Triple-phase abdominal computed tomography for detecting spontaneous portopulmonary shunts in cirrhotic patients

Background and Aim: Data on prevalence of portopulmonary shunts (PPS) are quite limited. Most studies have used cineportography or echocardiography for diagnosis. Only few recent case reports have reported the use of computed tomography (CT) for identifi- cation of PPS. This study tried to determine the prevalence of PPS in patients with cirrhosis using contrast-enhanced CT of the abdomen, and to determine their association with demographic and clinical characteristics. Methods: A total of 150 subjects with cirrhosis who had previously undergone triple- phase CT were analyzed. PPS was diagnosed when at least one esophageal varix met all of the following criteria: (i) it could be followed cephalad into the chest to the level of the inferior pulmonary vein or left atrium; (ii) it abutted the wall of either of these structures; (iii) it had luminal continuity with one of these structures; and (iv) it was no longer seen one slice above the level of contact. Results: Of 150 subjects, 18 were excluded for incomplete data. The prevalence of PPS was found to be 26/132 (19.7%). Of these, 14 (53.8%) patients had PPS draining into the left atrium and 12 (46.2%) had those draining into one of the pulmonary veins. Presence of PPS was associated with the presence of varices at endoscopy, ascites, thrombocytope- nia and splenomegaly. Conclusion: In our study, the largest study on PPS to date, the prevalence of PPS in cirrhotic patients using triple phase CT was found to be 19.7%. CT may be a useful technique to study PPS and their clinical implications.