Validation of biomarkers predicting clinical outcomes of umbilical cord blood transplantation

Patients with blood disorders, can be treated by stem cell transplants from a third party. Finding an adult match for such patients is not always possible, but alternatives such as umbilical cord blood (UCB) can be used for transplantation. UCB are readily available and stored in frozen cell banks around the world. However, UCB transplants show delayed and, sometimes, insufficient engraftment of the patient's haematopoietic system. In this study, we will investigate characteristics of UCB from different donors to find biomarkers, which are associated with better engraftment in a pre-clinical animal model and in patients who received UCB transplants. We will employ cutting-edge cellular and molecular biology analyses and implement in-depth artificial intelligence (machine learning) methodology in order to find the biomarkers and develop robust test for selection of UCB units, which would work best in patients and reduce the number of failed transplants.

Haematopoietic cell transplant (HCT) is a life-saving therapy for patients with haematological disorders. Umbilical cord blood (UCB) is a clinically accepted source of therapeutic cells and is a vital resource when HLA matched unrelated adult donors are in limited supply, which particularly affects some ethnic minority patients. Some UCB, however, show delayed haematopoietic reconstitution, which needs to be addressed to fully unlock this valuable resource. Our xenograft transplantation analysis revealed novel UCB traits that can serve as complementary selection criteria to identify UCBs with superior haematopoietic reconstitution capacity. Based on this, we aim to create a robust biomarker-based test system for identification of optimal UCB for clinical transplants. We will use two major complementary approaches: firstly, cellular and molecular traits of individual UCB units will be profiled and their reconstitution capacity monitored after xeno-transplantation into NSG mice, tracking production of clinically relevant human blood/immune cells. Secondly, we will analyse archived clinical samples and retrospective clinical data from patients that received UCB transplants in European and US centres. In both approaches, correlations between UCB properties and their transplantation outcomes, will be interrogated by bioinformatics and machine-learning methodologies. Our main deliverables will be 1) to generate data-rich computational model (classifier) linking molecular/cellular UCB traits with key features of hematopoietic reconstitution; which in turn will lead to 2) creation of a robust biomarker test system for selection of optimal UCBs. This classification system will increase success of UCB transplants and boost clinical uptake of UCBs, with great clinical and financial impacts. As a multi-national consortium of leading experts in haematopoietic stem cells, cord banking and clinical transplants, our team is in a strong position to deliver these goals.