Portal blood flow in chronic liver disease: measurement of portal blood flow using 2D-cine phase contrast magnetic resonance angiography

Abstract

Purpose: The evaluation of portal venous hemodynamics may help determine disease progression and prognosis in patients with chronic liver disease. Although pulsed Doppler ultrasound is used to measure portal venous flow, accuracy of this method is dependent on the proficiency of the operator, the obesity of the subject and the presence of gastrointestinal gas. We used two-dimensional cine phase contrast MR angiography (2D-CPC MRA), a technique not affected by the factors described above, to measure portal venous blood flow (PBF) in patients with chronic liver disease to test the clinical value of the measurement.
Material and Methods: This study consisted of 43 subjects: 35 patients with chronic liver disease (13 with chronic hepatitis, 22 with cirrhosis) and 8 healthy volunteers. A 1.5-T MR imaging unit was used. The cross-section, plethysmography-gated, multiphase images of 2D-CPC MRA were obtained from the center of the portal trunk. PBF was calculated from this area and the mean velocity of portal vein at each phase. Boundaries of the liver were traced manually on T_1- or T_2-weighted transverse images covering the whole liver to calculate the volume of the liver. PBF was divided by liver volume (PBF/LV) to obtain volume-corrected index. Doppler US of the abdomen was performed on the same day as 2D-CPC MRA. The indocyanine green retention at 15 minutes post-injection (ICG R₁₅％) was also measured when the subjects’condition permitted.
Results: PBF was significantly reduced in patients with cirrhosis compared with healthy volunteers (mean± SD; 511.0± 204.6 and 889.8± 289.5 ml/min, respectively).
 PBF/LV correlated positively with hepatic volume in healthy volunteers (r＝0.8, P<0.05) and in patients with chronic hepatitis (r＝0.20, P<0.05), but correlated negatively in patients with cirrhosis (r＝−0.60, P<0.001). PBF by 2D-CPC MRA correlated with that by Doppler US measurements (r＝0.5, P<0.05). PBF by 2D-CPC MRA weakly, but significantly correlated with ICG R₁₅％(r＝−0.16, P<0.05).
Conclusion: PBF was reduced in patients with cirrhosis in this study. PBF/LV correlated negatively with the liver’s volume in patients with cirrhosis, but positively correlated in healthy volunteers and in patients with chronic hepatitis. These results show that PBF and PBF/LV by 2D-CPC MRA may be used as an index for disease progression.