03:36 PM Abstract No. 173 Effect of shunt geometric parameters on transjugular intrahepatic portosystemic shunt patency

Abstract

To assess the effect of shunt geometric parameters on transjugular intrahepatic portosystemic shunt (TIPS) patency and to quantitatively analyzed change of stent configuration. A database of patients who had TIPS performed from November 2011 to June 2015 was generated. The angiographic images during initial TIPS creation were reviewed independently by two observers. Stent-to-inferior venous cava distance (SIVCD), hepatic vein to stent angle (HVSA) and portal vein to stent angle (PVSA) was measured in two orthogonal planes. The patients meet the standard that SIVCD < 15mm, HVSA < 45° and PVSA < 45° were classified into the optimal position (OP) group, others classified into non-optimal position (NOP) group. To quantitatively compare the configuration change of stent-graft, the angle between vertical line and tangent of proximal end of stent (AVTP), distal end (AVTD) and stent bending angle (SBA) were measured in portographic images during TIPS creation and follow-up, respectively. Cox regression analysis was performed to assess these parameters on primary patency of TIPS, hepatic encephalopathy and survival. Among all 495 patients, 327 patients were enrolled in OP group. During the follow-up of mean 698 days, the TIPS dysfunction rate among the patients in OP group was 6.1%, compared with 30.0% among the patients in NOP group (P<0.001). Portogram was conducted in 40 patients during the follow-up period. AVTP was significantly decreased by 5.1±8.7° (P = 0.001) and SBA was significantly increased by 9.2±8.5° (P < 0.001). It suggested that stent had tendency to straighten, but the variation was not correlated with time (P = 0.342). The cumulative TIPS patency rates at 1, 2 and 4 years were 93%, 87% and 73%. A Cox multivariate analysis revealed that optimal position of stent (HR = 4.948; 95% CI: 2.929-8.358; P < 0.001) and splenectomy (HR = 2.410; 95% CI: 1.335-4.350; P = 0.004) was associated with TIPS dysfunction. Initial stent position is predictive of TIPS patency, with stent extending to the hepatocaval junction and reducing the angle with vessels. The stent-graft will gradually straighten after TIPS, thereby impacting the shunt patency.Tabled 1Comparison of Geometric Parameters and Prognosis Between the OP and NOP GroupsOP GroupNOP GroupP ValuePatients (n)327168SIVCD (mm)8.8±3.411.6±4.5mm0.002HVSA-a (°)30.2±8.635.1±11.60.042HVSA-l (°)17.7±7.217.1±7.50.203PVSA-a (°)30.2±8.734.9±10.00.014PVSA-l (°)17.0±8.427.7±18.60.001Diameter of stent-graft 10 mm297 (90.8%)147 (87.5%)0.265 8mm36 (9.2%)21 (12.5%)0.134Length of stent-graft 6cm279 (85.3%)136 (81.0%)0.098 8cm48 (14.7%)32 (19.0%)0.148Target portion of PV Right branch248 (75.8%)90 (53.6%)0.021 Left branch42 (12.9%)67 (39.9%)0.002 Bifurcation36 (11.0%)10 (5.9%)0.017 Trunk1 (0.3%)1 (0.6%)0.428PSG decrease (mm Hg)15.7±5.216.5±6.20.266TIPS dysfunction20 (6.1%)47 (30.0%)<0.001 Variceal rebleeding17 (5.2%)36 (21.4%) Ascites recurrence1 (0.3%)0 (0) No symptom2 (0.6%)11 (8.6%)Hepatic encephalopathy103 (31.5%)48 (28.6%)0.403 One time41 (12.5%)20 (11.9%) Two times and above62 (19.0%)28 (16.7%)Mortality43 (13.1%)30 (17.9%)0.118 GI bleeding4 (1.2%)5 (3.0%) Other reason39 (11.9%)25 (14.9%) Open table in a new tab