Full article reprinted from "The Pink Sheet" February 10, 2009
FDA's handling of the advisory committee meeting for Lilly/Daiichi Sankyo's platelet inhibitor prasugrel is the first signal in some time that the agency is ready to approve a primary care drug to be used in large patient populations with relatively limited restrictions. Read more...
Full article reprinted from "The Pink Sheet" February 10, 2009
FDA's handling of the advisory committee meeting for Lilly/Daiichi Sankyo's platelet inhibitor prasugrel is the first signal in some time that the agency is ready to approve a primary care drug to be used in large patient populations with relatively limited restrictions.
If prasugrel is approved by FDA as expected, it would be the first approval of a new molecular entity into a primary care market with multi-billion dollar potential in more than a year.
The class of 2008 was represented by novel products in niche categories or crowded markets -lending credibility to the argument that the traditional blockbuster was indeed dead (1"The Pink Sheet," Jan. 12, 2009, p. 12).
Prasugrel (to be marketed as Effient) could breathe new life into the aging blockbuster model, providing healthy competition to Bristol-Myers Squibb's blockbuster Plavix (clopidogrel), which is the second-highest selling drug in the world (after Pfizer's Lipitor) with annual sales of $8 billion.
Analysts predict prasugrel could encroach on 15 percent to 20 percent of Plavix's current market, and garner sales of $1.3 billion (which Lilly would split with partner Daiichi Sankyo). That's nowhere near the numbers that Plavix pulls in, but is a healthy position for a traditional primary care drug.
A Drama-Free Meeting
Despite the drama that surrounded prasugrel, FDA came to the Feb. 3 advisory committee meeting with a clear recommendation of approval. That drama resurfaced, however, in the aftermath of the meeting in light of FDA's last-minute decision to remove an empanelled member.
The panel agreed with FDA's recommendation in a 9-0 vote - a green light that followed months of speculation over the fate of the blood thinner due to risks of major bleeding and a cancer signal. Prasugrel has been in limbo since FDA missed its revised Sept. 26 user fee deadline (2The IN VIVO Blog, Sept. 29, 2008).
But comments by advisory committee members throughout the meeting demonstrated the panel's comfort level with prasugrel's benefit-risk profile - so much as to recommend that FDA tone down its proposed safety warnings in labeling.
In the end, what they had to say may persuade FDA to keep warnings in the label for prasugrel modest - a move that would broaden access to the substantial acute coronary syndrome market in which prasugrel would compete.
There Will Be Blood
Lilly took an enormous risk in its Phase III program for prasugrel: it pitted the drug head-to-head against Plavix, the gold standard in platelet inhibition. And it paid off: the pivotal TRITON trial clearly demonstrated prasugrel's efficacy profile, especially in the prevention of non-fatal myocardial infarction.
But it also increased the risk of major bleeding, and there was a potential, although not substantiated, risk that prasugrel stimulated the growth of existing malignant tumors. Drugs have been taken down for less than that, but to many in FDA, prasugrel's strong efficacy profile outweighed those risks.
FDA clearly stated in briefing documents to the advisory committee that the risk was a "reasonable trade" for the strong cardiovascular efficacy profile: for every 1,000 patients treated with prasugrel rather than clopidogrel there would be a net reduction of 22 cardiovascular events (20 myocardial infarctions and two deaths) against two fatal bleeding events, four life-threatening events and an overall excess of five major bleeding events (4"The Pink Sheet" DAILY, Jan. 30, 2009).
Discussion of the bleeding risk took an interesting turn when several panel members pointed out that prasugrel's rapid onset of platelet inhibition could present another advantage over clopidogrel in the setting in which bleeding in both treatment groups was much more frequent: following coronary artery bypass graft surgery.
Those members suggested that using prasugrel would buy doctors time to conduct an angiogram to view the patient's coronary artery and determine whether CABG surgery is actually needed before initiating prasugrel, whereas initiation of clopidogrel must begin more immediately.
"With clopidogrel, you really have to decide that you're going to pre-treat them ahead of time," said committee member Michael Domanski, National Institutes of Health. "The rapid onset here actually gives you the option of at least seeing the anatomy before you commit yourself to [CABG surgery]. Prasugrel offers a big advantage here in terms of the usual patient we see."
In terms of withholding prasugrel in patients likely to need CABG surgery, panel member Mori Krantz, University of Colorado, said, "I think really this is in many ways a blessing ... in terms of looking at process of care, this manages to simplify it and create a better model for all our patients."
Acting committee chair Marvin Konstam, Tufts Medical Center, took a more balanced approach, noting that although pre-CABG initiation is the common practice with clopidogrel, "we've clearly seen that with prasugrel, the stakes have gone up with regard to inter-operative and peri-operative CABG bleeding - it's substantially worse with prasugrel than clopidogrel."
The committee ultimately agreed that because of higher bleeding risks in certain populations, labeling should discourage use of prasugrel in patients with a history of stroke/transient ischemic attack or in whom stroke/TIA developed during treatment, and that labeling should caution of increased risk in the elderly and patients under 130 pounds.
Cancer Signal Weak With Retrospective Analysis
Prasugrel also cleared its other significant risk hurdle during the advisory committee - cancer. FDA concluded that prasugrel did not cause cancer but could stimulate the growth of existing tumors, and asked the panel how that risk should be addressed in labeling.
FDA presented three options of varying severity to communicate the risk: (1) inclusion in the Warnings and Precautions section of labeling; (2) a "black box" warning; (3) or, the most serious option, that use be restricted "for a limited time," which some FDA reviewers thought should be defined as one week (5"The Pink Sheet" DAILY, Feb. 3, 2009).
But the committee unanimously chose "none of the above," advising FDA instead to merely mention the cancer risk in the "Adverse Reactions" section of prasugrel's label.
The panel's comfort seemed to stem from the rejection of FDA's cancer risk analysis by panel member Jim Neaton, University of Minnesota, following the agency's presentation. "I don't think, given the information that we've heard, that I have the same [cancer] concerns," Neaton said.
Neaton agreed with Lilly's argument that ascertainment bias could explain the higher number of cancer deaths in the prasugrel arm. TRITON subjects were not screened at baseline for cancer because the trial was not designed to ask or answer questions about cancer risk. When patients had bleeding events - which occurred more often in the prasugrel arm than the clopidogrel arm - more were evaluated for, and thus diagnosed with, cancer.
FDA asked Lilly to follow up with trial investigators as to the vital status of the subjects that had been diagnosed with a new cancer during the trial in order to find out the mortality rate in the prasugrel arm compared to the clopidogrel arm. The final tally was 33 compared to 21, respectively.
From those data, Deputy Director of the Cardiovascular & Renal Products division Ellis Unger told the panel that the relative risk of cancer for prasugrel over clopidogrel across the trial was the same if you subtracted those subjects diagnosed because of a bleeding episode.
Thus, "we don't buy ascertainment bias as accounting for the imbalance in cancers. What we can say is that bleeding led to the diagnosis, but it didn't account for the imbalance."
But Neaton said FDA's analysis has limitations. "I can't interpret that relative risk at all," he said. "The problem with the relative risk is that you don't know anything about the cancer deaths among the people that were not newly diagnosed."
Acting committee chair Konstam agreed, noting that he was "much more relaxed about [the cancer association] after that discussion."
"Once you start going beyond the pre-specified period of the trial and the data-set lock and you are preferentially following those patients who have an adverse event identified during a trial that could have been on an ascertainment bias and could have been by chance, it would seem natural that you would wind up having more deaths in that group as well," he said.
In terms of labeling, Konstam said "I agree with none of the above - no warnings or precautions or any of those things. ... I don't think it deserves anything more than a mention."
"I would make it as inconspicuous as you are willing to," Domanski added. And Neaton agreed: "I would say very little about it because whatever you say is probably going to be wrong."
- Jamie Hammon
Sign-Up for a free trial to "The Pink Sheet" today.
