The DZIF Transplant Cohort is a multicenter, prospective study within the framework of the German Center for Infection Research (DZIF) that focuses on transplant recipients and their infection risks [24]. This study specifically includes adult patients who received a kidney or simultaneous pancreas-kidney transplant at the University Hospital Heidelberg between January 2012 and December 2023. The Ethics Committee of the Medical Faculty of Heidelberg University approved for the study (No. S-585/2013), and all participants provided written informed consent. Clinical events were systematically recorded and evaluated, including all events up to December 2024. Follow-up examinations were performed at baseline, 3, 6, 9, and 12 months after transplantation, with additional annual visits thereafter or as clinically needed in case of complications. Clinical, laboratory, and demographic data were collected from medical records [24]. All recipients were followed for at least 1 year post-transplantation, unless death or graft loss occurred earlier.

The standard immunosuppressive protocol consisted of a calcineurin inhibitor - either tacrolimus (Tac) or ciclosporin A (CsA) - in combination with mycophenolate sodium (1.44 g/day) or mycophenolate mofetil (2 g/day), and methylprednisolone. Target trough (C₀) levels for Tac were: month 1 (6–9 ng/mL), month 3 (5–8 ng/mL), thereafter (4–7 ng/mL); and for CsA: month 1 (150–180 ng/mL), month 3 (100–150 ng/mL), thereafter (80–120 ng/mL). Induction therapy included either basiliximab or thymoglobulin. Patients with previous transplants or high levels of donor-specific anti-HLA antibodies (DSA) were classified as highly sensitized. Additional desensitization strategies such as plasmapheresis or immunoadsorption were used in these patients, as well as in AB0-incompatible recipients.

Prophylaxis and monitoring were performed according to the KDIGO 2009 guidelines [25]. For Pneumocystis jirovecii prophylaxis, trimethoprim-sulfamethoxazole (800 mg/160 mg) was routinely administered three times a week for the first 6 months after transplantation. Standard antiviral prophylaxis with valganciclovir was administered to all Cytomegalovirus (CMV) immunoglobulin G (IgG)–positive recipients and recipients of CMV IgG–positive donor organs for at least 3 months. In high-risk (D+/R–) cases, 6 months of antiviral prophylaxis was recommended. The dosage was adjusted according to renal function. Candida prophylaxis with oral nystatin was provided within the first one to 3 months if more than 20 mg of methylprednisolone was administered daily.

All infections requiring hospitalization were included. Diagnoses were made by the treating physician. Data collected included clinical presentation, laboratory and microbiological/virological findings, diagnostic procedures, treatment, disease course, and infection-related outcomes. Detailed infection definitions are provided in the (Supplementary Table S1).

Acute rejection was diagnosed by histopathological examination of renal biopsies by a trained local kidney pathologists based on morphological assessment and immunohistochemical markers. Rejection episodes were classified according to the Banff criteria [26].

Baseline characteristics were summarized using means ± standard deviations or medians with interquartile ranges for continuous variables and number with percentages for categorical variables. Group comparisons were performed using Pearson’s chi-squared test or Fisher’s exact test for categorical data, and Student’s t-test or Mann–Whitney U test for continuous variables, as appropriate. Survival probabilities were estimated using the Kaplan–Meier method and compared using the log-rank test. The association between recipient age (<40, 40–60, >60 years) was first explored through univariable Cox proportional hazards regression models. To estimate independent effects, propensity scores were derived from a multinomial logistic regression model including key baseline covariates. Inverse probability weighting (IPW) based on these propensity scores was then applied to create a weighted pseudo-population minimizing residual confounding. Subsequently, IPW-weighted multivariable Cox regressions were fitted to assess independent associations between recipient age and outcomes. Variables with a p-value <0.1 in the weighted univariate analyses and clinically relevant variables were entered into multivariable models, while age group was retained in all models. Hazard ratios (HRs) with 95% confidence intervals (CIs) were reported; p < 0.05 was considered statistically significant.

Sensitivity analyses included unweighted and trimmed (2.5th–97.5th percentile) IPW models, as well as models with and without adjustment for first-year infection burden (Supplementary Table S3). To account for potential COVID-19 bias, follow-up was additionally censored at 1 March 2020, or modeled with a pandemic indicator (March 2020–December 2022).

Death before graft failure was treated as a censoring event. Variables were handled by complete-case analysis within each model. Given the overall low proportion of missing data (<5% across key variables), multiple imputation was not deemed necessary.

Risk factors for infection burden within the first post-transplant year were analyzed using negative binomial regression.

All statistical analyses were conducted using R software version 2024.12.0 + 467.

The incidence of delayed graft function (DGF) increased significantly with recipient age, rising from 22.4% in recipients <40 years to 36.3% in those >60 years (p = 0.012, Table 1). Across all age groups, there was a significant improvement in estimated glomerular filtration rate (eGFR) from month 1 to month 3, with the greatest relative increase observed in recipients >60 years (median +50.2% from the baseline median of 24.5 mL/min/1.73 m²). Younger recipients maintained higher absolute eGFR levels throughout the first year (Figure 3a). In recipients >60 years, lower tacrolimus doses were required to achieve comparable and adequate drug levels (Figure 3b).

Short and long-term infection incidences are provided in Table 2. Older recipients had the highest overall infection rates, with a 5-year incidence of 94.1% (p < 0.001; Table 2). Bacterial infections were more frequent in older patients (>60: 77.0%, p < 0.0001), while viral infections occurred at similar rates across all age groups (p = 0.761). Fungal infections were significantly more common in older recipients (>60: 9.8% vs. <40: 0.8%, 40–60: 3.3%; p = 0.002).

UTIs were the most common infection type across all age groups, accounting for approximately 34%–35% of all infections. Although older age was initially associated with a higher risk of UTI, this association was no longer significant after multivariable adjustment (Figure 2; Supplementary Table S2).

The number of UTIs per infected patient was highest in recipients <40 years (3.6 episodes), followed by >60 years (3.2) and 40–60 years (2.7). Across groups, E. coli (42%, 38%, 35%) and Enterococcus spp. (30%, 22%, 30%) were the leading uropathogens. Pseudomonas aeruginosa prevalence increased with age (6% in <40 vs. 18% in >60), as did Enterobacter spp. (2%, 4%, 5%), while Klebsiella spp. decreased (21%, 18%, 12%) (Supplementary Figure S2). Resistant uropathogens were more frequent in recipients >60 years (17.6% vs. 6.7% in <40), but age was not an independent risk factor for resistant infections after full adjustment (Figure 2; Supplementary Table S2).

Older recipients more often had multidrug-resistant Gram-negative infections causing urosepsis or transplant pyelonephritis (45.4%). In recipients <40 years, resistant infections were limited to uncomplicated lower UTIs. Patients with resistant infections averaged 7.7 ± 5.8 episodes during follow-up and 3.9 ± 3.2 in the first post-transplant year. Their 5-year mortality was 28.4% (95%-CI, 19.3–38.4). Graft failure occurred in 33.3%. Cumulative incidence rose with age (from 4.8% to 11.6%), but age was not an independent predictor after full adjustment (HR = 1.66, p = 022.39; Table 2; Figure 2).

Other types of infection showed age-related patterns. Pneumonia accounted for a higher proportion of infections with increasing age: 6.9% (<40 years) vs. 11.8% (>60 years) with a higher incidence in the elderly (Table 2), but this difference lost significance after adjustment (Supplementary Table S2). In contrast, upper respiratory tract infections decreased with age (10.3% vs. 3.2%). Gastrointestinal infections, pyelonephritis, and infections of unknown origin remained stable across all age groups (Table 3; Figure 3). Viral GI infections were more common in younger recipients, while Clostridium difficile was more prevalent in older groups. Sepsis rates remained stable across all age groups (8%–9%). The majority of sepsis cases (89.2%) originated from UTIs, mainly caused by Gram-negative bacteria (89.2%), particularly E. coli (52%) and Klebsiella spp. (20.4%). Sepsis due to pneumonia or soft tissue infections was rare, both accounted for 3.1% of all sepsis cases.

BKV viremia was most common in the 40–60 years age group. In the first year post-transplant, the cumulative incidence ranged from 11.6% in recipients <40 years to 15.1% in those >60 years (Supplementary Table S2). New BKV viremia cases appeared in the >60 years group after the first year. Recipients aged 40–60 and >60 years had higher initial plasma viral loads than younger recipients, while those <40 years had the highest peak viral loads. Older recipients (40–60 and >60 years) also had shorter intervals between successive BKV viremia peaks (Supplementary Figure S1). The highest BKV nephropathy (BKVN) rate was found in the 40–60 years group (19.3%), compared to 5.3% in <40 years and 12.5% in >60 years.

CMV viremia was similar across age groups in the first post-transplant year: 13.6% in recipients <40 years, 14.6% in 40–60 years, and 16.4% in those >60 years. Beyond the first year, the CMV incidence continued to rise in older recipients, reaching 21.5% at 5 years (Supplementary Table S2). Peak viral load was highest in recipients >60 years (103,500 IU/mL), followed by <40 years (66,988 IU/mL) and 40–60 years (53,325 IU/mL). CMV-related organ complications occurred more frequent in recipients >60 years (19.2%) than in recipients between 40 and 60 years (17.4%) and <40 years (4.8%). Reinfection rates were significantly higher in recipients >60 years (46.2%) and 40–60 years (45.7%) than in those <40 years (28.6%, p = 0.03). De novo CMV infections occurred more frequently in >60 years (46.7%) than in 40–60 years (39.7%) and <40 years (39.3%). Reactivation rates followed a similar trend: 20.0% in >60 years, 14.8% in 40–60 years, and 11.9% in <40 years.

Non-CMV herpesvirus infections had a cumulative incidence of 6.1% at year one, increasing to 11.8% at 5 years. The overall incidence was 13.8%, with most cases occurring after the first year. VZV infections (all herpes zoster) increased from 2.5% at year one to 7.2% at 5 years. Notably, 90% of VZV infections in recipients >60 years occurred after the first year, with a median onset of 27 months (IQR 16–40), compared to 8 months (IQR 2–15) in those <40 years.

HSV infections had a cumulative incidence of 3.1% at year one, 4.5% at 5 years, and 6.0% overall. Nearly 46% of cases were diagnosed after the first year, particularly in older recipients. 61.5% of these cases presented with pneumonia, while all HSV cases in younger recipients were limited to herpes labialis.

Four cases of post-transplant lymphoproliferative disorder (PTLD) were reported, with a median onset of 266 days (IQR 241–335). Seventy-five percent of these cases occurred in recipients >60 years, including one who required transplant nephrectomy.

The 5-year cumulative incidence of opportunistic infections with organ involvement increased with age, from 3.4% in recipients <40 years to 12.7% in those >60 years (Table 2). Bacterial infections predominated in the youngest group (57.1%), while fungal infections were most common in the 40–60 years group (43.5%), followed by viral (34.8%) and bacterial (21.7%) infections. In recipients >60 years, fungal and viral infections occurred equally (40% each).

The median onset of infections was late (Md = 420 days, IQR 104–948), particularly in older recipients. Pneumonia was the most frequent manifestation, caused by pathogens including HSV (n = 11), Pneumocystis jirovecii (n = 11), Aspergillus spp. (n = 9), CMV (n = 4), Mycobacterium tuberculosis (n = 2), Legionella (n = 1), and Rhodococcus (n = 1). CMV colitis occurred in 10 cases. Although one case of disseminated Bartonella infection was identified in a 24-year-old, disseminated infections occurred predominantly in recipients >60 years and included severe disease such as mucormycosis (resulting in death within 26 days post-transplant), nocardiosis with brain abscess, invasive aspergillosis, multifocal CMV disease, and cryptococcal pyelonephritis requiring nephrectomy.

Age >60 years was initially associated with a higher risk of opportunistic infections but lost significance after adjustment (HR = 3.20, p = 0.1275; Figure 2), while high sensitization and high BMI remained significant predictors (Supplementary Table S2).

Aspergillus infections occurred exclusively in recipients >40 years of age, with an incidence of 1.8% in the 40–60 years group and 4.6% in those >60 years. Pneumocystis jirovecii pneumonia (PjP) occurred in all age groups but was least common in recipients <40 years (0.6% vs. 2.6% in older groups). Aspergillus infections appeared significantly earlier in the 40–60 years group (median 62 days, IQR 45–78) than in those >60 years (291 days, IQR 63–866; p = 0.037). The majority of PjP cases (90.9%) occurred after prophylaxis was stopped, with no age-related differences in time course. Intensive induction therapy was administered more frequently in PjP cases (33.3%) than in Aspergillus cases (8.3%). Recipients with Aspergillus infections had a higher infection burden in the first post-transplant year (mean 3.6 episodes) than in recipients with PjP (mean 2.2 episodes). Aspergillus-related mortality was 28.3% in recipients >60 years, while no deaths were observed in the 40–60 years group. Although 33.3% of PjP patients died, none of the deaths were directly attributable to PjP.

Acute rejection occurred in 22.6% of recipients (95% CI: 19.3–26.3), with no significant differences between age groups (Table 2; Figure 1c). However, the timing of rejection varied by age: the earliest rejections were observed in recipients >60 years (median 37 days, IQR 15–118), followed by 40–60 years (65 days, IQR 16–162), and the latest in recipients <40 years (78 days, IQR 13–414).

Recurrent rejection occurred most frequently in the youngest group (8.2%) compared to 5.4% in those >60 years. Biopsy findings revealed only minor age-related differences. Borderline rejection was the most common finding, occurring in 37.8%–47.3% of cases. Acute tubular necrosis increased with age, peaking at 13.3% in recipients >60 years. T-cell–mediated rejection (12.4%) and BKVN (11.8%) occurred most frequently in the 40–60 years group.

The 5-year graft failure rate increased significantly with recipient age, from 5.2% (95% CI, 3.0–8.9) in recipients <40 years to 21.0% (95%-CI, 14.6–33.2) in those >60 years (p 0.0001, Table 2; Figure 1b). The unadjusted analysis showed a higher risk of graft loss in recipients >60 years (Figure 2; Table 3). Graft failure also occurred earlier in older recipients, with a median time to failure of 34 months (IQR 18–56) in the >60 group, compared to 52 months (IQR 31–74) in <40 years.

Kidney transplant removal was more often necessary in older recipients (50.2% in >60 years vs. 18.2% in <40 years, p < 0.0001). Infections and acute or chronic rejection were the main causes of graft failure in all age groups. In recipients <40 years, infections accounted for more than half of graft losses, while infections contributed to 40% of graft failures in those >60 years. These infections included BKVN, parainfectious complications, and other graft-related infections.

After adjustment, recipient age >60 years was no longer a significant predictor of graft failure (HR = 0.67, p = 0.3924; Figure 2; Table 3). Independent predictors included donor age, male donor and the number of infections during the first post-transplant year (Table 3).

Five-year mortality rate rised significantly with increasing age, from 1.8% (<40 years) to 18.4% (>60 years) (p < 0.0001) (Table 2; Figure 1d). The unadjusted analysis showed a strong association between age >60 years and mortality risk (Figure 2; Table 3). Even after multivariable adjustment, it remained a significant predictor of mortality (HR = 1.15, p = 0.07), along with higher BMI, donor age and an increased infection burden in the first year after transplantation (Table 3). Sensitivity analyses confirmed the robustness of the association between older recipient age and mortality, whereas no consistent association was observed for graft failure (Supplementary Table S3). Analyses accounting for the COVID-19 pandemic period did not materially change the results (Supplementary Table S4).

Infections were the most common cause of death at 40% of cases. Pulmonary infections were the most frequent, affecting 55.6% of patients. Fatal cases included COVID-19 pneumonia, pulmonary aspergillosis, and Legionella pneumonia. Bloodstream infections and sepsis were the second most common causes of death. Pathogens were detected in approximately one-third of deaths, with Aspergillus spp., HSV, and E. coli being most common.

In recipients >60 years, the median time to death was shortest (24 months), compared with 28 months in the 40–60 years group and 33 months in <40 years (p = 0.034).

The first post-transplant year emerged as a critical period linking early infection burden to long-term outcomes. Higher infection burden during the first post-transplant year was independently associated with recipient age >60 years, CMV D+/R− serostatus, higher donor age, and greater HLA mismatch count (Supplementary Table S5).