Potential complications in patients undergoing an ethanol injection into the vein of Marshall

Abstract Background Ethanol injections into the vein of Marshall (VOM) (EIM) are considered to be a good therapeutic option for atrial tachyarrhythmias, however, the safety remains to be determined. To elucidate what would affect the safety and potential complications of an EIM, we investigated the anatomical features of the VOM and patient background. Methods We performed the EIM before the conventional pulmonary vein isolation for drug‐resistant atrial fibrillation in 88 patients and evaluated the anatomical features of the VOM and their background. Results All procedures were completed, however, other than myocardial staining, trivial contrast medium leaked out of the VOM into the pericardial space, that is, extravasation of contrast medium with capillary rupture, during the EIM in 20 patients (22.7%) regardless of the features of the VOM. No pericardial effusions requiring further intervention developed after the extravasation, which resolved by the next day on echocardiography in 18 of those patients. However, two patients who had extravasation other than during the initial contrast injection required additional therapeutic intervention for nonnegligible pericardial effusions. Their body weights were significantly lower and the latter two patients were also small lean women with heart failure and a preserved ejection fraction. Conclusions The physical constitution, regardless of the characteristics of the VOM, could be strongly associated with adverse events during the EIM. We must take extreme care in smaller patients with poor compliant hearts during the EIM.

create lesions by injecting ethanol directly into the vein. [1][2][3] However, what affects the safety and potential complications of the EIM remains to be determined. In this study, we investigated the anatomical features of the VOM and patient background to elucidate the safety of the EIM in relation to those parameters.

| Study population
This study included 88 consecutive patients who underwent an EIM and following RF ablation for drug-resistant AF (paroxysmal: 36, persistent: 23, and longstanding persistent: 29 patients) and were followed for at least 6 months. We defined paroxysmal AF as that terminating spontaneously and lasting for less than 1 week, persistent AF as that lasting for more than 1 week but less than 1 year, and long-standing persistent AF as that lasting for longer than 1 year. The subjects included 62 men and 26 women with a mean age of 65.4 ± 9.7 years. Written informed consent, including for the EIM, was obtained from each patient before the procedure according to the protocol approved by the Institutional Research Committee.

| Ablation procedure
All the procedures were performed under deep sedation with dexmedetomidine, propofol, and buprenorphine, and with esophageal temperature and direct blood pressure monitoring. The activated clotting time was kept at approximately 300 seconds throughout the procedure. All the patients underwent an EIM before the conventional pulmonary vein (PV) isolation (PVI).
The EIM procedures were performed similar to that previously described. [1][2][3][4] At first, we obtained a coronary sinus (CS) venography by injecting contrast medium via the infusion port of a duodecapolar CS electrode catheter (Abbott, St Paul, MN) without an occlusion balloon to confirm the VOM. Then, we performed a standard transseptal puncture, inserted three sheaths into the left atrium and drew the geometry of the left atrium with a voltage map using a circular catheter and ablation catheter with a 3D mapping system (CARTO; Biosense-Webster, Diamond Bar, CA, or Ensite NavX; Abbott, St Paul, MN) and placed the circular mapping catheter into the left superior and inferior PVs to record their potentials. After that, we inserted an angiographic catheter for the left internal mammary artery (6 Fr IM; Asahi Intec, Tokyo, Japan) into the CS close to the ostium of the VOM. Through that catheter, a peripheral angioplasty guidewire (0.014″ Cruise; Asahi Intec, Tokyo, Japan) was advanced into the VOM, and an angioplasty balloon catheter (8-mm length Apex OTW with a 2-mm nominal diameter and two radiopaque markers on both the distal and proximal ends of the balloon; Boston Scientific, Boston, MA) was then advanced gently over the wire as far as possible into the VOM. Then, following an injection of 0.2-0.3 mL of radiographic contrast medium for selective VOM venography through the wire lumen of the angioplasty balloon, 2.0 mL of dehydrated ethanol was injected after inflating the balloon with the minimal atmosphere pressure to occlude the VOM, which was commonly less than 4 atms.
Then, we pulled the balloon back 8 to 10 mm and repeated the ethanol injection while the distal tip remained inside the VOM. We used a 2-mL syringe to inject the contrast medium and 1 mL of the ethanol very slowly and gently over more than 1 minute. Throughout the EIM procedure, we carefully checked to confirm whether any contrast medium had leaked out of the VOM or its capillaries into the pericardial space. After the EIM procedure described above, we proceeded to the conventional RF ablation of both ipsilateral PV antra to complete bidirectional block of the PV electrograms.

| VOM evaluation
To evaluate the anatomical features of the VOM, we measured the distance from the CS ostium to the VOM ostium (a), length of the VOM measured on the CS imaging (b) and that of the direct image taken by injecting contrast medium from the balloon lumen (c), and the depth of the distal tip of the balloon penetrating into the VOM (d) using G-NAVI version 6.1.0 software (Goodman, Nagoya, Japan). We also checked how many injections of ethanol were administered, and whether or not the contrast medium had leaked out into the pericardial space other than that from myocardial staining, that is, extravasation of the contrast medium with capillary rupture (e) (Figure 1).

| Patient follow-up
We checked if any complications including a trivial pericardial effusion might have occurred using echocardiography soon after all the procedures including after the EIM and PVI had been completed. All patients were hospitalized for at least 3 days after the procedure with continuous monitoring of the ECG and underwent echocardiography at least once on the day after the procedure. We also performed a clinical follow-up at the cardiology clinic including echocardiography in all patients within 1 month after discharge.  (Table 1).

| DISCUSSION
Chemical ablation using an ethanol injection through a vessel has been widely accepted clinically and also applied to cardiac diseases. 5

| STUDY LIMITATIONS
First, this was a retrospective cohort study that included very few complications. Therefore, we could not perform multivariate analyses to determine the predictors of the Intervention(+) group, which required Second, we experienced two cases with complications during the early phase even when injecting the ethanol very slowly (1 mL over more than 1 minute). After that, we fortunately did not experience any further serious complications since we began performing the ethanol injection using the same size syringe but by delivering the ethanol more slowly and gently than before.
However, we encountered several patients with trivial extravasation associated with a capillary rupture that resolved without any further intervention. We did not measure the accurate pressure using a manometer during the injections, and therefore, we could not objectively determine the threshold related to the serious complications.