Abbreviations
ALT : alanine transaminase
cDCD: controlled donation after circulatory death
DBD: donation after brain death
DCD: donation after circulatory death
DGF: delayed graft function
FMN: flavin mononucleotide
HOPE: hypothermic oxygenated perfusion
ITBL: Ischemic-type biliary lesions
KT: kidney transplant
LT: liver transplant
NMP: normothermic machine perfusion
NRP: normothermic regional perfusion
RCT: randomized controlled trial
SCS: static cold storage
SRR: super-rapid recovery
uDCD: uncontrolled donation after circulatory death
UK: United Kingdom
INTRODUCTION
Donation after circulatory death (DCD) provides a viable means to deal with the ever-growing shortage of donors. Nevertheless, transplants from DCD are traditionally burdened by a higher rate of complications and graft loss. Dismal results have prompted the application of stricter donor and graft selection criteria in some settings and the use of machine perfusion technologies in others 1.
DCD inevitably involves compromised hemodynamics in the agonal phase, followed by an obligatory additional circulatory standstill for the declaration of death. In super-rapid recovery (SRR), this is hurriedly followed by incision, aortic cannulation, and organ perfusion with hypothermic preservation solution. Normothermic regional perfusion (NRP) temporarily re-establishes blood flow following declaration of death, through arterial and venous cannulae placed in the femoral vessels or directly in the aorta and vena cava after rapid laparotomy (Fig. 1). Supraceliac aortic balloon occlusion prevents cerebral reperfusion during NRP and restricts perfusion to the abdomen 2. This way, NRP restores previously depleted energy substrates, clears by-products of anaerobic metabolism, and induces endogenous antioxidants, thus helping to improve organ quality and maintenance before cold preservation 3. Moreover, in contrast to SRR, NRP also allows the assessment of organ function following the warm ischemic injury 4.
Although some ethical concerns have been raised, especially with premortem cannulation, heparin administration, and potential brain reperfusion, NRP is currently spreading in many European countries5. To maintain the permanence principle for death, Manara et al. have recently suggested inserting a cannula in the ascending aorta to identify inadequate occlusion and divert any collateral flow away from the brain 6. Moreover, antemortem interventions in the potential donor should follow national legislation and are ethically acceptable if they do not add risk, harm, or discomfort to the patient 3.
In this review, we first describe the diffusion of NRP worldwide. Next, the role of NRP in liver transplantation (LT) is discussed, with a particular focus on graft selection during perfusion and posttransplant outcomes. Finally, we review the clinical studies reporting on NRP in kidney transplantation (KT). The emerging use of NRP with complementary ex-situ machine perfusion is also described.
DIFFUSION OF NRP
Currently, NRP is mandatorily applied in DCD organ recovery in 3 European countries (Italy, France, and Norway) and is permitted in 5 (Spain, United Kingdom, Belgium, the Netherlands, and Switzerland) 1,3,7,8. Moreover, a progressive increase in NRP use has been registered over time even in those countries where NRP is not mandatory. In Spain, NRP is currently far more frequent than SRR for liver recovery 9. Nevertheless, this practice is still less embraced in the United States, although good results have been reported recently in a series of 13 DCD donors maintained on NRP 2. A few cases have also been reported from Russia and Korea 7,10.
NRP IN LIVER TRANSPLANTS
Clinical studies on LTs with NRP are reported in Table I.
Results of LT with NRP
NRP has originally allowed recovery and utilization of uncontrolled DCD liver grafts (uDCD; Maastricht category II and IV) in Spain, France, and Italy. Then its use has been shifted to controlled DCD grafts (cDCD; Maastricht category III). NRP turns the cDCD procedure into a more unhurried one compared to SRR, thus enabling graft evaluation and even warm dissection. Ischemic-type biliary lesions (ITBL) are the leading cause of patient morbidity and early graft loss in LTs from DCD. Many retrospective studies have shown that NRP treatment is effective in preventing the occurrence of ITBL compared to SRR, but randomized trials comparing these two techniques have not been reported yet 9,11. Nevertheless, given the ever-increasing use of NRP instead of SRR, such clinical trials are unlikely to be carried out in the future.
Selection of livers during NRP
Different criteria are used for graft selection during NRP, and progressive evolution of both parameters and thresholds has been noticed over time (Tab. II). A combination of macroscopic and microscopic assessment, alanine transaminase (ALT) levels, and lactate in perfusate are used to assess the suitability of the liver for transplantation in most protocols. All of these parameters inform about both liver viability and quality of the perfusion and have contributed to the selection of LT series with excellent results. Nevertheless, no strong correlation has been found between each parameter and the transplantation outcome 12,13. According to a recent review, the most used acceptation criterion in cDCD is the macroscopic aspect, while in uDCD ALT level is considered the most reliable 4. This attitude has been partially reflected in a recent Italian survey, where macroscopic assessment was highly considered in the opinion of the participants, along with the stability of NRP perfusion conditions 14. Although transaminase release is a widely accepted marker of liver injury, its cut-off has, however, been modified from initially 3-4 times the normal values to upper thresholds reported in the most recent series 11,12. Lactate clearance has been proposed as a parameter to assess liver function, as in normothermic machine perfusion, with a downward lactate trend indicating a well-functioning liver 4. However, Watson et al. noted that lactate leaking back from non-perfused areas in the donor decreases the reliability of this parameter as an indicator of liver function 15. Wang et al. have recently analysed flavin mononucleotide (FMN) in the perfusate during NRP and found that FMN levels were significantly higher in those livers that were declined for transplantation 16. Unfortunately, no correlation was made with the LT outcomes in the livers that were accepted. Further insight on this topic will probably be given by a nonrandomized trial on viability assessment during NRP, which is currently ongoing in France (NCT05361044).
NRP vs machine perfusion
Various dynamic ex-situ preservation strategies have been explored to ameliorate the outcomes of DCD livers, including hypothermic oxygenated perfusion (HOPE) and normothermic machine perfusion (NMP). Preference for in-situ NRP or ex-situ techniques varies among centres and countries. NRP is most frequently undertaken in tertiary hospitals with extracorporeal membrane oxygenation devices and cardiothoracic intensive care units, although good results have also been reported with mobile NRP teams 17,18. A few studies have compared NRP with HOPE or NMP, but no randomized trials have been published. In a large-scale international multicentric study, the utilization rate was significantly lower in the NRP group, despite shorter warm ischemic times and lower donor age compared to the HOPE group. However, after propensity-score adjustment of donor-recipient combinations, both strategies achieved similar posttransplant outcomes 19. In a single-centre retrospective analysis from the United Kingdom (UK), the NRP group had a lower incidence of cholangiopathy than static cold storage, but the same benefit was not achieved with NMP (NMP vs NRP: hazard ratio 3.5, p = 0.02) 17. Conversely, a multicentric study comparing NRP cases from France with NMP cases from the UK, Germany, Spain, and Belgium, failed to show any significant difference in the incidence of non-anastomotic biliary strictures (1.5 vs 2.9%; p > 0.99) and 30-day graft loss (4.4 vs 8.8%; p = 0.40) between the two groups 20.
NRP and subsequent HOPE or NMP
Prolonged cold ischemia and indication for retransplantation were found to be independent risk factors for graft loss among 545 DCD livers treated with NRP in Spain 9. These data support a potential role for complimentary ex-situ perfusion preservation for those cases with prolonged cold ischemia and/or technically complex recipients. The combined use of NRP with subsequent HOPE was first proposed in 2016 by our group to face the detrimental effects of the 20-min stand-off period in Italy, thus providing safe prolonged preservation and further reconditioning to the DCD livers 21. Although a direct comparison between NRP with subsequent HOPE and NRP only is still lacking in the Italian population, some indirect evidence exists to support this approach 14. In a recent multicentric analysis, the combined protocol has yielded good results compared to a static-preserved comparator group from the UK, despite the higher risk profile in Italy 12. Moreover, the same Italian cohort had shown similar results to the benchmark outcomes in LT from DCD22. The combined use of NRP and NMP was first reported by Pavel et al. 23. More recently, Ghinolfi et al. have proposed a flow-chart, where machine perfusion is used following the initial DCD liver evaluation during NRP. While HOPE is suggested for cases partially fulfilling the criteria, the authors recommend NMP when the criteria are not fulfilled 24. This proposal sounds promising but is still mainly based on a small number of uDCD cases, and only a few centres currently use both HOPE and NMP in Italy 14. The results of a currently ongoing randomized trial in Italy between sequential HOPE and NMP after NRP (NCT04744389) will hopefully provide further insight into this matter.
NRP IN KIDNEY TRANSPLANTS
A few clinical series on KT after NRP have been published, mainly from cDCD donors (Tab. III).
Results of KTs with NRP
Foss et al. compared the outcome of 14 DCD kidneys recovered with NRP with 163 transplants from donation after brain death (DBD) and observed no differences in delayed graft function (DGF) and 1-year graft survival between the groups 25. Similarly, Miñambres et al. reported nonsignificant differences in DGF and short-term graft survival comparing DCD kidneys treated with NRP with DBD controls 26. However, these studies did not compare the use of NRP with the widespread SRR. Ramirez et al. found that DCD kidneys treated with NRP had a lower rate of DGF than those with SRR 27. In a recent Spanish nationwide propensity score analysis, Padilla et al. have found that NRP was associated with improved rates of DGF and 1-year graft loss compared to SRR 28.
Selection of kidneys during NRP
For the acceptance of the kidneys during NRP, published reports mention macroscopic aspect, microscopic findings, and urine production. However, the absence of urine output is frequent during NRP and should not per se lead to organ discard 3. Rodríguez-Villar et al. investigated the evolution of biochemical parameters during NRP between accepted and discarded kidneys in 38 uDCD donors. Neither creatinine nor lactate sequential values was a useful tool to predict kidney allocation. Nevertheless, the authors did not correlate any posttransplant outcome with these variables 29. Ravaioli et al. reported the preliminary experience with sequential NRP and HOPE in 10 KTs from cDCD in Italy. They reported a 30% incidence of DGF and did not find any correlation with creatinine or lactate values during NRP. However, they found that lactate levels in the HOPE perfusate were significantly higher in those cases developing DGF 30. Marginal kidneys are usually assessed histologically with the Karpinski score, which however does not take into account the ischemic insult 30. In this context, Zagni et al. have recently reported that ischemic alterations of the proximal tubule are correlated with functional recovery in DCD kidneys 31. Centres using ex-situ hypothermic perfusion report using renal resistance for further selection, though studies outside the NRP field have shown kidneys should not be discarded based upon renal resistance only 32,33.
CONCLUSIONS
NRP is beneficial for both the liver and the kidneys from the same donor. Therefore, NRP improves organ quality and maintenance before cold preservation, turns the DCD procedure into a more unhurried one, and allows the assessment of organ function following the warm ischemic injury. Different parameters inform about both graft viability and quality of the perfusion, but no strong correlation has been found between each parameter and the transplantation outcome. Nevertheless, a combination of macro-microscopic assessment, biochemical, and perfusion parameters has contributed to the selection of liver and kidney transplant series with excellent results.
Acknowledgments
The Authors want to thank Elisa Lodi for her support in figure editing.
Conflict of interest statement
The Authors declare no conflict of interest.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Authors’ contributions
RDC performed data collection, interpreted data, and wrote the paper; LC, MM, LP, RC, and IV performed data collection and reviewed the paper; AL and LDC critically reviewed the paper.
Ethical consideration
The present study did not imply any direct investigation on humans or animals. Formal consent was not required.
Figures and tables
Figure 1. NRP circuit and advantages. The abdominal NRP circuit relies on extracorporeal membrane oxygenation technology and includes a pump, an oxygenator, and a heater. The blood is pumped through arterial and venous cannulae in the femoral vessels. Supraceliac aortic balloon occlusion prevents cerebral reperfusion. aNRP: abdominal normothermic regional perfusion.
| Author, year | Country | DCD type, n | Utilization rate (from NRP to LT) | EAD | PNF | AKI | ITBL | Graft survival (follow-up) | Patient survival (follow-up) |
|---|---|---|---|---|---|---|---|---|---|
| uDCD | |||||||||
| Fondevila, 2012 34 | Spain | uDCD: 34 | 34/290 (12%) | N/A | N/A | N/A | 8% | 70% (1y) | 82% (1y) |
| Savier, 2015 35 | France | uDCD: 13 | 13/183 (7%) | 31% | 23% | N/A | 8% | 69% (1y) | 85% (1y) |
| Justo, 2020 36 | Spain | uDCD: 75 | N/A | N/A | 8% | 20% | 4% | 78.3% (1y) excluding PNF | 82% (1y) excluding PNF |
| cDCD (multicentre/national) | |||||||||
| Watson, 2019 11 | UK | cDCD: 43 | 61% | 12% | 0% | N/A | 0% | 97.7% (90d) | 100% (90d) |
| Savier, 2020 37 | France | cDCD: 50 | 20% | 18% | N/A | 26% | 2% | 88% (2y) | 90% (2y) |
| De Carlis, 2021 12 | Italy | cDCD: 44 | 85% | N/A | 5% | 36% | 2% | 91% (2y) | 98% (2y) |
| Hessheimer, 20219 | Spain | cDCD: 545 | 70% | 15% | 3% | N/A | 1% | 90% (1y) | 92% (1y) |
| Sellers, 2022 2 | US | cDCD: 13 | N/A | 23% | 0% | N/A | 0% | 92% (439d) | 92% (439d) |
| Mixed series (uDCD and cDCD) | |||||||||
| De Carlis, 2018 38 | Italy | cDCD: 6 | 20/25 | 24% | 10% | 30% | 10% | 85% | 95% |
| uDCD: 14 | (80%) | (1y) | (1y) | ||||||
| Ghinolfi, 2020 24 | Italy | cDCD: 7 | 18/31 | 28% | 0% | 28% | 6% | 94% | 94% |
| uDCD: 11 | (58%) | (15mo.) | (15mo.) | ||||||
| Author, year | Type | Pump flow | Lactate | Transaminase | Macroscopical aspect | Liver biopsy |
|---|---|---|---|---|---|---|
| Fondevila, 200739 | uDCD | >1.7 L/min | - | Initial < 3 ULN | Before/after cold flush | - |
| (>4h) | Final > 4 ULN | |||||
| Oniscu, 201413 | cDCD | 1.7-4 L/min/m2 | - | Initial ALT < 3 ULN | - | - |
| (60-120 min) | Final ALT < 4 ULN | |||||
| Savier, 201535 | uDCD | efficient flow for > 240 min | - | ALT at 2h < 200 IU/L | - | MaS < 20% |
| De Carlis, 201740 | uDCD | - | Stable or downward | ALT < 1000 IU/L | Color, surface, margins, consistency | MaS < 30% |
| cDCD | Ishak 0-1 | |||||
| Watson, 201911 | cDCD | 2.5-3 L/min | Fall is encouraging | ALT < 200 IU/L | Steatosis | - |
| ALT < 500 IU/L | ||||||
| Savier, 202037 | cDCD | > 60 min | - | AST-ALT < 200 IU/L | - | MaS < 20% |
| Hesseimer, 20219 | cDCD | 2.2-2.4 L/min/m2 | Ideally downward | Ideally stable and < 200 IU/L | Liver, gallbladder, bile duct and bowel | - |
| (60-120 min) | ||||||
| De Carlis, 202112 | cDCD | 1.7-3 L/min/m2 (ideally 120 min, no upper limit if stable) | Ideally stable or downward | Ideally final ALT < 1000 IU/L | Perfusion, congestion | Ideally MaS < 30% Ishak 0-2 |
| Author, year | Study type | n | Donor fWIT (min) | MP | PNF | DGF | Graft survival (follow-up) | Patient survival (follow-up) |
|---|---|---|---|---|---|---|---|---|
| Foss, 2018 25 | Single-centre, retrospective, observational | NRP: 14 | NRP: 26.5 (20-49) | No | NRP: 0% | NRP:7.1% | NRP: 93% | N/A |
| DBD: 163 | DBD: 0% | DBD: 4.9% | DBD: 95% (1y) | |||||
| Miñambres, 2017 26 | Single-centre, retrospective, observational | NRP: 37 | NRP: 12 (10-19) | No | NRP: 5% | NRP: 27% | NRP: 91.8% | N/A |
| DBD: 36 | DBD: 0% | DBD: 33.3% | DBD: 97.2% (18mo.) | |||||
| Ravaioli, 2018 30 | Single-centre, retrospective, observational | NRP: 5 | NRP: 151.2 (40-325) | HOPE | NRP: 0% | NRP: 30% | NRP: 100% (6mo.) | NRP: 100% (6mo) |
| Padilla, 2020 28 | Nation-wide, retrospective, observational | NRP: 865 | NRP: 13 (10-17) | NRP: 15.9% | NRP: 4.8% | NRP: 30.3% | NRP: 93.1% | NRP: 97.6% |
| SRR: 1437 | SRR: 18 (13-3) | SRR: 7.3% | SRR: 4.4% | SRR: 48.4% | SRR: 91.5% (1y) | SRR: 95.6% (1y) | ||
| Ramirez, 2021 27 | Single-centre, retrospective, observational | NRP: 22 | NRP: 10 (10-35) | No | NRP: 4,55% | NRP: 36.36% | NRP: 91% | NRP: 77.27% |
| SRR: 62 | SRR: 15 (11-28) | SRR: 6,45% | SRR: 46.77% | SRR: 87% | SRR: 88.71% | |||
| DBD: 98 | DBD: 10,20% | DBD: 20.41% | DBD: 84,4% (15mo.) | DBD: 85.71% (1y) |
PDF
