| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Experimental Studies |
1 From the John S. Dunn Center for Radiological Sciences, Section of Vascular and Interventional Radiology, Department of Diagnostic Radiology (M.J.W., K.A., K.C.W.) and the Department of Veterinary Medicine and Surgery (L.C.S.), University of Texas M.D. Anderson Cancer Center, Unit 325, 1515 Holcombe Blvd, Houston, TX 77030-4009. Received July 26, 2002; revision requested September 24; revision received October 14; accepted December 10. Supported in part by a grant from the John S. Dunn Research Foundation and by grant NIH-NCI CA-16672 from the National Cancer Institute. Address correspondence to M.J.W. (e-mail: mwallace@mdanderson.org).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: PV access from the inferior vena cava (IVC) to the main PV was performed in eight pigs by using intravascular US guidance. Either a prototype stent-graft (n = 6) or Wallgraft (n = 2) was used to create the shunt. Intravascular US demonstrated the main PV to be in direct contact with the IVC in all animals. A mean of 1.75 needle passes were needed to enter the PV. Immediate postprocedure computed tomography (CT) of the abdomen helped identify and quantify the presence of hemoperitoneum. Shunt venography was performed at 2 weeks, followed by necropsy.
RESULTS: PV access and TEPS creation were successful in all animals. Contrast medium extravasation, due to inadequate coverage of the portacaval tract, was identified in four procedures and addressed by the placement of additional devices in three cases and prolonged balloon inflation in one. Abdominal CT demonstrated small amounts of hemoperitoneum in five animals and moderate to large amounts in three. Two animals did not live to the 2-week follow-up study. One animal was sacrificed on the day of the procedure owing to intraperitoneal hemorrhage; the second died of intussusception-related bowel necrosis 10 days after TEPS creation. Shunts were occluded or severely stenotic at venography and necropsy in the remaining six animals.
CONCLUSION: TEPS is technically feasible after intravascular US–guided PV access.
© RSNA, 2003
Index terms: Animals • Interventional procedures, experimental studies, 957.1268 • Shunts, portacaval, 957.1268, 957.711 • Stents and prostheses • Ultrasound (US), guidance, 957.12986
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Shunt stenosis and occlusion occur in about half of all cases after TIPS creation (1,4). Stenosis is believed to result from pseudointimal hyperplasia, most commonly in the outflow hepatic vein (5). TIPS dysfunction may be compounded by leakage of bile into the shunt due to trauma to the liver during the procedure (6,7) or bile duct proliferation within the tissue lining the shunt, producing stenosis (8). The transhepatic nature of this technique and the shunt length may prove to be the underlying sources of these problems. Early results (9,10) regarding the use of stent-graft technology for de novo TIPS creation to prolong shunt patency have been promising, but stenoses within the hepatic vein above the stent-graft have been encountered.
A new approach in which imaging guidance is used for portal vein access could simplify the process and reduce procedure-related complications. A short portacaval shunt that does not travel through the hepatic parenchyma may remain patent longer. The purpose of this investigation was to evaluate, in a swine model, the feasibility of intravascular ultrasonography (US)-guided access of the extrahepatic segment of the main portal vein to create a transvenous extrahepatic portacaval shunt (TEPS) as an easier and more durable alternative to TIPS.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Eight adult domestic pigs (M.D. Anderson Bastrop Facility, Bastrop, Tex) weighing 29.6–33.7 kg were sedated with an intramuscular injection of a solution containing 15 mg per kilogram of body weight (mg/kg) ketamine hydrochloride (Vetamine; Schering-Plough/Animal Health, Union, NJ), 0.15 mg/kg acepromazine maleate (Fermenta Animal Health, Kansas City, Mo), and 0.04 mg/kg atropine sulfate (American Pharmaceutical Partners, Los Angeles, Calif). Anesthesia was then induced with 5% isoflurane (Iso-Thesial; Abbot Laboratories, North Chicago, Ill) administered by means of a mask. Once anesthesia was established, an endotracheal tube was inserted and anesthesia was maintained with 1.5%–3.0% isoflurane and 0.8 L/min oxygen. The animals were given 5 mg/kg intramuscular doses of the antibiotic enrofloxacin (Baytril; Bayer, Shawnee Mission, Kan) before each procedure and once daily for 5 days after the procedure. All shunt procedures were performed by two of the authors (M.J.W., K.A.) working together.
Under aseptic conditions and with US guidance, the right and left common femoral veins of the animals were accessed by using a 21-gauge needle (Micropuncture set; Cook, Bloomington, Ind). The Seldinger technique was used to enlarge the access site. A 35-cm-long 9-F angled sheath (Flexor; Cook) was inserted into the right femoral vein, and a 25-cm-long 8-F sheath (Pinnacle; Boston Scientific/Medi-tech, Natick, Mass) was inserted into the left femoral vein. Both sheaths were advanced into the midportion of the inferior vena cava (IVC). All animals were given a single intravenous bolus of 100 U per kilogram of body weight of heparin sodium (American Pharmaceuticals, Los Angeles, Calif) after vascular access was established.
A 6-F, 12.5-MHz, intravascular US catheter (Sonicath Ultra; Boston Scientific) was then introduced via the left femoral vein access point and advanced into the upper portion of the IVC. The intravascular US catheter and an imaging system (ClearView Ultra; Boston Scientific) were used to identify the main portal vein at a level where it abuts the IVC (Fig 1). In all animals, intravascular US demonstrated that the main portal vein was in direct contact with the IVC, with no appreciable intervening hepatic parenchyma. A 14-gauge, blunt, curved cannula from a transjugular access set (AngioDynamics, Queensbury, NY) was positioned within the IVC just caudal to the intravascular US catheter. With real-time intravascular US guidance, the cannula was rotated anteriorly until tenting of the portal vein was identified (Fig 2). The guide wire was then removed, and a 67-cm-long 21-gauge needle with a tapered, back-loaded, 5-F catheter from the transjugular access set (AngioDynamics) was used to puncture the portal vein under intravascular US guidance (the needle extends several centimeters beyond the catheter). The needle was advanced until a "pop" was felt. This usually corresponded to through-and-through penetration of both walls of the portal vein. Gentle aspiration was applied to the hub of the needle as it was pulled back, until blood return was obtained. A mean of 1.75 needle passes (range, 1–4) per animal were required to enter the portal vein. A 0.018-inch control guide wire (V-18; Boston Scientific) was advanced into the portal vein, and the 5-F tapered catheter was then passed over the 21-gauge needle and guide wire into the portal system. The number of needle passes required to enter the portal vein was recorded for each procedure. After portography, the portacaval access site was enlarged over a 0.035-inch Amplatz wire (Boston Scientific) in a single step in which the 5-F catheter was exchanged for the original sheath dilator and was advanced across the tract.
|
|
Technical success of the procedure was defined as the ability to reliably identify and access the portal vein at the desired extrahepatic level and the ability to form a patent portacaval shunt without evidence of extravasation on the last venogram obtained prior to femoral vein hemostasis. Complications of hemorrhage during shunt creation were assessed with observation of extravasation at procedural venography (Fig 3) and documentation of hemoperitoneum at postprocedural computed tomography (CT).
|
|
All animals were extubated and allowed to recover from anesthesia. The experimental protocol included follow-up shunt venography at 2 weeks after TEPS creation. Sedation and anesthesia were induced as described above. The right femoral vein was accessed with US guidance, and a 5-F cobra catheter (Cook) was positioned in the body of the shunt or at its origin in the IVC and digital subtraction venography was performed.
Necropsy
Seven of the eight animals were euthanized by means of exsanguination under deep anesthesia at completion of the study period or sooner if procedural complications were encountered or if the animal did not recover appropriately from the procedure. One animal (animal 2) died 10 days after shunt creation. All animals were subjected to complete necropsy performed by two authors (M.J.W., L.C.S.) working together. Necropsy included removal of the involved portal vein, IVC, and adjacent connective tissues. The specimens were evaluated grossly to assess the position of the shunt with respect to the lumen of the portal vein and IVC and to identify the presence of intimal hyperplasia or thrombus within the shunt or adjacent venous structures. The peritoneal contents were also grossly inspected at the time of necropsy by a veterinary pathologist (L.C.S.) for secondary findings of prior substantial hemoperitoneum. The color and character of peritoneal fluid was evaluated, and the presence of fibrin deposits on the peritoneal surfaces and the color of nearby peritoneal and mesenteric lymph nodes were noted.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
Two weeks after the procedure, six animals underwent shunt venography. Two animals did not live long enough for the 2-week follow-up examination because of complications. Shunts were occluded or severely stenotic in the six animals that returned for venography at 2 weeks.
Necropsy Results
All but two pigs (animals 2 and 3) recovered appropriately from the anesthesia and procedure and resumed normal physical activities and oral intake within 24 hours of shunt creation. Animal 3 exhibited listlessness, pallor, and diminished body temperature in the hours after the procedure. Paracentesis performed 4 hours after the procedure demonstrated gross hemoperitoneum, a change from the CT findings of a small amount of peritoneal fluid immediately after TEPS. At necropsy, the distal legs of the prototype stent-graft were situated in the portacaval space, but it was not clear whether the device had retracted into the portacaval space, had been deployed incorrectly, or had been pulled out of the portal vein during necropsy. The lack of extravasation at completion venography and the lack of substantial peritoneal fluid at CT scanning favors retraction as the explanation. Moreover, as noted earlier, TEPS placement in this animal had been complicated by initial problems with stent-graft deployment, necessitating placement of a second prototype stent-graft to adequately cover the portacaval tract. Animal 2 failed to thrive (poor oral intake and physical activity) in the days after the procedure and died 10 days after shunt creation. Necropsy revealed bowel necrosis secondary to intussusception, unrelated to the TEPS procedure. Secondary findings of substantial hemoperitoneum, including fibrin tags on peritoneal surfaces, were identified at necropsy in one of six animals (animal 6) followed up for more than 2 weeks. This animal was one of the four animals in which extravasation was identified during TEPS and a large amount of hemoperitoneum was present on postprocedural CT scans.
The peritoneal fluid and regional lymph nodes were otherwise normal in color and character in all animals except the one (animal 3) that was euthanized within 4 hours of the procedure (Table).
At necropsy, the connection between the portal vein and the IVC was extrahepatic in all animals, with a well-defined portacaval space above and below the shunt (Fig 4). Severe intimal hyperplasia involving the portal venous and IVC ends of the prototype stent-graft was found to be the cause of shunt failure in animals 1 and 4 (Fig 5). Thrombus within the shunt and lesser degrees of intimal hyperplasia were present in the final four animals in which Wallgrafts were used. A minimal amount of intimal covering was present along the course of the prototype stent-graft in one animal (animal 2) examined 10 days after TEPS creation. The portal vein and IVC were patent at necropsy in all eight animals, but a small nonocclusive thrombus 2 cm long and 0.5 cm in diameter extended from the shunt into the IVC in animal 4. No inflammatory changes where identified in the region of the portacaval space where shunts were created.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Experience with TIPS over the past 12 years has uncovered substantial problems with shunt durability that require rigorous surveillance and reintervention to maintain respectable rates of primary assisted patency (87%) and secondary patency (89%) at 3 years (13). Initial investigations with stent-grafts in animals and humans are promising in terms of reductions in shunt stenosis due to intimal hyperplasia and premature shunt failure due to bile leakage. In animal investigations, Nishimine et al (14) in 1995 and Haskal et al (15) in 1997 reported substantial reductions in stenoses when minimally porous expanded-PTFE stent-grafts were used to create TIPS. Preliminary clinical studies by Otal et al (9) of the commercially manufactured (not FDA-approved) Viatorr stent-graft (Gore, Flagstaff, Ariz) showed that shunt stenosis was reduced when the Viatorr device was used to revise failing shunts in five patients and for de novo shunt creation in 15 patients with cirrhosis. Otal et al reported primary and secondary patency rates of 80% and 100% at 387 days after insertion of the stent-graft. Cejna et al (10), reporting their experience with the same stent-graft in 16 de novo TIPS procedures, showed that despite a mean 289-day "in-graft" patency rate of 100% in the 10 patients available for follow-up, 10 of the original 16 patients required revisions, predominantly because of hepatic vein stenosis above the stent-graft.
In addition to using stent-grafts to improve shunt patency, several investigators have looked at alternative transcaval approaches to avoid intimal hyperplasia in the hepatic vein and to shorten the shunt in hopes of improving its durability. Quinn et al (16) reported creating CT-guided transcaudate direct portacaval shunts in eight patients, and Petersen et al (17) created similar direct intrahepatic portacaval shunts under intravascular US guidance in five pigs and five patients. The authors of both of these feasibility studies demonstrated technical success with use of PTFE-covered balloon-expandable stents in all cases, but neither could comment on long-term shunt patency because of limited follow-up. The TEPS approach, like those described by Quinn et al and Peterson et al, avoids the problem of intimal hyperplasia in the hepatic vein that drains the shunt. An additional advantage of TEPS is the lack of parenchyma along the course of the shunt, which eliminates the potential for bile leakage and premature failure.
A key procedural advantage of intravascular US–guided TEPS over TIPS procedures is the opportunity for guided entry into the portal vein. In our study, portal vein entry was easily accomplished in a mean of fewer than two needle passes. Similar results were obtained by Petersen et al (17) when they used a modified Rosch-Uchida portal access set (RUPS-100; Cook) for their intravascular US–guided procedure. The 21-gauge needle in that access set was abraded to create an echogenic tip. We found that rotation of the unmodified curved cannula from the transjugular access set caused tenting of the portal vein (Fig 2), allowing ready identification of the entry site. Moreover, the tip of the intravascular US catheter is radiopaque and acts as a fluoroscopic marker during portal vein entry and stent-graft deployment.
In humans, the portacaval space has been evaluated as a potential site for percutaneous portacaval shunts. McLoughlin and Rankin (18) retrospectively reviewed contrast-enhanced CT scans in 100 patients and determined that the shortest distance between the middle portion of the main portal vein and the IVC was a mean of 6.1 mm ± 4.8. At this level, no intervening structures were present in 64% of cases, but liver (24%), nodes (10%), and hepatic artery (2%) were identified in the remaining cases. CT imaging before the procedure permits assessment of the feasibility of TEPS with regard to the presence of intervening structures and provides accurate measurements for selection of a stent-graft of appropriate length. Procedural complications related to accessing the portal vein can be reduced when intravascular US is used to guide TEPS placement to help avoid vital structures within the portacaval space.
The clinical application of an intentional extrahepatic portacaval shunt was described by Nyman et al (19), who created a direct mesocaval shunt between the superior mesenteric vein and the IVC with a handmade covered stent to bridge the two venous structures. A 20-gauge needle was inserted from the anterior abdominal wall through the transverse colon, across the superior mesenteric vein, and into a vascular retrieval device placed in the IVC. A 0.018-inch guide wire was inserted through the needle and captured in the IVC to create through-and-through guide wire access. The remainder of the procedure was performed with fluoroscopic monitoring. Although no intraperitoneal hemorrhage occurred, the shunt became occluded within 24 hours. We found the intravascular US-guided TEPS approach to be easier and less complex than the technique used by Nyman et al.
Inadvertent extrahepatic puncture of the portal vein is a complication of TIPS (20–22) that can result in fatal hemorrhage (22). Also, when the extrahepatic portal vein entry site is enlarged to 8 mm or more, hemorrhage and hypotension typically occur immediately after tract dilation, before the stent is deployed. These complications can be managed conservatively with rapid deployment of a Wallstent and completion of the TIPS procedure (20,21). TEPS does not involve a predilation step. One major limitation of the TEPS approach is the potential for major bleeding complications that may follow malpositioning of the stent-graft.
Extravasation occurred in our study when problems were encountered with stent-graft deployment. In three animals, the stent-grafts were too short. When constructing the prototype, we anticipated the portacaval shunt to be deployed horizontally; instead, the direction of portal venous access and the final shunt were angled cephalad from the IVC to the portal vein (Fig 3a), which necessitated deployment of a longer stent. This change in angle resulted in malpositioning of the devices, with the legs of the flanged portion of the stent-graft deployed in the portacaval space, and incomplete coverage of the tract. These issues of length and delivery orientation of the stent-graft will be addressed before future studies are undertaken. Extravasation due to incomplete coverage of the portacaval tract was managed by placing a Wallgraft within the prototype stent-graft in two animals. We chose to use the Wallgraft in the final two animals to complete the feasibility portion of our study.
In the six animals that returned for venography, all shunts were confirmed to be occluded at necropsy performed 18–21 days after TEPS creation. Intimal hyperplasia was severe and the cause of occlusion in two animals (Fig 5) in which the prototype stent-graft was used, and shunt thrombosis was the cause in the four animals in which polyethylene terephthalate–covered Wallgrafts were used. Unlike PTFE-covered stent-grafts, polyethylene terephthalate–covered stents have shown no improvements in shunt patency in investigations in animals (23,24). An alternative stent-graft or modifications to our prototype device will be required for future studies to adequately evaluate shunt patency for this alternative approach. The orientation of our portacaval shunt may have also played a role in the early occlusions encountered. The direction of venous flow would be retrograde in the shunt from portal vein (cephalad) to IVC (caudad), and the lack of portal hypertension in this model is believe to further hinder adequate shunt flow to maintain patency. We plan to address this issue in further studies and explore the jugular approach to reverse the orientation of venous flow through the shunt. Artificial creation of portal hypertension in this model may also be required to enable adequate evaluation of the durability of this technique.
In summary, we found that creation of a TEPS is technically feasible with intravascular US–guided access to the portal vein. Additional studies with further modified stent-grafts are needed to improve the reliability of stent-graft deployment before shunt durability can be evaluated.
Practical application: The TEPS procedure offers an alternative approach to the standard TIPS procedure for treating patients with portal hypertension. Intravascular US guidance can be used to simplify the initial crucial step of portal vein access, where most bleeding-related TIPS complications occur. The extrahepatic nature of the approach where the two large venous structures are closest together allows a short venovenous anastomosis and perhaps results in a more durable shunt. The TEPS approach can also be used as an alternative procedure in situations where standard TIPS creation would be more difficult or perhaps contraindicated due to portal venous thrombosis, advanced hepatic malignancy, polycystic liver disease, and Budd-Chiari syndrome.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, M.J.W.; study concepts, M.J.W.; study design, M.J.W., K.A., K.C.W.; literature research, M.J.W.; experimental studies, all authors; data acquisition and analysis/interpretation, M.J.W., K.A., L.C.S.; manuscript preparation, M.J.W., K.A., K.C.W.; manuscript definition of intellectual content, editing, and revision/review, M.J.W.; manuscript final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
