This blog aims to summarize the data available on how to safely perform invasive hospital procedures in patients with cirrhosis. Unfortunately, no prospective randomized controlled trials have been conducted to inform the practice of procedural medicine in cirrhotic patients. Therefore, our knowledge about the safety of performing bedside procedures in patients with cirrhosis is based on retrospective studies and expert opinion. Additionally, I will focus only on the common bedside procedures performed in the ED, ICU and on hospital wards such as thoracentesis, paracentesis, chest tube placement, central line placement and arterial line placement.
Most of the traditional laboratory tests used to assess the bleed risk for procedures are not valid in patient with chronic liver disease. The prothrombin time (PT/INR) and partial thromboplastin time (PTT) are coagulation tests that measure the levels of coagulation factors involved in the procoagulant arm of hemostasis. Patients with advanced cirrhosis will have decreased levels of clotting factors II, V, VII, IX, X and XI which will elevate the PT/INR. However, these same patients will also have increased levels of factor VIII, vWF and low levels of protein C, protein S and antithrombin that create a thrombotic tendency. The combination of changes leads to a rebalanced hemostasis in stable cirrhosis. The American Association for the Study of Liver Disease (AASLD) has clearly stated that PT/INR in patients with cirrhosis does not correlate with bleeding, and any attempt to correct with plasma is ineffective and likely to be detrimental.
Additionally, many patients with cirrhosis have chronic thrombocytopenia due to multiple factors: decreased thrombopoietin production, bone marrow suppression from HCV infection or active alcohol use, B12 or folate deficiency and splenomegaly with splenic sequestration. The bleeding risk by the quantitative thrombocytopenia when platelet counts fall below 50,000 is offset by the increased platelet adhesion by an elevation of vWF levels about 4-5x above normal.
One of the factors that can increase the bleed risk of acutely ill patients with cirrhosis is the development of accelerated intravascular coagulation and fibrinolysis (AICF). AICF is a new term that encompasses hyperfibrinolysis or low level DIC in patients with decompensated cirrhosis as a result of acute on chronic liver failure which can be manifested as grade 2 or 3 ascites, acute hepatic encephalopathy, gastrointestinal hemorrhage, or acute bacterial infections (eg, spontaneous bacterial peritonitis or bacteremia, urinary tract infection, pneumonia, or cellulitis) occurring within 2 weeks. AICF is marked by hypofibrinogenemia with fibrinogen levels <120 mg/dL.
The following are studies involving procedures performed in cirrhotic patients. One Italian study evaluated 852 Procedures carried out in 363 cirrhotics. Ten post-procedural bleeding events occurred and was unrelated to the platelet counts, to the degree of INR abnormalities, nor to the Child-Pugh-Turcotte classes.[i] A review of cirrhotic patients and severe coagulopathy (INR >1.8 or platelet count <50×109/L) undergoing invasive cardiovascular procedures found that bleeding complications happened in only 1 in 60 patients without platelet or plasma transfusions.[ii] In another multi-centre prospective study assessed the frequency of clinically significant bleeding in 380 cirrhotic patients with abnormal coagulation parameters, defined as an INR ≥1.5 and/or platelet count ≤50×109/L. In the low-risk procedure group, mostly entailing abdominal paracentesis, no patients had clinically significant post-procedure bleeding.[iii] In another study conducted on 90 Childs-Pugh class B or C cirrhotic patients undergoing central venous catheter placement, the authors found that bleeding events were minor and none of the patients had prolonged hospital stay due to complications of central venous catheter insertion. Additionally, the difference between transfused and controls was not statistically signiﬁcant. These studies also found that bleeding risk is increased if patients have decompensated portal hypertension, acute kidney injury or AICF with hypofibrinogenemia.[iv]
The following recommendations are a summary of the conclusions by the Society of Interventional Radiology (SIR), the AASLD, the American Gastroenterology Association (AGA) and the clinical studies that have been conducted in this area[v],[vi], [vii]:
- Low risk bedside procedures include ultrasound-guided thoracentesis, paracentesis, central line placement and radial arterial line placement and non-tunneled small-bore chest tube placement.
- Image-guided low risk bedside procedures can be performed safely if platelets >20K and fibrinogen >120 mg/dL irrespective of INR elevation in patients with stable cirrhosis.
- Global tests of clot formation (rotational thromboelastometry [ROTEM] or thromboelastography [TEG]) may eventually have a role in the evaluation of clotting in cirrhosis but currently lack validated targets.
- Blood products should be used sparingly because they increase portal pressure and carry a risk of transfusion-associated circulatory overload, transfusion-related acute lung injury, infection transmission, alloimmunization, and/or transfusion reactions and lack evidence to support benefit for historical threshold levels (INR>1.5 and platelets<50K).
- Anti-ﬁbrinolytic therapy may be considered in patients with persistent bleeding from mucosal oozing or puncture wound bleeding consistent with impaired clot integrity. Both ε-amino-caproic acid and tranexamic acid inhibit clot dissolution.
- Desmopressin may be given IV or subcutaneously at a dose of 0.3 mcg/kg (in 50 mL of saline over 15–30 minutes if IV) Desmopressin may be appropriate for administration to patients with end-stage renal disease and uremic platelet dysfunction.
We do know that the bleeding risk from invasive bedside procedures can be higher in patients with acute on chronic liver failure and in those with decompensated cirrhosis who are acutely ill from an infection, GI bleed or acute kidney injury. This is frequently due to AICF from hyperfibrinolysis and there are no studies in these patients to help inform decisions on the safe performance of invasive procedures in these patients. There is expert opinion that the fibrinogen level should be above 120 mg/dL, to consider antifibrinolytic agents (ε-amino-caproic acid or tranexamic acid [TXA]) for post-procedure bleeding and to consider DDAVP 0.3 mcg/kg IV for patients with renal insufficiency or presumed qualitative platelet dysfunction.
My personal practice for patients with decompensated cirrhosis who are acutely ill is to check a full DIC panel and to transfuse patients with cryoprecipitate to keep the fibrinogen >120 mg/dL, to use DDAVP 0.3 mcg/kg IV for any patient with AKI and to keep the platelets >30K for any patient with suspected AICF in whom I am performing an image-guided low-risk procedure. I am hoping that we will have large prospective RCTs soon to provide evidence-based recommendations for procedural medicine in patients with cirrhosis.
[v] Diagnosis, Evaluation and Management of Ascites, Spontaneous Bacterial Peritonitis, and Hepatorenal syndrome: 2021 Practice Guidance by the American Association for the Study of Liver Diseases. HEPATOLOGY. 2021;74, NO. 2: 1014-1048.
[vi] Society of Interventional Radiology Consensus Guidelines for the Periprocedural Management of Thrombotic and Bleeding Risk in Patients Undergoing Percutaneous Image-Guided Interventions—Part II: Recommendations. Indravadan J. Patel et al. J Vasc Interv Radiol 2019; 30:1168–1184