Randomized comparison of atrioventricular node re-entry tachycardia and atrial flutter catheter ablation with and without fluoroscopic guidance: ZeroFluoro study 

Abstract

Introduction

According to current guidelines, catheter ablation is recommended as the first-line therapy for symptomatic atrioventricular nodal re-entrant tachycardia (AVNRT) and typical atrial flutter (AFL).¹ Traditionally, electrophysiological procedures are performed under fluoroscopic guidance, which results in radiation exposure for both the patient and medical staff. The use of fluoroscopy differs widely among electrophysiological laboratories.^2,3 Owing to the stochastic effect of ionizing radiation, no safe dose exists. Thus, protective devices are used to minimize the fluoroscopy exposure of the operating team, but these devices are uncomfortable and may lead to musculoskeletal disorders.^4,5,6,7

The use of a three-dimensional (3D) electroanatomical mapping system and intracardiac echocardiography allows for non-fluoroscopic catheter visualization. Thus, catheter ablation of arrhythmias can be performed with a lower level of fluoroscopy exposure. The Near Zero Fluoroscopic Exposure during Catheter Ablation of Supraventricular Arrhythmias multicentre randomized (Abbott, formerly St. Jude Medical, St. Paul, Minnesota, USA) trial showed that the near-zero fluoroscopic approach with the EnSite Precision system during supraventricular tachycardia (SVT) ablation leads to a dramatic reduction in radiation exposure. This reduces the risk of cancer incidence and mortality as well as the number of years of life affected whilst maintaining procedural safety and efficacy.⁸ Whilst several observational studies have evaluated the use of the minimal fluoroscopic approach, no prospective randomized trial has been conducted to evaluate the zero-fluoroscopic approach for AVNRT and AFL ablation.^9,10,11,12

Therefore, the aim of this randomized study was to demonstrate a clinically significant reduction in ionizing radiation exposure during catheter ablation of SVTs using the EnSite Precision mapping system without compromising procedural efficacy and safety.

Methods

Trial design

The ZeroFluoro SVT ablation study is a prospective, randomized, open-label, multicentre study in which the zero-fluoroscopic approach was compared with conventional fluoroscopy-based ablation for AVNRT and AFL. The trial was investigator-initiated. The local ethics review committee of each centre approved the study.

Study participants

In this prospective randomized study, 123 participants were examined in centres in two European countries between May 2019 and August 2020. Consecutive patients with symptomatic AVNRT (94 patients) or AFL (29 patients) were enrolled in the study. Patients (male or female patients ≥ 18 years of age) receiving ablation for AVNRT or AFL were eligible for participation in the study. There is generally low incidence of SVTs other than AVNRT and AFL in an adult population. Thus, patients with arrhythmias other than the two were not considered eligible for participation in order to avoid possible outliers in the data and a decrease in statistical power. There was strict adherence to the analysis of ECG documentation during patient screening which led to preliminary exclusion of 10 patients due to AVRT and AT. Other exclusion criteria included pregnancy, contraindications to ionizing radiation exposure, life expectancy of less than 1 year, atrial fibrillation, ventricular tachycardia, presence of a congenital heart disease, history of open-heart surgery, cardiac implantable electric devices and known pathological venous access to the heart. All participants provided written informed consent. After enrollment, patients were randomly assigned in a 1:1 ratio to undergo catheter ablation either with conventional fluoroscopic guidance and no 3D mapping system (CF group) or without fluoroscopy using the EnSite system [zerofluoro (ZF) group]. Patients with different final diagnosis revealed by electrophysiology examination received ablation according to the protocol, but data were not included in the statistics (four patients with AVRT in total, 1:1 in both groups).

Catheter ablation procedure

Vascular access was chosen according to the operator’s preference. Catheter manipulation was guided by fluoroscopy in the CF group and by an electroanatomical mapping system in the ZF group. The electroanatomical map was created using the EnSite Precision Cardiac Mapping System, according to the study protocol (see Supplementary material). Briefly, after creating an anatomical map of the right atrium with a multipolar catheter focusing on the area of interest (e.g. slow pathway region in the case of AVNRT), other diagnostic catheters were placed in prespecified positions (e.g. right atrium, bundle of His, right ventricle). Minimum possible fluoroscopic use was allowed if the operator considered it necessary for the effective or safe continuation of the procedure. In patients with AVNRT, ablation was performed using a non-irrigated 4-mm tip catheter. Radiofrequency (RF) ablations in the slow pathway region were performed in the temperature control mode (temperature limit, 40–65°C; power, 20–40 W). In patients with AFL, an irrigated 4 mm tip ablation catheter was used for the ablation of the cavotricuspid isthmus (CTI). RF energy was applied in the temperature control mode (temperature limit, 43°C; power, 30–40 W). Acute procedural success was evaluated 20 min after the last RF application.

Study follow-up

Follow-ups were scheduled at 1 and 6 months after the index ablation procedure. At 1 month, follow-up through a phone call was allowed for reviewing the arrhythmia symptoms and adverse events. At 6 months, an outpatient visit was obligatory; arrhythmia symptoms and adverse events were reviewed through a physical examination and 12-lead electrocardiogram (ECG). Before the visit, either a 72 h Holter monitor recording or 1-week transtelephonic ECG recording was performed. All follow-up assessments were performed by the study personnel.

Endpoints

The primary hypothesis was that a clinically significant reduction or even total abolishment of ionizing radiation exposure would be achieved with RF catheter ablation of AVNRT and AFL using the EnSite Precision system without compromising efficacy and safety. The primary endpoints were acute procedural success and long-term success. Acute procedural success was defined as non-inducibility of arrhythmia in patients with AVNRT and bidirectional CTI block in patients with AFL. Long-term success was defined as no recurrence of clinical arrhythmia during the follow-up period. The safety endpoint was the occurrence of periprocedural and long-term adverse events.

The prespecified secondary endpoints reported in this article include skin-to-skin procedural time, fluoroscopy time, fluoroscopy dose, number of RF applications and total ablation time.

Professional level of involved physicians

Eight physicians were evenly represented between the centres participating in the study. Of them, seven had more than 5 years of experience in interventional electrophysiology, performing both fluoroscopic and electroanatomical mapping system-guided procedures. One of the physicians had less than 2 years of experience in interventional electrophysiology and is referred to as the young investigator in the following sections. Each participating centre belongs among specialized tertiary centre with a big referral territory performing between 600 and 1200 catheter ablation procedures per year, of which 400 to 600 are AF procedures.

Radiation dose risk models

Statistical analysis

Most of the variables showed skewed distributions. Continuous variables were expressed as mean ± standard deviation only if they followed a normal distribution after performing the Shapiro–Wilk test and as median and interquartile range otherwise. The significance of between-group differences was assessed using the two-tailed Student’s t-test or Mann–Whitney U test, when appropriate. As the default, a P-value <0.05 was considered statistically significant with an individual Bonferroni correction¹⁷ applied when required. Categorical variables were expressed as frequency and percentage. The significance of the difference in between-group distributions for both binary and ordinal variables was estimated by the chi square test. The Cochran–Mantel–Haenszel method was used to estimate the risk ratio of binary outcomes. The size of between-group differences between probability distributions in continuous variables were expressed using Cliff’s delta. The limits of delta indicating small and large effect sizes were 0.14 and 0.48, respectively.¹⁸

Statistical analyses were performed using IBM SPSS version 25.0 (Apache Software Foundation, USA) and GraphPad Prism 8.01 (GraphPad Software, Inc., USA).

Results

Patient characteristics

A total of 123 patients (58 in the Hungarian centre and 68 in the Czech centre) with symptomatic AVNRT (76.4%) or AFL (23.6%) were enrolled in the analysis. The distribution of baseline characteristics was comparable between the CF and ZF groups, with no significant differences (Table 1). This was a middle-aged cohort of patients with a balanced sex distribution (52.0% women) and predominantly normal left ventricular function.

Acute success and fluoroscopy parameters (primary endpoint)

Procedural parameters are listed in Table 2. All the patients received the first intervention. A bidirectional CTI conduction block was achieved in 100% of patients with AFL in both the CF and ZF groups. Non-inducibility of arrhythmia was achieved in 97.7% of patients with AVNRT, among whom five (three in the CF and two in the ZF group) had atypical or coexistent typical and atypical AVNRT. The unsuccessful procedures were performed by different physicians, and both were performed in patients with typical AVNRT in the CF group.

In the ZF group, we achieved total elimination of the need for fluoroscopy in each patient. In comparison, the median fluoroscopy time in the CF group was 240 s (interquartile range, 321 s) with an estimated median ED of 0.38 mSv. The minimal and maximal EDs were 0.013 and 3.412 mSv, respectively.

Procedural and cumulative ablation times (secondary endpoints)

The prespecified secondary endpoints reported in this study include skin-to-skin procedural time, number of RF applications and cumulative ablation time. The results corresponding to all these secondary endpoints are listed in Table 2. The comparison of procedural and RF ablation times between the ZF and CF groups is depicted in Figure 1. Procedural time (P = 0.794), number of RF applications (P = 0.220) and cumulative ablation time (P = 0.723) were not statistically different between the study groups. The median procedural time was 60.0 min (interquartile range, 35.0 min) in the CF group compared to 58.0 min (interquartile range, 39.0 min) in the ZF group. The median cumulative RF ablation time was 326 s in the CF group compared to 338.5 s in the ZF group. Cliff’s deltas showed almost negligible differences between probability distributions in both parameters in the CF and ZF groups (procedural time, 0.014; RF time, 0.025). There was a significant increase (P < 0.0001) in the procedural times of our young investigator compared to those of the experienced operators (one vs. all strategy). He also had slightly higher procedural times (Figure 1) in the ZF group than in the CF group (P = 0.136). The difference in the young investigator’s median procedural time between the groups was 22.5 min.

Figure 1

Comparison of (left) procedural and (right) cumulative radiofrequency ablation times between the ZF and CF groups for (A) all procedures and physicians, (B) AVNRT only, (C) atrial flutter only, (D) and the young investigator/electrophysiologist who had less than 2 years of experience. AVNRT, atrioventricular nodal re-entrant tachycardia; AFL, typical atrial flutter; CF, conventional fluoroscopy; ZF = zero fluoroscopy.

Open in new tabDownload slide

Procedural complications and safety endpoint

There were no cases of major complications or mortality during or after the procedure in either group. Intermittent asymptomatic atrioventricular block type I–II occurred during RF ablation of AVNRT in one patient in the CF group and was present at night during the follow-up period; however, there were no indications for pacemaker implantation. Further, one patient in the CF group with acute heart failure and 15% ejection fraction at admission experienced pulmonary oedema after RF ablation for AFL. A small groin haematoma was observed in one patient in the ZF group who was being treated with triple antithrombotic therapy and underwent ablation for AFL after a coronary intervention for severe stenosis of the left descending artery.

Follow-up data

After a median follow-up period of 182 days, 94.3% of patients (n = 123) were free from AVNRT or AFL. No procedure-related complications were reported during follow-up. Six patients with AVNRT had a documented recurrence of arrhythmia and underwent repeat ablation before the end of the 6-month follow-up period. Of these, one patient with AVNRT recurrence was newly diagnosed with AFL; both substrates were successfully treated with a single procedure. One patient who received AFL ablation experienced documented episodes of the same arrhythmia and underwent two additional ablation procedures to achieve a bidirectional CTI block. These were considered independent events in the descriptive statistics. One patient who underwent AFL ablation developed newly diagnosed left atrial flutter; nevertheless, a bidirectional isthmus block was confirmed by a repeated electrophysiological study. Five patients had episodes of paroxysmal atrial fibrillation. Six patients experienced two or fewer documented events of non-sustained irregular SVT. The overall results observed through the 6-month follow-up period are presented in Table 3.

The risk of arrhythmia recurrence was comparable between the groups; any differences in arrhythmia recurrence were unlikely to be due to the type of procedure. The number of patients using antiarrhythmic medication apart from beta-blockers was reduced by 61.5 and 68.7% in the CF and ZF groups, respectively (P = 0.412). Overall, 10.5% of patients observed palpitations during the 6-month follow-up period. Of these, eight were in the CF group and five in the ZF group.

LAR

At an average age of 55.5 years and median radiation exposure of 0.38 mSv, the estimate of LAR incidence in both sexes was approximately 1 in 14 084 (1 in 26 315 men and 1 in 30 303 women), see Table 4. These estimates are similar to the general risk estimates associated with stochastic effects at low doses, reported as an increase of approximately 5.5% in cancer cases for a dose of 1 Sv (13). The estimated mortality rate was 1 per 17 857 persons who were exposed. Mortality rates were not significantly different between the sexes (1 in 34 482 men and 1 in 37 037 women). Both the incidence and mortality rates were the highest in case of an exposure at ages below 40 years; however, the number of patients in this age group in our cohort was relatively small to make any general conclusion.

Discussion

The main finding in our prospective, randomized, open-label, multicentre study is that the procedural safety and efficacy of the zero-fluoroscopic approach are similar to those of conventional fluoroscopy-based ablation for AVNRT and AFL, while the lifetime risk of malignancy is markedly reduced.

Population exposure studies conducted over the last decades in the United States¹⁹ and European Union countries²⁰ reported an increase in the per capita ED in all medical imaging examinations by 33 and 27%, respectively. In this context, patients and staff can still be exposed to considerable radiation doses during interventional cardiology procedures due to prolonged fluoroscopy time. The consequence is an increase in health-related risk attributed to the stochastic effects of radiation based on a linear no-threshold cancer risk model.²¹

Efficacy and safety

In this prospective randomized study, which was carried out across two centres and involved several operators, we compared a completely fluoroscopy-free approach with conventional procedures involving the use of fluoroscopy. Within the limitations of this study, the use of a 3D navigation system (EnSite Precision) resulted in the same success rate and procedural and RF application times as those of conventional fluoroscopic guidance without compromising patient safety or long-term effects related to arrhythmia recurrence. This is in accordance with other studies^{8,9,11,12,22,23} confirming the safety, efficacy and feasibility of procedures exclusively guided by 3D navigation systems. Even though more complex SVTs were not included in this study, we were able to reach both efficacy and safety in ZF procedures focused towards left-sided substrates associated with AVRT and AT. In our centres, this has been limited almost exclusively to ICE-guided trans-septal approach. Together with time consuming mapping of more challenging substrates, such procedures might take substantially longer and should be handled by experienced operators. The risk of periprocedural complications was numerically larger in the CF group than in the ZF group although this difference did not reach statistical significance. This effect was not related to any particular centre or physician. Currently, there are conflicting results corresponding to the aforementioned phenomena in the literature. Similar recurrence rates were observed by Casella et al.⁸ and Earley et al.²² By contrast, a large prospective study by Chen et al.⁹ did not report any difference in long-term efficacy. In addition, the relative risk of periprocedural complications was reduced in both of the largest prospective studies^8,9 to a similar extent as that in our cohort (0.7 and 0.48, respectively) after the zero-fluoroscopic approach. These results do not indicate whether EnSite Precision-guided procedures are safer or more effective than fluoroscopic guided procedures. This can be explained by several factors, among which the study population size, probability distribution, and diversity of the arbitrarily rare periprocedural complications probably play the most important role. We can also hypothesize that a single-blinded prospective study design combined with reduced visual support may have led to the physicians taking greater precautions during the zero-fluoroscopic procedures. The extent to which such a trend would remain the same after the operators are adapted to the new (non-fluoroscopy) condition is still open to debate.

Learning curve

Among the main limitations of the zero-fluoroscopic approach, higher equipment prices and the time needed to gain adequate proficiency could be mentioned. The latter was previously demonstrated by Gist et al.²³ and Chen et al.⁹ during SVT ablation. By contrast, Wang et al.²⁴ documented a significant shortening of the procedural time only during the placement of the coronary sinus catheter via the subclavian vein. Chen et al.⁹ demonstrated that the procedural time in the zero-fluoroscopic approach substantially reduced after 40 cases and that the initial average procedural time was approximately 15 min. longer than that in the conventional approach. Despite the fact that our study design was not focused on evaluating procedure-specific learning curves, a similar prolongation (22.5 min) was observed in the median procedural time in the zero-fluoroscopic procedures performed by our young investigator. However, after switching to the zero-fluoroscopic approach, he was able to adapt very quickly, and the overall difference between the procedural times in the ZF (n = 13) and CF (n = 16) groups did not reach statistical significance. Regarding to safety outcomes, young investigator had none periprocedural complication in both groups and only a single unsuccessful procedure in the CF group. This suggests that it would be beneficial for younger operators to simultaneously begin with both the zero or near-zero fluoroscopic approach and conventional approach. However, more extensive study involving broader representation of less experienced operators would have had to be performed to fully support this statement.

Radiation exposure risk

Modern radiography systems allow for the use of low radiation doses while preserving adequate imaging quality. As shown by the results, a decrease in the LAR of cancer incidence and mortality compared with that in the study by Casella et al.⁸ can be achieved even by using conventional fluoroscopy. This effect is most likely due to a substantial reduction in fluoroscopic time and exposure. However, it should be noted that only patients with AVNRT or AFL were included in our cohort, and both fluoroscopic parameters may be significantly higher when dealing with demanding substrate-mapping procedures in arrhythmias, such as AVRT or AT. Relatively low LAR estimates should not hinder the efforts to further reduce radiation exposure. Limiting the operator’s reliance on fluoroscopic imaging with the use of 3D electroanatomical guidance systems helps in reducing radiation exposure for the medical professionals involved in interventional procedures. Over the last decade, several occupational hazards affecting interventional laboratory staff have been reported in the literature; among the most concerning hazards are the left-sided (85% of reported cases) brain and neck tumours developing in physicians performing interventional procedures.²⁵ Since the left side is usually more exposed to radiation, the disproportionate development of tumours on the left side suggests a relation to occupational radiation exposure. Another frequently discussed topic is the high prevalence of musculoskeletal pain and orthopaedic problems, which are assumed to be related to and worsened by wearing radiation protective aprons. In a multisite case–control study, Orme et al.⁶ reported that the probability of work-related pain in professionals involved in procedures with radiation exposure significantly increased with the duration of radiation exposure and time spent in the protective apron.

Procedure cost-effectiveness

Considering prices of the electrophysiology equipment common in regions of Central Europe, it is important to mention that the ZF procedure introduces necessary extra cost deriving from the 3D mapping system of approximately €840 per procedure using the Ensite Precision system and €670 vs. €1120 (AFL vs. AVNRT) using the Carto 3 system. Compared to average cost of the CF procedure of €3280, the ZF procedure is 20–35% more expensive. Affordability in different regions thus depends on healthcare costs paid by health insurance company for single conventional non-complex ablation procedure. An additional cost might arise for a particular centre when an acquisition of 3D mapping system is necessary. Even though the radiation exposure risk in our study is substantially low for patients in the CF group, we could assume there would be additional benefit in reducing occupational hazards discussed above. However, our study does not provide enough data to carry out such cost-effectiveness analysis.

Methodological issues

The computation of an ED depends on several assumptions and parameters, such as the configuration of the radiography system and patient characteristics. Thus, it is important to note that all reported EDs are statistical estimates based on generally used correction factors, which may result in biased cancer risk values. The estimation of LAR is based on empirical risk models from the life span study (LSS) of a survivor cohort of the atomic bombing in Japan. Despite the use of population-specific cancer incidence and mortality rates, the ability of the models to estimate potentially small risks of an individual patient exposed to low doses of radiation is significantly limited by a high degree of uncertainty. However, we must emphasize that, to our knowledge, so far, the LSS cohort study still includes the largest relevant dataset available on subjects exposed to low doses of radiation. All the zero-fluoroscopic procedures were performed using the EnSite Precision system in patients with non-complex arrhythmias (AVNRT and AFL). The applicability of the present findings should not be generalized to procedures targeting SVTs such as AVRT, AT and other complex arrhythmias or involving other mapping systems. Exclusion of other arrhythmias than AVNRT and AFL could possibly introduce selection bias into our cohort. By design itself, a prospective study does not allow for the elimination of the Hawthorne effect. The physician might have subconsciously exercised a greater degree of prudence or increased the demand for insurance based on the type of procedure, which might have led to bias in the results on safety endpoints.

Conclusion

In conclusion, in comparison with using the fluoroscopic approach, an absolute elimination of ionizing radiation exposure for both the patient and operator can be achieved through RF catheter ablation of AVNRT and AFL using the zero-fluoroscopic approach without compromising efficacy and safety. The LAR of cancer incidence and mortality caused by low-dose radiation exposure from fluoroscopy-guided interventional cardiac procedures should not be generally considered trivial. However, it can be reasonably reduced by proper radiation management and appropriate effort to decrease unnecessary radiation exposure. The use of a 3D navigation system might be considered as a valuable and safe alternative with long-term benefits to patients and mainly to electrophysiological laboratory staff. Our results indicate that there is no justified benefit of patients undergoing RF ablation for AVNRT or AFL under fluoroscopic guidance.

Supplementary material

Supplementary material is available at Europace online.

Funding

This work was supported by the European Regional Development Fund—Project ENOCH (No. CZ.02.1.01/0.0/0.0/16_019/0000868).

Data Availability

Data available on request due to privacy/ethical restrictions. The data are not publicly available due to General Data Protection Regulation.

References

Author notes