Portal vein thrombosis associates with high platelet-fibrin clot strength and platelet activation in decompensated cirrhosis: A retrospective study

Abstract

Background and aims

Alteration of platelet status associates with decompensation and death in cirrhosis, while its effect on portal vein thrombosis (PVT) remains unclear. We aimed to retrospectively investigate whether PVT associates with platelet-fibrin clot strength and platelet activation in decompensated cirrhosis.

Methods

Platelet-fibrin clot strength (G) was measured by thromboelastography (TEG). Platelet activation was reflected by plasma concentrations of soluble p-selectin (sPs) and a platelet aggregation test adjusted for platelet counts.

Results

Among 166 patients, 45 had PVT. The platelet count was significantly lower in PVT. While the G value was positively correlated with platelet count (ρ = 0.74, P < 0.01), increased G was associated with PVT after adjusting for platelet count in the logistic regression (P = 0.04). The normalized G value according to the linear relation with platelet count was calculated as follows: G_platelet = [(G - 2622)/platelet count]. This coefficient had no correlation with platelet count and was an independent risk factor of PVT (OR = 1.03, CI_95%: 1.01-1.05, P = 0.012). In two subanalyses, the collagen-induced platelet aggregation (n = 37, P = 0.029) and plasma concentration of sPs (n = 56, P = 0.001) adjusted for platelet count were significantly higher in PVT.

Conclusion

This study showed a positive correlation of high platelet-fibrin clot strength detected via TEG and platelet activation with PVT in decompensated cirrhosis.

Introduction

In the past decade, the concept that cirrhosis is a prohemorrhagic condition has undergone profound changes. The coagulation state in cirrhosis is rebalanced to be relatively hypercoagulable, as evidenced by the thrombin generation test and increased risk of portal vein thrombosis (PVT) [1]. The presence of PVT was found to correlate with variceal bleeding, worsening ascites and increased mortality [2], [3], [4]. Moreover, it will markedly increase the difficulties of surgeries for patients who need transjugular intrahepatic portosystemic shunt (TIPS) or liver transplantation [5].

The risk factors for PVT in cirrhosis are thus far not well defined. Studies have shown that decreased portal venous blood flow velocity [6] and large volume in the porto-collateral vessel [3] are significant predictive factors. Previous endoscopic therapy and abdominal surgery have been revealed as determinants of PVT due to vascular damage to the portal vein [2]. Decreased anticoagulant synthesis and increased procoagulant secretion during hemostatic rebalancing processes were supposed to favor PVT formation. Nevertheless, recent studies adjusted these single coagulation factors for the severity of liver disease and did not support their role in PVT formation [7], [8]. Thus, the association between hypercoagulability and PVT remains controversial.

Thromboelastography (TEG) is a more global evaluation of hemostasis that measures the viscoelastic changes of thrombus formation, reflecting the interaction between plasma, platelets, and other blood cells [9]. In recent years, TEG has been used among cirrhosis patients for guiding blood component use as well as assessing the severity of disease and risk of procedure-related bleeding [10], [11], [12]. It was shown to better reflect hemostatic abnormalities than traditional coagulation tests [13]. However, to date, the data are lacking to determine the association between the TEG profile and PVT, and only existing studies have conflicting conclusions [14], [15], [16]. An important reason is that a low platelet count is an established risk factor for PVT in cirrhosis [7], [17] and can dramatically affect the maximum amplitude (MA) value, one of the parameters of TEG reflecting the maximum clot strength. The actual platelet-fibrin clot strength, G value, is calculated by [(5000 × MA)/(100 – MA)], of which the platelet count and function contribute to 80%(18, 19). Therefore, taking platelet count into consideration is necessary when comparing the platelet-fibrin clot strength between PVT and non-PVT patients.

On the other hand, recent studies have demonstrated that increased platelet activation revealed by plasma biomarkers and platelet aggregation (PA) tests adjusted for platelet count in cirrhosis were related to further decompensation and liver-related death [20], [21], [22], [23]. However, other studies found that acute kidney injury (AKI) and bacterial infections in decompensated cirrhosis were associated with platelet dysfunction [[24], [25]]. These findings contribute to a deeper understanding of the alteration of platelet status in cirrhosis and provide new thoughts in exploring the association between coagulation and PVT.

With these findings in mind, we set out to perform a retrospective study in a cohort of hospitalized decompensated cirrhosis patients. The primary objective was to evaluate the association between platelet-fibrin clot strength and PVT. The influence of different severities of thrombocytopenia on the TEG results was overcome by normalizing to the platelet count. For the secondary objective, we performed subanalyses to determine whether increased platelet activation was related to PVT.