Affiliation:
1. Department of Surgery University Medical Center Groningen Groningen The Netherlands
2. Recanati/Miller Transplantation Institute, Mt. Sinai New York City New York USA
3. Department of Nephrology University Medical Center Groningen Groningen The Netherlands
4. Department of Critical Care University Medical Center Groningen Groningen The Netherlands
Abstract
AbstractBackgroundEx vivo normothermic machine perfusion (NMP) is a promising tool for assessing an isolated kidney prior to transplantation. However, there is no consensus on the perfusate's optimal oxygen‐carrying capacity to support renal function. To investigate the association of hemoglobin levels with renal function parameters, a retrospective analysis of isolated, normothermically, perfused porcine kidneys was performed.MethodsBetween 2015 and 2021, a total of 228 kidneys underwent 4 h of NMP with perfusates that varied in hemoglobin levels. A generalized linear model was used to determine the association of hemoglobin levels with time‐weighted means of renal function markers, such as fractional sodium excretion (FENa) and creatinine clearance (CrCl). Stratified by baseline hemoglobin level (<4.5, 4.5–6, or >6 mmol/L), these markers were modeled over time using a generalized linear mixed‐effects model. All models were adjusted for potential confounders.ResultsUntil a hemoglobin level of around 5 mmol/L was reached, increasing hemoglobin levels were associated with superior FENa and CrCl. Thereafter, this association plateaued. When hemoglobin levels were categorized, hemoglobin <4.5 mmol/L was associated with worse renal function. Hemoglobin levels were neither significantly associated with proteinuria during NMP nor with ATP levels at the end of NMP. Hemoglobin levels >6 mmol/L showed no additional benefits in renal function.ConclusionIn conclusion, we found an association between baseline hemoglobin levels and superior renal function parameters, but not injury, during NMP of porcine kidneys. Furthermore, we show that performing a retrospective cohort study of preclinical data is feasible and able to answer additional questions, reducing the potential use of laboratory animals.