Pharmacogenomics technologies may help scientists at Amgen and Johnson & Johnson determine why certain patients experience a paradoxical adverse event after being given a widely prescribed drug made by both firms.
Experiments with these technologies, if they are undertaken, may also be an important real-world test case for personalized medicine.
At the core of this epidemiological mystery is erythropoietin, the recombinant protein sold by Amgen and J&J, which is used to treat certain hematopoietic side effects tied to chemotherapy and kidney disease.
Around four years ago, physicians in Europe began noticing that certain patients with chronic renal failure who were given the drug, known as EPO in the United States, were experiencing a rare adverse event that mimics the blood disorder EPO was designed to treat. And, though early epidemiological studies conducted with Amgen linked the adverse event, pure red-blood-cell aplasia, to the EPO sold by J&J, a definitive explanation remains elusive.
“I think [in this case pharmacogenomics] has a shot” at determining the cause of the adverse event, said Mary Relling, a member of the NIH’s Pharmacogenetics Research Network. But first, she and others cautioned, both companies must rule out all other potential problems — from manufacturing and handling to administration and what other drugs may have concomitantly been prescribed.
EPO itself is at the center of a marketing grudge match between Amgen and J&J, each of which has been trying to win a greater portion of a global market with more than $7 billion in annual sales. (Amgen invented EPO in the 1980s with financing help from J&J. In exchange, Amgen awarded J&J a piece of the market, and the two companies have been in court over that split ever since.) Today, J&J manufactures its own erythropoietin in Europe, called Eprex, and markets in the United States a version made by Amgen.
It is in the European market, beginning in 1999, where approximately 180 patients with chronic renal failure undergoing erythropoietin therapy developed PRCA. Though some believe this number of adverse events to be relatively modest, Amgen and J&J consider the market to be critical: Amgen earned a little less than $3 billion from sales of EPO in 2002, which is almost half the company’s total sales that year. Eprex is also J&J’s most lucrative product, generating $4.3 billion in revenue last year — or almost 12 percent of the firm’s total sales in 2002.
To some, however, PRCA may be an important real-world trial of pharmacogenomics technologies. It may also be a tough trial considering the two dueling drug makers behind the scenes.
“It’s a pretty big haystack,” said Michael Liebman, director of computational biology at the University of Pennsylvania Cancer Center. Yet Liebman, a veteran of Roche and Wyeth, sees some pharmacogenomics potential: Assuming J&J can obtain genetic data from those patients who experienced PRCA, and “if you eliminate all [other probable causes] … then it will become a pharmacogenomics problem,” he explained.
A J&J spokesman said the company will, in fact, launch a pharmacogenomics study in this case, though he was unsure when. “It’s absolutely something we are looking into to help us understand this whole thing better,” he said. Amgen did not return phone calls seeking comment for this article.
Mary Relling, a member of the NIH’s Pharmacogenetics Research Network, told SNPtech Reporter there’s a likelihood the cases of PRCA were triggered by genetics. “There are very few genetic predictors of autoimmune reactions,” Relling said. “But, yes, it is possible to have germ-line polymorphisms that are correlated with susceptibility to an immune reaction.”
If the incidents of PRCA among those 180 European patients are “truly due to an autoimmune reaction — and I guess there’s at least circumstantial evidence that that’s true — it’s certainly possible there could be a germ-like adverse event that some patients develop it and others don’t,” Relling explained. “And that germ-line predisposing factor could be directly interacting with [Eprex]; and, for some reason, it is at a different frequency at a European population than at a US population.
Relling also doesn’t rule out polypharmacy. “A lot of the things we see that are drug reactions that are certainly and appropriately linked to one major drug are then subsequently linked to ... an interaction with additional medications or additional therapies,” she said.
Relling, who is also a researcher at the department of pharmaceutical science at St. Jude Children’s Research Center in Memphis, Tenn., has studied drug-induced secondary tumors, which she said have been linked by pharmacogenomics studies to topoisomerase-2 inhibitors, widely used chemotherapeutic agents. But, she stressed, not all patients given topo-2 inhibitors develop secondary tumors.
“Our question has been ‘Why?’” she said. “And, what we’ve found is that when you take patients that have a specific germ-line pharmacogenetic polymorphism who take another agent, thiopurine — which isn’t normally thougt to be very carcinogenic — plus the topo-2 inhibitor, then you can identify a subset of patients who are at high risk” for secondary tumors.
This relationship might also play out with erythropoietin: Patients in J&J’s European Eprex cohort who experienced PRCA were being treated primarily for chronic renal failure, a disease that requires broad polypharmacy.
According to Penn’s Liebman, end-stage renal disease “is a state that can be produced by up to 40 different diseases. Maybe some complex that’s causing this kidney disease” is also triggering the PRCA, he said.
“It’s certainly possible that some of the other [types of] supportive care that are used in Europe are different than what is used in the United States,” agreed Relling. “Although I’d imagine that someone is trying to look at that pretty carefully.”
So what’s a pharmacogeneticist to do? “What you’d really like to do is completely sequence all those people,” said Relling. “And even though you may say money is no object, that is too expensive for anybody to do.” So one does the next best thing.
“You use a combination of a target-gene approach … and maybe some kind of genome-wide scan,” said Relling. “That may lead me to a list of, say, something between 10 and 100 target genes. And I would very carefully assess the genetic status by sequencing or by some kind of genotyping method.”
Relling said pharmacogenomics “has a shot” in this case. But she cautioned, “given what appears to be a real difference in the frequency of [PRCA] based on either different formulations or geographical areas, it would lead me as sort of a pharmacoepidemiologist to look at those treatment-related factors very, very strongly” as opposed to host-related factors. After all, she said, Europeans and Americans share much of their genetic constitution.
But “in terms of finding all of the etiologies that underlie [PRCA] it is certainly reasonable to look at genetics in general,” she said.