Systemic lupus erythematosus (SLE) is a heterogeneous, autoimmune disease with highly variable morbidity and mortality. Disease heterogeneity is a major challenge to both the clinical care of SLE patients and successful clinical trials. With heterogeneous diseases, identifying clusters results in improved classification, biomarkers, and targeted therapies. Electronic health records (EHRs) represent a powerful tool to identify clusters and risk-stratify SLE patients. We have published an algorithm that accurately identifies SLE patients in the EHR and have assembled a cohort of 2,376 SLE patients with a mean follow-up of 9 years. This data is linked to one of the world's largest biobanks, BioVU, with 400 SLE patients already genotyped. The EHR and BioVU allow for novel methods such as phenome-wide association studies (PheWAS) that use billing codes for a comprehensive scan of the entire EHR. We have performed the first PheWAS in SLE. Our overall goal is to use readily available EHR data in an intelligent way to improve outcomes in SLE patients. We hypothesize that SLE is composed of multiple clusters of patients with different disease courses and comorbidities, and our EHR-based methodology that incorporates genetic information will serve as novel tools to risk-stratify SLE patients. Using PheWAS in Aim 1, we will uncover differences in comorbidities between SLE patients with and without autoantibodies and with and without pre-specified SLE susceptibility single nucleotide polymorphisms (SNPs).
In Aim 2, we will perform clustering analyses using demographics, autoantibodies, comorbidities, and SLE SNPs to risk-stratify SLE patients. We will assess renal outcomes, survival, and treatments received among the clusters.
In Aim 3, we will evaluate treatment response to induction therapy for SLE nephritis in the EHR and compare to published outcomes.
These aims are the necessary first steps to risk-stratify SLE patients and define treatment response in the EHR to then build models to predict treatment response and conduct EHR-based pragmatic clinical trials of targeted therapies. Additional mentored training and didactic coursework in genetics, biomedical informatics, and biostatistics will advance Dr. Barnado's career. Vanderbilt serves as an exceptional environment to support Dr. Barnado's transition to an independent physician scientist. Notable strengths include the Synthetic Derivative (SD), a de-identified EHR with over 2.7 million subjects, and BioVU, a genetic biobank linked to the SD. The Department of Medicine and Division of Rheumatology are in support of Dr. Barnado's career. Her mentors, Drs. Crofford and Denny, are internationally recognized in rheumatology and biomedical informatics with successful track records of mentoring. With Vanderbilt's institutional commitment to young investigators and expertise in biomedical informatics, Dr. Barnado's will successfully leverage her innovative proposal to independent R01 funding.
Systemic lupus erythematosus (SLE) is a heterogeneous disease with a highly variable disease course that causes major challenges to both the clinical care of SLE patients and successful clinical trials. The electronic health record (EHR) represents a power tool to risk-stratify SLE patients by using clinical data in real time to determine which patients are at risk of developing specific disease complications. This proposal focuses on developing novel, EHR-based methods to risk-stratify SLE patients and to define treatment response in the EHR.
|Barnado, A; Carroll, R J; Casey, C et al. (2018) Phenome-wide association study identifies dsDNA as a driver of major organ involvement in systemic lupus erythematosus. Lupus :961203318815577|