Indirect Doppler ultrasound abnormalities of significant portal vein stenosis after liver transplantation.

Abstract

To determine indirect Doppler ultrasound (DUS) abnormalities associated with significant portal vein (PV) stenosis (PVS) in recipients of liver transplantation (LT). This retrospective study was approved by our institutional review board. Between February 2006 and May 2017, 41 LT recipients were diagnosed with significant PVS, defined as having more than 50% narrowing of PV diameter for any reason, including thrombosis or flow disturbance associated with prominent collateral vessels on portal venography. We reviewed the DUS findings of hepatic arteries (HAs) as well as PVs of them, before and after treatment of PVS, and in comparison, with a one-to-one case-matched control. Inter-group comparison of frequency in DUS abnormalities was performed using Chi square (χ2)with Fisher's exact test and McNemar's test. Diagnostic values of each abnormal DUS finding and combinations were also evaluated. DUS of significant PVS showed "no demonstrable color flow," either at recipient PVs or anastomoses (26.7%), and showed turbulence (66.7%) and hepatofugal portal flow (HFPF; 20.0%) at the graft PVs. HFPF was more frequently observed in those with "no demonstrable color flow" at recipient PVs or anastomoses (p = 0.006). DUS of graft HAs revealed tardus-parvus waveforms (20.9%) and prolonged systolic acceleration times (16.3%), more commonly in the "no demonstrable color flow" group (p = 0.012). These indirect DUS abnormalities disappeared and resolved on follow-up DUS after treatment. In the control group, such Doppler abnormalities were less frequently shown than in the PVS group (p ≤ 0.01, respectively). When one of the portal-blood flow velocity (PFV)-related index abnormalities (such as increased time average velocity [TAV] at anastomosis and TAV ratio between recipient PV and anastomosis) or "no demonstrable color flow" were shown in DUS as well as one of the indirect DUS abnormalities, sensitivity, and specificity was 71.11 and 97.78%, respectively. In addition to PFV-related abnormalities, DUS occasionally shows "no demonstrable color flow" either at recipient PVs or anastomoses, and indirect Doppler abnormalities such as turbulence, HFPF at graft PVs, and abnormal waveforms at graft HAs in LT recipients with significant PVS. The combination of PFV-related abnormalities and indirect DUS abnormalities would be helpful for diagnosis of PVS.