Skip to main content

Conference Program

Subpage Hero


R. Rodney Howell Symposium: Looking Beyond the Lamppost: Genome-first Approaches Using EHR-linked Biobanks

16 Mar 2024
Venue: Metro Toronto Convention Center
Meeting Room: Exhibit Hall FG
Clinical Genetics and Therapeutics
  • Accredited:
    • Accredited
  • Primary Categories:
    • Population Genetics
  • Secondary Categories:
    • Population Genetics
  • Level of Learner:
    • Basic
The traditional approach of evaluating patients/families with a rare phenotype has included single-gene testing for those who meet certain clinical criteria. This phenotype-first single-gene approach may be subject to ascertainment bias towards the most highly penetrant clinical manifestations, miss non-penetrant cases, miss rare or unknown disorder manifestations, or over-estimate disease severity.  The increasing availability of large-scale, population-based exome/genome sequenced cohorts now allows for a genome-first strategy to determine a rare variant (gene) distribution and its clinical consequences in a population.  Such examinations have detected much higher-than-expected prevalence of purported pathogenic germline variation for a variety of disorders suggesting that these variants might cause unidentified or less severe clinical phenotypes.  Alternatively, our understanding and interpretation of the variant(s) may be incorrect.  Similarly, results from multigene panel tests for individuals who do not meet full syndromic criteria or have minimal family history have identified individuals harboring pathogenic variants in a variety of monogenic disorders. These changes in variant detection approaches will require fundamental adjustments in risk estimation and genetic counseling. 

The genome-first approach has potential to reduce ascertainment bias, identify a more complete phenotypic spectrum, and permit more precise gene and/or variant penetrance estimates. This session will illustrate challenges, opportunities, and lessons learned (to-date) in a variety of disorders using a genome-first approach. The four presenters will draw upon their published experience using population-scale cohorts (Geisinger, UK Biobank, Vanderbilt BioVU, NIH All of Us). It will start with an introduction by Dr. Justice, an epidemiologist at Geisinger with published experience in the analysis of the MyCode cohort. Dr. Kozel will then detail her use of the genome-first approach to identify individuals with elastin variation and what has been learned when those individuals are subject to deep phenotyping at the NIH Clinical Center. Dr. Mosley will then describe how genomically ascertained benign variation can improve clinical outcomes and decision-making. Dr. Stewart will show how variation in genomically ascertained tumor-predisposition genes is more common and often less penetrant, and associated with more varied phenotypes, than previously suspected.

Learning Objectives

  1. Describe benefits of leveraging clinical biobanks for gene discovery at the population-level and potential limitations
  2. Compare EHR vs. prospective deep phenotyping outcomes for individuals identified through gene-first approaches
  3. Explain how genetic variation unrelated to disease risk may prompt escalations in clinical care
  4. Relate how prevalence, phenotype and penetrance compare in phenotype- vs. genome-first approaches in cancer genetics