Comparing the predictive ability of portoatrial and portocaval gradient after transjugular intrahepatic portosystemic shunt creation for variceal rebleeding.

Abstract

Measuring the portal pressure gradient from the portal vein (PV) to the inferior vena cava (IVC) or to the right atrium (RA) remains controversial. The aim of our study was to compare the predictive ability of portoatrial gradient (PAG) and portocaval gradient (PCG) for variceal rebleeding. The data of 285 cirrhotic patients with variceal bleeding undergoing elective transjugular intrahepatic portosystemic shunt (TIPS) in our hospital were analyzed retrospectively. The variceal rebleeding rates were compared between groups categorized by established or modified thresholds. The median follow-up time was 30.0months. After TIPS, PAG was equal to (n = 115) or more than (n = 170) PCG. The pressure of IVC was defined as an independent predictor for a PAG-PCG difference of ≥ 2mmHg (p < 0.001, OR 1.23, 95% CI 1.10-1.37). Using a threshold of 12mmHg, PAG (p = 0.081, HR 0.63, 95% CI 0.37-1.06) could not predict variceal rebleeding but PCG could (p = 0.003, HR 0.45, 95% CI 0.26-0.77). This pattern was unchanged when a ≥ 50% reduction from baseline was also considered as a threshold (PAG/PCG: p = 0.114 and 0.001). Subgroup analyses showed that only in patients with post-TIPS IVC pressure < 9mmHg (p = 0.018), PAG could predict variceal rebleeding. Because PAG was on average 1.4mmHg higher than PCG, patients were classified by a PAG of 14mmHg, and there was no difference in rebleeding rates between these two groups (p = 0.574). For patients with variceal bleeding, the predictive ability of PAG is limited. The portal pressure gradient should be measured between the PV and IVC.