Koliogiannis, Dionysios

Abstract Background Machine perfusion represents the most significant breakthrough in liver transplantation (LT) since the discovery of cyclosporin. Yet, the optimal perfusion strategy remains unsettled. Randomized trials demonstrate that hypothermic oxygenated perfusion (HOPE) improves graft survival and function and reduces ischemic cholangiopathy, whereas normothermic machine perfusion (NMP) decreases post-transplant injury. Real-world comparisons are lacking and challenging due to heterogeneity in donor-recipient risk profiles and regional differences. Aims To compare real-world outcomes of HOPE and NMP in LT. Methods We analyzed consecutive NMP-preserved LTs from seven North American centers between 2021 and 2024. We compared NMP to the European HOPE-REAL cohort, comprising HOPE-treated LTs from 22 centers between 2012 and 2021. Analyses were stratifyied by graft type and risk category and adjusted for donor age, donor risk index, and balance of risk score using entropy balancing. Results A total of 470 NMP and 1202 HOPE-treated grafts were analyzed, revealing major differences in risk profiles (Figure 1). In NMP, death-censored graft survivals at 1, 2, and 3 years were >95% in donation after brain death (DBD) and >94% in donation after circulatory death (DCD) grafts. Comparable outcomes were observed in the HOPE cohort with >93%, in DBD, and >87% in DCD after up to 3 years, despite significantly higher donor age and longer functional warm ischemia in the HOPE DCD cohort (Figure 2). After risk adjustment, death-censored graft survival remained similar between both modalities (Table 1). Non-anastomotic biliary stricture rates after DCD-LT were comparable (8.5% vs. 12.4%), independent of whether continuous or end-ischemic NMP was performed. Conclusion This first transatlantic real-world report demonstrates excellent outcomes with both NMP and HOPE despite higher graft risk in Europe. Machine perfusion is the new standard of care in LT, while further understanding is needed regarding the distinct benefits of each modality across risk profiles.Figure 1.Graft risk in NMP and HOPE stratified by graft type. Benchmark case was defined as primary transplant with MELD <20 and BAR score <9, extended criteria as either BMI >30kg/m2, donor age >65 years, CIT >12h. DCD grafts were categorized as definedFor image description, please refer to the figure legend and surrounding text. Figure 2.Death-censored graft survival in liver transplantation following NMP and HOPE stratified by graft type. Abbreviations: NMP (Normothermic Machine Perfusion), HOPE (Hypothermic Oxygenated Perfusion), DBD (Donation after Brain Death), DCD (Donation after circulatory death)For image description, please refer to the figure legend and surrounding text. Table 1.Risk-adjusted comparison of death-censored graft survival between NMP and HOPE.For image description, please refer to the figure legend and surrounding text.

ABSTRACT Background Living kidney donation is a crucial option for addressing the global organ shortage and providing kidney transplantation for patients suffering from end‐stage kidney disease. Ensuring donor safety necessitates a comprehensive preoperative assessment of kidney anatomy and function. This study evaluates the relationship between kidney volumes derived from deep learning‐based MRI volumetry, intraoperative kidney volume measurements, split renal function measured by renal scintigraphy, and post‐donation eGFR. Deep learning‐based MRI volumetry is hypothesized to be a reliable method with good correlation. Methods This retrospective study analyzed 178 living kidney donors. Deep learning MRI volumetry‐based kidney volumes were compared with intraoperative volumes of the explanted donor kidneys obtained using the water displacement method. Additionally, MRI‐based volume ratios were compared with scintigraphy‐based split renal function ratios to determine their ability to predict the kidney with poorer renal function and post‐donation eGFR. Results Deep learning‐based MRI volumetry strongly correlated with intraoperatively measured kidney volumes (Pearson's correlation; r = 0.7671; p < 0.0001), confirming its precision in volume estimation. Although MRI‐based kidney volume ratios demonstrated only a moderate correlation with scintigraphy‐based split renal function ratios ( r = 0.4798), MRI volumetry correlated with 1‐year post‐donation eGFR. It tended to be better than renal scintigraphy ( r = 0.6829 versus r = 0.6191). Conclusion Deep learning‐based MRI volumetry is a reliable, non‐invasive tool for estimating kidney volumes in living donors, offering a radiation‐free alternative for preoperative assessment. While it differs from renal scintigraphy in evaluating split renal function ratios, its correlation with post‐donation eGFR tends to be better, supporting its potential role in living kidney donor assessment.