Reaping the Early Harvest of the Genomics Revolution

Reaping the Early Harvest of the Genomics Revolution

Gastroenterology
May 2010

Raymond T. Chung (Associate Editor)



Much fanfare heralded the successful sequencing of the human genome in 2003. Newly armed with a comprehensive understanding of the human genetic map, investigators proclaimed that personalization of medicine was imminent. Now, 7 years after this signal discovery, where do we stand in the delivery on this promise to clinical care?


In the past several years, there has been a breathtaking flurry of genome-wide association studies (GWAS) designed to identify susceptibility loci for complex, polygenic diseases. These studies have succeeded in identifying single nucleotide polymorphisms (SNPs) that increase predisposition to conditions incrementally (usually with odds ratios well under 1.5). For these disorders, which include Crohn's disease, rheumatoid arthritis, systemic lupus erythematosus, and type 2 diabetes mellitus, it is likely that only a small fraction of the genetic risk for these conditions has been identified. Although doubtless much will be learned regarding the functional significance of these susceptibility loci through fundamental biological investigation, and ultimately insight gained into new therapeutic targets or biomarkers, the dividends may be some years in the making. Furthermore, it seems that the probable requirement for interaction of several predisposition loci with environmental factors to trigger disease makes the prospect of a “slam dunk” genetic test to predict risk for complex disorders a somewhat remote one.


There are, however, notable exceptions to the premise that a given phenotype represents a complex interaction of genotypes. One such example was the recent discovery of an extraordinarily strong relationship between a SNP in the IL28B gene, which encodes IL28B, or interferon lambda-3 (IFNλ-3), and responsiveness to IFN-α–based antiviral therapy for hepatitis C virus (HCV) in persons harboring genotype 1 infection. In a recently published GWAS study of patients with chronic hepatitis C who had entered a clinical trial of pegylated IFN-α and ribavirin, possession of a favorable genotype (CC) at the rs12979860 SNP in the IL28B locus was associated with a 2.5-fold improved likelihood of experiencing sustained virologic clearance compared with the unfavorable genotype (CT or TT).1 To put this finding into genomics perspective, the significance threshold for GWAS studies is typically about 10−8. In the case of the IL28B allele, the overall significance level was a staggering 10−28. This level of significance may be as close as we get to a bankable finding.


More important, to put this finding into clinical perspective, the impact of IL28B genotype is on par with viral genotype as a determinant of IFN and ribavirin responsiveness, and eclipses viral load, liver histology, and ethnic group and race as a predictive factor. Indeed, the frequency of the favorable IL28B SNP explains in large measure (but not completely) the known, observed decrease in sustained virologic response in African Americans compared with Caucasians and Hispanics. To reinforce the strength of this data, these findings have been replicated in at least 2 other clinical cohorts of Asian and Caucasian descent.2, 3 Not surprisingly, the favorable IL28B genotype was also associated among persons with acute hepatitis C infection with a significant likelihood of spontaneous clearance compared with those who developed chronic infection.4


Is the finding of such a strong association with a single polymorphism surprising? Perhaps not, if one considers that, unlike complex disease traits, susceptibility to an infection that does not produce immediate life-threatening illness may not be under intensive selection pressure. The finding that a polymorphism in a gene that encodes a member of the type III IFN family (as IL28B is) raises a whole raft of fundamental biological questions that relate to the antiviral properties of IL28B and its interaction with type I IFNs, which include IFN-α. We can therefore anticipate that the uncovering of these interactions will remain an active area of investigation for years to come.


Although further investigation of much larger numbers of subjects will likely yield additional meaningful determinants of treatment-induced or spontaneous viral clearance, the fact remains that these findings are striking, and are likely to have implications for clinical practice among gastroenterologists and hepatologists who care for persons with HCV infection. How might such a gene test be used in practice? For starters, the use of an a priori test that can stratify response likelihood among persons with genotype 1 HCV will represent an important tool that will lead to the minimization of exposure of patients to IFN and ribavirin and their attendant toxicities. Possession of an unfavorable (non-CC) genotype will be taken as a measure of a very low likelihood of sustained virologic response, and, among those patients with limited liver disease who can afford to wait, could justify deferral until cocktails of direct antiviral agents become available, possibly as early as the next 5–10 years. On the other hand, those patients with the favorable IL28B genotype might be viewed as having a high likelihood of responding, and be encouraged to consider initiating therapy with current standard of care. In other words, genotype 1-infected patients with favorable IL28B genotype may behave clinically as do genotype 2- or 3-infected persons. Trials to examine the value of truncating therapy in this group of patients (as is the case for patients experiencing rapid virologic responses on pegylated IFN and ribavirin therapy) will no doubt also be warranted.


It is also not difficult to imagine combining this pharmacogenomic data with more conventional clinical variables (such as liver histology or viral load) to further refine our prediction rules for response. With this pharmacogenomic breakthrough, we in gastroenterology and hepatology will now firmly enter the arena of predictive medicine.


These findings have additional downstream implications for clinical trial design. It is highly conceivable that patients entering clinical trials of new therapies for HCV will be stratified by IL28B genotype. Furthermore, studies examining strategies of therapies thought to be of somewhat higher risk (such as combining direct acting antiviral agents without IFN) could be restricted to those with a favorable IL28B genotype who might then be “rescued” with IFN-based regimens in the event of treatment failure.


On the other hand, the counterargument could be made that those persons who have an unfavorable IL28B genotype should attempt to enter trials with oral cocktail therapy in light of their poor responsiveness to IFN. Either way, we seem to be poised to be able to dictate therapeutic choices in the management of HCV based on genetic information.


Although pharmacogenomics has already been productive in identifying alterations in the metabolism of some therapeutic agents (eg, azathioprine or 6-mercaptopurine), the power of the IL28B pharmacogenomic data lies in its potential ability to influence the subsequent choice of therapeutic agents themselves. Of course, an additional possible major benefit attending the IL28B gene discovery is that IFN-λ3 could itself become a useful therapeutic agent against HCV, either alone or in conjunction with IFN-α.


As we turn the corner on the first decade of the 21st century, the exploration of genomics in medicine has been made possible by efforts to pool very large numbers of patients into informative cohorts. Even so, we have only begun to peel away the outer, more apparent layers of risk, since it is clear that ever larger numbers of subjects will be required to identify less obvious but contributory loci to complex diseases. Yet as we embark on these lofty efforts, we can now be gratified that picking some of the lowest-hanging fruit of all has positioned us for a bountiful yield.



References


1. Ge D, Fellay J, Thompson AJ, et al. Genetic variation in IL28B predicts hepatitis C treatment-induced viral clearance. Nature. 2009;461:399–401



2. Suppiah V, Moldovan M, Ahlenstiel G, et al. IL28B is associated with response to chronic hepatitis C interferon-α and ribavirin therapy. Nat Genet. 2009;41:1100–1104


3. Tanaka Y, Nishida N, Sugiyama M, et al. Genome-wide association of IL28B with response to pegylated interferon-α and ribavirin therapy for chronic hepatitis C. Nat Genet. 2009;41:1105–1109


4. Thomas DL, Thio CL, Martin MP, et al. Genetic variation in IL28B and spontaneous clearance of hepatitis C virus. Nature. 2009;461:798–801