© 2005 American Thoracic Society doi: 10.1164/rccm.2412003
Pharmacogenetics and EthnicityMore Complexities of Personalized PrescribingDivision of Therapeutics Queens Medical Center Nottingham, United Kingdom Since the initial description of polymorphic variation in the coding region of the ß2 adrenoceptor by Reihsaus and colleagues in 1993 (1) and the subsequent description of complex haplotypes that include the promoter region for this gene (2, 3), the potential clinical importance of genetic variability at this locus has been the subject of extensive study. Functional responses in recombinant cell systems and primary cultured airway cell lines have revealed altered downregulation profiles for the Arg-Gly16 and Gln-Glu-27 variants of this receptor and altered coupling with the Thr-Ile-164 variant (4–6). This finding has led to a number of clinical studies, which initially had somewhat conflicting results, although there is now an increasing body of evidence suggesting that the Arg-16 variant is associated with increased acute bronchodilator responses to albuterol (7) but a reduced clinical response to chronic treatment with short-acting ß2 agonists (8). The latter effect has led to the first prospective randomized clinical trial examining pharmacogenetic responses in asthma (9), which confirmed the poorer response of Arg-Arg16 individuals to regular short-acting ß2-agonist therapy. Because any single genetic variant arises on a given genetic background, there may be marked differences between ethnic groups in the prevalence of allelic variants for a specific target. The ß2 adrenoceptor is a good example of such a gene; the Glu-27 variant is much less frequent in the black African population than in white populations (10). The other complicating factor in studying genetic variants at a given locus in terms of their functional and clinical consequences is the existence of linkage disequilibrium between the different single nucleotide polymorphisms (SNPs) at the locus: because the allelic frequency for given SNPs may vary in different populations, so will the haplotype frequencies. The study by Choudhry and colleagues in this issue of the Journal (pp. 563–570) (11) provides further insight into potential ethnic group differences in pharmacogenetic effects. The investigators studied bronchodilator responses to albuterol in a group of 274 Mexican and 393 Puerto Rican families recruited on the basis of having a child with asthma. The investigators found significant association of the Arg-16 allele with increased reversibility in the combined population; however, when stratified by ancestral country of origin, the association only remained significant in the Puerto Rican group. The investigators went on to examine whether haplotypes across the ß2-adrenoceptor locus provided additional information. They found that bronchodilator responses were greater in the group who had a certain combination of SNPs, which they designated the GALA haplotype 1. However, little additional information was gained over an SNP-based study assessing Arg-Gly 16 in isolation as almost all individuals with the Arg-16 allele were also GALA haplotype 1. These data contrast with the original report by Drysdale and coworkers (2), in which Drysdale haplotype 4 (which corresponds to GALA haplotype 1) was associated with lower measures of bronchodilator drug responsiveness. This discrepancy may be explained by the relatively small numbers included in the Drysdale study or by different linkage disequilibrium patterns in the different study groups. What are the implications of this study? First, the majority of studies examining ethnic group differences in responses have included all subjects of Hispanic descent as a single group for genetic analyses. Although the allelic frequency of the different polymorphisms at the ß2-adrenoceptor locus between the Puerto Rican and Mexican populations is relatively similar, it is clear that, at least for this pharmacogenetic effect, subjects of Mexican or Puerto Rican ancestry differ in their response to bronchodilators. It is interesting that there are marked differences in asthma morbidity and mortality between the two groups, with Puerto Ricans also having higher levels of asthma prevalence. It seems unlikely given the similar genetic backgrounds of the two populations that these differences can be entirely explained by genetic factors. One would therefore hypothesize that it is the interaction of socioeconomic and other environmental factors with genetic background that is responsible for these differences. Nonetheless, the clear differences based on genotype at the ß2-adrenoceptor locus in acute bronchodilator responsiveness suggest that there may be clinically important distinctions between these groups in long-term responses to administration of short-acting ß2 agonists, which could be predicted by knowledge of ancestry and genotype. Whether or not these differences extend to long-acting ß2 agonists is unknown and should be determined. This may be an issue because, in one previous study involving the retrospective assessment of whites, the Arg-16 allele was associated with greater loss of bronchoprotection after chronic treatment with long-acting ß2 agonists (12). These findings have potentially important implications for clinical trial design in that if substantial differences in response to ß2 agonists are present in other populations stratified by ethnicity and ancestral country of origin, future clinical trials will need to be stratified according to these factors as well as genotype. Such stratification will increase study complexity (and hence the required study size and cost) considerably. The concept of ethnic group differences in pharmacokinetic responses is far from new in the clinical pharmacology literature. One of the best-known examples is the absence of cytochrome P450 2 D6 null alleles in populations of Asian descent; there are many other examples (see Reference 13 for a recent review). However, the need to consider genetic variation in pharmacodynamic (as opposed to pharmacokinetic) targets by ethnic group is relatively new and will potentially cause additional problems for the regulators. With this in mind, the concept of accessible registries for deposition of pharmacogenetic data obtained during clinical trials is to be welcomed. The jury is still out on whether or not for the majority of targets such data will ultimately influence the way in which drugs are licensed or used, but this will undoubtedly be an important topic of study and debate over the next 5 years. There is a broader ethical dimension for the use of genetic profiling to stratify clinical trials that requires consideration. Most of the pharmacogenetic literature has concentrated on differences between populations from different ethnic backgrounds, and has ignored stratification within ethnic groups. However, there is increasing evidence that even within ethnic groups there are significant differences in SNP distribution and linkage disequilibrium profiles (14). As well as complicating the design of clinical trials, increased knowledge regarding genetic profiles of individuals within subpopulations may potentially lead to discrepancies between self-classification of ethnicity and ethnicity defined by genetic profile (15). Because of this issue, the need to define genetic profiles of individuals before individualizing treatment may cause unforeseen ethical issues for subjects and patients. In addition, we are at the start of an age where some drugs will only be licensed for use in specific groups of patients specified by genotype; the concept that ethnicity (or even ancestral country of origin) may also dictate prescribing guidelines, either independently or taken together with genotype information, should also be the subject of debate. FOOTNOTES Conflict of Interest Statement: I.P.H. has received an unrestricted grant for £140,000 for a study in vitro of ß2-adrenoceptor function and a lecture fee of $1,000 for speaking on pharmacogenetics from Kyorin. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
