EASL Clinical Practice Guidelines

Management of anticoagulation in patients with liver disease



The American College of Chest Physicians Evidence-Based Clinical Practice Guidelines [207] for acute VTE [including deep vein thrombosis, pulmonary embolism and SVT (i.e., mesenteric, portal and hepatic vein thromboses)] in non-liver disease patients, recommend initial parenteral anticoagulation. LMWH or fondaparinux is suggested over intravenous or subcutaneous unfractioned heparin (UFH). All the above anticoagulants require binding to endogenous antithrombin in order to exert their anticoagulant action. Unfractionated heparin upon binding to antithrombin neutralizes both FXa and thrombin, whereas LMWH and fondaparinux neutralizes predominantly FXa [208]. Unfractionated heparin is mainly given as an intravenous injection and requires laboratory monitoring by the activated partial thromboplastin time (APTT). The UFH therapeutic interval should be 1.5–2.5 times over the normal APTT. However, this therapeutic interval is uncertain because it has not been confirmed by controlled trials and is also dependent on the reagent/instrument used for testing (not yet standardized). The 1.5–2.5 APTT therapeutic intervals corresponds to approximately 0.3–0.7 U/ml anti-Xa. However, like the APTT also the standardization of anti-Xa assays is far reaching, so that also the definition of the anti-Xa therapeutic interval is problematic. Because of the above limitations and other complications associated with its uses (i.e., the risk of HIT, osteoporosis, etc.), unfractionated heparin has been gradually replaced by LMWH, notwithstanding that it can be rapidly reversed by protamine-sulphate.

LMWH is given subcutaneously once daily or twice daily in a fixed dose for thromboprophylaxis and in a weight-adjusted dose for therapeutic purposes. Laboratory monitoring is not generally required, but it has been suggested in obesity, renal insufficiency or pregnancy [208]. Fondaparinux is given subcutaneously once daily in a fixed dose, without laboratory monitoring.


Liver disease patients

The major concern on the use of heparins in cirrhosis is the reduction of antithrombin that is a typical feature of patients with advanced disease. The crucial question is whether fixed- or weight-adjusted LMWH doses that are effective/safe in non-liver disease patients are equally effective/safe in cirrhotic patients. The direct consequence of this question is whether or not LMWH in cirrhosis requires laboratory monitoring to adjust the dosage. Experience is still insufficient and is limited to a few non-randomized studies (reviewed in [209]), and to in vitro studies using the plasmas from cirrhotic patients treated with LMWH [[210], [211], [212]]. Briefly, the non-randomized studies showed that LMWH when used at fixed prophylactic dose (4000 IU/day SC) and without laboratory monitoring was effective and safe in patients with PVT [209]. The in vitro studies gave conflicting results that can be summarized as follows. The anti-Xa assay is not the assay of choice to measure the LMWH anticoagulant effect [[210], [211]]. Conversely, thrombin generation assays seem more suitable for this task, but are not available in every haemostasis laboratory and need to be evaluated [212]. Finally, plasma from cirrhotic patients seem more responsive to the anticoagulation mediated by LMWH, despite the fact that they have low antithrombin levels [212]. The only randomized trial so far carried out showed that LMWH when used at a fixed prophylactic dose and without laboratory monitoring was effective and safe in preventing PVT in cirrhotic patients [196].

Vitamin K antagonists


VKA are the drugs of choice for prevention of stroke and systemic embolism in atrial fibrillation, patients with mechanical prosthetic heart-valves, and treatment/prevention of VTE after the first event in non-liver disease patients. VKA are coumarin-like drugs that interfere with the carboxylation of vitamin K-dependent coagulation factors, thus limiting their activity. Owing to the relatively narrow therapeutic window, VKA need strict laboratory monitoring to adjust the dose and maintain the patients within the therapeutic interval [213]. The test of choice is the prothrombin time (PT) with results expressed as INR. Briefly, the INR is the ratio (patient-to-normal) of PT raised to a power equal to the international sensitivity index (ISI) according to the following equation:

The ISI, being characteristic for each thromboplastin/coagulometer, represents the slope of the relationship of the PT determined with the standard and the working thromboplastin/coagulometer for plasmas from healthy subjects and patients stabilized on VKA [214]. By definition the INR scale is valid only for patients on VKA and no other superimposed coagulation defect [214], and any other use warrants validation and possible modification. VKA dose-adjustment is aimed at maintaining patients within the therapeutic interval corresponding to 2.0–3.0 INR (target 2.5).

Liver disease patients

The main concern on the use of VKA in cirrhosis is that the baseline-PT is often prolonged. This implies that the attainment of the therapeutic interval requires probably smaller VKA doses, thus cirrhotic patients could be under-dosed. No studies whatsoever have yet to address this issue and cirrhotic patients are currently treated with VKA doses aimed at 2.0–3.0 INR. The second concern on VKA in cirrhosis is the use of the INR as the scale to express PT results. As demonstrated by independent groups (reviewed in [215]) the regular INR (here called INR-vka) is not valid for cirrhotic patients and hence cannot minimize the variability of the INR obtained in laboratories using different thromboplastins. As a consequence the INR-vka obtained in any given laboratory may or may not be representative of the real anticoagulation achieved with the specific dose. An alternative to this scale would be the modified-INR, valid for cirrhosis (called INR-liver) that has been developed [215], but not yet investigated for its value in assessing survival in patients with cirrhosis nor for monitoring patients with cirrhosis on VKA.

Direct oral anticoagulants (DOAC)

These are drugs that (unlike VKA) target directly specific activated factors such as thrombin (dabigatran) or FXa (rivaroxaban and apixaban) without intermediation by antithrombin or carboxylation. DOAC have been licensed for atrial fibrillation and treatment/prophylaxis of VTE (reviewed in [216]). Cirrhotic patients have been deliberately excluded from phase III trials and therefore (although they might have some theoretical advantages over heparins or VKAs) [182] no information is currently available in this setting. The main advantage of DOAC is that they do not require dose-adjustment by laboratory tests, thus the issue on the validity of the INR in this setting could be eliminated. Recently, it has been reported that treatment with rivaroxaban can be associated with severe symptomatic liver injury [217]. Caution should be exerted until this issue is addressed in clinical trials.