Assessment of the predictive value of non invasive markers for prediction of portal vein thrombosis in cirrhotic patients

Abstract

Portal vein occlusion by thrombosis typically occurs in cirrhotic patients. The aim of this study is to assessment of the predictive value of non invasive markers for prediction of portal vein thrombosis in cirrhotic patients. Materials and Methods: This study was cross-sectional study and the target population included all cirrhotic patients admitted to the gastrointestinal ward (or with a history of hospitalization in this ward) and all cirrhotic patients referred to the gastrointestinal clinic of Imam Reza Hospital in 2020. The sample size in this study was 126 people. The sampling method in this study was available. The course of liver function tests was recorded from the time of diagnosis of cirrhosis to the development of portal vein thrombosis. Finally, subjects were divided into two groups: cirrhotic subjects with portal vein thrombosis and cirrhotic subjects without portal vein thrombosis. These non-invasive markers obtained from patients' blood samples (INR, PT, PTT, Platelet, and Albumin) were recorded and then compared and analyzed between cirrhotic patients with portal vein thrombosis and cirrhotic patients without portal vein thrombosis to determine the predictive value of portal vein thrombosis. Results: In this study, the mean (standard deviation) of age in the two groups of cirrhotic patients with portal vein thrombosis and cirrhotic patients without portal vein thrombosis were 58.82 (±15.8) and 58.13 (±12.3) years, respectively. In this study, among the variables of age and sex and ultrasound result, only recanalization in the group of cirrhotic patients with portal vein thrombosis was significantly higher than the group of cirrhotic patients without portal vein thrombosis (P-value <0.001). Also, among blood factors, the variables of PLT (P-value = 0.003) and INR (P-value = 0.035) in the group of cirrhotic patients without portal vein thrombosis were significantly higher than the group of cirrhotic patients with portal vein thrombosis; On the other hand, the variable of Albumin in the group of cirrhotic patients with portal vein thrombosis was significantly higher than the group of cirrhotic patients without portal vein thrombosis (P-value = 0.001). Among the blood factors in predicting the formation of portal vein thrombosis in cirrhotic patients, the variables of Albumin (P-value = 0.001), PT (P-value = 0.001) and INR (P-value = 0.023) significantly had the highest predictive power of portal vein thrombosis, respectively. The optimal cut-off point of PT variable in predicting the formation of portal vein thrombosis in cirrhotic patients was 17.35, which sensitivity and specificity at this cut-off point was 68% and 67%, respectively, and the Yuden index was 0.35. Also, the optimal cut-off point of INR variable in predicting the formation of portal vein thrombosis in cirrhotic patients was 1.52, which sensitivity and specificity at this cut-off point was 62% and 56%, respectively, and the Yuden index was 0.18. The optimal cut-off point of Albumin variable in predicting the formation of portal vein thrombosis in cirrhotic patients was 3.45, which sensitivity and specificity at this cut-off point was 84% and 54%, respectively, and the Yuden index was 0.38.