Including genetic information in clinical trials: hepatitis C and IL28B
Online this week, Nature has published a genome-wide association study for response to treatment for chronic hepatitis C infection-the authors identify a polymorphism in an interleukin gene that is a strong predictor of how well an individual is able to clear the virus. Interestingly, the frequency of the polymorphism in different populations tracks the previously noted population difference in drug response, and the authors claim to explain half the difference in response rate between African- and European-Americans with this single polymorphism.
This paper is also interesting in that it represents one of the first (if not the first?) studies to coordinate a drug trial (in this cases, three different treatments for hepatitis C) with a genome-wide association study. This promises to lead to both important advances–as researchers are able to identify genetic subgroups of individuals who respond (or not) to a drug, even if it is ineffective in the population as a whole–as well as (my cynical side speaks) additional opportunities for misleading post hoc analyses by drug companies to try to salvage and market drugs that don’t work. Hopefully, mostly the former–this could be an important step towards legitimizing genetic information in the eyes of MDs, and an ever-so-slight step towards personalized medicine.
Labels: Genetics
I think genetic information is already considered legitimate by large subsets of physicians. Certainly for hematologists/oncologists, infectious disease specialists and pathologists genetic tests are already part of everyday practice, although often the genetic tests themselves are performed at reference labs due to patent restrictions and limited sample numbers requiring each test making it un-worthwhile to perform them on-site.
i guess i don’t know the field very well (I was basing that comment on a conversation about genetics and warfarin dosing, or the relative indifference of MDs to genetics when they can do a quick blood clotting test in-house); what tests are you referring to?
Where I trained, all patients with hepatitis C had their virus genotyped. Molecular tests for inherited thrombophilias like Factor V Leiden and prothrombin mutations, for microsatellite instability in colon cancers, for BCR-ABL expression in bone marrow and peripheral blood samples from patients with a history of chronic myeloid leukemia, etc. were also routinely performed. I don’t know why genetic testing for warfarin response hasn’t caught on, but my first guess would be expense. It is also still necessary to monitor the patient clinically and to follow PT’s and INR’s even after the genetic test has been performed.
The IL28B polymorphism is for SNP rs12979860. This SNP isn’t among the “basic” set that 23andMe tested for. (I don’t know about their current set, or those offered by Navigenics and others.)
The effect in terms of sustained viral response to ribavirin/IFN therapy shown in the paper is quite striking. For instance, Fig. 1 shows the T/T genotype for European-Americans leads to an SVR of ~33%, while C/C gives an SVR of ~81%. Heterozygotes are intermediate.
The minor T allele is quite common; the E-A breakdown is 38% C/C, 50% T/C, and 12% T/T.
Similar-scale effects are shown for other ethnicities.
Especially with the first novel agents from Vertex and others in late-stage Phase 3 trials, these results look very exciting, as a way to stratify patient populations and guide therapies.
*I’m an author on this study
This study is not the first report describing a GWAS using samples from patients enrolled in a clinical trials. GSK used a much more limited set of samples to identify the HLA locus associated with abacavir hypersensitivity reaction, (although I think the samples were culled from patients across a number of studies).
What I think is much more striking is that size of this effect and the failure of the numerous candidate gene studies to ever target this gene.
This paper (I hope) will open the door to many more genome-wide studies which will help identify genetic factors contributing to therapeutic response to pharmacotherapy.
WRT the warfarin PGx story, it is unfortunate that the utility of the CYP and VKORC1 assays has not infiltrated medical practice more effectively, but pharmacoeconomic studies have failed to show an overwhelming benefit to patients. This, coupled with glaring lack of education about the value of these tests in the primary care setting, and the comfort of MDs on the front line with titrating dose based on INRs (and the lack of easily accessible and fast test results coupled with clear dosing guidelines) combines to make warfarin PGx not ready for prime time.
Given the reality that chronically HCV infected patients wait on average 1 year after diagnosis before initiating therapy (due to the noxious nature of the 48 week regimen and the uncertainty related to response), I very much hope that this predictive test becomes a tool patients and caregivers will use to guide therapeutic decision-making.
Thanks Jason, interesting stuff.
Followup on this paper in the 8/18/09 GenomeWeb Biotech Transfer Week.
It turns out that the Schering-Plough Research Institute funded the Duke work, and S-P owns the IP to the SNP.
Fair use extract –
“[S-P] is contemplating how it might be able to further develop and market a diagnostic based on the biomarker to accompany its current marketed interferon-based HCV treatment.
[snip]
The company markets peginterferon alfa-2b… with ribavirin as an approved therapy for HCV, which affects an estimated 170 million people worldwide.
However, the treatment, which has a 40% cure rate, is expensive, takes four months to administer, and can lead to serious side effects.”