A Pfizer Clinical Trial And Unanswered Questions
12 CommentsBy Ed Silverman // April 29th, 2011 // 10:13 am
The big drugmaker yesterday released results from two eagerly awaited Phase III clinical trials for tofacitinib, a rheumatoid arthritis treatment that Wall Street bets will become a big seller and, significantly, compensate for a string of laboratory failures, not to mention a growing number of meds losing patent protection.
However, the results failed to answer a few key questions, at least according to one analyst. “Although the press release states that both studies met all their primary endpoints (beat placebo) at both doses of tofacitinib (5mg and 10mg), the press release doesn’t give us the critical pieces of information we are interested in, ” writes Wells Fargo Securities analyst Larry Biegelsen, in an investor note. “Unfortunately, the press releases did not include any key data” (read the press release).
What is he talking about? One study, which is called Oral Standard, compared both doses with every-other-week, weekly 40mg doses of Humira, a big-selling RA med from Abbott Laboratories. All of the patients, by the way, were also given methotrexate. The failure to offer comparative data was irksome, according to Biegelson. “That was obviously the purpose of the trial - to do a head-to-head comparison,” he writes. “If tofacitinib is even numerically but not statistically significantly inferior to Humira, it is likely to be perceived as a small positive for Humira in our view.”
Then there were results from the other study, called Oral-Step. This found that the 5mg dose of tofacitinib was significantly better than placebo in reducing RA signs and symptoms of RA when measured by a score known as ACR20 score. But Biegelsen points out that there was no mention of the ACR50 and ACR 70 scores, which are more important to clinicians.
Biegelsen goes on to complain that Pfizer included “a boiler-plate statement about safety…which leaves room for possible surprises when the actual data are presented at a medical meeting in the future.” But, he cautions, “in light of four deaths, only one of which was related to the drug, that were reported last week from another trial, “we think it would have been helpful if Pfizer had specifically addressed whether there were any deaths, and if so, probable cause.”
Another analyst, Sanford Bernstein’s Tim Anderson, took a slightly more benign view. In his own investor note, he writes, “This was an active comparator study required for European regulatory review, and is not a true head-to-head study that will allow for a label claim of non-inferiority or superiority, but prescribers will still clearly draw inferences about the comparative profiles of the two drugs. Obviously what is missing are the details from these results: for example, was tofacitinib better, worse, or the same as Humira?”
He adds that, “Many will rightly question what the safety profile of tofacitinib looks like in these new trials. While there was no mention of things like deaths, rates of serious infections, or heart failure, if there was an obvious imbalance this would have been called out…We continue to believe it is likely that tofacitinib will receive regulatory approval. What remains less clear is how exactly the label will read in the RA setting, but PFE is running a development program that theoretically will allow for the inclusion of all lines of therapy in the product label.”
Hat tip to the Medicine Show
Dan
Note that the comparison was to Humira every OTHER week, not weekly as the text above suggests.
ed silverman
Hi Dan,
Thanks for the note and pointing that out. Please note that I have corrected that fact.
Much obliged,
ed
Insider
THe fact they left all that information out does not surprise me. Now wait till they try to spin the results without the use of the data like they did with the Class study and Celebrex. THey made all kinds of superiority claims from the Class study that were using secondary endpoints finally correctd in the 2005 package insert. Then the Class study disappeared from our detail folders never to be heard of again.
industry insider
When I first started directing clinical trials, my boss said to me that with regard to efficacy. “the only thing that I care about is that the P-value is less than 0.05. Leave it to Marketing to do the rest”.
That’s largely still true today.
Jeff
Ed-
Nice summary of the different takes about this situation. The tofacitinib story has many complexities tied to whether this product will succeed. The safety signals are one issue (and I still want more details around the TBI). I also think we need more definitive data about tofacitinib’s ability to work in anti-TNF failed patients, particularly given that this is where it will likely earn its living during launch–prior to potentially achieving first-line utilization over biological DMARDs.
Insider
Dear Industry insider,
As long as the P valuse is really less than or equal to .05 that is fine and ther research followed the protocol. BUt often the marketing people just make it less than .05 when it is not which is what makes the sales rep less than a carpet salesman, their current status. DOJ has not been prosecutng misrepresenting data yet but they should.
industry insider
Insider, I agree. I’m a staunch opponent of data dredging and post hoc analyses of any sort, and I have made that perfectly clear to all of my Marketing colleagues over the year. All claims must be supportable by primary efficacy data. When I said let Marketing do the rest I meant that if you present them with strong efficacy data in the first place they cannot get in trouble. If you give them marginal efficacy data, as seems to be so often the case these days,then they will be tempted to “turn chicken s**t into chicken salad”, and they should be taken to task for theie efforts.
Salient point
industry insider-Then what do you do in the situation described in this post? ACR20 is the primary endpoint, presumably because it’s been established as the standard by FDA. But the medical community is clamoring for different data, ie, ACR50 & ACR70. Do you not share these data, even if available? Do you conduct additional research with those benchmarks as the primary endpoints, at the cost of years & millions of dollars (which get billed back to the consumer)? At what point do the specifics of the clinical trial design interfere with the physicians’ right to know?
Suzanne
ACR20 is a 20% improvement. People are definitely going to want to see the ACR50 and ACR70 data, unless they are already completely out of options.
industry insider
What I believe you are looking for from Pfizer is something like Table 2 in the Humira Package Insert, which would answer these questions.
http://www.rxabbott.com/pdf/humira.pdf
Salient point
industry insider-Thanks. ACR20 is the primary endpoint for the RA trials per the EPAR–the Humira PI includes all the ACR data without explicitly stating primary endpoints, which is a little strange.
The point I was trying to make is that the medical community would benefit from exposure to data beyond primary endpoints, which are often more interesting to FDA than to clinicians. But upon re-reading yours & others’ comments, it seems that the real objection is with post-hoc analyses & other forms of clinical trial data monitoring, which is admittedly harder to defend.
Animity
Sounds like another case of “P-ing” all over the place.