Accurate assessment of deceased-donor kidneys with regards to suitability for transplantation is central to effective organ use. In recent years, disparities between available donor organs and demand for kidney transplantation have resulted in an increasing reliance on suboptimal donors at higher risk of poor graft outcome (1, 2). Achieving primary function of the graft and sufficient graft survival for the transplant to be of net benefit to the recipient are arguably the two most important factors in the decision to accept or reject an organ. Risk of poor graft outcome needs to be balanced against the survival benefit to patients compared to remaining on the waiting list and on dialysis (3). To assist the transplant team in this decision, a number of tools and markers for predicting graft outcome following kidney transplantation have been developed, typically for application in kidneys from less than “ideal” donors. However, the evidence supporting the use of these methods and their predictive ability are at best patchy and most have not been validated outside the original study population (4). There are no national or international consensus guidelines or standards for deceased-donor kidney appraisal, and the final decision ultimately lies with the implanting surgeon and their team.
In the current climate of sustained organ shortage, inappropriate discard of donor kidneys also has serious consequences for wait-listed patients. It has been suggested that many of the 2,644 kidneys assessed as being not suitable for use in transplantation in the United States in 2011—some 18% of all donated kidneys—could have been transplanted with good outcomes (5). The tools or parameters cited as underlying the decision to discard a kidney have often not been clearly shown to impact on graft outcome (4, 6–8). Much of the 10-fold difference in discard rates observed between centers (9, 10) may be a result of the subjective nature of organ assessment as well as conflicting evidence regarding the value of appraisal tools.
Large-scale, multicenter, adequately powered prospective studies are ultimately required to answer many of the outstanding questions regarding kidney appraisal and utilization (4). In their absence, greater clarity among clinicians as to which tools are supported by robust, validated evidence and which are not is urgently needed. This is a first step in moving decision-making around the utilization of suboptimal kidneys from its current ad hoc practice to a more uniform, evidence-based approach. In this review, we sought to critically evaluate existing and promising emerging tools for deceased-donor kidney appraisal and outline the levels of evidence supporting their use. We also highlight where current accepted appraisal practice is supported by only scant evidence, as well as the challenges in conducting high-quality research in this field. We hope this will stimulate discussion and further study of this critically important aspect of the transplantation process.
STANDARD CRITERIA DONORS (SCD), EXPANDED CRITERIA DONORS (ECD), AND DONATION AFTER CIRCULATORY DEATH (DCD)
The “ideal” deceased donor is considered to be a person aged 10 to 39 years, without a history of hypertension, who did not die of a cerebrovascular accident and whose terminal pre-donation creatinine level was less than 1.5 mg/dL (11). Optimal short- and long-term graft outcomes are expected from donors with these clinical characteristics. However, increasing transplant waiting lists combined with persistent organ shortages over the past decade have provided a compelling reason for the transplant community to expand the criteria for organ acceptance from beyond the “ideal donor” (12). This has resulted in the widespread use of expanded criteria donor (ECD) kidneys. The criteria for ECD kidneys are based on national registry analysis from the United States for kidneys transplanted between 1995 and 2000. Age and a combination of three donor medical risk factors were found to be associated with such a relative risk (RR) for graft failure of greater than 1.7 when compared to an ideal donor reference group (11). On the basis of this analysis, ECD kidneys came to be defined as those 60 years or older or those aged 50 to 59 years with two of the following three features: hypertension, terminal serum creatinine greater than 1.5 mg/dL, or death from a cerebrovascular accident (11, 12).
A large number of studies have evaluated graft survival for kidneys meeting ECD criteria versus those from SCDs. These have typically been observational and retrospective. Several single-center studies have suggested that similar graft survival to SCD kidneys can be achieved by using ECD kidneys (13–17); however, a recent systematic review identified that all available multicenter registry analyses show significantly worse 1- to 15-year patient and graft survival after kidney transplantation with ECD versus SCD kidneys (17).
Although the SCD and ECD classification is easy to apply, has greatly assisted the transplant community to deal with kidneys at higher risk of graft failure in a consistent manner across organ sharing networks, and has shortened weighting times for patients consenting to receive such organs (17), it has several limitations. It applies a binary classification to what is a continuum of quality for donated kidneys and may therefore underestimate the variability in ECD kidneys. It was modeled on data from a period where reliance on ECD kidneys was not as high and when overall graft survival as well as graft survival for ECD kidneys was lower (18). The original modeling study did not include a validation cohort, making its predictive value unclear. Ensuring accurate predictive value of the ECD classification is important, as a kidney labeled as “ECD” is significantly less likely to be utilized than one labeled as SCD; between 1999 and 2005, 41% of ECD kidneys procured were discarded in the United States, including 28% of ECD kidneys with the lowest risk of poor outcomes (RR 1.7–1.99), compared to 8% of SCD kidneys (9).
Donation after circulatory death (DCD) has also increased in the past decade as a further means of expanding the pool of donor kidneys. Although delayed graft function (DGF) rates are significantly higher than when kidneys are procured following donation after brain death, graft survival is comparable (19). Despite this, kidneys labeled as “DCD” are more than three times as likely to be discarded than non-DCD kidneys (9).
In light of the different categories of donor quality and donor type, the predictive value of the different kidney appraisal tools should ideally be evaluated and validated for use in each of these categories. Unfortunately, this has rarely been done in studies evaluating appraisal tools and is further complicated by variation in working definitions of suboptimal organs (which the ECD designation was designed to overcome). It is therefore not possible to evaluate each tool separately for SCD, ECD, and DCD categories, though in instances where specific literature does exist for these we have highlighted this in this review.
General inspection (i.e., macroscopic appearance) of donor organs by the transplant team is widely practiced and often fundamental to the decision to accept or reject an organ. It is important for identifying renal tumors, vascular and anatomical variations or damage, thrombosis, atherosclerosis, infarction, fibrosis and scarring, and for evaluating the quality of perfusion of the organ after retrieval. Surgeons also typically consider donor factors, preservation variables, and recipient factors in their global assessment of the suitability of an organ for transplant. However, although there is some evidence from the liver transplantation literature that general inspection by the surgeon can be highly variable, particularly for organs at moderate risk of poor outcome (20, 21), there are no studies formally evaluating the strengths, weaknesses, interobserver variability, or the predictive value and validity of general inspection in the setting of kidney transplantation. There is an urgent need for research that addresses these questions, especially where surgeon appraisal is commonly used to inform organ utilization and allocation decisions.
Pretransplant (pre-implantation) biopsy of deceased-donor kidneys can be used to evaluate the quality of the donor kidney and determine pre-existing renal disease. It is typically applied selectively, predominantly in ECD and DCD kidneys; however, practice between centers and countries remains widely variable: in the United States, approximately 75% of extended criteria donor (ECD) kidneys are biopsied (9), whereas pre-implantation biopsy is rarely used, even in older donors, by European centers (4, 22).
Glomerulosclerosis (%), vascular disease, and interstitial fibrosis are the most commonly reported donor kidney parameters that have been shown to be associated with poor graft outcomes (23–30). However, there is no consensus as to the relative importance of each factor, nor agreement as to which threshold values should be applied to define limits of acceptability. In addition, it remains unclear how histological parameters interact with other donor and recipient risk factors; in one large registry study of ECD kidneys, the degree of glomerulosclerosis was significantly associated with the risk of discard, but not consistently associated with either DGF or graft failure (19).
Several composite histopathological scoring systems are in use—the Pirani score (also known as the Remuzzi score) and the Maryland Aggregate Pathology Index (MAPI)—and have been designed specifically for the purpose of donor-kidney evaluation (28, 29), whereas the Banff criteria and Chronic Allograft Damage Index score have been extrapolated from their original use in evaluation of chronic allograft nephropathy to the setting of pre-implantation assessment (24, 30). The key features of the MAPI and Remuzzi scores are summarized in Table 1.
The Pirani (Remuzzi) Score
The Pirani score (29) assigns 0 to 3 points each for the degree of chronic glomerular, vascular, tubular, and interstitial changes. The Pirani score was developed a priori from experimental work exploring the concept of a critical nephron mass required to sustain function, rather than from clinical data (31, 32). Applying the Pirani score, kidneys with a global score of 0 to 3 were recommended for single transplants, those with a score from 4 to 6 were recommended for use as dual transplants, and those with a score of 7 or greater were discarded (29). Remuzzi’s Dual Kidney Transplant Group has demonstrated that when the score is applied to marginal or older donors, outcomes are better than those achieved with older donors where biopsy was not utilized and comparable to those achieved using kidneys from standard criteria donors (27). However, follow-up in these studies has been relatively short: 6 months in the initial validation study (29) and a median of 23 months in the largest series (27). The merits of use of dual transplants to achieve comparable outcomes to single-kidney standard criteria have also been questioned, particularly for kidneys with an “intermediate” score of 4 (33). Nonetheless, application of the biopsy score to allocation resulted in a 24% increase in the number of available kidneys being transplanted (27). An important potential limitation of this score is that weighting for each histological parameter was arbitrarily and equally set based on the expected effect on nephron mass, rather than on the hazard ratio for graft function or graft survival. This may explain the conflicting results reported as to its discriminatory ability in predicting graft survival (26, 27, 34).
The Maryland Aggregated Pathology Index Score
The MAPI, developed from 371 biopsies from a single center, assigns a weighted score for arteriolar hyalinosis, periglomerular fibrosis, scarring, percentage glomerulosclerosis, and vascular wall narrowing (12). Biopsy features independently associated with an increased risk of graft failure were glomerulosclerosis greater than or equal to 15%, interlobular arterial wall-to-lumen ratio greater than or equal to 0.5, and presence of periglomerular fibrosis, arteriolar hyalinosis, or scarring. Five-year graft survival for kidneys with a MAPI score of 0 to 7 was 90%, 63% for those with a score of 8 to 11, and 53% for those with a score of 12 to 15. Glomerulosclerosis and vasculopathy were found to have significant negative impacts on graft survival, whereas interstitial fibrosis was not strongly associated with poorer outcome. In a model-validation cohort, there was an increase in RR of 21% for graft loss for every point increase in the score. The probability of concordance between the score’s prediction and the actual outcome was good, but not outstanding (area under the curve=0.74). One limitation of this study is that the cold ischemia times were long—on average over 33 hours and over 24 hours in 80% of kidneys studied. Furthermore, only a small number of kidneys with high levels of glomerulosclerosis were transplanted and the authors therefore advised against extrapolation of the score to biopsies with more than 25% glomerulosclerosis.
Pitfalls of Evaluation of Kidneys by Biopsy
While a single sample pre-implantation biopsy can provide reliable data on the state of the whole kidney in most cases (35, 36), interobserver variability remains a problem particularly in the rating of interstitial fibrosis. Accurate assessment of global glomerulosclerosis is dependent on the number of glomeruli sampled; reasonable precision can be obtained with as little as seven glomeruli on a needle biopsy (36), although a number of the histological scoring systems require a minimum of 25 glomeruli to be sampled (29). Shallow wedge biopsies can overestimate glomerulosclerosis, owing to the increased incidence of this in the subcapsular region (37). Frozen sections are not reliable for assessment of mesangial cellularity, glomerular capillary wall thickening, some diabetic lesions, microthrombi, and acute tubular necrosis (38).
Prolonged Cold Ischemic Time
Fixation of the biopsy specimen in formalin followed by paraffin sectioning overcomes the limitations of frozen sections (38), but typically takes 4 to 5 hours. The benefits of biopsy-guided decision-making must therefore be balanced against the risks emanating from increased cold ischemic time. ECD and DCD kidneys are most likely to require a biopsy to guide allocation, but are also most at risk from the negative effects of extended cold ischemic time (19). In one study of ECD kidney allocation, for example, not undertaking biopsies reduced the ischemic time by 9 hours and reduced the rate of DGF from 43% to 15%, although long-term outcomes for these kidneys were not reported (39).
In summary, pre-implantation biopsy is often employed in kidney appraisal on the basis that it is an objective measure of pre-existing donor kidney disease, especially in ECDs. However, the supporting literature is conflicting and it is important to be aware of the potential limitations when using biopsy to guide decision-making. This has recently been highlighted by U.S. registry data—where biopsy is routinely performed—which suggested that the use of pre-implantation biopsy in kidneys may actually increase the rate of inappropriate discards, especially in ECD kidneys, while making no measurable difference to allograft survival (10). However, there is also a body of evidence to support the role of pre-implantation histological scoring in predicting short-term to medium-term graft survival with fair to good precision, particularly in older donors. Parameters on biopsy independently associated with an increased risk of graft failure at 5 years are glomerulosclerosis greater than or equal to 15%, interlobular arterial wall to lumen ratio greater than or equal to 0.5, periglomerular fibrosis, arteriolar hyalinosis, or scar, but the relative importance of each parameter remains unclear. There is a paucity of studies investigating an association between pre-implantation biopsy and immediate graft function.
DONOR RISK SCORES
Donor risk indices are designed to identify pretransplant factors that are most strongly associated with an increased risk of long-term graft failure or delayed graft function (40). Over the past decade, the concept of the “sub-optimal” or “expanded criteria” donor has expanded from a purely binary classification system (SCD or ECD) towards recognition that graft quality is a spectrum in which donor, retrieval, preservation, and operative and recipient factors interact to determine short-term and long-term outcomes. Utilization of large registry-based datasets to model and test the pretransplant factors most strongly associated with an increased risk of graft failure has led to the development of a number of different “donor risk indices,” which are discussed in more detail below. An ideal donor risk score is one which utilizes information readily available to the transplant team at the time of donor organ offer, is simple to use, and has been shown to have strong predictive accuracy in both the model-validation cohort and in other donor populations (generalizability).
The history of the development of organ-specific donor risk indices has recently been reviewed in depth elsewhere (40). Two types of score have emerged: those which aim to predict long-term graft failure risk and those which aim to predict delayed graft function. The most commonly used long-term graft failure risk scores are the modified deceased donor score (DDS) of Nyberg et al. (41), the donor risk score (DRS) of Schold et al. (42), and the Kidney Donor Risk Index (KDRI) proposed by Rao et al. (43). These are all based on U.S. registry data. More recently, a simplified donor risk index based on registry data from the United Kingdom (UKKDRI) has also been published by Watson et al. (44). A comparison of the components of the four scores is shown in Table 2.
Anglicheau et al. have developed a composite clinical-histopathological scoring system for marginal donors (26). A composite score which included clinical and histopathological factors (donor serum creatinine ≥150 μmol/L, donor hypertension and glomerulosclerosis ≥10%) had the highest performance in predicting poor renal function (eGFR) at 1 year. This was greater than either the Nyberg DDS or the Pirani (Remuzzi) histopathology score alone, but did not increase the prediction of the composite model compared to the use of only glomerulosclerosis greater than or equal to 10% or less than 10%. Although the composite score is yet to be prospectively validated, such scores are likely to have greater predictive power if the clinical and histopathological factors they combine capture different aspects of graft quality, as this initial study would suggest.
Delayed graft function has become increasingly common with the use of ECD and DCD kidneys (19), and has significant clinical and economic implications. Although DGF should not in itself be used as a discriminatory factor for exclusion of a donor organ or a recipient from transplantation, quantification of its likelihood may nonetheless allow for modification of known modifiable risk factors that contribute to DGF such as cold ischemic time and immunosuppressant regimes. Using data from 13,846 patients from the United States Renal Data System, Irish et al. modeled and validated a nomogram for predicting the likelihood of DGF based on 16 independent donor and recipient risk factors known at the time of transplant (45). The most significant risk factors were a history of pre-implantation dialysis, followed by DCD donation and recipient of a single-organ transplant. A strong interaction with ethnicity was also noted. The nomogram had reasonable overall predictive power in the validation cohort (c-statistic=0.70), and it was later validated retrospectively in a separate population with a much lower overall rate of DGF, where it performed well despite the lower prevalence of DGF (46). However, a prospective validation carried out in another U.S. population in 2006 disputed the clinical utility of the nomogram for predicting risk in an individual patient (47), highlighting the difficulty in application of this approach to different donor populations.
Pitfalls of Donor Risk Scores
A significant limitation of the scoring systems in Table 2 is their low predictive accuracy, or concordance between predicted and observed (actual) outcomes for any given organ. Both the KDRI and the UKKDRI have overall concordance statistics of 0.63 and 0.62, respectively (43, 44), where a c-statistic of 0.5 indicates that the model has no more predictive value than that of chance. The KDRI c-statistic was 0.78 when comparing only the upper and lower quartiles of graft failure risk but only 0.58 when applied to only the middle two quartiles (43). The DGF nomogram in contrast had good overall predictive power in the validation cohort (c-statistic=0.70) (46) but was less valuable in a prospective study as outlined above (47).
The problem of generalizability with donor risk tools was highlighted by Moore et al. who performed a comparative analysis of donor quality scoring systems in a separate single-center renal transplant population (48). Long-term graft and patient outcomes are significantly influenced by factors that cannot be incorporated into models (49). These include the “center-effect” (50–52), which may explain as much as 30% to 40% of variation in 1-year graft survival (53) and is influenced by broader determinants of health outcomes such as health policy and funding systems and ethnic, geographic, and socioeconomic barriers to accessing specialist follow-up care (49, 51).
Presently therefore, donor risk indices remain an adjunct tool in clinical decision-making, which may inform, but should not direct, the course of action with regards to use and allocation of a graft. Donor risk scores can identify grafts at increased risk of poor long-term graft survival, but have poor predictive power and should not be used on their own in the decision to accept or discard a kidney. The risk nomogram for development of DGF by Irish et al. has fair predictive power and can be used to evaluate donor-recipient pairs at increased risk of DGF. There is no evidence that donor risk scores are useful in predicting whether a kidney will develop primary nonfunction (PNF).
MACHINE PERFUSION CHARACTERISTICS
Renewed interest in the use of hypothermic machine perfusion (HMP) for kidney preservation over the past decade has identified HMP perfusion characteristics and perfusate biomarkers as potential predictors of graft outcome. While initial studies were small, retrospective, and conflicting (54, 55), the recent large multicenter Eurotransplant machine perfusion randomized trial (56) identified renovascular resistance at the end of HMP (but not earlier) as an independent risk factor for DGF (OR 21.28) and 1-year graft survival (HR 12.33). However, the predictive value of renovascular resistance was poor (c-statistic of 0.58 for prediction of DGF), and attempting to identify a threshold value for resistance from the dataset did not improve the predictive value (55). Owing to the very low incidence of PNF in this series (2%), it was not possible to accurately determine the effect of renovascular resistance on this outcome. In another recent large robust retrospective series of DCD kidneys, renovascular resistance in the first hour (but not later) was independently associated with DGF (OR 2.35) as well as PNF (OR 2.04), but again the predictive power was not strong (c-statistic=0.61). In this series, renovascular resistance was not associated with graft or patient survival at 1 or 5 years (57). Owing to the suboptimal predictive value, both studies appropriately concluded that renovascular resistance should not be used as a stand-alone tool for appraisal of kidney viability (55). This conclusion is further supported by studies showing that kidneys declined based solely on their perfusion characteristics can be successfully transplanted, with outcomes similar to those of kidneys with “good” perfusion characteristics (58–60).
Biomarkers Within the Machine Perfusate
Molecules or cellular constituents released by the donor kidney into the preservation solution during HMP represent attractive potential biomarkers. These include lactate dehydrogenase (LDH), a nonspecific marker of cellular injury and glutathione-S-transferase (GST), an enzyme localized in the renal tubules (61, 62). The α-GST form of GST in particular is strongly associated with proximal tubular injury (63), a feature of DGF. Other candidate perfusate biomarkers include aspartate aminotransferase, N-acetyl-β-D-glucosaminidase (NAG), alanine aminopeptidase heart-type fatty acid binding protein (H-FABP), redox-active iron, and ionized calcium (64–66). As with renal resistance HMP parameters, interpretation of studies has been difficult owing to bias introduced by study design and methodology, and results have been conflicting (62). The most robust analysis of the role of perfusate biomarkers in predicting graft function and survival again comes from the Eurotransplant HMP randomized trial in which GST, NAG, and H-FABP concentrations at the end of HMP were shown to be independent predictors of DGF, but not of PNF or graft survival (64). The authors concluded that while the GST concentration (highest c-statistic of 0.67) in the perfusate may assist clinicians in identifying kidneys at higher risk of DGF when used in conjunction with other appraisal tools, it should not be used as a basis for kidney discard as it did not predict PNF or graft survival (64). Interestingly, concentrations of these biomarkers had no relevant correlation with either cold ischemia time or renal resistance measures, suggesting they may reflect different aspects of kidney damage and offer additional information in a rapidly expanding appraisal armamentarium.
On the basis of the current evidence, machine renovascular resistance or perfusate biomarkers should not be used as stand-alone tools for appraising kidney viability and assessing long-term graft outcome. The most promising perfusate biomarker is GST with moderate predictive value for development of DGF.
FUNCTIONAL KIDNEY WEIGHT (NEPHRON MASS)
A mismatch between the functional nephron mass of the donor kidney and recipient metabolic requirements has been proposed as a mechanism contributing to antigen-independent chronic allograft nephropathy and graft loss (67, 68). On this basis, methods for assessing and matching donor nephron mass to recipient requirements have been proposed in both living and cadaveric kidney donation, predominantly by assessing donor kidney weight (DKW) and recipient body weight (RBW) ratios (69–73). In one large study, a DKW-to-RBW ratio of less than 2 g/kg resulted in significantly higher rates of proteinuria. However, there was no effect of DKW-to-RBW ratio on graft survival up to 4 years (70). Other studies have reported conflicting results on the effect of nephron mass on graft survival, using kidney volume measurements (74), body surface area (BSA) (72), kidney ultrasonographic cross-sectional area to recipient body weight (73), and donor creatinine clearance to recipient BMI (71) as surrogate measures. There is no evidence that mismatch between the donor kidney size and the recipient is associated with development of DGF or PNF. Functional kidney weight does not therefore provide any additional insight into short-term or long-term graft function and for this reason does not appear to be widely used as a tool.
MOLECULAR DIAGNOSTICS AND BIOMARKERS
In the setting of renal transplantation, “-omics” approaches have allowed for the analysis of molecular profiles associated with DGF, acute rejection, and chronic allograft damage. These are covered in depth in two recent reviews of the topic (75, 76). Candidate gene approaches, genome-wide analysis, epigenomics, microarray gene expression screening, miRNA expression, and proteomics have all been used to investigate both ischemia-reperfusion injury and rejection in renal transplantation.
These have been most promisingly applied in the setting of urinary biomarkers for ischemia-reperfusion injury, where transcriptome-based studies have led to the identification of new kidney injury markers that can be detected in the urine. More than 15 different potential urinary biomarkers have been studied in the deceased-donor setting as potential predictors of graft outcomes (62), the most promising of which is urinary neutrophil gelatinase-associated lipocalin (NGAL) (77), optimistically labeled “the troponin of kidney injury” (78). NGAL is increasingly available as a (costly) clinical biochemical assay that can be completed within relevant timeframes for donation. In a recent prospective study of deceased donors, raised urinary NGAL was independently associated with prolonged (but not overall) DGF and reduced 1-year graft survival (90.3% vs. 97.4%, P=0.048), but had poor predictive power (c-statistic=0.58) and therefore should not be used as a stand-alone quality assessment tool.
Application of molecular diagnostics to pretransplant kidney biopsies has produced interesting new insights into the molecular basis of renal damage and disease during transplantation, but has yet to be translated into a clinically applicable test of kidney graft viability. One of the challenges of “-omics” approaches is that they generate large amounts of data per clinical sample, so studies using only very low sample numbers in each group are prone to bias (79). While molecular diagnostics hold promise for the future development of tools for predicting donor kidney outcomes—particularly in the combining of multiple molecular markers with histology, donor, and recipient risk factors (80)—currently there are no clinical applications for genomic, proteomic, or metabolomic approaches that can reliably predict graft survival, DGF, or PNF.
EX VIVO AND POSTMORTEM NORMOTHERMIC PERFUSION (VIABILITY ASSESSMENT)
One potential method of viability assessment uses an ex vivo perfusion system in which the kidney is perfused with blood or an analogous physiological solution with nutrient on an isolated organ perfusion system with an oxygenator and venous reservoir (81–83). An alternative method involves the use of extracorporeal membrane oxygenation (ECMO) or a cardiopulmonary bypass circuit in a donor following declaration of death (84). A number of potential viability measures for ex vivo perfusion have been examined in large animal transplant studies (81) and more recently in human kidneys (85, 86). These have mostly focused on assessing renal functional parameters, namely GFR, renal blood flow, and intra-renal resistance ex vivo. Their ability to predict short-term function in ECD kidneys appears promising but has yet to be subjected to a randomized clinical trial, where impact on long-term function remains unknown (85, 86). In addition to their potential in predicting graft outcomes, ex vivo normothermic perfusion techniques can also deliver therapeutic intervention to enhance preservation or restore function to marginal kidneys (82). Postmortem ECMO has been used in a number of small clinical studies as a bridge to successful kidney transplantation following donation after circulatory death (87, 88); however, less attention has been given to its use for developing and validating potential viability markers for donor kidneys.
Ex vivo normothermic perfusion represents a promising approach to pre-implantation evaluation of the donor kidney. However, there is insufficient clinical data on this approach to draw any meaningful conclusions at present. More research is required to clarify whether this method will better predict long-term graft survival, DGF, or PNF for ECD or DCD kidneys in particular, where functional assessment before implantation may be of most benefit.
Accurate appraisal of the deceased donor kidney is central to optimal organ utilization, but remains an inexact science. Despite a plethora of available appraisal techniques, only two of these approach acceptable predictive power to be clinically useful; pre-implantation biopsy for long-term graft outcome—which is of particular value in assessing organs from older donors—and the donor risk nomogram for development of delayed graft function (Table 3). Graft survival in kidneys who meet the ECD criteria appears to be generally worse than for SCD kidneys (17). However, the binomial nature of ECD classification underestimates the variability in kidney quality, and a reliance on retrospective and observational data to model the effects of ECD criteria and the absence of appropriate validation studies mean the predictive value remains unclear. Donor risk indices, developed to provide a more nuanced approach to risk stratification, have had disappointingly poor predictive value except at the extremes of risk. The validity or reliability of surgeon appraisal, despite being a widely used modality for appraising kidney quality, has not been characterized. Wide variations in practice between individual surgeons, centers, and regions have also made it difficult to compare the role of additional appraisal technologies against “standard practice”. In an era of personalized medicine, emerging technologies such as molecular diagnostics hold exciting potential for informing clinical decisions for an individual patient. This notion of personalized medicine based on molecular signatures is most close to being realized in the setting of immunosuppressive therapy, but could ultimately also be used to inform organ allocation for individual donor-recipient pairs. Ex vivo normothermic perfusion of kidneys as a form of viability testing before transplantation—which is already being investigated by some centers as a potential clinical therapy—also represents a promising opportunity for making a global functional assessment of the kidney before implantation.
Organ allocation decisions are complex, with multiple variables interacting to determine an outcome. None of the appraisal tools developed to date have been able to adequately account for donor, organ, and recipient risk interactions. In addition, organ shortages have resulted in an ongoing struggle to balance allocation efficiency at a population level with both benefit and equity at an individual level. This ultimately is what defines the threshold for an “acceptable” suboptimal organ and will vary according to waiting list characteristics, as well as the risk each individual recipient and their treating team is willing to accept. Given this complexity, it is likely that it will never be appropriate to rely solely on a single appraisal tool to make a decision as to whether to accept or reject an organ. Tools with proven predictive power, such as pre-implantation biopsy, can however help guide decision-making and better inform both clinicians and patients as to risks. More research, especially high-quality prospective multicenter studies is urgently required to ensure the ongoing development and optimization of appraisal tools with good predictive value.
1. Port F. Organ donation and transplantation trends in the United States, 2001. Am J Transplant
2003; 3 (Suppl 4): 7.
2. Nathan H, et al. Organ donation in the United States. Am J Transplant
2003; 3 (Suppl 4): 29.
3. Merion R, et al. Deceased-donor characteristics and the survival benefit of kidney transplantation. JAMA
2005; 294: 2726.
4. Shapiro R, et al. The `two, one, zero’ decision: what to do with suboptimal deceased donor kidneys. Am J Transpl
2010; 10: 1959.
5. Sack K. In Discarding of Kidneys, System Reveals its Flaws
. New York Times; 2012.
6. Kayler L, et al. Outcomes and utilization of kidneys from deceased donors with acute kidney injury. Am J Transplant
2009; 9: 367.
7. Edwards E, et al. Reasons for non-use of recovered kidneys: the effect of donor glomerulosclerosis and creatinine clearance on graft survival. Transplantation
2004; 77: 1411.
8. Cho Y, et al. Factors associated with discard of recovered kidneys. Transplant Proc
2008; 40: 1032.
9. Sung R, et al. Determinants of discard of expanded criteria donor kidneys: impact of biopsy and machine perfusion. Am J Transplant
2008; 8: 783.
10. Cecka J, Gritsch H. Why are nearly half of expanded criteria donor (ECD) kidneys not transplanted? Am J Transplant
2008; 8: 735.
11. Port F, et al. Donor characteristics associated with reduced graft survival: an approach to expanding the pool of kidney donors. Transplantation
2002; 74: 1281.
12. Metzger R, et al. Expanded criteria donors for kidney transplantation. Am J Transplant
2003; 3 (Suppl. 4): 114.
13. Collini A, et al. Long-term outcome of renal transplantation from marginal donors. Transplant Proc
2006; 38: 3398.
14. Fraser S, et al. Acceptable outcome after kidney transplantation using “expanded criteria donor” grafts. Transplantation
2010; 89: 88.
15. Morrissey P, et al. Renal transplant survival from older donors: a single center experience. Arch Surg
2004; 139: 384.
16. Carter J, et al. Expanded criteria donor kidney allocation: marked decrease in cold ischemia and delayed graft function at a single center. Am J Transplant
2005; 5: 2745.
17. Pascual J, Zamora J, Pirsch J. A systematic review of kidney transplantation from expanded criteria donors. Am J Kid Dis
2008; 52: 553–586.
18. Lamb K, Lodhi S, Meier-Kriesche H. Long-term renal allograft survival in the United States: a critical reappraisal. Am J Transplant
2011; 11: 450.
19. Summers D, et al. Analysis of factors that affect outcome after transplantation of kidneys donated after cardiac death in the UK: a cohort study. Lancet
2010; 376: 1303.
20. Imber C, et al. Current practice regarding the use of fatty livers: a trans-Atlantic survey. Liver Transpl
2002; 8: 545.
21. Dare A, et al. Appraisal of donor steatosis in liver transplantation: a survey of current practice in Australia and New Zealand. Transpl Res Risk Manag
2012; 4: 31.
22. Snoeijs M, et al. Histological assessment of preimplantation biopsies may improve selection of kidneys from old donors after cardiac death. Am J Transplant
2008; 8: 1844.
23. Karpinski J, et al. Outcome of kidney transplantation from high-risk donors is determined by both structure and function. Transplantation
1999; 67: 1162.
24. Lopes J, et al. Evaluation of pre-implantation kidney biopsies: comparison of Banff criteria to a morphometric approach. Kidney Int
2005; 67: 1595.
25. Cockfield S, et al. The prognostic utility of deceased donor implantation biopsy in determining function and graft survival after kidney. Transplantation
2010; 89: 559.
26. Anglicheau D, et al. A simple clinico-histopathological composite scoring system is highly predictive of graft outcomes in marginal donors. Am J Transplant
2008; 8: 2325.
27. Remuzzi G, et al. Long-term outcome of renal transplantation from older donors. N Engl J Med
2006; 354: 343.
28. Munivenkatappa R, et al. The Maryland Aggregate Pathology Index: a deceased donor kidney biopsy scoring system for predicting graft failure. Am J Transplant
2008; 8: 2316.
29. Remuzzi G, et al. Early experience with dual kidney transplantation in adults using expanded donor criteria. J Am Soc Nephrol
1999; 10: 2591.
30. Kahu J, et al. Donor risk score and baseline biopsy CADI value predict kidney allograft outcome. Clin Transplant
2011; 25: E276.
31. Chanutin A, Ferris E. Experimental renal insufficiency produced by partial nephrectomy. Arch Intern Med
1932; 49: 767.
32. Brenner B. Nephron adaptation to renal injury or ablation. Am J Physiol
1985; 249: F324.
33. Panesar M, Kumar J, Abul-Ezz S. Long-term outcome of renal transplantation from older donors. N Engl J Med
2006; 354: 2071.
34. Re L, et al. Comparison between clinical and histopathological scoring in cadaveric kidney transplantation and its correlation with posttransplant evolution. Transplant Proc
2006; 38: 903.
35. Snoeijs M, et al. Histological assessment of pre-transplant kidney biopsies is reproducible and representative. Histopathology
2010; 56: 198.
36. Mazzucco G, et al. The reliability of pre-transplant donor renal biopsies (PTDB) in predicting the kidney state. A comparative single-centre study on 154 untransplanted kidneys. Nephrol Dial Transplant
2010; 25: 3401.
37. Muruve N, Steinbecker K, Luger A. Are wedge biopsies of cadaveric kidneys obtained at procurement reliable? Transplantation
2000; 69: 2384.
38. Randhawa P. Role of donor kidney biopsies in renal transplantation. Transplantation
2001; 71: 1361.
39. Carter J, et al. Expanded criteria donor kidney allocation: marked decrease in cold ischemia and delayed graft function at a single center. Am J Transplant
2005; 5: 2745.
40. Akkina S, et al. Development of organ-specific donor risk indices. Liver Transpl
2012; 18: 395.
41. Nyberg S, et al. Improved scoring system to assess adult donors for cadaver renal transplantation. Am J Transplant
2003; 3: 715.
42. Schold J, et al. The broad spectrum of quality in deceased donor kidneys. Am J Transplant
2005; 5: 757.
43. Rao PS, Schaubel DE, et al. A comprehensive risk quantification score for deceased donor kidneys: the Kidney Donor Risk Index. Transplantation
2009; 88: 231.
44. Watson C, et al. A simplified donor risk index for predicting outcome after deceased donor kidney transplantation. Transplantation
2012; 93: 314.
45. Irish W, et al. Nomogram for predicting the likelihood of delayed graft function in adult cadaveric renal transplant recipients. J Am Soc Nephrol
2003; 14: 2967.
46. Kaisar M, Johnson D. Validation of a nomogram for predicting the likelihood of delayed graft function in Australian adult deceased donor renal transplant recipients. Nephrology
2006; 11: 78.
47. Grossberg J, et al. Utility of a mathematical nomogram to predict delayed graft function: a single-center experience. Transplantation
2006; 81: 155.
48. Moore J, et al. Identification of the optimal donor quality scoring system and measure of early renal function in kidney transplantation. Transplantation
2009; 87: 578.
49. Yost S, Kaplan B. Comparing kidney transplant outcomes; caveats and lessons. Nephrol Dial Transplant
2012; 28: 9.
50. Opelz G, Mickey M, Terasaki P. Comparison of kidney transplant survival among transplant centers. Transplantation
1975; 19: 226.
51. Ojo A. Influence of reporting methods of outcomes across transplant centers. Clin J Am Soc Nephrol
2012; 6: 2732.
52. Kim S, et al. Mortality after kidney transplantation: a comparison between the United States and Canada. Am J Transplant
2006; 6: 109.
53. Gjertson DW. Update: center effects. Clin Transpl
54. Wight J, et al. The clinical and cost-effectiveness of pulsatile machine perfusion versus cold storage of kidneys for transplantation retrieved from heart-beating and non heart-beating donors. Health Technol Assess
2003; 7: 1.
55. Jochmans I, et al. The prognostic value of renal resistance during hypothermic machine perfusion of deceased donor kidneys. Am J Transplant
2011; 11: 2214.
56. Jochmans I, et al. Machine perfusion versus cold storage for the preservation of kidneys donated after cardiac death: a multicenter, randomized, controlled trial. Ann Surg
2010; 252: 756.
57. de Vries E, et al. Renovascular resistance of machine-perfused DCD kidneys is associated with primary nonfunction. Am J Transplant
2011; 11: 2685.
58. Sonnenday C, et al. The hazards of basing acceptance of cadaveric renal allografts on pulsatile machine perfusion parameters alone. Transplantation
2003; 75: 2029.
59. Mozes M, Skolek R, Korf B. Use of perfusion parameters in predicting outcomes of machine-preserved kidneys. Transplant Proc
2005; 37: 350.
60. Guarerra J, Goldstein M, Samstein B. ’When good kidneys pump badly’: outcomes of deceased donor renal allografts with poor pulsatile perfusion characteristics. Transpl Int
2010; 23: 444.
61. Beckett G, Hayes J. Glutathione-S-transferase: biomedical applications. Adv Clin Chem
1993; 30: 281.
62. Bhangoo R, et al. Deceased-donor kidney perfusate and urine biomarkers for kidney allograft outcomes: a systematic review. Nephrol Dial Transplant
2012; 27: 3305.
63. Loeb W. The measurement of renal injury. Toxicol Pathol
1998; 26: 26.
64. Moers C, et al. The value of machine perfusion perfusate biomarkers for predicting kidney transplant outcome. Transplantation
2010; 90: 966.
65. Gok M, et al. Do tissue damage biomarkers used to assess machine-perfused NHBD kidneys predict long-term renal function post-transplant? Clin Chim Acta
2003; 338: 33.
66. Polyak M, et al. Calcium ion concentration of machine perfusate predicts early graft function in expanded criteria donor kidneys. Transpl Int
1999; 12: 378.
67. Brenner B, Cohen R, Milford E. In renal transplantation, one size may not fit all. J Am Soc Nephrol
1992; 3: 162.
68. Brenner B, Milford E. Nephron underdosing: a programmed cause of chronic renal allograft failure. Am J Kidney Dis
1993; 21 (5 Suppl 2): 66.
69. Tan J, et al. Validity of surrogate measures for functional nephron mass. Transplantation
2011; 92: 1335.
70. Giral M, et al. Impact of graft mass on the clinical outcome of kidney transplants. J Am Soc Nephrol
2005; 16: 261.
71. Stratta R, et al. Optimal use of older donors and recipients in kidney transplantation. Surgery
2006; 139: 324.
72. Kasiske B, Snyder J, Gilbertson D. Inadequate donor size in cadaver kidney transplantation. J Am Soc Nephrol
2002; 13: 2152.
73. Nicholson M, et al. Influence of allograft size to recipient body-weight ratio on the long-term outcome of renal transplantation. Br J Surg
2000; 87: 314.
74. Miles A, et al. The effect of kidney size on cadaveric renal allograft outcome. Transplantation
1996; 61: 894.
75. Naesens M, Sarwal M. Molecular diagnostics in transplantation. Nat Rev Nephrol
2010; 6: 614.
76. Mas V, et al. Identifying biomarkers as diagnostic tools in kidney transplantation. Expert Rev Mol Diagn
2011; 11: 183.
77. Hollmen M, et al. Deceased donor neutrophil gelatinase-associated lipocalin and delayed graft function after kidney transplantation: a prospective study. Crit Care
2011; 15: R121.
78. Devarajan P. Review: neutrophil gelatinase-associated lipocalin: a troponin-like biomarker for human acute kidney injury. Nephrology (Carlton)
2010; 15: 419.
79. Naesens M, Sarwal M. Looking into the crystal chip: can microarrays predict graft function? Transplantation
2008; 85: 499.
80. Kreepala C, et al. Comparing molecular assessment of implantation biopsies with histologic and demographic risk assessment. Am J Transplant
2013; 13: 415.
81. Hosgood S, et al. A pilot study assessing the feasibility of a short period of normothermic preservation in an experimental model of non heart beating donor kidneys. J Surg Res
2011; 171: 283.
82. Hosgood S, Nicholson M. Normothermic kidney preservation. Curr Opin Organ Transplant
2011; 16: 169.
83. Vogel T, et al. The role of normothermic extracorporeal perfusion in minimising ischemia reperfusion injury. Transplant Rev
2012; 26: 156.
84. Ko W, et al. Extracorporeal membrane oxygenation support of donor abdominal organs in non-heart-beating donors. Clin Transplant
2000; 14: 152.
85. Hosgood S, Nicholson M. First in man renal transplantation after ex-vivo normothermic perfusion. Transplantation
2011; 92: 735.
86. Hosgood S. Renal transplantation after ex vivo normothermic perfusion: the first clinical study. Am J Transplant
2013; 13: 1246.
87. Magliocca J, et al. Extracorporeal support for organ donation after cardiac death effectively expands the donor pool. J Trauma
2005; 58: 1095.
88. Hsieh C, et al. Extracorporeal membrane oxygenation support in potential organ donors for brain death determination. Transplant Proc
2011; 43: 2495.