Effects of massive transfusion (10-20 litres) versus ultramassive transfusion (≥20 litres) on mortality in adult liver transplant recipients: A propensity-score matched study

Abstract

Background Ultramassive perioperative fluid transfusion during orthotopic liver transplantation (OLT) identifies a high-risk recipient phenotype and is associated with substantially increased mortality compared with conventional massive transfusion. OLT frequently necessitates high-volume fluid resuscitation, yet the prognostic implications of ultramassive perioperative transfusion in this setting remain uncertain. We evaluated whether ultramassive transfusion (≥20 L total perioperative fluid) is independently associated with worse survival than massive transfusion (10– < 20 L) in adult OLT recipients. Methods In this single-centre retrospective cohort (2009–2023), we included adults undergoing primary OLT who received ≥10 L of total perioperative fluid (crystalloids, colloids, blood, and blood products administered intraoperatively and within 24 hours postoperatively). Propensity score matching was used to compare recipients receiving ultramassive (≥20 L) versus massive (10– < 20 L) transfusion, balancing recipient, donor, and intraoperative characteristics. Primary outcomes were 90-day, 3-year, and overall all-cause mortality; secondary outcomes included graft failure, graft and patient survival time, primary non-function, early allograft dysfunction, thrombotic complications, acute kidney injury, and hospital length of stay. Results Of 993 OLT recipients, 306 (30.8%) received ≥10 L perioperative fluid and comprised the unmatched cohort. In the propensity-matched cohort (n = 188), ultramassive transfusion was associated with significantly higher 90-day, 3-year and overall mortality compared with massive transfusion, with consistent effect estimates across multiple matching strategies and sensitivity analyses. In contrast, ultramassive transfusion showed no robust or consistent association with early allograft dysfunction, thrombotic complications, graft loss, or other secondary endpoints, likely reflecting low event rates and limited power. Conclusions Among adult OLT recipients exposed to high-volume perioperative resuscitation, ultramassive transfusion delineates a distinct high-risk phenotype characterised by substantially increased short- and long-term mortality. Although causal inference is precluded by the observational design, these findings suggest that reaching an ultramassive transfusion threshold may mark a particularly vulnerable subgroup that warrants intensified intraoperative optimisation strategies and tailored long-term surveillance.