Transhepatic Cardiac Catheterization in Children
Evaluation of Efficacy and Safety
Background In some children with congenital heart disease, conventional venous access is unavailable for cardiac catheterization. This study investigates a novel transhepatic venous approach to cardiac catheterization in children and evaluates its efficacy and safety.
Methods and Results Percutaneous transhepatic puncture was performed using a 22-gauge Chiba needle under fluoroscopic guidance. After wire exchanges were performed, a 5F to 8F sheath was positioned in the low right atrium and cardiac catheterization was performed. On completion of the catheterization, the sheath was withdrawn and a 3-mm steel coil was placed in the parenchymal tract between the hepatic vein and liver capsule. Liver enzyme studies were obtained before and after transhepatic catheterization, and an abdominal ultrasound was performed to evaluate the liver 24 hours after the procedure. Percutaneous transhepatic cardiac catheterization was performed successfully in 17 of 18 children in whom it was attempted. Patient age was 30±8 months (mean±SEM; range, 1 day to 9 years), weight was 10.5±1.5 kg (3.1 to 27.5 kg), and mean right atrial pressure was 10±1 mm Hg (5 to 19 mm Hg). Time from initial needle puncture to right atrial entry was 6.2±1.2 minutes. Diagnostic catheterization was performed successfully in all 17 children, and additional interventional procedures were performed in 5 children. The total catheterization time was 2.0±0.2 hours. Serum aspartate aminotransferase increased from 57±15 to 78±8 IU/L (P=.06), but alanine aminotransferase and gamma-glutamyl transpeptidase did not change. Ultrasound was performed 24 hours after transhepatic catheterization, and no evidence was found in any patient of hemorrhage or subcapsular hematoma.
Conclusions These data suggest that this novel transhepatic approach provides an effective and safe route for diagnostic and interventional cardiac catheterization in children.
In recent years, pediatric cardiac catheterization has evolved to include primarily therapeutic transcatheter procedures or complex diagnostic studies. In some children, however, traditional percutaneous venous (femoral, subclavian, or internal jugular veins) access may not be available for cardiac catheterization, as a consequence of previous indwelling central venous catheters, cardiac catheterization procedures, surgical procedures (eg, Glenn shunt), or congenital interruption of the inferior vena cava. In some children with complex congenital heart disease requiring multiple and staged interventions, more than one of these factors may adversely affect venous access. In such children, transvenous cardiac catheterization may not be feasible and, therefore, important diagnostic and/or therapeutic procedures may be difficult or impossible to perform.
The purpose of this prospective clinical investigation was to evaluate a novel technique for performing cardiac catheterization in children, the percutaneous transhepatic venous approach. The feasibility of the transhepatic approach, its effectiveness in providing access adequate for diagnostic and therapeutic catheter manipulations, and its safety were evaluated.
Subjects were selected from children who required transvenous cardiac catheterization at C.S. Mott Children’s Hospital at the University of Michigan Medical Center between February 1994 and January 1995. Entry criteria for transhepatic catheterization included children with no femoral venous access or with a congenitally interrupted inferior vena cava who had normal coagulation studies and no active liver or peritoneal disease. Femoral vein occlusion was documented angiographically at the present study, at prior catheterization attempts, or by inability of the cardiologist to gain conventional venous access at the present study (defined as failure to gain access despite >20 minutes of attempts to enter each femoral vein1 ). Three patients underwent transhepatic cardiac catheterization despite the availability of femoral venous access because the transhepatic route was thought to provide some advantage in performing therapeutic procedures.
Precatheterization studies included measurement of hemoglobin, hematocrit, platelet count, prothrombin and partial thromboplastin time, alanine amino transferase (ALT), aspartate amino transferase (AST), gamma-glutamyl transpeptidase (GGTP), two-dimensional echocardiogram, chest roentgenogram, and 12-lead ECG. Abdominal situs and liver location were noted from the chest roentgenogram. Informed consent was obtained from each child’s parent or legal guardian. The protocol was approved by the Institutional Review Board of the University of Michigan Medical Center, Ann Arbor, Mich.
Transhepatic Venous Cardiac Catheterization Procedure
The study was performed with general anesthesia in 10 children and with conscious sedation in 8 children. If an arterial line was required for the catheterization, this was inserted before transhepatic venous catheterization to monitor blood pressure. The right lateral rib cage was prepped and draped in a sterile fashion. The puncture site was identified by fluoroscopy in the midaxillary line, midway between the diaphragm and the lower margin of the liver. Local anesthesia (0.5% to 1% lidocaine) was provided at this interspace and extended deep to the liver capsule. A percutaneous transhepatic puncture was performed with a 22-gauge Chiba needle and stylet (Cook Inc) through this interspace under fluoroscopic guidance (Fig 1⇓). The needle was advanced horizontally (or very slightly posteriorly) to a depth of approximately half the distance to the midline. The stylet was removed, and the needle was withdrawn slowly as a small volume (0.5 to 2.0 mL) of nonionic contrast medium was infused gently. Once the needle tip was positioned in a hepatic vein, contrast was seen to fill a hepatic vein radicle and flow cephalad toward the right atrium. A 0.018-in soft-tipped mandril guide wire (Cook Inc) was advanced into the hepatic vein and into the right atrium. The needle was then removed, and a 4F or 5F coaxial dilator was advanced over the 0.018-in wire to the low right atrium. A 0.035-in J-wire was exchanged, the dilator was removed, and a 5F to 8F sheath was introduced until its tip was positioned in the low right atrium. Heparin (100 U/kg, maximum dose 3000 U) was administered, and transvenous cardiac catheterization was performed through this sheath with standard catheters. In the most recent cases, transhepatic catheter manipulation has been facilitated by the use of a custom sheath with a 45° bend and a radiopaque marker at its tip (RCFW-6.0-35-RB-MUOM-11-294, Cook Inc).
When cardiac catheterization was completed, the activated clotting time (ACT) was measured. If the ACT exceeded 200 seconds, protamine was administered (1 mg per 200 U heparin, maximum dose 25 mg, administered over a 5-minute period) to lower the ACT below 200 seconds before the transhepatic sheath was withdrawn. Then, the catheter was removed, the dilator was replaced, and the sheath was slowly withdrawn from the hepatic vein into the liver parenchyma. A very small volume (1 to 2 mL) of nonionic contrast was injected to ensure that the sheath was no longer within the hepatic vein. A steel Gianturco coil (0.035 in×4cm×3mm, Cook Inc) was placed in the parenchymal tract between the hepatic vein and the liver capsule to minimize the risk of bleeding. Embolization of the parenchymal tract is routine in percutaneous transhepatic portal venous catheterization.2 The sheath was then completely withdrawn and the puncture site covered with sterile dressing.
Patients were treated and monitored after catheterization. Vital signs were checked and the percutaneous transhepatic puncture site was inspected every 15 minutes for 1 hour, every 30 minutes for 2 hours, and then every 2 hours until hospital discharge on the following day. Antibiotic prophylaxis was not routinely provided.
Evaluation of Efficacy and Safety
To document the ease and effectiveness of the percutaneous transhepatic approach, a number of observations were made regarding the technique: the number of transhepatic punctures required for successful entry into the hepatic vein; time from first transhepatic puncture to successful entry into the right atrium; total volume of contrast used for the transhepatic entry procedure; and total duration of the transhepatic cardiac catheterization from the first transhepatic puncture to the completion of the study and removal of the sheath. The subjective ease of performing cardiac catheterization from this approach (“easy,” “moderately hard,” and “hard”) and any chambers or vessels that could not be successfully entered were also noted. The chambers and camera angles used for angiography were recorded, and whether the transhepatic sheath appeared in the field of view was documented. Finally, instances of inadvertent needle entry into the portal venous system, biliary tree, or hepatic artery were noted.
The safety of the transhepatic approach was documented by monitoring of the patients after catheterization and by laboratory tests. The following laboratory studies were obtained: hemoglobin and hematocrit 6 hours after catheterization and the following morning, chest roentgenogram 4 to 6 hours after catheterization and the following morning, and liver enzyme studies the morning after cardiac catheterization. Abdominal ultrasound was also performed the morning after cardiac catheterization to evaluate for possible subcapsular hepatic hematoma, intraperitoneal blood, or thrombosis of the hepatic or portal vein.
Continuous data obtained before and after transhepatic cardiac catheterization are compared by paired Student’s t test. A value of P<.05 was considered statistically significant. Data are presented as mean±SEM.
Our preliminary findings are based on all 18 children in whom transhepatic cardiac catheterization was attempted at our institution between February 1994 and January 1995. Patient age was 30±8 months (mean±SEM; range, 1 day to 9 years), weight was 10.5±1.5 kg (3.1 to 27.5 kg), and mean right atrial pressure was 10±1 mm Hg (5 to 19 mm Hg). The 18 children were diagnosed as having univentricular heart (n=9), tetralogy of Fallot (n=2), critical pulmonary stenosis (n=2), complete atrioventricular canal defect (n=2), pulmonary atresia with intact ventricular septum (n=1), Shone’s complex (n=1) or atrial septal defect (n=1). Conventional venous access was compromised by bilateral femoral venous occlusion (n=12), interrupted inferior vena cava (n=3), superior vena cava obstruction (n=2), and/or use of a bidirectional Glenn shunt (n=3). The transhepatic approach was used in 3 children for interventional procedures despite availability of femoral venous access.
Transhepatic central venous access was accomplished in 17 of the 18 patients in whom it was attempted, requiring 3±0.4 needle punctures (1 to 7 punctures) per patient. The total time from initial puncture to right atrial entry was 6.2±1.2 minutes (1.5 to 21 minutes). The total volume of contrast required for transhepatic entry was 3±1 mL. Transhepatic venous access was not accomplished in a 7-month-old infant with upper airway obstruction and marked hepatic movement with respirations; after 40 minutes, no further attempts at transhepatic entry were made. Diagnostic catheterization was performed successfully in the other 17 children and included catheterization of the superior vena cava, right atrium, right ventricle, and pulmonary arteries. In children with an atrial septal defect or patent foramen ovale, the left atrium, pulmonary veins, and left ventricle were also entered if present. In 1 child with a prosthetic aortic valve and obstructed femoral veins, the left atrium and left ventricle were entered with a transhepatic transseptal puncture (Fig 2⇓). In another child who had hypoplastic left heart syndrome with documented obstructed femoral veins and superior vena cava, a Fontan fenestration was easily crossed from the transhepatic approach. Interventional procedures were also performed in 5 children and included neonatal pulmonary balloon valvuloplasty (n=1), wire puncture and balloon dilatation of a membranous pulmonary atresia (n=1) (Fig 3⇓), pulmonary artery angioplasty (n=2), and successful closure of an atrial septal defect with a buttoned device (n=1). The total catheterization time in the 17 children was 2.0±0.2 hours.
The subjective ease of catheterization from the transhepatic approach was described as “easy” in 13 patients. In 4 patients, the catheterization was described as “moderately hard”; the difficulty encountered was catheter manipulation to the right ventricle (n=3) or superior vena cava (n=1). We have found that forming a catheter loop in the right atrium facilitates transhepatic catheter passage through the tricuspid valve. More recently, the use of a custom sheath with a 45° angle and radiopaque tip has facilitated intracardiac catheter manipulation from the transhepatic approach. The transhepatic sheath partially obscured the left ventricle in the hepatoclavicular view during cineangiography in 1 patient. Inadvertent needle entry occurred into other vessels, including small peripheral branches of the portal vein (n=3) and hepatic artery (n=1). In all four events, no deleterious effects were produced. No instances of bile duct entry occurred.
All patients were clinically stable during and after transhepatic cardiac catheterization. Serum AST increased from 57±15 to 78±8 IU/L (P=.06), but ALT and GGTP did not change (Table⇓). The hemoglobin concentration did not change after the procedure was performed. Hepatic ultrasound 24 hours after transhepatic catheterization (Fig 4⇓) revealed no subcapsular hematoma, no evidence of hemorrhage, and no thrombosis of the hepatic or portal vein in any patient (0% to 70% confidence limits, 0.0% to 5.6%). Chest roentgenograms obtained 6 and 24 hours after transhepatic catheterization revealed no effusions or pneumothoraxes.
Pediatric cardiac catheterization has undergone a major change during the last decade and is now performed most commonly in infants or children with complex cardiac disease. Many of these patients have undergone previous catheterizations and surgical procedures, and some have required extracorporeal membrane oxygenation support before or after surgery. It is our experience that many such children lose the venous access conventionally required for subsequent diagnostic or therapeutic cardiac catheterization procedures. In addition, some infants have congenital interruption of the inferior vena cava, which makes complete cardiac catheterization (especially entry into pulmonary veins) through the femoral vein difficult or impossible. Finally, children who have undergone a palliative Glenn procedure do not have cardiac access through the subclavian or internal jugular vein. In our center, approximately 20 children per year are unable to undergo successful transvenous cardiac catheterization as a result of limitations in peripheral venous access.
Percutaneous transhepatic cholangiography has been available as an effective diagnostic procedure for 2 decades.3 4 During attempts to insert the needle into the biliary tree, a hepatic vein radicle occasionally is entered inadvertently. There have been several previous reports of use of the transhepatic approach to obtain diagnostic information about the portal venous system,5 6 to localize occult neuroendocrine tumors,7 and to perform embolization procedures in patients with cirrhosis and bleeding varices.8 The transhepatic venous approach also has been used to obtain long-term central venous access in some patients.9 Transhepatic hepatic venous cannulation has been described in dairy cows.10 However, there have been no prior reports of the percutaneous transhepatic venous approach for the purpose of cardiac catheterization in children or adults.
The preliminary data presented in this study suggest that the transhepatic approach is an effective method for securing venous access for cardiac catheterization in children. Seventeen of 18 children underwent successful transhepatic cardiac catheterization through which complete diagnostic information was obtained. In addition, 5 children underwent successful transhepatic therapeutic catheterization procedures. In 2 neonates with critical pulmonary stenosis or membranous pulmonary atresia, we found that careful positioning of the transhepatic sheath in the right ventricular outflow tract provided improved directional control and support to the wire and balloon catheter and seemed to facilitate the procedure considerably. We also speculate that the transhepatic route may provide an improved approach to transcatheter atrial septal defect closure because the delivery sheath is more perpendicular to the atrial septum in this approach than in the femoral venous route. In 1 child in the present series, atrial septal defect closure was performed with ease from the transhepatic route.
Previous reports have indicated that percutaneous transhepatic cholangiography and portal venous catheterization can be performed safely. Rare complications have included intraperitoneal bleeding, pleural effusions, thrombosis of the portal venous system, perforation of the gallbladder and bowel, peritonitis, and bile peritonitis.11 12 The overall complication rate reported for the more traditional percutaneous transhepatic procedures has been <5%; in fact, these procedures are often performed on an outpatient basis. Our preliminary data indicate that the percutaneous transhepatic approach for cardiac catheterization in children is also quite safe. There was a small rise in AST, from 57±15 to 78±8 IU/L (P=.06); however, this seems unlikely to present a clinically significant adverse effect. Furthermore, there was no statistically significant increase in other liver enzymes, such as ALT or GGTP. There was no evidence by chest radiograph or abdominal ultrasound of bleeding, subcapsular hematoma, or hepatic or portal venous thrombosis.
These preliminary findings suggest that the transhepatic venous approach to cardiac catheterization is effective and safe in children. The transhepatic technique allows important diagnostic information to be obtained safely in children who have poor conventional venous access and in whom cardiac catheterization may not have been possible previously. We are also encouraged that the transhepatic approach may provide an advantage for certain interventional procedures, such as balloon pulmonary valvulotomy in infants and perhaps transcatheter atrial septal defect occlusion. As more experience and evidence of safety is obtained, the transhepatic approach may become the preferred route for percutaneous venous access in some children. We recommend, however, that pediatric cardiologists obtain initial training and guidance from an interventionalist familiar with percutaneous transhepatic procedures before attempting transhepatic cardiac catheterization.
Reprint requests to Robert H. Beekman III, MD, Division of Pediatric Cardiology, C.S. Mott Children’s Hospital, F1310 MCHC/Box 0204, 1500 E Medical Center Dr, Ann Arbor, MI 48109. Email RBeekman@umich.edu.
- Received February 2, 1995.
- Revision received March 21, 1995.
- Accepted March 26, 1995.
- Copyright © 1995 by American Heart Association
Schwartz SM, Beekman R, Crowley DC, Lloyd TR, Lindauer A, Boczar M. Randomized trial of a 20-gauge Doppler-guided introducer needle for vascular access in infants. Pediatric Res. 1993;33(pt 2):26A. Abstract.
Lunderquist A, Vang J. Sclerosing injection of esophageal varices through transhepatic selective catheterization of the gastric coronary vein: a preliminary report. Acta Radiol Diagn. 1974;15:546-550.
Slepetis R, Cohick WS, Bauman DE, Hackett R. Surgical cannulation of a hepatic vein in dairy cows utilizing diagnostic ultrasound. J Dairy Sci. 1987;70:571-575.
Hoevels J, Lunderquist A, Owman T. Complications of percutaneous transhepatic catheterization of the portal vein and its tributaries. Acta Radiol. 1980;21:593-601.