INTRODUCTION: The St. Jude Children’s Research Hospital PG4KDS clinical trial opened in 2011 with the goal of providing a process to clinically implement preemptive pharmacogenomics. Pharmacogenomic test results are placed in the electronic health record (EHR) and coupled with interpretations and clinical decision support (CDS) alerts to guide treatment decisions across a number of therapeutic areas.
METHODS: Genomic DNA obtained from peripheral blood is genotyped using the PharmacoScan™ array (Affymetrix, part of ThermoFisher Scientific, Santa Clara, CA, USA; from 2011 to 2017, we used the DMET™ Plus array supplemented with a CYP2D6 copy number assay). Clinical pharmacists place pharmacogenomic results and interpretative consults in the EHR. They also perform a medication assessment of the patient’s current drug therapy regimen to screen for high-risk drug use. Active CDS alerts are customized for each gene-drug pair to include prescribing recommendations. Problem list entries for high-risk phenotypes in the EHR serve as discrete fields from which CDS alerts are generated. Genedrug pairs are prioritized for implementation based on availability of CPIC guidelines, primary evidence review, and with approval of an institutional oversight committee. Participants are asked whether they wish to be informed of their pharmacogenomic test results when they are released into the EHR, and once they turn 18 years old are asked to re-consent in order to remain on the protocol.
RESULTS: As of May 2018, eight genes (CYP2D6, CYP2C9, CYP2C19, CYP3A5, DPYD, UGT1A1, TPMT, and SLCO1B1) and 25 drugs have been clinically implemented (see www.stjude.org/pg4kds). Out of 4735 patients approached for participation, 4471 patients (94%) have consented to enroll in the PG4KDS protocol. The median age of patients was 8.9 years (range 0.17-51.92). Of these patients, 4030 (92%) have at least one high-risk pharmacogenomic result posted in the EHR. Most participants (96%) have requested to be informed of their test results, and out of 550 patients who were approached for re-consent when they reached the age of majority 535 (97%) consented to remain on the study.
DISCUSSION: The PG4KDS model for implementation of preemptive pharmacogenomics provides a successful example for pharmacogenomic testing on an institution-wide level.
Integrating Big Data (genome data, pharmacogenomics, therapeutic applications of genome ed
