A presentation at the 2015 annual European Congress of Rheumatology meeting emphasized how creating electronic medical records (EMRs) for each patient and amassing “big data” derived from those records can help achieve personalized medicine.
Analysis of “big data” derived from EMRs can improve the care of individual patients, enable the discovery of new genotype-phenotype and phenotype-genotype associations, identify specific subsets of patients, be used for drug development or repurposing of drugs, and help implement genomic medicine, explained Joshua C. Denny, MD, of Vanderbilt University in Nashville, TN.1
In his talk, Denny demonstrated how EMRs linked to DNA repositories are being used for genetic research on many diseases (including rheumatoid arthritis [RA]) at Vanderbilt and other centers in the United States as well as around the world.
“This model allows the Vanderbilt investigators to assess between 20,000 and 30,000 samples per year from more than 200,000 people. This is one of the largest biorepositories in the world,” Denny said.
The first step is to develop an EMR for each patient that contains notes, text reports, clinical messaging, and laboratory and diagnostic data. The individual’s name is removed from the data. A very large DNA repository of more than 200,000 specimens is linked to the data from the EMR.
“The EMR is a complete record of disease, when the patient gets the disease, what drugs he or she takes, and what the responses are. We have used these data to investigate subtypes of seropositive and seronegative rheumatoid arthritis. You can only begin to investigate subtypes once you have large data sets,” he explained.
Trying to identify phenotypes from the EMR involves algorithm development and implementation. “You develop algorithms to refine the data,” Denny explained.
The investigators looked at a number of different disease states, including RA, and the association with markers and genes and genetic regions.
Vanderbilt is the coordinating center for the EMERGE network, a consortium of biorepositories linked to EMRs for conducting genomic studies. This effort includes 8 adult and 2 pediatric sites in the United States. The consortium is performing genome-wide association studies (GWAS) using EMR-derived phenotypes. The goal is to incorporate the identification of actionable genetic variants in the EMR, which can help determine if there is an available targeted treatment for a specific gene. They are looking at endophenotypes, pharmacogenomic phenotypes, and many disease states, including RA, and the association with markers, genes, and genetic regions.
The studies of RA phenotypes comprised more than 100,000 cases and controls from RA studies and biobanks that include multiple ethnicities.2 The researchers identified 101 RA loci and 98 candidate genes. Then they encoded known targets for RA drugs and drugs used in other diseases—two-thirds of the targets displayed associations for diseases beyond RA. In this way, the genetics of RA contribute to biology and drug discovery, Denny said.
The investigators analyzed disease risk and comorbidities associated with RA subtypes. They identified 2 phenotypes: seronegative RA, which is associated with fibromyalgia, and seropositive RA, which is more often found in cigarette smokers who are also more likely to have emphysema.
EMERGE now includes more than 400,000 DNA samples. Other centers in different parts of the world are also conducting GWAS and creating biobanks; for example, China’s has 500,000 samples and Denmark’s has 5,600,000 samples.
- Denny J. Electronic record keeping—use of big data. Presented at: 16th Annual European Congress of Rheumatology; June 10-13, 2015; Rome, Italy.
- Okada Y, Wu D, Trynka G, et al. Genetics of rheumatoid arthritis contributes to biology and drug discovery. Nature. 2014;506(7488):376-381.