Quality Assessment of Discarded Human Kidneys Using Normothermic Perfusion

Abstract

Background.

Ex vivo normothermic perfusion (EVNP) allows for comprehensive functional assessment; however, current scoring systems are validated only for short perfusion durations. This study investigates the utility of the EVNP quality assessment score for a 6-h period to predict viability in discarded human kidneys.

Methods.

Twenty human kidneys from marginal donors, previously declined for transplantation, underwent 6-h EVNP. Functional parameters (renal blood flow, resistance index, urine output), perfusate biochemistry, tissue injury biomarkers, and histology were assessed. Kidneys were stratified by EVNP score: group A (1–2), group B (3), and group C (4–5).

Results.

Group A demonstrated significantly higher renal blood flow (median 213.5 versus 145.1 [B] versus 81.7 mL/min [C], P < 0.0001), lower resistance index (0.32 versus 0.41 [B] versus 0.86 mL/min/mm Hg [C], P < 0.0001), greater urine output (241.1 versus 97.9 [B] versus 10.9 mL [C], P < 0.0001), and more favorable perfusate profiles (lactate: 14.2 versus 15.7 [B] versus 19.9 mg/dL [C], P = 0.02). Histological scores did not correlate with EVNP score groups at baseline or after perfusion. EVNP during 6 h showed trends consistent with functional graft performance, suggesting it may help inform decisions to reduce organ discard.

Conclusions.

The EVNP score may assist in identifying kidneys with potential for transplantation despite suboptimal histology, which could contribute to reducing the underutilization of marginal grafts.

INTRODUCTION

One of the current major challenges in kidney transplantation is the deteriorating quality of available deceased organs and the constantly growing gap between organ donor supply and demand.¹ Therefore, “marginal” grafts are nowadays commonly accepted for kidney transplantation as defined as grafts from donation after circulatory death (DCD) or from brain death donors (DBD) with expanded criteria donors.^1,2 There is existing evidence showing that these kidneys are more sensitive to ischemia/reperfusion injury (IRI) than standard criteria organs.^3-6 Consequently, many of these marginal kidneys are currently discarded because of concerns over organ quality. Therefore, the growing disparity between the demand for transplants and the available donor supply (as well as doubts concerning their quality) highlights the urgent need to develop technologies enabling to precisely assess kidney grafts and avoid organ underutilization.⁷

For these reasons, ex vivo organ perfusion strategies have seen an unprecedented interest in recent years and have gained a promising strategy: (1) to minimize IRI and its consequences (primary nonfunction and posttransplantation delayed graft function [DGF]),^8-10 and (2) in the decision process to discard the organs.

Currently, hypothermic machine perfusion is the most established ex vivo perfusion technique for deceased donor kidneys, and it has demonstrated improvements in posttransplant outcomes.^11,12 In contrast, ex vivo normothermic perfusion (EVNP) allows the restoration of kidney function by recirculating an oxygenated solution through the kidney at normal body temperature.¹³ Therefore, this technology provides a more precise metabolic evaluation and allows the assessment of the quality of a kidney to aid in the decision regarding its suitability for transplantation, even when the ex vivo time is prolonged.^13-18 Several assessment scores have been proposed during EVNP to predict the performance of a perfused kidney graft after transplantation. Among these, the Hosgood Score, introduced in 2015, is a concise tool for evaluating kidney grafts during EVNP and has been suggested as a potential adjunct for pretransplant assessment of marginal kidneys. The score classifies kidney grafts at 1 h of perfusion based on their macroscopic appearance, total urine output, and renal blood flow. However, it has only been validated for short perfusion periods (1 h).¹³ Such limited perfusion times may restrict the potential benefits of EVNP for kidney graft recovery and could contribute to higher discard rates if decisions are based on a score derived from short perfusion protocols^13,19

In this study, we report the outcomes of an EVNP program using discarded human kidneys developed at a tertiary center. We analyzed perfusion and biochemical parameters, as well as changes in markers of tissue injury, to assess whether the EVNP Hosgood Score at 1 h of perfusion correlates with kidney performance after a longer perfusion period (6 h), thereby assessing the potential value of this score in supporting kidney discard decisions.

MATERIALS AND METHODS

Human Discarded Kidneys

Human discarded kidneys from marginal donors (ie, expanded criteria donors and DCD) included in this study were retrieved initially for transplantation or research purposes. Kidney grafts were discarded according to the donor characteristics (including glomerular filtration rate), macroscopic evaluation, and high-grade kidney biopsy score (>4 according to Remuzzi Score²⁰). In accordance with the policy of our center, only DCD II kidney grafts underwent retrieval using normothermic regional perfusion. The study protocol was reviewed and approved by the Research Ethics Committee of our center (HCB/2021/1247). Written informed consent to participate in this study was provided by the patient’s relatives.

Normothermic Perfusion Set-up

The Ark kidney system (EBERS medical technology, Zaragoza, Spain) was used for ex vivo normothermic machine perfusion, without urine recirculation. The renal artery was cannulated with a 12-Fr or 16-Fr straight perfusion cannula; renal vein outflow was collected via a pump and directed to an oxygenator, whereby blood reentered the main circuit. The ureter was directly cannulated to the urine circuit, and urine was collected and monitored using a urine flow sensor. Instead of urine recirculation, replacement with a balanced solution (Isofundine, B Braun Medical, Barcelona, Spain) according to the urine output was conducted via a pump connected to the urine flow sensor. In addition to the urine outflow, the EVNP devices continuously monitor hemoglobin concentration, oxygen saturation, mean pressure, arterial flow, renal resistance index, and perfusate temperature. Perfusion was maintained for 6 h.

The EVNP protocol was similar to that 1 reported previously by our team (see also Supplemental Methods, SDC, https://links.lww.com/TXD/A811).¹⁸

Hemodynamics were pressure-controlled, establishing a pressure increase ramp until reaching a set mean arterial pressure of 70 mm Hg during perfusion. The temperature was set at 37 °C during perfusion. Oxygenation of the perfusate was performed by manual regulation of air (21% oxygen) at a flow rate of 1.5 mL/min to maintain an oxygen saturation level of >97% using an oxygenator. Biochemical parameters were measured hourly in the perfusate using the Epoc Blood Analysis System (Siemens Healthcare, Madrid, Spain) starting at the perfusion onset (T0).

EVNP Quality Assessment Score (Hosgood Score)

First described by Hosgood et al,^13,21,22 this score includes a macroscopic and functional assessment to provide a quantitative measure of functional parameters during the first hour of EVNP. The criteria score kidneys on a scale of 1–5 based on renal blood flow, total urine output, and macroscopic appearance, where 1 is the highest quality kidney and 5 is the poorest. Macroscopic appearance is graded, whereby grade I is a global pink appearance, grade II is a patchy appearance, and grade III is a global mottled and purple/black appearance (scored 1, 2, and 3, respectively). A urine output of <43 mL and a mean renal blood flow of <50 mL/mg/100 g both add 1 to the score.

The Hosgood Score was applied at 1 h of perfusion, consistent with its original description. This timepoint was selected because it allows for an early assessment of graft quality while providing sufficient time for the organ to stabilize. Kidney groups classified by the score at 1 h were then correlated with their subsequent performance, as evaluated through perfusate, biochemical, and histological parameters, during the following 6 h of perfusion. According to the initial score, kidneys were grouped as follows: group A, EVNP quality assessment score 1 or 2; group B, EVNP quality assessment score 3; and group C, EVNP quality assessment score 4 or 5.

Sampling Procedure

Perfusate samples were obtained hourly throughout the duration of perfusion. Hemodynamic perfusion parameters were recorded at corresponding timepoints.

Histological samples were collected 15 min after starting the perfusion (time point 0) and at 6 h. Two 1 × 1 × 1 mm tissue pieces of the kidney cortical were taken from the middle third of the graft. One biopsy was preserved in RNAprotect (Qiagen, Barcelona, Spain) for 24 h and then preserved at –80 °C until analysis. The other biopsy was fixed in 10% formal saline and then embedded in paraffin wax. Sections from the paraffin-embedded tissue were cut and stained with hematoxylin and eosin for histopathological scoring. A consultant pathologist graded the sections using the Remuzzi Score and assessed the level of acute tubular injury (ATI).²⁰ Sections were graded absent (0), mild (1), moderate (2), and severe (3) for the presence of ATI.

Real-time Quantitative Polymerase Chain Reaction

The samples were homogenized and total RNA was extracted using the Maxwell RSC Instrument (Promega, Madrid, Spain). Total kidney RNAs were extracted using the Maxwell RSC miRNA Tissue Kit (Promega, Madrid, Spain) in accordance with the manufacturer’s recommendations.

cDNA was synthesized from the RNA template using a cDNA synthesis kit from Invitrogen, according to the manufacturer’s instructions. The resulting cDNA was diluted and used to determine expression levels. The reference gene used was human β-actin, and the expression levels of kidney injury molecule 1 (KIM-1), caspase-3, vascular endothelial growth factor (VEGF), tumor necrosis factor α (TNF-α), transforming growth factor β (TGF-β), tissue inhibitor of metallopeptidase 1 (TIMP-1), β-catenin, and E-cadherin were measured.

Real-time quantitative polymerase chain reaction (PCR) was performed using the corresponding primers for each gene on a 384-plate using the PCR program provided by the supplier on a QuantStudio 7 device (Thermo Fisher Scientific, Waltham, MA). Samples were run in triplicate in 10 μL reaction volumes, and the mRNA expression of the target genes was normalized to β-actin mRNA and expressed as the fold change to time 0 (T0) kidney tissue using the 2-ΔΔCT (fold change) method.

Statistical Analysis

Quantitative data are presented as median (interquartile range [IQR] or whole range). Quantitative data were compared using either the Kruskal-Wallis test with the Dunn multiple comparisons test, the Mann-Whitney test, or the multiple Mann-Whitney tests with Holm-Sídák multiple comparisons test for nonnormally distributed data. Qualitative data are presented as number and proportion. Qualitative data were analyzed by the Fisher exact test or the chi-square test. A P value of <0.05 was considered statistically significant.

For real-time quantitative PCR, the mRNA expression of the target genes was normalized to β-actin mRNA and expressed as the fold change to time 0 (T0) kidney tissue using the 2-ΔΔCT (fold change) method. Graphical representation was conducted using S PRISM version 10.1.1 (GraphPad Software Inc, La Jolla, CA).

RESULTS

Donor and Kidney Baseline Parameters

Twenty kidneys were included. According to the EVNP quality assessment score, the macroscopic grades of kidneys were grade I (9), grade II (8), and grade III (3). Six kidneys had a renal blood flow below the threshold of 50 mL/min per 100 g and 16 had a total urine output of <43 mL during the first hour of perfusion. Therefore, the overall EVNP quality assessment scores were as follows: score 1 (n = 3), score 2 (n = 5), score 3 (n = 7), score 4 (n = 3), and score 5 (n = 2). The kidneys were grouped as follows: group A, EVNP quality assessment score 1 or 2 (n = 8); group B, EVNP quality assessment score 3 (n = 7); and group C, EVNP quality assessment score 4 or 5 (n = 5).

The donor and kidney baseline parameters are shown in Table 1. The median overall (IQR) donor age was 69.0 (55.0–78.3) y, without significant differences between all groups. Most (70%) of the donors were men. A majority (85%) of kidneys were from DCD donors. The median overall (IQR) serum creatinine level was 0.99 (0.86–1.20) mg/dL before organ procurement, with no significant differences between the groups (P = 0.90). The median cold ischemia time (CIT) was higher in group C without a statistical difference (P = 0.50). The most common reasons for decline were inadequate in situ perfusion (55%) and poor-quality organ at the biopsy (30%).

TABLE 1.

Donor characteristics and graft baseline parameters

	Group A (N = 8)	Group B (N = 7)	Group C (N = 5)	P
Donor’s characteristics
Donor age, y, median (IQR)	55.0 (40.0–79.0)	76.0 (55.0–88.0)	69.0 (44.5–74.0)	0.7
Donor male, n (%)	6 (75.0%)	5 (71.4%)	3 (60.0%)	0.8
Donor type, n (%)
uDCD	4 (50.0%)	1 (14.3%)	4 (80.0%)	0.2
cDCD	3 (37.5%)	4 (57.1%)	1 (20.0%)
DBD	1 (12.5%)	2 (28.6%)	0 (0%)
Cause of death, n (%)
CVA	4 (50.0%)	2 (28.6%)	4 (80.0%)	0.2
Stroke	4 (50.0%)	5 (71.4%)	1 (20.0%)
Previous medical history, n (%)
High blood pressure	6 (75.0%)	5 (71.4%)	3 (60.0%)	0.8
DM	2 (25.0%)	2 (28.6%)	4 (80.0%)	0.1
Cardiovascular disease	2 (25.0%)	4 (57.1%)	2 (40.0%)	0.4
Serum creatinine, mg/dL, median (IQR)	1.0 (0.9–1.3)	1.0 (0.9–1.1)	1.0 (0.9–1.2)	0.9
Graft baseline parameters
Cold ischemia time, h, median (IQR)	12.5 (9.8–17.8)	14.0 (11.5–15.1)	17.0 (12.0–25.0)	0.5
Right-sided graft, n (%)	5 (62.5%)	4 (57.1%)	3 (60.0%)	0.9
Organ preservation solution, n (%)
IGL-1	4 (50.0%)	3 (42.9%)	2 (40.0%)	0.4
Celsior	2 (25.0%)	3 (42.9%)	0 (0%)
HTK-Custodiol	2 (25.0%)	1 (14.3%)	3 (60.0%)
Reason for discard, n (%)
Inadequate in situ perfusion	5 (62.5%)	2 (28.6%)	4 (80.0%)	0.3
Donor previous medical history of metastatic prostate cancer	0 (0%)	2 (28.6%)	0 (0%)
Poor organ qualitya	2 (25.0%)	3 (42.9%)	1 (20.0%)
Kidney tumor on contralateral kidney	1 (12.5%)	0 (0%)	0 (0%)
No. of arteries, n (%)
1	6 (75.0%)	6 (85.7%)	4 (80.0%)	0.5
2	1 (12.5%)	1 (14.3%)	0 (0%)
3	1 (12.5%)	0 (0%)	0 (0%)
4	0 (0%)	0 (0%)	1 (20.0%)
No. of veins, n (%)
1	8 (100%)	7 (100%)	5 (100%)	0.9

^aRemuzzi Score: 4 or 5.

cDCD, controlled donation after circulatory death; CVA, cardiovascular arrest; DBD, donation after brain death; DM, diabetes mellitus; HTK, histidine-tryptophan-ketoglutarate; IGL-1, Institut Georges Lopez-1; IQR, interquartile range; uDCD, uncontrolled DCD.

Perfusion Parameters and Urine Production

The perfusion parameters are represented in Table 2 and Figure 1. Because the Ark kidney system is pressure-controlled, the median arterial pressure was similar in the 3 study groups. Kidneys in group A had improved perfusion parameters compared with groups B and C. The median renal blood flow was significantly higher in group A compared with groups B and C (213.5 versus 145.1 versus 81.7 mL/min in group A, B, and C, respectively, P < 0.0001). Correspondingly, the median renal resistance index was significantly lower in group A compared with group B and C (0.32 versus 0.41 versus 0.86 mL/min/mm Hg, P < 0.0001). At the end of EVNP, the median urine output was significantly higher in group A compared with groups B and C (241.1 versus 97.9 versus 10.9 mL, P < 0.0001; Figure 1D; Table S1, SDC, https://links.lww.com/TXD/A811).

TABLE 2.

Perfusion parameters and perfusate characteristics

	Group A (N = 8)	Group B (N = 7)	Group C (N = 5)	P
Perfusion parameters
Arterial pressure, mm Hg	69.8 (61.0–71.2)	69.1 (59.4–73.9)	69.9 (68.8–70.6)	0.9
Arterial flow, mL/min	213.5 (108.3–259.8)	145.1 (56.9–241.4)	81.7 (55.0–108.1)	<0.0001
Resistance index, mL/min/mm Hg	0.32 (0.27–0.61)	0.41 (0.26–0.58)	0.86 (0.64–1.23)	<0.0001
Perfusate characteristics
pH	7.50 (7.20–7.70)	7.39 (7.14–7.65)	7.41 (7.28–7.58)	0.8
Lactate, mg/dL	14.2 (10.5–19.3)	15.7 (10.7–19.6)	19.9 (14.0–20.0)	0.02
Arterial pO2, mm Hg	234.5 (88.3–435.5)	127.6 (79.4–215.1)	292.5 (111.9–344.0)	0.03
Arterial pCO2, mm Hg	5.0 (5.0–18.0)	15.80 (5.0–22.75)	5.0 (5.0–5.0)	0.0004
Hemoglobin, g/dL	8.0 (6.7–10.2)	6.7 (6.3–7.6)	6.7 (6.4–7.1)	<0.0001
Hematocrit, %	24.7 (20.7–31.4)	20.9 (19.5–23.4)	20.9 (19.7–22.0)	<0.0001
Sodium, mmol/L	151.0 (138.0–153.3)	147.0 (139.5–152.0)	144.0 (137.8–148.0)	0.3
Potassium, mmol/L	5.7 (3.5–7.3)	4.9 (4.0–6.1)	8.5 (6.3–9.1)	<0.0001
Calcium, mmol/L	1.1 (0.8–1.3)	1.1 (0.9–1.2)	0.9 (0.7–1.0)	0.002
Chloride, mmol/L	109.5 (101.5–115.8)	113.0 (103.5–117.0)	106.5 (99.8–114.3)	0.3
Glucose, mg/dL	80.0 (57.0–107.5)	83.0 (55.0–95.5)	107.0 (69.0–130.0)	0.049
Urine output at T6, mL	241.1 (0.0–801.6)	97.9 (23.5–896.3)	10.9 (6.5–54.6)	<0.0001

Data are expressed as median (interquartile range). Median values were calculated for each parameter across all measurements obtained during the 6-h perfusion.

FIGURE 1.

Perfusion parameters and urine production. A, Median arterial pressure in the 3 study groups. B, Median renal blood flow. C, Median renal resistance index. D, Median urine output. Data are expressed as median and interquartile range.

Perfusate Analysis

The perfusate characteristics are presented in Table 2 and Figure 2. The median perfusate pH was similar across the study groups, whereas the median perfusate lactate was significantly lower in group A compared with groups B and C (14.2 versus 15.7 versus 19.9 mg/dL in groups A, B, and C, respectively, P = 0.02). Regardless of using the same number of packed red blood cell units and similar T0 values of hemoglobin and hematocrit, the median perfusate hemoglobin and hematocrit were significantly higher in group A compared with groups B and C. Similarly, despite a uniform basic perfusate formulation and values at perfusion onset, kidneys exhibited different electrolyte levels in the perfusate, particularly as perfusion progressed. The median perfusate potassium was significantly lower in groups A and B compared with group C (5.7 versus 4.9 versus 8.5 mmol/L, P < 0.0001), whereas median perfusate calcium was significantly higher in groups A and B compared with group C (1.1 versus 1.1 versus 0.9 mmol/L, P < 0.0001), although these differences were not observed when hourly values were compared (Figure 2G). The median perfusate sodium and chloride were similar in all groups.

FIGURE 2.

Perfusate characteristics. A, Median perfusate pH. B, Median perfusate lactate. C and D, Median perfusate hemoglobin and hematocrit. E, Median perfusate sodium. F, Median perfusate potassium. G, Median perfusate calcium. H, Median perfusate chloride. I, Median perfusate glucose.

Tissue Injury Markers

The tissue injury markers are represented in Figure 3. The expression of acute kidney injury (KIM-1), apoptosis (caspase-3), inflammation (VEGF, TNF-α, TGF-β), and proliferation and repair (TIMP-1, β-catenin, E-cadherin) markers to assess tissue response was analyzed. The differences are not significant, although KIM-1 expression was higher in group A after 6 h of EVNP (Figure 3A). Caspase-3 expression was similar in all groups, with a lower level in group C (Figure 3B). Concerning inflammation markers, although the differences are not significant, TGF-β expression was higher in group B, TNF-α expression was higher in group A, and VEGF expression was higher in group C after 6 h of EVNP (Figure 3C–E). Expressions of TIMP-1, β-catenin, E-cadherin, and Ki67 were similar in all groups (Figure 3F–I).

FIGURE 3.

Tissue injury markers and histology. A, KIM-1 expression. B, Caspase-3 expression. C, TGF-β expression. D, TNF-α expression. E, VEGF expression. F, TIMP1 expression. G, β-Catenin expression. H, E-Cadherin expression. I, Ki67 expression. J, Total Remuzzi score at T0 and T6 for the 3 groups analyzed. K, ATI grade at T0 and T6. L, Representative HE image of kidney tissue at T0 showing a mild to moderate ATI. M, Representative HE image of kidney tissue at T6 showing a mild to moderate ATI, with no changes compared with T0. ATI, acute tubular injury; HE, hematoxylin and eosin; KIM-1, kidney injury molecule 1; TGF-β, transforming growth factor beta; TIMP-1, tissue inhibitor of metallopeptidase 1; TNF-α, tumor necrosis factor alpha; VEGF, vascular endothelial growth factor.

Histological Assessment

Kidney injury histological assessment by Remuzzi Score was not significantly different among the 3 groups at the same timepoints, with a median of 3 for the 3 groups at T0 and T6, except for group A, in which the median at T6 was 2 (Table S2, SDC, https://links.lww.com/TXD/A811). When comparing T0 to T6 scores for the same group, no significant differences were observed; also, a tendency for improvement in the score values was identified (Figure 3J). ATI was similar among the different groups, being mild to moderate in all cases (Table S2, SDC, https://links.lww.com/TXD/A811). No differences were observed after 6 h of perfusion compared with T0 biopsy for any of the groups assessed (Figure 3K–M).

DISCUSSION

Ex vivo kidney graft preservation strategies have become a promising strategy to minimize IRI with the aim of reducing the risks of primary nonfunction and posttransplant DGF.^8-10 Among them, EVNP has arisen as a platform to precisely perform a metabolic and functional evaluation of the kidney graft, thus aiding in the decision regarding its suitability for transplantation and reducing graft underutilization.^13-17

In 2015, Hosgood et al proposed an EVNP quality assessment score to predict the performance of a kidney graft before transplantation. However, it was only evaluated for a short perfusion period (1 h); a relatively brief perfusion time may limit the benefits of EVNP for kidney graft recovery and could lead to higher discard rates if decisions are based on a score derived from short perfusion protocols.^13,21,22 In this context, a recent randomized controlled trial reported that the Hosgood Score did not correlate with DGF after transplantation, contrasting with the findings of the original study.^13,19 It should be noted, however, that perfusion was limited to 1 h, which may be less beneficial than longer perfusion durations.^18,23 This shorter setting could have reduced the potential advantages of EVNP in graft recovery, possibly explaining the lack of a strong correlation between the score and DGF rates. Nevertheless, the strength of the Hosgood Score may lie in its ability to identify kidney grafts with a high risk of primary nonfunction.^13,21,22 Overall, we considered that its simplicity of application made it a useful tool for initial assessment in a longer perfusion scenario. Moreover, the Hosgood Score may provide early indications suggestive of primary nonfunction, although more data are needed to confirm its predictive accuracy.

Therefore, to further evaluate the performance of the Hosgood Score beyond short perfusion protocols, we examined whether it could anticipate kidney graft performance after 6 h of EVNP. In addition, we correlated the score with biochemical, perfusate, and histological assessments to explore its potential value in a prolonged perfusion setting. Irrespective of donor characteristics, reason for discard, and CIT, stable EVNP was achieved for the whole perfusion duration. Our results are consistent with those previously published, although in our case, CIT was similar among the 3 groups, and donor age was higher compared with that previously reported, which locates our evaluation and EVNP score in a nearer clinical setting.¹³ Furthermore, our analysis correlates the EVNP score at 1 h of perfusion with biochemical, molecular, histological, and hemodynamic monitoring results of discarded kidney grafts for 6 h of perfusion. We observed improved perfusion parameters and urine output in kidneys with a lower overall EVNP score during a longer perfusion time (6 h). Noticeably, groups A and B, classified as higher-quality grafts according to the Hosgood Score, exhibited a slight decrease in arterial flow after 3 h, although changes in the arterial resistance index remained minimal. This pattern is in line with that previously reported in prolonged kidney EVNP and likely represents a physiological response to IRI and the cytokine surge that has been reported to occur after 60–120 min of EVNP.^24-27 Such a response may predominantly affect initially less-inflamed kidneys, triggering a positive inflammatory reaction in metabolically active tissue and resulting in a transient reduction in arterial flow. In contrast, the continuous rise in arterial flow observed in group C, which includes lower-quality grafts, may reflect an absence of inflammatory response in tissue already affected by severe dysfunction and extensive cell death.

Kidneys with lower EVNP scores also demonstrated a more favorable perfusate biochemical profile, including lower lactate and more stable pH. Perfusate potassium levels were significantly lower in groups A and B compared with those in group C, which had the low-quality kidneys. Given that the perfusate composition was identical across experiments and none of the replacement solutions contained potassium, these findings may reflect differences in excretory capacity between groups. Specifically, the lower potassium concentrations observed in the higher-quality grafts (groups A and B) are likely related to a preserved ability for potassium clearance, whereas the accumulation observed in group C suggests impaired excretion and ongoing dysfunction.

When considering calcium values, kidneys with higher EVNP scores showed slightly higher ionic calcium levels compared with the lower scores. Although the differences were small and not evident in hourly comparisons, they may reflect a greater degree of calcium dysregulation in lower-quality grafts. Intracellular and mitochondrial calcium overload is a well-recognized mediator of IRI, and the longer CIT observed in group C may have impaired ATP-dependent ion pump function, exacerbating intracellular calcium accumulation. After reperfusion, this process could lead to further calcium influx into damaged cells, thereby reducing perfusate calcium levels in group C and indicating more severe tissue injury.^28,29

Histologically, the Remuzzi Score was similar within the 3 groups established according to the EVNP score. This finding suggests that histological evaluation does not necessary correlate with kidney viability and may lead to graft underutilization.²¹ In our study, to analyze the predictive capacity of the initial EVNP score, we performed a biopsy at the end of perfusion. A slight reduction in the Remuzzi Score was evidenced after 6 h of perfusion for all groups compared with their initial biopsy, although no statistically significant difference was reached. This improvement may reinforce the ability of the EVNP score to identify kidney graft viability.

Overall, given the observed correlations between biochemical, perfusate, and histological changes in the kidney groups classified according to the initial Hosgood Score, the accuracy of the score may improve if biochemical parameters (such as lactate, potassium, and calcium) as well as the histological assessment were incorporated. This could complement the evaluation provided by the score and enhance its discriminative capacity. In this context, based on an EVNP score <4 and 6-h perfusion monitoring, it is possible that up to 15 of 20 discarded kidney grafts (75%) might have been potentially suitable for transplantation, although this estimate should be interpreted with caution because no transplantation was performed.

When analyzing different tissue injury markers, KIM-1 expression was higher in kidneys with lower overall EVNP quality assessment scores, although some series have demonstrated that this marker does not depict a good correlation across the different EVNP grades and is not associated with donor acute kidney injury.^22,30 Inflammation markers (VEGF, TNF-α, and TGF-β) were increased in the different groups, with an enhancement of the inflammation-fibrosis pathway in all groups, probably due to the intrinsic characteristics of each kidney, especially considering the relatively long CIT associated with the analyzed discarded human kidneys.

There is limited evidence defining the optimal duration of EVNP for kidney graft assessment. Some studies suggest that short perfusion periods may be less beneficial than prolonged perfusion, whereas others have shown that brief periods of EVNP are sufficient to restore depleted cellular ATP levels without inducing additional inflammatory damage over time, which can be exacerbated by hemolysis from a blood-based perfusate.^{18,23,26,31,32} Our rationale for performing 6 h of perfusion is based on previous studies, as this duration is considered sufficiently prolonged to provide a meaningful assessment while remaining logistically feasible and avoiding excessive kidney graft inflammation.^27,32 Nevertheless, because the most significant changes in perfusate characteristics between kidney groups emerged after 2 h (particularly regarding biochemical parameters), a shorter perfusion time (eg, 3–4 h) might be adequate to evaluate the potential quality of perfused kidneys. Further studies are warranted to determine the optimal perfusion duration for assessing kidney quality.

Our study has some limitations. Therefore, while using the EVNP quality assessment score, the huge variety in organ quality and donor characteristics makes it difficult to compare tissue injury markers. In addition, because those kidneys were previously discarded and not transplanted, no correlation to clinical outcome can be made. Finally, a significant limitation of our study is the absence of a functional clearance readout, such as creatinine or inulin clearance. Due to technical and logistical constraints, these measurements could not be performed, and graft functionality was instead evaluated using measurable perfusion and biochemical parameters, along with histology and urine output.

CONCLUSIONS

Our study provides more data on the performance of the EVNP score during a 6-h normothermic perfusion time, suggesting that perfusion of discarded human kidney combined with the EVNP quality assessment scoring system is an innovative technology to evaluate the graft quality and avoid organ underutilization. Based on our findings, a further validation of the Hosgood Score in combination with additional markers (such as perfusate biochemical and histological parameters) to improve its discriminative capacity is needed. Although our study suggests that integrating these markers could enhance the evaluation of graft quality during prolonged EVNP, we believe that further prospective studies are required before considering the score ready for application in clinical settings involving marginal or discarded grafts. Such studies would help establish its predictive value for graft function and its potential to safely guide decisions regarding kidney utilization.