Forecasting of Patient-Specific Kidney Transplant Function With a Sequence-to-Sequence Deep Learning Model

A machine learning approach towards assessing consistency and reproducibility: an application to graft survival across three kidney transplantation eras

Background

In South Africa, between 1966 and 2014, there were three kidney transplant eras defined by evolving access to certain immunosuppressive therapies defined as Pre-CYA (before availability of cyclosporine), CYA (when cyclosporine became available), and New-Gen (availability of tacrolimus and mycophenolic acid). As such, factors influencing kidney graft failure may vary across these eras. Therefore, evaluating the consistency and reproducibility of models developed to study these variations using machine learning (ML) algorithms could enhance our understanding of post-transplant graft survival dynamics across these three eras.

Methods

This study explored the effectiveness of nine ML algorithms in predicting 10-year graft survival across the three eras. We developed and internally validated these algorithms using data spanning the specified eras. The predictive performance of these algorithms was assessed using the area under the curve (AUC) of the receiver operating characteristics curve (ROC), supported by other evaluation metrics. We employed local interpretable model-agnostic explanations to provide detailed interpretations of individual model predictions and used permutation importance to assess global feature importance across each era.

Results

Overall, the proportion of graft failure decreased from 41.5% in the Pre-CYA era to 15.1% in the New-Gen era. Our best-performing model across the three eras demonstrated high predictive accuracy. Notably, the ensemble models, particularly the Extra Trees model, emerged as standout performers, consistently achieving high AUC scores of 0.95, 0.95, and 0.97 across the eras. This indicates that the models achieved high consistency and reproducibility in predicting graft survival outcomes. Among the features evaluated, recipient age and donor age were the only features consistently influencing graft failure throughout these eras, while features such as glomerular filtration rate and recipient ethnicity showed high importance in specific eras, resulting in relatively poor historical transportability of the best model.

Conclusions

Our study emphasises the significance of analysing post-kidney transplant outcomes and identifying era-specific factors mitigating graft failure. The proposed framework can serve as a foundation for future research and assist physicians in identifying patients at risk of graft failure.

An observational cohort study examined the change point of kidney function stabilization in the initial period after transplantation

Baseline kidney function following kidney transplantation is often used in research and clinical decision-making yet is not well defined. Here, a method to determine baseline function was proposed and validated on three single-center retrospective cohorts consisting of 922 patients from Belgium (main cohort) and two validation cohorts of 987 patients from the Netherlands and 519 patients from Germany. For each transplant, a segmented regression model was fitted on the estimated glomerular filtration rate (eGFR) evolution during the first-year post-transplantation. This yielded estimates for change point timing, rate of eGFR change before and after change point and eGFR value at change point, now considered the "baseline function". Associations of eGFR evolution with recipient/donor characteristics and the graft failure rate were assessed with linear regression and Cox regression respectively. The change point occurred on average at an eGFR value of 43.7±14.6 mL/min/1.73m2, at a median time of 6.5 days post-transplantation. Despite significant associations with several baseline donor-recipient characteristics (particularly, donor type; living vs deceased), the predictive value of these characteristics for eGFR value and timing of the change point was limited. This followed from a large heterogeneity within eGFR trajectories, which in turn indicated that favorable levels of kidney function could be reached despite a suboptimal initial evolution. Segmented regression consistently provided a good fit to early eGFR evolution, and its estimate of the change point can be a useful reference value in future analyses. Thus, our study shows that baseline kidney function after transplantation is heterogeneous and partly related to pretransplant donor characteristics.

Stratification of Kidney Transplant Recipients Into Five Subgroups Based on Temporal Disease Trajectories

Background

Kidney transplantation is the treatment of choice for patients with end-stage renal disease. Considerable clinical research has focused on improving graft survival and an increasing number of kidney recipients die with a functioning graft. There is a need to improve patient survival and to better understand the individualized risk of comorbidities and complications. Here, we developed a method to stratify recipients into similar subgroups based on previous comorbidities and subsequently identify complications and for a subpopulation, laboratory test values associated with survival.

Methods

First, we identified significant disease patterns based on all hospital diagnoses from the Danish National Patient Registry for 5752 kidney transplant recipients from 1977 to 2018. Using hierarchical clustering, these longitudinal patterns of diseases segregate into 3 main clusters of glomerulonephritis, hypertension, and diabetes. As some recipients are diagnosed with diseases from >1 cluster, recipients are further stratified into 5 more fine-grained trajectory subgroups for which survival, stratified complication patterns as well as laboratory test values are analyzed.

Results

The study replicated known associations indicating that diabetes and low levels of albumin are associated with worse survival when investigating all recipients. However, stratification of recipients by trajectory subgroup showed additional associations. For recipients with glomerulonephritis, higher levels of basophils are significantly associated with poor survival, and these patients are more often diagnosed with bacterial infections. Additional associations were also found.

Conclusions

This study demonstrates that disease trajectories can confirm known comorbidities and furthermore stratify kidney transplant recipients into clinical subgroups in which we can characterize stratified risk factors. We hope to motivate future studies to stratify recipients into more fine-grained, homogenous subgroups to better discover associations relevant for the individual patient and thereby enable more personalized disease-management and improve long-term outcomes and survival.

Non-Traditional Non-Immunological Risk Factors for Kidney Allograft Loss-Opinion

Rates of late allograft loss have improved slowly in the last decades. Well described traditional risk factors that influence allograft survival include cardiovascular events, rejection, infections and post-transplant neoplasia. Here, we critically evaluate the influence of several non-immunological, non-traditional risk factors and describe their impact on allograft survival and cardiovascular health of kidney transplant recipients. We assessed the following risk factors: arterial stiffness, persistent arteriovenous access, mineral bone disease, immunosuppressive drugs residual levels variability, hypomagnesemia, glomerular pathological alterations not included in Banff criteria, persistent inflammation and metabolic acidosis.

When performance is not enough-A multidisciplinary view on clinical decision support

Scientific publications about the application of machine learning models in healthcare often focus on improving performance metrics. However, beyond often short-lived improvements, many additional aspects need to be taken into consideration to make sustainable progress. What does it take to implement a clinical decision support system, what makes it usable for the domain experts, and what brings it eventually into practical usage? So far, there has been little research to answer these questions. This work presents a multidisciplinary view of machine learning in medical decision support systems and covers information technology, medical, as well as ethical aspects. The target audience is computer scientists, who plan to do research in a clinical context. The paper starts from a relatively straightforward risk prediction system in the subspecialty nephrology that was evaluated on historic patient data both intrinsically and based on a reader study with medical doctors. Although the results were quite promising, the focus of this article is not on the model itself or potential performance improvements. Instead, we want to let other researchers participate in the lessons we have learned and the insights we have gained when implementing and evaluating our system in a clinical setting within a highly interdisciplinary pilot project in the cooperation of computer scientists, medical doctors, ethicists, and legal experts.

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.