INR-Triggered Transfusion In GI Bleeders From ER (I-TRIGER)

Transfusion-related acute lung injury (TRALI) is the most common cause of transfusion-related morbidity and mortality in the United States. It is very common and often unrecognized in the critically ill with the greatest incidence occurring in bleeding patients with liver disease. Plasma is the most blood component associated with this deadly complication and therefore patients with liver disease who frequently receive transfused plasma are at increased risk. The optimal plasma transfusion strategy for bleeding patients with liver disease is unknown and the investigators will evaluate this clinical question in a small pilot randomized controlled trial. The invstigators hypothesize that targetting a more restrictive INR Target (2.5) vs. an INR Target (1.8) will result in less hypoxemia, a TRALI surrogate without increasing bleeding complications.

Advances in the understanding of the coagulation imbalance in liver disease have experts questioning the clinical efficacy of current plasma transfusion practices in patients with liver disease. Having recently discovered a large previously unrecognized risk (TRALI) of plasma transfusion in this patient population, the investigators now believe the current clinical transfusion paradigm under-recognizes risk and overvalues the benefit of plasma transfusion in bleeding patients with liver disease. Though experts have recommended more judicious use of plasma, clinical practice remains variable. Transfusion triggers and thresholds are often arbitrarily set based on conventional coagulation studies and evidence to guide clinicians on plasma dosing required to achieve these laboratory thresholds does not exist. The investigators hypothesize that a restrictive plasma transfusion strategy in critically ill chronic liver disease patients with acute gastrointestinal bleeding will decrease a surrogate measure of TRALI without increasing bleeding complications (figure 1). With the collaborative support of the pulmonary/critical care, hepatology, and transfusion medicine services, the investigators will conduct a randomized controlled trial comparing a restrictive versus liberal strategy of plasma transfusion in bleeding patients with liver disease. In addition, investigators will refine and validate our plasma transfusion dosing algorithm so clinicians will have the tools to appropriately dose plasma to reach evidence-based transfusion targets. The development of TRALI is believed to require two pathophysiologic events. First, a pro-inflammatory stimulus, such as sepsis, leads to exposure of endothelial surface adhesion proteins and consequent capture of polymorphonuclear leukocytes (PMNs) within the pulmonary microvasculature. Second, these adherent PMNs are activated by mediators within transfused blood components, leading to neutrophilic inflammation and TRALI. Emerging evidence suggests that the process of neutrophil adhesion in the lung involves degradation of the endothelial glycocalyx, a thin layer of glycosaminoglycans (GAGs) lining the vascular lumen(S). In mice, sepsis results in pulmonary glycocalyx loss, neutrophil adhesion and subsequent development of ALI(S). Glycocalyx degradation is also associated with organ injury in humans, as evidenced by an increase in circulating GAG fragments (e.g. heparinoids) in septic shock. Circulating heparinoids can be detected quickly and accurately by a point of care heparinase-I modified thromboelastogram (TEG) study26-27. Detection of heparinoids by TEG may therefore indicate pulmonary microvasculature propensity for PMN adhesion (first event) and be utilized as a predictive biomarker for TRALI. Restrictive plasma transfusion strategies could then be individualized to high risk patients to decrease the probability of a second event resulting in the clinical syndrome of TRALI. In conjunction with the clinical trial, investigators will perform a translational observational study to assess whether detection of systemic heparinoids predict the subsequent development of a TRALI surrogate, post-transfusion hypoxemia. These clinical studies will pave the way for larger clinical trials guiding future plasma transfusion practice and decreasing the significant TRALI burden in the critically ill.

Respiratory Distress Syndrome, Adult, Gastrointestinal Hemorrhage, Liver Diseases, Transfusion Related Lung Injury

transfusion-related acute lung injury, Blood Component Transfusion, Gastrointestinal Hemorrhage, Cirrhosis, Fresh frozen plasma

Transfuse plasma to High INR target. Plasma will be transfused to reach a target INR=2.5 for 48 hours while patient is actively bleeding.

Transfuse plasma to Low INR target. Plasma will be transfused to reach a target INR=1.8 for 48 hours while patient is actively bleeding.

Using a dosing algorithm we will bolus plasma to reach an INR target (2.5) while patient is actively bleeding or 48 hours whichever comes first

Using a dosing algorithm we will bolus plasma to reach an INR target (1.8) while patient is actively bleeding or 48 hours whichever comes first

The development of hypoxemia will not distinguish between hydrostatic edema and TRALI, but investigators believe a significant change in oxygenation is clinically relevant and a more sensitive outcome variable for all transfusion-related pulmonary complications and therefore appropriate for use in this clinical trial.

The mean ∆S/F ratio immediately before and 60 minutes after transfusion of plasma vs. (RBCs or platelets) will allow investigators to analyze changes in oxygenation over time to further delineate which blood components are most temporarily associated with pulmonary edema.

Investigators will determine how many days a patient is alive and off mechanical ventilation at day 28 from enrollment.

Inclusion Criteria: Subjects will be eligible to participate in the study if they meet all of the following criteria: Admit to an ICU due to gastrointestinal bleeding AND an INR in first 12 hours >1.8; (INR ≥ 1.6 if received ≥ 2 units plasma) Patient has chronic liver disease defined as 1 or more of the three following diagnostic criteria: Previous diagnosis of chronic liver disease OR Imaging or biopsy diagnosis of cirrhosis Signs of portal hypertension (ascites, varices, hypersplenism) Laboratory evidence of synthetic dysfunction (INR>1.5, Bilirubin> 2.0, Albumin< 2.5) AND ≥2 physical exam findings on admission associated with chronic liver disease (palmar erythema, spider angiomata, asterixis, caput medusa, gynecomastia) Exclusion Criteria:Subjects will be ineligible to participate in the study if they meet any of the following criteria: Patient under age 18 OR pregnant OR incarcerated Patient meets criteria for acute respiratory distress syndrome (ARDS) (PaO2/FiO2<165)41 Patient admitted to ICU for re-bleed on same hospital admission OR has already received >4 units of plasma. Patient already underwent therapeutic endoscopy with noted hemostasis History of inheritable or acquired clotting or bleeding disorder (hemophilia A or B or acquired clotting factor inhibitor) Patient is being actively anticoagulated with vitamin K antagonists, direct thrombin inhibitors, heparins or anti-Xa antagonists Inability to obtain consent OR clinical team believes one of the transfusion strategies will be harmful to the patient Congestive heart failure (previous clinical diagnosis or Ejection Fraction (EF) <50%) Patient is do-not-resuscitate (DNR) or unexpected to live > 72 hours

Benson AB, Austin GL, Berg M, McFann KK, Thomas S, Ramirez G, Rosen H, Silliman CC, Moss M. Transfusion-related acute lung injury in ICU patients admitted with gastrointestinal bleeding. Intensive Care Med. 2010 Oct;36(10):1710-1717. doi: 10.1007/s00134-010-1954-x. Epub 2010 Jul 24.