accepted treatment modality for atrial fibrillation (AF).1 The procedure is technically and physically demanding for the operator and involves exposure of the patient and operator to radiation. Remote robotic navigation (RRN; Sensei X system, Hansen Medical Inc, Mountain View, CA) is a technology that may help meet the challenges of AF ablation. Only 1 small randomized trial comparing RRN to manual ablation (30 patients per study arm) has been published, including only patients with paroxysmal atrial fibrillation (PAF), with 6-month success rates of 73% for RRN and 77% for manual Prof Schilling has received research funding and research fellow support from Hansen Medical Inc. This research was supported by the National Institute for Health Cardiovascular Biomedical Research Unit at Barts. Address reprint requests and correspondence: Prof Richard Schilling, Cardiology Research Department, St Bartholomew’s Hospital, West Smithfield, London EC1A 7BE, United Kingdom. E-mail address: Richard. [email protected].
1547-5271/$-see front matter B 2014 Heart Rhythm Society. All rights reserved.
to manual ablation was 11/78 (14%), mainly secondary to technical problems with the RRN system. A learning curve was evident for RRN ablation: the fluoroscopy and procedure times were significantly lower after the first 10 cases in an operator’s experience. CONCLUSION This randomized trial showed no difference in the success rate for catheter ablation of AF between a RRN and manual approach. The results highlight the learning curve for RRN ablation and suggest that the use of this technology leads to an improvement in fluoroscopy times, catheter stability, and operator fatigue. KEYWORDS Atrial fibrillation; Catheter ablation; Remote robotic navigation; Manual ablation; Randomized controlled trial ABBREVIATIONS AAD ¼ antiarrhythmic drug; AF ¼ atrial fibrillation; AT ¼ atrial tachycardia; PAF ¼ paroxysmal atrial fibrillation; PeAF ¼ persistent atrial fibrillation; PV ¼ pulmonary vein; RRN ¼ remote robotic navigation; WACA ¼ wide area circumferential ablation (Heart Rhythm 2014;11:1862–1869) I 2014 Heart Rhythm Society. All rights reserved.
ablation (P ¼ .35).2 This as well as other studies suggest a benefit to RRN in terms of fluoroscopy time2–5 and catheter stability6 without affecting success rates.2,4,5 This prospective randomized study evaluated the hypothesis that the single-procedure AF ablation success rate is greater with RRN than with manual navigation.
Methods All patients gave informed consent to participate in the study, which was approved by the UK National Research Ethics Service and reported according to CONSORT guidelines.7 Consecutive patients listed for first-time catheter ablation of AF were allocated to manual or RRN ablation using computer-generated randomization. AF subtypes were classified as PAF, persistent atrial fibrillation (PeAF), or long-standing PeAF according to HRS guidelines,1 with a recruitment target of 50% patients with PAF and 50% patients with PeAF. Exclusion criteria were age o18 years, http://dx.doi.org/10.1016/j.hrthm.2014.06.026
Ullah et al
Manual vs Robotic AF Ablation
previous AF ablation, life expectancy o6 months, pregnancy, and procedural contraindications.
Ablation procedure Procedures were performed in the postabsorptive state under intravenous moderate sedation. In the manual group, 2 transseptal punctures were performed using an Endry’s coaxial (Cook Medical, Bloomington, IN) or Brockenbrough (St Jude Medical Inc, Saint Paul, MN) needle. A circular pulmonary vein (PV) mapping catheter and an irrigated ablation catheter of the operator’s choice were passed into the left atrium. In the RRN group, a 30-cm 14-F sheath was inserted into the right femoral vein (initially a shorter 14-F sheath was used, but this changed secondary to safety recommendations8). The ablation catheter was passed through an Artisan sheath (Hansen Medical Inc) and the assembly advanced under fluoroscopic guidance to the right atrium, with the catheter leading to reduced vascular trauma risk. A single transseptal puncture was performed and the sheath advanced through the septum for dilation and withdrawn, leaving only the outer needle across the septum. The ablation catheter and Artisan sheath were passed either manually or robotically into the left atrium through the transseptal puncture site and the original sheath passed alongside through the same puncture. Ablation was generally performed in a temperaturecontrolled mode (temperature limited to 481C and power to 30 W). The aim of each ablation was reduction in local bipolar electrogram amplitude by 480% or until o0.1 mV,9 or, failing this, delivery of energy for up to a minute. Procedures were guided by 3-dimensional navigation systems: EnSite Classic or EnSite Velocity (St Jude Medical Inc) and CARTO XP or CARTO 3 (Biosense Webster Inc, Diamond Bar, CA). The ablation protocol was the same in the 2 study arms. For PAF, wide area circumferential ablation (WACA) was performed to encircle ipsilateral PVs in pairs. If patients were in atrial tachycardia (AT) after WACA, the AT was mapped and ablated, but if in AF, they were electrically cardioverted. A cavotricuspid isthmus line was added only in patients with a history of typical atrial flutter. The end point for WACA was entry and exit block, which was assessed using the PV mapping catheter. Evidence of exit block was sought by pacing the PV catheter within the WACA line and confirming no conduction to the LA, although it was accepted that PV capture could not be documented in all cases. Acute WACA reconnection was assessed after a 1-hour waiting time from the completion of that WACA, with veins reisolated if necessary. In patients with PeAF, after WACA, fractionated electrograms were targeted throughout the left atria and then right atria (as described previously10) until all were abolished or sinus rhythm restored. If patients remained in AF, linear lesions were added at the mitral isthmus and roof. A cavotricuspid isthmus line was added in patients with a history of typical right atrial flutter. If at any point AF
1863 organized into AT, it was mapped and ablated. If sinus rhythm was not restored after these lesions, the patient was electrically cardioverted. The PVs were rechecked at the end of the case in sinus rhythm and reisolated where necessary. Linear lesions were checked and further lesions delivered where necessary to achieve block. During ablation, catheter stability was assessed in a semiquantitative manner similar to that described previously.11 If during ablation, the catheter position shifted, and this dislodgment was sufficient for the operator to prematurely terminate the ablation procedure or significantly readjust ablation catheter location, it was recorded as a catheter displacement. Catheter motion secondary to respiratory motion was not counted as displacement. At procedure completion, operators assessed their level of fatigue on a subjective scale ranging from 1 to 5, with 1 being no fatigue and 5 very fatigued. Four operators participated in the study. At the study’s commencement, the operators had performed a median of 400 (range 100–700) AF ablation procedures and 10 (range 2–20) RRN AF ablation procedures. No randomization was performed with respect to operator. To explore the presence of an RRN learning curve, procedures were coded in terms of how far along in each operator’s experience they were undertaken, on the basis of how many RRN ablation procedures they had performed before the trial. Therefore, if an operator had performed 5 RRN ablation procedures before the trial, their first case in the trial was coded as case 6 in their learning curve. For RRN cases, contact force measurement was done with the sheath-based IntelliSense system native to the RRN platform and ablation procedures were performed at 5–40 g of force.
Follow-up Antiarrhythmic drugs (AADs), including amiodarone, were discontinued predischarge, and all patients followed up at 3, 6, and 12 months. If the electrocardiogram at 3 months did not demonstrate AF/AT, a 7-day Holter monitor test was arranged before 6-month follow-up (unless the patient had a permanent pacemaker). Postprocedure, a 3-month blanking period was used. Patients with a recurrence of symptoms and/or documented AT/AF after this period were offered a repeat procedure. Complications were recorded from the time of the procedure until discharge and at follow-up visits. Complications prospectively investigated included cerebrovascular events, vascular access complications, pericardial effusion, tamponade, PV stenosis, the need for blood transfusion or surgery, hospital admission (or prolongation of hospital stay), and death. Complication severity is reported according to Heart Rhythm Society guidelines.1
End points The primary study end point was single-procedure success rate at 12 months, with success defined as freedom from symptomatic AF or asymptomatic AF or AT lasting Z30
1864
Heart Rhythm, Vol 11, No 11, November 2014
seconds off AADs.1 Secondary end points were comparisons of complication rates, catheter stability during ablation, procedure and fluoroscopy times and doses, and subjectively assessed operator fatigue.
subtype,1 sex, hypertension, ischemic heart disease, cerebrovascular disease, diabetes, left atrial diameter, treatment group, RRN learning curve, and mapping system) into models. For the multivariable analysis of success rates, a binary logistic regression model was used. SPSS 20 (IBM Corp, Armonk, NY) was used for statistical analysis.
Statistical analysis On the assumption of a 50% single-procedure success rate from ablation,12 the number of participants in the study who required to observe a 15% change in this success rate from RRN (power ¼ 80%; α ¼ 0.05) was 170 in each group. Patients were compared using an intention-to-treat analysis; accordingly, patients who crossed over between groups were analyzed on the basis of the treatment group to which they had been randomized. Patients randomized to a treatment group who had no procedure were excluded from the analysis because this was not believed to introduce bias into the assessment of the relative success rates of the 2 treatments.13 Patients who underwent procedures but were lost to followup were assumed treatment failures. Groups were compared using parametric and nonparametric testing based on the distribution of data (ordinal data were compared using the Mann-Whitney U test). Results are presented as mean ⫾ SD or as median (range). Categorical variables are described as count and percentage and compared using χ2 or Fisher exact tests. A P value of o.05 was considered statistically significant. Multivariable analyses was performed by simultaneously entering the chosen variables (operator, age, AF
Figure 1
Results On the basis of an interim analysis, the investigators suggested that the trial was unlikely to meet the end point of superiority of RRN over manual ablation for 12-month success rates. Consequently, the trial was terminated before targeted recruitment completion, with 166 patients randomized. For 156 patients ablated and followed up, the post hoc power of the study for the primary end point is 23.6% (α ¼ 0.05). Patient progression through the trial is displayed in Figure 1. One hundred sixty-six patients were randomized between April 2008 and August 2012. Of these, 9 were excluded because they had no ablation procedure performed. There were no significant differences in the baseline demographic characteristics of the 2 groups (Table 1). Patients with PAF had failed 3 (0–6) AADs at enrollment, and 14 of 41 patients (34%) had failed amiodarone therapy. Eleven RRN to manual crossovers occurred: in 2 cases, this occurred before the start of the procedure on the basis of the operator’s decision. Otherwise, crossovers were due to
Progression of patients from randomization to follow-up. RRN ¼ remote robotic navigation.
Ullah et al Table 1
Manual vs Robotic AF Ablation
1865
Baseline demographic characteristics
Characteristic Number of patients Age (y) Sex: male PeAF (all forms) Long-standing PeAF PAF AF duration from diagnosis (mo) Number of previous and current antiarrhythmic drugs Previous or current amiodarone use (all forms of AF) Previous or current amiodarone use (patients with PAF) Hypertension Ischemic heart disease Cerebrovascular disease Diabetes Left atrial diameter (cm) CHA2DS2VASC score
Table 2
Manual group
RRN group
78 59 ⫾ 10 62/78 (77) 58/78 (74) 29/58 (50) 20/78 (26) 48 (6–540)
79 59 ⫾ 11 61/78 (80) 58/79 (73) 38/58 (66) 21/79 (27) 48 (10–240)
.89 .85 1 .13 1 .57
2 (1–6)
3 (0–6)
.1
32/78 (41)
32/79 (41)
1
8/20 (40)
6/21 (29)
.52
10/78 (13) 6/78 (8) 6/78 (8) 6/78 (8) 4.1 ⫾ 0.7 0.9 ⫾ 1.2
8/79 (10) 4/79 (5) 6/79 (8) 6/79 (8) 4.2 ⫾ 0.7 1 ⫾ 1.2
.63 .53 1 1 .7 .47
Success rates Manual group RRN group P
Atrial Fibrillation Subtype
P
Values are presented as mean ⫾ SD, as median (range), or as n/N (%). AF ¼ atrial fibrillation; PAF ¼ paroxysmal atrial fibrillation; PeAF ¼ persistent atrial fibrillation; RRN ¼ remote robotic navigation.
RRN system malfunction (7 cases) or difficulty in usage (2 cases: one where there was difficulty crossing the transseptal puncture and the other where there was difficulty isolating the right PVs). Of those who underwent procedures, 155 patients completed follow-up, with 1 lost to follow-up. Another patient died several weeks after the procedure, with a postmortem suggesting a myocardial infarction in the context of known medically managed, triple-vessel coronary artery disease (with no atrioesophageal fistula or infarction in the ablation territory). This patient was excluded from the outcomes analysis. All patients were ablated using irrigated catheters. Twelve patients with PAF (29%: 4 manual and 8 RRN) were in arrhythmia after PV isolation and had further ablation. In patients with PeAF, a mitral isthmus line was successfully ablated in 35 of 116 (30%; manual: 13 of 58 vs RRN: 22 of 58; P ¼ .1), a roof line in 64 of 116 (55%; manual: 30 of 58 vs RRN: 34 of 58; P ¼ .6), and a cavotricuspid isthmus line in 37 of 116 (31%; manual: 20 of 58 vs RRN: 17 of 58; P ¼ .7). All attempted cavotricuspid isthmus lines were successful (bidirectional block demonstrated), and this was the case for 64 of 66 roof lines (97%) and 35 of 41 mitral isthmus lines (85%), with no significant differences between groups. Seven patients with PeAF were in sinus rhythm at the start of the procedure, and of the remainder, 27 of 109 (25%) were ablated to sinus rhythm, 9 of these 27 via an intermediary AT. There was no significant difference between groups in the number of patients ablated to sinus rhythm (manual: 18 of 54 [33%] vs RRN: 9 of 55 [16%]; P ¼ .13).
Overall 1-y success rate (off AADs) 26/78 (33) PAF 8/20 (40) PeAF 18/58 (31)
19/78 (24) .29 7/21 (33) .75 12/57 (21) .29
Values are presented as n/N (%). AAD ¼ antiarrhythmic drug; PAF ¼ paroxysmal atrial fibrillation; PeAF ¼ persistent atrial fibrillation; RRN ¼ remote robotic navigation.
There was no significant difference between the single procedure success rates for the 2 groups (Table 2). The major complication rate for the cohort was 7.6%. There was no significant difference between the groups in the combined rate of major and minor complications (Table 3). There was 1 procedure-related death in each group, both patients with PeAF. The manually ablated patient had a small cerebellar stroke and was transferred to a stroke unit. Three days after the procedure, he had a cardiac arrest and died. A postmortem demonstrated complete thrombotic occlusion of the circumflex artery at the site of a previous (10-year-old) stent. The patient in the RRN group had a posterior left atrial perforation and retroperitoneal hemorrhage secondary to a right common iliac vein perforation. Procedural factor comparisons are presented in Tables 4 and 5. Fluoroscopy and procedure times were significantly longer for patients whose procedures were performed within the first 10 cases of an operator’s learning curve (P = .012 and P = .018, respectively; Figure 2). Being within the first 10 cases did not increase the chances of crossing over and rates of major or minor complications or decrease success rates (P 4 .05, all measures). Importantly, however, the patient in the RRN group who died was early in the learning curve (10th RRN case for that operator). Table 3
Complications
Complication Type Major complications Death Stroke Retroperitoneal bleed Cardiac tamponade Pulmonary vein stenosis (major) Hematoma (major) Total patients with major complications Minor complications Pericardial effusion Hematoma (minor) Pulmonary vein stenosis (minor) Lower respiratory tract infection Total patients with minor complications
Manual group
RRN group
1/78 (1.3) 2/78 (2.6) 1/78 (1.3) 1/78 (1.3) 0
1/79 (1.3) 0 2/79 (2.5) 3/79 (3.8) 1/79 (1.3)
1 .2 .6 .4 1
1/78 (1.3) 5/78 (6.4)
2/79 (2.5) 7/79 (8.9)
1 .8
3/78 (3.8) 4/78 (5.1) 1/78 (1.3)
3/79 (3.8) 1/79 (1.3) 0
1 .2 1
2/78 (2.6)
2/79 (2.5)
.7
10/78 (12.8)
6/79 (7.7)
.3
Values are presented as n/N (%). RRN ¼ remote robotic navigation.
P
1866 Table 4
Heart Rhythm, Vol 11, No 11, November 2014 Procedural factors
Factor
Manual group
RRN group
P
Number of patients Procedure time (min) Fluoroscopy time (min) Radiation dose (cGy cm2) Time from local anesthetic administration to completion of second transseptal (min) Number of catheter displacements Operator fatigue (1–5: 1 ¼ not fatigued; 5 ¼ very fatigued)
78 273 (85–543) 50 (11–134) 4136 (474–21,243) 31 (10–90)
79 289 (159–548) 46 (17–154) 3354 (473–14,290) 43 (15–176)
.16 .34 .17 o.0005
5 (0–33) 3⫾1
o.0005 .001
1 (0–12) 2⫾1
Values are presented as median ⫾ SD or as median (range). RRN ¼ remote robotic navigation.
9% lower than those for manual ablation. Previous studies have compared success rates between RRN and manual ablation: a prior randomized trial (60 patients; RRN: 73% vs manual: 77%; P ¼ .35)2 and 2 registry studies (390 patients; RRN: 85% vs manual: 81%; P ¼ .264 and 38 patients; RRN: 91% vs manual: 81%; P ¼ .65). The lack of significant difference in success rates in the present trial is in keeping with previous results, represents only the second randomized data set, is the largest, and is first to include patients with PeAF. As with previous work,2–4 this study includes data from operators early in their RRN ablation experience and in so doing provides information regarding this phase of RRN adoption; it would be of interest to see how the results differ with exclusively highly experienced RRN operators. The single procedure success rates at 1 year (off AADs) in this trial were 37% for PAF and 26% for PeAF and overall 29% at 1 year. These are at the lower end of those reported previously but similar to those recently reported from a large European registry (40.7% for PAF, 30.2% for PeAF, and 36.7% for long-standing PeAF).15 Other reports from leading groups also describe similar success rates.14,19 Sixty percent of patients with PeAF in the present trial had been in continuous AF for at least 12 months (ie, long-lasting persistent). A longer duration of PeAF increases the risk of recurrent arrhythmia after the index procedure.17,20 Previous studies with higher success rates typically have a smaller proportion of patients with long-standing PeAF (usually under a third of the cohort)17,20,21: the higher proportion in the present study may explain the lower PeAF success rate. Patients with PAF in the present trial had failed 3 (0–6) AADs (34% failed on amiodarone). The number of failed AADs is higher than in previous studies reporting better 1year outcomes.20,22 Patients with PAF in the present trial were more advanced in the natural history of their disease than those in previous trials.
Multivariable analyses were performed for the success rates and procedure and fluoroscopy times (Table 6). Success rates and procedure times were not significantly affected by the treatment group. Fluoroscopy times were significantly shorter for patients in the RRN group. Being within the first 10 cases performed by an operator did not significantly affect the success rate from the procedure, but was associated with longer procedure and fluoroscopy times.
Discussion This randomized controlled trial showed no significant differences in single-procedure AF ablation success rates for RRN compared to manual ablation. No difference was observed in acute WACA reconnection or restoration of sinus rhythm by ablation. RRN was associated with a significant reduction in fluoroscopy times, catheter displacements, and subjective operator fatigue. No differences were observed for procedure times or complication rates. Operators in the trial had a spectrum of experience with RRN ablation, and a learning curve effect was seen in the results, with longer fluoroscopy and procedure times for cases performed in an operator’s first 10 RRN cases. The RRN to manual crossover rate was 14%. The baseline characteristics of the patients in the 2 study arms were evenly matched (Table 1). The incidence of hypertension was low than in previous studies,14,15 although not different between groups. Procedure3,16 and fluoroscopy3,17 times and acute WACA reconnection rates9 are similar to those reported for patients ablated contemporaneously in the present study. It remains to be seen whether use of more modern mapping systems, advanced contact force sensing technology and ablation targets18 will impact these parameters. There was no significant difference between the 2 groups in 1-year success rates, although rates for RRN were Table 5
Procedural factors for each wide area circumferential ablation
Factor
Manual—left side
RRN—left side
P
Manual—right side
RRN—right side
P
Successful isolation Time taken to isolate ipsilateral veins (min) Acute pulmonary vein reconnections
78 (100) 59 (12–130) 12 (15)
77 (97) 50 (15–155) 19 (25)
.5 .5 .16
78 (100) 50 (5–135) 15 (19)
75 (95) 55 (12–165) 14 (19)
.1 .44 1
Values are presented as median (range) or as n (%). RRN ¼ remote robotic navigation.
Ullah et al
Manual vs Robotic AF Ablation
1867
Figure 2 The RRN ablation learning curve for the fluoroscopy time (A) and procedure time (B). *P o .05 for comparison between patients within first 10 cases of the learning curve and those not within the first 10 cases. RRN ¼ remote robotic navigation.
The major complication rate of 7.6% is higher than the reported 4.5% in the worldwide survey,23 likely because of the more advanced AF cohort studied here. There was no significant difference in the complication rate observed Table 6
between groups. The occurrence of 2 deaths was surprising and not representative of our experiences of ablation outside the trial.24,25 Indeed, there have been more than 2000 ablation procedures for AF at our center and these are the
Multivariable analyses Dependent variable
Cofactor RRN treatment group (vs manual ablation) Left atrial diameter (cm) PAF (vs long-standing PeAF) PeAF (vs long-standing PeAF) Carto system (vs Ensite system) Within first 10 cases
Success rate
Log fluoroscopy time
Odds ratio of success (95% CI) P
Unstandardized β (95% CI) 0.2 (0.4 to 0.1)
Procedure time Unstandardized β (95% CI)
P
4.9 (34.8 to 25.1)
0.9 (0.4 to 2.5)
.87
0.3 2.9 2.4 1.8 1.4
.011 0.1 (0.03 to 0.2) .01 10.9 (9.2 to 3093) .28 .12 0.3 (0.5 to 0.1) .014 83.5 (121.6 to 45.5) o.005 .13 0.17 (0.4 to 0.01) .06 29.8 (62.9 to 3.3) .77 .31 0.4 (0.6 to 0.2) o.005 0.69 (42.2 to 28.4) .7 .72 0.4 (0.06 to 0.7) .02 62.7 (1.2 to 124.2) .046
(0.1 to 0.8) (0.8 to 10.8) (0.8 to 7.5) (0.6 to 6.0) (0.2 to 10.2)
.003
P .75
A positive β value for a cofactor means that if all other cofactors are unchanged, the dependent variable will increase as that cofactor increases (and vice versa). Model R2: success rate (Nagelkerke) 0.3; fluoroscopy time (adjusted) 0.5; procedure time (adjusted) 0.3. The results for cofactors significantly associated with the dependent variable are in bold type. Cofactors included but not significantly associated with any dependent variable: sex, age, hypertension, ischemic heart disease, cerebrovascular disease, and diabetes. The operator was also included in the models, with no significant effect in the success rate model, but with significant effects in the procedure and fluoroscopy time models. CI ¼ confidence interval; PAF ¼ paroxysmal atrial fibrillation; PeAF ¼ persistent atrial fibrillation.
1868 only 2 procedural deaths that have occurred to date. Of note, the death that occurred in the RRN arm was primarily due to atrial perforation (with concomitant retroperitoneal bleed) in a patient within the first 10 patients of an operator’s learning curve. Other groups have found an increased complication rate during their initial RRN cases.26 While no relationship has been demonstrated here between operator’s RRN experience and complication rates, increased vigilance at the start of one’s RRN experience seems prudent. Increased stability with RRN compared to manual ablation has previously been noted,6 as has a reduction in fluoroscopy times.2–5 The increased stability with RRN as well as catheter and sheath visualization provided by the system reduce the need for fluoroscopy. The present study demonstrated the presence of a learning curve for the procedure and fluoroscopy times for the first 10 cases. A learning curve for procedure times11 and fluoroscopy times4 has been noted previously, although the present study differs in reporting them at the individual operator level rather than at the institution level and in the context of a randomized cohort, which may be more informative to clinical practice. This study suggested a reduction in operator fatigue with RRN. This is intuitive since RRN procedures were performed while seated and, as the RRN control unit is outside the radiation field, without wearing a lead apron. The 14% crossover rate in this trial was driven by technical problems with RRN. The rate did not change over the course of the trial, suggesting it was not due to unfamiliarity with RRN. Crossover rates have previously been reported between 0% and 24%2,3,8,11—though only one of those reporting was a randomized trial, and in that no crossovers occurred. Combining the data from the present study and previous ones where a crossover rate was reported, the crossover rate is 8.5% (23 of 270).
Study limitations The study was terminated early compared to the recruitment target set by the a priori power calculation, as the interim analysis conducted half-way suggested no difference between groups, meaning the results in the second half of the trial would have had to be unfeasibly different to demonstrate an outcome benefit from RRN. Moreover, while the data available at the time of the study’s design indicated a 50% success rate for ablation,12 more recent data indicate that success rates are more modest,14,15 suggesting that an even larger study cohort would be needed to adequately power a study of a hypothesized superiority of RRN over manual ablation. Nevertheless, these data suggest that superiority of RRN over conventional manual ablation is improbable. While the data do not support superiority of RRN over manual ablation, the trial was not powered to assess noninferiority and this therefore is not proven by the data. The assessments of fatigue and catheter displacement were subjective and vulnerable to bias as the operators were unblinded to the treatment group. Patients were enrolled in
Heart Rhythm, Vol 11, No 11, November 2014 the study over a prolonged period of time (4 years) during which there were changes in the equipment and methods used for catheter ablation, including improvements in mapping systems. To control for the influence of factors such as the learning curve and AF subtype on the results, multivariate analyses were conducted—this would not have been necessary in a completely homogeneous population. Some of the limitations of this trial will be addressed by the ongoing Man and Machine trial comparing robotic to manual PV isolation27: the operators for this noninferiority trial will be beyond RRN learning curves and will be using contemporary ablation equipment and methods.
Conclusion This trial showed no significant difference between RRN and manual ablation with regard to single procedure AF ablation success. RRN was found to confer an advantage with respect to fluoroscopy times, subjectively assessed operator fatigue, and catheter stability. There was an initial learning curve for RRN ablation after which procedure and fluoroscopy times improved. There was a 14% crossover rate from RRN to manual ablation. Although there appear certain advantages to the RRN system, the crossover rate and lack of outcomes benefits are important considerations for the uptake of this technology.
References 1. Calkins H, Kuck KH, Cappato R, et al. 2012 HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for patient selection, procedural techniques, patient management and follow-up, definitions, endpoints, and research trial design. Heart Rhythm 2012;9:632–696. (e21). 2. Steven D, Servatius H, Rostock T, Hoffmann B, Drewitz I, Müllerleile K, Sultan A, Aydin MA, Meinertz T, Willems S. Reduced fluoroscopy during atrial fibrillation ablation: benefits of robotic guided navigation. J Cardiovasc Electrophysiol 2010;21:6–12. 3. Thomas D, Scholz EP, Schweizer PA, Katus HA, Becker R. Initial experience with robotic navigation for catheter ablation of paroxysmal and persistent atrial fibrillation. J Electrocardiol 2012;45:95–101. 4. Di Biase L, Wang Y, Horton R, et al. Ablation of atrial fibrillation utilizing robotic catheter navigation in comparison to manual navigation and ablation: single-center experience. J Cardiovasc Electrophysiol 2009;20:1328–1335. 5. Kautzner J, Peichl P, Čihák R, Wichterle D, Mlčochová H. Early experience with robotic navigation for catheter ablation of paroxysmal atrial fibrillation. Pacing Clin Electrophysiol 2009;32:S163–S166. 6. Malcolme-Lawes LC, Lim PB, Koa-Wing M, Whinnett ZI, Jamil-Copley S, Hayat S, Francis DP, Kojodjojo P, Davies DW, Peters NS, Kanagaratnam P. Robotic assistance and general anaesthesia improve catheter stability and increase signal attenuation during atrial fibrillation ablation. Europace 2013;15:41–47. 7. Schulz KF, Altman DG, Moher D, for the CONSORT Group. CONSORT 2010 Statement: updated guidelines for reporting parallel group randomised trials. BMJ 2010;340:c332–c332. 8. Wazni OM, Barrett C, Martin DO, et al. Experience with the Hansen robotic system for atrial fibrillation ablation—lessons learned and techniques modified: Hansen in the real world. J Cardiovasc Electrophysiol 2009;20:1193–1196. 9. Rajappan K, Kistler PM, Earley MJ, Thomas G, Izquierdo M, Sporton SC, Schilling RJ. Acute and chronic pulmonary vein reconnection after atrial fibrillation ablation: a prospective characterization of anatomical sites. Pacing Clin Electrophysiol 2008;31:1598–1605. 10. Elayi CS, Verma A, DiBiase L, et al. Ablation for longstanding permanent atrial fibrillation: results from a randomized study comparing three different strategies. Heart Rhythm 2008;5:1658–1664. 11. Schmidt B, Tilz RR, Neven K, Chun KRJ, Fürnkranz A, Ouyang F. Remote robotic navigation and electroanatomical mapping for ablation of atrial fibrillation: considerations for navigation and impact on procedural outcome. Circ Arrhythm Electrophysiol 2009;2:120–128.
Ullah et al
Manual vs Robotic AF Ablation
12. Cappato R. Worldwide survey on the methods, efficacy, and safety of catheter ablation for human atrial fibrillation. Circulation 2005;111:1100–1105. 13. Hollis S, Campbell F. What is meant by intention to treat analysis? Survey of published randomised controlled trials. BMJ 1999;319:670–674. 14. Weerasooriya R, Khairy P, Litalien J, et al. Catheter ablation for atrial fibrillation: are results maintained at 5 years of follow-up? J Am Coll Cardiol 2011;57: 160–166. 15. Arbelo E, Brugada J, Hindricks G, et al. The atrial fibrillation ablation pilot study: an European Survey on Methodology and results of catheter ablation for atrial fibrillation conducted by the European Heart Rhythm Association. Eur Heart J 2014;35:1466–1478. 16. Oral H, Chugh A, Yoshida K, et al. A randomized assessment of the incremental role of ablation of complex fractionated atrial electrograms after antral pulmonary vein isolation for long-lasting persistent atrial fibrillation. J Am Coll Cardiol 2009;53:782–789. 17. Rostock T, Salukhe TV, Steven D, et al. Long-term single- and multipleprocedure outcome and predictors of success after catheter ablation for persistent atrial fibrillation. Heart Rhythm 2011;8:1391–1397. 18. Ullah W, Hunter RJ, Baker V, Dhinoja M, Sporton S, Earley MJ, Schilling RJ. Target indices for clinical ablation in atrial fibrillation: insights from contact force, electrogram and biophysical parameter analysis. Circ Arrhythm Electrophysiol 2014;7:63–68. 19. Morillo CA, Verma A, Connolly SJ, et al. Radiofrequency ablation vs antiarrhythmic drugs as first-line treatment of paroxysmal atrial fibrillation (RAAFT-2): a randomized trial. JAMA 2014;311:692–700.
1869 20. Winkle RA, Mead RH, Engel G, Patrawala RA. Long-term results of atrial fibrillation ablation: the importance of all initial ablation failures undergoing a repeat ablation. Am Heart J 2011;162:193–200. 21. Hussein AA, Saliba WI, Martin DO, et al. Natural history and long-term outcomes of ablated atrial fibrillation. Circ Arrhythm Electrophysiol 2011;4:271–278. 22. Medi C, Sparks P, Morton J, Kistler P, Halloran K, Rosso R, Vohra J, Kumar S, Kalman J. Pulmonary vein antral isolation for paroxysmal atrial fibrillation: results from long-term follow-up. J Cardiovasc Electrophysiol 2011;22:137–141. 23. Cappato R, Calkins H, Chen S-A, et al. Updated worldwide survey on the methods, efficacy, and safety of catheter ablation for human atrial fibrillation. Circ Arrhythm Electrophysiol 2010;3:32–38. 24. Hunter RJ, Berriman TJ, Diab I, et al. Long-term efficacy of catheter ablation for atrial fibrillation: impact of additional targeting of fractionated electrograms. Heart 2010;96:1372–1378. 25. Tayebjee MH, Creta A, Moder S, Hunter RJ, Earley MJ, Dhinoja MB, Schilling RJ. Impact of angiotensin-converting enzyme-inhibitors and angiotensin receptor blockers on long-term outcome of catheter ablation for atrial fibrillation. Europace 2010;12:1537–1542. 26. Bai R, DI Biase L, Valderrabano M, et al. Worldwide experience with the robotic navigation system in catheter ablation of atrial fibrillation: methodology, efficacy and safety. J Cardiovasc Electrophysiol 2012;23:820–826. 27. Rillig A, Schmidt B, Steven D, et al. Study design of the man and machine trial: a prospective international controlled noninferiority trial comparing manual with robotic catheter ablation for treatment of atrial fibrillation. J Cardiovasc Electrophysiol 2013;24:40–46.
Introduction Catheter ablation has become an accepted treatment modality for atrial fibrillation (AF).1 The procedure is technically and physically demanding for the operator and involves exposure of the patient and operator to radiation. Remote robotic navigation (RRN; Sensei X system, Hansen Medical Inc, Mountain View, CA) is a technology that may help meet the challenges of AF ablation. Only 1 small randomized trial comparing RRN to manual ablation (30 patients per study arm) has been published, including only patients with paroxysmal atrial fibrillation (PAF), with 6-month success rates of 73% for RRN and 77% for manual Prof Schilling has received research funding and research fellow support from Hansen Medical Inc. This research was supported by the National Institute for Health Cardiovascular Biomedical Research Unit at Barts. Address reprint requests and correspondence: Prof Richard Schilling, Cardiology Research Department, St Bartholomew’s Hospital, West Smithfield, London EC1A 7BE, United Kingdom. E-mail address: Richard. [email protected].
1547-5271/$-see front matter B 2014 Heart Rhythm Society. All rights reserved.
to manual ablation was 11/78 (14%), mainly secondary to technical problems with the RRN system. A learning curve was evident for RRN ablation: the fluoroscopy and procedure times were significantly lower after the first 10 cases in an operator’s experience. CONCLUSION This randomized trial showed no difference in the success rate for catheter ablation of AF between a RRN and manual approach. The results highlight the learning curve for RRN ablation and suggest that the use of this technology leads to an improvement in fluoroscopy times, catheter stability, and operator fatigue. KEYWORDS Atrial fibrillation; Catheter ablation; Remote robotic navigation; Manual ablation; Randomized controlled trial ABBREVIATIONS AAD ¼ antiarrhythmic drug; AF ¼ atrial fibrillation; AT ¼ atrial tachycardia; PAF ¼ paroxysmal atrial fibrillation; PeAF ¼ persistent atrial fibrillation; PV ¼ pulmonary vein; RRN ¼ remote robotic navigation; WACA ¼ wide area circumferential ablation (Heart Rhythm 2014;11:1862–1869) I 2014 Heart Rhythm Society. All rights reserved.
ablation (P ¼ .35).2 This as well as other studies suggest a benefit to RRN in terms of fluoroscopy time2–5 and catheter stability6 without affecting success rates.2,4,5 This prospective randomized study evaluated the hypothesis that the single-procedure AF ablation success rate is greater with RRN than with manual navigation.
Methods All patients gave informed consent to participate in the study, which was approved by the UK National Research Ethics Service and reported according to CONSORT guidelines.7 Consecutive patients listed for first-time catheter ablation of AF were allocated to manual or RRN ablation using computer-generated randomization. AF subtypes were classified as PAF, persistent atrial fibrillation (PeAF), or long-standing PeAF according to HRS guidelines,1 with a recruitment target of 50% patients with PAF and 50% patients with PeAF. Exclusion criteria were age o18 years, http://dx.doi.org/10.1016/j.hrthm.2014.06.026
Ullah et al
Manual vs Robotic AF Ablation
previous AF ablation, life expectancy o6 months, pregnancy, and procedural contraindications.
Ablation procedure Procedures were performed in the postabsorptive state under intravenous moderate sedation. In the manual group, 2 transseptal punctures were performed using an Endry’s coaxial (Cook Medical, Bloomington, IN) or Brockenbrough (St Jude Medical Inc, Saint Paul, MN) needle. A circular pulmonary vein (PV) mapping catheter and an irrigated ablation catheter of the operator’s choice were passed into the left atrium. In the RRN group, a 30-cm 14-F sheath was inserted into the right femoral vein (initially a shorter 14-F sheath was used, but this changed secondary to safety recommendations8). The ablation catheter was passed through an Artisan sheath (Hansen Medical Inc) and the assembly advanced under fluoroscopic guidance to the right atrium, with the catheter leading to reduced vascular trauma risk. A single transseptal puncture was performed and the sheath advanced through the septum for dilation and withdrawn, leaving only the outer needle across the septum. The ablation catheter and Artisan sheath were passed either manually or robotically into the left atrium through the transseptal puncture site and the original sheath passed alongside through the same puncture. Ablation was generally performed in a temperaturecontrolled mode (temperature limited to 481C and power to 30 W). The aim of each ablation was reduction in local bipolar electrogram amplitude by 480% or until o0.1 mV,9 or, failing this, delivery of energy for up to a minute. Procedures were guided by 3-dimensional navigation systems: EnSite Classic or EnSite Velocity (St Jude Medical Inc) and CARTO XP or CARTO 3 (Biosense Webster Inc, Diamond Bar, CA). The ablation protocol was the same in the 2 study arms. For PAF, wide area circumferential ablation (WACA) was performed to encircle ipsilateral PVs in pairs. If patients were in atrial tachycardia (AT) after WACA, the AT was mapped and ablated, but if in AF, they were electrically cardioverted. A cavotricuspid isthmus line was added only in patients with a history of typical atrial flutter. The end point for WACA was entry and exit block, which was assessed using the PV mapping catheter. Evidence of exit block was sought by pacing the PV catheter within the WACA line and confirming no conduction to the LA, although it was accepted that PV capture could not be documented in all cases. Acute WACA reconnection was assessed after a 1-hour waiting time from the completion of that WACA, with veins reisolated if necessary. In patients with PeAF, after WACA, fractionated electrograms were targeted throughout the left atria and then right atria (as described previously10) until all were abolished or sinus rhythm restored. If patients remained in AF, linear lesions were added at the mitral isthmus and roof. A cavotricuspid isthmus line was added in patients with a history of typical right atrial flutter. If at any point AF
1863 organized into AT, it was mapped and ablated. If sinus rhythm was not restored after these lesions, the patient was electrically cardioverted. The PVs were rechecked at the end of the case in sinus rhythm and reisolated where necessary. Linear lesions were checked and further lesions delivered where necessary to achieve block. During ablation, catheter stability was assessed in a semiquantitative manner similar to that described previously.11 If during ablation, the catheter position shifted, and this dislodgment was sufficient for the operator to prematurely terminate the ablation procedure or significantly readjust ablation catheter location, it was recorded as a catheter displacement. Catheter motion secondary to respiratory motion was not counted as displacement. At procedure completion, operators assessed their level of fatigue on a subjective scale ranging from 1 to 5, with 1 being no fatigue and 5 very fatigued. Four operators participated in the study. At the study’s commencement, the operators had performed a median of 400 (range 100–700) AF ablation procedures and 10 (range 2–20) RRN AF ablation procedures. No randomization was performed with respect to operator. To explore the presence of an RRN learning curve, procedures were coded in terms of how far along in each operator’s experience they were undertaken, on the basis of how many RRN ablation procedures they had performed before the trial. Therefore, if an operator had performed 5 RRN ablation procedures before the trial, their first case in the trial was coded as case 6 in their learning curve. For RRN cases, contact force measurement was done with the sheath-based IntelliSense system native to the RRN platform and ablation procedures were performed at 5–40 g of force.
Follow-up Antiarrhythmic drugs (AADs), including amiodarone, were discontinued predischarge, and all patients followed up at 3, 6, and 12 months. If the electrocardiogram at 3 months did not demonstrate AF/AT, a 7-day Holter monitor test was arranged before 6-month follow-up (unless the patient had a permanent pacemaker). Postprocedure, a 3-month blanking period was used. Patients with a recurrence of symptoms and/or documented AT/AF after this period were offered a repeat procedure. Complications were recorded from the time of the procedure until discharge and at follow-up visits. Complications prospectively investigated included cerebrovascular events, vascular access complications, pericardial effusion, tamponade, PV stenosis, the need for blood transfusion or surgery, hospital admission (or prolongation of hospital stay), and death. Complication severity is reported according to Heart Rhythm Society guidelines.1
End points The primary study end point was single-procedure success rate at 12 months, with success defined as freedom from symptomatic AF or asymptomatic AF or AT lasting Z30
1864
Heart Rhythm, Vol 11, No 11, November 2014
seconds off AADs.1 Secondary end points were comparisons of complication rates, catheter stability during ablation, procedure and fluoroscopy times and doses, and subjectively assessed operator fatigue.
subtype,1 sex, hypertension, ischemic heart disease, cerebrovascular disease, diabetes, left atrial diameter, treatment group, RRN learning curve, and mapping system) into models. For the multivariable analysis of success rates, a binary logistic regression model was used. SPSS 20 (IBM Corp, Armonk, NY) was used for statistical analysis.
Statistical analysis On the assumption of a 50% single-procedure success rate from ablation,12 the number of participants in the study who required to observe a 15% change in this success rate from RRN (power ¼ 80%; α ¼ 0.05) was 170 in each group. Patients were compared using an intention-to-treat analysis; accordingly, patients who crossed over between groups were analyzed on the basis of the treatment group to which they had been randomized. Patients randomized to a treatment group who had no procedure were excluded from the analysis because this was not believed to introduce bias into the assessment of the relative success rates of the 2 treatments.13 Patients who underwent procedures but were lost to followup were assumed treatment failures. Groups were compared using parametric and nonparametric testing based on the distribution of data (ordinal data were compared using the Mann-Whitney U test). Results are presented as mean ⫾ SD or as median (range). Categorical variables are described as count and percentage and compared using χ2 or Fisher exact tests. A P value of o.05 was considered statistically significant. Multivariable analyses was performed by simultaneously entering the chosen variables (operator, age, AF
Figure 1
Results On the basis of an interim analysis, the investigators suggested that the trial was unlikely to meet the end point of superiority of RRN over manual ablation for 12-month success rates. Consequently, the trial was terminated before targeted recruitment completion, with 166 patients randomized. For 156 patients ablated and followed up, the post hoc power of the study for the primary end point is 23.6% (α ¼ 0.05). Patient progression through the trial is displayed in Figure 1. One hundred sixty-six patients were randomized between April 2008 and August 2012. Of these, 9 were excluded because they had no ablation procedure performed. There were no significant differences in the baseline demographic characteristics of the 2 groups (Table 1). Patients with PAF had failed 3 (0–6) AADs at enrollment, and 14 of 41 patients (34%) had failed amiodarone therapy. Eleven RRN to manual crossovers occurred: in 2 cases, this occurred before the start of the procedure on the basis of the operator’s decision. Otherwise, crossovers were due to
Progression of patients from randomization to follow-up. RRN ¼ remote robotic navigation.
Ullah et al Table 1
Manual vs Robotic AF Ablation
1865
Baseline demographic characteristics
Characteristic Number of patients Age (y) Sex: male PeAF (all forms) Long-standing PeAF PAF AF duration from diagnosis (mo) Number of previous and current antiarrhythmic drugs Previous or current amiodarone use (all forms of AF) Previous or current amiodarone use (patients with PAF) Hypertension Ischemic heart disease Cerebrovascular disease Diabetes Left atrial diameter (cm) CHA2DS2VASC score
Table 2
Manual group
RRN group
78 59 ⫾ 10 62/78 (77) 58/78 (74) 29/58 (50) 20/78 (26) 48 (6–540)
79 59 ⫾ 11 61/78 (80) 58/79 (73) 38/58 (66) 21/79 (27) 48 (10–240)
.89 .85 1 .13 1 .57
2 (1–6)
3 (0–6)
.1
32/78 (41)
32/79 (41)
1
8/20 (40)
6/21 (29)
.52
10/78 (13) 6/78 (8) 6/78 (8) 6/78 (8) 4.1 ⫾ 0.7 0.9 ⫾ 1.2
8/79 (10) 4/79 (5) 6/79 (8) 6/79 (8) 4.2 ⫾ 0.7 1 ⫾ 1.2
.63 .53 1 1 .7 .47
Success rates Manual group RRN group P
Atrial Fibrillation Subtype
P
Values are presented as mean ⫾ SD, as median (range), or as n/N (%). AF ¼ atrial fibrillation; PAF ¼ paroxysmal atrial fibrillation; PeAF ¼ persistent atrial fibrillation; RRN ¼ remote robotic navigation.
RRN system malfunction (7 cases) or difficulty in usage (2 cases: one where there was difficulty crossing the transseptal puncture and the other where there was difficulty isolating the right PVs). Of those who underwent procedures, 155 patients completed follow-up, with 1 lost to follow-up. Another patient died several weeks after the procedure, with a postmortem suggesting a myocardial infarction in the context of known medically managed, triple-vessel coronary artery disease (with no atrioesophageal fistula or infarction in the ablation territory). This patient was excluded from the outcomes analysis. All patients were ablated using irrigated catheters. Twelve patients with PAF (29%: 4 manual and 8 RRN) were in arrhythmia after PV isolation and had further ablation. In patients with PeAF, a mitral isthmus line was successfully ablated in 35 of 116 (30%; manual: 13 of 58 vs RRN: 22 of 58; P ¼ .1), a roof line in 64 of 116 (55%; manual: 30 of 58 vs RRN: 34 of 58; P ¼ .6), and a cavotricuspid isthmus line in 37 of 116 (31%; manual: 20 of 58 vs RRN: 17 of 58; P ¼ .7). All attempted cavotricuspid isthmus lines were successful (bidirectional block demonstrated), and this was the case for 64 of 66 roof lines (97%) and 35 of 41 mitral isthmus lines (85%), with no significant differences between groups. Seven patients with PeAF were in sinus rhythm at the start of the procedure, and of the remainder, 27 of 109 (25%) were ablated to sinus rhythm, 9 of these 27 via an intermediary AT. There was no significant difference between groups in the number of patients ablated to sinus rhythm (manual: 18 of 54 [33%] vs RRN: 9 of 55 [16%]; P ¼ .13).
Overall 1-y success rate (off AADs) 26/78 (33) PAF 8/20 (40) PeAF 18/58 (31)
19/78 (24) .29 7/21 (33) .75 12/57 (21) .29
Values are presented as n/N (%). AAD ¼ antiarrhythmic drug; PAF ¼ paroxysmal atrial fibrillation; PeAF ¼ persistent atrial fibrillation; RRN ¼ remote robotic navigation.
There was no significant difference between the single procedure success rates for the 2 groups (Table 2). The major complication rate for the cohort was 7.6%. There was no significant difference between the groups in the combined rate of major and minor complications (Table 3). There was 1 procedure-related death in each group, both patients with PeAF. The manually ablated patient had a small cerebellar stroke and was transferred to a stroke unit. Three days after the procedure, he had a cardiac arrest and died. A postmortem demonstrated complete thrombotic occlusion of the circumflex artery at the site of a previous (10-year-old) stent. The patient in the RRN group had a posterior left atrial perforation and retroperitoneal hemorrhage secondary to a right common iliac vein perforation. Procedural factor comparisons are presented in Tables 4 and 5. Fluoroscopy and procedure times were significantly longer for patients whose procedures were performed within the first 10 cases of an operator’s learning curve (P = .012 and P = .018, respectively; Figure 2). Being within the first 10 cases did not increase the chances of crossing over and rates of major or minor complications or decrease success rates (P 4 .05, all measures). Importantly, however, the patient in the RRN group who died was early in the learning curve (10th RRN case for that operator). Table 3
Complications
Complication Type Major complications Death Stroke Retroperitoneal bleed Cardiac tamponade Pulmonary vein stenosis (major) Hematoma (major) Total patients with major complications Minor complications Pericardial effusion Hematoma (minor) Pulmonary vein stenosis (minor) Lower respiratory tract infection Total patients with minor complications
Manual group
RRN group
1/78 (1.3) 2/78 (2.6) 1/78 (1.3) 1/78 (1.3) 0
1/79 (1.3) 0 2/79 (2.5) 3/79 (3.8) 1/79 (1.3)
1 .2 .6 .4 1
1/78 (1.3) 5/78 (6.4)
2/79 (2.5) 7/79 (8.9)
1 .8
3/78 (3.8) 4/78 (5.1) 1/78 (1.3)
3/79 (3.8) 1/79 (1.3) 0
1 .2 1
2/78 (2.6)
2/79 (2.5)
.7
10/78 (12.8)
6/79 (7.7)
.3
Values are presented as n/N (%). RRN ¼ remote robotic navigation.
P
1866 Table 4
Heart Rhythm, Vol 11, No 11, November 2014 Procedural factors
Factor
Manual group
RRN group
P
Number of patients Procedure time (min) Fluoroscopy time (min) Radiation dose (cGy cm2) Time from local anesthetic administration to completion of second transseptal (min) Number of catheter displacements Operator fatigue (1–5: 1 ¼ not fatigued; 5 ¼ very fatigued)
78 273 (85–543) 50 (11–134) 4136 (474–21,243) 31 (10–90)
79 289 (159–548) 46 (17–154) 3354 (473–14,290) 43 (15–176)
.16 .34 .17 o.0005
5 (0–33) 3⫾1
o.0005 .001
1 (0–12) 2⫾1
Values are presented as median ⫾ SD or as median (range). RRN ¼ remote robotic navigation.
9% lower than those for manual ablation. Previous studies have compared success rates between RRN and manual ablation: a prior randomized trial (60 patients; RRN: 73% vs manual: 77%; P ¼ .35)2 and 2 registry studies (390 patients; RRN: 85% vs manual: 81%; P ¼ .264 and 38 patients; RRN: 91% vs manual: 81%; P ¼ .65). The lack of significant difference in success rates in the present trial is in keeping with previous results, represents only the second randomized data set, is the largest, and is first to include patients with PeAF. As with previous work,2–4 this study includes data from operators early in their RRN ablation experience and in so doing provides information regarding this phase of RRN adoption; it would be of interest to see how the results differ with exclusively highly experienced RRN operators. The single procedure success rates at 1 year (off AADs) in this trial were 37% for PAF and 26% for PeAF and overall 29% at 1 year. These are at the lower end of those reported previously but similar to those recently reported from a large European registry (40.7% for PAF, 30.2% for PeAF, and 36.7% for long-standing PeAF).15 Other reports from leading groups also describe similar success rates.14,19 Sixty percent of patients with PeAF in the present trial had been in continuous AF for at least 12 months (ie, long-lasting persistent). A longer duration of PeAF increases the risk of recurrent arrhythmia after the index procedure.17,20 Previous studies with higher success rates typically have a smaller proportion of patients with long-standing PeAF (usually under a third of the cohort)17,20,21: the higher proportion in the present study may explain the lower PeAF success rate. Patients with PAF in the present trial had failed 3 (0–6) AADs (34% failed on amiodarone). The number of failed AADs is higher than in previous studies reporting better 1year outcomes.20,22 Patients with PAF in the present trial were more advanced in the natural history of their disease than those in previous trials.
Multivariable analyses were performed for the success rates and procedure and fluoroscopy times (Table 6). Success rates and procedure times were not significantly affected by the treatment group. Fluoroscopy times were significantly shorter for patients in the RRN group. Being within the first 10 cases performed by an operator did not significantly affect the success rate from the procedure, but was associated with longer procedure and fluoroscopy times.
Discussion This randomized controlled trial showed no significant differences in single-procedure AF ablation success rates for RRN compared to manual ablation. No difference was observed in acute WACA reconnection or restoration of sinus rhythm by ablation. RRN was associated with a significant reduction in fluoroscopy times, catheter displacements, and subjective operator fatigue. No differences were observed for procedure times or complication rates. Operators in the trial had a spectrum of experience with RRN ablation, and a learning curve effect was seen in the results, with longer fluoroscopy and procedure times for cases performed in an operator’s first 10 RRN cases. The RRN to manual crossover rate was 14%. The baseline characteristics of the patients in the 2 study arms were evenly matched (Table 1). The incidence of hypertension was low than in previous studies,14,15 although not different between groups. Procedure3,16 and fluoroscopy3,17 times and acute WACA reconnection rates9 are similar to those reported for patients ablated contemporaneously in the present study. It remains to be seen whether use of more modern mapping systems, advanced contact force sensing technology and ablation targets18 will impact these parameters. There was no significant difference between the 2 groups in 1-year success rates, although rates for RRN were Table 5
Procedural factors for each wide area circumferential ablation
Factor
Manual—left side
RRN—left side
P
Manual—right side
RRN—right side
P
Successful isolation Time taken to isolate ipsilateral veins (min) Acute pulmonary vein reconnections
78 (100) 59 (12–130) 12 (15)
77 (97) 50 (15–155) 19 (25)
.5 .5 .16
78 (100) 50 (5–135) 15 (19)
75 (95) 55 (12–165) 14 (19)
.1 .44 1
Values are presented as median (range) or as n (%). RRN ¼ remote robotic navigation.
Ullah et al
Manual vs Robotic AF Ablation
1867
Figure 2 The RRN ablation learning curve for the fluoroscopy time (A) and procedure time (B). *P o .05 for comparison between patients within first 10 cases of the learning curve and those not within the first 10 cases. RRN ¼ remote robotic navigation.
The major complication rate of 7.6% is higher than the reported 4.5% in the worldwide survey,23 likely because of the more advanced AF cohort studied here. There was no significant difference in the complication rate observed Table 6
between groups. The occurrence of 2 deaths was surprising and not representative of our experiences of ablation outside the trial.24,25 Indeed, there have been more than 2000 ablation procedures for AF at our center and these are the
Multivariable analyses Dependent variable
Cofactor RRN treatment group (vs manual ablation) Left atrial diameter (cm) PAF (vs long-standing PeAF) PeAF (vs long-standing PeAF) Carto system (vs Ensite system) Within first 10 cases
Success rate
Log fluoroscopy time
Odds ratio of success (95% CI) P
Unstandardized β (95% CI) 0.2 (0.4 to 0.1)
Procedure time Unstandardized β (95% CI)
P
4.9 (34.8 to 25.1)
0.9 (0.4 to 2.5)
.87
0.3 2.9 2.4 1.8 1.4
.011 0.1 (0.03 to 0.2) .01 10.9 (9.2 to 3093) .28 .12 0.3 (0.5 to 0.1) .014 83.5 (121.6 to 45.5) o.005 .13 0.17 (0.4 to 0.01) .06 29.8 (62.9 to 3.3) .77 .31 0.4 (0.6 to 0.2) o.005 0.69 (42.2 to 28.4) .7 .72 0.4 (0.06 to 0.7) .02 62.7 (1.2 to 124.2) .046
(0.1 to 0.8) (0.8 to 10.8) (0.8 to 7.5) (0.6 to 6.0) (0.2 to 10.2)
.003
P .75
A positive β value for a cofactor means that if all other cofactors are unchanged, the dependent variable will increase as that cofactor increases (and vice versa). Model R2: success rate (Nagelkerke) 0.3; fluoroscopy time (adjusted) 0.5; procedure time (adjusted) 0.3. The results for cofactors significantly associated with the dependent variable are in bold type. Cofactors included but not significantly associated with any dependent variable: sex, age, hypertension, ischemic heart disease, cerebrovascular disease, and diabetes. The operator was also included in the models, with no significant effect in the success rate model, but with significant effects in the procedure and fluoroscopy time models. CI ¼ confidence interval; PAF ¼ paroxysmal atrial fibrillation; PeAF ¼ persistent atrial fibrillation.
1868 only 2 procedural deaths that have occurred to date. Of note, the death that occurred in the RRN arm was primarily due to atrial perforation (with concomitant retroperitoneal bleed) in a patient within the first 10 patients of an operator’s learning curve. Other groups have found an increased complication rate during their initial RRN cases.26 While no relationship has been demonstrated here between operator’s RRN experience and complication rates, increased vigilance at the start of one’s RRN experience seems prudent. Increased stability with RRN compared to manual ablation has previously been noted,6 as has a reduction in fluoroscopy times.2–5 The increased stability with RRN as well as catheter and sheath visualization provided by the system reduce the need for fluoroscopy. The present study demonstrated the presence of a learning curve for the procedure and fluoroscopy times for the first 10 cases. A learning curve for procedure times11 and fluoroscopy times4 has been noted previously, although the present study differs in reporting them at the individual operator level rather than at the institution level and in the context of a randomized cohort, which may be more informative to clinical practice. This study suggested a reduction in operator fatigue with RRN. This is intuitive since RRN procedures were performed while seated and, as the RRN control unit is outside the radiation field, without wearing a lead apron. The 14% crossover rate in this trial was driven by technical problems with RRN. The rate did not change over the course of the trial, suggesting it was not due to unfamiliarity with RRN. Crossover rates have previously been reported between 0% and 24%2,3,8,11—though only one of those reporting was a randomized trial, and in that no crossovers occurred. Combining the data from the present study and previous ones where a crossover rate was reported, the crossover rate is 8.5% (23 of 270).
Study limitations The study was terminated early compared to the recruitment target set by the a priori power calculation, as the interim analysis conducted half-way suggested no difference between groups, meaning the results in the second half of the trial would have had to be unfeasibly different to demonstrate an outcome benefit from RRN. Moreover, while the data available at the time of the study’s design indicated a 50% success rate for ablation,12 more recent data indicate that success rates are more modest,14,15 suggesting that an even larger study cohort would be needed to adequately power a study of a hypothesized superiority of RRN over manual ablation. Nevertheless, these data suggest that superiority of RRN over conventional manual ablation is improbable. While the data do not support superiority of RRN over manual ablation, the trial was not powered to assess noninferiority and this therefore is not proven by the data. The assessments of fatigue and catheter displacement were subjective and vulnerable to bias as the operators were unblinded to the treatment group. Patients were enrolled in
Heart Rhythm, Vol 11, No 11, November 2014 the study over a prolonged period of time (4 years) during which there were changes in the equipment and methods used for catheter ablation, including improvements in mapping systems. To control for the influence of factors such as the learning curve and AF subtype on the results, multivariate analyses were conducted—this would not have been necessary in a completely homogeneous population. Some of the limitations of this trial will be addressed by the ongoing Man and Machine trial comparing robotic to manual PV isolation27: the operators for this noninferiority trial will be beyond RRN learning curves and will be using contemporary ablation equipment and methods.
Conclusion This trial showed no significant difference between RRN and manual ablation with regard to single procedure AF ablation success. RRN was found to confer an advantage with respect to fluoroscopy times, subjectively assessed operator fatigue, and catheter stability. There was an initial learning curve for RRN ablation after which procedure and fluoroscopy times improved. There was a 14% crossover rate from RRN to manual ablation. Although there appear certain advantages to the RRN system, the crossover rate and lack of outcomes benefits are important considerations for the uptake of this technology.
References 1. Calkins H, Kuck KH, Cappato R, et al. 2012 HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for patient selection, procedural techniques, patient management and follow-up, definitions, endpoints, and research trial design. Heart Rhythm 2012;9:632–696. (e21). 2. Steven D, Servatius H, Rostock T, Hoffmann B, Drewitz I, Müllerleile K, Sultan A, Aydin MA, Meinertz T, Willems S. Reduced fluoroscopy during atrial fibrillation ablation: benefits of robotic guided navigation. J Cardiovasc Electrophysiol 2010;21:6–12. 3. Thomas D, Scholz EP, Schweizer PA, Katus HA, Becker R. Initial experience with robotic navigation for catheter ablation of paroxysmal and persistent atrial fibrillation. J Electrocardiol 2012;45:95–101. 4. Di Biase L, Wang Y, Horton R, et al. Ablation of atrial fibrillation utilizing robotic catheter navigation in comparison to manual navigation and ablation: single-center experience. J Cardiovasc Electrophysiol 2009;20:1328–1335. 5. Kautzner J, Peichl P, Čihák R, Wichterle D, Mlčochová H. Early experience with robotic navigation for catheter ablation of paroxysmal atrial fibrillation. Pacing Clin Electrophysiol 2009;32:S163–S166. 6. Malcolme-Lawes LC, Lim PB, Koa-Wing M, Whinnett ZI, Jamil-Copley S, Hayat S, Francis DP, Kojodjojo P, Davies DW, Peters NS, Kanagaratnam P. Robotic assistance and general anaesthesia improve catheter stability and increase signal attenuation during atrial fibrillation ablation. Europace 2013;15:41–47. 7. Schulz KF, Altman DG, Moher D, for the CONSORT Group. CONSORT 2010 Statement: updated guidelines for reporting parallel group randomised trials. BMJ 2010;340:c332–c332. 8. Wazni OM, Barrett C, Martin DO, et al. Experience with the Hansen robotic system for atrial fibrillation ablation—lessons learned and techniques modified: Hansen in the real world. J Cardiovasc Electrophysiol 2009;20:1193–1196. 9. Rajappan K, Kistler PM, Earley MJ, Thomas G, Izquierdo M, Sporton SC, Schilling RJ. Acute and chronic pulmonary vein reconnection after atrial fibrillation ablation: a prospective characterization of anatomical sites. Pacing Clin Electrophysiol 2008;31:1598–1605. 10. Elayi CS, Verma A, DiBiase L, et al. Ablation for longstanding permanent atrial fibrillation: results from a randomized study comparing three different strategies. Heart Rhythm 2008;5:1658–1664. 11. Schmidt B, Tilz RR, Neven K, Chun KRJ, Fürnkranz A, Ouyang F. Remote robotic navigation and electroanatomical mapping for ablation of atrial fibrillation: considerations for navigation and impact on procedural outcome. Circ Arrhythm Electrophysiol 2009;2:120–128.
Ullah et al
Manual vs Robotic AF Ablation
12. Cappato R. Worldwide survey on the methods, efficacy, and safety of catheter ablation for human atrial fibrillation. Circulation 2005;111:1100–1105. 13. Hollis S, Campbell F. What is meant by intention to treat analysis? Survey of published randomised controlled trials. BMJ 1999;319:670–674. 14. Weerasooriya R, Khairy P, Litalien J, et al. Catheter ablation for atrial fibrillation: are results maintained at 5 years of follow-up? J Am Coll Cardiol 2011;57: 160–166. 15. Arbelo E, Brugada J, Hindricks G, et al. The atrial fibrillation ablation pilot study: an European Survey on Methodology and results of catheter ablation for atrial fibrillation conducted by the European Heart Rhythm Association. Eur Heart J 2014;35:1466–1478. 16. Oral H, Chugh A, Yoshida K, et al. A randomized assessment of the incremental role of ablation of complex fractionated atrial electrograms after antral pulmonary vein isolation for long-lasting persistent atrial fibrillation. J Am Coll Cardiol 2009;53:782–789. 17. Rostock T, Salukhe TV, Steven D, et al. Long-term single- and multipleprocedure outcome and predictors of success after catheter ablation for persistent atrial fibrillation. Heart Rhythm 2011;8:1391–1397. 18. Ullah W, Hunter RJ, Baker V, Dhinoja M, Sporton S, Earley MJ, Schilling RJ. Target indices for clinical ablation in atrial fibrillation: insights from contact force, electrogram and biophysical parameter analysis. Circ Arrhythm Electrophysiol 2014;7:63–68. 19. Morillo CA, Verma A, Connolly SJ, et al. Radiofrequency ablation vs antiarrhythmic drugs as first-line treatment of paroxysmal atrial fibrillation (RAAFT-2): a randomized trial. JAMA 2014;311:692–700.
1869 20. Winkle RA, Mead RH, Engel G, Patrawala RA. Long-term results of atrial fibrillation ablation: the importance of all initial ablation failures undergoing a repeat ablation. Am Heart J 2011;162:193–200. 21. Hussein AA, Saliba WI, Martin DO, et al. Natural history and long-term outcomes of ablated atrial fibrillation. Circ Arrhythm Electrophysiol 2011;4:271–278. 22. Medi C, Sparks P, Morton J, Kistler P, Halloran K, Rosso R, Vohra J, Kumar S, Kalman J. Pulmonary vein antral isolation for paroxysmal atrial fibrillation: results from long-term follow-up. J Cardiovasc Electrophysiol 2011;22:137–141. 23. Cappato R, Calkins H, Chen S-A, et al. Updated worldwide survey on the methods, efficacy, and safety of catheter ablation for human atrial fibrillation. Circ Arrhythm Electrophysiol 2010;3:32–38. 24. Hunter RJ, Berriman TJ, Diab I, et al. Long-term efficacy of catheter ablation for atrial fibrillation: impact of additional targeting of fractionated electrograms. Heart 2010;96:1372–1378. 25. Tayebjee MH, Creta A, Moder S, Hunter RJ, Earley MJ, Dhinoja MB, Schilling RJ. Impact of angiotensin-converting enzyme-inhibitors and angiotensin receptor blockers on long-term outcome of catheter ablation for atrial fibrillation. Europace 2010;12:1537–1542. 26. Bai R, DI Biase L, Valderrabano M, et al. Worldwide experience with the robotic navigation system in catheter ablation of atrial fibrillation: methodology, efficacy and safety. J Cardiovasc Electrophysiol 2012;23:820–826. 27. Rillig A, Schmidt B, Steven D, et al. Study design of the man and machine trial: a prospective international controlled noninferiority trial comparing manual with robotic catheter ablation for treatment of atrial fibrillation. J Cardiovasc Electrophysiol 2013;24:40–46.
