Liver Transplantation at the Extreme: Analysis of Outcomes and Predictors of Futility in Acute-on-Chronic Liver Failure Grade 3 Patients on Maximal Life Support.

BACKGROUND & AIMS: While patients with grade 3 acute-on-chronic liver failure (ACLF-3), have high mortality, there is a subset requiring maximal life support. We report post liver transplant (LT) outcomes and identify factors associated with futility in extreme ACLF-3 (eACLF-3) recipients, all requiring intubation, vasopressors, and dialysis at LT. METHODS: We identified adult eACLF-3 LT recipients between January 2010 and December 2021. Pre-LT characteristics were compared among eACLF-3 recipients with and without the futile outcome (one-year or in-hospital mortality), predictors of futility were identified, and a logistic regression model was evaluated using 5-fold cross validation. Findings were validated using the Multi-Organ Dysfunction and Evaluation for Liver Transplantation (MODEL) Consortium. RESULTS: Of 1,608 adult LT recipients at our center, 177 (11%) were eACLF-3. Thirty-five (20%) eACLF-3 recipients experienced futility. Predictors of futility were previous upper abdominal surgery (OR=4.6, P=0.002), sepsis within 14 days prior to LT (OR= 5.9, P= 0.013), intubation due to pneumonia or acute respiratory distress syndrome (OR= 3.5, P= 0.010), worsening PaO2:FiO2 ratio (OR= 3.9, P=0.008), and pre-LT base deficit (OR=1.27, P=0.007). Futility was predicted with an AUROC of 0.74 (SD 0.15). When compared to the MODEL Consortium, our cohort had improved three-, six-, and twelve-month survivals of 89%, 85%, and 80%, compared to 83%, 79%, and 70% (p=0.048); however overall 5-year survival was similar (63% vs 59%, p=0.14). CONCLUSIONS: We report the largest single-center experience of LT for eACLF-3 patients, demonstrating acceptable survival despite high acuity, validated in the MODEL Consortium dataset. Prospective validation of the predictive model and identified risk factors is necessary, and may aid in appropriate selection of ACLF-3 patients requiring maximal life support prior to LT. IMPACT AND IMPLICATIONS: There has been persistent interest in the liver transplant community regarding transplant for acute-on-chronic liver failure grade 3 (ACLF-3) recipients. In this work, we report on a large single-center experience transplanting ACLF-3 recipients, specifically investigating an extreme subset of these patients who were all mechanically ventilated, receiving vasopressors, and receiving hemodialysis at time of transplant. We characterized their outcomes, identifying clear risk factors associated with futility in this hyper-acute subgroup and demonstrating that with appropriate selection, even patients in this extreme subgroup can experience acceptable five-year post-transplant survival. While we could not externally validate our model due to data availability, the obtained external patient cohort of extreme ACLF-3 patients broadly validates our findings. We feel that these results represent an important contribution to the liver transplant community given concerns about transplanting this hyper-acute subpopulation. These results may help guide physicians in selecting appropriate candidates for transplant.