19
Invasive Procedures in Patients with Liver Disease

Will Lester

University Hospitals Birmingham, Birmingham, UK

KEY POINTS

Many patients with liver disease have changes in pro‐ and anticoagulant factors and in platelet numbers and function, resulting in a rebalancing of hemostasis.
While the prothrombin time and platelet count have been historically used to determine the safety and management of invasive procedures, these parameters do not accurately reflect the risk of bleeding or the need for prophylactic transfusion of blood products.
For low‐risk procedures (such as therapeutic paracentesis or endoscopy), clinicians should not routinely correct thrombocytopenia and coagulopathy.
For higher‐risk procedures (such as liver biopsy), thresholds of hematocrit ≥25%, platelet count > 50 × 10⁹/l, and fibrinogen > 1.2 g/l may be used as guide to the risk of bleeding associated with invasive procedures pending further evidence.
The role of fresh frozen plasma (FFP) and of platelet transfusions is unclear; FFP may increase the risk of bleeding and complications from invasive procedures.
Thrombopoietin receptor agonists (TPO‐RA) are now licensed for preprocedure use in patients with thrombocytopenia associated with liver disease but may be associated with an increased risk of thrombosis.
Clinicians should seek expert advice before undertaking invasive procedures in patients with acute on chronic decompensated liver disease.

Introduction

Historically, it was assumed that the raised prothrombin time (the ratio commonly expressed as an international normalized ratio, INR) seen in liver disease reflected a bleeding risk and that correction with plasma products, specifically FFP, would reduce/correct this hemostatic failure. However, more recent studies have shown quite the reverse; there is often a prothrombotic state and transfusion of FFP is usually unnecessary, ineffective, and potentially hazardous. Published cohort studies of patients with liver disease undergoing invasive procedures demonstrate low bleeding risks.

Coagulation Testing in Patients with Liver Disease

In patients with chronic liver disease, the use of the INR to predict bleeding risk following invasive procedures in liver patients is not supported by clinical evidence. Although there is reduced synthesis/consumption of many procoagulant proteins and reduced numbers of platelets, there is an increase in factor VIII and von Willebrand factor, and a reduction in natural anticoagulants. Hemostasis in chronic stable liver disease is often referred to as “rebalanced,” with a net effect of normal hemostasis or even a procoagulant state. The INR is sensitive only for fibrinogen, FII, FV, FVII, and FX, and does not test the hemostatic balance in a patient with liver disease. Therefore, an INR of 2.0 in a patient with chronic liver disease is not comparable to the same result in a patient taking warfarin. There appears to be a similar rebalancing in acute liver disease.

The INR is, however, a surrogate marker of liver fibrosis and portal hypertension, which is the likely explanation for the weak association of a higher INR with an increased bleeding risk from liver biopsy. Consistent with this understanding that the INR is not representing a bleeding diathesis, preprocedural FFP, and/or platelet transfusions prior to transcutaneous liver biopsy, have not been shown to have a significant effect on hemorrhagic complication rates, and will just serve to increase intravascular volume and portal pressure, which is likely to be counterproductive.

More global assays of coagulation in patients with liver disease have been used to assess hemostasis; a few studies have examined their use to predict bleeding prior to invasive procedures. In one study of thrombin generation in patients with cirrhosis, FFP only slightly improved coagulation test values in a limited number of patients and even appeared to worsen them in one third of cases.

Viscoelastic testing (thromboelastography or rotational thromboelastometry) is generally normal, or at least near normal, in patients with chronic liver disease, including in the presence of a prolonged INR, and limited studies have used them to reduce blood product usage compared with transfusion triggers using standard assays of coagulation and platelets. However, there is still insufficient evidence to show that global tests of coagulation can be used routinely to predict bleeding prior to invasive procedures in patients with liver disease.