Skip to main content
Renal Failure logoLink to Renal Failure
. 2026 Nov 11;48(1):2755138. doi: 10.1080/0886022X.2026.2755138

Perioperative dynamics of selenoproteins and risk of cardiac surgery–associated acute kidney injury

Xian Xie a,b,#, Wenbo Xue c,#, Zeyang Wang d, Zhi Li a,b, Wei Feng c, Hao Zhang a,b, Wei Zhang a,b,✉
PMCID: PMC13757821  PMID: 43015035

Abstract

Cardiac surgery–associated acute kidney injury (CSA-AKI) affects over 20% of patients and is associated with perioperative mortality and long-term renal impairment. Reliable early biomarkers remain limited. Selenoproteins, including glutathione peroxidases (GPX1, GPX3) and selenoprotein P (SEPP1), are critical for redox homeostasis, but their perioperative changes in CSA-AKI remain unclear. This study aimed to characterize selenoprotein dynamics and explore their association with renal function and CSA-AKI. In this prospective observational cohort study, 99 adult patients undergoing cardiac surgery were enrolled, serum and urine samples were collected preoperatively and at 24/48 h postoperatively. SEPP1, GPX1, GPX3, and KIM-1 were measured. Mixed-effects, correlation, and logistic regression assessed selenoprotein dynamics and associations with CSA-AKI, with an exploratory model including age and preoperative SEPP1. Among 89 analyzed patients, 22 (24.7%) developed AKI. Univariable analyses showed that lower perioperative SEPP1 levels, attenuated urinary SEPP1 increases, and greater serum GPX1 decreases were associated with higher odds of CSA-AKI; older age, hypertension, and longer operation time were also associated with CSA-AKI. Postoperatively, serum SEPP1 declined markedly, remaining lower in AKI patients. Urinary SEPP1 increased significantly in non-AKI patients. Changes in GPX1 and GPX3 differed between AKI and non-AKI groups. Serum SEPP1 correlated positively with eGFR and inversely with creatinine and cystatin C. An exploratory model incorporating age and preoperative serum SEPP1 yielded an apparent AUC of 0.818. In summary, perioperative alterations in selenoproteins (GPX1, GPX3, SEPP1) are closely linked to CSA-AKI. Preoperative SEPP1 combined with clinical variables may enable early risk prediction and support timely intervention.

Keywords: Cardiac surgery-associated acute kidney injury, selenoproteins, SEPP1, GPX3, GPX1

LAY SUMMARY

Acute kidney injury is a common complication after cardiac surgery. Selenoproteins help regulate antioxidant defense and responses to physiological stress, but their changes around cardiac surgery are not well understood.

In this prospective study of 89 adults undergoing cardiac surgery, 22 developed acute kidney injury. In 54 patients with biomarker measurements, we assessed three selenoproteins—SEPP1, GPX1, and GPX3—in blood and urine before and after surgery.

Patients who developed acute kidney injury showed different perioperative selenoprotein patterns, including lower serum SEPP1 levels and altered GPX1 and GPX3 responses. Higher serum SEPP1 was generally associated with better kidney function, and preoperative serum SEPP1 showed potential value for risk assessment when combined with age.

These findings suggest that selenoproteins may provide additional information about kidney vulnerability after cardiac surgery. Larger independent studies are needed before these markers can be used clinically.

KEY MESSAGES

  1. What is known: CSA-AKI is a frequent and serious complication, but the perioperative dynamics of selenoproteins and their association with renal vulnerability remain poorly understood.

  2. What this study adds: This study reveals that CSA-AKI patients exhibit lower serum SEPP1, altered GPX1/GPX3 dynamics, and a blunted urinary SEPP1 rise, and that a preoperative model combining age with SEPP1 shows promising discriminative ability.

  3. Potential impact: Preoperative SEPP1 may help characterize renal risk and inform future stratification research, though its clinical value requires validation in larger independent cohorts.

Introduction

Acute kidney injury (AKI), defined by a rapid decline in renal function through elevated serum creatinine or reduced urine output, is a common perioperative complication, particularly following cardiopulmonary bypass (CPB) or major surgery [1,2]. Postoperative AKI, occurring within 7 days after surgery, represents a serious clinical challenge [3]. Among its forms, cardiac surgery-associated AKI (CSA-AKI) affects over 20% of patients and is linked to increased perioperative mortality, prolonged hospitalization, and higher risks of complications and readmission [4]. In the long term, it can lead to persistent renal impairment and chronic kidney disease, significantly worsening prognosis and quality of life [5]. Early recognition is difficult, especially in high-risk patients such as the elderly or those with comorbidities, as AKI often progresses with few early symptoms [6]. Therefore, improving early identification and prevention of postoperative AKI remains a critical clinical priority.

The pathogenesis of postoperative AKI is multifactorial, involving renal hypoperfusion, ischemia-reperfusion injury (IRI), inflammation, oxidative stress, nephrotoxic exposures, and immune activation [4,7]. In cardiac surgery, CPB exacerbates these mechanisms by impairing renal perfusion and oxygen delivery, resulting in medullary ischemia and tubular damage. Contributing factors such as intraoperative hypoperfusion, postoperative volume depletion, systemic inflammatory responses, and nephrotoxic medications further increase AKI risk [1,8,9]. Surgical technique, individual patient susceptibility, and perioperative management also significantly influence AKI incidence and severity [1,7]. Early identification of key risk factors for CSA-AKI is therefore critical for risk stratification, targeted intervention, and ultimately, improved patient outcomes.

Recent research has increasingly focused on the protective role of selenoproteins in kidney injury [10–12]. This family of proteins, characterized by the incorporation of selenocysteine, is integral to cellular antioxidant defense and immunoregulation [13,14]. Key members such as glutathione peroxidase 1 (GPX1), glutathione peroxidase 3 (GPX3), and selenoprotein P (SEPP1) are vital for scavenging reactive oxygen species (ROS) and maintaining redox homeostasis [15–19]. Experimental models of AKI, particularly those involving ischemia-reperfusion and inflammation, show that selenoprotein expression changes dynamically and correlates with the degree of renal damage [20–23]. However, perioperative selenoprotein dynamics and their relationship to postoperative AKI risk—especially in cardiac surgery patients undergoing CPB—are not well understood. Therefore, this study prospectively evaluates perioperative changes in GPX1, GPX3, and SEPP1 in CPB patients, analyzes their association with renal function and injury biomarkers, and explores their potential role in the development and early risk stratification of CSA-AKI.

Methods

Study design and population

This single-center prospective observational study enrolled 99 adult patients who underwent elective cardiac surgery in the Department of Cardiothoracic Surgery at the Third Xiangya Hospital of Central South University between 2024 and 2025. The sample size was based on the number of eligible patients enrolled during the study period. Inclusion criteria were: (1) age 18–80 years; and (2) undergoing cardiac surgery requiring CPB, including coronary artery bypass grafting (CABG), valvular surgery, and aortic surgery. Exclusion criteria were: (1) severe infections; (2) diagnosed malignancy under active chemotherapy; (3) a history of chronic kidney disease (CKD); (4) liver cirrhosis or severe hepatic dysfunction; (5) previous dialysis treatment; (6) history of kidney transplantation; (7) occurrence of AKI within 7 days before surgery; and (8) absence of preoperative baseline serum creatinine (Supplementary Figure 1).

Patients were classified into AKI and non-AKI groups according to the KDIGO 2012 criteria. AKI was defined as an increase in serum creatinine of ≥26.5 µmol/L (≥0.3 mg/dL) within 48 h after surgery, or an increase to ≥1.5 times baseline within 7 days, or a urine output <0.5 mL/(kg·h) for more than 6 consecutive hours [24]. All participants provided written informed consent prior to enrollment. This study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Ethics Committee of the Third Xiangya Hospital of Central South University (approval No. 241073). This study was reported in accordance with the STROBE guidelines (Supplementary File 2). The primary objective of this study was to characterize the perioperative dynamics of serum and urinary SEPP1, GPX1, and GPX3 and to determine their associations with CSA-AKI. The secondary objective was to evaluate the relationships between these selenoproteins and perioperative renal function and injury markers. An additional exploratory objective was to assess the potential value of preoperative serum SEPP1 for predicting CSA-AKI risk.

Data collection and sample processing

Baseline clinical data were systematically collected from the hospital electronic medical record system, including age, sex, body mass index (BMI), medical history (e.g. hypertension, diabetes mellitus, and coronary artery disease), preoperative laboratory parameters, intraoperative anesthesia records, and postoperative complications. Surgery-related variables included surgical type, American Society of Anesthesiologists (ASA) classification, total operation time, CPB time, aortic cross-clamp time (ACCT), reperfusion duration, intraoperative medications, intraoperative fluid input and output, central venous pressure (CVP), arterial partial pressure of oxygen (PO2), and lowest core body temperature.

Venous blood and urine samples were collected from all patients at 24 h before surgery (preoperative 24h, pre 24h), 24 h after surgery (postoperative 24h, post 24h), and 48 h after surgery (postoperative 48h, post 48h). Blood samples were allowed to stand and then centrifuged to obtain serum, which was stored at −80 °C until subsequent analysis.

For all 99 enrolled patients, routine clinical laboratory parameters were recorded, including complete blood counts, liver function tests, kidney function indices, estimated glomerular filtration rate (eGFR), and markers of tubular function. The eGFR was calculated using the 2021 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation without a race coefficient [25]. Missing data were handled as follows: variables with a missing rate greater than 20% were excluded from linear mixed-effects models and included only in descriptive analyses, whereas variables with a missing rate of ≤20% were imputed using single median imputation.

Biomarker measurements

Among the 89 patients included in the clinical analyses (22 with AKI and 67 without AKI), complete paired preoperative and postoperative serum and urine samples were available for biomarker measurements in 54 patients (20 with AKI and 34 without AKI). The reduced sample size was attributable to the absence of specimens at predefined time points or insufficient sample volume for all assays. In these 54 patients, preserved serum and urine samples were used to quantitatively measure biomarkers related to oxidative stress and tubular injury using enzyme-linked immunosorbent assays (ELISA), including GPX1 (Shanghai Enzyme-linked Biotechnology Co., Ltd., Shanghai, China; Catalog No. ml057611; Lot No. Nov 2024), GPX3 (Wuhan Fine Biotech Co., Ltd., Wuhan, China; Catalog No. EH2531; Lot No. FN241127), SEPP1 (CUSABIO, Wuhan, China; Catalog No. CSB-EL021018HU; Lot No. W05039053), and kidney injury molecule-1 (KIM-1; Wuhan Fine Biotech Co., Ltd., Wuhan, China; Catalog No. EH0210; Lot No. FN241127). According to the manufacturers’ validation data, the intra-assay and inter-assay coefficients of variation for all ELISA kits were <10%. All assays were performed strictly in accordance with the manufacturers’ instructions.

Statistical analysis

All statistical analyses in this study were performed using R software (version 4.4.1).

For baseline characteristics, continuous variables were presented as mean ± standard deviation or median (interquartile range), as appropriate. Student’s t-test or Welch’s t-test was used for normally distributed variables, depending on variance homogeneity, whereas the Mann–Whitney U test was used for non-normally distributed variables. Categorical variables were presented as numbers and percentages and compared using the chi-square test or Fisher’s exact test, as appropriate.

For correlation analyses, Spearman’s rank correlation was used to evaluate the relationships between surgery-related indicators and biomarker changes. Correlation coefficients and corresponding p values were reported, and correlations were visualized using scatter plots with regression lines and 95% confidence intervals (CIs).

A linear mixed-effects model (LMM) was used to assess the dynamic changes of biomarkers in the AKI and non-AKI groups at pre 24h, post 24h, and post 48h time points. Group (TYPE), time (TIME), and their interaction were set as fixed effects, with sex, age, and total operation time as covariates. Subject ID was set as a random intercept. The model was fitted using the lmerTest package, and significance testing was performed using the Satterthwaite method. Predicted trends and 95% CI were extracted using ggeffects and visualized with ggplot2. The model is specified as: Value ∼ TYPE * TIME + SEX + AGE + TotalOpTime + (1|ID)

Univariable logistic regression was used to explore associations between individual clinical variables or biomarkers and CSA-AKI. Odds ratios and 95% confidence intervals were reported. Given the limited number of CSA-AKI events, an exploratory parsimonious logistic regression model incorporating age and preoperative serum SEPP1 was constructed to assess whether SEPP1 provided additional information beyond age. Adjusted odds ratios and 95% confidence intervals were reported for both variables. Model discrimination was evaluated using the area under the receiver operating characteristic curve.

For exploratory model performance and internal validation, an age-only logistic regression model was first constructed, followed by a prespecified exploratory model incorporating age and preoperative serum SEPP1, with SEPP1 scaled per 100-unit increase. Model discrimination was assessed using the area under the receiver operating characteristic curve (AUC) with 95% confidence intervals. AUCs between the two models were compared using the paired DeLong test, and improvement in model fit was evaluated using the likelihood ratio test. Calibration was assessed using the calibration intercept and slope, and overall model performance was evaluated using the Brier score. Internal validation was performed using 1,000 bootstrap resamples to estimate optimism-corrected model performance.

Results

Baseline characteristics and routine clinical parameters

A total of 99 patients undergoing elective cardiac surgery were initially enrolled. Following exclusions (severe postoperative infection, severe liver dysfunction, or sample unavailability), 89 patients were included in the final analysis (Figure 1, Supplementary Figure 1). The median age was 56 years (IQR: 51–61), and 51.7% were male. Based on the KDIGO criteria, 22 patients (24.7%) developed CSA-AKI, while 67 (75.3%) constituted the non-AKI group.

Figure 1.

Flowchart detailing research methodology for cardiac surgery patients and AKI risk analysis. This flowchart outlines the clinical study involving cardiac surgery patients, starting with preoperative sample collection (Pre24h), followed by cardiopulmonary bypass. Post-surgery, patients are categorized into "Non-AKI" and "AKI," with data collected in categories: "Patient Characteristics," "Operative Data," "Clinical Lab Tests," and "Serum/Urine: ELISA (Se-proteins)". The analysis section displays methods like "Post-Pre changes," "LME model," "Correlation analysis," "Univariable Logit," and "Parsimonious logistic model," focusing on AKI risk associations and predictions.

Study design and analytical workflow. Patients undergoing cardiac surgery with cardiopulmonary bypass were assessed preoperatively and at 24 and 48 h postoperatively. Clinical variables and serum/urinary selenoproteins were analyzed to characterize perioperative changes, associations with kidney injury, and CSA-AKI risk.

Baseline characteristics (Table 1) revealed that patients in the AKI group had a higher prevalence of hypertension (45.5% vs. 14.9%, p = 0.006) and worse preoperative cardiac function (predominantly NYHA Class III, p = 0.025). Intraoperatively, the AKI group experienced significantly longer total operative times (320.4 ± 68.5 vs. 284.2 ± 64.1 min, p = 0.026) and exhibited higher post-CPB central venous pressures (p = 0.024). Postoperatively, endotracheal intubation was more frequent in the AKI cohort (p = 0.014), whereas other complications did not differ significantly (Supplementary Table 1).

Table 1.

Baseline patient characteristics.

  Total
(n = 89)
non-AKI
(n = 67)
AKI
(n = 22)
p
Age (years) 56 (51–61) 56.0 (51.5-59) 58.5 (51.25–65.75) 0.1014
Male sex, no. (%) 46 (51.7%) 34(50.7%) 12(54.5%) 0.9493
Total hospital stay (days) 22 (19–27) 22 (18–27) 22 (19.5-25) 0.8713
Preoperative days (days) 8 (7–12) 8 (7–12) 8.5 (6-11.75) 0.5181
Postoperative days (days) 12 (10–16) 12 (10-15.5) 12 (10-15.75) 0.9619
BMI 23.31 (21.88–24.97) 23.31 (21.63–24.83) 23.67 (21.92–25.6) 0.3156
NYHA classification, no. (%)
I 3 (3.4%) 3 (4.5%) 0  
II 44 (49.4%) 37 (55.2%) 7 (31.8%)  
III 36 (40.4%) 23 (34.3%) 13 (59.1%)  
IV 6 (6.7%) 4 (6%) 2 (9.1%)  
NYHA (median) II II III 0.0250
Preoperative comorbidities
Coronary heart disease 26 (29.2%) 16 (23.9%) 10 (45.5%) 0.0968
Pulmonary Hypertension 15 (16.9%) 13 (19.4%) 2 (9.1%) 0.3406
Atrial fibrillation 21 (23.6%) 17 (25.4%) 4 (18.2%) 0.6892
Infective endocarditis 6 (6.7%) 4 (6%) 2 (9.1%) 0.6344
Rheumatic Heart Disease 30 (33.7%) 22 (32.8%) 8 (36.4%) 0.9651
Hypertension 20 (22.5%) 10 (14.9%) 10 (45.5%) 0.0064
Fatty liver disease 3 (3.4%) 3 (4.5%) 0 (0%) 0.5717
Diabetes mellitus 15 (16.9%) 9 (13.4%) 6 (27.3%) 0.1870
Pneumonia 7 (7.9%) 6 (9%) 1 (4.5%) 0.6762
History of surgery 5 (5.6%) 4 (6%) 1 (4.5%) 1.0000
Hyperlipidemia 4 (4.5%) 2 (3%) 2 (9.1%) 0.2544
Cerebrovascular disease 4 (4.5%) 3 (4.5%) 1 (4.5%) 1.0000
Smoking 24 (27%) 19 (28.4%) 5 (22.7%) 0.8107
Surgical data
Surgery type, no. (%) 0.1050
Isolated valve surgery 70 (78.7%) 56 (83.6%) 14 (63.6%)  
Isolated CABG 12 (13.5%) 6 (9%) 6 (27.3%)  
Combined valve and CABG surgery 2 (2.2%) 1 (1.5%) 1 (4.5%)  
Other procedures 5 (5.6%) 4 (6%) 1 (4.5%)  
ASA physical status class, no. (%) 0.5288
I 0 0 0  
II 3 (3.4%) 2 (3%) 1 (4.5%)  
III 36 (40.4%) 29 (43.3%) 7 (31.8%)  
IV 49 (55.1%) 35 (52.2%) 14 (63.6%)  
V 1 (1.1%) 1 (1.5%) 0  
CPB Duration(min) 145.6 ± 53.4 143.8 ± 54.9 151.1 ± 49.4 0.5800
Cross clamp time(min) 91.1 ± 32.6 90.6 ± 34.0 92.6 ± 28.6 0.8009
Reperfusion time(min) 35 (30–50) 35 (26–50) 41 (32–48.75) 0.1309
Total operative time(min) 293.1 ± 66.7 284.2 ± 64.1 320.4 ± 68.5 0.0262
Intraoperative medication, no. (%)
Vasopressors 41 (46.1%) 29 (43.3%) 12 (54.5%) 0.5010
Diuretics 10 (11.2%) 6 (9.0%) 4 (18.2%) 0.2553
Steroids 29 (32.6%) 22 (32.8%) 7 (31.8%) 1.0000
Nitroglycerin 18 (20.2%) 12 (17.9%) 6 (27.3%) 0.3678
Tranexamic Acid 6 (6.7%) 6 (9.0%) 0 (0%) 0.3297
Beta-blockers 17 (19.1%) 15 (22.4%) 2 (9.1%) 0.2214
Anticoagulants 54 (60.7%) 41 (61.2%) 13 (59.1%) 1.0000
Intraoperative fluid therapy
Intraoperative fluid (mL) 2309.4 ± 905.5 2234.7 ± 772.8 2536.9 ± 1219.5 0.2843
Crystalloid (mL) 1050 (850–1500) 1100 (850–1500) 1000 (850-1312.5) 0.7350
Colloid (mL) 0 (0–500) 0 (0–500) 375 (0–500) 0.2537
Blood product (mL) 500 (300–1000) 500 (300–975) 704 (300–975) 0.5068
Intraoperative fluid loss
Blood Loss (mL) 400 (200–510) 400 (200–500) 400 (300-577.5) 0.6281
Urine output (mL) 900 (600–1000) 900 (650–1000) 700 (325–1000) 0.2749
Intraoperative Indicators
Pre-CPB CVP (mmHg) 9 (7–13) 9 (7–13) 9 (7–12) 0.8819
Post-CPB CVP (mmHg) 11 (7–13.2) 9 (7–12.2) 12 (10.8-14) 0.0235
Lowest MAP (mmHg) 45 (40–50) 45 (40–50) 46.5 (40.8-50) 0.8923
Time with MAP <60 mmHg (min) 20 (10–35) 20 (10-33.8) 20 (10–45) 0.4795
PO₂ at end of CPB (mmHg) 241.4 ± 99.1 230.0 ± 94.5 276.2 ± 106.7 0.0572
Lowest temperature (°C) 32 (31–33) 33 (31–33) 32 (31–33) 0.0870

Notes: Normally distributed continuous variables are presented as mean ± standard deviation, non-normally distributed continuous variables as median (interquartile range), and categorical or ordinal variables as n (%).

Abbreviations: AKI, acute kidney injury; ASA, American Society of Anesthesiologists; BMI, body mass index; CABG, coronary artery bypass grafting; CPB, cardiopulmonary bypass; CVP, central venous pressure; CVD, cerebrovascular disease; MAP, mean arterial pressure; NYHA, New York Heart Association; PO₂, partial pressure of oxygen.

Among the 89 patients, a sub-cohort of 54 patients (20 AKI, 34 non-AKI) had complete specimens for novel biomarker quantification (Table 2, Supplementary Tables 2–4). Their baseline characteristics were comparable to the unmeasured cohort (Supplementary Tables 5 and 6). As expected, conventional renal injury markers (such as serum KIM-1, cystatin C, and uric acid) demonstrated significantly higher postoperative trajectories in the AKI group (Supplementary Figure 2).

Table 2.

Comparison of clinical indicators between AKI and non-AKI groups at different perioperative time points.

Variables Time Total non-AKI AKI p
Blood routine examination
WBC (×109/L) Pre-op 6.34 (5.25–7.83) 6.34 (4.99–7.99) 6.27 (5.64–7.59) 0.9280
  post 24h 11.83 ± 3.06 11.5 ± 3.03 12.5 ± 3.08 0.2238
  post 48h 10.45 ± 3.28 10.25 ± 3.36 11.08 ± 3.02 0.3172
RBC (×1012/L) Pre-op 4.4 ± 0.56 4.38 ± 0.56 4.45 ± 0.55 0.6360
  post 24h 3.13 (2.74–3.58) 3.26 (2.83–3.65) 3.07 (2.67–3.5) 0.2523
  post 48h 3.07 (2.8–3.49) 3.14 (2.88–3.65) 2.83 (2.53–3.17) 0.0072
Hb (g/L) Pre-op 129.43 ± 19.45 128.55 ± 19.74 132.09 ± 18.73 0.4621
  post 24h 94.17 ± 15.58 95.67 ± 14.86 91.19 ± 16.9 0.2861
  post 48h 92 (82–103) 94 (84-106.75) 82 (72–95) 0.0140
PLT (×109/L) Pre-op 204 (170–255) 208 (166-255.5) 203.5 (178-247.75) 0.9091
  post 24h 140.35 ± 49.88 144.64 ± 44.71 131.76 ± 59.15 0.3380
  post 48h 133 (107–182) 144.5 (112.5–199.5) 120 (88–151) 0.0464
Neu (%) Pre-op 64.83 ± 11.03 64.61 ± 11.19 65.52 ± 10.76 0.7389
  post 24h 86.7 (83.2–90.05) 86.9 (83.3–89.9) 85.6 (82.9–90.1) 0.6304
  post 48h 83.6 (79.9–86.7) 82.95 (79.4–86.12) 85 (81.7–89.2) 0.0967
Lym (%) Pre-op 24.84 ± 9.59 25.3 ± 9.95 23.42 ± 8.45 0.4274
  post 24h 5.7 (3.9–8.5) 5.8 (3.87–7.38) 5.7 (4-8.6) 0.6565
  post 48h 9 (6.1–11.8) 8.7 (6.3–12.2) 9.3 (5.9–10.5) 0.6761
Eos (%) Pre-op 1.8 (0.9–2.7) 1.5 (0.9–2.55) 2 (1.52–3.42) 0.0540
  post 24h 0 (0-0) 0 (0-0) 0 (0-0) 0.8004
  post 48h 0.2 (0-0.7) 0.2 (0-0.72) 0.1 (0-0.6) 0.1165
Blood electrolytes
Na (mmol/L) Pre-op 141.6 (140.25–143.12) 141.65 (140.1–143.02) 141.5 (140.33–143.4) 1
  post 24h 141.85 (139.8–144.93) 141.5 (138.25–144.9) 142.2 (140.7–144.9) 0.4104
  post 48h 138.54 ± 4.78 137.18 ± 4.23 142.41 ± 4.18 < 0.0001
K (mmol/L) Pre-op 4.08 ± 0.37 4.06 ± 0.38 4.15 ± 0.36 0.3557
  post 24h 4.24 ± 0.34 4.32 ± 0.34 4.08 ± 0.3 0.0309
  post 48h 4.25 (4.03–4.53) 4.24 (4.02–4.57) 4.26 (4.2–4.51) 0.9249
Ca (mmol/L) Pre-op 2.33 ± 0.12 2.31 ± 0.12 2.38 ± 0.13 0.0247
  post 24h 2.14 ± 0.17 2.15 ± 0.15 2.13 ± 0.21 0.7437
  post 48h 2.22 ± 0.17 2.23 ± 0.18 2.2 ± 0.15 0.5844
Liver function tests
Alb (g/L) Pre-op 39.07 ± 4.51 39 ± 4.51 39.3 ± 4.59 0.7964
  post 24h 36.7 (33.38–39.6) 36.75 (34.9–40.42) 36.4 (32.92–38.53) 0.2683
  post 48h 36.65 ± 3.88 36.38 ± 4.06 37.5 ± 3.21 0.2514
Glob (g/L) Pre-op 25.5 (22.35–28.2) 25.2 (22.35–28.05) 26.6 (22.4–30.1) 0.3908
  post 24h 19.55 (16.52–22.48) 20.05 (16.98–22.63) 17.4 (14.43–20.77) 0.0808
  post 48h 22.3 (19.4–24.4) 22.65 (19.98–24.6) 21 (17.5-23) 0.0659
ALT (U/L) Pre-op 21 (15-31.5) 20.5 (15-28.75) 25 (17–35) 0.1969
  post 24h 28 (19–41) 25 (18.75–40.75) 30 (21.25-41) 0.3892
  post 48h 25 (17–38) 22 (17-36.25) 31 (17–41) 0.3431
AST (U/L) Pre-op 23 (17-29.5) 23.5 (17-29.75) 23 (18–27) 0.5251
  post 24h 67.5 (46–107) 69 (45.75–98.25) 66 (47.75–120.5) 0.8793
  post 48h 34 (24–51) 34.5 (23.75–50.25) 31 (25–51) 0.7949
TBIL (μmol/L) Pre-op 11.8 (8.45–17.35) 13.55 (8.8–18.55) 10.2 (7.5–10.7) 0.0190
  post 24h 19.2 (12.75–25.15) 19.55 (12.67–23.7) 19 (13.98–30.03) 0.6996
  post 48h 16.3 (12.1–22.9) 17.4 (12.6–23.32) 14.4 (12-20.3) 0.3279
DBIL (μmol/L) Pre-op 3.9 (2.6–6.3) 4.3 (2.6–6.8) 3.1 (2.1–4.2) 0.0240
  post 24h 7.95 (5.32–11.55) 7.3 (5.38–11.03) 11 (5.22–18.98) 0.1659
  post 48h 7 (4.4–10.1) 7.45 (4.4–10.2) 6.1 (4.8–7.9) 0.7637
Glomerular Function Indicators
BUN (mmol/L) Pre-op 5.05 (4.32–6.57) 5.32 (4.36–6.68) 4.74 (4.3–5.84) 0.4439
  post 24h 8.88 ± 2.73 8.18 ± 2.38 11.02 ± 2.64 <0.0001
  post 48h 6.93 (5.19–9.18) 6.22 (4.99–8.11) 8.44 (7.42–13.82) <0.0001
Scr (μmol/L) Pre-op 67 (56–79) 66 (55 − 78.5) 71.5 (67 − 82.5) 0.2269
  post 24h 74 (58–98) 67 (50 − 84.5) 108.5 (95 − 137.25) <0.0001
  post 48h 68 (51–82) 63 (47–74) 88 (80.25 - 132) <0.0001
UA (μmol/L) Pre-op 327.34 ± 119.95 324.78 ± 123.99 335.14 ± 109.08 0.7275
  post 24h 267.08 ± 106.6 238.28 ± 96.94 350.86 ± 89.02 <0.0001
  post 48h 200 (139.75 - 286) 174.5 (121.75–253.5) 289.5 (213.25–326.5) 0.0002
eGFR (mL/min/1.73 m²) Pre-op 101.19 (88.5–107.95) 103.63 (87.94–108.7) 96.5 (89.39–103.88) 0.0999
  post 24h 88.37 ± 26.69 98.09 ± 19.28 59.68 ± 25.11 <0.0001
  post 48h 101.16 (86.73–109.81) 105.76 (98.01–114.68) 68.91 (45.67–96.31) <0.0001
Tubular Function and Injury Markers
uKIM1 (ng/mL) Pre-op 0.44 (0.25–0.75) 0.43 (0.27–0.70) 0.45 (0.25–0.82) 0.8661
  post 24h 1.12 ± 0.66 1.09 ± 0.64 1.18 ± 0.71 0.6008
  post 48h 1.12 (0.64–1.87) 1.08 (0.62–1.64) 1.16 (0.69–1.94) 0.6003
sKIM1 (ng/mL) Pre-op 0.16 (0.14–0.18) 0.15 (0.13–0.16) 0.18 (0.16–0.20) 0.0104
  post 24h 0.26 (0.18–0.39) 0.21 (0.17–0.32) 0.38 (0.20–0.58) 0.0236
  post 48h 0.24 (0.16–0.33) 0.20 (0.15–0.26) 0.27 (0.19–0.67) 0.0079
sRBP (mg/L) Pre-op 39.06 (28.12–48.34) 36.45 (24.6–46.38) 46.89 (39.22–51.08) 0.0161
  post 24h 24.51 (19.69–31.28) 23.38 (18.32–28.37) 30.7 (26.41–38.36) 0.0010
  post 48h 25.42 ± 8.58 23.47 ± 7.29 30.91 ± 9.68 0.0004
sNGAL (ng/mL) Pre-op 176.9 (118.4–278.78) 211.55 (138.18 - 299) 163.25 (112.32–198.82) 0.0828
  post 24h 293 (224.3–430.3) 293.7 (230.87–462.72) 291.2 (207.3–351.5) 0.2646
  post 48h 292.2 (220.7–454.75) 292.6 (217.8–428.02) 286.2 (234.1–462.3) 0.5828
uNGAL (ng/mL) Pre-op 65.5 (15.5–157.8) 61.2 (15.62–161.18) 69.2 (15.5–156.1) 0.9909
  post 24h 94.95 (33.6–233.67) 109.25 (28.85–237.85) 80.8 (43.77–219.9) 0.8798
  post 48h 60.3 (29.7–116.55) 52.6 (28.45 - 123) 73.05 (40.2–97.72) 0.2863
CyC (mg/L) Pre-op 1.01 (0.83–1.2) 0.98 (0.8–1.18) 1.04 (0.97–1.31) 0.0635
  post 24h 0.92 (0.75–1.15) 0.88 (0.68–1.04) 1.12 (1.02–1.5) 0.0001
  post 48h 1 (0.78–1.26) 0.95 (0.75–1.17) 1.21 (1.07–1.54) 0.0001
uRBP (mg/L) Pre-op 0.21 (0.11–0.44) 0.2 (0.11–0.44) 0.21 (0.16–0.41) 0.9960
  post 24h 13.15 (11.34–14.01) 12.94 (11.36–13.87) 13.41 (10.53–16.06) 0.4278
  post 48h 11.83 (2.47–13.98) 11.57 (2.46–14.01) 12.2 (3.89–13.55) 0.7772
α1-MG (ug/mL) Pre-op 8.65 (4.78–19.5) 9.52 (4.45–21.4) 8 (6.4–17.4) 0.6876
  post 24h 35.3 (30.25–47.85) 34.9 (29.65–46.27) 41.2 (32.35–49.42) 0.4526
  post 48h 41.2 (22.55–49.35) 40.9 (22.55–49.05) 41.8 (22.7–50.05) 0.8678
β2-MG (mg/L) Pre-op 0.26 (0.12–0.54) 0.26 (0.12–0.53) 0.33 (0.17–0.62) 0.6696
  post 24h 3.86 (1.24–7.64) 4.08 (1.38–8.98) 3.75 (1 − 5.54) 0.4780
  post 48h 4.4 (1.69–6.7) 4.53 (1.63–6.69) 3.99 (1.77–6.61) 0.8591
TRF (mg/L) Pre-op 1.63 (0.99–2.7) 1.69 (1.07–2.88) 1.16 (0.86–1.87) 0.0710
  post 24h 4.11 (1.56 − 7) 4.46 (2.39–6.85) 2.78 (1.46–7.03) 0.1863
  post 48h 3.3 (2.2–5.46) 3.87 (2.5–5.72) 2.33 (1.43–3.01) 0.0210
NAG (U/L) Pre-op 9.2 (5.2–14.15) 9.9 (5.15–14.65) 7.8 (5.6–13.6) 0.7434
  post 24h 23 (12.05–32.75) 23.2 (12.73–33.33) 14.1 (10.4–31.35) 0.4401
  post 48h 22 (14.95–37.95) 21.95 (15.17–48.58) 24.9 (13.6–29.6) 0.4707

Notes: Normally distributed continuous variables are presented as mean ± standard deviation, and non-normally distributed continuous variables as median (interquartile range). Because of missing observations, the effective sample size varies across variables and time points. Serum and urinary KIM-1 measurements were available for 54 patients, including 34 non-AKI patients and 20 patients with CSA-AKI.

Abbreviations: AKI, acute kidney injury; Pre-op: preoperative; post 24h: 24 h postoperatively; post 48h: 48 h postoperatively. WBC: white blood cell count; RBC: red blood cell count; Hb: hemoglobin; PLT: platelet count; Neu, neutrophil percentage; Lym, lymphocyte percentage; Eos, eosinophil percentage; Na: sodium ion; K: potassium ion; Ca: calcium ion; Alb: albumin; Glob: globulin; ALT: alanine aminotransferase; AST: aspartate aminotransferase; TBIL: total bilirubin; DBIL: direct bilirubin; BUN: blood urea nitrogen; Scr: serum creatinine; UA: uric acid; eGFR: estimated glomerular filtration rate (calculated using the 2021 CKD-EPI equation); uKIM1: urinary kidney injury molecule-1; sKIM1: serum kidney injury molecule-1; sRBP: serum retinol-binding protein; sNGAL: serum neutrophil gelatinase-associated lipocalin; uNGAL: urinary neutrophil gelatinase-associated lipocalin; CyC: cystatin C; uRBP: urinary retinol-binding protein; α1-MG: alpha-1 microglobulin; β2-MG: beta-2 microglobulin; TRF: transferrin; NAG: N-acetyl-β-D-glucosaminidase.

Perioperative dynamic trajectories of selenoproteins

Longitudinal changes in serum and urinary selenoproteins were evaluated using linear mixed-effects models (Tables 3 and 4, Figure 2).

Table 3.

Comparison of selenoprotein levels between AKI and non-AKI groups at different perioperative time points.

Variables Time Total (n = 54) non-AKI (n = 34) AKI (n = 20) p
sGPX1 pre-op 2.18 (0.80–3.31) 1.42 (0.62–3.00) 2.80 (2.21–3.65) 0.0092
  post 24h 1.97 (1.12–2.8) 1.31 (0.78–2.56) 2.26 (1.69–3.12) 0.0458
  post 48h 1.90 (1.00–3.13) 1.40 (0.78–2.99) 2.36 (1.72–3.21) 0.0940
uGPX1 pre-op 4.02 ± 1.83 4.14 ± 1.59 3.81 ± 2.21 0.5365
  post 24h 3.35 (2.54–4.86) 3.33 (2.56–4.53) 3.40 (1.99–6.02) 0.9073
  post 48h 3.76 (2.95–5.35) 3.62 (2.89–5.31) 4.09 (3.05–5.96) 0.6738
sGPX3 pre-op 24.01 (13.36–31.94) 18.60 (12.86–34.28) 25.37 (21.13–30.05) 0.5366
  post 24h 30.70 (26.47–45.20) 36.18 (27.80–48.53) 27.76 (25.58–31.44) 0.0403
  post 48h 28.41 (22.91–42.63) 30.46 (21.06–48.50) 27.53 (23.44–31.39) 0.4792
uGPX3 pre-op 15.53 (8.29–32.27) 17.67 (10.32–40.14) 11.75 (6.63–21.63) 0.0975
  post 24h 63.61 ± 30.78 61.50 ± 33.11 67.19 ± 26.77 0.5175
  post 48h 61.02 ± 29.89 58.95 ± 29.46 64.54 ± 31.04 0.5126
sSEPP1 pre-op 4455.42 (3013.76–6997.4) 5798.97 (3790.48–8378.24) 2823.18 (2053.36–4211.56) 0.0005
  post 24h 3623.32 (1637.42–5131.67) 4181.83 (3322.21–5455.18) 1601.68 (1191.06–2480.40) 0.0005
  post 48h 3597.52 (2366.07–4710.38) 3971.78 (3260.81–4942.57) 2325.73 (1470.34–3118.66) 0.0013
uSEPP1 pre-op 1.79 (0.53–4.41) 1.94 (0.84–4.84) 0.44 (0.28–2.51) 0.0152
  post 24h 7.40 (1.92–11.91) 9.70 (4.73–13.05) 2.30 (1.17–7.59) 0.0082
  post 48h 8.16 (2.06–11.86) 10.59 (5.85–13.23) 1.64 (0.61–8.58) 0.0004

Notes: Normally distributed continuous variables are presented as mean ± standard deviation, and non-normally distributed continuous variables as median (interquartile range).

Abbreviations: AKI, acute kidney injury; GPX1, glutathione peroxidase 1; GPX3, glutathione peroxidase 3; SEPP1, selenoprotein P; s, serum; u, urinary; pre-op, preoperative; post 24 h, postoperative 24 h; post 48 h, postoperative 48 h. Units: ng/mL.

Table 4.

Comparison of postoperative changes in selenoprotein levels between AKI and non-AKI groups.

Variables Time Total (n = 54) non-AKI (n = 34) AKI (n = 20) p
sGPX1 Δ24h–Pre 0.13 (−0.54–0.58) 0.17 (−0.12–0.57) −0.54 (−1.16–0.78) 0.1054
  Δ48h–Pre −0.01 (−0.49–0.2) 0.08 (−0.25–0.30) −0.44 (−1.51–0.13) 0.0286
uGPX1 Δ24h–Pre −0.38 (−1.07–0.54) −0.41 (−1.25–0.41) −0.03 (−0.58–0.82) 0.1837
  Δ48h–Pre −0.14 (−0.86–1.02) −0.31 (−0.86–0.59) 0.24 (−1.00–2.66) 0.3145
sGPX3 Δ24h–Pre 8.59 (−0.23–18.45) 12.97 (5.79–19.84) 1.42 (−1.83–12.06) 0.0492
  Δ48h–Pre 7.71 (−2.79–15.6) 9.10 (−0.06–15.63) 2.01 (−3.51–11.71) 0.2289
uGPX3 Δ24h–Pre 42.79 (16.65–59.13) 33.26 (10.15–55.33) 46.24 (33.10–64.73) 0.1135
  Δ48h–Pre 44.3 (9.94–64.9) 30.55 (1.26–52.41) 56.68 (28.88–72.33) 0.0807
sSEPP1 Δ24h–Pre −1278.75 (−2185.18 to −560.12) −1278.75 (−2185.18 to −636.69) −1264.8 (−1933.48 to −429.84) 0.5880
  Δ48h–Pre −831.55 (−2026.26 to 37.57) −997.13 (−3162.02 to −112.65) −557.15 (−1241.39 to −1.16) 0.1405
uSEPP1 Δ24h–Pre 4.26 (0.67–8.56) 5.24 (1.65–9.52) 1.06 (0.49–5.83) 0.0807
  Δ48h–Pre 4.72 (0.96–8.45) 6.27 (2.32–10.36) 0.82 (0.16–4.69) 0.0005

Notes: Changes were calculated as the postoperative value minus the corresponding preoperative value. Normally distributed continuous variables are presented as mean ± standard deviation, and non-normally distributed continuous variables as median (interquartile range).

Abbreviations: AKI, acute kidney injury; GPX1, glutathione peroxidase 1; GPX3, glutathione peroxidase 3; SEPP1, selenoprotein P; s, serum; u, urinary; Δ24h–Pre: change from preoperative to 24 h postoperatively; Δ48h–Pre: change from preoperative to 48 h postoperatively. Units: ng/mL.

Figure 2.

Twelve-panel figure displaying predicted trends for serum and urinary GPX1, GPX3, and SEPP1 across time points (pre, post24h, post48h) in AKI and non-AKI groups, along with effect estimate plots. The figure features twelve panels arranged in three rows and four columns, illustrating predicted trends and effect estimates for serum and urinary biomarkers (GPX1, GPX3, SEPP1) related to AKI status. The first two columns of each row show line graphs with time on the x-axis and biomarker levels on the y-axis for both AKI and non-AKI groups, accompanied by confidence intervals. The last two columns display forest plots indicating the interaction effects of AKI and time on each marker, showing confidence intervals and effect estimates that vary for post-24h and post-48h across the biomarkers.

Dynamic Trends of Selenoproteins in Serum and Urine During the Perioperative Period. Predicted perioperative trajectories of serum and urinary GPX1, GPX3, and SEPP1 in patients with and without AKI are shown based on linear mixed-effects models. Individual dots represent observed measurements at each time point (preoperative, postoperative 24 h, and postoperative 48 h), while solid lines indicate model-predicted marginal means with shaded areas representing 95% confidence intervals. Forest plots on the right display the estimated effects of the AKI × TIME interaction on each biomarker, with horizontal lines representing 95% confidence intervals. Units: ng/mL.

SEPP1 showed distinct perioperative changes in serum and urine. Serum SEPP1 levels decreased significantly at 24 h and 48 h postoperatively compared with preoperative levels (both p < 0.001); however, preoperative serum SEPP1 levels were significantly lower in patients who subsequently developed AKI than in those who did not (p < 0.001). In addition, urinary SEPP1 increased significantly at 24 h and 48 h postoperatively in the non-AKI group (both p < 0.001). Urinary SEPP1 also increased in the AKI group, but the magnitude of increase was attenuated. Net change analysis further showed that the change in urinary SEPP1 from baseline (ΔSEPP1) was greater in the non-AKI group at both time points, with the between-group difference reaching statistical significance at 48 h (p < 0.001; Table 4). These findings suggest that renal handling or tubular secretion of SEPP1 may be impaired under severe stress.

GPX3 also showed distinct perioperative changes in serum and urine. Serum GPX3 increased significantly postoperatively (p < 0.010). Notably, the magnitude of this early systemic response was smaller in the AKI group, and net change analysis further showed that the increase in serum ΔGPX3 at 24 h was significantly greater in the non-AKI group (p < 0.050; Table 4). Urinary GPX3 increased postoperatively in both groups.

GPX1 exhibited a different perioperative pattern. Urinary GPX1 remained relatively stable, with no significant between-group differences, whereas serum GPX1 showed a gradual decline. Compared with the non-AKI group, the AKI group showed a greater decline in serum GPX1 from baseline to 48 h postoperatively, although the overall difference did not reach statistical significance (p = 0.082). Net change analysis showed that serum ΔGPX1 at 48 h differed significantly between the AKI and non-AKI groups, with a greater decline in the AKI group (p < 0.050).

Correlation between selenoproteins and renal vulnerability

To ascertain the biological relevance of these novel biomarkers, we mapped their correlations with surgical stress indices and established renal injury markers (Figures 3–6).

Figure 3.

Nine scatterplots displaying correlations between biomarkers and surgical time variables, complete with regression lines and statistics. This figure features eight scatterplots arranged in a 2x4 grid illustrating correlations between various biomarkers and time variables related to surgery and reperfusion. The plots include: "Urinary KIM1 Post24h vs Total Op Time" (positive correlation, r = 0.282, p = 0.039), "Serum GPX1 Post24h vs Total Op Time" (positive correlation, r = 0.335, p = 0.013), "Serum GPX1 Post48h vs Total Op Time" (positive correlation, r = 0.334, p = 0.014), "Urinary SEPP1 Post48h vs Total Op Time" (negative correlation, r = -0.271, p = 0.048), "Serum GPX3 Post48h vs Reperfusion Time" (negative correlation, r = -0.304, p = 0.025), "Urinary GPX3 Post48h vs Reperfusion Time" (positive correlation, r = 0.276, p = 0.043), "Serum ΔKIM1 at 48 h vs Total Op Time" (positive correlation, r = 0.270, p = 0.048), and "Serum ΔSEPP1 at 24 h vs Cross-Clamp Time" (negative correlation, r = -0.305, p = 0.025). Each plot displays data points, trend lines, and correlation coefficients with p-values.

Correlation Between Surgical Parameters and Renal Injury Markers. Scatter plots illustrating the associations between key surgical parameters and renal injury markers or selenoproteins. Spearman’s rank correlation coefficients (r) and corresponding p values are shown in each panel. Red lines indicate trend lines for visualization, with shaded areas representing 95% confidence intervals.

Figure 4.

Nine scatterplots showing correlations of GPX1 with various serum and urinary markers, including CysC, Bun, Scr, Ua, and eGFR. The figure features nine scatterplots organized in a 3x3 grid, illustrating how GPX1 levels correlate with clinical markers. The top row plots include Serum GPX1 Pre24h vs CysC (r = 0.438, p = 0.001), Post24h vs Bun (r = 0.264, p = 0.056), and Post24h vs Scr (r = 0.356, p = 0.009). The middle row displays Post24h vs CysC (r = 0.392, p = 0.004), Post24h vs Ua (r = 0.313, p = 0.024), and Post24h vs eGFR (r = -0.377, p = 0.005). The bottom row shows Post48h vs CysC (r = 0.406, p = 0.003), Post48h vs eGFR (r = -0.265, p = 0.053), and Urinary ΔGPX1 at 24 h vs ΔCysC at 24 h (r = 0.343, p = 0.015). Each plot includes fitted regression lines and shaded confidence intervals.

Correlation Between SEPP1 and Renal Function. Scatter plots illustrating the associations between SEPP1 and Renal Function. Spearman’s rank correlation coefficients (r) and corresponding p values are shown in each panel. Red lines indicate trend lines for visualization, with shaded areas representing 95% confidence intervals.

Figure 5.

Grid of scatterplots showing correlations between GPX3 levels and various biomarkers over different time points. The image shows a grid of scatterplots analyzing relationships between various biomarkers, with blue data points and red trend lines for each plot. The upper row features correlations such as Serum GPX3 Pre24h vs sNGAL (r=-0.268, p=0.053) and Serum GPX3 Pre24h vs Serum KIM1 Pre24h (r=0.567, p<0.001). The middle row includes Urinary GPX3 Post24h vs Urinary KIM1 Post24h (r=0.376, p=0.005) and Urinary GPX3 Post48h vs Bun (r=0.323, p=0.017). The lower row compares Serum ΔGPX3 at 24 h vs ΔeGFR at 24 h (r=0.325, p=0.018) and Urinary ΔGPX3 at 48 h vs ΔBun at 48 h (r=0.292, p=0.032), highlighting significant correlations and trend lines across different time points.

Correlation Between GPX3 and Renal Function. Scatter plots illustrating the associations between GPX3 and Renal Function. Spearman’s rank correlation coefficients (r) and corresponding p values are shown in each panel. Red lines indicate trend lines for visualization, with shaded areas representing 95% confidence intervals.

Figure 6.

Sixteen scatterplots showing correlations between serum and urinary SEPP1 levels and various clinical markers across different time points. The figure displays 16 scatterplots in a grid format, illustrating correlations between serum and urinary SEPP1 levels and clinical markers across different time points (Pre24h, Post24h, Post48h). Each plot shows data points, regression lines, and confidence intervals, detailing relationships such as negative correlations for serum SEPP1 Post24h vs. Scr (r = -0.339, p = 0.012) and positive correlations for serum SEPP1 Post48h vs. eGFR (r = 0.356, p = 0.008). Other correlations are presented, including metrics such as CysC and UA, with coefficients and significance levels indicated for clarity.

Correlation Between GPX1 and Renal Function. Scatter plots illustrating the associations between GPX1 and Renal Function. Spearman’s rank correlation coefficients (r) and corresponding p values are shown in each panel. Red lines indicate trend lines for visualization, with shaded areas representing 95% confidence intervals.

Extended surgical duration and reperfusion times correlated with both an exacerbation of traditional injury markers (KIM-1) and a depletion of protective selenoproteins (e.g. negative correlation with serum/urinary SEPP1 and serum GPX3) (Figure 3).

Crucially, SEPP1 consistently demonstrated a “renoprotective” correlation profile across the perioperative period: higher serum and urinary SEPP1 levels were robustly correlated with higher eGFR and lower levels of detrimental markers including Scr, Cystatin C, and uric acid (Figure 4). Concurrently, alterations in GPX1 and GPX3 closely mirrored tubular injury severity, exhibiting positive correlations with serum and urinary KIM-1 and blood urea nitrogen (BUN) dynamics (Figures 5 and 6).

Univariable associations with CSA-AKI

Univariable logistic regression confirmed that diminished selenoprotein reserves and impaired dynamic responses were significantly associated with an increased risk of CSA-AKI (Supplementary Table 7). Specifically, lower absolute serum SEPP1 levels preoperatively and postoperatively, an attenuated increase in urinary SEPP1 (baseline to 48 h, p = 0.023), and a more severe depletion of serum GPX1 (baseline to 48 h, p = 0.030) were all strong individual risk factors (Supplementary Table 7). Additional clinical risk factors included older age, hypertension, longer operation time, and specific routine laboratory deviations (Supplementary Tables 8 and 9).

Exploratory predictive value of preoperative SEPP1

As SEPP1 was a biomarker of primary interest in this study, showed a significant perioperative decline, and was lower preoperatively in patients who subsequently developed CSA-AKI, we hypothesized that baseline serum SEPP1 might help identify renal vulnerability before surgical insult. To reduce the risk of model overfitting given the limited number of outcome events, we constructed a parsimonious logistic regression model incorporating only older age, an established clinical risk factor for CSA-AKI [1], and preoperative serum SEPP1.

Compared with the age-only reference model, the addition of preoperative serum SEPP1 improved model performance (Figure 7). The apparent AUC increased from 0.710 (95% CI, 0.569–0.852) to 0.818 (95% CI, 0.689–0.946). Although the paired DeLong test for the AUC increment (ΔAUC = 0.107) did not reach statistical significance (p = 0.142), the likelihood ratio test showed a significant improvement in overall model fit after the inclusion of SEPP1 (χ2 = 8.792, p = 0.003), and the apparent Brier score decreased from 0.205 to 0.164. Following internal validation using 1,000 bootstrap resamples, the combined model showed good stability, with an optimism-corrected AUC of 0.802, a calibration intercept of −0.017, and a calibration slope of 0.941. These findings suggest that preoperative serum SEPP1 may provide incremental predictive value for CSA-AKI and may have potential for early risk prediction, although further validation in larger cohorts is required.

Figure 7.

ROC curve comparing sensitivity vs. 1-specificity for models: Age (gray) and Age + SEPP1 (blue), showing AUC values. The image displays a ROC curve illustrating sensitivity against 1-specificity, with both axes ranging from 0 to 1. The gray line represents the Age model with an AUC of 0.710, while the blue line for Age + SEPP1 has a higher AUC of 0.818. Both lines are stepped, indicating changes in sensitivity at various specificity levels, with the blue line remaining above the gray line. A diagonal line indicates no-discrimination performance, aiding in visual comparison. Each AUC value includes 95% confidence intervals for clarity.

Exploratory ROC Analysis of Preoperative Serum SEPP1 for CSA-AKI. ROC curves comparing the age-only model and the model incorporating age and preoperative serum SEPP1. AUC, area under the receiver operating characteristic curve; CI, confidence interval.

Discussion

Our study showed that preoperative levels and perioperative changes in key selenoproteins, including SEPP1, GPX1, and GPX3, were associated with CSA-AKI and renal function. In particular, preoperative serum SEPP1 showed potential value for preoperative risk assessment, whereas postoperative changes in these proteins provided insights into the perioperative redox alterations accompanying CSA-AKI. CSA-AKI is a severe complication involving interconnected mechanisms, including altered renal perfusion, inflammation, and oxidative stress, with oxidative stress serving as an important mediator [1,26]. Although conventional renal function markers, such as serum creatinine and blood urea nitrogen, and emerging biomarkers, including NGAL, KIM-1, and cystatin C, contribute to the detection of kidney injury [1,27,28], our findings suggest that selenoproteins may provide complementary information regarding perioperative redox homeostasis in CSA-AKI.

Consistent with previous reports, our univariable analysis confirmed that advanced age and hypertension were associated with an increased risk of AKI [1,29]. Lower preoperative serum SEPP1 was also associated with higher odds of CSA-AKI, extending previous evidence linking perioperative selenium depletion to adverse outcomes after cardiac surgery [30]. Beyond selenium-related markers, previous studies have reported that plasma IL-8, plasma TNF-α, and urinary NGAL normalized to creatinine excretion (NGAL/Cr) are significantly associated with AKI and serve as potential predictors of CSA-AKI. In our study, conventional tubular injury markers, including NGAL and KIM-1, increased postoperatively (Supplementary Figure 2); however, these changes may also reflect the generalized stress response to surgery [31,32]. These findings support further evaluation of selenoproteins as complementary markers of perioperative redox responses and CSA-AKI susceptibility. In addition, higher urinary SEPP1 levels at 48 h postoperatively and a less pronounced decline in serum GPX1 were associated with lower odds of CSA-AKI. Overall, selenoproteins may provide complementary information on perioperative oxidative stress, but their clinical relevance requires confirmation in larger prospective studies with appropriate multivariable adjustment.

Oxidative stress is a cornerstone of CSA-AKI pathogenesis [1,33]. Selenoproteins are crucial components of the endogenous antioxidant defense. SEPP1 is the primary selenium transport protein, supplying substrate for the synthesis of glutathione peroxidases like GPX1 (intracellular) and GPX3 (kidney-derived, secreted) [34,35]. These enzymes detoxify peroxides, maintaining redox balance. Impaired selenoprotein function is linked to increased AKI susceptibility, as evidenced by experimental models where deficiency exacerbates injury [10,15]. A key finding involves the dynamics of GPX1. While raw data showed higher levels in AKI patients at some time points, mixed-effects modeling—adjusting for confounders like age and operative time—revealed a relative GPX1 deficiency in the AKI group. The more pronounced postoperative decline in serum GPX1 in patients with AKI may reflect reduced antioxidant capacity or increased GPX1 consumption under oxidative stress. This direction is consistent with preclinical evidence supporting a renoprotective role of GPX1. In hypoxia/reoxygenation-exposed renal tubular epithelial cells and IRI-AKI mice, selenium nanoparticle treatment increased GPX1 expression, suppressed NLRP3 inflammasome and Caspase-1 activation, and attenuated renal injury and subsequent fibrosis [10]. These findings support the GPX1/NLRP3/Caspase-1 axis as a potential protective pathway in AKI, although the effects of selenium-based interventions cannot be attributed exclusively to GPX1.

GPX3 exhibited a characteristic ‘rise-and-fall’ postoperative pattern. Although absolute levels did not differ significantly between groups, the postoperative increase in serum GPX3 was attenuated in patients with AKI, which may reflect a less pronounced adaptive antioxidant response. In contrast, urinary GPX3 showed a greater postoperative rise in the AKI group, possibly reflecting altered renal handling, injury-related release, or an adaptive response.

Preclinical evidence supports a renoprotective role for GPX3. Our previous study showed that renal GPX3 expression was markedly reduced in patients with AKI and in experimental AKI models [15]. Consistent with this observation, our previous work and other preclinical studies have supported a renoprotective role for GPX3. In renal IRI, targeted restoration of GPX3 attenuated kidney injury by reducing ROS-mediated mitochondrial apoptosis, whereas tubular-specific Gpx3 deletion aggravated renal dysfunction, tubular injury, oxidative stress, apoptosis, and mitochondrial disturbance [15,21,36]. Notably, in our preliminary bilateral renal IRI experiments, 30 min of ischemia followed by 48 h of reperfusion resulted in increased serum GPX3 levels despite reduced GPX3 expression in renal tissue. Given that GPX3 is an extracellular selenoprotein produced predominantly by renal proximal tubular cells and secreted mainly into the circulation, this discordant pattern may reflect compartmental redistribution of GPX3 after kidney injury or increased release from the kidney into the circulation. However, whether the circulating increase reflects altered secretion, injury-related release, or changes in GPX3 clearance requires further investigation.

Our observations contrast with those of Zou et al. who reported a decline in serum GPX3 at 6 h postoperatively in AKI patients, with a lower preoperative-to-postoperative ratio predicting higher AKI risk [37]. Differences in sampling time may partly explain these findings. Their 6-h measurement may have captured an earlier phase of the perioperative response, whereas our measurements at 24 and 48 h reflected later changes. This interpretation is supported by data from an IRI mouse model (GSE98622), which showed an initial decrease in renal GPX3 mRNA at 2–4 h post-ischemia, followed by recovery at 24 h (Supplementary Figure 3) [38]. Although renal GPX3 mRNA expression cannot be directly equated with circulating or urinary GPX3 protein levels, this temporal pattern suggests phase-dependent regulation of GPX3 following ischemic injury, with early decreases potentially accompanying acute oxidative injury and later recovery or increases reflecting restoration or compensatory adaptation. Accordingly, longitudinal assessment may better capture CSA-AKI pathophysiology than single time-point measurements. Further studies are required to define its predictive and therapeutic value.

Our findings on SEPP1 present an intriguing contrast to some prior reports. While other studies noted an early postoperative surge [39], we observed a consistent decline in serum SEPP1 levels in AKI patients from 24 to 48 h. This discrepancy may relate to sampling time, as database validation (GSE98622) shows SEPP1 mRNA peaks early (2–4 h) after ischemia before declining (Supplementary Figure 3) [38]. Our later measurements (24-48 h) likely capture a phase of exhausted compensation. The attenuated rise in urinary SEPP1 in patients with AKI may reflect altered renal handling of SEPP1 or a limited adaptive response to perioperative stress, although the underlying mechanism requires further investigation. Importantly, urinary SEPP1 measured at 48 h was not intended as an early diagnostic marker or a substitute for serum creatinine. Rather, its distinct perioperative trajectory was examined to characterize the biological changes associated with CSA-AKI.

These dynamics imply distinct stress adaptation patterns: non-AKI patients effectively redistribute selenium (decreased circulating SEPP1) to upregulate protective enzymes like GPX1 and GPX3, whereas AKI patients exhibit a maladaptive pattern of low baseline SEPP1 and a limited enzymatic response. Although selenium redistribution was not directly measured, these findings suggest that lower baseline SEPP1 and an inadequate adaptive selenoprotein response may identify patients who are more vulnerable to postoperative kidney injury.

Clinical evidence from other cardiorenal settings also supports a potential association between low SEPP1 and renal vulnerability. A recent cohort study in patients with acute heart failure found that lower circulating SEPP1 levels were associated with early renal dysfunction and an increased subsequent risk of AKI-related hospitalization [40]. Although this study was not conducted in a cardiac surgery population, the direction of this association is consistent with our finding of lower preoperative serum SEPP1 levels in patients who subsequently developed CSA-AKI. Taken together, these observations suggest that low circulating SEPP1 may mark increased susceptibility to renal stress, although its mechanistic role remains to be established.

Previous in vivo and in vitro studies have shown that sodium selenite can increase SEPP1 expression and improve tubular cell survival under hypoxic stress, while selenium-related interventions have attenuated renal dysfunction and tubular injury in experimental AKI models [10,20,41–44]. However, these experimental findings have not translated into consistent renal or overall clinical benefits in randomized trials of perioperative high-dose sodium selenite in cardiac surgery [45,46]. This discrepancy may reflect the broader effects of sodium selenite on multiple selenoproteins and redox pathways, as well as the lack of stratification by baseline selenium or SEPP1 status in previous trials. Thus, although low preoperative SEPP1 and an attenuated postoperative selenoprotein response may characterize renal vulnerability, our findings do not support selenium supplementation as a preventive strategy for CSA-AKI. Further mechanistic and biomarker-stratified prospective studies are warranted.

In this exploratory analysis, adding preoperative serum SEPP1 to the age-only model resulted in greater apparent discrimination, which was largely retained after bootstrap correction for optimism. This finding suggests that preoperative serum SEPP1 may provide information on individual susceptibility beyond that captured by age alone and supports further evaluation of its potential contribution to preoperative risk assessment for CSA-AKI. Unlike approaches relying on postoperative inflammatory markers [47] or combinations of functional and tubular injury biomarkers [48], our analysis focused on SEPP1 measured before surgery, highlighting its potential value for identifying high-risk patients before the clinical recognition of CSA-AKI. Nevertheless, given the limited sample size and number of CSA-AKI events, this model was intended only as an initial assessment of the potential contribution of SEPP1 to preoperative risk evaluation and should not be regarded as an established clinical prediction tool.

Several limitations should be acknowledged. First, this was a single-center study with a relatively limited sample size and number of CSA-AKI events, which constrained the complexity of the multivariable analyses and prediction modeling. Second, biomarkers were assessed only preoperatively and at 24 and 48 h postoperatively, and therefore may not have fully captured the early postoperative dynamics of selenoproteins. Third, given the observational design, the associations observed between selenoprotein alterations and CSA-AKI require further validation. Finally, although the exploratory prediction model was internally validated using bootstrap resampling, external validation was not performed. Future studies in larger, multicenter prospective cohorts, with more intensive perioperative sampling, are warranted to validate these findings and more comprehensively characterize the temporal dynamics of selenoproteins.

In conclusion, perioperative alterations in SEPP1, GPX1, and GPX3 were associated with CSA-AKI occurrence and perioperative renal function, suggesting distinct redox response patterns in patients with and without CSA-AKI. Preoperative serum SEPP1 may represent a candidate biomarker for early risk stratification of CSA-AKI. These findings further identify the selenoprotein pathway as a biologically relevant pathway warranting mechanistic and clinical investigation. Larger multicenter prospective studies incorporating serial sampling from the preoperative period through the early postoperative period are needed to validate these associations, characterize their temporal patterns, and clarify their clinical significance.

Supplementary Material

Supplemental Material
IRNF_A_2755138_SM7730.docx (103.9KB, docx)
Supplementary File 2_STROBEchecklist.docx
Supplementary Figure2.jpg
Supplementary Figure3.tif
IRNF_A_2755138_SM0069.tif (105.3MB, tif)
Supplementary Figure1.jpg

Acknowledgments

The authors thank the Central Laboratory of the Third Xiangya Hospital, Central South University, for providing experimental equipment and technical support. We also thank the Department of Cardiothoracic Surgery of the Third Xiangya Hospital, Central South University, for its assistance with patient recruitment and clinical sample collection.

Funding Statement

This work was funded by the ‘co-PI’ project from the Third Xiangya hospital of Central South University (Grant No. 202430) and the National Natural Science Foundation of China (Grant No. 82303525). Additionally, it was funded by the Major Research Project Fund for High-level Talents in Health and Wellness of Hunan Province (Grant No. 20230591).

Declaration of generative AI use

Generative AI was used solely for limited language editing and polishing of the manuscript. All AI-assisted revisions were reviewed and confirmed by the authors, who take full responsibility for the final content of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  • 1.Wang Y, Bellomo R.. Cardiac surgery-associated acute kidney injury: risk factors, pathophysiology and treatment. Nat Rev Nephrol. 2017;13(11):697–711. doi: 10.1038/nrneph.2017.119. [DOI] [PubMed] [Google Scholar]
  • 2.Fliser D, Laville M, Covic A, et al. A European renal best practice (ERBP) position statement on the kidney disease improving global outcomes (KDIGO) clinical practice guidelines on acute kidney injury: part 1: definitions, conservative management and contrast-induced nephropathy. Nephrol Dial Transplant. 2012;27(12):4263–4272. doi: 10.1093/ndt/gfs375. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Prowle JR, Forni LG, Bell M, et al. Postoperative acute kidney injury in adult non-cardiac surgery: joint consensus report of the acute disease quality initiative and perioperative quality initiative. Nat Rev Nephrol. 2021;17(9):605–618. doi: 10.1038/s41581-021-00418-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Boyer N, Eldridge J, Prowle JR, et al. Postoperative acute kidney injury. Clin J Am Soc Nephrol. 2022;17(10):1535–1545. doi: 10.2215/CJN.16541221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Pickering JW, James MT, Palmer SC.. Acute kidney injury and prognosis after cardiopulmonary bypass: a meta-analysis of cohort studies. Am J Kidney Dis. 2015;65(2):283–293. doi: 10.1053/j.ajkd.2014.09.008. [DOI] [PubMed] [Google Scholar]
  • 6.Kidney Disease: Improving Global Outcomes (KDIGO) . Scope of work: KDIGO clinical practice guideline for acute kidney injury (AKI) and acute kidney disease (AKD): update 2023. 2023. [Google Scholar]
  • 7.Goren O, Matot I.. Perioperative acute kidney injury. Br J Anaesth. 2015;115 Suppl 2(Suppl 2):ii3–14. doi: 10.1093/bja/aev380. [DOI] [PubMed] [Google Scholar]
  • 8.Yang X, Zhu L, Pan H, et al. Cardiopulmonary bypass associated acute kidney injury: better understanding and better prevention. Ren Fail. 2024;46(1):2331062. doi: 10.1080/0886022X.2024.2331062. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Do-Nguyen CC, Sturmer DL, Yang G, et al. Oxygen delivery thresholds during cardiopulmonary bypass and risk for acute kidney injury. Ann Thorac Surg. 2023;116(3):607–613. doi: 10.1016/j.athoracsur.2023.04.049. [DOI] [PubMed] [Google Scholar]
  • 10.Wang S, Chen Y, Han S, et al. Selenium nanoparticles alleviate ischemia reperfusion injury-induced acute kidney injury by modulating GPx-1/NLRP3/Caspase-1 pathway. Theranostics. 2022;12(8):3882–3895. doi: 10.7150/thno.70830. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Zhao W, Nikolic-Paterson DJ, Li K, et al. Selenium binding protein 1 protects renal tubular epithelial cells from ferroptosis by upregulating glutathione peroxidase 4. Chem Biol Interact. 2024;393:110944. doi: 10.1016/j.cbi.2024.110944. [DOI] [PubMed] [Google Scholar]
  • 12.Huang J, Bao D, Lei C-T, et al. Selenoprotein T protects against cisplatin-induced acute kidney injury through suppression of oxidative stress and apoptosis. Faseb J. 2020;34(9):11983–11996. doi: 10.1096/fj.202000180RR. [DOI] [PubMed] [Google Scholar]
  • 13.DeAngelo SL, Győrffy B, Koutmos M, et al. Selenoproteins and tRNA-Sec: regulators of cancer redox homeostasis. Trends Cancer. 2023;9(12):1006–1018. doi: 10.1016/j.trecan.2023.08.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Davis CD, Tsuji PA, Milner JA.. Selenoproteins and cancer prevention. Annu Rev Nutr. 2012;32:73–95. doi: 10.1146/annurev-nutr-071811-150740. [DOI] [PubMed] [Google Scholar]
  • 15.He J, Wu X, Qiao J, et al. Tubular specific glutathione peroxidase 3 deletion exacerbates kidney damage in IRI-AKI mice. Biochim Biophys Acta Mol Basis Dis. 2025;1871(6):167895. doi: 10.1016/j.bbadis.2025.167895. [DOI] [PubMed] [Google Scholar]
  • 16.He W, Tian X, Zhou Q, et al. Selenium nanoparticles ameliorate lumbar disc degeneration by restoring GPX1-mediated redox homeostasis and mitochondrial function of nucleus pulposus cells. J Nanobiotechnology. 2024;22(1):634. doi: 10.1186/s12951-024-02890-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Li L, Lu M, Peng Y, et al. Oxidatively stressed extracellular microenvironment drives fibroblast activation and kidney fibrosis. Redox Biol. 2023;67:102868. doi: 10.1016/j.redox.2023.102868. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Song Y, Zhu M, Islam MA, et al. Glutathione peroxidase 3 is essential for countering senescence in adipose remodelling by maintaining mitochondrial homeostasis. Redox Biol. 2024;77:103365. doi: 10.1016/j.redox.2024.103365. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Barrett CW, Reddy VK, Short SP, et al. Selenoprotein P influences colitis-induced tumorigenesis by mediating stemness and oxidative damage. J Clin Invest. 2015;125(7):2646–2660. doi: 10.1172/JCI76099. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Coppolino G, Celano M, Musolino M, et al. Selenoprotein-P1 (SEPP1) expression in human proximal tubule cells after ischemia-reperfusion injury: an in vitro model. Medicina (Kaunas). 2024;60(6):875. doi: 10.3390/medicina60060875. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Wu X, Tang S, Dai Q, et al. Vitamin D-vitamin D receptor alleviates oxidative stress in ischemic acute kidney injury via upregulating glutathione peroxidase 3. Faseb J. 2023;37(2):e22738. doi: 10.1096/fj.202201400R. [DOI] [PubMed] [Google Scholar]
  • 22.Xia Y, Pan W, Xiao X, et al. MicroRNA-483-5p accentuates cisplatin-induced acute kidney injury by targeting GPX3. Lab Invest. 2022;102(6):589–601. doi: 10.1038/s41374-022-00737-3. [DOI] [PubMed] [Google Scholar]
  • 23.Wang M, Zhou Y, Hao G, et al. Recombinant Klotho alleviates vancomycin-induced acute kidney injury by upregulating anti-oxidative capacity via JAK2/STAT3/GPx3 axis. Toxicology. 2023;499:153657. doi: 10.1016/j.tox.2023.153657. [DOI] [PubMed] [Google Scholar]
  • 24.Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120(4):c179–84. doi: 10.1159/000339789. [DOI] [PubMed] [Google Scholar]
  • 25.Inker LA, Eneanya ND, Coresh J, et al. New creatinine- and cystatin C-based equations to estimate GFR without race. N Engl J Med. 2021;385(19):1737–1749. doi: 10.1056/NEJMoa2102953. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Scurt FG, Bose K, Mertens PR, et al. Cardiac surgery-associated acute kidney injury. Kidney360. 2024;5(6):909–926. doi: 10.34067/KID.0000000000000466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Udzik J, Waszczyk A, Wojciechowska-Koszko I, et al. The utility of novel kidney injury biomarkers in early detection of CSA-AKI. IJMS. 2022;23(24):15864. doi: 10.3390/ijms232415864. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Schaer DJ, Schaer CA, Humar R, et al. Navigating hemolysis and the renal implications of hemoglobin toxicity in cardiac surgery. Anesthesiology. 2024;141(6):1162–1174. doi: 10.1097/ALN.0000000000005109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Molina Andújar A, Lucas A, Escudero VJ, et al. Risk factors for acute kidney injury following cardiac surgery and performance of Leicester score in a spanish cohort. J Clin Med. 2022;11(4):904. doi: 10.3390/jcm11040904. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Stoppe C, Schälte G, Rossaint R, et al. The intraoperative decrease of selenium is associated with the postoperative development of multiorgan dysfunction in cardiac surgical patients. Crit Care Med. 2011;39(8):1879–1885. doi: 10.1097/CCM.0b013e3182190d48. [DOI] [PubMed] [Google Scholar]
  • 31.de Godoy Torso N, Visacri MB, Quintanilha JCF, et al. Assessment of renal function in head and neck cancer patients treated with cisplatin: different biomarkers and acute kidney injury classifications. Int J Mol Sci. 2022;24(1):141. doi: 10.3390/ijms24010141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Brozat JF, Harbalioğlu N, Hohlstein P, et al. Elevated serum KIM-1 in sepsis correlates with kidney dysfunction and the severity of multi-organ critical illness. Int J Mol Sci. 2024;25(11):5819. doi: 10.3390/ijms25115819. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Haase M, Bellomo R, Haase-Fielitz A.. Novel biomarkers, oxidative stress, and the role of labile iron toxicity in cardiopulmonary bypass-associated acute kidney injury. J Am Coll Cardiol. 2010;55(19):2024–2033. doi: 10.1016/j.jacc.2009.12.046. [DOI] [PubMed] [Google Scholar]
  • 34.Olson GE, Winfrey VP, Hill KE, et al. Megalin mediates selenoprotein P uptake by kidney proximal tubule epithelial cells. J Biol Chem. 2008;283(11):6854–6860. doi: 10.1074/jbc.M709945200. [DOI] [PubMed] [Google Scholar]
  • 35.Kurokawa S, Eriksson S, Rose KL, et al. Sepp1(UF) forms are N-terminal selenoprotein P truncations that have peroxidase activity when coupled with thioredoxin reductase-1. Free Radic Biol Med. 2014;69:67–76. doi: 10.1016/j.freeradbiomed.2014.01.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Wu Y, Shi H, Xu Y, et al. Targeted restoration of GPX3 attenuates renal ischemia/reperfusion injury by balancing selenoprotein expression and inhibiting ROS-mediated mitochondrial apoptosis. Transplantation. 2024;108(12):2351–2365. doi: 10.1097/TP.0000000000005068. [DOI] [PubMed] [Google Scholar]
  • 37.Zou Z, Ren T, Li Y, et al. The association between serum glutathione peroxidase-3 concentration and risk of acute kidney injury after cardiac surgery: a nested case-control study. Am J Cardiol. 2023;209:29–35. doi: 10.1016/j.amjcard.2023.08.141. [DOI] [PubMed] [Google Scholar]
  • 38.Liu J, Kumar S, Dolzhenko E, et al. Molecular characterization of the transition from acute to chronic kidney injury following ischemia/reperfusion. JCI Insight. 2017;2(18):e94716. doi: 10.1172/jci.insight.94716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Bolignano D, Jiritano F, Zicarelli M, et al. Selenoprotein P-1 (SEPP1) as an early biomarker of acute kidney injury in patients undergoing cardiopulmonary bypass. Rev Cardiovasc Med. 2022;23(5):170. doi: 10.31083/j.rcm2305170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Ohlsson MA, Molvin J, Holm Isholth H, et al. Selenoprotein P deficiency is associated with early signs of kidney disease and hospitalization risk in heart failure. Nutrients. 2026;18(5):721. doi: 10.3390/nu18050721. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Brodin O, Hackler J, Misra S, et al. Selenoprotein P as biomarker of selenium status in clinical trials with therapeutic dosages of selenite. Nutrients. 2020;12(4):1067. doi: 10.3390/nu12041067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Zhou X, Zhao J, Cheng W, et al. Selenomethionine alleviates LPS-induced septic kidney injury by regulating mitochondrial dynamics changes. Front Pharmacol. 2025;16:1606365. doi: 10.3389/fphar.2025.1606365. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Ayari F, Abdollahzade Fard A, Chodari L.. Selenium pretreatment improve renal function, autophagy signaling pathway and miR21a gene expression in renal ischemia reperfusion injury model in male rat. J Trace Elem Med Biol. 2025;88:127610. doi: 10.1016/j.jtemb.2025.127610. [DOI] [PubMed] [Google Scholar]
  • 44.Zuo Z, Luo M, Liu Z, et al. Selenium nanoparticles alleviate renal ischemia/reperfusion injury by inhibiting ferritinophagy via the XBP1/NCOA4 pathway. Cell Commun Signal. 2024;22(1):376. doi: 10.1186/s12964-024-01751-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Schmidt T, Pargger H, Seeberger E, et al. Effect of high-dose sodium selenite in cardiac surgery patients: a randomized controlled bi-center trial. Clin Nutr. 2018;37(4):1172–1180. doi: 10.1016/j.clnu.2017.04.019. [DOI] [PubMed] [Google Scholar]
  • 46.Sarhan AM, Awad AK, Alassiri AK, et al. Efficacy of selenium on patients undergoing cardiac surgery: a meta-analysis of randomized controlled trials. J Cardiothorac Surg. 2024;19(1):264. doi: 10.1186/s13019-024-02761-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Wang Z, Xu J, Zhang Y, et al. Prediction of acute kidney injury incidence following acute type A aortic dissection surgery with novel biomarkers: a prospective observational study. BMC Med. 2023;21(1):503. doi: 10.1186/s12916-023-03215-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Neyra JA, Hu M-C, Minhajuddin A, et al. Kidney tubular damage and functional biomarkers in acute kidney injury following cardiac surgery. Kidney Int Rep. 2019;4(8):1131–1142. doi: 10.1016/j.ekir.2019.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Material
IRNF_A_2755138_SM7730.docx (103.9KB, docx)
Supplementary File 2_STROBEchecklist.docx
Supplementary Figure2.jpg
Supplementary Figure3.tif
IRNF_A_2755138_SM0069.tif (105.3MB, tif)
Supplementary Figure1.jpg

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.


Articles from Renal Failure are provided here courtesy of Taylor & Francis

RESOURCES