• Abacavir (Ziagen), an HIV drug, produces a hypersensitivity reaction in about 5%-8% of patients. About 3-4 years after abacavir receive marketing approval, pharmacogenomic research identified a HLA allele (HLA-B*5701) that appeared associated with the reaction. A prospective trial, PREDICT-1, assessed the relationship between this HLA marker and abacavir hypersensitivity, resulting in revised labeling in 2008 that included a strong recommendation for HLA screening of patients before prescribing the drug (N. Engl. J. Med. 2008;358:568-79).
• Clopidogrel (Plavix), an antiplatelet drug widely used to reduce thrombotic events in patients with cardiovascular disease. Clopidogrel is a prodrug that requires multiple cytochrome P450 enzymes for conversion into its active form. Postmarketing studies of clopidogrel showed that a genetic allele of CYP2C19 linked to reduced production of the active metabolite and poorer clinical responses in patients. This led to revised clopidogrel labeling in 2009 and again in 2010 for patients with reduced CYP2C19 function.
• Warfarin, a widely used antithrombotic drug, requires ongoing monitoring of the coagulation state of patients receiving the drug. Variations in the cytochrome P450 enzyme CYP2C9 and in the VKORC1 gene, which encodes the vitamin K epoxide reductase, account for a substantial part of the variability in warfarin’s effect in patients. In 2010, the FDA updated the labeling for warfarin to include a table to guide initial dosing with the drug based on CYP2C9 and VKORC1 genotypes.
"The FDA guidance will make significant headway for U.S.-based studies, and our hope is that potentially the ICH [International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use] will pick up the topic and talk about harmonization of at least the collection of future-use specimens across the United States, Europe, and Japan," Dr. Warner said at the meeting, sponsored by the Drug Information Association. Even now, before the finalized version of the guidance, "communicating the FDA draft guidance has helped." The draft guidance "already seems to be [affecting] decisions" made by trial oversight groups both in the United States and elsewhere, she noted. "When an IRB has questions about the scope of research, we [Merck] have sent them the FDA’s draft guidance, and we’re getting positive responses."
Recent cancer-drug development stands in sharp contrast, having largely avoided the pharmacogenomic issues. "Oncology is often considered an exception. Pharmacogenomics has been most beneficial [to date] in oncology." Cancer drug development has already undergone "a paradigm change," that embraced pharmacogenomics, Dr. Warner said in an interview.
Some recent examples of the now tight link between clinical specimen collection, pharmacogenomic analysis, and cancer-drug development include vemurafenib (Zelboraf), a recently approved drug that specifically acts against cases of advanced melanoma that feature a mutation in the BRAF gene, and crizotinib (Xalkori), which is particularly active against advanced-stage lung cancer that features an ALK fusion mutation.
Other steps the workshop participants agreed on included having companies and industry associations formalize best practices for specimen collection and use, partnering with patient advocacy groups, and, above all, better education aimed at all parties – government authorities, academic institutions, researchers, ethics committees, and IRBs – on why specimens are collected, how they’re handled, and the safeguards in place to ensure confidentiality and proper use. Another development that should ease restrictions going forward is the increasing integration of pharmacogenomics into routine medical practice, and the resultant increased familiarity and comfort that physicians and patients have with these analyses.
"There has been frustration that [drug companies] have identified subpopulations [defined by pharmacogenomics] that hasn’t yet translated yet into clear dosing strategies," Dr. Warner said. The drug industry "needs to do better defining dosing strategies [when physicians] collect pharmacogenomic information.
"Industry needs to build trust for how we collect, store, and use specimens for research. We need to make transparent our processes and how we put them into practice to ensure patient protection and ensure we use specimens responsibly. A realistic goal is a global success rate [for clinical-specimen collection 5 years from now] of 80%. A goal of more than 90%" would be ideal, but probably is not realistic given current issues and challenges, she said.
"Pharma is moving toward more trials in developing countries, where there is more restriction and concern about specimen collection and storage, which will hold us back from a 5-year goal of more than 90%. We’re going to have to make targeted efforts in key countries, such as China. Local storage and analysis [of specimens] will be the strategy that most companies will need to implement."
Dr. Warner is an employee of Merck. Dr. Burckart said he had no disclosures.