Author(s): John Jack [*] aff1 aff2 , Tammy M Havener aff3 , Howard L McLeod aff4 aff5 , Alison A Motsinger-Reif aff1 aff2 , Matthew Foster aff6
admixture; cancer; Caucasian; cytotoxicity; drug; genome-wide association study; GWAS; hispanic; in vitro ; lymphoblastoid cell lines
Pharmacogenomics research has increased the understanding of genetic variability and its relationship to individual drug response [1 ]. It is well established that the majority of drugs/doses to combat diseases and disorders do not exhibit uniform effects for individuals. In particular, chemotherapeutic agents often have a narrow therapeutic range, which blurs the line between treatment and inefficacy or drug-related toxicity [ 2,3 ]. While the interindividual variation in drug response is undeniable, the proportion of this variation that is due to genetic variation is an open question. With family-based design cohorts available, it is unusual to be able to formally assess the heritability of differential drug response [ 4 ]. Pharmacogenomics investigation often relies on indirect evidence of a genetic etiology, such as differential response in animal models, and racial differences in clinical outcomes.
For antineoplastic therapies, there is clear evidence of racial disparities in outcome. While a portion of this variation is undoubtedly due to socioeconomic factors, studies with consistent and high access to therapy and adherence across racial and ethnic subgroups have also demonstrated disparate outcomes [5 ]. Most pharmacogenomics studies of genetic determinants of cancer disparities have focused on racial differences - between subjects of African versus European ancestry. Genetic determinants of disparities between ethnicities (e.g., Hispanic versus non-Hispanic Caucasian) are less well investigated. One reason for this paucity is that ethnic groups are typically defined by common geography and culture rather than ancestry. Such groups are typically highly admixed, featuring recent ancestry from multiple continental populations. However, with the refinement of high-throughput genomic methods, the potential to understand the relationships between complex diseases and admixed populations is becoming possible.
For example, admixture mapping in African-Americans successfully identified the genetic locus responsible for a form of neutropenia [6 ]. It has also been implicated in finding important variations linked to prostate cancer [7,8 ], hypertension [9 ] and renal disease [10 ]. However, highly admixed ethnic groups such as the Hispanic population have offered greater challenges, despite well-described disparities in outcome. To this point, the well-described [11 ] higher rates of relapse and poorer survival of Hispanic children with acute lymphoblastic leukemia (ALL) have recently been linked to percentage of Native American ancestry using principal component analysis (PCA) [ 12 ]. Furthermore, genome-wide association studies (GWASs) have linked polymorphisms in the ARID5B gene to both elevated risk of developing ALL and risk of relapse after multiagent chemotherapy in this population [13,14 ]. Despite these advances, whether these genetic associations are responsible for variations in response to particular antineoplastic agents remains unclear. As ALL chemotherapy regimens typically contain seven or more chemotherapeutic agents, the contribution of any individual drug to genetically determined disparities is unlikely to be determined in analyses of clinical trials. For this reason, preclinical models of genetically determined drug susceptibility...