Last week, AstraZeneca received some bad news from an FDA advisory committee, which voted 14 to 3 not to approve a respiratory syncytial virus (RSV) drug called Rezield, or motavizumab, for high-risk infants due to side effect and risk-benefit concerns. The drug is a planned successor to Synagis, a $1 billion-a-year blockbuster that loses patent protection in 2015. And if Peter Roelvink is to be believed, public health officials may breathe easier over the panel's decision.
Here's why: Roelvink joined AstraZeneca as a director in its Infectious Disease department shortly before the drugmaker, which purchased both drugs as part of its 2007 acquisition of MedImmune, received a complete response letter from the FDA. He had been assigned to speed the biologics license application and, later, review RSV nasal wash samples that were obtained from patient studies and had to be analyzed in response to the FDA's request for more info.
However, Roelvink soon learned of a discrepancy between samples tested in different locations and, after investigating further, he discovered that inappropriate filters had been used by a technician to distinguish viruses from 111 patient samples. Yet his supervisor allegedly did not intend to submit the conflicting data to the FDA, according to a lawsuit filed by Roelvink, who claims he was terminated wrongfully for objecting to the decision to include incorrect results. As the lawsuit describes it, he suffered retaliation because he refused to "sweep problems under the rug," and blame the test-kit manufacturer.
"For reasons not made known to Roelvink, (his supervisor) said that not all differing results would be reported to the FDA. Roelvink knew or reasonably believed that the failure to report all test discrepancies would constitute false reporting in violation of FDA law, and indicated this" to his supervisor, according to the lawsuit, which suggests "serious concerns" over the data in the original BLA. Roelvink's staff, by the way, later found 58 percent and 65 percent of the samples from two different sample groups were incorrect.
By August 2009, he was fired for "the flimsiest reasons," according to the suit, which states his supervisor alleged he was not up to being a department director. Only five months earlier, he received a $19,000 bonus. And so he is seeking $5 million in compensatory damages and $5 million in punitive damages. We are awaiting a reply from AstraZeneca (see below for the reply).
UPDATE: On Tuesday, June 8, at about 10:30 am, AstraZeneca conveyed this response: "Bahija Jallal, MedImmune’s Senior VP of Research and Development, provided (this statement) at last week’s advisory committee meeting: MedImmune is confident that the data submitted to the FDA and presented to the Advisory Committee are accurate. The pending lawsuit contains unsubstantiated allegations by a former employee. MedImmune is confident that all data submitted for review for this application are true and accurate."
8 Comments
My guess is that in this case a central laboratory was not used. A central lab should be used for all clinical trials, even small ones because a central lab will use a single validated protocol that will use the same size filters on all samples. There may be other possibilities, but if that was the case here it was an avoidable error. Even if separate labs were used the protocols should have been validated and submitted to the sponsor for review prior to the study start. That might have also caught the problem. At the same time I would say that most lab validation studies are more likely to wind up serving as a coaster for the sponsor's apres dinner brandy than be actually read.
It appears Astra Zeneca once more is doing everything possible to live up to those many CIA's they have signed; or is this just another episode of "As The Pharmaceutical Industry World Turns' part VI "More Smoke and Mirrors"?
As it's said -BUSINESS AS USUAL-
If this is true it may preclude this drug from ever being approved.
FDA has a rule that implemented after the generic drug scandals that any falsification of data in a submission will result in the withdrawal of marketing approval. This rule was implemented via the public comment mechanism and so according to my understanding has the effect of law.
Falsification includes the failure to submit material information.
Salmon
"Falsification includes the failure to submit material information."
Interesting......
Seroquel anyone?
SO, he was wrongfully terminated for upholding the law. Could he end up being classified as a whistleblower now? Would the DOJ like to tell us why CIA's and AstraZeneca just don't seem to gel?
Was the $520million dollar hand slap for the Seroquel scandal not enough to cause the FDA to investigate AstraZeneca and it's business practices?
Is anyone ever going to do prison time for these behaviors? knowing and hiding and burying data all just seem par for the course, which is NOT acceptable.
Just how corrupt can one company be!?
I know let's play a game, which company IS the most corrupt? Do tell.Seems to be the goal besides the income and profit.
Ah AZ! Had to get out because I couldn't stand it anymore! I think that it is not a place hwere truth, honesty and integrity are welcome. I hope they prosceute the offenders, but I doubt it.
Salmon, you are correct about withdrawal of ANDA approval if subsequent falsification is found. However, 80% of the profit on the first ANDA is made in year one post-approval. Conservatively speaking it would take that long to discover the problem. Let's say for example that the first ANDA on Zyprexa makes about one billion dollars in year one sales, not out of the realm of possiblity. Based on the level of fines paid in the 1980's generic drug scandal (generally less than five million) the resulting fine would probably not exceed 1% of sales. Still a good ROI from the company's perspective. Remember that in that scandal, the only person who wound up incarcerated, if I'm not mistaken, was actually from the FDA.
The policy also applies to NDAs. There the profit comes later.
Salmon