Normal and altered three-dimensional portal venous hemodynamics in patients with liver cirrhosis.

Purpose: To compare time-resolved three-dimensional (3D) phase-contrast magnetic resonance (MR) imaging with three-directional velocity encoding (flow-sensitive four-dimensional [4D] MR imaging), with Doppler ultrasonography (US) as standard of reference, for investigating alterations in 3D portal venous hemodynamics in patients with liver cirrhosis compared with healthy age-matched control subjects and healthy young volunteers. Material & Methods: This prospective study was approved by the local ethics committee, and written informed consent was obtained from all participants. Three-dimensional portal venous hemodynamics was assessed, employing flow-sensitive 4D MR imaging with a 3-T MR system (spatial resolution, approximately 2 mm(3); temporal resolution, approximately 45 msec) in 20 patients with hepatic cirrhosis, 20 healthy age-matched control subjects, and 21 healthy young volunteers. Flow characteristics were analyzed by using 3D streamlines and time-resolved particle traces. Quantitative analyses were performed by retrospectively evaluating regional peak and mean velocities, flow volume, and vessel area. Doppler US was used as standard of reference. Independent-sample t tests or Wilcoxon-Mann-Whitney tests were applied for comparing each subject group. Paired-sample t tests or Wilcoxon tests were applied when comparing MR imaging and US. Results: Three-dimensional visualization of portal venous hemodynamics was successful, with complete visualization of the vessels in 18 patients and 35 volunteers, with limitations in the left intrahepatic branches (87%, reader A; 89%, reader B). A moderate but significant correlation was observed between 4D MR imaging and Doppler US in nearly all maximum and mean velocities, flow volumes, and vessel areas (r = 0.24-0.64, P = .001-.044). With MR imaging, significant underestimation was observed of intrahepatic flow velocities and flow volumes, except vessel area, which Doppler US represented as even lower (P < .001 to P = .045). Six patients had collateralization with reopened umbilical vein, while one had flow reversal in the superior mesenteric vein visible at MR imaging only. Conclusion: Flow-sensitive 4D MR imaging may constitute a promising, alternative technique to Doppler US for evaluating hemodynamics in the portal venous system of patients with liver cirrhosis and may be a means of assessing pathologic changes in flow characteristics.