Transfusing common sense instead of blood products into coagulation testing in patients with cirrhosis: Overtreatment not equal safety.
Robert S. Rahimi, M.D.
Rahimi, R. S. and J. G. O’Leary (2016). “Transfusing common sense instead of blood products into coagulation testing in patients with cirrhosis: Overtreatment not equal safety.” Hepatology 63(2): 368-370.
In patients with cirrhosis the imbalance of procoagulants and anticoagulants combined with potential alterations in fibrinolysis and platelet number and function can alter standard laboratory coagulation testing. Prothrombin time and platelet count are frequently abnormal and, in our risk-averse health care system, often result in preprocedure transfusions to achieve “safer” thresholds. But what are we actually achieving: a risk of portal hypertensive bleeding, transfusion reaction, transfusion-related acute lung injury, infection transmission, human leukocyte antigen (HLA) antibody development, superior test results, or improved coagulation? According to the International Symposium on Coagulopathy and Liver Disease, rebalanced hemostasis is what we need to measure, which stems from both the prohemostatic and antihemostatic pathways working in concert. Disproportionately fewer nonportal hypertension-related bleeding complications occur in patients with cirrhosis relative to their prothrombin time and platelet values. This results from relatively rebalanced hemostasis. Despite this, many health care professionals continue to transfuse blood products (i.e., fresh frozen plasma [FFP] and platelets) prophylactically to improve laboratory profiles, regardless of function and prior data demonstrating a low risk of bleeding complications in patients with cirrhosis, especially for inimally invasive procedures like paracentesis. Fortunately, the study by De Pietri and colleagues in this issue of Hepatology offers an alternative to unnecessary preprocedural use of blood products in patients with cirrhosis with an elevated international normalized ratio (>1.8) and/or thrombocytopenia (<50 × 103/µL) using thromboelastography (TEG) to guide transfusions.4 TEG was adopted into clinical practice in 1985, more than 35 years after the procedure was first pioneered for research purposes. Using approximately 0.35 mL of whole blood in an oscillating cup at 37oC, this test uses mechanically transduced waves at different rates to graphically illustrate the patient's coagulation function. TEG is already used clinically in liver transplant, cardiovascular, trauma, and obstetric surgery where dynamic changes in blood volume and coagulation occur. This semiautomated bedside instrument easily yields an overall coagulation assessment in 30 minutes, with prolonged times resulting in blood product transfusions. A prolonged reaction time measures clot formation, indicative of coagulation factor function and results in transfusion (i.e., usually FFP), while a prolonged maximum amplitude measures clot strength, indicative of platelet function and results in transfusion. The authors of this open-label, intention-to-treat trial randomized patients to TEG-guided transfusion where patients only received transfusion if their functional coagulation was altered versus standard of care preprocedural prophylactic transfusion therapy.4 Of note, only 16.7% of TEG-guided patients received transfusions compared to 100% of patients in the standard of care arm (P = 0.009). Although platelet levels >50,000/µL have sufficient thrombin production, the use of FFP and platelet transfusions in patients with cirrhosis rarely, if ever, achieves complete normalization of coagulation parameters by standard testing. Of note, although allogenic transfusions at times are necessary, their cost and more importantly risk must also be taken into account (Table 1). The risk of overtransfusion was highlighted in a recent randomized controlled trial comparing liberal (hemoglobin <9 g/dL) versus restrictive (hemoglobin <7 g/dL) pack red blood cell (RBC) transfusion strategies in acute upper gastrointestinal bleeding. This trial subsequently resulted in a lower hemoglobin threshold (<7 g/dL) for RBC transfusion in general clinical care. Furthermore, liberal blood transfusions in patients with cirrhosis, compared to restrictive strategies, had the additional risk of worsening portal hypertension (P = 0.03) and resulted in higher bleeding complication rates (P = 0.01), a longer length of stay (P = 0.01), and more adverse events (P = 0.02; especially transfusion associated overload, P = 0.001), culminating in an increased all-cause 45-day mortality (P = 0.02).