Elderly kidney transplant recipients have favorable outcomes but increased infection-related mortality

Patients

Patients aged >18 years who underwent KT at Kyungpook National University Hospital in Daegu, Korea between January 2000 and December 2019 were retrospectively analyzed. Retransplantation patients were also included. Old patients were defined as those 60 years or older, and we divided KT recipients into four groups according to combination of ages of recipient and donor, as follows (Fig. 1):

Figure 1.

Flow diagram of the study.

(1) Elderly recipients: old-to-old, defined as both donor and recipient aged ≥60 years; young-to-old, defined as donor aged <60 years and recipient aged ≥60 years.

(2) Young recipients: old-to-young, defined as donors aged ≥60 years and recipients aged <60 years; young-to-young, defined as both donor and recipient aged <60 years.

Data collection

Patient data were retrospectively collected from electronic medical records. The demographics of the recipients include age, sex, body mass index (BMI), primary renal diseases, comorbid diseases, pretransplant dialysis vintage, pretransplant desensitization, number of human leukocyte antigen (HLA) mismatches, and type of immunosuppressants. Serial follow-up serum creatinine and tacrolimus trough levels were also collected, as well as demographic characteristics of donors, including age, sex, and BMI. The estimated glomerular filtration rate (eGFR) was calculated using the MDRD equation (derived from the Modification of Diet in Renal Disease study) [13]. Delayed graft function was diagnosed in KT recipients who underwent dialysis during the first week after transplant because of poor graft function [8]. Finally, annual mean tacrolimus trough level was calculated as the average value of all measured tacrolimus trough levels yearly after transplant.

Immunosuppressive treatment

Induction and maintenance immunosuppressive agents were used according to our center’s protocol. The detailed protocol is described in our previous studies [8,14]. Briefly, intravenous basiliximab was used as induction immunosuppression in normal-risk patients, and anti-thymocyte globulin was used for high-risk patients. For maintenance immunosuppression, standard triple immunosuppressive therapy (calcineurin inhibitor, mycophenolate, and corticosteroid) was applied. The target tacrolimus trough level was 5–10 ng/mL until 3 months after transplant, then it was tapered to 3–7 ng/mL. The target cyclosporine trough level was 200–300 ng/mL during the 3 months after transplant, after which it was tapered to 50–150 ng/mL for the maintenance period. Mycophenolate mofetil was administered daily at a fixed dose of 1.0 to 2.0 g. For corticosteroid treatment, 500 mg intravenous methylprednisolone was administered at the time of transplant and was tapered to 5 mg per day oral prednisolone within six months after transplant.

Outcomes

The study outcomes were patient survival, death-censored graft survival, and biopsy-proven acute rejection (BPAR)-free survival. Patient survival was defined as the time from KT to death from all causes. Death-censored graft survival was defined as the time from transplant to the restart of RRT. In cases of patient death with a functioning graft, graft survival was censored at the time of death. BPAR was diagnosed based on the Banff 07 classification [15] and was subdivided according to the time of diagnosis after transplant (early, 1 year or less after transplant; late, more than 1 year after transplant) and type of rejection.

Statistical analysis

Continuous variables were expressed as mean ± standard deviation, and categorical variables were expressed as number (percentage). Analysis of variance or Student t test was used to determine differences between continuous variables. Pearson chi-square test or Fisher exact test was used to determine differences among categorical variables. Patient survival, death-censored graft survival, and BPAR-free survival were analyzed using Kaplan-Meier analysis with log-rank tests. Multivariable Cox proportional hazard regression analyses were conducted for survival analyses. The results of Cox regression analysis were presented as hazard ratio (HR) with 95% confidence interval (CI). To avoid confounding effects, variables that were significantly different at baseline (p < 0.05; comorbid diabetes, BMI of recipient and donor, tacrolimus use, and number of HLA mismatches) or were associated with clinical outcomes of transplant (sex, transplant type, immunosuppressants given) were entered in the multivariable Cox regression analysis. Here, Model 1 was adjusted for sex and diabetes; Model 2 for Model 1 variables plus recipient BMI, donor BMI, and transplant type; Model 3 was adjusted for Model 2 variables as well as tacrolimus use, number of HLA mismatches, anti-thymocyte globulin induction, and maintenance mycophenolate use. Statistical analyses were conducted using IBM SPSS version 22.0 (IBM Corp., Armonk, NY, USA) and R (version 3.6.2; The R Foundation for Statistical Computing, Vienna, Austria). The p-values of <0.05 were considered statistically significant.

Ethics statement

The Institutional Review Board of Kyungpook National University Hospital reviewed and approved the study protocol (No. 2021-07-069). The requirement for informed consent was waived because the study was conducted as a retrospective review of electronic medical records. All patient information was anonymized and deidentified before analyses. This study was conducted according to the 2000 Declaration of Helsinki and the Declaration of Istanbul, 2008.

Baseline characteristics

Of the total 1,023 KT recipients examined during the study period, 129 (12.6%) were ≥60-years-old at the time of transplant. Among these elderly recipients, 46 (4.5% of the total) received KT from old donors (age, ≥60 years), which is defined as the old-to-old group; 83 elderly recipients (8.1% of the total) were included in the young-to-old group. There were 893 KT recipients (87.4%) aged <60 years at the time of transplant; of them, 98 (9.6% of the total) received KT from old donors (age, ≥60 years) and were defined as the old-to-young group. There were 796 cases (77.8% of the total) in which both donor and recipient were aged <60 years, defined as the young-to-young group and set as the reference. The median follow-up duration for all patients was 69.2 months (interquartile range [IQR], 39.8–115.9 months); that for old-to-old was 52.0 months (IQR, 38.0–77.0 months), that for young-to-old was 57.7 months (IQR, 32.9–86.2 months), that for old-to-young was 55.8 months (IQR, 29.8–81.9 months), and that of the young-to-young groups was 73.4 months (IQR, 44.8–131.6 months). Baseline characteristics of each patient group are shown in Table 1. The mean age was 64.8 ± 3.5 years in the old-to-old group, 63.8 ± 3.2 years in the young-to-old group, 45.1 ± 3.2 years in the old-to-young group, and 43.6 ± 10.2 years in the young-to-young group. The proportion of males was higher among elderly recipients (old-to-old and young-to-old) compared with young recipients (old-to-young and young-to-young) (p = 0.007). More than half of the elderly recipients had ESKD due to diabetes, while more than half of the young recipients had ESKD due to glomerulonephritis (p < 0.001). Comorbid diabetes was more frequent in the elderly recipients compared with the young recipients (p < 0.001). Pretransplant dialysis vintage and pretransplant desensitization rates did not differ among groups. The mean ages of donors were 66.7 ± 4.9 years in old-to-old, 43.7 ± 12.0 years in young-to-old, 64.2 ± 3.7 years in old-to-young, and 41.9 ± 12.3 years in young-to-young (p < 0.001). Pretransplant donor serum creatinine levels were not different among groups. The proportion of living or deceased donors and type of induction immunosuppression did not differ among groups.

Table 1.

Baseline characteristics of patients

Tacrolimus immunosuppression after transplant

The majority of KT recipients (83.9%, 858 of 1,023) were on immunosuppressive therapy, including tacrolimus. Therefore, we compared mean tacrolimus trough levels to identify degree of immunosuppression after transplant (Fig. 2). The mean tacrolimus trough level was significantly lower for the old-to-old group than the young-to-young group at one year after transplant. However, there were no differences in annual mean tacrolimus trough level among groups at 2, 3, 4, 5, and 10 years after transplant.

Figure 2.

Comparison of annual mean tacrolimus trough levels.

*Old-to-old vs. young-to-young; p < 0.05.

Outcomes and graft function

Patient and graft outcomes are summarized in Table 2. The incidence of delayed graft function, graft failure, and BPAR including early and late events was not different among groups. The proportion of KT recipients whose serum creatinine level was higher than 1.5 mg/dL at 1 year after transplant was higher in the old-to-old and old-to-young groups. The proportions of patients who were hospitalized to treat infection did not differ among groups. The incidence of patient death was 13.0% (6 of 46) in the old-to-old group and 16.9% (14 of 83) in the young-to-old group. However, young recipients showed lower mortality (5.1% [5 of 98; old-to-young] and 6.0% [48 of 796; young-to-young]) (p = 0.001). Infection-related mortality was higher among elderly recipients (p < 0.001). The incidence of cardiovascular and cancer-related deaths was not different among the four groups. Supplementary Table 1 (available online) shows detailed information about infection-related death, the most common cause of which is pneumonia.

Table 2.

Comparison of graft and patient outcomes

In subgroup analysis according to donor type, elderly recipients showed higher all-cause mortality and infection-related mortality than young recipients in both living donor KT and deceased donor KT groups (all ps < 0.05; Supplementary Table 2, available online).

We increased the cutoff age of the elderly group to 65 years and compared outcomes among the four groups (Supplementary Table 3, available online). The number of patients in the old-to-old group was 16, in the young-to-old group was 33, and in the old-to-young group was 48. The results were similar to those of the analysis with an elderly cutoff age of 60 years. Both all-cause mortality and infection-related mortality were higher in the elderly recipient groups. When patients were divided into two groups based on the recipient or donor age of 70 years (age of ≥70 and <70 years), neither extremely elderly recipient nor donor KT groups showed higher incidence of graft failure and death of patients than the young recipient or donor KT groups as follows. 1) Extremely elderly recipients (n = 7) vs. young recipients (n = 1,016): graft failure, 0 vs. 111 (10.9%) (p > 0.99); death, 1 (14.3%) vs. 72 (7.1%) (p = 0.41). 2) Extremely elderly donors (n = 21) vs. young donors (n = 1,002): graft failure, 1 (4.8%) vs. 110 (11.0%) (p = 0.37); death, 1 (4.8%) vs. 72 (7.2%) (p = 0.67).

Serial changes in eGFR after transplant are shown in Fig. 3. The old-to-old group had significantly lower eGFR than groups with young donors (young-to-old and young-to-young) at 14 days and 1, 2, 3, and 5 years after transplant (p < 0.05).

Figure 3.

Serial changes in eGFR during 10-year follow-up.

eGFR, estimated glomerular filtration rate; POD, postoperative day.

^*Old-to-old vs. young-to-old; p < 0.05. ^§Old-to-old vs. young-to-young; p < 0.05.

Patient survival

In the Kaplan-Meier curve for patient survival, elderly recipients showed significantly shorter survival times compared with young recipient groups (log-rank p < 0.001) (Fig. 4). In univariable Cox proportional hazard regression analysis, both elderly recipient groups showed a significantly higher risk of patient death than the young-to-young group (Table 3). In multivariable analysis to adjust for confounding effects of baseline characteristics, elderly recipients were consistently at significantly higher risk of patient death than was the young-to-young group (Model 3; old-to-old: aHR, 2.89; 95% CI, 1.14–7.32; p = 0.03; young-to-old: aHR, 3.06; 95% CI, 1.51–6.20; p = 0.002).

Figure 4.

Kaplan-Meier curve for patient survival.

A p-value less than 0.008 indicates significant difference in post hoc Bonferroni correction.

Table 3.

Cox regression analysis of patient death

Graft survival and biopsy-proven acute rejection-free survival

Death-censored graft survival was significantly different among the four groups in the Kaplan-Meier survival curve (log-rank p = 0.02) (Fig. 5A). In univariable and multivariable Cox proportional hazard regression analyses, there was no significant difference in graft failure in elderly recipients compared with the young-to-young group (p > 0.05) (Table 4). However, the old-to-young group showed a consistently increased risk of death-censored graft failure compared with the young-to-young group (Model 3: aHR, 2.41; 95% CI, 1.30–4.46; p = 0.005).

Figure 5.

Kaplan-Meier curves for graft outcomes.

(A) Death-censored graft survival. (B) Biopsy-proven acute rejection (BPAR)-free survival. A p-value less than 0.008 indicates significant difference in post hoc Bonferroni correction.

Table 4.

Cox regression analysis of death-censored graft failure

BPAR-free survival did not differ among the four groups as per the Kaplan-Meier survival curve (log-rank p = 0.389) (Fig. 5B). Table 5 shows a risk of BPAR among groups in univariable and multivariable Cox proportional hazard regression analyses. There were no significant differences in BPAR risk in groups other than the young-to-young group (p > 0.05).

Table 5.

Cox regression analysis of biopsy-proven acute rejection