Abstract: Objective This study was conducted to evaluate predictive performance of thromboelastogram (TEG)parameters and blood coagulation indexes for portal vein thrombosis (PVT)in patients with liver cirrhosis (LC) after splenectomy. Methods 129 patients with hepatitis B-induced LC were recruited in our hospital between January 2020 and June 2024, and all underwent splenectomy. Baseline thromboelastography was performed for measurement of coagulation reaction time (R), blood coagulation time (K), solidification angle and maximal amplitude (MA). Portal vein diameter (PVD) and splenic vein diameter (SVD) were measured by endoscopic ultrasonography(EUS). Multivariate Logistic regression analysis was applied to assess risk factors, and area under receiver operating characteristic curve (AUC) was used to evaluate predicting efficacy. Results Of the 129 patients with LC, the incidence of PVT after splenectomy as confirmed by imaging was 29.5%; baseline platelet count in PVT group was 77.4(50.3, 101.2)×10⁹/L, much lower than [107.5(81.8, 159.7)×10⁹/L,P<0.05], while MELD score, PVD and SVD were 22.3(20.1, 24.8)points, 17.0(16.2, 17.8)mmand 14.0(13.2,15.4)mm, all much greater than [18.2(16.9, 20.0)points, 15.1(13.7, 16.4)mmand 11.9(10.9, 13.5)mm, respectively, P<0.05] in non-PVT group; prothrombin time (PT) and activated partial thromboplastin time (APTT) in PVT group were 16.5(15.1,18.0)s and 43.1(39.7, 45.9)s, both significantly longer than [15.1(14.0, 16.3)s and 40.3(37.4, 43.0)s, respectively, P<0.05], whilie serum fibrinogen (FIB) level was 2.0(1.6, 2.4)g/L, much lower than [2.3(1.9, 3.0)g/L, P<0.05] in non-PVT group, and K was 2.6(2.0, 3.4)min, much less than [3.7(1.8, 5.7)min,P<0.05], while MA was 52.7(43.7, 57.9)mm, much greater than [41.6(36.0, 53.1)mm, P<0.05] in non-PVT group; multivariate Logistic regression analysis showed that PVD, SVD, K and MA were all the independent risk factors for PVT occurrence in patients with LC after splenectomy (P<0.05);ROC analysis demonstrated thatthe AUC was 0.90, with sensitivity of 95.6% and specificity of 81.6%, when combination of PVD, SVD, K and MA in predicting PVT occurrence in LC patients after splenectomy, much superior to any parameter did alone (P<0.05). Conclusion Surveillance of TEG parameters, e.g., K and MA in patients with LC after splenectomy might help predict PVT occurrence, and an appropriate interventional measures should be given early and prevent it happening.

Key words: Liver cirrhosis, Splenectomy, Portal vein thrombosis, Thromboelastography, Blood coagulation time, Maximal amplitude, Diagnosis