Decreased portal vein velocity is predictive of the development of portal vein thrombosis: A matched case‐control study

• Published in: Liver international Vol. 38; no. 1; pp. 94 - 101
• Main Authors: Stine, Jonathan G., Wang, Jennifer, Shah, Puja M., Argo, Curtis K., Intagliata, Nicolas, Uflacker, Andre, Caldwell, Stephen H., Northup, Patrick G.
• Format: Journal Article
• Language: English
• Published: United States Wiley Subscription Services, Inc 01.01.2018
• Subjects: Adult; Aged; Blood Flow Velocity; Case-Control Studies; Cirrhosis; coagulopathy; Computed Tomography Angiography; Databases, Factual; Disease control; Female; Hazards; Health risk assessment; Humans; Liver; Liver Circulation; Liver cirrhosis; Liver Cirrhosis - complications; Liver Cirrhosis - diagnostic imaging; Liver Cirrhosis - physiopathology; Liver diseases; Magnetic Resonance Imaging; Male; Middle Aged; Patients; Phlebography - methods; Portal vein; Portal Vein - diagnostic imaging; Portal Vein - physiopathology; Predictive control; prohaemostasis; Risk Factors; Risk groups; Statistical models; Thromboembolism; Thrombosis; Ultrasonography, Doppler, Color; Velocity; Venous Thrombosis - diagnostic imaging; Venous Thrombosis - etiology; Venous Thrombosis - physiopathology; thrombosis; coagulopathy; prohaemostasis; cirrhosis
• ISSN: 1478-3223; 1478-3231; 1478-3231
• DOI: 10.1111/liv.13500

Background & Aims Portal vein thrombosis (PVT) in cirrhosis may lead to hepatic decompensation and increased mortality. We aimed to investigate if decreased portal vein (PV) velocity is associated with future PVT. Methods Data on adult patients with cirrhosis and PVT between January 1, 2005 and July 30, 2015 were obtained. Cases with PVT were matched by age, gender and Model for End‐stage Liver Disease (MELD) score to corresponding controls without PVT. Cox proportional hazards models, receiver operator curves and Kaplan Meier curves were constructed. Results One hundred subjects (50 matched pairs) with mean age 53.8±13.1 y and MELD score 14.9±5.5 were included in our analysis. Sixty‐four percent were male and 76% were Child‐Turcotte‐Pugh Class A or B. Baseline characteristics (prior to development of PVT) were similar, except for baseline PV velocity (16.9 cm/s, 95% CI 13.9‐20.0 PVT vs 25.0, 95% CI 21.8‐28.8 no PVT, P<.001). 30 PVT subjects had PV velocity <15 cm/s compared to five without PVT (P<.001). On adjusted multivariable analysis, PV velocity was the strongest independent risk factor predicting PVT development (HR 0.86, 95% CI 0.80‐0.93). The predictive value for PVT development was greatest for flow <15 cm/s (c‐statistic 0.77). PV velocity <15 cm/s had a highly significant association with future PVT (HR 6.00, 95% CI 2.20‐16.40, P=<.001). Conclusions Decreased PV velocity is associated with increased risk of future PVT. Patients with cirrhosis and decreased PV velocity are a high‐risk subgroup that warrants further investigation with prospective study.