Multimodal biomarker approach using urinary prostaglandin E2, kidney injury molecule-1 and resistive index for assessing kidney adaptation in transplant recipients

Abstract

Background

Kidney transplantation is the standard treatment for end-stage renal disease (ESRD). Traditional monitoring with serum creatinine and renal biopsy is limited by delayed detection and invasiveness. Urinary biomarkers including prostaglandin E2 (PGE2) and kidney injury molecule-1 (KIM-1), together with doppler resistive index (RI) may detect early allograft adaptation and subclinical injury. This study aims to evaluate the potential of PGE2, KIM-1 and RI in reflecting post-transplant kidney adaptation.

Methods

We conducted a prospective study of 59 adult living-donor kidney transplant recipients in 2023–2024. Urine and blood were collected pre-transplant and serially post-transplant within 3 months. PGE2 was measured at baseline, day 2, and 1 month, while KIM-1 was analyzed at 1 week. Primary outcome was functional delayed graft function (fDGF). Associations with glomerular filtration rate and resistive index were analyzed using correlation and multivariate regression.

Results

Urinary PGE2 on postoperative day 2 was significantly higher in patients with immediate graft function (IGF) and showed the best diagnostic value for fDGF (AUC 0.655; sensitivity 82%; specificity 52%). Urinary KIM-1 was not significantly different between groups and limited predictive value (AUC 0.647). RI at both segmental and arcuate arteries during the first postoperative week was significantly higher in recipients with fDGF (p = 0.045 and p = 0.028, respectively). Day-2 PGE2 correlated with 1st month RI whereas preoperative PGE2 correlated with 1st month postoperative RI at segmental but not arcuate arteries. Multivariate regression reveals KIM-1 and preoperative biomarkers were not predictive while 1-month PGE2 independently predicted 3-month eGFR.

Conclusions

Elevated urinary PGE2 may reflect adaptive hyperfiltration and showed modest predictive value for early graft function. The resistive index was associated with fDGF, while KIM-1 showed inconsistent associations. These findings suggest potential complementary roles of PGE2 and resistive index in the early assessment of kidney transplant adaptation.

Background

Kidney transplantation is the treatment of choice for end-stage renal disease, offering improved survival and quality of life compared to dialysis [1, 2]. Despite advancing immunosuppression strategies, clinical challenges persist to improve long-term graft survival. Studies have shown that early posttransplantation kidney function is essential for long-term graft survival [3, 4]. Multiple factors, including donor age, sex, and immunologic compatibility, showed significant impact on kidney function after transplantation. In addition, perioperative parameters such as cold and warm ischemia times were also identified as factors influencing graft survival [5, 6].

To measure changes in allograft function, serum creatinine and renal biopsy is primarily used. However, these measurements have several drawbacks. Elevated serum creatinine occurs later in injury and has limited ability to predict or evaluate injury progression, and is not specific for the type of injury [7]. In patients with organ transplantation, prediction of glomerular filtration rate (GFR) is often overestimated, as creatinine metabolism varied due to a spectrum of biological and therapeutic conditions in this population [8]. Renal biopsy is the gold standard for detecting pathological changes in kidney transplant patients, but it is invasive and cannot be performed serially, so it has limited use in routine transplant monitoring [7, 9].

In order to evaluate dynamic changes after renal transplantation, there has been growing interest in urinary biomarkers and radiological diagnostics. Prostaglandin E2 (PGE2) is a lipid mediator with vasodilatory and immunomodulatory properties, which plays a key role in maintaining renal perfusion and tubular function [10]. Increased production in PGE2 has been associated with hyperfiltration and renal fibrosis, which can be detected even before the onset of clinical albuminuria [11, 12]. Meanwhile, kidney injury molecule-1 (KIM-1) is a transmembrane protein upregulated in proximal tubular injury, which serves as an early marker of subclinical tubular stress [13]. In chronic kidney disease patients, KIM-1 has been shown to outperform blood urea nitrogen and serum creatinine as predictors of histopathological changes in proximal tubules, and several studies have shown association between KIM-1 with delayed graft function. However, timing of biomarker measurements and its predictive role in graft outcome measures still needs further exploration [7, 14, 15].

Ultrasonography is a noninvasive imaging modality that can be used to evaluate renal allografts. Doppler ultrasonography enables the measurement of the intrarenal resistive index (RI), which has been proposed as a useful parameter to evaluate graft function. Several studies have suggested that an elevated RI is associated with an increased risk of kidney graft dysfunction and early graft loss following transplantation [16, 17].

Understanding their dynamic changes in the early post-transplant phase could offer novel insights into graft adaptation, early dysfunction, or subclinical injury, which could guide early therapeutic strategies and improve long-term outcomes. This study aims to investigate the urinary levels of PGE2 and KIM-1 and the resistive index of kidney transplant recipients within the first three months post-transplant, and to evaluate their potential as novel biomarkers reflecting kidney adaptation during this critical period.

Methods

Study design and setting

This was a prospective observational following a cohort of living donor kidney transplantation recipients at Dr. Cipto Mangunkusumo Hospital, a tertiary hospital in Jakarta, Indonesia. Patients were followed up for 3 months and primary data was utilized. The study was conducted in accordance with the Declaration of Helsinki and was approved by the local ethics board of Universitas Indonesia, ND- 302/UN2.F1/ETIK/PPM.00.02/2025.

Participants

Eligible participants were adult kidney transplant recipients aged ≥ 18 years who received a kidney from living donors between 2023 and 2024. Donors with a history of proteinuria, estimated glomerular filtration rate (eGFR) < 75 ml/min/1.73 m², uncontrolled hypertension, diabetes, renal cysts, or recurrent renal stones were excluded. Recipients who met the inclusion criteria and consented to a 3-month follow-up were included in the study.

Procedure

Serial urinary samples and fresh blood samples were obtained before transplant and in the posttransplant period at 2 days, 1 week, 1 month, and 3 months. These time points were selected to broadly reflect different phases of post-transplant adaptation, including early hemodynamic changes, subacute injury and recovery, and later functional stabilization [18–21]. All samples were submitted to the biochemical laboratory for analysis shortly after collection and stored at − 80 °C until further analysis. Routine pathology tests were performed on the available samples, including serum creatinine, urea, urine creatinine, and albumin-to-cereatinine ratio (ACR). Urinary PGE2 was measured at earlier post-transplant time points, whereas KIM-1 and renal Doppler resistive index were assessed starting from day 7 post-transplantation based on clinical and logistical considerations. Doppler ultrasonography measurements were performed by a single experienced radiologist in accordance with standardized protocols to minimize interobserver variability.

PGE2 levels were analyzed using competitive enzyme linked immunosorbent assay (ELISA) (Prostaglandin E2 SimpleStep, Abcam, USA) from preoperative urinary samples and at 2 days and 1 month posttransplant. Urinary KIM-1 was measured at 1 week posttransplant using sandwich ELISA assay (R&D Corp, USA). A standard curve was generated to calculate the concentration of samples.

Variables and outcomes

Patient characteristics were obtained at enrollment, including demographic features (age, sex), physical examinations (body mass index/BMI), medical history (etiology of kidney disease, duration of kidney disease, presence of comorbidities), and donor characteristics (relation with donors, gender of donors). Furthermore, renal imaging was obtained before transplant to calculate for baseline kidney volume and renal resistive index. Renal resistive index (RI) was calculated using peak systolic volume (PSV) and end diastolic volume (EDV) from doppler ultrasonography using the formula RI = . Kidney volume was calculated using ellipsoid formula . During transplant, factors such as donated kidney weight, warm ischemic time (WIT), and cold ischemic time (CIT) were also obtained. However, due to the limited sample size and number of outcome events, inclusion of a large number of covariates was not feasible. Therefore, only selected clinically relevant variables were included in the multivariable analysis to minimize the risk of overfitting. Variable selection was guided by clinical relevance and data availability.

The primary outcome variable was functional delayed graft function (fDGF), defined as serum creatinine decreased by less than 10% per day immediately after surgery during three consecutive days for more than one week, regardless of dialysis. Immediate graft function (IGF) was defined as recovery of serum creatinine > 70% in the first week posttransplantation [22]. This definition was selected to better capture early graft functional recovery in living donor kidney transplant recipients, as dialysis-based definitions may fail to capture early or mild impairments in graft function. Secondary outcome was allograft function, measured by estimated glomerular filtration rate (eGFR) as calculated using the 2021 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula. Furthermore, to measure hyperfiltration, RI was also assessed to present renal blood flow at different time points.

Statistical analysis

Statistical analyses were performed using the IBM SPSS statistical package version 27.0 (IBM Corp, USA). Shapiro-Wilk test was used to determine data distribution. Continuous variables were expressed as mean ± standard deviation (SD) for normally distributed data, and median with interquartile range (IQR) for non-normally distributed data, while categorical variables were expressed as frequency with percentages. Independent t-test was used to determine differences between groups for parametric continuous variables, while Mann-Whitney test and Chi-Square test were performed for nonparametric and categorical data respectively. Additionally, receiver operating characteristic (ROC) curve analysis was performed to evaluate the diagnostic performance of urinary PGE2, KIM-1, and Doppler resistive index (RI) at different postoperative time points for predicting fDGF. The area under the ROC curve (AUC) was calculated with corresponding 95% confidence intervals (CI) to assess overall discriminatory ability.

Biomarker concentrations of PGE2 and KIM1 were analyzed using Spearman correlation analysis to assess association with the resistive index as continuous variable at different time points. A p-value < 0.05 was regarded as statistically significant. Multivariate linear regression was performed to estimate independent associations of PGE2 and KIM1 with eGFR. To this purpose, variables with a correlation p-value < 0.25 in univariable analyses, as well as clinically relevant covariates were entered into a multiple linear regression model. Model assumptions were evaluated by inspecting residual plots for linearity, homoscedasticity, and normality. Regression coefficients (β) with 95% confidence intervals (CIs) were reported, representing the change in eGFR per unit change in the predictor variable.

Results

A total of 59 patients who had received living donor renal transplant and consented to the 3 month follow up of the study were recruited. At baseline, the median age of subjects was 32 (26–43) years old. CKD causes were diverse, with hypertensive nephropathy being the most common (49.2%), followed by glomerular disease (17%) and diabetic nephropathy (15.3%). Other causes encompasses polycystic kidney disease, reflux nephropathy, single kidney, and drug-induced kidney disease. The majority of patients received a kidney from biologically related donors (57.6%). All patients received triple-drug immunosuppressive regimens consisting of tacrolimus, mycophenolate mofetil, and corticosteroids. There were statistically significant differences in sex, preoperative urea, serum creatinine, eGFR, and anuria between IGF and fDGF recipients. Table 1 showed characteristics of patients.

Table 1.

Characteristics of participants

Variables	Baseline (n = 59)	IGF (n = 45)	fDGF (n = 14)	p-value
Recipient characteristics
Age, years	32.0 (26.0–43.0)	30.0 (26.0–42.0)	38.0 (30.5–47.5)	0.076
Sex
Male	37 (62.7)	25 (55.6)	12 (85.7)	0.042*
Female	22 (37.3)	20 (44.4)	2 (14.3)
Etiology of kidney disease†				0.144
Hypertension	29 (49.2)	22 (48.9)	2 (14.3)
Diabetes mellitus	9 (15.3)	5 (11.1)	4 (28.5)
Glomerular disease	10 (17.0)	7 (15.6)	3 (21.4)
Autoimmune disease	7 (11.9)	6 (13.3)	1 (7.1)
Others	9 (15.3)	4 (18.2)	5 (13.5)
Comorbidities
Hypertension	38 (64.4)	28 (62.2)	10 (71.4)	0.530
Diabetes mellitus	9 (15.3)	5 (11.1)	4 (28.6)	0.113
Chronic liver disease	10 (16.9)	58 (17.8)	2 (14.3)	0.761
Body mass index, kg/m2	21.64 (20.16–24.60)	21.58 (19.37–24.44)	22.81 (21.12–26.07)	0.233
Underweight (< 18.5)‡	10 (16.9)	9 (20)	1 (7.1)	0.437
Normal (18.5–22.9)	28 (47.5)	22 (48.9)	6 (42.9)
Overweight (23.0–24.9)	7 (11.9)	4 (8.9)	3 (21.4)
Obese (≥ 25.0)	14 (23.7)	10 (22.2)	4 (28.6)
Duration of chronic kidney disease, months	25.0 (15.0–54.0)	23 (15.5–57.0)	25.0 (11.25–33.75)	0.336
Donor characteristics
Donated kidney weight, grams	182.5 (151.80–210.51)	183.8 (150.05–213.16)	178.95 (154.98–198.83)	0.402
Donor age, years	47.64 ± 11.73	46.52 ± 11.27	51.23 ± 12.94	0.209
Biological relationship with recipient				0.057
Related	34 (57.6)	29 (64.4)	5 (35.7)
Unrelated	25 (42.4)	16 (35.6)	9 (64.3)
Gender mismatch				0.942
Yes	30 (50.8)	23 (51.1)	7 (50)
No	29 (49.2)	22 (48.9)	7 (50)
Baseline laboratory values
Urea, mg/dL	64.2 (49.2–85.6)	70.6 (52.45–102.75)	51.35 (43.85–67.93)	0.044*
Serum creatinine, mg/dL	6.5 (5.0–8.6)	6.9 (5.40–8.65)	4.95 (4.0–7.23)	0.028*
eGFR, ml/min/1.73m2		9.36 ± 3.62	12.98 ± 4.78	0.004*
Urine creatinine, mg/dL§	48.39 (25.23–101.41)	38.91 (18.24–101.41)	58.93 (39.07–124.57)	0.092
Albumin-to-creatinine ratio§	1548.25 (373.4–5314.61)	2073.46 (276.11–5463.94)	1152.99 (405.82–1706.99)	0.524
Anuria	24 (40.7)	22 (48.9)	2 (14.3)	0.021*
Baseline imaging values
Kidney volume	118.59 ± 29.60	118.03 ± 32.78	120.39 ± 16.41	0.721
RI (a. segmental)	0.61 ± 0.06	0.61 ± 0.06	0.62 ± 0.06	0.365
RI (a. arcuate)	0.58 ± 0.06	0.59 ± 0.06	0.58 ± 0.06	0.586
Intraoperative factors
Warm ischemic time (WIT), minutes	41.68 (36.67–45.03)	40.8 (36.14–46.18)	43.84 (37.18–45.56)	0.454
Cold ischemic time (CIT), minutes	37.73 (33.02–49.87)	38.35 (34.64–48.78)	34.17 (27.7–34.39)	0.247

Continuous variables are presented as mean ± SD for normally distributed data and median (interquartile range) for non-normally distributed data. Categorical variables are expressed as count (%)

^†Patients may have more than one

^‡BMI are categorized based on World Health Organization Asia-Pacific Task Force

^§Calculated from patients without anuria, n = 35

*Denotes significant value (p < 0.05)

Comparison of biomarkers and resistive index between recipients with and without fDGF

Urinary PGE2 measured at the 2nd postoperative day was significantly higher in the group with IGF. However on the 1st month postoperative, median PGE2 was higher in the group with fDGF, although this difference is not statistically significant. KIM-1 was not significantly different between the two groups. Ultrasonography measurements revealed resistive index at both segmental artery and arcuate artery were significantly associated with fDGF at first week post operative, but no significant difference was found at later time points. Table 2 shows the comparison of urinary biomarkers levels between recipients with IGF and fDGF. Association between urinary KIM-1, PGE2 and RI with fDGF status post transplantation can be seen in Fig. 1.

Table 2.

Comparison of urinary biomarkers levels between recipients with IGF and fDGF

	Total (n = 59)	IGF (n = 45)	fDGF (n = 14)	p-value
Urinary biomarkers
PGE2, preoperative§	89.25 (64.0–125.9)	91.0 (70.5–132.75)	79.10 (55.10–97.80)	0.186
PGE2, 2nd day postoperative	69.5 (55.8–98.5)	69.70 (56.35–106.35)	62.20 (48.10–69.50)	0.030*
PGE2, 1st month postoperative	86.6 (67.2–119.8)	81.90 (65.30–117.10)	94.40 (78.20–147.20)	0.845
KIM-1, 1st week postoperative	0.80 (0.39–1.16)	0.54 (0.33–0.92)	0.93 (0.62–1.41)	0.099
Doppler resistive index, a. segmental
1st week postoperative	0.68 (0.61–0.74)	0.67 (0.60–0.74)	0.71 (0.69–0.75)	0.045*
1st month postoperative	0.67 ± 0.07	0.66 ± 0.07	0.70 ± 0.05	0.141
3rd month postoperative	0.67 ± 0.06	0.67 ± 0.06	0.68 ± 0.07	0.889
Doppler resistive index, a. arcuate
1st week postoperative	0.64 (0.55–0.69)	0.62 (0.55–0.67)	0.68 (0.61–0.75)	0.028*
1st month postoperative	0.61 ± 0.07	0.61 ± 0.07	0.64 ± 0.05	0.061
3rd month postoperative	0.62 ± 0.06	0.62 ± 0.06	0.62 ± 0.08	0.890

^§Calculated from patients without anuria, n = 35

Fig. 1.

Association between urinary KIM-1 (A), PGE2 (B), and resistive index (C) with fDGF status posttransplantation. IGF, immediate graft function; fDGF, functional delayed graft function; KIM-1, kidney injury molecule-1; PGE2, prostaglandin E2

Diagnostic value of urinary biomarkers for fDGF

The area under the curve (AUC) values were 0.640, 0.655, and 0.396 for PGE2 preoperative, 2nd day postoperative, and 1st month postoperative, and 0.647 for KIM-1 (Fig. 2). ROC curve analyses revealed that PGE2 on the 2nd postoperative day has a sensitivity of 82% and specificity of 52% for diagnosing fDGF, which was the highest sensitivity and specificity compared to PGE2 at other time points and KIM-1. For resistive index, measurements at both arteries were of poor discriminatory value, with measurement at 1st week showing the greatest AUC (Table 3).

Fig. 2.

ROC curve for PGE2 preoperative, PGE2 2nd day postoperative, PGE2 1 month postoperative, KIM-1 2nd day postoperative, and RI at 1st week postoperative

Table 3.

Predictive value of urinary biomarkers for fDGF

Biomarker	Time after transplant	AUC (95% CI)	p-value	Cutoff (ng/mL)	Sensitivity	Specificity
PGE2	Preoperative	0.640 (0447–0.833	0.186	116.2	90.9%	36%
	2nd day postoperative	0.655 (0.472–0.837)	0.144	69.6	81.8%	52%
	1st month postoperative	0.396 (0.181–0.611)	0.328	141.4	81.8%	12%
KIM-1	1st week postoperative	0.709 (0.542–0.876)	0.048	0.48	90.9%	48%
RI, a. segmental	1st week postoperative	0.631 (0.437–0.824)	0.099	0.675	90.9%	52%
RI, a. arcuate	1st week postoperative	0.649 (0.465–0.833)	0.094	0.59	90.9%	48%

Correlation between urinary biomarkers and resistive index

Table 4 showed correlation coefficients between urinary biomarkers and renal resistive index. Only PGE2 at 2nd day postoperative was significantly correlated with 1st month resistive index (a. segmental r= − 0.349; 0.007, a. arcuate r= – 0.262; p = 0.047). Preoperative PGE2 was associated with 1st month postoperative RI at a. segmental, but not at a. arcuate (r= – 0.419; p = 0.011).

Table 4.

Correlation coefficients between urinary biomarkers and renal resistive index at different timepoints posttransplantation

	Time after transplant	PGE2, preoperative	PGE2, D2 postoperative	PGE2, 1 mo postoperative	KIM-1, 1 wk postoperative
Resistive index, a. segmental	1st week postoperative	–0.219	–0.173	0.020	0.280*
	1st month postoperative	–0.419*	–0.349*	–0.029	0.077
	3rd month postoperative	–0.138	–0.109	0.128	0.136
Resistive index, a. arcuate	1st week postoperative	–0.201	–0.157	0.083	0.169
	1st month postoperative	–0.274	–0.262*	–0.007	–0.088
	3rd month postoperative	–0.273	–0.204	–0.054	0.145

*p < 0.05; **p < 0.001

Correlation between urinary biomarkers and resistive index to kidney function

Multivariate linear regression was performed for correlation between KIM-1 and PGE2 with eGFR at multiple time points as the dependent variable. KIM-1 was only weakly correlated with eGFR at 1 week postoperative, but not on any other time points. PGE2 at 1 month posttransplant was shown to be weakly negatively correlated with serum creatinine at all time points, with the only significant correlation was with 3 months postoperative eGFR (β = 0.104 [95% CI 0.007–0.200). Preoperative PGE2 and resistive index did not show statistically significant correlations with eGFR at any time points. Multivariate linear regression of eGFR at multiple time points can be seen in Table 5.

Table 5.

Multivariate linear regression of eGFR at multiple time points

	Regression coefficient from multiple linear regression model†
	Estimate	95% CI	p-value
eGFR, 1 week postoperative
PGE2, preoperative	0.026	–0.244–0.642	0.363
PGE2, 2nd day	0.116	–0.030–0.262	0.117
PGE2, 1st month	0.144	0.001–0.288	0.049
KIM-1, 1st week	–4.844	–10.325–0.637	0.082
RI a. segmental, 1st week	–39.564	–129.475–50.347	0.380
RI a. arcuate, 1st week	–18.822	–106.772–69.128	0.668
RI a. segmental, 1st month	–121.105	–246.520–4.311	0.058
RI a. arcuate, 1st month	–37.590	–171.885–96.704	0.576
RI a. segmental, 3rd month	34.311	–103.391–172.014	0.502
RI a. arcuate, 3rd month	11.559	–120.813–143.931	0.861
eGFR, 1 month postoperative
PGE2, preoperative	0.007	–0.127–0.141	0.915
PGE2, 2nd day	0.023	–0.096–0.142	0.702
PGE2, 1st month	0.103	–0.012–0.218	0.079
KIM-1, 1st week	–2.651	–7.087–1.785	0.235
RI a. segmental, 1st week	4.291	–67.776–76.359	0.905
RI a. arcuate, 1st week	–6.362	–76.380–63.655	0.856
RI a. segmental, 1st month	–41.701	–144.723–61.322	0.815
RI a. arcuate, 1st month	–56.032	–161.811–49.746	0.292
RI a. segmental, 3rd month	–32.153	–141.474–77.169	0.557
RI a. arcuate, 3rd month	–26.509	–131.448–78.430	0.613
eGFR, 3 months postoperative
PGE2, preoperative	–0.031	–0.142–0.081	0.575
PGE2, 2nd day	0.056	–0.044–0.156	0.267
PGE2, 1st month	0.104	0.007–0.200	0.036*
KIM-1, 1st week	–2.412	–6.180–1.356	0.204
RI a. segmental, 1st week	–19.730	–80.073–40.613	0.514
RI a. arcuate, 1st week	–22.668	–81.195–35.858	0.439
RI a. segmental, 1st month	–50.684	–136.648–35.280	0.241
RI a. arcuate, 1st month	–69.745	–157.378–17.888	0.116
RI a. segmental, 3rd month	–33.903	–125.660–57.854	0.461
RI a. arcuate, 3rd month	–52.201	–139.331–34.930	0.234

^†Adjusted for age, sex, etiology, hypertension, diabetes, BMI, biological relationship with recipient, kidney weight, kidney volume, WIT, CIT, and baseline serum creatinine

Discussion

Early posttransplant renal functions have emerged as an important predictor of long-term graft survival [23]. DGF has a critical impact on both graft function and long-term graft survival [24]. There is growing interest in urinary biomarkers and radiological diagnostics for detecting early kidney injury and predicting prognosis in kidney transplant recipients. While serum creatinine has its limitations in estimating renal function – notably as it is influenced by various factors such as liver function, muscle mass, fluid status, and medications – it remains as the most widely used marker of GFR following kidney transplantation due to its convenience and low cost [25, 26]. Several studies have shown that early post-transplant serum creatinine is a strong independent predictor of graft survival. A study by Hariharan et al. reported that a value of > 1.5 mg/dL is a significant cut off predictor for both graft half-life and rejection episodes [23, 27].

This study showed that urinary KIM-1 at 1 week post-transplant was higher in fDGF patients, but the result is not significant. KIM-1 is a transmembrane glycoprotein that is upregulated in injured renal tubules. Thus, increased urinary KIM-1 is associated with the degree of renal damage [28–30]. Our result aligned with a study by Tabarnero et al. in which there is no significant difference between KIM-1 values at 1 week and the incidence of DGF. However, previous studies have also found that KIM-1 in the early postoperative period is significantly associated with graft function, with higher KIM-1 on the first postoperative day linked to fDGF [31]. Since urinary biomarkers of renal injury rise at different time points, these findings may suggest that KIM-1 is an early marker of tubular damage. Its predictive value was not significant at one week post-transplant, likely due to repair mechanisms or overlapping contributions from other pathways [19]. Interestingly, KIM-1 at day 7 showed the highest AUC among the evaluated biomarkers and was the only parameter that reached statistical significance, suggesting its potential discriminative ability in predicting fDGF. This finding is consistent with previous studies reporting that urinary KIM-1 demonstrates moderate to good diagnostic performance in kidney injury, with AUC values ranging from approximately 0.62 to 0.86 depending on the population and clinical context. KIM-1 is a well-established marker of tubular injury, and its expression correlates with the severity of renal damage [32]. In addition, urinary KIM-1 has been reported to be more sensitive than conventional markers such as creatinine in detecting early kidney injury. Beyond its role as a biomarker, KIM-1 may also be involved in renal recovery and tubular regeneration, further supporting its biological relevance in the post-transplant setting. In addition, previous studies have demonstrated that urinary KIM-1 levels increase during the first post-transplant week and are associated with post-transplant outcomes, including renal function stabilization, supporting the clinical relevance of this time window [23]. While the correlation between KIM-1 and renal damage has been established, its precise function remains incompletely understood. KIM-1 has also been shown to increase during the recovery phase of kidney damage, playing a role in the removal of apoptotic cells and necrotic tissue by transforming tubular cells into phagocytic cells. KIM-1 also increases phagocytosis of albumin, which alleviates tubular damage. Furthermore, upregulation of KIM-1 can promote repair and proliferation of renal epithelial cells after kidney ischemia damage in in vitro studies [5].

PGE2 is a lipid mediator which can be produced by renal cells. It is a major signalling molecule in the kidney that has been implicated in chronic kidney disease as a marker of hyperfiltration [11]. To the best of the author’s knowledge, this is the first study to investigate the evolution of PGE2 after renal transplantation. Previous studies in single functioning kidney patients showed an increase of PGE2 as an early marker of adaptive hyperfiltration in response to increase fluid flow shear stress and tensile stress [12]. Hyperfiltration is a physiological response following nephron loss, characterized by increasing glomerular capillary pressure and glomerular enlargement. This response is initially adaptive, as seen in adult living kidney donors with increasingly GFR following kidney donation [33].

Our study showed that PGE2 at day 2 post-transplant was lower in patients with fDGF. Furthermore, PGE2 within 1 month post-transplant was positively correlated with eGFR across all time points. To determine whether this association was due to hyperfiltration, we performed a correlation analysis with renal resistive index to represent intrarenal arterial blood flow. Notably, negative correlation was found between RI and PGE2, indicating that its protective effect may not only be attributed to hyperfiltration. PGE2 is a potent vasodilator through activation of EP2 receptors, which increases renal blood flow and GFR. It also regulates sodium and water reabsorption, promoting natriuresis along with increased renal perfusion [34]. In vivo study has shown that activation of EP4 receptors by PGE2 modulates immune function of dendritic cells in ischemia-reperfusion injury. The PGE2-EP4 pathway is protective towards reactive oxygen species (ROS) induced by reperfusion injury, reduces cell apoptosis, and suppresses the production of pro-inflammatory factors by dendritic cells [35]. This protective effect has also been shown in acute myocardial injury, where endogenous PGE2 protects cardiomyocytes from ischemia-reperfusion injury mediated by the EP4 receptors [36]. This alludes that PGE2 is not only a marker of adaptive hyperfiltration in renal transplant recipients, but also protective against ischemia reperfusion injury, which is inevitable in kidney transplantation where blood supply is temporarily reduced, then restored [37].

Serial measurements in this study indicated that PGE2 values decreased on postoperative day 2, followed by an increase after one month. Whether PGE2 continues to rise and its implications towards long-term renal transplant outcomes still need further elaboration. Despite being beneficial in the early postoperative period, persistent hyperfiltration may become pathological as increased mechanical stress on glomeruli can lead to glomerulosclerosis and tubular damage [33]. Geurts et al. shows that during a 15-year follow-up period, increased urinary PGE2 is associated with higher risk of incident eGFR < 60 ml/min/1.73m².¹¹ Rise in PGE2 also precedes albuminuria in children with single kidney, indicating that PGE2 can be an early biomarker for hyperfiltration-mediated injury. While the individual predictive power of PGE2 was limited by modest AUC values, this biomarker provides non-invasive insights into early graft hemodynamics and tubular injury that standard parameters, such as serum creatinine, may lack during the hyperacute phase [38]. Notably, the 82% sensitivity observed for PGE2 suggests significant clinical utility. In this hyperacute setting, prioritizing sensitivity is vital to ensure that potential graft dysfunction is not overlooked, even if further confirmatory testing is required.

Resistive index in doppler ultrasonography is a hemodynamic index used to measure blood flow resistance in the kidney [16]. A decrease in renal blood flow is associated with prolonged cold ischemic time and DGF [39]. Our study shows that RI at both segmental artery and arcuate artery were significantly associated with fDGF at first week post operative. Our result is aligned with a study by Bakirdogen et al. in which RI measurements in the first week post transplant are significantly associated with higher DGF [40]. High RI during the first week may be associated with DGF as it may reflect early hemodynamic alterations in the transplanted kidney such as acute tubular necrosis due to ischemia-reperfusion injury, acute rejection, obstruction and vascular complications such as arterial stenosis and renal vein thrombosis [41]. These acute changes may impair renal perfusion and are linked to the development of DGF, making RI a significant predictor in the first few postoperative weeks.

Our study found that RI has no statistically significant correlation with eGFR at any time point during multivariate linear regression analysis. Previous studies have reported inconsistent findings, Gupta et al. reveals a significant inverse correlation between RI and eGFR, whereas Heine et al. reported no significant association between renal resistance indices and glomerular filtration rate (GFR) [42, 43]. RI is strongly influenced by systemic factors such as pulse pressure, vascular compliance and perioperative status rather than solely reflecting intrarenal vascular resistance which may confound its associated with eGFR. RI is known to predict DGF more reliably than eGFR values, as it primarily reflects hemodynamic disturbances more compared to direct filtration capacity [43]. In addition, previous studies have shown that renal hemodynamic adaptation following transplantation, including increases in resistive index due to hyperfiltration, tends to peak around day 7 and gradually decline toward baseline by day 30. Hyperfiltration represents an adaptive response characterized by increased intraglomerular pressure and renal blood flow, which may transiently elevate resistive index values without directly reflecting stable renal function [44, 45]. Beyond the first month, particularly up to 3 months post-transplantation, resistive index values tend to stabilize and may better reflect steady-state graft function rather than transient adaptive changes [40]. This may explain why RI measurements at day 7 primarily reflect transient hemodynamic changes rather than sustained renal function, thereby limiting its correlation with eGFR. The predictive value of RI for eGFR may also appear weaker later as renal function may also be influenced by other factors besides vascular factors such as immune response, rejection, infections, and drug toxicity [46]. Given the inconsistency across studies and the relatively small sample size of our study, which may limit the statistical power and reduce the ability to detect an independent contribution of RI to eGFR, further research with larger cohorts is needed to clarify the relationship between RI and eGFR.

In summary, while the biomarkers and resistive index investigated in this study currently preclude their use as standalone predictors due to modest AUC values, they offer valuable adjunctive insights alongside established clinical markers. Their primary utility lies in enhancing the early screening process for allograft injury, providing a non-invasive ‘early warning’ during the hyperacute phase. Future multi-center studies with larger cohorts are warranted to further validate these findings and refine the optimal ‘cut-off’ values necessary for standardized clinical implementation.

This study has several strengths. It provides a comprehensive evaluation of multiple noninvasive biomarkers, including urinary PGE2, KIM-1, and Doppler-derived resistive index, within the early post-transplant period. The integration of both biochemical and hemodynamic parameters offers a more holistic assessment of graft adaptation. In addition, the prospective follow-up and standardized data collection enhance the internal validity of the findings. The inclusion of only living donor kidney transplant recipients also represents a strength, as it provides a relatively homogeneous population and minimizes potential confounding factors such as prolonged cold ischemia time and higher rates of delayed graft function, thereby allowing a clearer assessment of correlation between PGE2, KIM-1, and post-transplant outcomes.

However, several limitations should be acknowledged. First, the biomarkers were measured at different time points, which may limit direct comparability of their diagnostic performance. Urinary PGE2 was assessed earlier, whereas KIM-1 and resistive index were measured starting from day 7. Although this may reflect distinct pathophysiological processes, it may also have influenced the observed performance of each biomarker [15, 47, 48]. In particular, KIM-1 may have reached peak levels at earlier time points that were not captured in this study, which could have led to an underestimation of its predictive performance [15]. Second, the relatively small sample size and limited number of outcome events reduce statistical power and increase the risk of type II error and model overfitting. In addition, the number of covariates included in the multivariable models relative to the sample size may affect model stability [49]. However, all covariates were selected a priori based on clinical relevance and existing literature to ensure appropriate control for potential confounding. Furthermore, the multivariable analyses were exploratory in nature and not intended for predictive modeling; therefore, the findings should be interpreted with caution, particularly with respect to effect size estimates and statistical significance. Third, not all established risk factors for delayed graft function, including donor-related, perioperative, and immunological factors, were included in the multivariable analysis, which limits the ability to assess the incremental predictive value of the studied biomarkers. Fourth, the definition of functional delayed graft function (fDGF) used in this study is not fully standardized, which may limit comparability with prior studies. Fifth, conventional clinical markers such as serum creatinine and blood urea nitrogen were not included in the comparative diagnostic analysis due to inconsistent availability across time points, which may limit clinical interpretability. In addition, the single-center design, absence of an external validation cohort, and exclusive inclusion of living donor recipients may limit the generalizability of the findings, particularly to deceased donor kidney transplant populations.

Conclusion

Elevated urinary PGE2 may reflect adaptive hyperfiltration in renal transplant recipients and demonstrated modest predictive performance for early graft function. Urinary KIM-1 showed a trend toward association with transplant outcomes, although findings were not consistent across time points. The resistive index in the first postoperative week was associated with fDGF.

Overall, these findings suggest that PGE2 and resistive index may have potential as complementary biomarkers in the early assessment of kidney transplant adaptation. However, given the relatively small sample size and modest discriminatory performance, these results should be interpreted with caution, and further studies with larger cohorts are warranted to validate these findings.

Abbreviations

ESRD

End stage renal disease

PGE2

Prostaglandin E2

KIM-1

Kidney injury molecule − 1

RI

Resistive index

IGF

Immediate graft function

fDGF

Functional delayed graft function

eGFR

Estimated glomerular filtration rate

ELISA

Enzyme linked immunosorbent assay

ROC

Receiver operating characteristic

Author contributions

M. B. H. M., A. R., N. R., P. B., D. A., S. B. R. E. M, A. R. P. contributed to the conceptualization of the study. B. A., M. M. A., D. S., A. I. K., E. S., D. E., A. S., K. W., T. K. A., J. M., were responsible for data collection. P. G. N., M. C. H., J. M., A. R. T. S. H were responsible for manuscript writing. Supervision was provided by M. B. H. M., A. R., N. R., P. B., D. A., S. B. R. E. M, A. R. P., A. I. K., D. E. All authors reviewed and approved the final version of the manuscript.

Funding

No financial support was received for the writing or publication of this manuscript.

Data availability

All data supporting the findings of this study are available within the paper.

Declarations

Ethics approval and consent to participate

This study conforms to the principles outlined in the 1964 Declaration of Helsinki and its subsequent amendments. Ethical approval was obtained from the local Ethics Committee of Universitas Indonesia (Protocol No. KET-1382/UN2.F1/ETIK/PPM.00.02/2022). Written informed consent to participate was obtained from all participants prior to their inclusion in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.