Heilman, Raymond L.

Importance Accurate prediction of patient mortality after kidney transplant is an unmet need. Objective To develop and validate an integrative prediction model that predicts short- and long-term patient mortality for kidney recipients. Design, Setting, and Participants This international cohort study included patients who underwent transplants between 2004 and 2023 from 14 academic medical centers from Europe and the US. The derivation cohort consisted of 1566 consecutive adult kidney recipients in a deeply phenotyped cohort prospectively recruited in 1 French center between 2004 and 2014. External validation cohorts consisted of 10 951 kidney recipients from 7 centers in France, 3 centers in Europe, and 3 centers in North America who underwent kidney transplants between 2006 and 2023. Data were analyzed from January 2023 to June 2025. Main Outcome and Measures All-cause mortality was assessed, and 121 candidate prognostic factors, comprising demographics and clinical, biological, imaging, and immunological parameters, were collected. Results Among 12 517 kidney transplant recipients, including 1566 in the derivation cohort (mean [SD] age, 50.05 [14.31] years; 942 male [60.15%]) and 10 951 in validation cohorts (mean [SD] age, 53.32 [13.97] years; 6766 male [61.78%]), 2486 patients (19.9%) died after a median (IQR) follow-up of 5.08 (2.97-7.00) years. Fourteen prognostic factors (including clinical, biological, and imaging risk factors) were independently associated with patient death (eg, patient age: hazard ratio per 1-year increase in age, 1.07 [95% CI, 1.06-1.08]; P < .001) and were combined into a risk prediction model (mBox). The model exhibited accurate calibration and discrimination in the derivation cohort, with C statistics of 0.82 (95% CI, 0.77-0.87) and 0.80 (95% CI, 0.78-0.82) at 1 and 10 years, respectively, after the transplant. Abbreviated models were developed and validated to ensure model generalizability. Performance of abbreviated models was confirmed in external validation cohorts from France (C statistic, 0.76 [95% CI, 0.73-0.78]), Europe (C statistic, 0.74 [95% CI, 0.72-0.76]), the US (C statistic, 0.74 [95% CI, 0.70-0.78]), the Greater Paris University Hospital database (C statistic, 0.79 [95% CI, 0.77-0.81]), and the University of California at San Francisco database (C statistic, 0.70 [95% CI, 0.65-0.74]). The model was also validated in a series of subpopulations (eg, women vs men: C statistic, 0.81 [95% CI, 0.77-0.84] vs 0.79 [95% CI, 0.77-0.82]) and clinical scenarios (eg, before COVID-19 era: C statistic, 0.79 [95% CI, 0.77-0.81]) with good and stable performance. Conclusions and Relevance In this study, an accurate prediction model for mortality among kidney recipients, computable at the time of transplant, was developed and externally validated. Results suggest that this model may help stratify patient risk of death, allowing for improved medical decisions.

ABSTRACT Predicting delayed graft function (DGF) relies on donor, recipient, and perioperative factors. Despite the growing recognition that DGF duration strongly influences patient outcomes, no models currently address or predict its impact—highlighting an important gap in current research and clinical practice. This study aimed to develop an ensemble‐based machine learning model using perioperative data to predict DGF occurrence and duration. The gradient‐boosted decision trees model was trained and validated on 2725 patients, with k‐fold cross‐validation in an external cohort of 284 patients. Model performance was evaluated based on accuracy, ROC‐AUC, and other metrics using R and Python libraries. The binary DGF prediction model achieved a ROC‐AUC of 0.77, while the DGF duration classification model had an accuracy of 79.2%. DGF duration AUC values by time interval were 0.76 for 0–1 weeks, 0.81 for >1–2 weeks, 0.80 for >2–3 weeks, and 0.87 for >3 weeks; the overall macro‐averaged AUC was 0.81. The mean Brier score for multi‐class predictions was 0.14. External validation showed a 78% accuracy for DGF duration prediction. Acute kidney injury (AKI) and donor donation after circulatory death (DCD) status were key DGF predictors. The use of gradient‐boosted decision trees (GBDT) improves the prediction of both the likelihood and duration of DGF, addressing a current gap in kidney transplant patient care. By facilitating personalized transplant care, this model supports more effective perioperative planning and timely interventions, which may contribute to better patient outcomes.

Background The primary objective of this analysis was to assess the association between non-invasive biomarker results within the first 2-years post-kidney transplant and 10-year All Cause Graft Loss (ACGL). Methods CTOT-08 was linked with the Scientific Registry of Transplant Recipients (SRTR) to assess 10-year ACGL. Cox Regression was used to determine the association between gene expression profile (GEP) and donor-derived cell-free DNA (dd-cfDNA) results and ACGL. A Cox Regression with time-varying covariates was used to test associations with ACGL with simultaneous and repeated biomarker results. Results The final landmarked Cox model included clinical acute rejection (aHR 2.58 (1,30,5.13) for 1 rejection, aHR 3.41 (1.07,10.85) for 2 rejections), diabetes (aHR 1.71 (0.99,2.96)), recipient age (aHR 1.02 (1.00,1.04)), the highest quartile of positive GEP (aHR 2.44 (1.45,4.10)) and the highest quartile of positive dd-cfDNA (aHR 2.91 (1.77,4.78)). In the time-varying model, having both biomarkers positive at the same time carried the greatest association with ACGL (aHR 4.29 (2.10,8.76)). Conclusions In this multicenter, prospective analysis of GEP and dd-cfDNA on patient and graft outcomes, an independent association with long-term ACGL was demonstrated. Further, a time-varying analysis demonstrated that two biomarkers with independent methodologies tested at the same time can provide more information than either one alone.

Background. Long-term outcomes in kidney transplantation remain suboptimal, with subclinical acute rejection (subAR) frequently going undetected by conventional methods. We hypothesized that subAR represents a molecular precursor to clinical acute rejection (cAR). Leveraging a large National Institutes of Health-funded cohort with matched kidney tissue and peripheral blood samples, our primary objective was to determine whether subAR and cAR share molecular patterns, establishing subAR as an early stage of cAR. As a secondary aim, we compared biopsy and blood gene expression profiles to identify unique and shared pathways and to assess their diagnostic utility. Methods. This was a retrospective analysis of gene expression data from an observational clinical trial. We analyzed 578 blood and biopsy samples from 129 kidney transplant recipients enrolled in a multicenter observational study. Gene expression differences were analyzed using ANOVA with false discovery rate correction and molecular pathways analysis. Samples were matched and classified based on histologic findings and blood-based molecular diagnostics. Results. Gene expression and pathway analyses revealed that subAR exhibits a molecular profile consistent with a milder form of clinical rejection in both tissue and blood. Molecular signatures distinguished immune-mediated injury from nonimmune injury. Notably, tissue gene expression profiling in cases that were misclassified by peripheral blood diagnostics were like correctly classified cases, suggesting that histology alone may not reliably define true clinical phenotypes. Conclusions. Molecular profiling of tissue and blood reflects distinct but complementary aspects of transplant biology and confirms that subAR and cAR share overlapping molecular profiles, supporting the hypothesis that subAR represents an early stage in the progression toward clinical rejection. These findings highlight the limitations of relying solely on histological evaluation and support the use of molecular tissue profiling as an adjunct to histology, particularly in diagnostically ambiguous cases. Molecular profiling provides a supporting framework for identifying subclinical rejection and guiding personalized immunosuppression to improve long-term outcomes.

ABSTRACTBackgroundDespite many people awaiting kidney transplant, kidney allografts from acute kidney injury (AKI) donors continue to be underutilized. We aimed to cluster kidney transplant recipients of AKI kidney allografts using an unsupervised machine learning (ML) approach.MethodsUsing Organ Procurement and Transplantation Network–United Network for Organ Sharing (OPTN/UNOS) data, a consensus clustering cohort analysis was performed in 12 356 deceased donor kidney transplant recipients between 2015 and 2019 in whom donors had a terminal serum creatinine ≥1.5 mg/dL. Significant cluster characteristics were determined, and outcomes were compared.ResultsThe median donor terminal creatinine was 2.2 (interquartile range [IQR] 1.7–3.3) mg/dL. Cluster analysis was performed on 12 356 AKI kidney recipients, and three clinically distinct clusters were identified. Young, sensitized kidney re‐transplant patients characterized Cluster 1. Cluster 2 was characterized by first‐time kidney transplant patients with hypertensive and diabetic kidney diseases. Older diabetic recipients characterized Cluster 3. Clusters 1 and 2 donors were young and met standard kidney donor profile index (KDPI) criteria; Cluster 3 donors were older, more likely to have hypertension or diabetes, and meet high KDPI criteria. Cluster 1 had a higher risk of acute rejection, 3‐year patient death, and graft failure. Cluster 3 had a higher risk of death‐censored graft failure, patient death, and graft failure at 1 and 3 years. Cluster 2 had the best patient‐, graft‐, and death‐censored graft survival at 1 and 3 years. Compared to non‐AKI kidney recipients, the AKI clusters showed a higher incidence of delayed graft function (DGF, AKI: 43.2%, 41.7%, 45.3% vs. non‐AKI: 25.5%); however, there were comparable long‐term outcomes specific to death‐censored graft survival (AKI: 93.6%, 93.4%, 90.4% vs. non‐AKI: 92.3%), patient survival (AKI: 89.1%, 93.2%, 84.2% vs. non‐AKI: 91.2%), and overall graft survival (AKI: 84.7%, 88.2%, 79.0% vs. non‐AKI: 86.0%).ConclusionsIn this unsupervised ML approach study, AKI recipient clusters demonstrated differing, but good clinical outcomes, suggesting opportunities for transplant centers to incrementally increase kidney utilization from AKI donors.

ABSTRACTThis single‐center retrospective study investigated subclinical rejection prevalence and significance in simultaneous pancreas and kidney transplant (SPKT) recipients. We analyzed 352 SPKT recipients from July 2003 to April 2022. Our protocol included pancreas allograft surveillance biopsies at 1, 4, and 12months post‐transplant. After excluding 153 patients unable to undergo pancreas biopsy, our study cohort comprised 199 recipients. Among the 199 patients with protocol pancreas biopsies, 107 had multiple protocol pancreas biopsies in the first year, totaling 323. Subclinical rejection was identified in 132 episodes (41%). Of these, 72% were Grade 1, 20% were indeterminate, and 8% were Banff Grade 2 or higher. All episodes of subclinical rejection were treated. Rates of pancreas graft loss (10% vs. 7%) and clinical rejection (21% vs. 20%) at 3 years were similar between those with and without subclinical rejection. Subclinical rejection Banff Grade 2 or more was associated with poor pancreas graft survival HR of 5.5 (95% CI: 1.24–24.37, p = 0.025). Of 236 simultaneous protocol kidney and pancreas biopsies, 102 (43%) showed pancreas subclinical rejection, while only 17% had concurrent kidney subclinical rejection. Our findings suggest limited predictive value of pancreatic enzymes and euglycemia in detecting pancreas rejection. Furthermore, poor concordance existed between pancreas and kidney subclinical rejection.

Key Points Peripheral blood biomarkers may have value in monitoring kidney transplant recipients after treatment of acute rejection. The donor-derived cellfree DNA may be more sensitive to identifying antibody-mediated rejection and gene expression profile may be more sensitive to identifying acute cellular rejection. Background Persistent rejection is an increasingly recognized barrier to long-term kidney allograft survival. A noninvasive method to help identify patients with persistent rejection in need of biopsy would be valuable. Methods This was a post hoc analysis of a multicenter observational study. Patients who had a biopsy-proven acute rejection, had another biopsy within 9 months (270 days), and had a biopsy-paired biomarker sample were included. Results A total of 64 index rejections in 58 patients with repeat biopsies were identified with a median time to repeat biopsy of 100 days. Persistent rejection was present in 61%; 69% of follow-up biopsies were performed in clinically stable patients. Peripheral blood gene expression profile (GEP) demonstrated a sensitivity of 59%, specificity of 76%, positive predictive value (PPV) of 79%, and negative predictive value of 54%. Donor-derived cellfree DNA (dd-cfDNA) demonstrated a sensitivity of 62%, specificity of 86%, PPV of 88%, and negative predictive value of 56%. For repeat biopsies within 90 days of rejection in clinically stable patients (63% of repeat biopsies), both GEP and dd-cfDNA had specificities and PPVs of 100%. GEP was more likely to be positive in T-cell–mediated rejection while dd-cfDNA was more likely to be positive in antibody-mediated rejection. Conclusions Both GEP and dd-cfDNA may have utility in identifying clinically stable patients with persistent rejection in need of biopsy; however, they identify different types of rejections. Clinical Trial registration number: NCT01531257.