Preemptive genotyping: More than 90% of patients have actionable result

Published: May 15th, 2014

Category: Stories

Out of approximately 10,000 patients genotyped in the Pharmacogenomic Resource for Enhanced Decisions in Care and Treatment (PREDICT) program at Vanderbilt University Medical Center, 91% had at least one actionable genotype. Of these, over 40% had evidence of exposure to at least one related medication or medication class. A reactive genotype approach would have resulted in 1.7-fold more tests than those conducted in a preemptive model.

Main Points

Although clinical implementation of pharmacogenetics is increasing, challenges exist to widespread adoption. One such challenge is the practical and economic feasibility of incorporating panel-based preemptive genotyping into clinical care. More data are needed regarding the value of a preemptive genotyping approach, as compared with reactive, “one at a time” genotyping to gather single nucleotide polymorphism (SNP) or other genetic data to guide individual drug therapy.

In the current descriptive report from the Vanderbilt University Medical Center’s PREDICT program, authors summarized data for patients who underwent preemptive clinical genotyping. Patient selection criteria for preemptive genotyping included anticipated coronary artery stenting, likelihood of receiving a pharmacogenomic medication (as estimated by prognostic score), or provider preference. Patients were genotyped for five drug-gene interactions that have been implemented clinically: clopidogrel (CYP2C19), simvastatin (SLCO1B1), warfarin (CYP2C9 and VKORC1), thiopurines (TPMT), and tacrolimus (CYP3A5). Investigators defined “actionable” genotypes as those that would prompt clinical decision support to suggest a drug therapy change (complete definitions available in online supplementary materials); medication exposure data was determined through electronic medical records.

Investigators obtained complete genotype data for 9,589 individuals (median age 63 years [55–71]; 59% male) since September 2010, with analyzed genotype frequencies being consistent with published allele frequency data. Of these patients, 91% of all and 96% of African American patients had at least one actionable genotype. A total of 42% of all patients and 23% of African-American patients had an actionable genotype and evidence of exposure to at least one related medication or medication class.

Investigators compared these preemptive data to a theoretical reactive genotyping approach in the same population. Researchers calculated that use of serial single-gene testing in the overall patient cohort would have led to 14,656 tests, a 1.7-fold increase over the number of tests conducted in the preemptive model.

Clinical Implications

This report provides important data in support of the potential clinical value of a predictive, preemptive genotyping model. Investigators noted the importance of a clinical infrastructure to support provider notification and follow-up of actionable pharmacogenetic test results.

Researchers also pointed out potential implications of study findings for stratifying risk based on ancestry. Programs such as this one “have the potential to improve medication outcomes for non-[European-American] patients by personalizing therapeutic doses and medication choices, just as incorporation of genetic ancestry has been shown to increase precision in evaluating pulmonary function,” investigators wrote.

However, a preemptive pharmacogenetic testing approach is not without its critics. In a 2013 Letter to the Editor in response to a previous publication from researchers of the current study, Koch et al cited conflicting evidence about clinical benefits of genotype-guided therapies, financial concerns, and difficulties in interpreting and applying data clinically as potential downsides to a preemptive model.


Van Driest SL et al. Clinically actionable genotypes among 10,000 patients with preemptive pharmacogenomic testing. Clin Pharmacol Ther. 2014;95:423–31.