Ji Won Min

The aim of this study is to characterise the immunological features of preformed HLA‐DQ DSAs and investigate their impact on post‐transplant outcomes. Among 1540 ABO‐compatible KT recipients with available high‐resolution HLA typing at Seoul St. Mary's Hospital (January 2010–December 2024), 179 had preformed DSAs. Patients were classified into DQ group ( n = 58), comprising those with DQ DSAs alone or in combination with other DSAs, and non‐DQ ( n = 121) group. We compared baseline DSA characteristics, desensitisation (DSZ) responses and post‐transplant outcomes. The DQ group had a significantly higher proportion of re‐transplant recipients. At baseline, DQ DSAs exhibited markedly higher MFI values compared to non‐DQ DSAs. Following DSZ, 45.2% of DQ DSAs remained strong (MFI > 10,000), compared to only 2.8% of non‐DQ DSAs. While the overall incidence of acute ABMR did not differ significantly between groups, a significant difference was observed in the temporal distribution and the characteristics of the associated DSAs. In early‐onset ABMR, cases in the DQ group developed after the complete resolution of preformed DSAs, whereas those in the non‐DQ group were associated with preformed DSAs. Late‐onset ABMR occurred more frequently in the DQ group (41.7% vs. 10.0%, p = 0.024), and it developed with preformed DQ DSAs. Furthermore, allograft function showed a more rapid deterioration in the DQ group. Preformed DQ DSAs carry a higher immunological risk and exhibit greater resistance to treatment than non‐DQ DSAs, and tend to cause late‐onset acute ABMR rather than immediate post‐transplant rejection. Therefore, more stringent immunological monitoring and immunosuppressive therapy are required.

Background: Sensitization acts as an immunological barrier to successful kidney transplantation (KT). We aim to investigate the efficacy and safety of bortezomib-based desensitization (BOZ-DSZ) in both highly sensitized living donor KT (LDKT) and deceased donor KT (DDKT). Methods: We applied BOZ-DSZ to 20 highly sensitized patients—14 LDKT and six DDKT candidates—and analyzed the change in anti-human leukocyte antigen (HLA) antibody, the success rate of transplantation, and posttransplant outcomes including biopsy-proven allograft rejection (BPAR) rate, infectious complication, and allograft survival. Results: Among 14 LDKT candidates, the peak mean fluorescence intensity (MFI) level of donor-specific anti-HLA antibody (HLA-DSA) decreased in 10 patients (p < 0.05), and the success rate of KT was 92.9% (13 of 14). Incidence of BPAR within the first posttransplant year was 53.8% (7 of 13), and all such cases were rescued by antirejection treatment. One case resulted in mortality due to pneumonia, and there was one allograft failure during the follow-up of 34 months (range, 6–129 months). Among the six DDKT candidates, the peak MFI level of HLA-DSA showed a significant decrease after DSZ in five patients (p = 0.098), and the success rate of KT was 50.0% (3 of 6). One BPAR case (33.3%) occurred within the first posttransplant year and was successfully treated. There was one case of Cytomegalovirus viremia, and there was no allograft failure during the 36-month follow-up (range, 17–42 months).Conclusion BOZ-DSZ is effective and safe in terms of successful DSZ, infectious complications, and allograft outcomes for both highly sensitized LDKT and DDKT candidates.

Background: This study addresses the gap in knowledge regarding the long-term mortality implications of postoperative acute kidney injury (PO-AKI) utilizing advanced machine learning techniques to predict outcomes more accurately than traditional statistical models.Methods: A retrospective cohort study was conducted using data from seven institutions between March 2009 and December 2019. Machine learning models were developed to predict all-cause mortality of PO-AKI patients using 23 preoperative variables and one postoperative variable. Model performance was compared to a traditional statistical approach with Cox regression analysis. The concordance index was used as a predictive performance metric to compare prediction capabilities among different models.Results: Among 199,403 patients, 2,105 developed PO-AKI. During a median follow-up of 144 months (interquartile range, 99.61–170.71 months), 472 in-hospital deaths occurred. Subjects with PO-AKI had a significantly lower survival rate than those without PO-AKI (p < 0.001). For predicting mortality, the XGBoost with an accelerated failure time model had the highest concordance index (0.7521), followed by random survival forest (0.7371), multivariable Cox regression model (0.7318), survival support vector machine (0.7304), and gradient boosting (0.7277). Conclusion: XGBoost with an accelerated failure time model was developed in this study to predict long-term mortality associated with PO-AKI. Its performance was superior to conventional models. The application of machine learning techniques may offer a promising approach to predict mortality following PO-AKI more accurately, providing a basis for developing targeted interventions and clinical guidelines to improve patient outcomes.

Background: The aim of this study is to investigate the impact of sex on the clinical outcomes of spousal donor kidney transplantation. Methods: We analyzed 456 spousal donor kidney transplantation recipients and categorized them into standard or high immunological risk groups according to panel-reactive antibody ≥50% or less. There were 366 recipients in the standard-risk group and 89 recipients in the high-risk group. Results: When comparing biopsy-proven allograft rejection within 1 year from kidney transplantation, husband-to-wife recipients showed significantly higher incidence than wife-to-husband recipients in the high-risk group. By contrast, there was no significant difference between wife-to-husband and husband-to-wife recipients in the standard-risk group. Allograft function recovery was better in husband-to-wife recipients than in wife-to-husband recipients in each group, while husband-to-wife recipients in the high-risk group showed a more rapid decline than other recipients. The long-term patient and allograft survival rates showed no difference between husband-to-wife recipients and wife-to-husband recipients within the same groups. Conclusion: The husband-to-wife recipients with high immunological risk showed a higher risk of biopsy-proven allograft rejection compared to wife-to-husband recipients, so careful monitoring and management may be required.

Objectives: The aim of this study was to stratify cardiovascular risk based on the mean and variability of non-high-density lipoprotein cholesterol (non-HDL-C) in patients undergoing hemodialysis. Methods: Data were analyzed for 453 hemodialysis patients without a history of myocardial infarction (MI) or stroke, who underwent at least five lipid profile measurements at any one of seven university hospitals in the Republic of Korea between March 2009 and December 2020. Visit-to-visit non-HDL-C variability was calculated using the coefficient of variation. The endpoints of the study were newly diagnosed MI, stroke, or all-cause death. Patients were divided into four groups according to quartiles of the mean and variability of non-HDL-C. Results: During a median follow-up of 97.0 months, there were 39 cases of MI, 99 cases of stroke, and 96 deaths. The cumulative incidence rate of MI was significantly highest in the low mean/high variability group (log-rank p = 0.0296). However, there were no significant differences between groups in the incidence rates of stroke (log-rank p = 0.9939) or all-cause mortality (log-rank p = 0.9373). In the multivariable Cox regression analysis, age and low mean/high variability (HR: 3.311, 95% CI: 1.380–7.944) were identified as independent risk factors for MI. However, for stroke and all-cause mortality, age was the only independent risk factor. Moreover, neither the mean nor the variability of non-HDL-C alone was associated with MI, stroke, or all-cause mortality. Conclusions: Our results suggest that the coexistence of low mean non-HDL-C and high variability is associated with an elevated risk of MI in hemodialysis patients.

Background: Continuous renal replacement therapy is crucial for treating critically ill patients with acute kidney injury. This study aimed to ascertain the association between urine output at continuous renal replacement therapy discontinuation and 1-year allcause mortality and the development of end-stage renal disease (ESRD) in patients with acute kidney injury. Methods: This study included 851 patients with acute kidney injury who underwent continuous renal replacement therapy at three tertiary hospitals from 2012 to 2020 and categorized them into four groups based on 24-hour urine output (mL/day) on the day of continuous renal replacement therapy discontinuation: <100, 100-499.9, 500-1,499.9, and ≥1,500. Results: During the 1 year after continuous renal replacement therapy discontinuation, 333 (39.1%) of 851 patients died, and 150 (30.1%) of 499 patients progressed to ESRD. Multivariable analysis revealed that decreased urine output was associated with a higher risk for all-cause mortality (<100 mL/day: hazard ratio [HR] 3.07, 95% confidence interval [CI] 2.22-4.24, p < 0.01; 100-499.9 mL/day: HR 2.25, 95% CI 1.58-3.19, p < 0.01; 500-1,499.9 mL/day: HR 1.67, 95% CI 1.16-2.41, p < 0.01), and for the development of ESRD (<100 mL/day: odds ratio [OR] 16.01, 95% CI 7.54-33.99, p < 0.01; 100-499.9 mL/day: OR 7.17, 95% CI 3.30-15.56, p < 0.01; 500-1,499 mL/day: OR 3.11, 95% CI 1.40-6.93, p = 0.01). Conclusion: This study highlights the importance of monitoring urine output in predicting the long-term risks of ESRD and mortality in patients with acute kidney injury undergoing continuous renal replacement therapy.

Background: The aim of this study is to investigate the specific pathway involved in human leukocyte antigen (HLA) sensitization using single-cell RNA-sequencing analysis and an allo-sensitized mouse model developed with an HLA.A2 transgenic mouse. Methods: For sensitization, wild-type C57BL/6 mouse received two skin grafts from C57BL/6-Tg(HLA-A2.1)1Enge/J mouse (allogeneic mouse, ALLO). For syngeneic control (SYN), skin grafts were transferred from C57BL/6 to C57BL/6. We performed single-cell RNA-sequencing analysis on splenocytes isolated from ALLO and SYN and compared the gene expression between them.Results: We generated 9,190 and 8,890 single-cell transcriptomes from ALLO and SYN, respectively. Five major cell types (B cells, T cells, natural killer cells, macrophages, and neutrophils) and their transcriptome data were annotated according to the representative differentially expressed genes of each cell cluster. The percentage of B cells was higher in ALLO than it was in SYN. Kyoto Encyclopedia of Genes and Genomes enrichment analyses indicated that the highly expressed genes in the B cells from ALLO were mainly associated with antigen processing and presentation pathways, allograft rejection, and the Th17 cell differentiation pathway. Upregulated genes in the T cells of ALLO were involved in the interleukin (IL)-17 signaling pathway. The ratio of Th17 cluster and Treg cluster was increased in the ALLO. On flow cytometry, the percentage of Th17 (IL-17+/CD4+ T) cells was higher and regulatory T cells (FOXP3+/CD4+ T) was lower in the ALLO compared to those in the SYN.Conclusion: Our results indicate that not only the B cell lineage but also the Th17 cells and their cytokine (IL-17) are involved in the sensitization to HLA.

Malnutrition is a common and serious complication in patients with chronic kidney disease (CKD), significantly impacting morbidity, mortality, and health-related quality of life. Recognizing the close association between nutritional status and clinical outcomes, recent clinical practice guidelines have emphasized proactive nutritional assessment and individualized intervention as key components of CKD management. Accurate nutritional assessment remains challenging in CKD patients due to fluid shifts, altered body composition, and laboratory variability. While various nutritional assessment tools are available, their interpretation requires careful consideration due to each tool’s characteristics and patient’s CKD stage. Nutritional interventions must be tailored to the patient’s CKD stage, dialysis status, and comorbidity profile. Strategies include individualized dietary counseling and oral nutritional supplements. Also, patient education and multidisciplinary collaboration—particularly involving nephrologists and renal dietitians—are essential to improving adherence and long-term outcomes. This review focuses on evaluating the nutritional status and intervention of CKD patients, highlighting key aspects based on the latest literature and clinical guidelines.

Background: Postoperative acute kidney injury (AKI) is a significant risk associated with surgeries under general anesthesia, often leading to increased mortality and morbidity. Existing predictive models for postoperative AKI are usually limited to specific surgical areas or require external validation. Objective: We proposed to build a prediction model for postoperative AKI using several machine learning methods. Methods: We conducted a retrospective cohort analysis of noncardiac surgeries from 2009 to 2019 at seven university hospitals in South Korea. We evaluated six machine learning models: deep neural network, logistic regression, decision tree, random forest, light gradient boosting machine, and naïve Bayes for predicting postoperative AKI, defined as a significant increase in serum creatinine or the initiation of renal replacement therapy within 30 days after surgery. The performance of the models was analyzed using the area under the curve (AUC) of the receiver operating characteristic curve, accuracy, precision, sensitivity (recall), specificity, and F1-score. Results: Among the 239,267 surgeries analyzed, 7935 cases of postoperative AKI were identified. The models, using 38 preoperative predictors, showed that deep neural network (AUC=0.832), light gradient boosting machine (AUC=0.836), and logistic regression (AUC=0.825) demonstrated superior performance in predicting AKI risk. The deep neural network model was then developed into a user-friendly website for clinical use. Conclusions: Our study introduces a robust, high-performance AKI risk prediction system that is applicable in clinical settings using preoperative data. This model’s integration into a user-friendly website enhances its clinical utility, offering a significant step forward in personalized patient care and risk management.

Eum SH, et al. J Korean Med Sci. 2025 Nov;40:e308. https://doi.org/10.3346/jkms.2025.40.e308