OUP user menu

Percutaneous transhepatic access for catheter ablation of cardiac arrhythmias

Duy Thai Nguyen, Rajan Gupta, Joseph Kay, Thomas Fagan, Christopher Lowery, Kathryn K. Collins, William H. Sauer
DOI: http://dx.doi.org/10.1093/europace/eus315 494-500 First published online: 5 February 2013


Aims Femoral venous access may be limited in certain patients undergoing electrophysiology (EP) study and ablation. The purpose of this study is to review a series of patients undergoing percutaneous transhepatic access to allow for ablation of cardiac arrhythmias.

Methods and results Six patients with a variety of cardiac arrhythmias and venous abnormalities underwent percutaneous transhepatic access. Under fluoroscopic and ultrasound guidance, a percutaneous needle was advanced into a hepatic vein and exchanged for a vascular sheath over a wire. Electrophysiology study and radiofrequency ablation was then performed. All tachycardias, including atrial tachycardia, atrial flutter, atrioventricular nodal tachycardia, and atrial fibrillation, were ablated. Procedural times ranged from 227 to 418 min. Fluoroscopy times ranged from 32 to 95 min. There were no complications. All six patients have been arrhythmia-free in follow-up (5–49 months, mean 23.1 months).

Conclusion Percutaneous transhepatic access is safe and feasible in patients with limited venous access who are undergoing EP study and ablation for a range of cardiac arrhythmias.

  • Transhepatic access
  • Cardiac arrhythmia
  • Ablation


Patients with congenital heart disease are at increased risk of developing a variety of arrhythmias. However, their cardiac anomalies are oftentimes associated with other structural anomalies that limit their venous access. Furthermore, they frequently undergo multiple venous accesses throughout their lifetime, leading to occlusions of their venous system.

Catheter ablation of cardiac arrhythmias frequently depends on access from the femoral venous system. Although access and ablation from a superior approach via the superior vena cava is possible, catheter manipulation and control is better accomplished from the inferior approach, especially in cardiac anatomies with prior surgical repair or structural anomalies. The superior approach, however, has been reported to be successful in patients who have no inferior access.1,2

In patients with no femoral vein access, a percutaneous transhepatic approach is also possible.35 Such an approach has been applied in the paediatric population, initially in interventional cardiac catheterizations69 and then translated to paediatric electrophysiologic (EP) procedures.10,11 Little is known about this approach in the adult population, although its feasibility was recently shown in two reported cases.12 The purpose of this study is to provide further data on this technique and to report the largest series to date of adult patients undergoing transhepatic access for ablation of a variety of cardiac arrhythmias.

What's new?

• Percutaneous transhepatic access offers an approach that confers similar catheter stability and manipulation as routine ablation procedures from the iliofemoral veins.

• This study presents a series of patients who underwent successful ablation of their clinical arrhythmias after transhepatic access, which can be done safely through the use of non-invasive imaging including fluoroscopy and ultrasound guidance.

• Several types of arrhythmias can be adequately targeted and successfully ablated using a transhepatic approach.

Patients and methods

All patients referred for ablation from June 2007 to February 2012 were evaluated. Six patients were identified who had limited venous access and underwent transhepatic access. Electrophysiologic studies with three-dimensional electroanatomic mapping (Biosense-Webster, CARTO) and radiofrequency (RF) ablation procedures were performed in all six patients with a variety of atrial arrhythmias and venous abnormalities. Patient characteristics, details of their procedures, and clinical follow-up were ascertained (Table 1).

View this table:
Table 1

Patient characteristics, procedural details, target arrhythmia, and clinical follow-up

PatientPatient 1Patient 2Patient 3Patient 4Patient 5Patient 6
ArrhythmiaCTI atrial flutterTypical AVNRTCTI atrial flutterAtrial tachycardiaCTI atrial flutterAtrial fibrillation/CTI atrial flutter
Congenital cardiac anomaliesD-TGA with Mustard baffle; situs inversus totalisNoneD-TGA with Mustard baffleSingle atrium with baffle; dextrocardiaD-TGA with Mustard baffleNone
Reason for hepatic accessBilateral iliofemoral occlusionsSurgical ligation of inferior vena cavaBilateral iliofemoral occlusionsOccluded right iliofemoral vein. Left femoral vein drains to azygous veinBilateral iliofemoral occlusionsCongenital absence of inferior vena cava with femoral veins draining to azygous vein
CathetersDecapolar catheter; ablation catheterDecapolar coronary sinus catheter; ablation catheterDecapolar catheter; ablation catheterDecapolar coronary sinus catheter; ablation catheterDecapolar catheter; ablation catheterDecapolar coronary sinus catheter; intracardiac ultrasound; ablation catheter
Procedural anticoagulationHeparinCoumadin INR 1.7HeparinNoneHeparinHeparin
Total procedural time (min)257227420a305345418
EPS time223176260204203358
Fluoroscopy time5632112a365995
Hepatic access closureVascular closure device; gel foamVascular closure device; gel foamVascular closure device; gel foamVascular closure device; gel foamVascular coilVascular closure device; gel foam
Ablation SuccessYesYesYesYesYesYes
Follow-up34 months36 months49 months8 months7 months5 months
  • aCombined ablation procedure with interventional catheterization to assess patient's anatomy and haemodynamics and to repair a baffle leak.

  • CTI, cavotricuspid isthmus; AVNRT, atrioventricular nodal reentrant tachycardia; D-TGA, D-transposition of the great vessels.

Patients were admitted to the cardiac EP laboratory in the fasting state. Informed consent for the procedure, including transhepatic access, was obtained. Advantages and disadvantages of transhepatic access, compared with a superior approach, were discussed with patients. In one patient, a combined superior and transhepatic approach was performed. Our centre has extensive experience in transhepatic access for a variety of indications, including non-cardiac pathology. Antiarrhythmic drugs were discontinued at least five half-lives before the procedure. After induction of general anaesthesia and endotracheal intubation (two patients underwent conscious/deep sedation only), right internal jugular vein access and femoral arterial access were obtained, if needed. Percutaneous transhepatic access was obtained using an 18-gauge or 22-gauge echo-tip Chiba needle (EMcision Ltd, London, UK) and ultrasound guidance preferentially into the right hepatic vein (Figure 1). The right or middle hepatic veins were chosen for percutaneous access because they both are easily visible from an intercostal right-sided approach, and they direct the sheath appropriately to the midline.

Figure 1

Ultrasound guidance of transhepatic access. Transhepatic ultrasound access into the right hepatic vein with an 18 gauge echo-tip Chiba needle. Note that a moderately long tract length was chosen to allow adequate room for tract embolization. Care was taken to avoid other portal venous and biliary structures in the needle path. Hepatic veins can be differentiated from portal venous branches on ultrasound by less echogenic borders and the ability to trace the venous branches centrally to the superior aspect of the liver. Differentiation can also be made with colour Doppler imaging.

Initial entry site was through a low intercostal or subcostal approach to avoid a transpleural puncture. An intrahepatic tract of moderate length (≥5 cm) was chosen to allow for sufficient space for tract embolization at the conclusion of the procedure. Care was taken to choose an appropriate needle path to avoid transgressing portal or biliary structures and limit access to a single capsular puncture. Appropriate access was confirmed with aspiration of venous blood and contrast injection. After confirmation of hepatic venous access, a fluoroscopic spot image was obtained of the needle entry position to provide fluoroscopic landmarks for tract embolization at the conclusion of the procedure (Figure 2). An Amplatz stiff guidewire was advanced into the right atrium under fluoroscopic visualization and the needle was exchanged for a long vascular sheath (Figure 2). Intravenous heparin bolus and drip was then administered if entry into the systemic arterial circulation was planned.

Figure 2

Fluoroscopy of transhepatic access. Fluoroscopic view of transhepatic access with a needle placed in the right hepatic vein (A). This image was saved to provide a reference for the exact point of hepatic venous entry to aid in tract embolization. Fluoroscopic image of advancement of the guidewire and sheath to the right atrium (B). After exchanging a vascular sheath over the wire, an ablation catheter is advanced to the right atrium (C).

Programmed stimulation, arrhythmia induction, and EP studies were performed as indicated by the patient's clinical arrhythmia. Diagnosis of the arrhythmia was performed using the usual EP manoeuvres. Electroanatomical mapping (CARTO; Biosense Webster) and entrainment mapping was performed, where necessary. Radiofrequency ablation was delivered to terminate the clinical arrhythmia, and appropriate EP manoeuvres were performed to confirm non-inducibility of the tachycardia or electrical conduction block across ablation lines.

At the end of the procedure, protamine was administered to reverse heparin, if it was used during the procedure. The hepatic sheath was withdrawn under fluoroscopic guidance while injecting contrast to position the sheath within the hepatic parenchymal tract but outside the hepatic venous entry site. Correlation was made to the fluoroscopic spot image of the needle entry site into the hepatic vein saved at the time of the initial puncture in the same fluoroscopic plane. This was performed as an added confirmation of appropriate location within the parenchymal tract prior to embolization given that contrast injection of the hepatic tract during sheath pullback may appear similar to a hepatic venous branch leading to inadvertent sheath removal from the liver precluding embolization. An Amplatzer Vascular Plug (AGA Medical, Golden Valley, MN, USA) was placed through the sheath into the hepatic tract and deployed in a controlled manner. The vascular plug was chosen for tract embolization due to its low risk of distal embolization into the hepatic vein, ease of deployment, and ability for repositioning if necessary. In most cases, there was persistent but decreased back bleeding from the sheath side port after deployment of the plug. The remainder of the hepatic tract was packed with large gelfoam pledgets (Pfizer, New York, NY, USA). These were delivered through the cut end of the sheath and were pushed in with the sheath dilator with the goal of complete cessation of blood flow through the parenchymal tract prior to sheath removal. In one patient, a vascular coil (Cook Cardiology, Bloomington, IN, USA) was deployed instead, due to physician preference. Haemostasis using coil or plug occlusion of a tract is not immediate. Therefore, use of gelfoam pledgets of reasonable size (2–3 mm) will impart immediate haemostasis, which is important in patients in whom anticoagulation is resumed. The amplatzer plug further serves as a backstop to prevent the pledgets from embolizing centrally into the hepatic veins. Ultrasound of the liver was performed immediately after sheath removal to ensure there was no major bleeding, although this was a not critical diagnostic study and was adjunctive to haemodynamic monitoring. No periprocedural complications occurred. When necessary, full anticoagulation with intravenous heparin and warfarin therapy was resumed 6 h post-sheath removal. Serial haematocrits were obtained every 8 h for 24 h, or more often as clinically indicated, to ensure that there was no active bleeding prior to the administration of anticoagulation. Patients remained overnight in an inpatient setting for a 24 h period. Anticoagulation was closely monitored.


Three patients had bilateral iliofemoral venous occlusions. One patient had congenital absence of the inferior vena cava. Another patient had a right iliofemoral occlusion and anomalous return of the left iliofemoral vein into the azygous vein. The last patient had a prior surgical ligation of his inferior vena cava to prevent recurrent pulmonary embolism. Transhepatic access allowed for six arrhythmias to be ablated in these patients, including typical atrioventricular nodal reentrant tachycardia (AVNRT), a focal atrial tachycardia, three cavotricuspid isthmus macroreentrant flutters, and atrial fibrillation (AF).

Four patients required intravenous heparin after transhepatic access. Three of these patients had D-transposition of the great vessels with surgical baffles; they required systemic arterial circulation access to complete a cavotricuspid isthmus ablation line from the tricuspid valve (which was the systemic atrioventricular valve) to the inferior vena cava (via the baffle). Figure 3 shows an MRI reconstruction of a patient's atrial anatomy and baffle merged with the electroanatomic mapping of counterclockwise macroreentrant atrial flutter around the tricuspid valve (CARTO-MERGE, Biosense Webster). In the patient with drug-refractory paroxysmal AF, a single transseptal puncture was performed using fluoroscopic and intracardiac echocardiogram guidance (placed via the right internal jugular vein). Pulmonary vein isolation was confirmed using a multielectrode Lasso (Biosense Webster) and AF non-inducibility was accomplished (Figure 4).

Figure 3

Closure of transhepatic access. Fluoroscopic image of vascular plug placed into the hepatic parenchymal tract just outside the hepatic vein.

Figure 4

Three-dimensional electroanatomic mapping after transhepatic access. Left anterior oblique view of an magnetic resonance imaging reconstruction of both atria in a D-transposition of the great vessels patient with Mustard baffle, which is merged with an electroanatomic mapping of counterclockwise macroreentrant flutter circulating around the tricuspid valve (CARTO-MERGE, Biosense Webster). The colour scale represents a spectrum of atrial activation, with red as the region of earliest activation and purple as latest activation. The circular dots (mostly in red) represent ablation points along the cavotricuspid isthmus on either of the baffle (posterior isthmus was within the baffle while the anterior isthmus was accessed via a retrograde aortic approach).

Total procedure time varied from 227 to 418 min, depending on the targeted arrhythmia. One patient's total procedure time was 420 min but this included an interventional procedure that assessed the patient's anatomy by angiography, evaluated his haemodynamics with right heart catheterization, and repaired a baffle leak with vascular coiling. Procedural time duration was determined by the complexity of the arrhythmia and its catheter ablation, rather than by the transhepatic access. All arrhythmias were ablated successfully without further inducibility. There were no complications. All patients have remained arrhythmia-free, with one patient (focal atrial tachycardia) requiring an antiarrhythmic drug for symptomatic premature atrial ectopy without sustained arrhythmias. Follow-up period ranged from 5 to 49 months (mean 23.1 months) (Figure 5).

Figure 5

Atrial fibrillation ablation after transhepatic access. Left anterior oblique (A) and right anterior oblique (B) fluoroscopic images of the ablation catheter, after transseptal access, in the right superior pulmonary vein during atrial fibrillation ablation.


Patients with cardiac arrhythmias may present with limited venous access. Many patients with congenital heart disease have concurrent anomalous venous anatomy, including absence of the inferior vena cava or obstruction of their iliofemoral venous system due to multiple catheterization procedures or indwelling catheters during prior hospitalizations. This is becoming a more prevalent issue for adult electrophysiologists, as patients with congenital heart disease live longer and develop arrhythmias, and their venous abnormalities are found to be occluded from childhood procedures. Alternative access strategies have been employed, including utilizing a superior venous approach via the jugular or subclavian veins.1,13 However, disadvantages of this approach include less compressible sites for haemostasis; less familiarity with catheter manipulation and stability in these positions; more difficult transseptal access given the non-traditional approach and lack of sheath support and stability.

Transhepatic access allows for an inferior approach, which is more familiar to electrophysiologists and allows for greater degrees of manoeuvrability and catheter manipulation. This approach, however, is also a less compressible site for haemostasis. This study shows that a transhepatic approach is feasible for catheter ablation of a variety of cardiac arrhythmias, including AVNRT, atrial tachycardia, atrial flutter, and AF. The more posterior approach of a transhepatic access did present a more challenging transseptal puncture in the AF ablation case, since it directed the transseptal needle more anteriorly. The angle of the transhepatic approach is posterior and this therefore points the access sheath anteriorly into the heart. The ideal puncture site for a transseptal access across the fossa ovalis is more posterior along the interatrial septum; when the transseptal needle is directed anteriorly, there is the potential for inadvertent aortic root puncture. In this situation, a bidirectional ablation catheter was utilized to direct the SL1 long sheath more posteriorly. The ablation catheter was then exchanged for the SL1 sheath's dilator and a Brockenborough needle, which was then utilized to access the left atrium.

Percutaneous transhepatic access has been utilized in non-cardiac procedures by interventional radiologists for placement of central access.4,14 It has also been utilized for various paediatric interventional cardiac cases.68,1518 The reported complication rates are relatively low (<5%), and major complications include bleeding, infection, thrombosis, and pneumothorax.3,17,19 Intraperitoneal bleeding is a possible major complication that has been reported; we followed haemodynamics carefully after every case and did not experience this complication. Subcutaneous haematomas have also been reported, especially in patients on warfarin. As with other complex techniques and procedures, low complication rates are related to a centre's experience and high volume, and thus this procedure should be reserved at centres with experienced interventionalists who have performed transhepatic access for a variety of indications, including cardiac procedures.

The use of a transhepatic approach in an adult population for EP study and ablation is rare and has been previously reported in only two cases;12 however, it has been used for a variety of cardiac procedures in the paediatric population, including for diagnostic and interventional cardiac catheterizations, pacemaker placement, and electrophysiological procedures. We report here the largest adult series, to date, of transhepatic access in order to facilitate EP study and ablation, which can be performed safely and with full anticoagulation on heparin. Judicious selection of patients and careful planning are important to minimize risks of complications; transhepatic access should only be performed in a centre familiar with the approach. Use of landmarks, fluoroscopy, and ultrasound are important to guide access to the hepatic veins, which are of sufficient size to accommodate the large sheaths required for certain ablation procedures. While none of the patients in this series have required a second procedure, the presence of more than one hepatic vein and ensuring that the closure device is outside of the hepatic vein will allow for repeat transhepatic access in the future, if necessary. Figure 6 shows a patent hepatic vein in patient #1, 3 years after her transhepatic access and closure. Repetitive transhepatic access has been performed safely, including in one patient who was accessed 11 times for post-transplant endomyocardial biopsies.20

Figure 6

Three-year follow-up computed tomography scan post-transhepatic access. Coronal multiplanar reformat (MPR) of contrasted computed tomography scan 3 years after the procedure. The Amplatzer Vascular Plug (blue arrow) is visible in the parenchymal tract outside of the right hepatic vein (black arrow), which remains patent. Note the presence of inferior vena cava occlusion (white arrow), which necessitated the transhepatic access.

Haemostasis can be achieved using a variety of mechanisms, including manual pressure, vascular coiling, vascular plugs and gelfoam, and cauterization with RF energy.3,12,17 Regardless of the technique, it is important that any closure device be within the tract towards the hepatic vein but outside of it, to minimize the risks of thrombosis or embolization. Furthermore, continued monitoring for bleeding is paramount, with all patients being monitored in an intensive care unit (ICU). Intensive care unit monitoring is not obligatory, since many patients undergoing non-cardiac transhepatic access are not admitted to the ICU. However, following serial haematocrits is important, particularly in those patients requiring continued anticoagulation. In patients who required continued anticoagulation, they were started on a heparin infusion, without a bolus, 6 h after sheaths were pulled. They were monitored overnight and if they remained stable, warfarin was started the next day and the heparin infusion was stopped.

In summary, in patients with limited venous access from an inferior approach and who require EP study and possible ablation for cardiac arrhythmias, percutaneous transhepatic access offers an approach that confers similar catheter stability and manipulation as routine ablation procedures from the iliofemoral veins. This study presents a series of patients who underwent successful ablation of their clinical arrhythmias after transhepatic access, which can be done safely through the use of non-invasive imaging including fluoroscopy and ultrasound guidance.21 Furthermore, this series demonstrates that the transhepatic approach can be used to target and successfully ablate several types of arrhythmias.





Conflicts of interest: none.


View Abstract