Utilizing Large Electronic Medical Record Data Sets to Identify Novel Drug–Gene Interactions for Commonly Used Drugs

Abstract

Real-world prescribing of drugs differs from the experimental systems, physiologicalpharmacokinetic (PK) models and clinical trials used in drug development and licensing, with drugs often used in patients with multiple comorbidities with resultant polypharmacy. The increasing availability of large biobanks linked to electronic healthcare records enables the potential to identify novel drug-gene interactions in large populations of patients. In this study we used 3 Scottish cohorts and UK Biobank to identify drug-gene interactions for the 50 most commonly used drugs and 162 variants in genes involved in drug pharmacokinetics. We defined two phenotypes based upon prescribing behaviour -drug-stop or dose-decrease. Using this approach, we replicate 11 known drug-gene interactions including, for example, CYP2C9/CYP2C8 variants and sulphonylurea/thiazolidinedione prescribing and ABCB1/ABCG2 variants and statin prescribing. We identify 8 novel associations after bonferroni correction, 3 of which are replicated or validated in the UK biobank or have other supporting results: The C-allele at rs4918758 in CYP2C9 was associated with a 25% (15-44%) lower odds of dose reduction of quinine, p=1.6×10 -5 ; the A-allele at rs9895420 in ABCC3 was associated with a 46% (24-62%) reduction in odds of dose reduction with doxazosin, p=1.2×10 -4 , and altered blood pressure response in the UK Biobank; the CYP2D6*2 variant was associated with a 30% (18 %-40%) reduction in odds of stopping ramipril treatment, p=1.01×10 -5 , with similar results seen for enalapril and lisinopril and with other CYP2D6 variants. This study highlights the scope of using large population bioresources linked to medical record data to explore drug-gene interactions at scale.

Computational drug repurposing based on electronic health records: a scoping review

“…The most common repurposing target was diabetes-related, consisting of 10 out of the 33 publications 17 , 26 , 29 , 32 , 37 – 42 , including type 2 diabetes 37 , 41 , 42 , gestational diabetes 17 , diabetes (unspecified) 26 , 29 , diabetes-related tests including glycated hemoglobin 32 and Fasting Blood Glucose 38 – 40 . Six publications did not focus on any specific diseases 21 , 22 , 27 , 28 , 43 , 44 . For example, Dang et al 21 aimed to establish a generic process and method to integrate phenomic data in EHR with omic and drug data.…”

“…In contrast, others indicated open software, which they used, without their practical implementations 20 , 27 , 45 . Lastly, some of the studies shared the analysis and results in the form of supplements or separated links 19 , 26 , 27 , 31 , 33 , 35 , 36 , 41 , 42 , 44 – 47 , 49 – 52 .…”

CYP2C19 Loss‐of‐Function Variants Associated With Long‐Term Ischemic Stroke Events During Clopidogrel Treatment in the Chinese Population

“…19,20 However, such a new pharmacogenetic paradigm has not been exploited in China, in which the LoF variants are carried by around 60% of this ethnic group. 14,21 Here, we leveraged a biobank linked to a Chinese EMR database to ascertain clopidogrel users in the primary care settings. By retrieving EMR information for each individual, their drug exposure and incidence of ischemic stroke over a maximum of a 3-year follow-up period were examined.…”

“…Several studies of this kind have demonstrated biological resources linked to EMR systems are valuable platforms that confirmed about 25% of the White population carrying CYP2C19 LoF variants indeed had a higher risk of stroke upon clopidogrel treatment 19,20 . However, such a new pharmacogenetic paradigm has not been exploited in China, in which the LoF variants are carried by around 60% of this ethnic group 14,21 …”

Drug–Drug–Gene Interactions in Cardiovascular Medicine

“…For example, Malki et al used three Scottish cohorts and the UK Biobank (with linked electronic healthcare records) to identify novel DGIs for 50 most commonly used drugs and 162 variants in 35 genes involved in drug pharmacokinetics. 185 In addition to an efficacy endpoint (systolic blood pressure reduction), they used two phenotypes based on prescribing behaviour (drug-stop and dose-decrease, which are proxies for altered efficacy or tolerability), which enabled them to replicate 11 known DGIs and identify eight novel ones. On the other hand, McInnes and Altman searched the UK Biobank for DGIs among 200 drugs and 9 genes using maintenance dose (the average milligrams of drug per day for the last five prescriptions of each drug) and differential drug response phenotypes (diagnosis codes in primary care data, eg risk of developing a specific side effect).…”