Tim van Smaalen

Kidney transplantation remains the treatment of choice for patients with end stage renal disease. While patients die and suffer on the waiting list, transplantation provides the best chance of restoring a normal lifestyle. In 65 years of kidney transplantation there have been many medical, technical and logistic improvements which led to a tremendous increase in patient and graft survival. However, since the first human renal transplant, the same barriers still exist; organ shortage and graft failure. The ongoing shortage of kidneys and the limitations to perfect functional outcome warrant studies which may aid in overcoming these problems. In this thesis “Marginal donor kidneys; novel strategies to assess and improve viability” we have employed novel strategies to assess and improve viability. Good renal allograft viability is a prerequisite to perform transplantation of a kidney graft. However, the definition of renal viability is subjective, criteria unclear, and different per country, transplant centre or doctor. It is still impossible to exactly determine a kidney's viability and thus, transplant outcome. Evidently, kidneys with minor injury and of ideal donors (young healthy living donors) will likely function after transplantation. Therefore, we have focussed in this thesis on a group of kidneys of which it is hard to estimate viability and transplant outcome: marginal donor kidneys. These are kidneys from patients of old age and/or with increased morbidity or kidneys which have sustained substantial (ischemic) injury. It is currently almost impossible to predict transplant outcome for marginal donor kidneys. Thus, choosing which marginal kidney can be safely transplanted remains difficult and a struggle for transplant clinicians. Although more patients may receive a transplant kidney when a liberal acceptance strategy for marginal kidneys is used, it may also lead to poor transplant outcomes including post-transplant graft failure. Patients are unnecessarily exposed to risks of surgery, immunosuppression, sensitization of donor antigens or even death. On the other hand, a conservative strategy may lead to unnecessary discard of possibly viable kidneys and a smaller number patients will be transplanted. This results in an increase in waiting time and an increase of patients suffering and dying on the waiting list. ‘Primum non nocere – First, to do no harm (Hippocrates)’ But how do we do no further harm, when we do not know what to choose? In this thesis we aim to get insight in the pathophysiological processes of kidney transplantation and which factors influence viability and transplant outcome, so that transplant clinicians may have tools to make a well informed decision. To do so we have performed translational studies, using both human and animal models. Furthermore, we tried to translate our results from the laboratory to the clinic and the other way around, to explain wat we have observed in a clinical situation. The marginal kidney donor To expand the donor pool, kidneys of marginal quality have been increasingly used. The introduction of donation after brain death (DBD) donors, the use of expanded criteria donors (ECD) and the increasing use of donation after circulatory death (DCD) donors have shown to be successful strategies with improvement of graft and patient survival. There is a general consensus that, because of the survival benefit and positive effect on the waiting list, kidneys from these donors should be used for transplantation. Unfortunately, there is no general consensus on the use of marginal donor kidneys, but there are studies, which define those kidneys which are acceptable for transplantation and those that are not. Marginal donor kidneys includes a very large and heterogenous group of kidneys. These include kidneys from old living donors or those with comorbidities, such as obesity, hypertension, microalbuminuria or borderline-normal glomerular filtration rates. Furthermore, they include DBD and DCD donors with comorbidity and of old age or DCD donors themselves may be seen as marginal quality donors, especially uncontrolled donors, because of the large amount of warm ischemic injury. Uncontrolled donation after circulatory death Maastricht category 1 and 2 DCD donors, uncontrolled DCD donors (uDCD), have been part of ongoing discussion whether they should be part of transplant programs throughout the world. In Europe, there is a variation in DCD protocols and not all countries have a uDCD program. In uDCD, after unsuccessful resuscitation, organ perfusion is promptly initiated by either cold or normothermic perfusion. The process of unsuccessful resuscitation and prompt recirculation for organ preservation purpose has been subject of ethical discussion and an argument for procurement organizations not to implement uDCD programs. Furthermore, logistic difficulties make it hard to implement uDCD programs. It is advised that, when implementing a uDCD program, strict national and regional protocols are needed. Although there may be (some) reluctance towards uDCD, outcome has proven to be acceptable. A study of our own cohort in 2011 showed similar results comparing uDCD and DCD in terms of primary non-function (PNF): 22% vs. 21%, respectively) and delayed graft function (DGF, 61% vs. 56%, respectively). Ten-year graft and recipient survival were also similar in uDCD and DCD groups (50% vs 46% and 61% vs 60%, respectively). The Madrid group reported similar results in the largest uDCD group ever reported (N=774). Ten-year survival rates for uDCD were better than those of ECD DBD donors (72% vs 61%, p=0.02, respectively). The main cause of graft loss was PNF of which donor death by pulmonary embolism, extra-hospital cardiopulmonary resuscitation (CPR) ≥75 min and in-hospital CPR >50 min were predictive factors. The results of multiple studies were recently reviewed in a systematic review and meta analysis and showed that uDCD is inferior in the short-term, but outcomes are comparable to controlled DCD in the long-term. In this thesis, in Chapter 2, we show that for uncontrolled DCD, age was the only risk factor for PNF and strongly associated with graft failure, while resuscitation times were not recorded. The differences between our results and those in Spain are a good example of the heterogenicity of protocols and laws regarding uDCD, determination of death and transport of deceased patients. Furthermore, the mean donor age in Madrid was only 40 years, while this is a risk factor in our own cohort, especially for the group older than 54 years. Unfortunately, while general uDCD results are good, the Maastricht group had to abandon their uDCD protocol, as it is costly, time consuming, requires a lot of expertise and effort and uDCD donors rates were continuously low. A promising intervention that may increase the use of uDCD donors is the introduction of a new in situ preservation technique, normothermic regional perfusion (NRP). This is in line with the increasing use of normothermic perfusion in organ preservation. NRP uses extra corporal membrane oxygenation techniques to perfuse and resuscitate organs in situ after circulatory death while at the same time the brain is excluded from perfusion. The number of clinical studies is still low, but is increasing. A study 2019 reported a cohort of 499 kidneys transplantations of which 50% were preserved with NRP. In situ preservation with a double-balloon triple lumen catheter (DBTL) resulted in a higher risk of PNF (OR 4.5, p=0.015) and worse renal function after one year compared to NRP. The Spanish also showed that NRP led to superior results compared to in-situ preservation, with an increased risk of graft loss (OR 5.6; 95% CI 2.7-11.5) after in-situ preservation during the first year after transplantation. Because of the promising results, NRP has now also been used in controlled DCD donors with success. Next to impact on graft quality, NRP has the advantage of quality assessment during perfusion, already in the donor. Hemodynamic parameters and markers of injury, inflammation, haemostasis, coagulation perturbation may be measured to estimate graft viability. When Maastricht decided to stop their uDCD program, the use of uDCD in the Netherlands came to a halt. This is unfortunate, because transplant waiting lists tend to grow and studies show favourable outcome. The reluctance towards uDCD is attributable to the uncertainty of its cost-effectives as well as the clinical doubt in a good outcome, which can be very subjective and differs per team or doctor. However, the technical and clinical developments in extra-corporal membrane oxygenation (ECMO) in general, may lead to new insights regarding the use of uDCD in the Netherlands. First, the clinical use of ECMO (or similar techniques) is increasing and the array of indications to use ECMO has broadened. Thus, there is an increasing number of medical personnel with expertise and training programs are successfully developed to facilitate efficient use of oxygenation machines and proper management of organ or donor viability. Because uDCD programs require a specialist team to be available 24/7, the increase of trained staff may lower the burden for centres who are willing to (re)start a uDCD program. Second, the increasing use of ECMO and ex-vivo perfusion of all transplantable organs, leads to a growing pool of evidence in favour of its clinical use. Furthermore, the use of these techniques in multiple settings and for multiple organs, leads to cross-pollination of knowledge and confidence therein. Hopefully, this may lead to the renewed and successful use of uDCD in the Netherlands in the future. In addition, viability assessment may contribute too, as it may be used as final tool to decide whether or not to transplant an organ. Therefore, studies on viability assessment are warranted. Viability assessment Viability assessment of donor kidneys has been performed for many years, with the aim to predict post-transplant graft function and survival and thereby simplifying the decision to accept or decline a donor kidney. Furthermore and theoretically, it may provide valuable knowledge on a graft quality, so that to recipient management can be customized per case. Also, the obtained knowledge regarding graft quality and especially graft injury may be used to develop new interventions to prevent further injury or even targeted resuscitation strategies. Unfortunately, while viability assessment theoretically has great potential and seems harmless (most test are non-invasive), its clinical implementation has been difficult. Not only did studies show that many viability tests have a poor predictive value on transplant outcome, it was also argued that assessment may have negative affect on the waiting lists. Improper use of viability assessment may lead to unnecessary discard of organs which may aggravate the organ shortage. Furthermore, viability test are time consuming and may lead to delay and increased detrimental ischemia times. Viability assessment has most often been performed by means of machine perfusion viability testing, histologic procurement biopsy assessment and donor risk scores. In chapter 3 we have reviewed the current methods of viability assessment during machine perfusion and come to a similar conclusion as many. There were (and still are) no singular viability tests which have a desired predictive value to be of clinical use. This repetitive negative conclusion has always been accompanied by another statement, which is the need for further studies towards the understanding of the mechanisms involved in graft failure and the need for new techniques to improve prediction accuracy. We share the same conclusions and have therefore performed multiple studies addressing mechanisms of kidney injury during the transplant process and viability assessment. Molecular mechanisms of graft injury differ in DCD and DBD In chapter 1 the differences between DCD and DBD were explained in terms of types, categories and their impact on donor logistics and organ perfusion. To get a better understanding of kidney viability, it is of importance to understand the underlying mechanisms that are associated to the development of graft failure. They are complex, because both donor and recipient factors contribute to the decay in graft viability. These mechanisms include immunological and non-immunological processes and differ per type of deceased donation, i.e. DCD vs. DBD. Several clinical studies have shown that the pathophysiological processes of organ injury and repair differ for DCD and DBD. This was foremost shown by the finding that graft survival was not affected by DGF in patients who received a kidney from DCD donors, while it was affected by those who received kidneys from DBD donors. Furthermore, although DCD kidneys have twice the risk on DGF, long-term graft survival is equivalent to DBD kidneys. These clinical differences reflect a variation of pathologic processes during the process of kidney transplantation and thereafter. Although DBD is preferred over DCD due to presumed better organ quality, several specific detrimental processes have been attributed to DBD. Brain death is induced by increased intracranial pressure and herniation of the brain stem, which triggers the release of catecholamines and increases parasympathetic activity. The negative impact on hemodynamic donor stability will result in decreased organ perfusion and the need for fluid resuscitation and inotropic support in the donor, with negative consequences for the grafts. Furthermore, the impaired function of the hypothalamus and pituitary gland causes hormone imbalance in the donor, such as decreased cortisol, ADH, T3 and insulin release. Brain death may induce an inflammatory response through several mechanisms. The ischemic brain releases inflammatory mediators and catecholamine release induces aerobic metabolism or causes shear stress of the endothelium and cytokine release. Changes in hormones, like insulin, are directly related to the inflammatory response. The pathophysiologic changes after brain death have detrimental effect on kidneys. Circulation of cytokines (TNF-α and interleukins), chemokines and adhesion molecules (VCAM-1, ICAM-1, E-selectin, P-selectin) promote leukocyte influx in kidney grafts, which is aggravated after reperfusion. Histological examination of kidneys after brain death showed arterial intimal proliferation and degenerative changes, such as vacuolar degeneration, atrophy and necrosis, especially in distal tubuli. A study on the effect of brain death duration on kidney graft function surprisingly showed that prolonged brain death duration may be somewhat protective. This may indicate that after the direct changes of brain death a resuscitative period may arise in the donor. Indeed, in prolonged brain death cytoprotective genes are upregulated. These findings instigate the implementation of specialized donor management to adequately preserve organs in the donor. For DCD kidneys, the cessation of blood after circulatory arrest has both advantages and disadvantages. As compared to DBD, the absence of circulation prevents toxins and immunoactive compounds to circulate through the body. However, traumatic injury, cardiovascular resuscitation and brain injury may also lead to detrimental effects on kidneys in DCD donors, similar to DBD, such as an inflammatory response. Furthermore, an agonal phase in controlled donors may lead to irreversible inflammatory and ischemic injury. For this reasons there are strict protocols as to under which conditions an agonal phase may exist. The high risk of DGF and worse short-term graft survival in DCD kidney transplantation has mostly been attributed to warm ischemic injury (chapter 1, Figure 1.1). Ischemia-reperfusion injury (IRI) was discussed in depth in chapter 1 and further explains the biological difference between DCD to DBD kidneys. Thus, the biological differences between DCD and DBD kidneys are significant, the understanding and awareness of these principles of importance and thus they should be further studied. Not only do these differences lead to a different transplant outcome, they change the way on how we should treat donors, employ preservation strategies and treat kidney recipients. As consequence, we may need to revise our thoughts regarding viability assessment of deceased donor kidneys and perhaps employ different strategies for different kidney types in the future. Moreover, with the development of molecular based viability assessment, as we will discuss later, this may be of special interest. Novel viability assessment strategies in human renal transplantation The difficulties with the current viability assessment strategies is highlighted in Chapter 4, as we discuss the controversies of the use of procurement and implantation biopsies. In Maastricht, colleague Snoeijs studied the role of histologic assessment of pre-implantation biopsies with the use of the Remuzzi score and Banff classification. He concluded that there is an association with kidney function and graft survival in DCD donors older than 60 years, although the predictive value on graft loss was poor. For this older group it was suggested that careful histologic assessment may improve donor selection. Thereafter, the use of pre-implantation biopsies decreased, due to the risk of discarding kidneys which would otherwise have functioned. Therefore, biopsies are currently only taken on indication. One of the difficulties of histologic assessment is the interobserver variability, which can be lowered when there is a special team trained for histopathologic analysis. Also, there is discussion as to which technique (needle vs wedge biopsy) is best. However, optimizing quality by implementation of a special ‘donor kidney analysis team’ may present clinical difficulties due to the acute setting of transplantation and the need for on-call pathologists 24/7. The use of a first digital histopathology service in the United Kingdom is currently studied and may overcome such problems. In addition and to anticipate on future developments, the first papers have been published to address the role of artificial intelligence in renopathology and transplantation. They state that artificial intelligence has the potential to significantly improve transplant outcome. In chapter 7 we have used mass-spectrometry imaging (MSI) assisted molecular assessment in combination with digital histopathologic examination. The additional value of these techniques should be further studied in a clinical setting. Furthermore, studies on pre-transplant donor kidney biopsies are limited and include retrospective studies with a high risk of selection bias and prospective clinical studies are lacking. Thus, there is debate on the quality of evidence based selection and decision making and the use of biopsies in clinical situations. Studies on implantation biopsies (at the time of reperfusion) are even more limited and of questionable quality. Post-reperfusion biopsies have a different meaning, as they are not used for the decision to accept or decline a kidney for transplantation. However, they may show changes after reperfusion, like neutrophil influx and alloimmune processes, as compared to biopsies before reperfusion. These insights may aid in the development of targeted therapy of a donor organ or recipient or lead to identification of viability markers. One implantation biopsy study on short-term function showed that arteriolar hyaline thickening and interstitial fibrosis were the only factors associated with outcome in multivariate analyses. In addition, the combination of a histologic score and donor risk score may have the best predictive value on graft loss so far. Although sceptics on procurement and implantation biopsies argue their uselessness and the recent published British guidelines state that “None of the kidney transplant biopsy scoring systems - alone or in combination - have sufficient predictive value to mandate organ discard”, Angeletti et al. stated that the lack of good quality studies should prompt more studies aimed at optimizing the strategies to score renal samples. In chapter 4 we have aimed to broaden the tools for histopathologic examination of kidney biopsies as regularly performed. We introduced a set of histologic markers, related to IRI, in contrast to the regularly performed analysis according to the Banff criteria and show that these are associated to DGF. As such, they may be potential predictive markers, but prospective randomized studies need to be performed. The increased presence of histologic IRI parameters in chronically injured donor kidneys also indicates a higher susceptibility to acute injury. These findings are in line with clinical studies, which show that ECD kidney transplantation has worse outcome than those of ‘regular’ DCD or DBD kidneys. Furthermore, this emphasizes that the increased susceptibly of marginal donor kidneys to IRI warrants the need for resuscitative strategies to prevent them from further injury, like NRP or normothermic machine perfusion (NMP). Next we focussed on viability assessment by the use of biomarkers during machine perfusion. As reviewed in chapter 3, current biomarkers are associated with transplant outcome but the predictive power is unsatisfactory. In collaboration with the department of biochemistry we studied the association of extracellular histones and transplant outcome in an unselected cohort of 390 DCD kidneys in Chapter 5. We showed that extracellular histones were independent risk factors for DGF and 1 year graft failure. The predictive value on DGF was equivalent to the predictive value of most other machine perfusate biomolecules, thus we had to conclude that this marker (also) cannot be used as stand-alone clinical predictive marker of outcome. Extracellular histones had not yet been studied in the pathophysiologic process of kidney injury and organ viability in a clinical setting of organ donation and transplantation before. Therefore, these findings show that histones reflect organ injury and further studies may test whether extracellular histones can be of value in viability assessment. Furthermore, because our study was performed with DCD kidneys, further studies should also focus on DBD kidneys. Extracellular histones are known to be cytotoxic and could have a different effect on DBD donors and kidneys, while cardiovascular perfusion is intact and renal injury ongoing. Also, because histones are associated with cell death and plasma membrane rupture, they may be of value for monitoring the quality of organ resuscitation after procurement. This prompted us to go back to the laboratory and perform additional tests in an animal model to get insight in the kinetics of extracellular histone release after renal injury. Novel viability assessment strategies in an ex-vivo porcine model In chapter 6 and 7 we developed a new perfusion model, using porcine kidneys from the slaughterhouse which, after death, were flushed several times so that all intravascular content (proinflammatory molecules and toxins) was removed. Thereafter, we were able to perform paired kidneys analysis comparing kidneys with severe and mild ischemic injury. The kinetics of extracellular histones release were studied in chapter 6. We showed that histones are exponentially released over time and depend on the amount of ischemic injury. The used techniques were able to measure histones 1 hour after perfusion. Additionally, in vitro tests showed that histones are more cytotoxic to renal cells in increased temperature and concentration. Thus, ischemia alone may lead to cytotoxic histone release, which, especially in a warm environment, may trigger a vicious circle of cell death, increased release of cytotoxic molecules and loss of graft quality. These findings confirm that warm ischemia should be minimized. Furthermore, during intact circulation in the donor or during NMP, extracellular histones can cause renal injury, therefore it may be of interest to either minimize the cytotoxic effects or decrease their release. The Maastricht biochemistry department earlier showed that heparin can neutralize cytotoxic histones both in vitro an in vivo and Sedigh et al. showed beneficial effects of a heparin conjugate in ischemically injured kidneys. Therefore, we tested whether the addition of non-anticoagulant heparin would decrease histone release. Addition of heparin did not change the total histone H3 levels. However, we did observe a continuous increase in the level of a lower molecular weight histone H3 variant, which we presume is an effect of increased protease activity by histone-heparin binding. All in all, histone release is directly related to ischemic kidney injury and further studies should be performed to explain their contribution to renal injury and ways to neutralize their adverse effects in order to optimally preserve transplant kidneys. In collaboration with The Maastricht Multimodal Molecular Imaging Institute (M4I), Division of Imaging Mass Spectrometry, we attempted to use mass spectrometry imaging (MSI) for the purpose of organ quality assessment, which had never been done before (Chapter 7). This was made possible by the introduction of a new type of imaging mass spectrometer, which allows to assess spatial distribution of biomolecules 20 to 50 times faster than other commercial instrumentation available at the time. Therefore, high-throughput analysis of samples was possible and moreover, because of the speed, it becomes possible to perform tissue analysis in a clinical acceptable time frame for viability assessment. Thus, we used MSI to identify ischemic injury between mild and severely injured pig kidneys. While our first tests, focussing on proteins were not successful due to the difficulty to assess and identify molecules with high molecular mass and their diversity. The subsequent analysis of lipids (having a much smaller molecular mass) was successful. First, a blinded MSI analyst could rapidly and easily differentiate between severe and mild ischemic injury. Interestingly, histopathologic examination, being the gold standard, could not differentiate between amount of ischemic injury, which may be explained by the following factors. First, it has repetitively been shown that microscopic assessment is not accurate for assessment of ischemic injury after kidney donation (as described above). Furthermore, MSI allows molecular based identification of injury during donation, in a time frame in which microscopic and morphologic changes may not be detectable yet. Of clinical interest is that the speed of MSI may be faster than histopathologic examination and that, being a digital technique, interpretation of MSI output may be performed on a topographic different location. These promising results should be interpreted with caution and translation to the clinical setting is preliminary, because the analysis on human tissue is expected to be more difficult. This is because of the heterogeneity of human renal tissue. Human kidneys have sustained injury over years (senescence) and suffer from injury different per type of donor (i.e. DCD and DBD), compared to a relatively homogenous group of young pig kidneys. Next to the capacity of MSI to differentiate between mild and severely injured kidneys, this study also led to insights in the underlying biochemical processes of acute ischemic kidney injury. We found that a suite of lipid degradation products is elevated after severe ischemia, such as lysocardiolipins, lysophosphatidylcholines and lysophosphatidylinositol. These lipids are associated to both mitochondrial and cell membrane integrity. The increased concentration of these lipids confirms that they have a role in renal injury. Furthermore, using MSI, we were able to appoint these injury markers to distinct tubular areas of the renal cortex, which may make future analysis of renal injury distribution after ischemia possible. In conclusion, MSI may be a new promising technique in the field of viability testing. It may provide information in an early stage of injury on a molecular level. While new preservation techniques, such as NMP emerge, the use of high-throughput MSI for viability testing is of even more interest, because it may test the effect of interventions on tissue in a pre-clinical setting and reveal pathophysiologic processes which are altered. This needs to be further investigated. Furthermore, future clinical studies may provide insight in the role of MSI in human tissue analysis and its possible clinical value. The next novel method, with a potential to test viability in a porcine perfusion model, is the use of renovascular circulating volume (chapter 8). We used this dilution technique, used in circulating volume assessment of hemodialyzer fibres, to test renovascular circulating volume during machine perfusion with different perfusion time, pressure and different ischemia times. Because this test was not performed before, we focussed on its repeatability and validation. The dilution technique was reproducible. Furthermore, renovascular circulating volumes differ per group of ischemia (15 vs 45 min), showing that circulating volume decreases when the ischemic injury is more severe. Although the results of this study show that the technique is reproducible, it did not result in a follow-up study. This is mostly due to our focus on different techniques, which we felt are more promising, such as MSI or pathologic examination. However, the dilution technique remains non-invasive and easily to perform. As such, it may be of interest to perform further studies with an improved technical setup; even better it would be when the technique is incorporated into perfusion machines. Furthermore, with the introduction of NMP (with higher flows, temperature and circulating of blood), the sensitivity and specificity of the dilution technique may be improved. It may also be an addition to monitor organ quality during these resuscitative preservation techniques. In chapter 9 we performed capsulotomy of ischemically injured porcine kidneys. The idea arose from the observation that severe ischemically injured kidneys were often swollen and renovascular circulating volume and perfusion reduced. We showed, using 16 kidneys, that capsulotomy along the greater curvature of the kidney immediately alleviates the compromised kidney and parameters, such as flow, resistance, renovascular circulating volume and intraparenchymal pressure, changed in all kidneys for the better. Surprisingly, we did not find a significant difference between kidneys with minor (20 min) or severe (45 min) ischemia. The positive effect of this intervention on renal perfusion seems evident, but its clinical use is however debatable and needs further study on its safety. Although we did not observe any parenchymal injury of the intervention in our model, transplant surgeons feel reluctant to perform this manoeuvre. Especially, because there are no reliable viability tests to select those kidneys which are viable and thus need no intervention with possible complications. Therefore, capsulotomy should only be performed in kidneys susceptible to renal compartment syndrome and the risk of further decline in viability. Irrespective of the risk, capsulotomy may improve renal perfusion and thereby preservation quality by means of faster cooling, improved perfusion of therapeutics and better wash-out of waste products. As result of the latter, viability assessment may be improved when injury markers in the perfusate are measured. Furthermore, the study was performed when NMP was not performed yet. It would be of interest to repeat the study using NMP, which offers a better setting to observe relevant changes in a metabolic active organ as well as the possible risk of parenchymal injury and haemorrhage. Future prospects Since January 2016 hypothermic machine perfusion (HMP) has been implemented as standard preservation technique for all kidneys donated after death in the Netherlands. From a clinical point of view this was necessary because it was concluded that HMP is superior to cold storage. However, the use of viability assessment during the perfusion period is not implemented in protocols, because, as we have reviewed, there is no scientific basis to do so. In this thesis we have used multiple novel strategies to improve viability testing, but were not able to find a single test with clinical useful predictive power. In addition, we have used novel strategies, like the use of renovascular circulating volume or the use of MSI, in animal studies. Especially MSI seems promising, but results are preliminary and the technique should be further developed and studied to estimate its potential. We conclude that the future of viability testing is unclear and no one has shown otherwise so far. After more than 20 years of use and research of HMP, we may argue that HMP is not suitable for viability assessment of donor kidneys. Donor kidneys in a hypothermic hypometabolic state may not properly reflect their viability and chance of post-transplant function well enough. Also, the aetiology of graft failure is mostly multifactorial and prediction of which kidneys will function or fail, especially with regards to marginal donor kidneys, is difficult when kidneys are preserved at 4° C. The use of NMP has theoretically many advantages of HMP with regard to viability assessment. Compared to HMP, kidneys are perfused with blood-based perfusates and with higher flows and they are metabolically fully active. Of all viability assessments strategies during HMP we discussed, i.e. measurement of perfusion parameters, perfusate biomarkers and histologic and molecular tissue analysis, all are theoretically more promising when performed in a normothermic state. The improved perfusion mimics in-vivo perfusion and parameters like renal resistance and renovascular circulating volume may be much more reliable. If perfusion is compromised it will be easier to pin-point the problem, maybe even by macroscopic inspection. Furthermore, increased perfusion leads to a better inflow of therapeutics and/or oxygen and better outflow of waste-products. As results, biomarkers may give a better reflection of kidney viability. Using blood-based perfusate, more biomarkers may be measured Additionally, the kidney will produce urine if properly perfused, and production of and biomarkers therein can be measured to estimate viability. As result, normothermic machine perfusion, as a new platform to perform pre-transplant viability assessment, has led to a rapid increase in studies using interesting novel strategies to determine renal viability. Using a porcine perfusion model, it was shown that laser speckle contrast imaging linearly correlates to renal blood flow and microcirculation of absent perfusion can be visualized and semiquantified. The group of Thomas Minor used the metabolic turnover of 13C-acetate as marker during pre-transplant isolated perfusion and could discriminate between “good”, “acceptable” and “critical” quality grafts (AURC area under the ROC curve 0.91; p<0.001). The group from Rotterdam also used photoacoustic imaging in a porcine model, which is easily performed using an modified ultrasound devices, to assess cortical oxygenation. They showed that it effectively distinguishes between functional and non-functional kidneys and conclude that this is a promising tool that has a high potential for supporting organ selection for kidney transplantation. In line with non-invasive assessment during NMP, Hamerlink et al. used magnetic resonance imaging to assess renal function. They found different blood flow distribution, comparing in-vivo and ex-vivo with minimal or severe warm ischemic injury. Analysis of tissue after NMP has not been performed much. However, in theory, tissue analysis after a normothermic resuscitative period may give a better reflection of organ viability than after a hypothermic period. After the first kidney transplantation after NMP in 2011 in Leicester, only a few groups have been able to achieve the same results. Clinical data is scarce, almost solely coming from the Nicholson group. They recently performed a high-quality randomized controlled trial, comparing static cold storage (SCS) to SCS plus one hour NMP in a total of 338 DCD kidneys. The found no beneficial nor detrimental effect of NMP, with a DGF rate of 60.7% for NMP and 58.5% in SCS (adjusted odds ratio (95% confidence interval) 1.13 (0.69–1.84); p=0.624). Interestingly, for viability assessment, they only used a NMP binary assessment score, including macroscopic appearance, renal blood flow and urine output, and found no difference in observed DGF for high or low scores. In the near future, more clinical studies are expected, comparing NMP to HMP or NMP tot SCS. In contrast to clinical studies, there are multiple study groups in a pre-clinical phase. They focus on improving NMP technical set-up, optimising perfusion solution and devising new renoprotective strategies and interventions for NMP, using animal or discarded human kidneys and comparing them to HMP or SCS. For viability assessment, the widespread implementation of NMP in renal transplantation has just begun. Therefore, clinical data on viability assessment is still hardly available. In conclusion, NMP is a new technique which needs further development and study to clarify its additional clinical value. Viability assessment slowly follows and enters a new ‘warm era’, of which the theoretical advantages seem obvious, but the data to support this are still lacking. Irrespective of the technical developments in donor kidney viability improvement and testing, their use heavily depends on the developments in clinical kidney transplantation. A recent paper describes the successful developments in pre-clinical xenotransplantation. This led to the first pig-to-human genetically modified kidney transplant on the 16th of March 2024 and it demonstrates that, at least in the short term, the kidney is safe and functions. Regretfully, the recipient died after 52 days from sudden cardiac arrest, but with a kidney without serologic of histologic evidence of antibody-mediated rejection. If xenotransplantation proves to yield clinical acceptable long-term results, it may greatly change the availability of transplantable organs. However, these follow-up results are not available, thus this remains unclear. Another interesting future prospect was recently discusses by Hosgood et al. With the use of vitrification (the rapid cooling of organs to a stable, ice-free, glass-like state) and nanowarming (the heating of nanoparticles within organ vasculature to achieve a rapid and uniform warming) it was shown that rat kidneys sustained function up to 30 days after transplantation after they had been cryogenically stored for 100 days. The concept of cryopreservation and organ banking would allow a more elective approach to kidney transplantation. The concepts of successful vitrification, nanowarming, cryopreservation, organ banking and xenotransplantation will inevitably impact the field of organ viability improvement and testing, a field which will have to adept and advance to these possible clinical changes. These upcoming developments are exiting and hopeful for future patients with end-stage renal disease.