Causes for the absence of thrombocytopenia in patients with liver cirrhosis and portal vein thrombosis: A case-control study

Abstract

Background: Complications of liver cirrhosis (LC), such as thrombocytopenia and portal vein thrombosis (PVT), have similar pathophysiology. However, the association between PVT and platelet count in LC patients is contradictory. Aim: To assess factors affecting the platelet count in patients with LC and PVT. Materials and methods: This was a retrospective case-control study. The cases were 114 patients with LC of various etiologies and newly diagnosed PVT unrelated to invasive hepatocellular carcinoma. From the database of LC patients without PVT, 228 controls were randomly selected with stratification by gender, age and etiology of cirrhosis. The patients from both groups were divided into subgroups with thrombocytopenia ( 150 × 109/L) and without thrombocytopenia (≥ 150 × 109/L). We analyzed the LC etiology, portal hypertension severity (ascites, hepatic encephalopathy, gastroesophageal varices and associated bleedings, the spleen length, and portal vein diameter), laboratory parameters (white blood cell counts, neutrophils, lymphocytes, hemoglobin levels, total protein, albumin, total bilirubin, fibrinogen, neutrophil-to-lymphocyte ratio, and prothrombin); also, the rates of newly diagnosed malignant tumors was assessed. The statistical analysis included calculation of odds ratios (OR) and 95% confidence intervals (CI), logistic regression models with assessment of the model accuracy, and the area under the ROC curve (AUC). Results: There were no differences in the severity of thrombocytopenia between the case and control groups: thrombocytopenia was severe in 15.8% (18 patients) vs 13.6% (31 patients, p = 0.586); moderate, in 41.2% (47 patients) vs 46.1% (105 patients, p = 0.398) and mild, in 31.6% (36 patients) vs 24.5% (56 patients, p = 0.168). The proportion of the patients without thrombocytopenia was 11.4% (13 patients) in the case group and 15.8% (36 patients) in the control group, with the between-group difference being non-significant (p = 0.276). In the subgroups of patients without thrombocytopenia (both in the cases and in the controls), the proportion alcoholic etiology of LC, white blood cells counts, neutrophils, lymphocytes, and fibrinogen concentrations were significantly higher (p 0.05) than in those with thrombocytopenia. The model based on the outcome "absence of thrombocytopenia" included white blood cells counts, hemoglobin and albumin levels, the presence of newly diagnosed malignant tumors in the case group (model accuracy 90.4%, AUC 0.873), and neutrophil counts and spleen length in the control group (model accuracy 86.4%, AUC 0.855). In the patients with PVT and platelet counts of ≥ 150 × 109/L, the OR for all newly diagnosed malignant tumors was 26.3 (95% CI 7.37–93.97, р 0.0001), for newly diagnosed hepatocellular carcinoma without portal vein invasion 17.42 (95% CI 4.84–62.65, р 0.0001). Conclusion: In LC patients, the prevalence and severity of thrombocytopenia are not different depending on the PVT presence or absence. The absence of thrombocytopenia in PVT patients is associated with a higher risk of malignant tumors identification, primarily that of hepatocellular carcinoma.