RESUMO
PURPOSE: To determine the frequency and potential importance of findings initially interpreted as portal vein occlusion with "cavernous transformation" at transjugular intrahepatic portosystemic shunt (TIPS) placement with hepatic parenchymal CO(2) injection. MATERIALS AND METHODS: One hundred forty-seven patients underwent hepatic parenchymal CO(2) injections in the setting of fine-needle CO(2) TIPS procedures. Hepatic parenchymal CO(2) injections were retrospectively reviewed for findings suggestive of cavernous transformation in which direct portal venography confirmed portal vein patency. Direct portography was performed by injecting CO(2) via the fine needle/guide wire complex or a 5-F catheter in a branch of the intrahepatic portal vein. RESULTS: Hepatic lymphatic vessels mimicked cavernous transformation of the portal vein in 29 of the 147 patients (19.7%). One patient with portal vein occlusion showed profuse lymphatic filling without filling the extrahepatic portal vein. Hepatic parenchymal CO(2) injections safely permitted targeting of the portal vein in all patients. CONCLUSIONS: During fine-needle CO(2) TIPS procedures, what was believed to represent reversed flow of hilar collateral vessels represented hepatic lymphatic vessels. Despite the lack of visualization of the portal vein in nearly one-fifth of patients, targeting of the portal vein was facilitated by the fact the hepatic lymphatic vessels showed a periportal distribution. The high incidence of hepatic lymphatic filling suggestive of cavernous transformation necessitates direct portography or wedged hepatic venography to verify portal vein patency before the procedure is aborted.
