FDA official defends integrity of prasugrel review

Responding to criticism, FDA official Tom Marciniak has defended the methodology used in his review of prasugrel for the Cardiorenal advisory committee meeting in February.

Many CardioBrief readers will be aware of the long and complicated history behind this controversy. (Click here for a detailed chronology.)

This latest episode begins with our story last week, which summarized Victor Serebruany’s Cardiology editorial. The editorial used data from the FDA briefing document to criticize the TRITON-TIMI 38 trial. We posted a response from Lilly and Serebruany’s rebuttal. We then received an off-the-record communication that was highly critical of the FDA review, alleging that the MI data in the review was erroneous, based on a flawed method of looking at the case reports. It also appeared that similar criticisms of the FDA review had been circulating informally within the cardiology research community.

We then forwarded the criticism to Marciniak, who provided the following response:

“The statistics on site-reported MIs provided in my review included in the public briefing materials for the FDA Cardiovascular and Renal Drugs Advisory Committee meeting on February 3, 2009, are correct and not subject to the problem alleged in the response you quote. The TRITON adverse event forms do not have a checkbox for MI–to identify the type of event they only have the free text field for the event description. There is a checkbox for MI on a cardiac ischemic events form to indicate that the event should be considered for MI adjudication. There is also a field on that form providing the AEID number that links to the corresponding adverse event form entry. All but one of the MI cardiac ischemic events have a link to a valid adverse event entry and all but two of the linked adverse event entries have the free text event description completed. My search of these free text fields was far more comprehensive than searching “for the word MI” as your response alleges. Contrawise, while the free text entries of MI adverse events are virtually complete, the records of clinical histories, serial ECGs, and serial biomarkers used for MI adjudication are frequently incomplete. Anyone who alleges that my site-reported MI statistics are inaccurate or do not provide valid information should answer the following two questions: (1) If my statistics are wrong, why haven’t the investigators published the correct statistics rather than listing “Site Reported Myocardial Infarction” in a table with the disqualification “CEC-adjudicated events . . .” buried in a footnote? (2) Why do the site-reported MIs correlate much better with mortality than the adjudicated events?”

One knowledgeable expert who did not wish to be quoted emphasized to CardioBrief that even the Serebruany and FDA analyses found a beneficial effect for prasugrel in the reduction of MI. The debate, then, is over the degree of benefit, and the precise calculations of the risk/benefit equation.

Marciniak also clarified to CardioBrief his position regarding the Serebruany editorial. In the acknowldgement section at the end of the editorial, Serebruany thanked Marciniak and another FDA official, Norman Stockbridge, for “constructive dialog.” Asked whether he agrees with the conclusions of the editorial, Marciniak wrote:

“On a drug under review we can not say anything more.  The acknowledgement is a statement of fact, not an endorsement.”


  1. […] FDA official defends integrity of prasugrel review […]

Speak Your Mind