Acute and mid-term outcome of ethanol infusion of vein of Marshall for the treatment of perimitral flutter

Abstract

Aims 

We hypothesized that an epicardial approach using ethanol infusion in the vein of Marshall (EIVOM) may improve the result of ablation for perimitral flutter (PMF).

Methods and results 

We studied 103 consecutive patients with PMF undergoing high-resolution mapping. The first 71 were treated with radiofrequency (RF) ablation alone (RF-group), and the next 32 underwent EIVOM followed by RF on the endocardial and epicardial mitral isthmus (EIVOM/RF-group). Contact force was not measured during ablation. Acute and 1-year outcomes were compared. Flutter termination rates were similar between the RF-group (63/71, 88.7%) and EIVOM/RF-group (31/32, 96.8%, P = 0.27). Atrial tachycardia (AT) terminated with EIVOM alone in 22/32 (68.6%) in the EIVOM/RF-group. Bidirectional block of mitral isthmus was always achieved in the EIVOM/RF-group, but significantly less frequently achieved in the RF-group (62/71, 87.3%; P = 0.05). Median RF duration for AT termination/conversion was shorter [0 (0–6) s in the EIVOM/RF-group than 312 (55–610) s in the RF-group, P < 0.0001], as well as for mitral isthmus block in the EIVOM/RF-group [246 (0–663) s] than in the RF-group [900 (525–1310) s, P < 0.0001]. Pericardial effusion was observed in 1/32 (3.2%) in EIVOM/RF-group and 5/71 (7.0%) in RF-group (P = 0.66); two in RF-group required drainage and one of them developed subsequent ischaemic stroke. One-year follow-up demonstrated fewer recurrences in the EIVOM/RF-group [6/32 (18.8%)] than in the RF-group [29/71 (40.8%), P = 0.04]. By multivariate analysis, only EIVOM was significantly associated with less AT recurrence (hazard ratio = 0.35, P = 0.018).

Conclusion 

Ethanol infusion in the vein of Marshall may reduce RF duration required for PMF termination as well as for mitral isthmus block without severe complications, and the mid-term outcome may be improved by this approach.

Graphical Abstract

Open in new tabDownload slide

Introduction

With the development of extensive substrate catheter and surgical ablation strategies for persistent atrial fibrillation (AF), there has been a concurrent rise in the incidence of iatrogenic atrial tachycardias (ATs), in particular demonstrating complex mechanisms and anatomical circuits.^1–4 Perimitral flutter (PMF) in particular frequently occurs as a recurrent type of AT. Although high-resolution mapping displays the precise activation sequence of an AT,⁵ allowing the operator to modify their strategy and select the most effective ablation line, PMFs still have a high incidence of recurrence due to epicardial connections, including the coronary sinus (CS) and vein of Marshall (VOM) system.⁶ Our recent paper described a high incidence of residual epicardial connections such as the VOM, which was responsible for refractory AF ablation-related AT.⁷ It may be difficult to create durable lesions with complete block endocardially because these structures are distributed epicardially with some connections to the endocardium.⁸ Báez-Escudero et al.⁹ reported the effectiveness of ethanol injection in the VOM (EIVOM) to create complete block on the mitral isthmus (MI). However, no direct comparison of strategies has been conducted between VOM ethanol injection followed by radiofrequency (RF) application vs. RF application alone as a treatment for PMF. The present study was conducted to elucidate this point.

Methods

Study population

Between January 2015 and August 2018, 103 patients with post-AF AT demonstrated PMF during the procedure, diagnosed and treated using the ‘Rhythmia’ high-resolution mapping system (Boston Scientific, Marlborough, MA, USA). The first 71 patients, with a procedure performed prior to June 2017, were treated by RF alone (RF-group), while the next 32 patients, with a procedure performed after July 2017, were treated by EIVOM followed by RF (EIVOM/RF-group). Acute procedure results and follow-up outcome were compared. Prior written informed consent to participate was obtained from all participants according to institutional guidelines.

Electrophysiological study and mapping with Rhythmia

Electrophysiological study and mapping with the Rhythmia™ system was performed as previously described.^2–4 Antiarrhythmic medications were discontinued >5 half-lives prior to ablation with the exception of amiodarone. A steerable decapolar catheter was introduced within the CS. If patients were in sinus rhythm at the beginning of the procedure, AT was induced with burst pacing regardless of reconduction or block of previous pulmonary vein (PV) or linear lesions. Atrial tachycardias were mapped with the Orion™ multipolar basket catheter with support from a steerable long sheath (Agilis: St Jude Medical or Zurpaz: Boston Scientific). An activation map was created under standard automatic beat acceptance criteria based on (i) cycle length (CL) variation, (ii) activation time difference variations between the CS electrograms, (iii) catheter motion, (iv) electrogram stability, (v) catheter tracking quality, and (vi) respiration gating. Although the dense scar threshold was nominally set at 0.03 mV as a confidence mask parameter, it was manually lowered to be as close as possible to the noise threshold, to allow visualization of the entire circuit.^2–4,6,7,10 After the activation map was completed, entrainment pacing was performed to confirm the diagnosis and circuit unless there were unstable ATs or ATs with short CL (<200 ms), where entrainment pacing might result in the conversion or termination of the index AT. postpacing interval tachycardia cycle length (TCL) ≤20 ms was used to prove that the pacing site was in the circuit.

Catheter ablation and ethanol infusion in the vein of Marshall for perimitral flutter

In the RF-group, the endocardial MI between the lateral mitral annulus and left PVs was targeted as an ablation site. In contrast, in the EIVOM/RF-group, 2–10 mL ethanol infusion inside the VOM was first performed via a femoral approach as previously reported by our institution¹¹ as shown in Figure 1, followed by RF application. Additionally, RF application was performed until complete bidirectional block was achieved. In both groups, when MI block was not achieved by endocardial ablation, the interior CS was also targeted. The endpoint for ablation was bidirectional conduction block by differential pacing with one catheter in the left appendage. Pulmonary veins were re-isolated in cases of reconnection. On completion of the procedure, entrance and exit PV block was confirmed in all four veins and bidirectional conduction block was confirmed by differential pacing. In addition, non-inducibility of the tachycardia after ablation was confirmed. Atrial tachycardia induction was performed by incremental burst pacing up to a pacing CL of 200 ms or the local effective refractory period of the atrial tissue. Radiofrequency energy was delivered using an irrigated 3.5 mm-tip ablation catheter (Thermocool^® SF catheter, Biosense Webster, Diamond Bar, CA, USA), incapable of measuring the contact force, with a power of 30–40 W and a cut-off temperature of 45°C.

In the redo procedure, when the MI was found to be reconnected, ethanol infusion was first attempted, followed by RF since July 2017.

Comparison between radiofrequency alone vs. ethanol infusion in the vein of Marshall/radiofrequency

Patient characteristics, mapping data, acute results, and follow-up were compared between the RF-group EIVOM/RF-group.

Follow-up

All patients were seen at 3, 6, and 12 months’ post-ablation in the outpatient clinic. At each visit, the patients were investigated with a 24 h Holter monitor, exercise stress test, and echocardiography. Follow-up visits at 1 and 9 months were organized by the referring physicians (with 24 h Holter monitoring). In case of symptoms, an additional 12-lead electrocardiogram was acquired. Antiarrhythmic drug treatment, including amiodarone, was discontinued after the procedure in all patients. Patients with more than one recurrence of AT were scheduled for redo procedures and antiarrhythmic drug treatment was reinitiated (unless contraindicated), as a bridge to the redo ablation procedure.

Statistical analysis

Data are expressed as mean ± standard deviation or/and median (25th–75th percentile) for continuous variables, and percentages for frequencies. To compare groups, χ² analysis or Fisher’s exact test was used for categorical variables, and an Student's t tests or Wilcoxon analysis was used for continuous variables. Variables with P-value <0.25 by univariate analysis was included in the multivariate analysis. P-values <0.05 were considered statistically significant.

Results

Patient characteristics

Patient characteristics of 103 consecutive patients [63 (58–69) years, 25 females] with PMF undergoing ablation procedure are shown in Table 1. The initial 71 patients treated by RF only and the latter 32 patients treated by EIVOM/RF were compared. Median number of prior AF/post-AF ablation AT was three, which was similar between the two groups. Although most parameters did not differ between the two groups, the EIVOM/RF-group included a higher incidence of diabetes and hypertension (HT).

Mapping data

Mapping data using the high-resolution/high-density Rhythmia™ system were compared in Table 2. Although the total number of mappable ATs per case and their ATCLs were similar between the two groups, left atrium (LA) volume and the mapping points were significantly larger in the EIVOM/RF-group.

Acute results

Acute periprocedural results are compared in Table 3. Although AT termination or conversion was achieved to a similar extent in both groups (P = 0.27), time for termination or conversion by RF was significantly shorter in the EIVOM/RF-group [0 (0–6) s in EIVOM/RF-group vs. 312 (55–610) s in RF-group, P < 0.0001], as ATs were terminated or converted by ethanol infusion alone in 22/32 (68.8%) patients in the EIVOM/RF-group (Figure 2). Bidirectional block was always achieved in the EIVOM/RF-group, but significantly less frequently achieved in the RF-group (62/71, 87.3%: P = 0.05). Time for complete block was significantly shorter in the EIVOM/RF-group [246 (0–663) s in EIVOM/RF vs. 900 (525–1310) s in RF-group, P < 0.0001], and complete MI block was achieved with ethanol infusion alone in 9/32 (28.1%) in the EIVOM/RF-group. Pericardial effusion was more frequently observed in RF-group [1/32 (3.1%) in EIVOM/RF vs. 5/71 (7.0%) in RF-group], but the difference did not reach significance (P = 0.66). Additionally, two patients in RF-group required drainage and one of them subsequently developed ischaemic stroke. Radiofrequency time for the achievement of complete MI block in RF-group was significantly longer in patients with pericardial effusion than without it [2040 (870–3640) s vs. 895 (523–1285) s, P = 0.05]. EIVOM/RF-group did not significantly reduce, but was weakly associated with a median procedure time (248 min in EIVOM/RF-group vs. 300 min in RF-group, P = 0.09).

1-Year follow-up

One-year follow-up is demonstrated in Figures 3 and 4. In the EIVOM/RF-group, AT recurrence was observed in 5/32 (15.6%) at 6 months and 6/32 (18.8%) at 1 year. In contrast, the RF-group showed more frequent AT recurrence: 22/71 (31.0%) at 6 months and 29/71 (40.8%) at 1 year. Kaplan–Meier analysis demonstrated significantly higher AT recurrence-free survival rate in the EIVOM/RF-group compared to the RF-group (P = 0.03) as shown in Figure 4.

Predictor of atrial tachycardia recurrence

As shown in Table 4, multivariate analysis demonstrated that EIVOM + RM were significantly associated with AT recurrence at 1 year in patients with PMF, and LA volume might have had a weak association with the AT recurrence. Ethanol infusion in the VOM reduced the risk of AT recurrence by 65%. Every 10 mL increase in LA volume increased the risk of AT recurrence by 6%.

Mitral isthmus reconduction in the redo procedure

The prevalence of the MI reconduction was described in Figure 3. In the EIVOM/RF-group, six patients had recurrence of AT although the MI had been bidirectionally blocked in the first procedure, and five of them underwent a redo procedure. Perimitral flutter was clinically observed in 2 (40.0%) patients with endocardial MI reconduction, but MI conduction was not recovered in the other 3 patients.

In the RF-group, 29 patients had recurrence of AT; MI block had been achieved during the first procedure in 23 patients, and 14 of them underwent a redo procedure. Although PMF was clinically observed in 8/14 (57.1%) patients, MI conduction recovered in 13/14 (92.9%) patients; 6/13 (46.2%) of them showed only epicardial conduction.

In the six patients without MI block in the first procedure, four of them underwent a redo procedure. Only epicardial MI conduction was observed in 3/4 (75.0%), and PMF was clinically observed in 2 patients.

Discussion

Major findings

This study demonstrates that:

1. Acutely, EIVOM may reduce the RF duration required for termination of PMF as well as for MI block, and EIVOM/RF may more frequently achieve complete MI bock.

2. EIVOM/RF may provide better outcomes at 1 year as a treatment for PMF.

3. Multivariate analysis shows that EIVOM is a significant predictor to avoid the AT recurrence with a reduction of the risk of recurrence at 1 year by 65%.

Acute results

The acute impact of EIVOM has been reported in several studies. One multicentre study⁹ reported the efficacy and safety of EIVOM in 50 patients undergoing redo procedures and 21 patients with de novo AF, showing that mitral isthmus block was achieved in all patients with 2 min additional RF applications after EIVOM, and RF application inside the CS was required in only one patient.

We have achieved a relatively higher rate of complete MI block even in the RF-group (87.3%), but this is probably because patients frequently received MI ablation in multiple prior procedures. However, the achievement of complete MI block has been generally reported to be challenging (32–92% success)¹² and frequently requires RF applications inside the CS (59–91%). Several anatomical limitations may be associated with the difficulty in achieving complete MI block: a thicker and longer MI,¹³ anatomical variation between the CS, circumflex artery, epicardial structures such as the CS/VOM system and the MI.^6,7,14 To overcome these limitations, several techniques have been reported including using a steerable sheath and high-power RF application,¹⁵ an anterior transseptal approach, or different linear lines¹⁶ such as an anterior line or anteroseptal line. Although superiority of contact force sensing catheter has not been reported in creating MI linear lesion,¹⁷ it may theoretically support more effective and safer RF applications. Physiologically, the ideal option is to create the line of block at the MI region as this is usually activated at the end of sinus rhythm (SR), and block here should not disturb normal conduction during SR and should not increase the likelihood of biatrial ATs.⁴ However, increasing the RF power or contact force to create transmural and durable lesion on this site may also increase the risk of complications such as cardiac effusion or tamponade. It is clear from the present study that longer RF duration on the MI may result in the higher incidence of pericardial effusion. Furthermore, EIVOM/RF may lead to the lower risk of pericardial effusion and cardiac tamponade than RF only due to the reduction of extensive RF application on the MI, even though the difference did not statistically reach significant in this study because of the lower number of the complication in total. Ablation of epicardial structures on the mitral isthmus from the epicardial side by EIVOM may be a reasonable choice in terms of the efficacy and safety. Although it may be still controversial to take EIVOM as the first-line therapy in all PMF, this approach should be considered before alternative epicardial approach¹⁸ in refractory PMF, as it is at least less invasive and likely safer.

Outcome of ethanol infusion in the vein of Marshall

There have been several reports of the outcome of EIVOM in persistent AF and AT patients. Although most publications detail the methodology of EIVOM,¹⁹ or represent single-arm data²⁰ without a comparator, one study has reported the long-term efficacy of EIVOM as an adjunctive treatment to conventional ablation strategy for non-paroxysmal AF²¹ and concluded that EIVOM reduced the risk of AF recurrence by 80%. No randomized controlled study has been published comparing outcomes between EIVOM and other strategies, but there are now two prospective, multicentre, randomized, controlled trials in progress: VENUS-AF and MARS-AF.²² VENUS-AF will enrol patients undergoing their first catheter ablation of persistent AF. MARS-AF will enrol patients undergoing ablation after previous ablation failure(s). The present study is the first prospective study to examine the superiority of EIVOM/RF to RF alone in treating PMF. Although the population was not randomized, patient backgrounds were similar or even with greater comorbidity in the EIVOM/RF-group including patients with diabetes mellitus, HT, and large LA. In addition, we switched our ablation procedure for PMF in July 2017. Even considering the learning curve of the new strategy with EIVOM, the present study demonstrated that the outcome at 1 year was still better in the EIVOM/RF-group, and EIVOM/RF remained as a sole significant predictor of the absence of AT recurrence. Additionally, MI block recovered more frequently in the RF alone group in patients undergoing a redo procedure. These findings may indirectly show that durable lesions on the mitral isthmus region are more frequently achieved with EIVOM/RF.

Clinical implications

In the present study, we compare the efficacy and safety of EIVOM/RF vs. RF alone to treat PMF. We demonstrate that EIVOM may reduce RF duration required for PMF termination as well as for mitral isthmus block without severe complications, and the acute and mid-term outcome may be improved by this approach.

Although EIVOM has potentials in decreasing complications and the procedure time in treating PMF, further prospective randomized study with a larger population may be required to elucidate this point.

Limitations

A major limitation of our study is the absence of formal randomization. Although we acknowledge this, this comparison of two strategies has been done in consecutive patients and using a prospective design. The ablation strategy for PMF was changed from RF alone to EIVOM/RF from July 2017. Patients treated with EIVOM/RF showed a higher prevalence of HT and diabetes with larger LA, which might be associated with worse outcomes. The initial cases treated with EIVOM/RF might have relatively worse outcomes because physicians were not familiar with the new strategy. Even with these negative factors, the EIVOM/RF-group still showed improved outcomes compared to the RF-group.

Second, the prevalence of PMF associated with epicardial conduction during the first procedure was not clear because an epicardial catheter through the VOM was not always placed. However, we may surmise that the prevalence is similar between the two groups because the study population was consecutively collected and the median prior procedure number and rate of ablation on the mitral isthmus was similar between the two groups.

Third, contact force sensing catheter is widely used nowadays. The benefit of this technology may be proved in PV isolation in paroxysmal AF²³ and may theoretically improve the safety and efficacy of RF in general.¹⁷ However, the superiority of contact force measurement has not been elucidated so far in mitral isthmus ablation. In addition, the Rhythmia™ system, unfortunately, does not provide catheters capable of measuring a contact force.

Finally, the difference in the number of ATs in the two groups may have influenced the outcome. However, both the number of ATs and study population may be too small to show the significance difference, influencing the outcome.

Conclusion

Ethanol infusion in the vein of Marshall may reduce RF duration required for PMF termination as well as for MI block. Additionally, EIVOM may improve the 1-year outcome.

Funding

This research was partially funded by a grant from ‘Investissement d’avenir’: IHU LIRYC ANR-10-IAHU-04.

Conflict of interest: M.Haïssaguerre, M.Hocini, P.J., and F.S. have received lecture fees from Biosense Webster and Abbott. A.D., N.D., P.J., and F.S. have received speaking honoraria and consulting fees from Boston Scientific. All other authors declared no conflict of interest.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

References

Author notes