Personalized prediction of delayed graft function for recipients of deceased donor kidney transplants with machine learning

First experiences with machine learning predictions of accelerated declining eGFR slope of living kidney donors 3 years after donation

BACKGROUND

Living kidney donors are screened pre-donation to estimate the risk of end-stage kidney disease (ESKD). We evaluate Machine Learning (ML) to predict the progression of kidney function deterioration over time using the estimated GFR (eGFR) slope as the target variable.

METHODS

We included 238 living kidney donors who underwent donor nephrectomy. We divided the dataset based on the eGFR slope in the third follow-up year, resulting in 185 donors with an average eGFR slope and 53 donors with an accelerated declining eGFR-slope. We trained three Machine Learning-models (Random Forest [RF], Extreme Gradient Boosting [XG], Support Vector Machine [SVM]) and Logistic Regression (LR) for predictions. Predefined data subsets served for training to explore whether parameters of an ESKD risk score alone suffice or additional clinical and time-zero biopsy parameters enhance predictions. Machine learning-driven feature selection identified the best predictive parameters.

RESULTS

None of the four models classified the eGFR slope with an AUC greater than 0.6 or an F1 score surpassing 0.41 despite training on different data subsets. Following machine learning-driven feature selection and subsequent retraining on these selected features, random forest and extreme gradient boosting outperformed other models, achieving an AUC of 0.66 and an F1 score of 0.44. After feature selection, two predictive donor attributes consistently appeared in all models: smoking-related features and glomerulitis of the Banff Lesion Score.

CONCLUSIONS

Training machine learning-models with distinct predefined data subsets yielded unsatisfactory results. However, the efficacy of random forest and extreme gradient boosting improved when trained exclusively with machine learning-driven selected features, suggesting that the quality, rather than the quantity, of features is crucial for machine learning-model performance. This study offers insights into the application of emerging machine learning-techniques for the screening of living kidney donors.

Cherry on Top or Real Need? A Review of Explainable Machine Learning in Kidney Transplantation

Research on solid organ transplantation has taken advantage of the substantial acquisition of medical data and the use of artificial intelligence (AI) and machine learning (ML) to answer diagnostic, prognostic, and therapeutic questions for many years. Nevertheless, despite the question of whether AI models add value to traditional modeling approaches, such as regression models, their "black box" nature is one of the factors that have hindered the translation from research to clinical practice. Several techniques that make such models understandable to humans were developed with the promise of increasing transparency in the support of medical decision-making. These techniques should help AI to close the gap between theory and practice by yielding trust in the model by doctors and patients, allowing model auditing, and facilitating compliance with emergent AI regulations. But is this also happening in the field of kidney transplantation? This review reports the use and explanation of "black box" models to diagnose and predict kidney allograft rejection, delayed graft function, graft failure, and other related outcomes after kidney transplantation. In particular, we emphasize the discussion on the need (or not) to explain ML models for biological discovery and clinical implementation in kidney transplantation. We also discuss promising future research paths for these computational tools.

A Machine Learning Prediction Model for Immediate Graft Function After Deceased Donor Kidney Transplantation

BACKGROUND

After kidney transplantation (KTx), the graft can evolve from excellent immediate graft function (IGF) to total absence of function requiring dialysis. Recipients with IGF do not seem to benefit from using machine perfusion, an expensive procedure, in the long term when compared with cold storage. This study proposes to develop a prediction model for IGF in KTx deceased donor patients using machine learning algorithms.

METHODS

Unsensitized recipients who received their first KTx deceased donor between January 1, 2010, and December 31, 2019, were classified according to the conduct of renal function after transplantation. Variables related to the donor, recipient, kidney preservation, and immunology were used. The patients were randomly divided into 2 groups: 70% were assigned to the training and 30% to the test group. Popular machine learning algorithms were used: eXtreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine, Gradient Boosting classifier, Logistic Regression, CatBoost classifier, AdaBoost classifier, and Random Forest classifier. Comparative performance analysis on the test dataset was performed using the results of the AUC values, sensitivity, specificity, positive predictive value, negative predictive value, and F1 score.

RESULTS

Of the 859 patients, 21.7% (n = 186) had IGF. The best predictive performance resulted from the eXtreme Gradient Boosting model (AUC, 0.78; 95% CI, 0.71-0.84; sensitivity, 0.64; specificity, 0.78). Five variables with the highest predictive value were identified.

CONCLUSIONS

Our results indicated the possibility of creating a model for the prediction of IGF, enhancing the selection of patients who would benefit from an expensive treatment, as in the case of machine perfusion preservation.

Toward generalizing the use of artificial intelligence in nephrology and kidney transplantation

With its robust ability to integrate and learn from large sets of clinical data, artificial intelligence (AI) can now play a role in diagnosis, clinical decision making, and personalized medicine. It is probably the natural progression of traditional statistical techniques. Currently, there are many unmet needs in nephrology and, more particularly, in the kidney transplantation (KT) field. The complexity and increase in the amount of data, and the multitude of nephrology registries worldwide have enabled the explosive use of AI within the field. Nephrologists in many countries are already at the center of experiments and advances in this cutting-edge technology and our aim is to generalize the use of AI among nephrologists worldwide. In this paper, we provide an overview of AI from a medical perspective. We cover the core concepts of AI relevant to the practicing nephrologist in a consistent and simple way to help them get started, and we discuss the technical challenges. Finally, we focus on the KT field: the unmet needs and the potential role that AI can play to fill these gaps, then we summarize the published KT-related studies, including predictive factors used in each study, which will allow researchers to quickly focus on the most relevant issues.

Postoperative day 1 serum cystatin C level predicts postoperative delayed graft function after kidney transplantation

Background

Delayed graft function (DGF) commonly occurs after kidney transplantation, but no clinical predictors for guiding post-transplant management are available.

Materials and methods

Data including demographics, surgery, anesthesia, postoperative day 1 serum cystatin C (S-CysC) level, kidney functions, and postoperative complications in 603 kidney transplant recipients who met the enrollment criteria from January 2017 to December 2018 were collected and analyzed to form the Intention-To-Treat (ITT) set. All perioperative data were screened using the least absolute shrinkage and selection operator. The discrimination, calibration, and clinical effectiveness of the predictor were verified with area under curve (AUC), calibration plot, clinical decision curve, and impact curve. The predictor was trained in Per-Protocol set, validated in the ITT set, and its stability was further tested in the bootstrap resample data.

Result

Patients with DGF had significantly higher postoperative day 1 S-CysC level (4.2 ± 1.2 vs. 2.8 ± 0.9 mg/L; P < 0.001), serum creatinine level (821.1 ± 301.7 vs. 554.3 ± 223.2 μmol/L; P < 0.001) and dialysis postoperative (74 [82.2%] vs. 25 [5.9%]; P < 0.001) compared with patients without DGF. Among 41 potential predictors, S-CysC was the most effective in the parsimonious model, and its diagnostic cut-off value was 3.80 mg/L with the risk score (OR, 13.45; 95% CI, 8.02–22.57; P < 0.001). Its specificity and sensitivity indicated by AUC was 0.832 (95% CI, 0.779–0.884; P < 0.001) with well fit calibration. S-CysC yielded up to 50% of clinical benefit rate with 1:4 of cost/benefit ratio.

Conclusion

The postoperative day 1 S-CysC level predicts DGF and may be used as a predictor of DGF but warrants further study.

Predicting the Kidney Graft Survival Using Optimized African Buffalo-Based Artificial Neural Network

A variety of receptor and donor characteristics influence long-and short-term kidney graft survival. It is critical to predict the effectiveness of kidney transplantation to optimise organ allocation. This would allow patients to choose the best accessible kidney donor and the optimal immunosuppressive medication. Several studies have attempted to identify factors that predispose to graft rejection, but the results have been contradictory. As a result, the goal of this paper is to use the African buffalo-based artificial neural network (AB-ANN) approach to uncover predictive risk variables related to kidney graft. These two feature selection approaches combine to provide a novel hybrid feature selection technique that could select the most important elements to improve prediction accuracy. The feature analysis revealed that clinical features have varied effects on transplant survival. The collected data is processed in both training and testing methods. The prediction model's performance, in terms of accuracy, precision, recall, and F-measure, was examined, and the results were compared with those of other existing systems, including naive Bayesian, random forest, and J48 classifier. The results suggest that the proposed approach can forecast graft survival in kidney recipients' next visits in a creative manner and with more accuracy compared with other classifiers. This proposed method is more efficient for predicting kidney graft survival. Incorporating those clinical tools into outpatient clinics' everyday workflows could help physicians make better and more personalised decisions.

Artificial Intelligence-A Tool for Risk Assessment of Delayed-Graft Function in Kidney Transplant

Delayed-graft function (DGF) might be responsible for shorter graft survival. Therefore, a clinical tool predicting its occurrence is vital for the risk assessment of transplant outcomes. In a single-center study, we conducted data mining and machine learning experiments, resulting in DGF predictive models based on random forest classifiers (RF) and an artificial neural network called multi-layer perceptron (MLP). All designed models had four common input parameters, determining the best accuracy and discriminant ability: donor’s eGFR, recipient’s BMI, donor’s BMI, and recipient–donor weight difference. RF and MLP designs, using these parameters, achieved an accuracy of 84.38% and an area under curve (AUC) 0.84. The model additionally implementing a donor’s age, gender, and Kidney Donor Profile Index (KDPI) accomplished an accuracy of 93.75% and an AUC of 0.91. The other configuration with the estimated post-transplant survival (EPTS) and the kidney donor risk profile (KDRI) achieved an accuracy of 93.75% and an AUC of 0.92. Using machine learning, we were able to assess the risk of DGF in recipients after kidney transplant from a deceased donor. Our solution is scalable and can be improved during subsequent transplants. Based on the new data, the models can achieve better outcomes.

Standardized Reporting of Machine Learning Applications in Urology: The STREAM-URO Framework

The Standardized Reporting of Machine Learning Applications in Urology (STREAM-URO) framework was developed to provide a set of recommendations to help standardize how machine learning studies in urology are reported. This framework serves three purposes: (1) to promote high-quality studies and streamline the peer review process; (2) to enhance reproducibility, comparability, and interpretability of results; and (3) to improve engagement and literacy of machine learning within the urological community.

Technology-Enabled Care and Artificial Intelligence in Kidney Transplantation

Purpose of Review

Artificial intelligence (AI), machine learning, and technology-enabled remote patient care have evolved rapidly and have now been incorporated into many aspects of medical care. Transplantation is fortunate to have large data sets upon which machine learning algorithms can be constructed. AI are now available to improve pretransplant management, donor selection, and post-operative management of transplant patients.

Recent Findings

Changes in patient and donor characteristics warrant new approaches to listing and organ acceptance practices. Machine learning has been employed to optimize donor selection to identify patients likely to benefit from transplantation of higher risk organs, increasing organ discard and reducing waitlist mortality. These models have greater precisions and predictive ability than currently employed metrics including the Kidney Donor Profile Index and the expected posttransplant survival models. After transplant, AI tools have been developed to optimize immunosuppression management, track patients adherence, and assess graft survival.

Summary

AI and technology-enabled management tools are now available throughout the transplant journey. Unfortunately, those are frequently not available at the point of decision (patient listing, organ acceptance, posttransplant clinic), limiting utilization. Incorporation of these tools into the EMR, the Donor Net® organ offer system, and mobile devices is vital to ensure widespread adoption.

Artificial intelligence and kidney transplantation

Artificial intelligence and its primary subfield, machine learning, have started to gain widespread use in medicine, including the field of kidney transplantation. We made a review of the literature that used artificial intelligence techniques in kidney transplantation. We located six main areas of kidney transplantation that artificial intelligence studies are focused on: Radiological evaluation of the allograft, pathological evaluation including molecular evaluation of the tissue, prediction of graft survival, optimizing the dose of immunosuppression, diagnosis of rejection, and prediction of early graft function. Machine learning techniques provide increased automation leading to faster evaluation and standardization, and show better performance compared to traditional statistical analysis. Artificial intelligence leads to improved computer-aided diagnostics and quantifiable personalized predictions that will improve personalized patient care.

Perioperative Normal Saline Administration and Delayed Graft Function in Patients Undergoing Kidney Transplantation: A Retrospective Cohort Study

BACKGROUND

Perioperative normal saline administration remains common practice during kidney transplantation. The authors hypothesized that the proportion of balanced crystalloids versus normal saline administered during the perioperative period would be associated with the likelihood of delayed graft function.

METHODS

The authors linked outcome data from a national transplant registry with institutional anesthesia records from 2005 to 2015. The cohort included adult living and deceased donor transplants, and recipients with or without need for dialysis before transplant. The primary exposure was the percent normal saline of the total amount of crystalloids administered perioperatively, categorized into a low (less than or equal to 30%), intermediate (greater than 30% but less than 80%), and high normal saline group (greater than or equal to 80%). The primary outcome was the incidence of delayed graft function, defined as the need for dialysis within 1 week of transplant. The authors adjusted for the following potential confounders and covariates: transplant year, total crystalloid volume, surgical duration, vasopressor infusions, and erythrocyte transfusions; recipient sex, age, body mass index, race, number of human leukocyte antigen mismatches, and dialysis vintage; and donor type, age, and sex.

RESULTS

The authors analyzed 2,515 records. The incidence of delayed graft function in the low, intermediate, and high normal saline group was 15.8% (61/385), 17.5% (113/646), and 21% (311/1,484), respectively. The adjusted odds ratio (95% CI) for delayed graft function was 1.24 (0.85 to 1.81) for the intermediate and 1.55 (1.09 to 2.19) for the high normal saline group compared with the low normal saline group. For deceased donor transplants, delayed graft function in the low, intermediate, and high normal saline group was 24% (54/225 [reference]), 28.6% (99/346; adjusted odds ratio, 1.28 [0.85 to 1.93]), and 30.8% (277/901; adjusted odds ratio, 1.52 [1.05 to 2.21]); and for living donor transplants, 4.4% (7/160 [reference]), 4.7% (14/300; adjusted odds ratio, 1.15 [0.42 to 3.10]), and 5.8% (34/583; adjusted odds ratio, 1.66 [0.65 to 4.25]), respectively.

CONCLUSIONS

High percent normal saline administration is associated with delayed graft function in kidney transplant recipients.

EDITOR’S PERSPECTIVE

Toward Advancing Long-Term Outcomes of Kidney Transplantation with Artificial Intelligence

After decades of pioneering advances and improvements, kidney transplantation is now the renal replacement therapy of choice for most patients with end-stage kidney disease (ESKD). Despite this success, the high risk of premature death and frequent occurrence of graft failure remain important clinical and research challenges. The current burst of studies and other innovative initiatives using artificial intelligence (AI) for a wide range of analytical and practical applications in biomedical areas seems to correlate with the same trend observed in publications in the kidney transplantation field, and points toward the potential of such novel approaches to address the aforementioned aim of improving long-term outcomes of kidney transplant recipients (KTR). However, at the same time, this trend underscores now more than ever the old methodological challenges and potential threats that the research and clinical community needs to be aware of and actively look after with regard to AI-driven evidence. The purpose of this narrative mini-review is to explore challenges for obtaining applicable and adequate kidney transplant data for analyses using AI techniques to develop prediction models, and to propose next steps in the field. We make a call to act toward establishing the strong collaborations needed to bring innovative synergies further augmented by AI, which have the potential to impact the long-term care of KTR. We encourage researchers and clinicians to submit their invaluable research, including original clinical and imaging studies, database studies from registries, meta-analyses, and AI research in the kidney transplantation field.