scientific session 23 portal venous intervention
TRANSCRIPT
200
(6.1% vs 15.7% at 1 year, 6.1% vs 27.3% at 2 years,respectively) .
CONCLUSION Catheter fixation method improves catheter stability and HA patency, and enables patients to
undergo long-term arterial infusion.
Scientific Session 23Portal Venous InterventionModerator: Robert K. Kerlan, Jr., MD
1:00 pm
Effect of IV Vasopressin on the PortosystemicGradient During TIPSMichael C. Soulen, Philadelphia, PA
PURPOSE: Systemic vasopressin decreases portal venous
pressure 200/0-30% and is used to treat acute variceal
bleeding. This study evaluated whether pharmacologic
suppression of portal pressure alters the portosystemicgradient, which is used as an endpoint during TIPS
placement.
METHODS: Nine patients underwent TIPS for variceal
bleeding while receiving IV vasopressin. Portal vein
and right atrial pressures were measured after TIPS
placement, and again 20 minutes after weaning off the
vasopressin infusion. Splenic venograms were alsoperformed before and after withdrawal of vasopressin.
RESULTS: Eight patients were receiving vasopressin at
0.4 units/min and one at 0.2 u/min. TIPS was success
ful in all cases. After stopping vasopressin the porto
systemic gradient increased in six of the nine patients.
In three patients, the gradient rose from ~12 mmHg to
14-17 mmHg. The mean portosystemic gradient afterTIPS placement was 9.3 ± 4.4 mmHg (S.D.), range4-15 mmHg. After weaning off vasopressin, the meanportosystemic gradient was 10.3 ± 4.6 mmHg, range4-17 mmHg (p = 0.45). There was no qualitativechange in variceal opacification after stopping vasopressin.
CONCLUSION On average, the portosystemic gradientacross a TIPS is not significantly altered by systemic
vasopressin. However, in some cases the gradient rises
above the endpoint of 12 mmHg when vasopressin is
withdrawn.
1:15 pm
Intraluminal Irradiation for TIPS Stenosis: Resultsin a Swine ModelTodd Lessie, MD, Salt Lake City, UT. Howard Nelson,MD • Franklin j. Miller, MD. David Fillmore, MD.
Hyo-Chun Yoon, MD, PhD
PURPOSE: To evaluate the ability of 32Phosphorus in
traluminal irradiation to reduce pseudointimal hyperpla
sia (PIB) in TIPS.
MATERIALS AND METHODS: TIPS were successfullyplaced in eleven swine with normal portal pressures. Six
animals received 16 Gy intraluminal irradiation to the
hepatic parenchyma and venous outflow tract at the timeof TIPS placement using a Sodium 32Phosphorus e2p)
filled balloon angioplasty catheter. Five control animalsunderwent TIPS and balloon angioplasty with saline. All
animals were followed for 28 days, at which time per
cutaneous portography was performed, animals sacrificed, and the tissue around the TIPS stent processed for
histologic analysis. Maximum PIH as a percentage of
estimated TIPS diameter was calculated for each animal.
RESULTS: At the time of sacrifice, all five control TIPS and
all but one irradiated TIPS were occluded. Histologic
analysis demonstrated considerable variability in the de
gree of pseudointimal hyperplasia within each TIPS and
between animals. No significant difference was found in
the maximum PIH between control and irradiated animals.
CONCLUSIONS: Irradiation of TIPS with 16 Gy 32p deliv
ered at the time of TIPS placement did not significantly
improve TIPS patency or reduce the degree of pseudo
intimal hyperplasia in swine with normal portal pressures.
1:30 pm
Bile Leak in TransjuguJar IntrahepaticPortosystemic Shunt: Effect on SMC ProliferationGao-jun Teng, MD, Lebanon, NH. Michael A.
Bettmann, MD. PaulJack Hoopes, DVM; PhD.
Robert j. Wagner, MA • Byeong-Ho Park, MD. LiYang, MD
PURPOSE: To evaluate the effect of bile on smoothmuscle cell (SMC) proliferation in vitro and in vivo in aporcine TIPS model.
MATERIAL AND METHODS: In vitro, SMCs explantedfrom porcine thoracic aorta were cultured utilizing standard technique, and were then subcultured in one ofthree groups: 1% PS [porcine serum] + 1% bile; 10% PS+ 1% bile; 10% PS. Cells were harvested at 3, 10 or 14days with measurement of DNA, Protein and DPM. Invivo, TIPS was successfully performed in 45 swine. The
proliferative response within the stent was histologically
quantified and correlated to evidence of bile leak.
RESULTS: From pilot studies, 2.5-10% bile solutions
caused 100% SMC mortality by 3 days. With 1% solutions
(with and without PS), both DNA and protein production
decreased significantly compared to PS alone (P < .05).
In the in vivo model, 13 of 45 specimens (29%) showedbile leakage by gross or microscopic examination. There
was less SMC proliferation in animals with than in those
without bile leak. Thrombus formation was the primary
component of TIPS occlusion or stenosis in specimens
with bile leak, while myofibroblastic proliferation was
the major component in the group without bile leak.
CONCLUSIONS: These data, from both in vitro and invivo studies, suggests that the proliferative response inTIPS is not primarily due to bile leak.
1:45 pm
Comparison of Polyethelene Teraphthalate (PET)Covered Wallstents Versus Uncovered Wallstentsfor TIPS Creation in an Animal ModelZiv Haskal, MD, Philadelphia, PA
PURPOSE: To evaluate the safety and efficacy of a polyethelene teraphthalate (PET) covered Wallstent for TIPScreation in an animal model.
MATERIALS AND METHODS: Thirteen TIPS were createdin 13 minipigs: 8 lined with PET-covered Wallstents (water permeability: 750-2000 cc/cmjeP2/min), 5 with standard Wallstents. Shunt venography was performed at5-8 weeks, necropsy at 7-8 weeks. Histology, immunohistopathology, and electron microscopy was performedon the explants.
RESULTS: The devices proved equally easy to deploy. At7-8 week venography, the mean shunt stenoses of thecontrol and graft groups were 45% (range 26-83%) and53% (range 0-90%), respectively. There was 1 graft TIPSocclusion. All stenoses developed within the liver tract.One stenosis was due to trans-stent proliferation of normalporcine liver tissue. Typical myofibroblast and collagenmatrix proliferation encompassed both control and graftTIPS. There was no evidence of bile staining within thistissue. A neointima of varying thickness lined the lumina inboth groups. Slightly more pronounced inflammation waspresent at the graft ends (site of PET material fixation).
CONCLUSION This TIPS stent-graft is easy and safe toplace. Its healing response and patency, in this specificanimal model, are similar to bare stent TIPS.
2:00 pm
Percutaneous Retroperitoneal Splenorenal Shunt:An Experimental Study in SwineToshio Kaminou, MD, PhD, Portland, OR. Barry T.
Uchida, BS. ShOji Sakaguchi, MD • Dusan Pavcnik,MD, PhD. Hans A. Timmermans, BFA. Frederick S.Keller, MD
PURPOSE: To evaluate percutaneous retroperitonealsplenorenal shunt (PRESS) in swine, which potentiallycan be a new minimal invasive treatment of portal hypertension.
MATERIAL AND METHODS: Eight domestic swine underwent PRESS. A modified TIPS puncture set wasintroduced from the right femoral vein into the leftrenal vein. The proximal splenic vein was punctured.A Wallstent and a PTFE covered stent-graft were thenplaced connecting both veins through the retroperitoneal cavity. The animals were followed up weekly bysplenic venography until the shunt occluded.
RESULTS: Puncture of the splenic vein and stent placement was successfully completed in all cases. In seven ofeight animals, good flow from the splenic vein to the leftrenal vein through the shunt tract was observed onehour after PRESS. No changes in vital signs were noticedduring the procedure. Five of eight tracts showed obstruction one week after PRESS. At necropsy minimalretroperitoneal hematoma along the tract were found.Stents and stent grafts were positioned in the retroperitoneal cavity connecting both veins in all cases. Remarkable intimal hyperplasia with clots was present in thetracts.
CONCLUSIONS: Creation of a percutaneous splenorenalshunt is technically possible and further experimentalstudies of its efficacy are warranted.
2:15 pm
Portal Vein Embolization for Extension ofHepatectomy IndicationsThierry de Baere, MD, Villejuif, France. Alain Roche,
PhD. Dominique Elias, MD. Philippe Lasser, MD
PURPOSE: To render hepatectomy feasible in patientswith an initially deficient volume of the future remnantliver (FRL).
MATERIALS AND METHODS: Portal blood flow wasredistributed towards the FRL with preoperative embolization of the portal veins (POPE) feeding the future resected liver. POPE was performed in 38 patientswith malignant tumors and in 2 with benign lesions.POPE was performed under fluoroscopic guidance,via a percutaneous transhepatic access to the portalvein.
RESULTS: FRL volume values were 90-560 cc (m = 271cc) before POPE and 160-933 cc (m = 455 cc) afterPOPE, which represents a volume increase of 83 ± 49%.Hypertrophy of the FRL was 92% ± 62% after 31 dayswith Cyanoacrylate, 53% ± 6% after 43 days with Gelfoam, and 44 ± 30% after 35 days with coils. Slightshrinkage was obtained in the volume of the embolizedliver. POPE was well tolerated and surgery was possiblein 32 patients. Seven operations were cancelled due totumor spread, but only one due to insufficient hypertrophy of the FRL.
CONCLUSION Although exclusively applicable in a limited subset of patients, POPE widens the possibilities ofcurative hepatectomies, because it induces hypertrophyof unembolized liver segments. Among embolic materials, cyanoacrylate seems to ensure better and faster hypertrophy.
Scientific Session 24Venous ThrombolysisModerators: Martin Richard Crain, MDj Julien 1.
Struyven, MD201