Relationship Between Conventional Coagulation Tests and Bleeding for 600 Consecutive Liver Transplantations

Historically, orthotopic liver transplantation (OLT) has been associated with major blood loss and the need for massive blood product transfusions (1); however, a significant decrease in blood loss and blood product requirement has been observed during OLT over the last two decades (2). Despite these improvements, there is a large variability of blood product transfusion rates between liver transplantation centers. A substantial body of evidence suggests that the use of blood products during an OLT is associated with increased morbidity and mortality (2). Ironically, approximately one third of liver recipients acquire their liver disease through transfusion of blood products. Treatment approaches that reduce blood loss and transfusion volume are desirable; however, which patients would benefit from these treatments? A number of liver transplantation centers have tried to identify variables that predict blood transfusion rate, with conflicting results. Previous studies found that the only laboratory variable linked to transfusion of red blood cells (RBC) units was the baseline hemoglobin (Hb) value. A second variable, preemptive phlebotomy (PPh) was found to be linked to nontransfusion and an increased survival (2). Historically, blood loss was difficult to accurately measure during OLT but the recent widespread use of the cell saver allowed for a more precise assessment. Coagulation defects identified in traditional laboratory screening tests have long been believed to predict bleeding in liver recipients, but this point of view has been challenged (3). The goal of this present work was to predict bleeding and provide guidance on specific therapy (fibrinogen concentrate, prothrombin complex, and coagulation concentrates) to prevent bleeding and reduce the need for conventional blood product transfusions (RBC, plasma, platelets). Six hundred consecutive OLTs performed on adults between October 2002 and June 2012 were investigated. In the absence of uncontrollable bleeding, coagulation disorders were not corrected before or at the time of transplantation. No fresh-frozen plasma, platelets, or cryoprecipitate were given; a “wait and see” approach of rescue therapy was used instead of prophylactic or preventive interventions. The triggering Hb level for RBC transfusion was set between 60 and 70 g/L. Neither thromboelastrography nor thromboelastometry monitoring was used. Each anesthesiologist tried to lower the central venous pressure, before the anhepatic phase by approximately 33%, by using a restrictive intravenous fluid infusion approach or phlebotomy without volume replacement (to decrease the portal venous pressure) or by using a combination of both techniques. For the logistic regression analysis, the dependent variable “blood loss” was considered a dichotomic variable, using the median to split the group (820 mL). Table 1 compares the demographic, health, and surgical characteristics for the whole group of 600 patients: the group was split into two halves according to the median blood loss (<820 mL vs. ≥820 mL), and both groups were essentially the same. For the whole group of 600 patients, the mean intraoperative transfusion rate per OLT was: 0.5±1.3 RBC units, 0.2±1.2 fresh-frozen plasma, and 0.2±1.2 platelet units. Neither albumin nor cryoprecipitate was used. The average final Hb value for the whole group was 94±24 g/L. Table 2 summarizes the logistic regression analysis and odds ratio (OR) for blood loss, for no RBC transfusion, and for survival. Univariate and multivariate analyses demonstrated a correlation between no blood loss and PPh. When PPh was used, the risk of bleeding in excess of the median (820 mL) decreased by 32% (OR, 0.679; confidence index (CI), 0.484–0.955). Two other analyses were made on transfusion of RBC and survival as a dependent variable. For each increase of 1 g/L of starting Hb concentration, the risk of RBC transfusion was decreased by 1.1% (OR, 0.950; CI, 0.935–0.996). Similarly, the risk of RBC transfusion was increased by a factor of 4 when PPh was not performed (OR, 4.211; CI, 2.127–7.353). For each intraoperative transfusion of one RBC unit, the risk of dying within a year increased by 25% (OR, 1.252; CI, 1.071–1.465). When PPh was performed, the risk of dying was decreased by 50% (OR, 0.497; CI, 0.267–0.923).TABLE 1: Demographic, health, and surgical characteristics for the whole group of 600 patients, and the ones who bleed less or more than the median (820 mL)TABLE 2: Summary of the logistic regression model and odd ratios for blood lossUnfortunately, this work did not find any variable linked to blood loss except a negative correlation with the use of PPh. Traditional variables describing disease severity and the degree of coagulopathy in cirrhosis failed to predict bleeding. Preoperative Hb concentration was the only laboratory value, and PPh was the only clinical intervention that showed correlation with RBC transfusion rate. Moreover, PPh and “bloodless surgery” were linked to the 1-year postoperative survival in OLT. As was demonstrated in our previous work, phlebotomy decreased portal venous pressure and decreased bleeding during the dissection phase (4, 5); thus, decreasing the need of blood product transfusion and increasing the 1-year patient survival. Conventional tests (international normalized ratio, platelets count, fibrinogen) are poor predictors of clinical bleeding tendency. Hemostasis in patients with cirrhosis is characterized by a decline in both procoagulant and anticoagulant proteins (6). Interestingly, in patients with cirrhosis, a low platelet count is frequently counterbalanced in part by highly elevated levels of the platelet adhesive protein von Willebrand factor and a low plasma level of the von Willebrand factor cleaving protease ADAMTS 13 (7). Is there place for conventional coagulation tests (international normalized ratio, platelet count, and fibrinogen) to guide transfusion therapy? There is no documented benefit from aggressive prophylactic correction of laboratory values that have no correlation with bleeding during OLT. Excessive and unnecessary blood component transfusion should be abandoned to reduce morbidity and mortality. For 600 consecutive liver transplantations, the mean intraoperative blood loss was 1127±1050 mL with a median of 820 mL. The mean intraoperative transfusion rate was 0.5 RBC unit per patient and nearly 80% of patients who did not receive any blood product. There was no link between the severity of the recipients’ disease or the preexisting coagulation defects and bleeding. Preemptive phlebotomy was the only procedure linked with a decreased blood loss, a decrease transfusion rate, and an increased 1-year survival rate after liver transplantation. Baseline Hb value was the only laboratory test correlating with intraoperative RBC transfusion. Luc Massicotte 1 André Y. Denault1 Lynda Thibeault2 Zoltan Hevesi3 Anna Nozza4 André Roy5 1 Department of Anesthesiology Centre Hospitalier de l’Université de Montréal Hôpital St-Luc., Montreal QC, Canada 2 Epidemiology Department Centre Hospitalier de l’Université de Montréal, Montreal QC, Canada 3 Department of Anesthesiology University of Wisconsin Madison, WI 4 Institut de Cardiologie de Montréal Montreal QC, Canada 5 Hepatopancreatobilliary Surgery Centre Hospitalier de l’Université de Montréal, Montreal QC, Canada

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.