The US Food and Drug Administration should rewrite the label for warfarin to recommend testing patients for variants of two genes in order to better gauge starting doses, according to a recommendation made last week by the FDA's Clinical Pharmacology Subcommittee of the Advisory Committee on Pharmaceutical Science.
Following a two-day meeting in Rockville, Md., the subcommittee voted 8-2 to recommend relabeling the blood-thinning drug to include information on both genes. The decision follows on the heels of a similar suggestion to relabel the colorectal cancer drug irinotecan, which is manufactured by Pfizer under the trade name Camptosar. The FDA followed the panel's advice in June and relabeled the drug accordingly.
As part of the relabeling effort dubbed "retrofitting" by Larry Lesko, director of the Office of Clinical Pharmacology and Biopharmaceuticals at the US FDA's Center for Drug Evaluation and Research the FDA updates the information available about currently marketed drugs to take advantage of pharmacogenomics technologies or discoveries. If the warfarin relabeling follows the irinotecan precedent, it could be followed by the development and commercial launch of a warfarin-specific diagnostic that will be submitted to the FDA for clearance. The development is especially noteworthy because warfarin, a blood thinner, is one of the biggest-selling prescription drugs of all time.
"I fully expect companies
to begin developing commercial assays."
By updating Camptosar's label, the FDA encouraged companies to develop and submit their own assays for interrogating a gene related to the drug's adverse events. The only FDA-cleared diagnostic for detecting the gene, called UGT1A1, is an Invader test made by Third Wave Technologies. The newer Camptosar label recommends a lower starting dose of the drug in patients homozygous for the UGT1A1*28 allele.
Asked whether the FDA is in contact with diagnostic companies concerning a test intended to aid warfarin dosing, Lesko said, "That is the case. There were many companies who make diagnostics [that attended] … the advisory committee, and we did have conversations with them after the meeting, and I do know that this is going to be in a commercial-development mode with several companies over the next four to six months."
Lesko declined to say specifically which companies were developing warfarin-related diagnostics, but he said representatives from Roche Diagnostics, Third Wave Technologies, and a trade association for medical device manufacturers attended the subcommittee meeting.
"My expectation is that the [drug] relabeling and the [clearance] of a test for these genes would be done, like we did with irinotecan, in a coordinated way, so that doctors who want to order a test have available an approved test for the enzyme test we recommend in the label," Lesko told Pharmacogenomics Reporter this week.
Since 30 million prescriptions are written every year for warfarin in the United States, and given the high rate and severity of adverse events, "I fully expect companies to begin developing commercial assays," Lesko added.
Deciding What to Do
The subcommittee's task was to review the evidence on CYP450 2C9 and VKORC, both of which have a function in regulating the dose-response relationship for warfarin. The members looked at whether genotyping patients before beginning warfarin therapy would reduce adverse events and "improve achievement of stable" international normalized ratios, according to FDA draft questions prepared for the subcommittee meeting.
The subcommittee was also charged with advising whether lower starting doses of warfarin should be used in patients having certain genotypes. If it found reason to suggest both genetic testing and lower doses, the subcommittee was asked to consider whether and how the drug's label should be changed.
"My expectation is that the [drug] relabeling and the [clearance] of a test for these genes would be done, like we did with irinotecan, in a coordinated way, so that doctors who want to order a test have available an approved test for the enzyme test we recommend in the label."
Clinicians currently use a variety of clinical factors, such as weight, height, age, gender, and other current medications, to estimate a patient's maintenance dose of warfarin. After starting a patient on a certain dose, physicians will adjust it over time according to international normalized ratios, or INR, which is a measure of clinical response. But a given patient's INR is unreliable during the first 30 days of therapy, and adverse events associated with warfarin, including embolisms, can be life threatening.
"If you look at both [2C9 genotype and VKORC haplotype], you account for a greater percentage of the variability" than the typical clinical factors can, Lesko said. "The two genes allow you to account for up to 50 to 60 percent of the variability in dose response," and thereby reduce the "very high" adverse event rate and more quickly get the patient into the target INR range to avoid embolisms, he said.
"We did not, at this meeting, discuss specific information that would go into the label, so what we're going to do is begin to look into the label and determine what specific recommendations we're going to put into the label," said Lesko. That process usually involves discussions with the drug maker on the language to be used in a new label, and the company generally files a supplement to its drug application that contains the labeling recommendation, he said. The committee would also be required to brief the FDA's medical division covering hematology and ask for advice on the label's wording, he added.
Since warfarin is a generic drug the FDA will probably contact the maker of the reference-listed drug, which is Bristol-Myers Squibb, which manufactures the compound under the trade name Coumadin.
Lesko estimated that the entire process should take between four and six months.
Chris Womack ([email protected])