Outcomes in pediatric liver transplantation (pLT) are not limited by the donated organ supply. Kids are dyingwaiting for organs even when these deaths are completely preventable through proper organ selection. Instead ofdying, children can live a full and active lifetime with a properly selected liver graft for transplant. Critical toachieving zero waitlist mortality and long-term transplant benefit is the capacity to intervene in a timely mannerwith a suitable organ and graft type. Decisions to proceed with pLT are complicated, ultimately based on thealignment of transplant team experience, clinical assessment, and organ availability. In an era of organ shortages,the use of technical variant (TV) grafts, including split liver transplantation and living donor liver transplant, hasthe potential to expand graft choice and enable timelier surgical intervention. Most transplant programs that haveprior experience with TV grafts have low patient mortality and excellent transplant outcomes. However, sometransplant programs that have limited prior experience with TV grafts have reported many poor outcomes forpatients receiving TV transplants. Despite improvements in overall outcomes, national registry data haveconfirmed significant variation among transplant centers in waitlist mortality, TV graft use, and post-transplantoutcomes. Integrally linked to this variation is the intricacy of transplant decision making. Collectively, donorand graft acceptance, prioritization of candidates, and allocation policies depict a complex scenario. More than100 variables can be considered in a single donor-recipient ""best matching'' decision, with a risk of subjectivityand mismatch because of human limitations that should not be underestimated. Recognizing these limitations,artificial intelligence classifiers, including machine learning and deep learning, have been recognized for theirpotential to support or confirm decision making within the field of transplantation. Still, overall data-drivensupport for optimal graft selection and dissemination of graft decision support is lacking. Opportunities for, andthe impact of, discovery are high. This project will result in a composite decision support software tool that usesmachine learning to predict and model the best survival for the patient using pre-transplant mortality, post-transplant outcomes, and prior center experience. The decision support tool can be established to supplementcurrent graft selection practices in pLT. We anticipate that modeling based on composite measures willdemonstrate equivalent outcomes in recipients of TV grafts. We will develop an algorithm for optimal pediatricgraft-type selection that will be commercialized for use through the Starzl Network for Excellence in PediatricTransplantation and after further multi-center validation it will be available for all pediatric transplant programs.We will accomplish our objective through the following three aims. One, determine the optimal feature space forpredictive variables for patient and pLT graft survival. Two, develop survival prediction models, "PSELECT," forremaining on the waitlist or receiving various graft types. Three, demonstrate the simulated technical feasibilityto eliminate the waitlist mortality based on the PSELECT performance on previously held-out data.
Public Health Relevance Statement: PROJECT NARRATIVE
Kids are needlessly dying waiting for an organ transplant because donor organ selection is difficult, and this is
not due to an organ shortage. Optimal organ selection and waitlist management strategies have not been fully
developed or disseminated, contributing to observed variability in waitlist mortality which varies from 0-2
deaths per 100 waitlist years to >20 deaths per 100 waitlist years at individual centers. The proposed project
solves this problem and ends waitlist mortality by development of novel machine learning algorithms and real-
time clinical decision support software to determine the optimal transplant graft for each child based upon
predicted survival.
Project Terms: <21+ years old><0-11 years old>
