Plasma transfusion trials and tribulations

Frozen plasma (FP) transfusions are commonly used to correct clotting factor deficiencies especially among critically ill patients.1 In such “coagulopathic” patients with bleeding, FP transfusions are justified to restore hemostasis and avoid excess morbidity or mortality.2 However, most FP transfusions are administered to prevent, rather than to treat, bleeding, typically in patients with an upcoming invasive procedure.3 In that setting, it remains unclear whether FP transfusions make any difference.4 Complicating matters further is the way in which intensive care unit (ICU)-acquired factor deficiencies and, in turn, response to FP transfusions, are measured. The international normalized ratio (INR) has been misused for so long that its interpretation has become obscure. Developed as a measure of the anticoagulant effect of warfarin, the INR is a standardized way of reporting the prothrombin time (PT) normalized for different types of PT reagents used across laboratories.5 The INR has excellent correlation with effective warfarin dose,6, 7 but it may not accurately assess global impairments in hemostasis due to liver disease, sepsis, or coagulation factor consumption, which are so commonly encountered in critically ill patients. Equipoise exists around the use of prophylactic FP transfusion to correct a prolonged INR in preparation for an invasive procedure: On the one hand, it may reduce bleeding and improve clinical outcomes; on the other hand, it is associated with a risk of infection,8 allergic reactions, transfusion-associated circulatory overload, transfusion-associated lung injury, and multiorgan failure.9 The pressing need for randomized controlled trials (RCTs) in this area has been recognized for some time; however, logistic and methodologic challenges including difficulties with blinding, patient selection, and buy-in have hampered progress. The RCT by Müller and colleagues,10 published in this issue of TRANSFUSION was a brave attempt to address this enduring and important clinical question. In this trial, ICU patients with a mild INR elevation (1.5-3.0) were randomly assigned to receive or not receive FP transfusion before a planned invasive procedure (central venous catheter placement, percutaneous tracheostomy, chest tube, or abscess drainage procedures). The primary outcome was procedure-related bleeding. The trial faced several challenges that ultimately resulted in its early termination. Enrollment was slow because of strict eligibility criteria. The screening process was complex and involved identifying potential participants with an INR that was within study range, applying exclusion criteria, and capturing those patients for whom a procedure was planned. On top of that, a significant proportion of patients were missed, refused to participate, or were denied participation by their treating physician. Thus, 81 of 263 eligible patients (30.8%) with a planned procedure, or 81 of 1478 patients (5.5%) with a suitable INR, were ultimately randomized over 3 years across four centers. Because of a low event rate, the investigators changed the primary outcome from major procedure-related bleeding to procedure-related bleeding of any severity. The initial sample size was calculated based on an estimated event rate of major procedure-related bleeding of 1%; however, estimates of bleeding rates in patients with an elevated INR who undergo invasive procedures have not been well established due to heterogeneous definitions of coagulopathy and bleeding and diverse patient populations.8, 11-17 After 81 patients were enrolled (of the planned 400), one major bleed had occurred. While this event was consistent with their prediction, this low event rate was too small to conduct the planned inferiority analysis. In their final conclusion, the authors reported that there was no difference in bleeding complications with or without FP; however, the true meaning of these results is uncertain since most bleeds were minor and the post hoc analysis was underpowered. ICU trials are no easy task, because of the high acuity of illness among the participants, frequent reliance on substitute decision makers for consent, and rotating staff caring for the patients.18 Moreover, FP transfusion practice is habitual, often rooted in tradition and associated with strong beliefs, making its omission potentially unpalatable for clinicians even in the context of an RCT. A previous RCT using prophylactic FP transfusion to prevent hepatobiliary procedure-related bleeding was also stopped early because of difficulties with accrual.19 Thus, many of these challenges could have been anticipated by the investigators, but the extent to which these and other barriers would impact the trial's success was underestimated. Ultimately, the trial was truncated after only one-fifth of the target sample was enrolled thus leaving us still uncertain as to whether withholding FP transfusion is no worse than giving it. Incomplete trials also present ethical dilemmas since they expose patients to potential risks without the benefit of knowledge gained, and they consume limited research funds. The trial by Müller and colleagues identified important challenges in the conduct of a prophylactic FP transfusion trial in the ICU, particularly relating to narrow eligibility criteria and missed patients on study. Several solutions can be considered for the next attempt including broadening the range of acceptable INR values, close screening of all potentially eligible patients even outside of daytime hours, and where applicable, waived consent. These conditions would require intense educational efforts for all staff. In hindsight, a pilot trial of FP transfusions may have identified some of these limiting feasibility challenges beforehand.18 This strategy has been successful for other ICU transfusion trials: For example, a pilot trial of restrictive versus liberal RBC transfusion strategies for older mechanically ventilated critically ill patients demonstrated that compliance with the protocol was high and hemoglobin separation between groups was good;20 a pilot trial of fluid resuscitation with 5% albumin versus normal saline in early septic shock demonstrated that the target rate of recruitment was achievable;21 and the Age of Blood Evaluation Study (ABLE) pilot trial helped define strategies to ensure an adequate supply of fresh RBCs throughout the trial.22 The trial by Müller and coworkers is among the first RCTs to test the effect of prophylactic FP transfusion in the ICU to prevent bleeding. Although it did not provide an answer to this question, it has taught us some important lessons about the design and conduct of such a trial in the future. The authors have disclosed no conflicts of interest.

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.