Genomic Medicine Implementation Unit

National Human Genome Research Institute

Investigators

Abstract

This project is focused on translating genomic knowledge into improved diagnosis, treatment, and prevention in rapidly evolving areas of genomic medicine such as pharmacogenomics, undiagnosed Mendelian conditions, and actionable secondary findings from genomic sequencing and other genetic testing. While we plan to integrate our work with that of other NCHPR Principal Investigators, serving as the implementation arm for their translatable findings, in this first year we have focused on establishing our small footprint with one post-baccalaureate fellow. She has completed FAES courses in basic statistical analysis and R programming, as well as a university course in human physiology to help her better understand the concepts and approaches being applied in genomic medicine. Our initial work has focused on assessment of actionable variants in pharmacogenes (genes whose products control critical pathways in drug absorption, distribution, metabolism, excretion, and response) among sequenced patients in the ClinSeq cohort and other Clinical Center patients. We have reviewed Clinical Pharmacogenetic Implementation Consortium (CPIC) guidelines for using these pharmacogenomic (PGx) variants to identify those variants most likely to be relevant in these patients and most likely to be assayed by exome sequencing. We found this approach to have a number of limitations: 1) published literature suggests that exomes may fail to detect more than half of actionable PGx variants; 2) software for extracting PGx variants is still in development; and 3) any variants to be used in clinical care would still need validation with Clinical Laboratory Improvement Act (CLIA) testing. To address these problems, we are developing a collaboration with William (Doug) Figg and Tristan Sissung of the National Cancer Institute (NCI) to compare PGx variants detected by exome and genome sequencing with those identified through a CLIA-validated commercial product, PharmacoScan (Thermo Fisher Scientific) in a small series of NCI lung cancer patients. We are also exploring this problem in exome and genome sequencing data derived from a series of National Institute of Allergy and Infectious Diseases patients with Steve Holland, Morgan Similuk, and Magda Walkiewicz. Once we can identify a viable assay method for actionable PGx variants, we expect to develop user-friendly educational material to share with Clinical Center patients, their NIH providers, and their referring primary providers. We hope to provide PGx consultations for these patients and providers, and perform follow-up assessment of the use and usefulness of this information in these patients subsequent clinical care, whether related to their primary condition for referral to the Clinical Center or to other existing or subsequent conditions. We have also examined race/ethnic differences in frequency of PGx variants in Latino and Indigenous American populations as catalogued by CPIC guidelines for 11 PGx genes classified for normal or altered function (CACNA1S, CYP2B6, CYP2C9, CYP2C19, CYP2D6, CYP3A5, DPYD, NUDT15, SLCO1B1, TPMT, UGT1A1). Not surprisingly, we determined that these populations are significantly underrepresented among the 632,000 persons of European, Latino, and Indigenous American descent included in the data used to generate these guidelines. More importantly, we observed that alleles with altered function were detected at a nearly 5-fold higher frequency in Latino and Indigenous American populations than in European populations. This could have important clinical implications and widen health disparities experienced by these two groups as PGx recommendations are brought into clinical practice. These data were submitted as an abstract to the October 2021 virtual meeting of the American Society for Human Genetics that has been accepted for poster presentation. We are expanding our exploration of these race/ethnic differences to other underrepresented groups reported in CPIC guidelines and have begun work on a manuscript reporting these findings. We also collaborated with Les Biesecker, Jennifer Johnston, and their colleagues to assess the yield in the ClinSeq cohort of actionable variants from the newly released American College of Medical Genetics and Genomics (ACMG) ACMG SF 3.0 addition of 14 genes (TTN, FLNC, TRND, CASQ2, PALB2, MAX, TMEM127, GAA, BTD, ACVRL1, ENG, RPE65, HFE, HNF1A) meeting the ACMG Secondary Findings Working Group criteria for actionability, severity, penetrance, and impact or burden of treatment and/or screening. We performed manual curation of several of these genes and jointly found the returnable P/LP variant frequency increased by 25%, from 3.7% to 4.6%. These data were also submitted as an abstract to the October 2021 virtual meeting of the American Society for Human Genetics that has been accepted for poster presentation.