EASL Clinical Practice Guidelines on prevention and management of bleeding and thrombosis in patients with cirrhosis

Introduction

Patients with cirrhosis frequently acquire substantial alterations in their haemostatic system that are apparent during screening with basic tests of haemostasis (mainly international normalised ratio [INR], activated partial thromboplastin time [APTT] and platelet count). Historically, these changes were thought to induce a haemostasis-related bleeding tendency.
Nowadays, however, it is well accepted that basic haemostasis tests, such as prothrombin time and APTT, do not truly represent the haemostatic system operating in patients with liver disease who remain in haemostatic balance, as both pro- and anti-haemostatic systems change simultaneously.
 
Furthermore, it is also acknowledged that although patients with liver disease may experience bleeding complications, many of these bleeds are unrelated to haemostatic failure but are a consequence of portal hypertension or mechanical vessel injury, which could be caused by inadvertent vessel puncture during invasive procedures, for example. It has also been established that patients with cirrhosis are not protected from the occurrence of thrombosis and may require anticoagulant therapy for prevention or treatment of thrombotic episodes. The occurrence of both bleeding and thrombosis may pose difficult clinical questions, which will be addressed in these Clinical Practice Guidelines (CPGs).
The absence of high-quality clinical studies represents an additional challenge in constructing guidance. We have therefore attempted to base our recommendations not only on available clinical literature, but also on biochemical studies, so that our recommendations, when insufficiently supported by clinical data, have a mechanistic basis.

Clinical studies on bleeding and thrombosis in patients with liver disease have been of poor to moderate quality for reasons outlined elsewhere.

 

 As an example, the question of how to manage haemostatic abnormalities prior to invasive procedures has not been addressed by adequately powered randomised studies. Ideally, a pragmatic study addressing this issue should include an arm where blood product transfusions are administered on the basis of pre-selected coagulation test cut-offs and a comparison arm where no prophylactic blood product transfusion is mandated, and only rescue treatment is allowed. However, although the results of such a study would provide a definite answer to this long-standing question, it is quite unlikely that it will ever be performed; taking into account the generally low incidence of bleeding (e.g., approximately <1.5%) following the most common procedures, it would require the enrolment of a cohort of 1,531 patients per arm in order to detect a statistically significant increase (i.e., from 1.5% to 3.0%, at least 100% of the baseline risk) in peri-procedural bleeding rate. Thus, although the low level of evidence for many of our recommendations should encourage the hepatology community to organise better quality studies, we should accept that many of these studies will not be performed in the foreseeable future. Nevertheless, we are convinced that there is reason to change some widely accepted treatment dogmas as we have outlined in this document.

 
Multiple recommendations in this document relate to interventions the panel responsible for drafting this document feel are not useful, even though they are currently widely applied in clinical practice. Even though there is no definitive evidence from well-designed clinical studies for many of the statements, the combination of potential harm, cost, and biochemical support for a lack of effect has led to these recommendations that are supported by both the writing panel and the Delphi panel (details provided in the next section). An example of such a recommendation is the advice not to attempt to correct prolonged prothrombin time in a patient with cirrhosis prior to a procedure, as there is evidence that a prolonged prothrombin time does not predict bleeding, likely because patients are in a ‘rebalanced’ haemostatic state. Furthermore, there is no clinical evidence that fresh frozen plasma (FFP) infusion reduces bleeding risk, and there is evidence for a lack of a biochemical response to FFP infusion in this setting. The writing panel proposes that all statements are valid for any patient with cirrhosis, including those who are critically ill. However, there are limited data on bleeding and thrombosis risk in critically ill patients with cirrhosis, and very little is known about the efficacy of pro- and anti-haemostatic interventions in these patients. Therefore, we propose that deviations from our recommendations may be justified in individual cases. For example, it may be justified to provide pro-haemostatic therapy in a patient with extreme changes in the haemostatic system when there is particular concern regarding bleeding among the treating physicians.
 

Spontaneous bleeding prevention

In patients with cirrhosis and abnormal laboratory tests (INR, APTT, platelet count, fibrinogen), is correction of these tests by blood products or factor concentrates indicated to prevent spontaneous bleeding?

Recommendation

  1. In patients with cirrhosis and abnormal laboratory tests (INR, APTT, platelet count, fibrinogen), attempting to correct these tests by administering blood products or factor concentrates, with the aim of preventing spontaneous bleeding, is not recommended (LoE 3, strong recommendation).

Future perspectives

  1. Large observational studies aiming to define the precise incidence of spontaneous bleeding events and their impact on clinical course and survival of patients with cirrhosis are recommended.

Procedural bleeding prevention

Do traditional haemostasis tests (INR, APTT, platelet count, fibrinogen), or viscoelastic tests, predict bleeding in patients with cirrhosis undergoing invasive procedures at low or high risk of bleeding?

Statement

  1. INR and APTT do not predict post-procedural bleeding in patients with cirrhosis undergoing invasive procedures (LoE 3).
  2. Studies do not consistently demonstrate a link between thrombocytopenia, hypofibrinogenaemia, or viscoelastic test results and the risk of post-procedural bleeding, although there may be subgroups in whom thrombocytopenia is related to procedural bleeding risk and there is initial evidence suggesting that viscoelastic tests might help to address this issue (LoE 4).

Recommendation

  1. In patients with cirrhosis, the use of traditional haemostasis tests, or viscoelastic tests, cannot be generally indicated to predict procedural bleeding risk, although they can be used to assess severity of disease or haemostatic status and to provide an initial benchmark to guide management in the case of post-procedural bleeding (LoE 3, strong recommendation).

Future perspectives

  1. On the basis of initial evidence provided by the results of viscoelastic tests, their potential to predict post-procedural bleeding should be further explored in prospective, adequately powered studies including different categories of patients with cirrhosis (compensated, decompensated, acute-on-chronic liver failure) undergoing high-risk procedures.
In patients with cirrhosis undergoing invasive procedures with a low risk of clinically relevant bleeding, is laboratory evaluation of haemostasis indicated to predict procedure-related bleeding?
Recommendation
  1. In patients with cirrhosis undergoing invasive procedures with a low risk of bleeding, laboratory evaluation of haemostasis with the aim of predicting post-procedural bleeding is not indicated (LoE 4, strong recommendation).
In patients with cirrhosis undergoing invasive procedures with a high risk of clinically relevant bleeding, is laboratory evaluation of haemostasis indicated to predict procedure-related bleeding?
Statements
  1. There is weak evidence that the measurement of platelet count might be indicated to identify patients at increased procedural bleeding risk. No solid data are available for fibrinogen (LoE 4).
  2. As evidence supporting viscoelastic tests as predictors of procedure-related bleeding in patients with acute decompensation of cirrhosis, with or without organ failure, is weak, it is not possible to advise for or against their use (LoE 4).
Recommendation
  1. In patients with cirrhosis undergoing invasive procedures associated with a high risk of bleeding, laboratory evaluation of haemostasis is generally not indicated to predict post-procedural bleeding, although it may serve to provide a baseline status of the patient and to assist the physician in the case of bleeding events (LoE 4/5, weak recommendation).
In patients with cirrhosis undergoing invasive procedures, does correction of a prolonged INR by infusion of FFP decrease the rate of procedure-related clinical relevant bleeding?
 when the term “invasive procedure” is used, it means both low- and high-risk procedures
Recommendation
  1. In patients with cirrhosis undergoing invasive procedures, correction of a prolonged INR with FFP is not recommended to decrease the rate of procedure-related clinically relevant bleeding (LoE 2, strong recommendation).
In patients with cirrhosis undergoing invasive procedures, does administration of prothrombin complex concentrates (PCCs) decrease the rate of procedure-related clinically relevant bleeding?
Recommendation
  1. In patients with cirrhosis undergoing invasive procedures, routine use of PCCs to decrease the rate of procedure-related clinically relevant bleeding is discouraged (LoE 3, weak recommendation).
Future perspectives
  1. Trials – preferably randomised placebo-controlled trials should assess the safety and efficacy of PCCs and their optimal dosage for decreasing procedure-related clinically relevant bleeding episodes.
In patients with cirrhosis undergoing invasive procedures, does correction of thrombocytopenia by infusion of platelet concentrates or by thrombopoietin receptor (TPO-R) agonists decrease the rate of procedure-related clinically relevant bleeding?
Statement
  1. In patients with cirrhosis undergoing invasive procedures, no studies have specifically evaluated whether the infusion of platelet concentrates or TPO-R agonists decrease the rate of procedure-related clinically relevant bleeding (LoE 1).
Recommendation
  1. In patients with cirrhosis undergoing invasive procedures, infusion of platelet concentrates or use of TPO-R agonists is not recommended when platelet count is above 50 × 109/L or when bleeding can be treated by local haemostasis (LoE 3/4, strong recommendation).
  2. In patients undergoing high-risk procedures in whom local haemostasis is not possible and platelet count is between 20 × 109/L and 50 × 109/L infusion of platelet concentrates or TPO-R agonists should not be routinely performed but may be considered on a case-by-case basis (LoE 3/4, strong recommendation).
  3. In patients undergoing high-risk procedures in whom local haemostasis is not possible and platelet count is very low (<20 × 109/L) infusion of platelet concentrates or TPO-R agonists should be considered on a case-by-case basis (LoE 3/4, strong recommendation).
Future perspectives
  1. Large observational cooperative studies are recommended to collect data, in patients with cirrhosis undergoing invasive procedures, on real-life incidence of bleeding and on complications associated with these events and with inherent therapies. Placebo-controlled, randomised trials including a large proportion of patients with cirrhosis and very low platelet counts undergoing high-risk invasive procedures should be performed, with clinically significant bleeding as a primary endpoint, to assess the role of platelet transfusions in extreme situations. Similar trials should be performed to assess the utility of TPO-R agonists
In patients with cirrhosis undergoing invasive procedures, does correction of acquired fibrinogen deficiency by administration of fibrinogen concentrate or cryoprecipitate decrease the rate of procedure-related clinically relevant bleeding?
Recommendation
  1. In patients with cirrhosis undergoing invasive procedures, routine correction of fibrinogen deficiency to decrease the rate of procedure-related clinically relevant bleeding is discouraged (LoE 4, strong recommendation).
Future perspectives
  1. Large prospective studies aiming at evaluating the qualitative and quantitative role of fibrinogen deficiency in patients with cirrhosis undergoing invasive procedures are warranted.
In patients with cirrhosis undergoing invasive procedures, does correction of anaemia decrease the rate of clinically relevant bleeding?
Recommendations
  1. In patients with cirrhosis, every effort should be made to optimise haemoglobin levels by treating iron, folic acid, vitamin B6, and vitamin B12 deficiencies, especially in those patients likely to undergo invasive procedures (LoE 5, weak recommendation).
  2. In the setting of invasive procedures, prophylactic red blood cell transfusion with the aim of decreasing the risk of procedure-related bleeding is not recommended (LoE 5, weak recommendation).
    Future perspectives
    1. Large observational studies are recommended to evaluate whether anaemia (alone or combined with thrombocytopenia) is associated with bleeding events in patients with cirrhosis.
    In patients with cirrhosis undergoing invasive procedures, does administration of antifibrinolytic drugs such as tranexamic acid decrease the rate of clinically relevant bleeding?
    Recommendation
    1. In patients with cirrhosis undergoing invasive procedures, routine use of tranexamic acid to decrease the rate of procedure-related clinically relevant bleeding is discouraged (LoE 4, weak recommendation).
    Future perspectives
    1. Prospective studies are needed to evaluate the role of the hyperfibrinolytic state on post-procedural bleeding in patients with cirrhosis undergoing invasive procedures. Large clinical studies are needed to evaluate antifibrinolytic therapy in patients with cirrhosis undergoing invasive procedures, especially in those with an established (or suspected) hyperfibrinolytic state.
    In patients with cirrhosis with abnormal laboratory tests (prothrombin time, APTT, platelet count, fibrinogen) undergoing prophylactic band ligation, is correction of these tests by blood products or factor concentrates indicated to prevent bleeding?
    Recommendation
    1. In patients with stable cirrhosis and abnormal laboratory tests (prothrombin time, APTT, platelet count, fibrinogen) undergoing prophylactic band ligation, administration of blood products or factor concentrates with the aim of avoiding post-ligation bleeding is not recommended (LoE 5, strong recommendation).
    Should antiplatelet and/or anticoagulant agents be discontinued in patients with cirrhosis before invasive procedures to decrease the rate of procedure-related clinically relevant bleeding?
    Recommendation
    1. In patients with cirrhosis, antiplatelet and/or anticoagulant agents should be managed following the same guidelines as in patients without cirrhosis before invasive procedures (LoE 4, strong recommendation).
    In patients with cirrhosis, should invasive procedures be performed with specific modalities (e.g. experienced operators, imaging guidance) to reduce procedure-related bleeding?
    Recommendation
    1. In patients with cirrhosis, imaging guidance is recommended for liver biopsy, central venous line placement and jugular puncture for TIPS placement (LoE 3, strong recommendation).
    Future perspectives
    1. Any new specific modality aimed at reducing the risk of procedure-related bleeding should be evaluated in prospective randomised trials.
    Should patients with cirrhosis undergoing invasive procedures be monitored differently for bleeding complications than patients without cirrhosis?
    Recommendation
    1. Patients with cirrhosis undergoing invasive procedures should be monitored for bleeding complications in the same way as patients without cirrhosis (LoE 3, strong recommendation).
    Future perspectives
    1. Any change in standard surveillance after invasive procedures in patients with cirrhosis (e.g. shortened hospital surveillance) should be evaluated prospectively.
    In patients with cirrhosis and active variceal bleeding, besides vasoactive treatment and endoscopic therapy, is correction of haemostatic alterations indicated to stop bleeding?
    Recommendations
    1. In patients with cirrhosis and active variceal bleeding, if haemostasis is achieved with portal hypertension-lowering drugs and endoscopic treatment, correction of haemostatic abnormalities is not indicated (LoE 3, strong recommendation).
    2. In case of failure to control haemorrhage, the decision to correct haemostasis should be considered on a case-by-case basis (LoE 3, strong recommendation).
    3. In patients with cirrhosis and active variceal bleeding, tranexamic acid should not be used (LoE 2, strong recommendation).
    In patients with cirrhosis and active bleeding related to portal hypertension, but not to varices (e.g. portal hypertensive gastropathy), is correction of a prolonged prothrombin time, fibrinogen deficiency or thrombocytopenia by transfusion of FFP, fibrinogen concentrate/cryoprecipitate, PCCs, or platelet concentrate indicated to stop bleeding?
    Statement
    1. No studies evaluating correction of haemostasis in patients with cirrhosis and active bleeding related to portal hypertension, but not to varices (e.g., portal hypertensive gastropathy), are available (LoE 5).
    Recommendations
    1. In patients with cirrhosis and active bleeding related to portal hypertension, but not to varices (e.g., portal hypertensive gastropathy), bleeding should be managed with portal hypertension-lowering measures (LoE 5, weak recommendation).
    2. In the case of failure to control haemorrhage with portal hypertension-lowering drugs, the decision to correct haemostasis should be considered on a case-by-case basis (LoE 5, weak recommendation).
    Future perspectives
    1. Large observational cooperative studies are recommended to collect data on the treatment (including correction of haemostatic abnormalities) of patients with overt portal hypertensive gastropathy-related bleeding and/or chronic anaemia due to portal hypertensive gastropathy-related bleeding in whom vasoactive treatment and endoscopic therapy failed to stop bleeding.
    In patients with cirrhosis that are actively bleeding from a non-portal hypertensive cause, is correction of a prolonged prothrombin time, fibrinogen deficiency, or thrombocytopenia by transfusion of FFP, fibrinogen concentrate/cryoprecipitate, PCCs, or platelet concentrate indicated to stop bleeding?
    Recommendations
    1. In patients with cirrhosis who are actively bleeding from a non-portal hypertensive cause, active bleeding should first be addressed by local measures and/or interventional radiology procedures (LoE 4, strong recommendation).
    2. In those patients in whom local measures fail to stop the bleeding, addressing contributing factors (renal failure, infection or sepsis, and anaemia) may reduce bleeding while correction of haemostatic abnormalities can be considered on a case-by-case basis (LoE 5, weak recommendation).
    Future perspectives
    1. Preferably randomised, placebo-controlled studies should explore the best strategies to correct haemostasis in patients with non-portal hypertension-related bleeding which is unresponsive to local measures. Depending on the clinical context, low-volume products (fibrinogen concentrates and PCCs) or platelet concentrates are the agents of choice for such studies.
    In patients with cirrhosis that are actively bleeding from a non-portal hypertensive cause, is administration of antifibrinolytic drugs such as tranexamic acid indicated to stop bleeding?
    Recommendation
    1. In patients with cirrhosis, routine use of antifibrinolytic agents to treat active bleeding from a non-portal hypertensive cause is discouraged (LoE 5, weak recommendation).
    Future perspectives
    1. Preferably randomised, placebo-controlled trials should assess the safety and efficacy of antifibrinolytic agents in treating active non-portal hypertensive bleeding episodes that are unresponsive to local measures.
    In patients with cirrhosis who are actively bleeding, is pro-haemostatic management better guided by viscoelastic tests than by routine coagulation tests?
    Statement
    1. There is initial evidence that the use of viscoelastic tests is associated with decreased blood product use in patients with cirrhosis and active upper gastrointestinal bleeds, without differences in bleeding control and mortality (LoE 3).
    Recommendation
    1. Given the benefits of reducing blood transfusion, viscoelastic tests can be used when available (LoE 1, strong recommendation).
    Future perspectives
    1. Viscoelastic tests’ ability to guide pro-haemostatic management in patients with cirrhosis and active bleeding should be prospectively evaluated in randomised controlled trials.
    Are patients with cirrhosis at risk of developing venous thromboembolism (VTE/deep vein thrombosis [DVT]/pulmonary embolism [PE])?
    Statement
    1. Based on clinical observations and laboratory findings, it can be concluded that the risk of developing DVT/PE is at least as high in patients with cirrhosis as in the general population (LoE 2).
    Future perspectives
    1. Data from large observational studies will provide the basis for prophylactic treatment of venous thromboembolism in patients with cirrhosis.
    Can clinical prediction scores be used to decide which patients with cirrhosis are at risk of VTE (DVT/PE)?
    Recommendation
    1. Clinical prediction scores, such as the Padua prediction score (>3 or ≥4 or IMPROVE score (≥4) can be used to predict which patients with cirrhosis are at high risk of developing lower limb DVT and/or PE (LoE 3, strong recommendation).

    Future perspectives

    1. Large observational cooperative studies are recommended to collect data on the real-life ability of clinical prediction scores to assess the risk for VTE in patients with cirrhosis.
    Can viscoelastic tests or other laboratory tests be used to decide which patient with cirrhosis is at risk of VTE (DVT/PE)?
    Recommendation
    1. The use of viscoelastic tests or other laboratory tests to identify which patients with cirrhosis are at risk of VTE (DVT/PE) is not recommended (LoE 5, strong recommendation).
    Future perspectives
    1. Prospective studies should explore the predictive value of laboratory tests alone or in combination with clinical scores to predict the risk of VTE during hospitalisation.
    Is administration of thromboprophylaxis with LMWH or DOACs indicated to decrease the incidence of lower limb DVT (DVT/PE) in high-risk patients with cirrhosis?
    Recommendations
    1. In patients with cirrhosis at risk of DVT/PE, thromboprophylaxis with LMWH can be recommended as it has a reasonable safety profile, but efficacy is unclear based on available data (LoE 3, weak recommendation).
    2. In patients with Child-Pugh class A and B cirrhosis at risk of DVT/PE, thromboprophylaxis with DOACs can be recommended as DOACs have a reasonable safety profile in these patients, but efficacy data are still limited. In patients with Child-Pugh C cirrhosis, DOACs are not recommended (Safety: LoE 2; Efficacy: LoE 4; weak recommendations).
    Future perspectives
    1. Large observational cooperative studies, or ideally randomised controlled trials, are recommended to determine the safety and efficacy of LMWH or DOACs in hospitalised patients with cirrhosis at high risk of VTE.
    Can vitamin K agonists or LMWH be used in the treatment of DVT/PE in patients with cirrhosis?
    Recommendation
    1. For treatment of DVT/PE, vitamin K antagonists should be used with caution in patients with cirrhosis, as these patients can have baseline altered INR and thus target INR remains unknown. In patients with Child-Pugh A, LMWH, and vitamin K antagonists are reasonable options. Until more data become available, LMWH is recommended for treatment of DVT/PE in patients with Child-Pugh B and Child-Pugh C cirrhosis, whereas UFH is the treatment of choice in case of renal failure (LoE 4, weak recommendation).
    Future perspectives
    1. Large observational cooperative studies are needed to assess safety and efficacy of vitamin K antagonists and LMWH in the treatment of DVT/PE in patients with cirrhosis. Tools able to monitor their effects on haemostasis should be refined.
    Can DOACs be used in the treatment of DVT/PE in patients with cirrhosis?
    Recommendations
    1. For the treatment of DVT/PE in patients with cirrhosis, currently available data suggest that there are no major concerns regarding the safety of DOACs in patients with Child-Pugh class A cirrhosis. Due to the possibility of accumulation, DOACs should be used with caution in Child-Pugh class B patients, as well as in patients with creatinine clearance below 30 ml/min. The use of DOACs in Child-Pugh class C patients is not recommended (LoE 4, strong recommendation).
    Future perspectives
    1. Large observational cooperative studies are needed to assess the safety and efficacy of DOAC therapy in patients with Child-Pugh B or C cirrhosis.

    Texto completo:

     Journal of Hepatology 2022 vol. – j 1–34.  |     https://doi.org/10.1016/j.jhep.2021.09.003  

    Compartir

    Leave A Comment