Uncontrolled donation after circulatory death: comparison of two kidney preservation protocols on graft outcomes

We conducted a retrospective study in an academic hospital where patients received kidney grafts from uDCD. All cases were consecutive recipients and we collected data on each uDCD recipient transplanted between September 2006 and September 2013 using medical records. We included 27 patients from a previous study, all preserved by ISP [11]. Kidneys were retrieved from two French hospitals: Edouard Herriot public teaching hospital in Lyon, and the public hospital of Annecy, France. We did not include recipients transplanted outside our centre with grafts from those 2 retrieving centres. The end of the follow-up period was October 2015; all patients were followed for 2 years.

uDCD inclusion criteria were based on the French uDCD program, as follows: out-of-hospital cardiac arrest (Maastricht I) or in-hospital cardiac arrest (Maastricht II), with a unsuccessful resuscitation, a precise time of cardiac arrest, a duration without cardiopulmonary resuscitation less than 30 min (no flow period), age ≥ 18 to ≤55 years and an interval before preservation protocol initiation <150 min [20]. Donors with history of chronic kidney disease, hypertension, diabetes, sepsis, neoplasm, intravenous drug addiction, or traumatic cardiac arrest were not included. We also did not include patients eligible for extracorporeal life support or prolonged cardiopulmonary resuscitation (no flow less than 5 min, capnography greater than 15 Torr (2 kPa) after 20 min of resuscitation, hypothermia, drug intoxication, or signs of life during cardiopulmonary resuscitation).

Recipient criteria were age ≤ 60 years, first kidney transplantation, ABO compatibility, no previous HLA-sensitisation, and agreement to potentially receive a graft from an uDCD.

Patients with out-of-hospital cardiac arrest were handled on-site by the emergency medical services with advanced life support. If fulfilling the criteria for uDCD, patients were referred to the uDCD center of our institution, under mechanical ventilation and continuous external cardiac massage. Upon arrival, death was certified by a 5-min electrocardiogram confirming the absence of spontaneous cardiac activity. Standard blood tests were then performed as a conventional prerequisite for donation. The entire uDCD procedure of our institution is described in Fig. 1. From September 2006 to January 2010, after confirmation of death, the ISP preservation protocol was initiated. An intraaortic double-balloon catheter (Gillot catheter) and a venous vent were surgically inserted through the right side of the groin, with an injection of 25,000 IU of heparin. The intraaortic catheter was perfused with a heparinized (5000 IU/L) preservation solution at 4 °C (IGL-1, Institut Georges Lopez, Saint-Didier-au-Mont-d’Or, France) at a rate of 250 to 500 mL/min until blood washout, and maintained at 100 mL/min. Finally, peritoneal refrigeration with 4 L isotonic saline at 4 °C was performed. ISP duration had to be less than 180 min. From February 2010, we used cardiopulmonary bypass for organ procurement as follows: after cannulation of the femoral artery and vein, cannulas were connected to a blood oxygenator, a heat exchanger and a non-pulsatile roller pump. This NR was run at the minimum pump flow rate of 2 L/min, with an oxygen concentration set to 100%. NR duration had to be less than 240 min. The automated national registry for organ donation refusal was consulted (registre national des refus), and donor family consent was obtained for all uDCD kidneys.

Cold ischemia time after kidney retrieval was managed with a pulsatile hypothermic (1–4 °C) machine perfusion (RM3, Waters Medical Systems, Rochester, MN, USA, or LifePort, Organ Recovery System, Des Plaines, IL, USA) and the organ preservation solution was the Belzer MPS Solution (Waters Medical Systems) or KPS-1 (Organ Recovery System). Grafts were preserved by this machine perfusion until the end of the cold ischemia time. Organ viability was assessed by measuring ex vivo intrarenal vascular resistance index.

Patients received antithymocyte globulin (Thymoglobulin, Genzyme, Cambridge, MA, USA) for induction therapy (1.25 mg/kg/day for 10 days). Maintenance immunosuppression included three types of drug. Calcineurin inhibitors were introduced on day 6–8; cyclosporine (Neoral, Novartis Pharma AG, Basel, Switzerland) or tacrolimus (Prograf, Astellas Pharma Inc., Tokyo, Japan) doses were adjusted to obtain trough levels between 100 and 150 ng/mL and 8–12 ng/mL, respectively, during the first year. Mycophenolate mofetil 2 g/day adapted to patient tolerance (Cellcept, Roche Pharmaceuticals, Basel, Switzerland). Steroids were quickly tapered to 5 mg/day within the first 2 months after transplantation. Cyclosporine was used for patients with body mass index >25 kg/m² and/or past medical history of diabetes, and tacrolimus for the others. This protocol remained unchanged during the study period.

Graft and patient survival at 24 months (M24) were analysed, graft loss being defined by requirement of chronic dialysis. Short-term outcome was evaluated by PNF, DGF rates, duration of dialysis after transplantation, and number of days to obtain a urine output above 1000 mL. PNF was defined as an immediate and permanent non-function of the graft after transplantation requiring chronic dialysis (and chronic dialysis was defined by more than 3 months of dialysis). DGF was defined as the need for at least one dialysis session during the first week after transplantation, except for post-transplantation hyperkalaemia. Measured GFR (mGFR) was performed with inulin or iohexol clearance at M12. We assessed graft function by estimated GFR (eGFR) with simplified modification diet renal disease (MDRD) formula (at 1, 3, 6, 12, and 24 months after transplantation: M1, M3, M6, M12, and M24) [21]. Clinical acute rejection was defined as a serum creatinine increase of more than 20% over baseline. These acute rejections were biopsy-proven where possible. Subclinical rejection without creatinine increase was diagnosed with protocol biopsies at M3 and M12.

Protocol graft biopsies were performed at M3 and M12. Kidney core biopsies were screened by the same pathologist according to the Banff 07 classification [22]. Masson-stained biopsy slides were subsequently digitised to quantify interstitial fibrosis by automatic colour image analysis, as previously described in a robust and reproducible manner [23, 24].

Categorical variables were compared using the Chi square test or Fisher’s exact test when the conditions of application of Chi square test were not met. Quantitative variables were compared between groups using Student’s t test after verification of equality of variances when data were normally distributed, and with the Wilcoxon nonparametric test when the hypothesis of normality of distribution was not verified.

Survival curves for graft and patient survival were obtained using the Kaplan-Meier method and compared using the log-rank test. Graft survival was determined between date of transplantation to date of chronic dialysis requirement. Patient overall survival was defined by date of death.

Comparison of mean mGFR values at 12 months between the two procedures were adjusted for recipient age and gender, cold ischemia time, and NR/ISP duration using a MANOVA (GLM SAS procedure).

A total of 64 renal transplant recipients from uDCD were included, all Maastricht I category. They were classified according to their kidney donor preservation group: ISP (n = 32) and NR (n = 32). Donor characteristics were not statistically different between the two groups. The number of HLA mismatches were more frequent in the NR group (mean ± SD = 4.8 ± 1.2 vs. 4.1 ± 1.1; p = 0.04), and duration of NR was longer than that of ISP (mean ± SD = 203 ± 46 min vs. 165 ± 44 min; p = 0.001), whereas duration of cold ischemia was longer in the ISP group (mean ± SD = 1027 ± 250 min vs. 817 ± 211 min; p = 0.001; Table 1).

Overall survival at M24 was not statistically different (log-rank test: p = 0.14) between the ISP and NR groups, respectively 100 and 92.6% (one patient died at M8 from pneumonia, another one at M15 after severe trauma). Kidney graft survival at M24 was also not statistically different (log-rank test: p = 0.27) between the ISP (96.8%) and NR (96.5%).

We observed one PNF in each group. The incidence of DGF was not statistically different between groups (ISP: n = 27, 84% and NR: n = 23, 72%; p = 0.23). There were five complications in the ISP group (one arterial stenosis, two ureteral stenoses, two venous thromboses) and six in the NR group (five arterial stenoses and one postoperative haemorrhage). The median [IQR] duration of hospital stay for patients receiving kidneys preserved by NR was significantly 1 week shorter than that of patients receiving kidneys preserved by ISP (17 [15–21] vs. 24 [18–33] days, p = 0.003; Table 2).

The one-year mean ± SD mGFR of the NR group (n = 24) was 53.8 ± 12.8 mL/min/1.73 m² and that of the ISP group (n = 28) was 43.0 ± 12.8 mL/min/1.73 m² (p = 0.007). The multivariate analysis confirmed this finding; the difference remained significant after adjusting for recipient age and gender, cold ischemia time and NR/ISP duration (p = 0.03), see Additional file 1. Mean eGFR was statistically different between groups at 1 year (p = 0.01) and 2 years (p = 0.03) favouring NR, but not at day 15 (p = 0.36; Figure 2).

The total number of adequate protocol biopsies was 34/64 at M3 and 32/64 at M12. These were not performed for 36% of patients at M3 and 42% of patients at M12 due to patient refusal, death, anticoagulant treatment, or post-transplantation arterio-venous fistula; other biopsies were not sufficient to perform histological assessment and were discarded.

The incidence of clinical acute rejection did not statistically differ between the ISP and the NR groups (19% vs. 18%; p = 0.93), but the incidence of borderline changes observed on the systematic biopsies was statistically more frequent in the ISP group than in the NR group (56% vs. 22%, p = 0.002). The mean interstitial fibrosis score was not different between groups at M3, or at M12. However, there was a trend towards lower interstitial fibrosis in the NR group at M12 (mean ± SD = 0.30 ± 0.10 vs. 0.37 ± 0.12). Concerning the Banff 07 score, the distribution of cv score at M3 and at M12 was not significantly different between the 2 groups (Table 3).