Day 1 of the FDA Avandia Panel- Live Blogging the Morning

The afternoon will be blogged in a separate post.

Lunch break until 1 PM.

12:05: Ungar: to answer the malfeasance question we may have to audit the entire trial

12:02: Kaul: Did Marciniak unearth malfeasance in RECORD? How many missing MIs in RECORD were biomarker-only or clinical events?

11:59: Ungar says he is concerned about the overall data quality in Avandia trials. Mortality is reassuring, but he agrees that the followup issues raised by Marciniak creates problems for this view. > This is his defense?

11:57: Ungar, agreeing for once with Marciniak, says he doesn’t like CV deaths in a trial that is open label.

11:55: Ungar not surprised by low rate of CV deaths in RECORD.

11:54: The hypothesis generated by Nissen metanalysis is not supported by RECORD.

11:50: Ungar not concerned if patients aren’t properly followed in RECORD if they’ve been off study drug for long time.

11:49: Ungar: one would have to adjudicate all the case reports in RECORD– this is his defense of the trial? The MI findings are inconclusive. Not reassuring but don’t support Nissen meta-analysis.

11:47: Ungar: ascertainment bias in RECORD means that only mortality is reliable

11:46: FiercePharma: Unger :Marciniak reinterpreted info, adding and subtracting endpoint events ; office doesn’t agree w/ this approach

11:41: Matt Herper: Ellis Unger seems more concerned with defending FDA, not Avandia.

11:39: Can we trust the sponsor of RECORD? Marciniak should not have reanalyzed RECORD.

11:38: Ungar: Marciniak is one tenacious reviewer. But…

11:36: Ellis Ungar now up. Will disagree with Marciniak’s conclusions. Here are his slides.

11:35: GSK disagrees with Marciniak. Says MI definition was prespecified.

11:33: Fleming: any concern that decision to exclude silent MI by GSK was based on  knowing the impact of the decision? Marciniak doesn’t know.

11:31: Fleming to Marciniak: so many issues, I wish we had more time!

11:28: Chairman to Marciniak: don’t just try to wrap up, wrap up!

11:24: Marciniak: RECORD investigator says no one knew when patients died!

11:22: Even if you’re not a Twitter person you might want to follow this on Twitter:

11:19: Marciniak: if I had been consulted in advance I would have rejected the design of RECORD out of hand

11:16: RT Pharmalot: An Avandia Study Author…Or So We Thought Steve Haffner, who leaked Nissen NEJM mss to GSK, had paper ghosted by GSK.

11:09: Marciniak ripping into case report forms in RECORD. Here are his posted slides.

11:07: Marciniak up. Was not involved with rosi until last fall. Says he’s biased in favor of new drugs but some would disagree. Also says that many in FDA would be happy to see Nissen disproved and were hoping his review would do that.

11:04: FDA’s Parks says most PPAR developments put on hold because of cancer, not CV, concerns

11:02: Nissen: since rosi approved 50 PPARs have been developed and all have failed! This is not a class effect.

11:00: Question about absolute versus relative risk. Very small absolute difference. Nissen: raw absolute #s not useful for calculating hazard ratios.

10:58: Question about disparity in CV mortality and skewed data. “I would be terribly cautious of subgroup analyses.”

10:57: Konstam asks Nissen if he’s worried about sulfas also. Yes.

10:44: Nissen: “this trial (RECORD) was not properly handled and should not be used for regulatory purposes.”

10:42: TheHeart.Org: Nissen slams RECORD for “extraordinary unblinding” GSK had unrestricted access to treatment codes. “Ever heard of that?”

10:39: Nissen really doesn’t like RECORD. Who knew?

10:37: GSK issues press release in response to senate finance committee.

10:31: Nissen: n 2006 GSK observational study did not include pio data… if it had the FDA might have acted.

10:25: Nissen: Several years later ezetemibe approved only for LDL decreases… so why approve Avandia if it raises LDL?

10:24: Nissen: Avandia should never have been approved in 1999. Increases in LDL were enough to stop it.

10:09 AM: Panel on break until 10:20. Nissen is next!

10:05 AM: Fierce Pharma asks: Will FDA experts get last-minute Avandia documents?

10:03: Question about dropouts in RECORD though Homes doesn’t like the term dropouts.

9:58 AM: Sanjay Kaul starts out the questioning. No surprise. Also no surprise it’s a technical question I don’t quite understand.

9:55 AM: GSK playing the glucose control card. Wow. I guess that settles it.

9:43: Grassley/Baucus letter to FDA leaders (including internal GSK documents mentioned in NY Times story)

9:37: Senate finance committee press release: Baucus, Grassley Find Company Failed to Promptly Alert FDA about Drug Risks

9:35: Good perspective on GSK defense by Matt Herper at Forbes.

9:03: Ramsey Baghdadi writes (on Twitter): “Gsk says most appropriate to focus on long term clinical trials when looking at cv risks b/c of limitations with observational studies.” I agree– too bad we don’t have any good reliable long term clinical trials!

9:00 AM: Brilliant GSK strategy: put the panel to sleep!

8:55 AM: As far as I’m concerned the REAL meeting starts at 9:45 when the committee questions GSK.

8:49 AM: GSK presentation seems a bit slow. Someone should tell them trying to run out the clock won’t work.

8:33 AM: GSK sponsor presentation up next. Here is a link to their briefing document. A lion in a den of Daniels?

8:30 AM: Talk about walking a tightrope. FDA’s Mary Parks trying to give an objective overview.

7:54 AM: The FDA panel is now underway. It looks very crowded! (I’m following remotely on I will try to refrain from reporting the boring stuff. Click here for links to FDA briefing documents.


  1. […] Filed under: Uncategorized « Day 1 of the FDA Avandia Panel- Live Blogging the morning […]

  2. […] drug Avandia is generating lots of buzz on Twitter and various health blogs. For a sample, try by longtime medical journalist Larry Husten, who’s posting updates with a touch of […]

  3. […] Blog of the 2 Day Panel (Part 1, Part 2, Part 3, Part 4) (July 13-14, […]

Speak Your Mind