This retrospective study involved deceased donors and recipients who had undergone successful kidney transplantation in our transplant center between 2016 and 2019. No organs were sourced from executed prisoners. Donors were excluded if they fulfilled any of the following criteria: (a) lacked a blood specimen, (b) had contaminated blood specimens affecting CIRBP levels, (c) were recipients who underwent combined multi-organ transplants, (d) were participating in other clinical trials, (e) had a history of organ transplantation, or (f) had other conditions deemed unsuitable for enrollment by the investigator. Blood specimens were collected from donors by Organ Procurement Organizations and used in various studies with the donors’ consent. All recipients were followed-up for a long period post-transplantation.

This study was conducted in accordance with the ethical standards outlined in the Declaration of Helsinki (22) and approved by the Medical Ethics Committee of The Third Affiliated Hospital of Sun Yat-sen University. Organs were allocated in a fair and transparent manner using the Chinese Organ Transplant Response System (23). The donor’s family voluntarily determined the type of donation, and the recipients underwent follow-up after transplantation. All participants provided informed consent before being included in the study, and their confidentiality was ensured throughout the research process. Data were anonymized to protect personal identities, and no patient-identifiable information was used in any part of the study.

Donor data were obtained from the Chinese Computerized System for Human Organ Distribution and Sharing, completed by members of organ transplant organizations. Donor information was obtained from clinical electronic medical records. Recipient data were derived from electronic clinical records alone. Recipients were categorized into the DGF or immediate graft function (IGF) group according to their early renal function status. Donors were similarly classified into DGF and IGF groups, depending on the early renal function recovery of their paired recipients. DGF was defined as a reduction in blood creatinine concentration of less than 10% for three consecutive days within 1 week postoperatively, or if serum creatinine (SCr) did not decrease to 400 μmol/L within 1 week post-surgery (24, 25). IGF was defined as the fast postoperative recovery of renal function with satisfactory diuresis and no further need for dialysis. Graft loss was defined as end stage renal disease requiring dialysis or retransplantation. ECD donors were defined as those over 60 years-old, or between 50 and 59 years old who met at least two of the following criteria: final serum creatinine exceeding 133 μmol/L, cause of death being cerebrovascular accident, or a history of hypertension (26). We converted human leukocyte antigen (HLA) mismatches into hierarchical profiles to facilitate data analysis: 0 to 1 mismatch as Level 1, 2 to 4 mismatches as Level 2, and 5 to 6 mismatches as Level 3.

Prior to kidney procurement, 2 mL of fresh blood was drawn from the donor in the operating room using a vacuum blood collection needle and placed in a sterile test tube. The blood was left at room temperature for 1 hour and then centrifuged at 4°C, 2,500 rpm for 10 min. The resulting supernatant was separated, while the sediment was discarded. The supernatant was subsequently divided into aliquots, frozen, and stored at -80°C in the hospital. Before testing, the serum samples were thawed on crushed ice and centrifuged again at 1,000 g for 15 min at 4°C.

Serum CIRBP concentrations were measured using an enzyme-linked immunosorbent assay (ELISA) kit (CSB-EL005440HU). Firstly, the protein standard was diluted per the instructions, and 100 µL of different concentrations of standard and samples (1:10 dilution) were added to a 96-well plate in duplicates, and the plates were incubated at 37°C for 2 h in the dark. Following incubation, the liquid from the wells was discarded, and 100 µL of pre-diluted anti-CIRBP biotin-labeled antibody solution was introduced into each well. After incubating at 37°C for 1 h, each well was washed at least three times with a washing solution before 100 μL of diluted horseradish peroxidase-labeled antibody was added to each well, following the same procedure as that in the previous step. After another 1-hour incubation at 37°C, we washed each well at least five times, to prevent false positive results. Thereafter, we added 90 µL of substrate solution to each well, and add 50 µL of termination solution was added to stop the reaction after a 15-min color development at 37°C. The optical density (OD) at 450 nm was measured within 5 minutes using an enzyme marker. A standard curve model was then generated from the OD values of the standard wells, and the CIRBP concentration of each sample was determined using this model.

Age is presented as the mean ± standard deviation (SD). Student’s t-test was applied to compare the ages between the DGF and IGF groups. Body mass index (BMI), donor terminal creatinine, donor serum CIRBP, kidney donor risk index, KDPI, dialysis duration, and post-transplantation renal function values are presented as median [interquartile range], and these variables were compared between the DGF and IGF groups using the Mann-Whitney U test. Frequencies and percentages were used to present categorical variables, such as sex, cause of death, and dialysis modality. Comparisons between two groups were performed using the Chi-square tests when the sample size was ≥40 and Fisher’s exact tests when the sample size was <40. KDPI was calculated using standard methods (27, 28). Factors associated with DGF and graft loss were confirmed using the Spearman correlation analysis.

Each factor potentially associated with the occurrence of DGF after renal transplantation was initially analyzed using univariate logistic regression to statistically determine its influence on DGF. Independence of these factors was further evaluated through stepwise backward multivariate logistic regression, with entry criteria set at 0.05 and removal criteria at 0.1. This approach ensured that only meaningful parameters were retained in the final model. The predictive ability of CIRBP for DGF and graft loss was evaluated through receiver operating characteristics (ROC) curve analyses, with optimal cutoff points derived from Youden’s index, J, which is determined as J=sensitivity + specificity – Eqn. 1 (29). In addition, we applied forward stepwise multivariate linear regression analyses to identify the factors influencing recipient SCr concentrations at 6 months post-kidney transplantation. The Kaplan-Meier analysis was used to evaluate the impact of DGF occurrence and elevated donor plasma CIRBP levels on allograft survival time following kidney transplantation.

All statistical tests were performed using SPSS version 27.0 (SPSS Inc., Chicago, IL, USA), and significance was determined by a P-value of less than 0.05.

Figure 1 illustrates the process of assembling the cohort, while Tables 1 and 2 present the baseline characteristics of the 207 donor-recipient pairs engaged in the study. The cohort was stratified according to allograft function; 51 (25%) recipients developed DGF after kidney transplantation. The DGF and IGF groups were then compared. The mean ± SD donor age was 42.31 ± 14.25 years, and the median [interquartile range] BMI was 23.00 [21.40–25.40] kg/m². Among the donors, 86% were male, 20% had hypertension, 46% died of cerebral hemorrhage, and 17% were extended-criteria donors; however, no factor significantly differed between the groups. As expected, the median plasma CIRBP level in the DGF group was 6.82 ng/mL, significantly higher than that in the IGF group (3.44 ng/mL; P<0.001). Similar results were observed for donor terminal SCr levels (158.00 vs. 109.00; P<0.001). However, we failed to observe the variation in donor KDPI values amidst the DGF and IGF groups (59% vs. 46%; P=0.215), consistent with our prediction that KDPI does not reliably predict early post-transplant renal function. The mean ± SD recipient age was 42.56 ± 10.80 years, with a median [interquartile range] BMI of 21.26 [19.84–22.86] kg/m². Comparing the DGF and IGF groups, we found that male recipients (88% vs. 66%; P=0.002) and those with a higher BMI (22.64 vs. 20.96; P<0.001) had a greater likelihood of developing DGF. Additionally, recipients with DGF had distinctly higher SCr levels at 6 months post-transplant compared to those in the IGF group (171.00 vs. 121.00; P<0.001). No differences were found between the groups in terms of dialysis type, dialysis duration, HLA mismatch level, induction therapy, or immunosuppressant dosage.

DGF occurred in recipients paired with 51 (25%) donors following kidney transplantation. Testing of donor plasma samples revealed that the median CIRBP concentration in the DGF group was 6.82 ng/mL, which was 1.98 times higher than that in the IGF group (6.82 vs. 3.44; P<0.001). Recipients with CIRBP concentrations exceeding 5.484 ng/mL were four times more likely to develop DGF than those with concentrations below 5.484 ng/mL (53% vs. 12%; P<0.001). When CIRBP levels exceeded 7.92 ng/mL, the probability of developing DGF was 100%, whereas levels below 2 ng/mL corresponded to 0% probability of developing DGF ( Figure 2B ). Although the median terminal SCr concentration was notably higher in the DGF group than that in the IGF group, 19% of recipients developed DGF postoperatively even when the SCr concentration was < 100 μmol/L ( Figure 2A ). Additionally, the scatter distribution of CIRBP resembled that of SCr ( Figure 2 ), and Spearman’s correlation demonstrated a significant relationship between the two factors (r=0.140, P=0.045). Furthermore, the proportion of female recipients was notably lower in the DGF group than that in the IGF group (12% vs. 34%; P=0.002) ( Figure 2D ), while no significant variation in KDPI was found between the two groups ( Figure 2C ). Graft loss occurred in 10 recipients in the DGF group, showing a higher rate, compared to the IGF group (19.6% vs. 5.8%; P=0.003).

To further clarify the role of CIRBP in DGF, a logistic regression analysis was conducted. Univariate analysis revealed that for every 1 μmol/L increase in donor terminal SCr, the risk of developing DGF increased by 0.7% (95% confidence interval [CI]: 1.004–1.010; P<0.001), for every 1 ng/mL rise in plasma CIRBP concentration, the risk increased by 73.3% (95% CI: 1.446–2.076; P<0.001), and female recipients had a 0.259-fold lower probability of developing DGF compared to males (95% CI: 0.104–0.646; P=0.004). Donor plasma CIRBP (OR=1.660; 95% CI: 1.376–2.004; P<0.001) and donor terminal SCr (OR=1.005; 95% CI: 1.001–1.009; P=0.025) were identified by multivariate analysis as independent risk factors for DGF development, while a protective factor was female sex of the recipient (OR=0.302; 95% CI: 0.103–0.885; P=0.029) ( Table 3 ).

The area under the ROC curve (ROC-AUC) of CIRBP was 0.801 (95% CI: 0.728–0.874; P<0.001), and the optimal cut-off value was 5.548 ng/mL, calculated using the Youden index, with a sensitivity of 65% and 83% specificity ( Figure 3B ). This performance was markedly superior to that of donor-terminal creatinine (ROC-AUC=0.654; 95% CI: 0.560–0.749; P=0.001) and KDPI (ROC-AUC=0.558; 95% CI: 0.471–0.645; P=0.215) ( Figures 3A, C ). Enhancing the predictive ability of CIRBP, ROC curve analysis was performed via logistic regression, separately fitting donor plasma CIRBP to donor terminal SCr levels and recipient sex. The best predictive model was represented by the formula: y=1.660×donor plasma CIRBP+3.312×male+1.005×donor terminal SCr-5.365 (ROC-AUC=0.835; 95% CI: 0.771–0.898; P<0.001). The optimal cutoff value was 0.308, having a sensitivity of 63% and 88% specificity ( Figure 3F ) and was superior to both the fitting model of plasma CIRBP and donor terminal SCr (ROC-AUC=0.808; 95% CI: 0.733–0.884; P<0.001) and the fitting model of plasma CIRBP and recipient sex (ROC-AUC 0.821; 95% CI: 0.753–0.888; P<0.001) ( Figures 3D, E ).

At 6 months post-transplantation, recipient SCr concentration increased with increasing plasma CIRBP levels in donors ( Figure 4A ). Meanwhile, the SCr concentration at 6 months post-transplantation increased with the KDPI value ( Figure 4C ), consistent with previous findings, suggesting that the KDPI may help predict long-term donor kidney survival. However, no significant trend was noted for donor terminal SCr levels ( Figure 4B ).

Univariate linear regression analysis revealed significant linear correlations between SCr concentration at 6 months and donor plasma CIRBP (R²=0.08, P<0.001), KDPI (R²=0.07, P<0.001), female recipients (R²=0.06, P=0.002), donor age (R²=0.054, P=0.003), as well as donor cause of death being cerebral hemorrhage (R²=0.035, P=0.016) ( Table 4 ). Ultimately, only donor plasma CIRBP (R²=0.08, P=0.003), KDPI (R²=0.137, P=0.002), and female recipient status (R²=0.174, P=0.008) were retained in the multiple linear regression analyses ( Table 4 ). Conversely, donor terminal SCr levels were not found to be associated with recipient creatinine at 6 months post-transplantation.

Recipients were categorized into high plasma CIRBP (≥5.484 ng/mL, n=60) and low plasma CIRBP (<5.484 ng/mL, n=147) groups based on the optimal cutoff value ( Figure 3B ). We hypothesized that donor kidneys with higher plasma CIRBP concentrations would exhibit significantly shorter survival times compared to those with lower concentrations. Although our survival analysis showed a trend of reduced survival time for donor kidneys with plasma CIRBP concentrations >5.484 ng/mL, further statistical analysis did not support this hypothesis (P=0.167; Figure 5A ). However, when comparing graft survival times between the DGF and IGF groups, the Kaplan-Meier curves revealed that recipients who developed DGF had shorter long-term graft survival (P=0.002; Figure 5B), consistent with findings from related studies.

Univariate and multivariate analyses were conducted to identify the risk factors for kidney transplant loss. The results indicated that donor plasma CIRBP and donor terminal SCr were not significantly correlated with transplant loss (P=0.057 and P=0.231, respectively). Recipient female sex was the only significant factor correlated with reduced transplant loss risk (OR=0.125; 95% CI: 0.016–0.955; P=0.045) ( Table 5 ). Subsequently, ROC curve analyses were conducted to evaluate the association between transplanted kidney survival and donor-terminal SCr (ROC-AUC=0.420; 95% CI: 0.283–0.557; P=0.250), plasma CIRBP (ROC-AUC=0.601; 95% CI: 0.469–0.732; P=0.149), and KDPI (ROC-AUC=0.593; 95% CI: 0.457–0.728; P=0.183) ( Figures 5C–E ). However, these analyses did not reveal any statistically significant differences.