COLUMBUS, Ohio — Pfizer’s commitment to personalized medicine has been tempered by five external “barriers,” according to a recent presentation by a senior R&D official at the drug giant.
An untenable reimbursement infrastructure, divergent Rx/Dx-development timelines, ongoing public suspicion that genetic data may be misused, insufficient consumer and healthcare professional education, and the lack of electronic medical records currently must be “addressed in order for the science to move forward,” according to Dennis Van Liew, senior director of Pfizer’s Global R&D Strategic Management Group.
“We see those as five key barriers and we have some ideas about how to address those barriers,” Van Liew said. “But we actually think that personalized medicine is reality.”
These five barriers to personalized medicine were identified six weeks ago at an internal corporate meeting in New York, where Pfizer brought together 80 employees from its branches around the world to discuss the extent to which the commercial market for genetically guided medicine has matured, as well as the company’s investment in the field.
Van Liew described this internal meeting at a conference on personalized medicine last week at Ohio State University Medical Center.
Pfizer has identified potential solutions to some of the five challenges, according to Van Liew. On the reimbursement front, he noted that Pfizer is developing its R&D strategy with input from payors, including their reimbursement outlook for personalized medicines.
“What we’re starting to do is … talk to payors, inviting them to meet our research teams, and saying these are the kinds of programs we’re thinking of developing, these are the kinds of unmet medical needs there are, what evidence do you need to see in order for these to get reimbursed?”
On the Rx/Dx challenge, Van Liew echoed the commonly held view among industry observers that pharmaceutical and diagnostic development timelines need to be aligned in order for personalized medicine to move forward.
“We have found it a real conundrum to try to align product development for a therapy with the companion diagnostic development life cycle,” Van Liew said.
He estimated that the companion diagnostic market would be worth around $11 billion by 2010, and attributed the “miniscule market” to the fact that the development timelines for drugs and diagnostics are not in sync.
The US Food and Drug Administration issued a white paper on Rx/Dx co-development in 2005. Since then, the agency has said it intends to issue a formal guidance on the matter, but has yet to do so [see PGx Reporter 11-21-2007].
At the Ohio conference, Decode Genetics Chief Scientific Officer Jeffrey Gulcher noted that one of the major problems in developing a drug/diagnostic combination product is designing a clinical trial that is appropriately powered to validate the efficacy of both the therapeutic and the assay.
“[O]ur work in [the pharmacogenomics] field … has largely been inward-seeking.”
“When you look at the fact that the FDA would like for you to enroll those patients who are biomarker-negative, in addition to enriching your trial with biomarker-positive patients, and then come up with a strategy to justify a companion diagnostic that will be presumably on the label, you might end up in a predicament,” Gulcher noted.
“Your drug works really well in the patients who are positive for your genetic markers, [and] works half as well or better than placebo in the group that is biomarker-negative,” he added. “But when you do the final analysis [to see if] your diagnostic picks success, then you’re underpowered, because there are some patients who are biomarker-negative.
“So, the worst case scenario is that you have a drug that worked great overall, but wouldn’t have the companion diagnostic because we couldn’t completely validate that using the clinical trial results,” Gulcher noted.
With the development timelines and clinical trial designs for a drug and diagnostic askew, the incentives to pharma and test makers for investing in combination products are also thrown off.
“I hear from payors … [that] 60 percent of our burn on our payments is for chronic disease,” Van Liew noted. “Most of that I can get at without personalized medicine. So, I’m really driven to look at that as the best use of my premium dollars.”
And then there are lingering privacy concerns diminishing the advance of personalized medicine by discouraging participation in clinical trials. Despite the Genetic Information Nondiscrimination Act, Van Liew noted that the public is still fearful of how their genetic information might be misused.
GINA “helps a lot but patients are very nervous about how their genetic information is going to be used,” he said, adding that there needs to be clearer legislation on how genetic samples are used as part of clinical trials.
Pfizer collects genetic samples as part of all its clinical trials, Van Liew noted.
Finally, he added that adoption of personalized medicine could be improved by investing in education for healthcare professionals and consumers, and by implementing electronic health records to identify patients with a particular genetic profile and to track patient compliance.
At the meeting, Van Liew suggested that the pharmaceutical industry would do well to explore pharmacogenomics to improve the success rate of investigational products.
According to industry observers, approximately 60 percent of investigational drugs fail Phase II clinical trials while 56 percent of drug candidates fail in Phase III. Conventional wisdom about the clinical applications of pharmacogenomics holds that big pharma will have the most use for the discipline in their attempt to learn about investigational products earlier in the developmental continuum to make go/no-go decisions, reduce risk, and improve the success of their products.
“Historically, at Pfizer, our company’s molecular scientists and our molecular medicine group are using techniques with biomarkers and tests to de-risk drug discovery and development,” Van Liew said. Pfizer uses pharmacogenomics strategies to “design trials, and if we have an unexpected event in a trial, to try to characterize that event in the redesign of that trial.
“So, our work in this field … has largely been inward-seeking,” he noted.
Van Liew pointed out that in the few instances in which Pfizer has launched drugs with companion diagnostics, it has done so because those drugs could only be brought to market for a very specific subset of the population.
For instance, in 2006, Pfizer and Monogram entered into a non-exclusive agreement to market a test that will gauge the tropism of status for patients prior to treatment with Pfizer’s CCR5-antagonist HIV drug Selzentry [see PGx Reporter 12-06-06].
Pfizer has said that it identified early on the need for a diagnostic partner in order to bring Selzentry to the market. In order for Selzentry to meet FDA’s efficacy standards, a phenotypic assay was needed to determine whether an individual’s virus exclusively uses the CCR5 co-receptor (R5-tropic) and thus will respond to the drug. As part of its collaboration with Monogram, Pfizer invested $25 million in the diagnostic firm, used its Trofile assay in clinical trials for Selzentry, and even submitted information about the test to FDA as part of Selzentry’s master file.
The FDA granted accelerated approval for Selzentry in August 2007. The drug’s label recommends doctors establish the tropism status of their patients prior to administering the treatment [see PGx Reporter 08-08-2007].
“If we think about personalized medicine looking forward, we think about diseases in terms of unmet medical need,” Van Liew said at the conference.
He also acknowledged that an investment in personalized medicine might mean lower revenues for big pharma.
“Medicines we’ve brought to market recently have had lower peak sales than blockbuster drugs we have,” Van Liew commented. “But we are convinced that as long as you meet an unmet medical need that there will always be a commercial place for that.”
To that, Decode’s Gulcher quipped, “So, cardiovascular disease is not an unmet medical need anymore at Pfizer?” He was referring to the broad R&D restructuring currently ongoing at the pharma giant.
PGx in R&D Shuffle?
In late September, a leaked internal Pfizer memo revealed that the marketer of the mega-blockbuster statin Lipitor plans to exit the cardiovascular disease space, in particular the markets for hyperlipidemia/atherosclerosis, heart failure, obesity, and peripheral arterial disease. Lipitor is used to treat or prevent all of these conditions.
Pfizer’s decision to leave the space follows the costly late-stage failure in 2006 of torcetrapib, a CETP inhibitor it hoped would become a blockbuster, and the loss of patent protection of its CV drug Norvasc last year.
Another reason to abandon CVD is increased competition in the space from generic versions of Bristol-Myers Squibb’s Pravachol and Merck’s Zocor, which have hurt sales of all statins. Also, Lipitor will lose its patent protection in 2011.
In the memo, Pfizer said it plans to redirect its R&D dollars into treatments for cancer, diabetes, Alzheimer's, pain, and mental health. Although it is not immediately clear how this restructuring could affect Pfizer’s investment in pharmacogenomics-guided personalized medicine, Van Liew’s comments in Ohio and some recent Pfizer R&D alliances with diagnostic firms suggest it is cautiously considering PGx strategies as a way to reduce attrition in its pipeline and improve the efficacy of marketed drugs in certain subpopulations.
Within Pfizer’s shifting R&D focus, cancer remains one of its “high priority areas.” In this market, the pharmaceutical giant has partnered with at least one diagnostic company to invest in personalized medicine.
Genomic Health has said it has begun gene identification work to develop a prognostic test for clear cell type kidney cancer under its collaboration with Pfizer [see PGx Reporter 08-06-2008].
Pfizer did not respond to requests for an interview.