This story has been updated to add a sentence that stresses that the FDA does not plan to do away with label changes when it comes to recommending the use of diagnostics. Rather, it will update labels when it deems it necessary.
Simultaneously developing and marketing targeted drugs with companion diagnostics may be the best way to encourage physicians to incorporate genetic testing in their clinical care, according to a US Food and Drug Administration official.
The official also said he does not believe that including language in labels for existing drugs encouraging the use of third-party gene-based tests will have the same effect.
“What is really going to change the way we practice medicine, what is really going to be needed to incorporate genetic and genomic knowledge, not only in drug development but also in drug use, is something that we call drug/test co-development,” Felix Frueh, associate director of genomics at the FDA, recently told Pharmacogenomics Reporter.
“I am convinced that the only way to introduce genetic and genomic tests is by combining the test with the drug,” he said.
Separately, the FDA, hoping to help physicians learn the basics of pharmacogenomics, plans to launch two continuing education courses in mid-2007 through the American Medical Association and the American College of Clinical Pharmacology (see related story).
Industry observers continually note that physician reluctance to adopt genetic tests has been a drag on the introduction and adoption of targeted medicines and other pharmacogenomics products currently on the market.
Diagnostics companies have traditionally believed that having the FDA change the label of a drug to show that a certain genetic test can improve the drug’s efficacy or reduce its risk of adverse events could nudge physicians into using such tests. Similarly, insurers have said they will look for labeling language recommending genetic testing to legitimize reimbursement decisions.
However, Frueh told PGx Reporter that labeling updates have not shown to meaningfully change physicians’ prescribing behavior and that the FDA “might not be” pursuing label changes “as aggressively” as it had in the past.
However, he stressed that it is not the FDA’s policy to do away with labeling changes when it comes to recommending the use of diagnostics. The agency will update labels when it deems it necessary.
FDA No Fan of Labeling Updates, Favors Rx/Dx Co-development
Citing label changes to the cancer drugs 6-mercaptopurine and irinotecan in 2006, Frueh said the agency’s experience with updates so far has been “extremely difficult.” He noted that “even if [the FDA] updated the label, the uptake in clinical practice has been extremely slow …
“If you look at how these label changes affected clinical practice, it’s almost non-existent.”
Frueh said that internally the process “has told us something about the educational needs and the expectations that the medical officers at the FDA put on genomics are sometimes not realistic because they maybe don’t always understand what genomics does.”
“Perhaps we should focus less on what is already out there — maybe label changes isn’t the way we change medical practice,” Frueh said, adding that the agency should focus more on biomarker research prospectively rather than retroactively adding genomic suggestions into an existing drug’s label.
In Frueh’s view, co-developing drugs and companion diagnostics is the best way forward.
The most prominent case in favor of the simultaneous launch of therapeutics and companion diagnostics is Genentech’s Herceptin and Dako’s HercepTest. The immunohistochemistry-based test looks to see if a woman with breast cancer overexpresses the Her-2 protein, which would make her eligible for Herceptin treatment.
According to Frueh, the simultaneous availability in 1998 of Herceptin and the HercepTest made use of the diagnostic “just a given.”
Information correlating Her-2 testing using the HerecepTest and patients’ performance to the drug in clinical trials was included in the original label for Herceptin.
“Herceptin should only be used in patients whose tumors have Her-2 protein overexpression,” the 1998 label stated, adding that the expected incidence of Her-2 overexpression gauged by HercepTest correlated with the results for a clinical trial assay of over 500 breast cancer histology specimens obtained from the National Cancer Institute Cooperative Breast Cancer Tissue Resource.
Vysis’ PathVysion Her-2 DNA Probe – a laboratory test utilizing FISH technology to target the stable DNA and assess Her-2 – became incorporated in Herceptin’s label several years later. The FDA revised Herceptin’s label in August 2002 to include the PathVysion Probe as a method to select patients for Herceptin therapy.
“No one really questions the fact that you have to take a test prior to prescribing Herceptin,” Frueh said. “So at the moment that Herceptin hit the market each and every physician prescribing the drug got educated about the fact that there is a test to be taken prior to prescribing it.”
Introducing a new drug with a companion diagnostic allows for “a very natural process that we don’t necessarily have if we update a label because, again, you’re fighting against established medical practice,” Frueh said. “That’s much harder to do than saying upfront there is a new drug and with this drug there is a test, and both of them come to the market at the same time.”
Frueh indicated that the FDA has noted through recent IND, NDA, and voluntary data submissions that more companies are taking the Rx/Dx co-development route. “Over the next five or so years we will see a whole lot more of these types of personalized medicine approaches for medicines,” Frueh said.
Pfizer and Monogram have such an alliance ongoing in which they are co-developing Pfizer’s investigational HIV drug Maraviroc with Monogram’s CCR-5 tropism assay. Industry stakeholders are touting this companion set as a model for Rx/Dx partnerships. If the drug is approved, the relationship between Pfizer and Monogram will extend from a symbiotic development agreement to a profit-sharing commercial partnership [see PGx Reporter 12-06-06].
While the FDA is planning to back off of making labeling updates regarding genetic tests, sponsors had been hoping that agency recommendations in this regard would encourage physicians to use diagnostics. Affymetrix recently said it believes tamoxifen’s updated label will foment demand from physicians who had been reluctant to adopt Roche’s AmpliChip test, which uses its GeneChip platform to screen for the CYP2C19 and CYP2D6 genes [see PGx Reporter 11-08-06].
FDA’s move away from labeling updates recommending genetic tests may also make reimbursement prospects for companion diagnostics more difficult. Payors have said that a nod from the agency suggesting the utility of specific genetic tests may make it easier to reimburse for a product.
Russell Teagarden, vice president of clinical practices and therapeutics at pharmacy benefits manager Medco, recently said that payors would feel more comfortable incorporating genetics into reimbursement decisions if tests were recommended in a drug’s label or by credible independent institutions [see PGx Reporter 12-06-06].
Warfarin’s Battle: Genetic Tests Vs. 50 Years of Trial and Error
Replication of the Herceptin-HercepTest dynamic may be the gold standard for companies venturing into personalized medicines. Genentech and Dako, however, had an advantage. When their companion products began hitting formularies in 1998 the health care community was in the midst of the Human Genome Project.
But for sponsors of older products, the experience of introducing companion diagnostics has been an uphill struggle marked by deep-rooted conventional wisdom among physicians and long-standardized practices.
“You are kind of stuck in an environment where [it is] very hard to convince physicians that the new way is better than the old way,” Frueh said.
It has been particularly difficult to encourage physicians to use CYP2C9 and VKORC1 genetic tests to dose warfarin, an anticoagulant that became available in the mid-1950s. “If you think about Herceptin and how it got introduced in the market, it’s very different than the warfarin example,” Frueh said.
Physicians currently use a trial-and-error method to figure out the right dose for patients. However, this dosing method can be dangerous, since too much or too little warfarin can lead to serious adverse reactions. Despite the existence of several tests designed to help doctors dose the drug, physicians have been slow to incorporate them into their prescribing practice.
Although the FDA is currently in the process of updating warfarin’s label [see PGx Reporter 11-15-06] to include recommendations for genetic testing, label changes are unlikely to change physician behavior with regard to warfarin, Frueh suggested.
“The updates to labels happen after the fact. The drug is on the market with the label, meaning that physicians that prescribe the drug have been exposed to a drug and the use of the drug without pharmacogenetic information for a certain period of time. What you are doing with label update is that you try to change medical or clinical practice. That is very hard to do,” Frueh said.
“I am convinced that the only way to introduce genetic and genomic tests is by combining the test with the drug.”
“Every physician would agree that once they learn to use a particular drug, that’s the way they are going to use it. Much like a molecular biologist who has learned to extract DNA a certain way, it’s very hard to convince him to use a column from a different vendor, for example,” he said.
However, given the widespread use of warfarin in the US and the high potential for adverse reactions, there are strong arguments in support of using genetic tests. Approximately 2 million new patients are prescribed warfarin each year in the US and between 5 million and 6 million people are on the drug at any given time.
“The public health impact is enormous,” Frueh said. “The genetic variability in the genes that are responsible for the metabolism or action of the drug … seems to have such a large impact that it seems difficult for me to understand why people are so reluctant to take [genetic tests] into consideration knowing that there is a great chance of every new patient having a bleeding episode”.
Frueh said adverse reactions to warfarin are still one of the main reasons for adverse events in the US. “There is hardly any argument against testing except that you have to do the testing and it costs money,” he said.
In fact, today several disparate groups within health care are studying the pharmacoeconomics of genetic tests. For instance, one collaborative study between the Mayo Clinic and the pharmacy benefit manager Medco is looking into the pharmacoeconomics of genetic testing in warfarin therapy [see PGx Reporter 12-06-06]. The Medco/Mayo alliance comes on the heels of a Harvard Partners Center for Genetics and Genomics study investigating the clinical and economic utility of using genetic data in warfarin therapy [see PGx Reporter 11-15-06].
Different Outlook for Genetic Tests in Oncology
Physicians’ outlook on genetic tests also differs by therapeutic area.
In oncology, for instance, patients are willing to tolerate greater toxicities for a few months of added survival. Therefore, with cancer drugs, the use of genetic drugs to gauge which patients will experience adverse reactions is “counter-intuitive” for many doctors who use toxicity to gauge patients’ response to treatment.
“You are talking to a physician that might use what you call an adverse event that you’re trying to prevent as the measure [of efficacy] with which he felt comfortable,” Frueh explained.
“So that’s a fairly tough argument to counterweigh blood levels or mechanistic explanations of why the metabolism of the drug is affected by a genetic variation,” he continued. “That doesn’t sound quite as real as having the patient in front of you that’s dying and you don’t want to give him the full dose.”