Kidney transplantation (KT) remains the best available therapeutic option for patients suffering from end-stage renal disease. The universal scarcity of deceased donor organs as well as the demographic change have resulted in a continuous increase in the proportion of elderly donors and donors with comorbidities over the last decades [1]. Nowadays, more than 40% of donors in the Eurotransplant area are expanded criteria donors (ECD) and more than 25% of donors are aged ≥65 years [2–4]. Despite these developments, waiting times remain a challenge so that further measures to expand the donor pool are necessary.

Cardiopulmonary resuscitation (CPR) is a major reason for admission to intensive care units and - despite best efforts - it is often followed by irreversible brain damage and brain death. Several studies have demonstrated that kidneys from younger donors (aged typically 18–50 years) with a history of CPR can be safely transplanted without compromising long-term allograft survival [5, 6]. Whether the same is true for the particularly vulnerable but growing group of ≥65-year-old donors is largely unknown even though nowadays transplant centers regularly receive such offers. A recent, single-center analysis from our group had suggested no negative effect of a history of CPR on allograft survival and function in KTs from ≥65-year-old donors [7]. However, additional data are needed to better guide management decisions in this challenging group of high-risk CPR donors. To this end, we performed a multi-center retrospective study to analyze outcomes of KTs from ≥65-year-old donation after brain death (DBD) donors with and without a history of CPR.

A total of 680 KTs from 569 DBD donors were included in this study. Of these, 81 KTs (11.9%) were from 68 donors with a history of CPR prior to organ donation (CPR group). The remaining 599 KTs originated from 501 donors that were not resuscitated prior to transplantation (no-CPR group).

Donor demographics are shown in Table 1. Donor age was slightly lower in the CPR group compared to the no-CPR group (median 70.0 vs. 72.0 years, p = 0.02). Donors in the CPR group died more often from non-cerebrovascular and non-traumatic causes (primarily anoxic brain damage; 42.6%), whereas donors in the no-CPR group died mostly from intracranial bleeding (58.9%, p < 0.001). No significant differences were observed regarding a donor history of arterial hypertension, diabetes, or smoking.

Donors with a history of CPR had significantly lower admission, minimum and final eGFR levels compared to the no-CPR group (on admission: 66.0 mL/min vs. 84.4 mL/min, p < 0.001; minimum: 60.5 mL/min vs. 71.3 mL/min, p < 0.001; final: 67.6 mL/min vs. 81.0 mL/min, p = 0.004). Consequently, a donor AKI was seen significantly more often in the CPR group (45.6% vs. 23.4%, p = 0.001). Diuresis in the last 24h prior to organ recovery was also slightly lower in the CPR group (3.0 L vs. 3.6 L, p = 0.03) but no donor in either group was oliguric (i.e., diuresis <500mL/24h), anuric or on hemodialysis. Interestingly, an ongoing AKI at organ recovery was present significantly more often in the no-CPR group than in the CPR group (64.0% vs. 39.4%, p = 0.02).

The median duration of CPR was 20 min with 80% of all CPRs lasting ≤30 min. The vast majority of CPRs occurred out-of-hospital (85%). Stratification of the CPR group according to the median CPR duration revealed that donors in the long CPR group were younger than donors in the short CPR group and the no-CPR group (70 vs. 72 vs. 72 years, p = 0.04). Also, a donor history of diabetes and smoking were each significantly more frequent in the long CPR group (Supplementary Table 1). The other results were similar as observed for the whole CPR group.

Recipient characteristics are displayed in Table 2. The median recipient age was 67.0 years in the CPR group and 67.0 years in the no-CPR group (p = 0.88). No significant differences were found regarding the duration of dialysis prior to KT, underlying renal disease, history of diabetes mellitus or arterial hypertension, number of HLA mismatches, cold ischemia time (CIT) and baseline immunosuppression. The results for the subgroup analysis stratified according to the duration of donor CPR revealed the same results and are displayed in Supplementary Table 2.

The incidence of DGF was comparable between the two groups (CPR group: 27.2% vs. no-CPR group 33.1%, p = 0.40). PNF also occurred at similar rates (CPR group: 9.9% vs. no-CPR group 12.2%, p = 0.19). Both groups had similar median follow-up time (CPR group 37.5 months vs. no-CPR group 41.0 months; p = 0.30). During this follow-up time, graft function did not show significant differences at any assessed time point (3 months, 1 year, 3 years and 5 years) between the two groups (Table 3; Figure 1). The incidence of BPAR within the first three years post KT was also very similar (CPR group 41.9% vs. no-CPR group 42.0%, p = 0.31). The subgroup analysis stratified according to duration of CPR revealed essentially the same results (Supplementary Table 3). The DGF rate was 25.6% in the short CPR group, 28.6% in the long CPR group and 33.1% in the no-CPR group (p = 0.38). Graft function and incidence of BPAR were also not different between the short CPR group, the long CPR group, and the no-CPR group.

Five-year death-censored graft survival and patient survival were comparable between donors with and without a history of CPR (p = 0.24 and p = 0.61 respectively, see Figures 2A,B). After stratification into donors with short and long CPR we neither observed differences regarding 5-year death-censored graft nor patient survival between the groups (Figures 3A,B).

For further validation of our results, we subsequently performed a multivariable Cox regression analysis using a multiple imputation approach with adjustment for known risk factors and confounding factors of death-censored graft loss and death (Table 4). Random effects were included to capture the influence of the three transplant centers and to account for the non-independence of kidney transplants originating from the same donor. CPR did not increase the risk of death-censored graft loss (Hazard Ratio [HR] 0.76, 95% confidence interval [CI] 0.48–1.21, p = 0.25) or death (HR 0.77, 95% CI 0.47–1.26, p = 0.30). The same multivariable Cox regression analyses were repeated after subgroup stratification according to the duration of CPR. Neither short nor long duration of CPR were found to be a risk factor for death-censored graft loss or death (Supplementary Table 4).

Even when performed optimally, CPR results in global hypoperfusion and ischemia. In turn, successful resuscitation is frequently accompanied by a multisystem injury and a systemic inflammatory response resembling severe sepsis [14, 15]. This response constitutes an additional insult to the inevitable ischemia-reperfusion injury intrinsic to organ transplantation. Increased susceptibility of organs from elderly donors has been postulated, and grafts from elderly donors with preceding CPR have therefore been subjected to heightened concerns regarding organ acceptance [16, 17].

Prior studies on the impact of CPR in KT had almost exclusively focused on younger donors (aged 18–50 years) and had observed no negative effect of a history of CPR on long-term allograft survival and function [5, 6]. In response to the ongoing organ shortage and the demographic change in our society, organ offers from far elder donors with a history of CPR have become common place. However, so far information on the outcome of such donors was very scarce. This provided the rationale for our study.

Our results demonstrate comparable allograft function as well as comparable death-censored graft survival and patient survival for KTs from ≥65-year-old DBD donors with and without a history of CPR. The incidence of BPAR was also not different. Stratification according to duration of CPR did not change the results–also for the long CPR group in which CPR duration was ≥20 min.

While these results are encouraging, we feel it is necessary to point out several selection criteria among the CPR donor group that we believe to be of high importance for the positive results of our study. First, all KTs assessed in this study were from DBD donors as DCD is prohibited in Germany. This is important as for DCD donors, worse allograft kidney function has been demonstrated also for younger CPR donors in some previous studies [18, 19]. Second, regarding CPR itself, median duration was fairly short (20 min), and the large majority of CPR donors had been resuscitated for ≤30 min (80%). While we did not observe a negative effect of CPR on allograft survival and function in the long CPR group, it is our belief that a CPR duration of >30 min in a ≥65-year-old DBD donor requires a very close scrutiny of the additional donor information before acceptance. This leads us to the third selection criteria. In our view, CPR as a donor characteristic cannot be regarded in isolation. Instead, we feel it is crucial to assess for CPR-associated organ damage (in the setting of KT especially severity of AKI and residual diuresis). In this regard, it is important to point out that even though the incidence of AKI was fairly high among CPR donors in our study (45.6%), most donors (84%) had only stage I AKI and none were oliguric, let alone anuric or on hemodialysis. Consequently, the median diuresis in the last 24 h prior to organ recovery was 3.0 L among CPR donors. Of note, we recently assessed the impact of donor AKI in ≥65-year-old DBD donors and found no difference in allograft survival and function for KTs from (primarily stage I) AKI donors with preserved diuresis [20]. Next to these CPR-associated aspects, we also consider CIT, donor comorbidities (especially arterial hypertension/diabetes) and donor age carefully in the selection process. Concerning these aspects, CIT in our study was kept short (median 10.2 h for CPR donors) in order to minimize the magnitude of ischemia-reperfusion injury (see section below) [4]. Also, CPR donors were slightly but significantly younger than no-CPR donors (70 years vs. 72 years) and almost all CPR donors were <80 years old (97%). We believe that these factors represent important donor selection criteria among ≥65-year-old DBD donors with a history of CPR that should be taken into account when considering a kidney offer of this kind. Therefore, our results do not imply that every ≥65-year-old CPR donor is a suitable candidate for KT. It has to be stressed that every organ offer needs to be assessed on a case-by-case basis after assessing the aforementioned factors as well as all other available donor information.

Another important aspect is recipient selection. The majority of kidneys in this study (>80%) were allocated through the Eurotransplant Senior Program (ESP). Within the ESP, kidneys from ≥65-year-old donors are allocated to ≥65-year-old recipients preferentially in the same geographical region and without HLA-matching [21, 22]. Therefore, recipients and donors were of comparable age (median donor age: 70 years in the CPR group and 72 years in the no-CPR group; median recipient age: 67 years in the CPR group and 67 years in the no-CPR group). The regional allocation also explains the short CIT in both groups which is a typical characteristic of the ESP [21, 23, 24]. Hence, we would suggest to offer a kidney from a ≥65-year-old CPR donor preferentially to a similarly aged recipient. Of note, the regional allocation is also the reason why, in 19.5% of donors, we were offered both kidneys from the same donor and subsequently transplanted them into two separate recipients. In turn, the number of KTs (n = 680) was higher than the number of kidney donors (n = 569).

There have been previous studies on the effect of donor CPR on kidney allograft survival [5, 6]. The most recent meta-analysis by Sandroni et al. showed no difference in allograft survival between 950 DBDs with a history of CPR and 12,719 DBDs without a history of CPR after 1 year and actually even a small survival advantage at the longest follow-up (even though this time point was not defined more clearly; p = 0.04) [6]. However, virtually all studies included in the meta-analysis had focused on younger donors (aged primarily 18–50 years) which is why the median donor age of 70 years in our CPR group was 20 years higher than in any other study from the meta-analysis. Yet, nowadays, kidney transplant physicians in countries with low organ donation rates (such as Germany) regularly receive kidney offers from ≥65-year-old DBD donors with a history of CPR which highlights the knowledge gap about this vulnerable donor group. The only other publication specifically analyzing the effect of CPR in ≥65-year-old DBD donors came from our own group [7]. However, this previous paper had several shortcomings as it was a single-center analysis with a considerably smaller sample size (n = 185) and covering a longer time span (1999–2019) in which mycophenolate-based immunosuppression and induction therapy was not yet standard of care. In contrast, the current data set was derived from three transplant centers comprising more than three times the previous sample size (n = 680) and covering a different time period (2006–2023), in which mycophenolate-based immunosuppression and induction therapy had become standard of care. These facts make the current paper more reliable and more relevant to everyday practice.

Of note, we also report no significant difference in the incidence of DGF and PNF between the CPR and no-CPR groups (27.2% vs. 33.1% and 9.9% vs. 12.2%, respectively). These numbers are comparable to previously published data on ESP donors and ECDs aged ≥65 years [21, 25, 26]. The aforementioned meta-analysis also showed no significant difference in the incidence of DGF between the CPR and the no-CPR groups. Mechanistically, one could speculate that successful CPR allows the restoration of physiological organ function already during the stabilization phase in the intensive care unit, when considerations regarding organ donation are being undertaken. This may in turn reduce the incidence of DGF. Ultimately, we can only speculate about the reasons.

It is important to realize that the results of this study may have been influenced by donor selection bias because we can only report on the results of those kidney organ offers that had been accepted for KT in the first place. Offers that had been declined were not available for outcome analysis.

In summary, we demonstrate comparable 5-year death-censored graft survival as well as up to 5-year allograft function between KT from ≥65-year-old DBD donors with and without a history of CPR. These encouraging results indicate that KT from ≥65-year-old CPR donors is a viable option especially when the donors have no or only stage I AKI with preserved diuresis and both the duration of CPR as well as the expected CIT are short (ideally <30 min and <12 h, respectively). These results may help to further enhance the pool of suitable elderly kidney donors.