Prediction of post-donation renal function using machine learning techniques and conventional regression models in living kidney donors.

BACKGROUND: Accurate prediction of renal function following kidney donation and careful selection of living donors are essential for living-kidney donation programs. We aimed to develop a prediction model for post-donation renal function following living kidney donation using machine learning. METHODS: This retrospective cohort study was conducted with 823 living kidney donors between 2009 and 2020. The dataset was randomly split into training (80%) and test sets (20%). The main outcome was the post-donation estimated glomerular filtration rate (eGFR) 12 months after nephrectomy. We compared the performance of machine learning techniques, traditional regression models, and models from previous studies. The best-performing model was selected based on the mean absolute error (MAE) and root mean square error (RMSE). RESULTS: The mean age of the participants was 45.2 ± 12.3 years, and 48.4% were males. The mean pre-donation and post-donation eGFRs were 101.3 ± 13.0 and 68.8 ± 12.7 mL/min/1.73 m2, respectively. The XGBoost model with the eGFR, age, serum creatinine, 24-h urine creatinine, 24-h urine sodium, creatinine clearance, cystatin C, cystatin C-based eGFR, computed tomography volume of the remaining kidney/body weight, normalized GFR of the remaining kidney measured through a diethylenetriaminepentaacetic acid scan, and sex, showed the best performance with a mean absolute error of 6.23 and root mean square error of 8.06. An easy-to-use web application titled Kidney Donation with Nephrologic Intelligence (KDNI) was developed. CONCLUSIONS: The prediction model using XGBoost accurately predicted the post-donation eGFR after living kidney donation. This model can be applied in clinical practice using KDNI, the developed web application.

Impact of anemia requiring transfusion or erythropoiesis-stimulating agents on new-onset cardiovascular events and mortality after continuous renal replacement therapy.

Anemia is common in critically ill patients undergoing continuous renal replacement therapy (CRRT). We investigated the impact of anemia requiring red blood cell (RBC) transfusion or erythropoiesis-stimulating agents (ESAs) on patient outcomes after hospital discharge in critically ill patients with acute kidney injury (AKI) requiring CRRT. In this retrospective cohort study using the Health Insurance Review and Assessment database of South Korea, 10,923 adult patients who received CRRT for 3 days or more between 2010 and 2019 and discharged alive were included. Anemia was defined as the need for RBC transfusion or ESAs. Outcomes included cardiovascular events (CVEs) and all-cause mortality after discharge. The anemia group showed a tendency to be older with more females and had more comorbidities compared to the control group. Anemia was not associated with an increased risk of CVEs (adjusted hazard ratio [aHR]: 1.05; 95% confidence interval [CI]: 0.85-1.29), but was associated with an increased risk of all-cause mortality (aHR: 1.41; 95% CI 1.30-1.53). For critically ill patients with AKI requiring CRRT, anemia, defined as requirement for RBC transfusion or ESAs, may increase the long-term risk of all-cause mortality.

Acute kidney injury in bortezomib-treated patients with multiple myeloma.

BACKGROUND: The nephrotoxicity of bortezomib, a proteasome inhibitor, has not yet been elucidated, although tumor lysis syndrome (TLS) associated with multiple myeloma (MM) has been reported to increase after introduction of the drug. This study compared the incidence and risk factors for acute kidney injury (AKI) and TLS in patients with MM after bortezomib-based chemotherapy to investigate drug-related nephrotoxicity. METHODS: From 2006 to 2017, 276 patients who underwent a first cycle of bortezomib-based chemotherapy for MM were identified in a single tertiary hospital. Laboratory TLS was defined according to the Cairo-Bishop definition. Development of AKI was assessed by AKI Network criteria within 7 days of the first chemotherapy. RESULTS: The median (interquartile range) age was 65 (56-72) years, and baseline estimated glomerular filtration rate (eGFR) was 61.3 (34.1-89.1) mL/min/1.73 m2. The incidences of AKI and laboratory TLS were 17% (n = 47) and 13% (n = 36), respectively. Ten (3.6%) subjects met both AKI and TLS criteria. Multivariate analyses showed that lower eGFR category [30-59, odds ratio (OR) 3.005 (95% confidence interval 1.163-7.976); 15-29, OR 4.225 (1.183-15.000); <15, OR 16.154 (3.831-70.920) vs ≥60, P < .001], lower serum albumin level [per 1 increase, OR 0.479 (0.256-0.871), P = .018], renal amyloidosis [OR 13.039 (4.108-44.041), P < .001] and use of acyclovir during bortezomib treatment [OR 3.689 (1.133-14.469), P = .042] were predictors of AKI. MM stages and ß-2-microglobulin were not associated with increased risk of AKI. Regarding laboratory TLS, MM stage and ß-2-microglobulin were higher in those with TLS than in others. In multivariate analyses, ß-2-microglobulin level [OR 1.204 (1.005-1.461), P = .038] and absence of high-risk chromosome abnormalities [OR 0.143 (0.022-0.588), P = .016] were associated with higher risk of TLS. CONCLUSIONS: Development of AKI was often observed in the absence of TLS in patients with MM after treatment with bortezomib. In addition, the risk factors for AKI and TLS varied widely. These findings indicate the potential nephrotoxicity of bortezomib irrespective of TLS in patients with decreased kidney function.

Recoverability of Diabetic Nephropathy of Donor Kidney After Kidney Transplantation.

Some kidney donors have diabetes, and little of their natural course of diabetic nephropathy (DN) is known. The aim of this study was to analyze the changes in pathologic lesions in the diabetic donor kidney after KT by performing protocol biopsy two weeks and one year after KT. This retrospective study included 103 patients who underwent KT, with kidneys from donors with a history of diabetes mellitus (DM). Among them, data of 34 patients who underwent biopsy two weeks and one year after KT were reviewed. Biopsy specimens were reviewed using light microscopy and electron microscopy. Glomerular basement membrane (GBM) thickness at 2 weeks and 1 year was compared. Biopsy showed that DN occurred in 29 of the 34 patients. Only trivial histological changes were observed in 22 patients (64.7%), including 5 patients who did not show DN. At one year after transplantation, there was no change in the DN histologic class in 26 patients (76.5%), and there was no statistically significant difference in the change in GBM thickness. This pattern was observed regardless of the recipient's DM or glycemic control. With this understanding, clinicians can use kidneys from DM donors with more comfort, thereby reducing the kidney discard rate.

Depression is a main determinant of health-related quality of life in patients with diabetic kidney disease.

Low health-related quality of life (HRQOL) is associated with adverse outcomes in diabetic kidney disease (DKD) patients. We examined the modifiable factors associated with low HRQOL in these patients. We enrolled 141 DKD patients. HRQOL was assessed with the Short Form 36 (SF-36) questionnaire. Low HRQOL was defined as a score > one standard deviation below the mean. Depression and anxiety were assessed with the Hospital Anxiety and Depression Scale (HADS-D and HDAS-A, respectively). The patients' median age was 65 years, and 73% were men. The prevalence rates of anxiety and depression were 8% (n = 11) and 17% (n = 24), respectively. Forty (28%) patients were identified as poor sleepers, and 40 (28%) had low physical activity levels. Anxiety, depression, and poor sleep quality were negatively correlated with SF-36 scores. Higher levels of physical activity and the estimated glomerular filtration rate (eGFR) were correlated with higher SF-36 scores, which indicated better health status. Higher depression scores (HADS-D scores) were associated with low HRQOL, independent of factors including age, sex, smoking status, comorbidities, eGFR, anemia, sleep quality, anxiety levels, and physical activity levels (odds ratio, 1.43; 95% confidence interval, 1.17-1.75). Among the clinical and psycho-physical factors, depression was a main determinant of low HRQOL in DKD patients.

Clinical value of urinary cytokines/chemokines as prognostic markers in patients with crescentic glomerulonephritis.

Crescentic glomerulonephritis (CrGN) usually requires urgent immunosuppressive treatment. However, aggressive immunosuppressive treatment is often difficult because of the patients' medical conditions or comorbidities. Prognostic markers including urinary cytokines/chemokines as noninvasive biomarkers were explored in CrGN patients. This prospective cohort study included 82 patients with biopsy-confirmed CrGN from 2002 to 2015 who were followed up for 5 years. Urine and serum cytokines/chemokines on the day of kidney biopsy were analyzed in 36 patients. The median age was 65 years and 47.6% were male. Baseline estimated glomerular filtration rate (eGFR) and interstitial fibrosis and tubular atrophy (IFTA) scores were identified as significant prognostic factors. Among patients with cytokines/chemokines measurement, increased IL-10 level was identified as an independent predictor of good prognosis, and increased levels of urinary MCP-1 and fractalkine tended to be associated with good prognosis after adjusting for baseline eGFR and IFTA score. However, semiquantitative analysis of intrarenal leukocytes did not show prognostic value predicting renal outcome or correlation with urinary cytokines/chemokines. This study supports the clinical importance of baseline eGFR and IFTA scores and suggests potential usefulness of urinary IL-10, MCP-1, and fractalkine as prognostic markers for predicting renal outcomes in patients with CrGN.

Repair phase modeling of ischemic acute kidney injury: recovery vs. transition to chronic kidney disease.

The repair mechanism after ischemic acute kidney injury (AKI) involves complex immunologic processes, which determine long-term renal outcomes. Through investigating two murine ischemia-reperfusion injury (IRI) models: bilateral IRI (BIRI) and unilateral IRI (UIRI), we aimed to determine an appropriate murine model that could simulate the recovery phase of ischemic AKI. Changes in renal function, phenotypes of kidney mononuclear cells, renal fibrosis, and intrarenal cytokine/chemokine expression were serially analyzed up to 12 weeks after IRI. Plasma creatinine and BUN concentrations increased and remained elevated in the BIRI group until 7 days but decreased to comparable levels with the sham control group at 2 weeks after surgery and thereafter, whereas plasma creatinine and BUN concentrations remained unchanged in the UIRI group. Intrarenal total leukocytes, and effector memory and activated phenotypes of CD4 and CD8 T cells markedly increased in the postischemic kidneys in both IRI groups. Expression of proinflammatory cytokines/chemokines and TGF-ß1 was enhanced in the postischemic kidneys of both IRI groups with a higher degree in the UIRI group. Importantly, intrarenal immunologic changes of the BIRI group persisted until 6 weeks despite full functional recovery. The postischemic kidneys of the UIRI group showed earlier and more pronounced proinflammatory conditions as well as more severe atrophic and fibrotic changes compared to the BIRI group. These findings support the utility of longer follow-ups of BIRI and UIRI models for investigating the adaptive repair process, which facilitates recovery of ischemic AKI and maladaptive repair process may result in AKI to CKD transition, respectively.

Poly (ADP-Ribose) Polymerase Inhibitor Treatment as a Novel Therapy Attenuating Renal Ischemia-Reperfusion Injury.

Intrarenal robust inflammatory response following ischemia-reperfusion injury (IRI) is a major factor in the pathogenesis of renal injury in ischemic acute kidney injury (AKI). Although numerous studies have investigated various agents of immune modulation or suppression for ischemic AKI, few showed reproducible effects. We hypothesized that poly (ADP-ribose) polymerase (PARP) inhibitor may favorably change post-ischemic intrarenal immunologic micromilieu by reducing damage-associated molecular pattern (DAMP) signals and improve renal outcome in ischemic AKI. The effects of JPI-289 (a PARP inhibitor) on early renal injury in a murine IRI model and hypoxic HK-2 cell model were investigated. Bilateral IRI surgery was performed in three groups of 9-week-old male C57BL/6 mice (control, JPI-289 50 mg/kg, and JPI-289 100 mg/kg; n = 9-10 in each group). Saline or JPI-289 was intraperitoneally injected. Renal function deterioration was significantly attenuated in the JPI-289 treatment groups in a dose-dependent manner. Inflammatory cell infiltration and proinflammatory cytokine/chemokine expressions in the post-ischemic kidneys were also attenuated by JPI-289 treatment. JPI-289 treatment at 0.5 and 0.75 µg/ml facilitated the proliferation of hypoxic HK-2 cells. PARP inhibition with JPI-289 treatment showed favorable effects in ischemic AKI by attenuating intrarenal inflammatory cascade in a murine model and facilitating proliferation of hypoxic HK-2 cells.

The combination of area under the curve and percentage change in estimated glomerular filtration rate predicts long-term outcome of kidney transplants.

The development of surrogate markers for long-term outcomes of kidney transplant (KT) is a focus of attention. We examined the possibility of using a combination of the area under the curve of estimated glomerular filtration rate (eGFR) for 2 years (AUCeGFR2yrs ) and percent change in eGFR between 1 and 2 years after KT (% changeeGFR1/2yr ) as a surrogate marker. We compared the predictive power of death-censored graft failure with various combinations. The combination of >2% vs ≤2% for % changeeGFR1/2yr and >1300 vs ≤1300 mL/min/month for AUCeGFR2yr had the highest Harrell C-index (0.647; 95% confidence interval [95% CI], 0.604-0.690). The death-censored graft survival rate of the group with ≤2% changeeGFR1/2yr and ≤1300 mL/min/month AUCeGFR2yr was significantly lower than those of other groups. The AUC/% change eGFR had comparable predictive power to the previously identified marker ≥30% decline in eGFR between years 1 and 3 after KT (≤-30% changeeGFR1/3yr ) (Harrell's C-index = 0.645 [95% CI 0.628-0.662] for ≤-30% changeeGFR1/3yr ). The proposed combination might be useful as a surrogate marker in KT trials because it requires a shorter surveillance period than the established marker while having comparable predictive power.

Risk Prediction for Contrast-Induced Nephropathy in Cancer Patients Undergoing Computed Tomography under Preventive Measures.

BACKGROUND: Contrast-induced nephropathy (CIN) is a major cause of acute kidney injury in chronic kidney disease. Many cancer patients have risk factors for CIN and frequently undergo contrast-enhanced computed tomography (CECT). We aimed to develop a risk prediction model for CIN in cancer patients undergoing CECT. METHODS: Between 2009 and 2017, 2,240 cancer patients with estimated glomerular filtration rate (eGFR) < 45 mL/min/1.73 m2 who underwent CECT with CIN preventive measures were included in a development cohort. Primary outcome was development of CIN, defined as 25% increase in serum creatinine within 2-6 days after contrast exposure. A prediction model was developed using logistic regression analysis. The model was evaluated for prognostic utility in an independent cohort (N = 555). RESULTS: Overall incidence of CIN was 2.5% (55/2,240). In multivariable analysis, eGFR, diabetes mellitus, and serum albumin level were identified as independent predictors of CIN. A prediction model including eGFR, serum albumin level, and diabetes mellitus was developed, and risk scores ranged from 0 to 6 points. The model demonstrated fair discriminative power (C statistic = 0.733, 95% confidence interval [CI] 0.656-0.810) and good calibration (calibration slope 0.867, 95% Cl 0.719-1.015). In the validation cohort, the model also demonstrated fair discriminative power (C statistic = 0.749, 95% CI 0.648-0.849) and good calibration (calibration slope 0.974, 95% CI 0.634-1.315). CONCLUSIONS: The proposed model has good predictive ability for risk of CIN in cancer patients with chronic kidney disease. This model can aid in risk stratification for CIN in patients undergoing CECT.

Association between diuretics and successful discontinuation of continuous renal replacement therapy in critically ill patients with acute kidney injury.

BACKGROUND: Despite aggressive application of continuous renal replacement therapy (CRRT) in critically ill patients with acute kidney injury (AKI), there is no consensus on diuretic therapy when discontinuation of CRRT is attempted. The effect of diuretics on discontinuation of CRRT in critically ill patients was evaluated. METHODS: This retrospective cohort study enrolled 1176 adult patients who survived for more than 3 days after discontinuing CRRT between 2009 and 2014. Patients were categorized depending on the re-initiation of renal replacement therapy within 3 days after discontinuing CRRT or use of diuretics. Changes in urine output (UO) and renal function after discontinuing CRRT were outcomes. Predictive factors for successful discontinuation of CRRT were also analyzed. RESULTS: The CRRT discontinuation group had a shorter duration of CRRT, more frequent use of diuretics after discontinuing CRRT, and greater UO on the day before CRRT discontinuation [day minus 1 (day - 1)]. The diuretics group had greater increases in UO and serum creatinine elevation after discontinuing CRRT. In the CRRT discontinuation group, continuous infusion of furosemide tended to increase UO more effectively. Multivariable regression analysis identified high day - 1 UO and use of diuretics as significant predictors of successful discontinuation of CRRT. Day - 1 UO of 125 mL/day was the cutoff value for predicting successful discontinuation of CRRT in oliguric patients treated with diuretics following CRRT. CONCLUSIONS: Day - 1 UO and aggressive diuretic therapy were associated with successful CRRT discontinuation. Diuretic therapy may be helpful when attempting CRRT discontinuation in critically ill patients with AKI, by inducing a favorable fluid balance, especially in oliguric patients.

Clinical relevance of urinary angiotensinogen and renin as potential biomarkers in patients with overt proteinuria.

Urinary angiotensinogen (AGT) and renin have been reported to reflect the intrarenal renin-angiotensin system (RAS) activity. However, the adequacy and clinical significance of these markers have not been evaluated in overtly proteinuric patients. In patients with biopsy-proven glomerulonephritis, plasma and urinary AGT and renin were analyzed. A cohort of 75 patients treated with RAS inhibitors was followed for 1 year. Among the 207 patients, 105 had subnephrotic and 102 had nephrotic-range proteinuria. Mean age, estimated glomerular filtration rate (eGFR), and urinary protein-to-creatinine ratio (P/Cr) of all patients were 48 years, 79.7 mL/min/1.73 m(2), and 5.66 mg/mg, respectively. Both natural logarithm of urinary AGT/creatinine (ln [urinary AGT/Cr]) and ln (urinary renin/Cr) showed positive correlations with urinary P/Cr. There was a positive correlation between ln (urinary AGT/Cr) and ln (urinary renin/Cr). Ln (urinary renin/Cr) was not affected by ln (plasma renin) regardless of the degree of proteinuria. The treatment response to RAS inhibitors was greatest in patients with high urinary AGT and renin. However, the predictive value of those parameters was no longer present when the values were adjusted by the degree of proteinuria. Ln (urinary renin/Cr) and initial eGFR were independently associated with the changes in renal function for 1 year. Ln (urinary AGT/Cr) was associated with persistent overt proteinuria after 1 year. Our study suggests that urinary renin may be a better marker in heavy proteinuria, and the treatment response to RAS inhibitors may be enhanced in patients with high urinary renin and AGT. Further studies will be necessary to explore the value of urinary AGT and renin.