Extrahepatic portal vein stenosis in recipients of living-donor allografts: Doppler sonography.

Abstract

The aim of this study was to describe the appearances obtained and the pitfalls involved with the use of Doppler sonography for detecting portal vein stenoses after surgery in 198 recipients of pediatric reduced-size transplants. We analyzed sonographic and Doppler studies after surgery for 167 children (average, 2.5 years old) who were recipients of 198 left lobe or left lateral segment liver segments (79 living-donor allografts and 119 cadaveric grafts). Sonographic and Doppler studies were performed either on the basis of clinical evidence of portal hypertension or as part of a screening protocol. Demographic and surgical data were compared with the incidence of portal vein structure. We calculated pressure gradients from Doppler jet velocities and compared them with gradients measured manometrically from direct portography in 12 patients. Imaging criteria that indicated portal vein stenoses were (1) a visualized portal vein diameter of 2.5 mm or less, (2) an acceleration of flow at the stricture or a poststenotic jet of portal vein flow revealed by Doppler imaging, or (3) both. Stenoses meeting these criteria were verified by surgical or angiographic identification. Seventeen (22%) portal vein stenoses were detected in recipients of the 79 living-donor liver transplants, whereas three (3%) were detected in recipients of the 119 cadaveric grafts (p < .005). The use of cryopreserved venous extension grafts was the most significant parameter of correlation (p < .025). Doppler sonography predicted the stenoses in all cases, although it overestimated the pressure gradients in all but one of the verified cases. Intrahepatic portal vein flow was frequently normal in the presence of significant extrahepatic portal vein stenosis. Diagnosis of portal vein stenosis in recipients of living-donor allografts requires real-time visualization of the entire length of the portal vein, combined with spectral and color Doppler investigations of the portal and splenic veins and a search for collateral vessels. Visualization of each component alone may be insufficient. In our study, when care was taken to follow this procedure, sonography accurately showed all angiographically verified portal vein stenoses, although pressure gradients frequently were inaccurate. A protocol for periodic follow-up with real-time and Doppler sonography is crucial for pediatric patients to permit early identification of portal vein stenoses.