Graft and patient survival with living donation (LD) is better than donation after brain death (DBD) [1–6]. Data from a UK transplant registry showed that all-cause mortality was lower for recipients of older LD kidneys (aged ≥60 years) than standard-criteria DBD (DBD-SC), but graft and overall survival were lower for LD recipients with older living donors rather than younger (aged < 60 years) [7]. DBD and LD transplantations have significant differences that may affect post-transplantation survival. One of the major advantages of LD transplantation is that it allows for pre-emptive transplantation. The age and immunological profile of LD recipients may also differ from those of DBD recipients.

At first glance, a short cold ischemia time and very good health of the donor seem to result in higher eGFR after LD than DBD transplantation. Alternatively, these expected benefits of LD over DBD may be counterbalanced by the better age and immunological matching between the donor and recipient in DBD than in genetically and emotionally related LD. Altogether, LD and DBD have different graft access procedures and clinical characteristics that may affect the outcomes of kidney transplantation (KT) independent of the donation type [1].

From our annual medical and scientific report [8], for 44% of LD recipients, the eGFR at 1 year was >60 mL/min as compared with 51% for DBD-SC recipients. This unfavorable outcome for LD prompted us to conduct this study.

In the context of the various pros and cons for each of the two strategies, we compared the impact of DBD and LD on eGFR at 1 year after transplantation using propensity score (PS) matching to attempt to mimic a randomized trial [9]. Because age is an important element in the choice of donor and eGFR interpretation, we conducted several sensitivity analyses to explore this confounding factor. We analyzed eGFR with DBD and LD by recipient age and donor subgroup, namely, standard criteria and expanded criteria. Secondary outcomes were graft and patient survival at 4 years.

LD transplantation is the preferred treatment for kidney failure, offering a priori superior patient and graft survival as compared with DBD transplantation. However, LD recipients are usually not comparable to DBD recipients, LD recipients typically being younger and having no or less pre-transplant dialysis duration. We attempted to emulate a target trial by creating a PS-matched cohort to investigate eGFR at 1 year after KT for DBD and LD recipients, ensuring comparability between the LD and DBD groups in recipient age, sex, blood group, pretransplant dialysis duration, and cPRA. The eGFR at 1-year post-transplant is widely considered the most relevant marker for predicting graft and patient survival after transplantation and is extensively used in randomized clinical trials [15]. At 1 year after KT, eGFR was significantly higher for LD than DBD recipients. Specifically, the OR for attaining an eGFR ≥45 mL/min/1.73 m² was 2.12 times higher and an eGFR ≥60 mL/min/1.73 m² was 1.40 times higher for LD than DBD recipients.

However, our study suggests that the superiority of LD over DBD in terms of eGFR is not consistent across all recipient age groups. Among recipients <40 years, the OR for an eGFR ≥60 mL/min/1.73 m² did not significantly differ between DBD and LD recipients. For recipients ≥60 years, the OR for an eGFR ≥60 mL/min/1.73 m² was 5 times higher for LD than DBD recipients (0.91 vs. 0.18 for younger LD recipients).

Several factors may explain this difference in eGFR between DBD and LD recipients based on recipient age. LD recipients <40 years frequently receive a kidney from an older LD donor, a situation that contrasts with the DBD allocation strategy, which often favors age matching. As a result, younger DBD recipients typically receive kidneys from young DBD donors [16]. This situation is supported by the greater delta donor age–recipient age in the LD than DBD group. In contrast, DBD recipients ≥60 years mainly receive kidneys from donors within the same age group, whereas LD recipients ≥60 years may receive kidneys from younger donors [16]. Additionally, DBD recipients ≥60 years may receive kidneys from DBD-EC donors, whereas LD recipients usually have few or no comorbidities. Hypertension and kidney aging are associated with a higher proportion of sclerotic glomeruli and nephron loss, leading to lower eGFR after KT [17–19]. Therefore, kidneys from an older LD donor may result in a lower eGFR than kidneys from a younger DBD donor. Our sensitivity analyses showed that a threshold eGFR of 80 mL/min/1.73 m² was more common among DBD than LD recipients, who are often matched with younger donors. In contrast, at 1 year after KT, the mean eGFR was approximately 50 mL/min/1.73 m² for recipients from LD donors ≥60 years old, regardless of recipient age.

At 1 year after KT, donor age has been found correlated with renal function as well as long-term graft and patient survival [20–24]. Lim et al. [25] reported an association between donor age, 1-year eGFR, and overall graft loss. However, eGFR <60 mL/min/1.73 m² at 1 year with a kidney from an older LD is expected and should not preclude the selection of an LD donor as suitable [26, 27].

In our matched cohort, the analysis of graft and patient survival at 4 years post-KT highlights the inferior outcomes of transplantation from a DBD-EC. Indeed, at 4 years, the graft survival was approximately 15% lower for recipients of a DBD-EC than for other recipients. The graft acceptance criteria are more extensive in France than in the United States [28, 29]. However, graft survival from a DBD-EC donor in France is comparable to literature data and is considered satisfactory as compared with dialysis maintenance [17, 30]. In older patients, long-term results were found better with LD than DBD-SC or DBD-EC and suggested the use of LD transplants in older patients whenever possible [18]. However, in Japan, the age of the oldest LD is high (70–89 years) and outcomes were found to be poor in terms of graft survival and eGFR for older recipients from very old LD donors [31]. In the same way, from UK registry data, all-cause mortality was greater for recipients of older LD (donor age ≥70 years) than DBD-SC and was equivalent to that for DBD-EC recipients [7].

Our study has some limitations. It was conducted in France, where national practices for DBD allocation and organ acceptance differ from those in other countries. Different allocation scores and stricter kidney acceptance criteria in other regions may yield different results between DBD and LD recipients. The particular strength of our work lies in its methodology, effectively used in other studies [32, 33], emulating a target trial that is not feasible in the real world. This method allows for defining all the conditions required for the target trial and precisely describing the deviations from it. However, some important factors may not have been taken into account in our PS. Furthermore, we excluded recipients without timely follow-up or who died in the first week before matching. Because they represented less than 10% of the target population and their distribution was not different between LD and DBD recipients, we considered the resulting selection bias to be negligible.

A second limitation is the lack of long-term follow-up. Nevertheless, we needed to begin inclusions in 2015 to have both a homogeneous donor population (sharp increase in age of donors since 2000) and the same kidney allocation system throughout the cohort (the current kidney allocation system was implemented in 2015).

A third limitation could be our choice to conduct stepwise regressions based on p-value and involving multiple comparisons; we opted for this choice because of our sample tail that allowed for a high number of degrees of freedom. Furthermore, we opted to categorize all our continuous variables, which implies loss of information and precision [34, 35]. We made this choice after discussion with clinicians who preferred having information directly usable with the references they used in practice. To test the robustness of our results, we conducted a sensitivity analysis performed with Harrell’s recommendations and found consistent results.

A final limitation is the use of eGFR at 1 year after KT as a single surrogate marker of graft outcome. Indeed, predicted graft survival based on this surrogate marker is correlated with observed graft survival [36] and this parameter is used in studies testing new treatments [37]. However, graft injuries may develop slowly over time and eGFR at 1 year fails to capture ongoing disease process [38]. In our matched cohort on recipient characteristics, we tested eGFR at 1 year after KT depending on donor type stratified by age. So, we discussed eGFR at 1 year as a marker potentially reflecting nephron loss more than nephron injury that leads to graft failure.

Our study showed that older recipients derive significant benefits from LD transplants, which emphasizes the importance of evaluating living donors ≥60 years old. Conversely, transplants from DBD donors can yield good outcomes for young adult recipients, provided that there is a suitable age and HLA match, along with prompt access to transplantation. Of note, our matched cohort reflects a notably short pre-transplant dialysis period, a characteristic often associated with LD transplants. Paired-exchange programs offer a viable avenue to explore improved age matching, particularly with a significant age gap between the donor and recipient. However, KT for compatible donor-recipient pairs seeking a better match in terms of age should not be delayed too long, and the search for a better match should be carried out early.

Our study is the first published research to use an emulated target trial to compare LD and DBD recipients. These findings offer valuable insights for healthcare professionals, empowering them to make well-informed decisions regarding the suitability of different donor types for specific recipients.