Chapter 14: The Difficult Transjugular Portosystemic Shunt Procedure: Unconventional Techniques Although transjugular intrahepatic portosystemic shunts (TIPS) have become a mainstream procedure, they can still be technically challenging. Technical success rates are generally 90% to 100%,1,2 although this tends to overestimate the ease with which these procedures are accomplished. This chapter covers anatomic problems that can increase the difficulty of a TIPS and strategies for achieving technical success despite these obstacles. Indications for variations in standard tips technique include any case in which anatomic factors preclude performing TIPS in the standard way. These include but are not limited to: • Venous access stenosis or occlusion • Difficult angulation from the inferior vena cava (IVC) to the hepatic veins (HVs) • HV occlusions • Hepatic parenchymal problems (cysts, tumors, or fibrosis) • Portal vein (PV) occlusions Contraindications to specialized TIPS techniques include: • The usual contraindications for any TIPS procedures (see other chapters) • Lack of operator experience because some of these specialized techniques should not be attempted unless the operator is comfortable with standard TIPS procedures The initial step of any TIPS requires getting to the HVs from the venous access site. Occlusion of the right internal jugular (RIJ) vein can make TIPS more difficult right from the onset. It may be possible to puncture a thrombosed RIJ and recanalize the vein. Alternatively, a collateral vein may be useable if it communicates in a straight line down to the superior vena cava (SVC). If the right-sided access is not possible, the left IJ can be used. Hausegger et al3 reported a 92% technical success rate for 12 TIPS done from a left IJ approach. If the SVC is occluded, preventing access from above, performing the TIPS from a femoral approach has been described. One technique is to advance the puncture needle from an inferior HV into a portal branch.4,5 Alternatively, a fine needle can be advanced percutaneously through a main portal branch into the IVC using CT or ultrasound guidance. A wire passed through this needle can then be snared from a femoral approach, and the rest of the case can then be done from the femoral access.6 The angle between the HV and the IVC can be very acute, particularly if the liver is small and pushed cephalad by significant ascites. This can make it difficult to advance the access needle into the HV. Adding some additional curve to the puncture needle can make it easier to advance the needle into the HV. The needle should be bent with a stiff wire in the lumen to avoid kinking the needle. Also, it is best to add the curve over several centimeters rather than trying to bend the needle too acutely in any one area. An acute angle from the IVC into the HV can also increase the chances that the needle may pop out into the IVC as the needle is positioned to make passes. Because breathing can cause caudal motion of the liver, light continuous pressure should be maintained on the needle to avoid the needle’s popping out of the HV. Hepatic vein occlusion can also present a major challenge. In Budd-Chiari syndrome, there is often still an HV stump. If this stump can be engaged with the access needle, passes toward the PV can be initiated from this stump ( Because there is a higher incidence of hepatocellular carcinoma (HCC) in patients with cirrhosis, the radiologist may encounter patients with HCC who need TIPS. There are theoretic concerns about creating a TIPS tract through a tumor, including an increased risk of bleeding and hematogenous seeding of the tumor. Careful analysis of prior cross-sectional imaging may allow one to choose a path that does not transgress the tumor. This may mean using an atypical pathway such as the left HV to the left PV to avoid a large right lobe HCC. Polycystic liver disease has been previously considered to be a contraindication to TIPS, presumably for fear of potential bleeding from the parenchymal tract into one of the cysts. However, several reports have described safe creation of TIPS in this setting.8–11 Again, analysis of prior computed tomography or magnetic resonance imaging scans may help choose a path through a portion of liver that may be relatively free of cysts. As the access needle is advanced through the liver parenchyma, if a cyst is encountered, the needle may be redirected around the cyst, although it has never been proven that it is necessary to avoid going through cysts. With modern use of stent grafts for TIPS, the cyst would presumably be excluded from the blood flow through the shunt. The cysts cause expansion of the liver; thus, the tract from the HV to PV will be longer than for a TIPS in a liver without polycystic disease. Thus longer or overlapping stent grafts will be needed to line the tract. In small, shrunken, cirrhotic livers, the PV is more cephalad relative to the HV compared with the relationship seen in normal livers. In this situation, it is difficult to advance the access needle anteriorly enough, and the standard needle curve tends to pass caudal to the right portal trunk. Adding additional curve to the access needle can compensate and allow access into the PV. Alternatively, a long 21-gauge Chiba needle (Cook Inc., Bloomington, Indiana) can be advanced through the larger access needle, and this Chiba needle can be curved near its tip to provide additional curvature to the needle pass.12 If the PV is exceptionally high and cannot be punctured from the HV, the gunsight technique can be used.13 With this technique, a snare is placed into the right PV via a percutaneous access, and a second snare is advanced from the jugular access into the HV. The two snares are fluoroscopically lined up by angling the image intensifier. A needle is then passed percutaneously through both of these snares. A wire passed through the percutaneous needle is pulled out the jugular access by the HV snare. This provides through-and-though wire access, which allows catheters to be passed from the jugular access across the tract into the PV.
Introduction
Indications and Contraindications
Procedures
Access Problems
Hepatic Vein Problems
Fig. 14.1). If no HV stump can be catheterized, needle passes can be made directly from the IVC. It is important to start in the upper 5 or 6 cm of the IVC because this is the intrahepatic segment. Puncturing from below this level risks starting from an extrahepatic location, raising the chance of hemorrhage. Because wedged hepatic venography cannot be performed, one may not know where to aim the needle. Boney landmarks can provide a rough guide. Using the patient’s own vertebral body width as a unit of measure, 90% of right portal trunks are approximately one vertebral body width lateral to the spine.7 Alternatively, the needle may be partially advanced to engage the liver parenchyma, and an intraparenchymal injection of CO2 will often opacify the portal system (
Fig. 14.2).
Liver Parenchymal Problems
Portal Vein Problems
Thoracic Key
Fastest Thoracic Insight Engine
