Brief Normothermic Machine Perfusion Rejuvenates Discarded Human Kidneys

Abstract

Normothermic machine perfusion (NMP) may allow resuscitation and improved assessment of kidneys before transplantation. Using discarded human kidneys, we investigated the mechanistic basis and translational potential of NMP compared with cold static storage (CS).

Methods: Discarded deceased donor kidneys (n = 15) underwent 1-hour NMP following CS. Renal perfusion, biochemical, and histologic parameters were recorded. NMP was directly compared with CS in paired donor kidneys using simulated transplantation with allogeneic whole blood, followed by assessment of the aforementioned parameters, in addition to RNA sequencing.

Results: Kidneys were successfully perfused, with improved renal blood flows and resistance over the course of perfusion, and evidence of urine output (median 21 mL), in all but one kidney. NMP completely resolved nonperfused regions in discarded donation after circulatory death kidneys. In paired kidneys (n = 4 pairs), transcriptomic analyses showed induction of stress and inflammatory pathways in NMP kidneys, with upregulation of pathways promoting cell survival and proliferation. Furthermore, the NMP pairs had significantly better renal perfusion (1.5-2 fold improvement) and functional parameters, and amelioration of cell death, oxidative stress, and complement activation.

Conclusions: In this pilot preclinical study using simulated transplantation of paired kidneys, NMP of discarded marginal kidneys demonstrated some significant mechanistic benefits in comparison to CS alone. NMP may have potential to reduce organ discards and enhance early graft function in such kidneys.

FIGURE 1.

Experimental flow diagram. Experimental pathways followed after kidney retrieval. Please note one set of paired kidneys (from donor 3) did not undergo ex vivo whole blood reperfusion as whole blood was not available. CS, cold static storage; NMP, normothermic machine perfusion; SWIT, second warm ischemic time.

FIGURE 2.

Perfusion, functional, and biochemical changes during normothermic machine perfusion (NMP). A, Renal blood flow and intra-renal resistance (IRR) during NMP, with a mean arterial pressure maintained between 75–85 mm Hg. B, Left panel—Urine output (UO) per h of NMP for each donor kidney. Right panel—Relationship between UO, creatinine clearance (CrCl), and fractional excretion of sodium (FeNa) in each donor kidney. C, Perfusate acid-base balance (pH, lactate, and bicarbonate) and electrolyte (sodium and osmolality) concentrations over the course of NMP, plotted for each individual kidney. DBD, donation after brain death; DCD, donation after circulatory death.

FIGURE 3.

Leukocyte efflux from the donor kidney during normothermic machine perfusion (NMP). Absolute cell counts for granulocytes, monocytes, NK cells, and lymphocytes were calculated by flow cytometry using the preperfusion sample as a baseline, followed by NMP arterial line sampling at selected time points. n = 3 kidneys. NK, natural killer.

FIGURE 4.

Normothermic machine perfusion (NMP) is possible in the presence of anatomical variation, and useful in poorly perfused kidneys. A and B, NMP is feasible and safe for kidneys with multiple renal arteries, achieves good renal blood flows, and intra-renal resistance. Kidneys shown are from (A) donor 7 (DBD-D4) and (B) donor 8 (DBD-D5). C, NMP is ideal for the assessment of kidneys discarded for poor in situ perfusion. Left panel—Kidney (donor 2; DCD-D1) visualized at end-cold storage, discarded due to a nonperfused lower pole and patchy middle region. Right panel—The lower pole of the same kidney is pictured 5 min after the commencement of NMP, showing complete resolution of the previously nonperfused area. DBD, donation after brain death; DCD, donation after circulatory death.

FIGURE 5.

Comparison of whole transcriptome gene expression following cold static storage (CS) or normothermic machine perfusion (NMP). Sequential biopsies were taken immediately before the commencement of NMP (end-CS), after 1 h of NMP (end-NMP), and after 1 h of simulated transplantation (end-ex vivo). A, Left panel—Gene expression changes (heatmap) after NMP in comparison to the end-CS period (for the same kidneys). Right panels—Pathway analyses, displaying the most up- (top) and down-regulated (bottom) pathways after NMP. B, Left panel—Heatmap showing differentially expressed genes in the NMP group of kidneys after ex vivo whole blood reperfusion (in comparison to the end-NMP samples from the same kidneys). Right panels—(top) HMGA1 is the most differentially expressed gene between end-NMP and end-ex vivo samples. Bottom, Scatter plot showing the most up- and down-regulated genes after ex vivo reperfusion (simulated transplantation) in comparison to end-NMP samples from the same kidneys. DBD, donation after brain death; DCD, donation after circulatory death.

FIGURE 6.

Changes in whole transcriptome gene expression following normothermic machine perfusion (NMP). Comparisons were conducted after a simulated second warm ischemic period of 30 min, followed by reperfusion of each kidney with whole allogeneic blood at a mean arterial pressure of 85 mm Hg and temperature of 37°C. A, Left panel—Gene expression heatmap comparing paired kidneys having cold static storage (CS) or NMP, after ex vivo whole blood reperfusion. Right panels—(top) Scatter plot outlining the most significantly up- and down-regulated genes in the NMP group in comparison to CS paired kidneys. Bottom, Most differentially expressed gene between NMP and CS kidneys (HSPH1), with an obvious difference in expression at the end-ex vivo time point for all 3 kidney pairs. B, Ingenuity Pathway Analysis showing top canonical pathways significantly up- or down-regulated in the NMP group of kidneys in comparison to CS kidneys. Pathways are ordered by magnitude of –log (P value). Left panel—Indication of pathway activation or repression based on the z score. A positive z score suggests pathway induction, and a negative z score denotes pathway suppression. Right panel—Percentage (indicated by colored bars) of total number of genes (indicated by numbers to right of bars) in a specific pathway that are differentially expressed in NMP compared with CS kidneys. DBD, donation after brain death; DCD, donation after circulatory death; IL, interleukin.

FIGURE 7.

Perfusion and functional parameters following cold static storage (CS) or normothermic machine perfusion (NMP) and subsequent simulated transplantation. A, Upper panels—Renal blood flow and intra-renal resistance (IRR) graphed for each individual donor kidney. Lower panels—Cumulative flow and IRR for the kidneys stored by CS alone in comparison to contralateral kidneys having CS followed by NMP. B, Comparison of renal functional parameters between the groups after simulated transplantation—urine output (UO), creatinine clearance (CrCl), fractional excretion of sodium (FeNa), oxygen consumption, and perfusion fluid levels of lactate dehydrogenase (LDH) and aspartate aminotransferase (AST). RFN, reperfusion.

FIGURE 8.

Histopathology and ischemia-reperfusion injury in kidney pairs having normothermic machine perfusion (NMP) or cold static storage (CS) followed by simulated transplantation. A, Representative photomicrograph (pair 2; DBD-D3) and cumulative comparison of renal cell death/apoptosis in both study groups as determined by TUNEL staining (×40). Similar immunofluorescence-based comparisons of (B) oxidative stress (using dihydroethidium [DHE] staining) (pair 3; DBD-D4), and (C) complement C9 staining (pair 2; DBD-D3) after ex vivo whole blood reperfusion (×20). D, Representative photomicrograph of a kidney pair (pair 2; DBD-D3) after simulated transplantation following either CS or NMP; Periodic acid-Schiff stain (×20). DBD, donation after brain death; RFN, reperfusion; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling.