Development of Genomic Risk Prediction Models for T-cell Malignancies

About This Grant

PROJECT SUMMARY/ABSTRACT Relapsed acute lymphoblastic leukemia (ALL) and lymphoblastic lymphoma remain leading causes of cancer- related death in children. For many pediatric cancers, risk classification using clinical and genomic features has reduced risk of relapse by identifying high-risk patients and allocating intensified therapy, while simultaneously avoiding treatment related mortality and toxicity for children with low-risk disease. However, no risk classifiers exist in clinical practice for T-cell ALL (T-ALL) and T-cell lymphoblastic lymphoma (T-LBL). Clinically useful risk prediction models are urgently needed to prevent more children from relapsing and dying of T-ALL and T-LBL. The overall objective of the proposed research is to improve outcomes in T-ALL/T-LBL by 1) identifying high risk genomic features by genetic ancestry, and b) developing and validating risk prediction models integrating clinical features, early disease response, ancestry informative markers, and genomic features to identify high- and low- risk patients. In Specific Aim 1, the candidate will identify recurrent, prognostic genomic features associated with relapse and death in T-LBL by genetic ancestry and compare this to ancestry-informed prognostic genomic alterations in T-ALL. In Specific Aim 2, the candidate will develop and validate risk prediction models in T-ALL and T-LBL, integrating clinical features, disease response, multi-omics (whole genome sequencing, RNA sequencing), and ancestry informative markers. Notably, DNA-based genetic ancestry is not included in risk allocation for any cancer, despite growing data demonstrating its impact on the prognostic value of somatic alterations. This work will promote a novel approach to improve outcomes by incorporating germline markers of genetic ancestry a priori into risk prediction modeling. The long-term goal of this work is to better understand differences in the biology of T-ALL and T-LBL and develop and validate risk prediction models which can be implemented into phase 3 clinical trials to allocate high-risk patients for intensified/alternative therapies, and low- risk patients for de-intensified treatment. The candidate is a pediatric oncologist committed to harnessing multi-omics and predictive modeling to improve outcomes and reduce toxicity for individuals with leukemia and lymphoma. The proposed training plan and research project will be conducted at the Children’s Hospital of Philadelphia (CHOP) and the University of Pennsylvania (Penn). CHOP and Penn are leaders in multi-omic research will provide an outstanding environment in which to conduct the proposed work. The candidate will be mentored by Dr. David Teachey along with a very strong multidisciplinary advisory committee. The long-term goal of the candidate is to develop a research program devoted to advancing translational multi-omic epidemiologic research and provide the base for her future R01 proposal implementing the risk prediction model in a prospective phase 3 clinical trial. Completion of the proposed research project and the complementary training plan will provide a robust foundation to achieve this goal.