A Translational Bioinformatics Approach for Phenotyping and Genetic Risk of Severe Cutaneous Adverse Drug Reactions - PROJECT SUMMARY Drug reaction with eosinophilia and systemic symptoms (DRESS) is a life-threatening severe cutaneous adverse drug reaction (SCAR). Vancomycin is the first-line treatment for multidrug-resistant gram-positive bacteria and is the most common drug associated with DRESS in the United States (US). Vancomycin DRESS is strongly associated with HLA-A*32:01, and approximately 20% of patients carrying HLA-A*32:01 develop DRESS after more than 2 weeks of vancomycin exposure. This underscores a vital knowledge gap that exists for all HLA- associated SCAR: what co-morbidity, genetic, and other multi-omic risk factors that interact with and without HLA confer susceptibility or resistance. By leveraging electronic health record (EHR) linked DNA biobanks, this project will discover co-morbidity, genetic, multi-omic, and other risk factors for vancomycin DRESS through an innovative translational bioinformatics approach to inform mechanisms of drug hypersensitivity and tolerance. In this K08 proposal, I will undertake critical training for my long-term career goal of bringing precision medicine to SCAR and other allergic/immunologic conditions by developing and applying expertise in allergy/immunology, immunogenomics, and translational bioinformatics. The project will be completed under the guidance of my primary mentor Dr. Elizabeth Phillips, and a Research Advisory Committee of experts in bioinformatics and immunogenomics. By integrating targeted immunogenomic typing, including HLA, KIR, and ERAP and GWAS- level genotyping, I will discover epistatic interactions of HLA/KIR/ERAP and multi-omic variation in vancomycin DRESS, which has the potential to apply to other allergic and immune-mediated diseases such as cancer, autoimmune and infectious diseases. My mentored research project has two specific aims: Aim 1: Develop EHR phenotyping algorithms to identify DRESS cases and utilize machine learning on routine laboratory values to develop predictive models for earlier diagnosis. Aim 2A: Confirm the association of HLA-A*32:01 and vancomycin DRESS in a large prospective cohort; discover additional genetic variation outside of HLA through GWAS and targeted immunogenomic typing; and, importantly, assess the generalizability of the HLA-A*32:01 association to diverse populations in other EHR biobanks. Aim 2B: Perform multi-omic analyses and causal discovery methods to test the hypothesis that genetically regulated transcriptomic, proteomic, and metabolic traits vary between vancomycin DRESS cases and vancomycin-tolerant controls. Through the skills and mechanistic insights gained by this career development award, I will develop a research niche to design and implement future studies on SCAR and other allergic/immunologic diseases that leverage translational bioinformatics in diverse EHR biobanks, such as the NIH’s All of US cohort, eMERGE Network and others across the US and globally. Through this career development award, I will gain the foundation to independently lead a multi-disciplinary research team in future R01 proposals to bring precision medicine approaches to SCAR that will translate to strategies for prevention, earlier diagnosis, and targeted therapies.
