Short- and long-term hemodynamic effects of transjugular intrahepatic portosystemic shunts: a Doppler/manometric correlative study.

Abstract

The purposes of this study were to evaluate the effect of a well-functioning transjugular intrahepatic portosystemic shunt (TIPS) on the splanchnic and intrahepatic circulation, to determine if sonographic measurements can predict shunt dysfunction before clinical manifestations of portal hypertension occur, and to compare Doppler sonographic findings with portocaval gradient measurements before and after shunt revision. Forty-four patients with cirrhosis (n = 43) and myelofibrosis (n = 1) who underwent successful TIPS insertion were included in this prospective study. Indications for TIPS placement were: refractory ascites (24 patients), bleeding esophageal varices (17 patients), portal hypertensive gastropathy (two patients), and bleeding colonic varices (one patient). The portal vein and the inferior vena cava were catheterized; and the portocaval gradient was recorded before TIPS placement, at 2 and 12 months after TIPS placement, and when clinical or Doppler findings suggested shunt dysfunction. Doppler studies were done within 1 week before TIPS placement, within 2 days after TIPS placement, every 2-3 months thereafter, and before and after a TIPS revision. The Doppler studies included flow volume measurements in the portal vein and in the stent, as well as determination of the direction of flow in the segmental branches of the portal vein, in the splanchnic veins, and in portosystemic collaterals. Changes in Doppler findings and in catheter pressure measurements were compared using Spearman's rank correlation test. Significance was set at the .05 level. A marked decrease (-51%) in portocaval gradient was observed after TIPS placement. At Doppler sonography, portal vein velocity and diameter were both higher after TIPS placement, resulting in a marked increase in portal venous flow (170%). Mean flow velocity in the shunt was 55.8 +/- 3.6 cm/sec, and flow volumes in the shunt and in the main portal vein were 1596 ml/min and 1731 ml/min, respectively (p = nonsignificant). Dysfunction of the stent occurred in 27% of the patients. Changes in stent blood flow volume were closely related to changes in the portocaval gradient (r = -0.67, p < .001). Reduction of blood flow volume in the stent or change of direction of flow in intrahepatic portal veins or in collateral veins signaled shunt dysfunction (84% sensitivity, 89% specificity). Marked hemodynamic changes in the portal venous system occur soon after a TIPS procedure. Monitoring of shunt function with periodic Doppler sonography, including calculation of shunt blood flow, is useful in detecting shunt dysfunction before clinical signs occur.