Validation of mitochondrial FMN as a predictor for early allograft dysfunction and patient survival measured during hypothermic oxygenated perfusion

Hypothermic oxygenated machine perfusion (HOPE) preconditions liver grafts before transplantation. While beneficial effects on patient outcomes were demonstrated, biomarkers for viability assessment during HOPE are scarce and lack validation. This study aims to validate the predictive potential of perfusate flavin mononucleotide (FMN) during HOPE to enable the implementation of FMN-based assessment into clinical routine and to identify safe organ acceptance thresholds. FMN was measured in perfusate samples of 50 liver grafts at multiple time points. After transplantation, patients were followed up for development of early allograft dysfunction (EAD), transplantation, and 1-year survival. FMN concentrations were significantly higher for grafts that developed EAD at 5 and 60 minutes into HOPE ( p = 0.008, p = 0.026). The strongest predictive potential of FMN for EAD was observed at 5 minutes of HOPE with an AUC of 0.744. Similarly, 5-minute FMN was predictive for 1-year mortality ( p < 0.001), reaching a remarkable AUC of 0.890. Cutoffs for prediction of EAD (10.6 ng/mL) and early mortality (23.5 ng/mL) were determined and allowed risk stratification of grafts. Particularly, patients receiving low-risk grafts developed EAD in 9% of cases, while all patients survived the first postoperative year. In contrast, high-risk organs developed an incidence of EAD at 62%, accompanied by the necessity of retransplantation in 38% of cases. One-year mortality in the high-risk cohort was 62%. Evaluation of FMN as early as 5 minutes during HOPE allows for risk stratification of liver grafts. Low-risk grafts, according to FMN, display a negligible risk for patients. Yet, high-risk grafts are associated with increased risk for EAD, transplantation, and early mortality and should not be used for transplantation without further assessment.