Predictive Genetic Testing Using Common Variants: Evidence and Translation

Abstract

Starting with the publication of the first genome-wide association study (GWAS) in 2005, the last decade has seen the identification of thousands of novel genetic susceptibility loci. These common, low-penetrance single nucleotide polymorphisms (SNPs) have been associated with myriad phenotypes, ranging from benign (e.g., skin and hair pigmentation) to medically informative (e.g., risk of Alzheimer’s disease), and, in a few cases, clinically actionable (e.g., genotype-guided dosing of the anti-coagulant warfarin). While the clinical applications of GWAS findings remain limited, the commercial industry that has grown up around this research – providing testing for ancestry-, trait-, and disease-associated SNPs directly to consumers – continues to expand. In Sections 1 and 2 of this dissertation, I evaluate the impact of direct-to-consumer personal genomic testing (DTC-PGT) on its customers. First, I quantify the effect of providing genetic risk information on consumers’ perceived risk of four common cancers, and consider the implications of my findings in the context of clinical cancer risk counseling. Second, I describe post-PGT prescription medication behaviors, evaluate the relationship between these behaviors and the pharmacogenomic information provided to consumers, and finally interpret my findings against the backdrop of recent efforts by the Food and Drug Administration (FDA) to regulate the DTC-PGT industry. In Section 3, I take a step back and evaluate the evidence underlying a hypothesized concurrence between the genetic susceptibility loci for late-onset Alzheimer’s disease (LOAD) and those for early subjective memory complaints (SMCs) in cognitively intact adults. I conclude by discussing how my findings contribute to our understanding of the biological relationship between these phenotypes and to efforts to target risk SNPs and/or SMCs in screening tools for LOAD.