It turns out to be a little of both. A number of reviews have addressed this. [1] [2] [3] Here are some of the key points:
The old maxim that cirrhotic patients are auto-anticoagulated is a myth. Severe liver disease is associated with a delicate balance between bleeding and clotting. In general, cirrhosis tends to be a hypercoagulable state. The relative risk for VTE in such patients has been estimated at around 2.
What are the mechanisms for the hypercoagulability of liver disease?
In primary hemostasis, while platelet numbers are often decreased these platelets tend to be hyper-functional. This is due to low levels of ADAM TS 13, correspondingly increased levels of VW factor and multifactorial endothelial dysfunction.
In secondary hemostasis, reasons for hypercoagulability include decreased levels of liver dependent natural anticoagulants such as protein C, protein S, and antithrombin. Factor VIII (not synthesized in the liver) tends to be increased.
There is also impaired fibrinolysis with increased levels of PAI-1 and decreased levels of plasminogen.
What are the clinical implications?
Traditional hemostatic tests are generally used but are of limited reliability. There's been increasing interest in global hemostatic tests such as viscoelastic assays which are conceptually more valid but are not yet ready for translation into clinical practice.
For low risk procedures, prophylactic hemostatic products peri-procedure are generally not indicated.
Antithrombotic treatments should be given in accordance with standard clinical indications recognizing a potentially increased risk of bleeding.
DOACs can be used in many patients but certain published restrictions apply.
