FDA HAP/VAP Discussion March 30

Image via Wikipedia

As I noted in my last blog, I did have a conversation with the FDA on March 30 regarding their draft guidance on trial design in HAP/VAP.  For my meeting, I constructed a series of talking points.  The FDA team asked me to submit the document to the docket which I did.  Because the submission has not yet appeared on regulations.gov, I reproduce it below. I tried to summarize key issues identified almost uniformly in the docket submissions and to provide solutions going forward.

•  NI Margin – as this becomes more relaxed – lots of good reasons to do this – everything else gets easier.
• Please provide numbers required for studies where mortality is lower than 20%.  (they provided an estimate which was, as expcted, staggering).
• APACHEII now underestimates mortality and scores higher than 15 will be required to assure mortality greater than 20%.
• Proscription against antibiotics – most commonly cited problem for infeasibility and probably unnecessary.  Also will exclude those patients you most want to include.
•  Mortality as an endpoint is insensitive and irrelevant to clinicians treating the disease.
• VAP is a disappearing disease – probably related to better care plus significant recent underreporting in the US under pressure from CMMS reimbursement guidelines.  Therefore, special consideration should be given to the fact that there is a small and shrinking population available for study even though the medical need in this population is greatest.  SEE J&J DOCKET SUBMISSION.
•   Inclusion of HCAP per Cubist?

SOLUTIONS

• 1.     Admit that diagnosis is difficult and a perfect rationale for the NI margin in this disease does not exist nor does a perfect AND feasible trial design and go on from there. Allow the use of endpoints - mortality, OR clinical outcome OR clinical outcome plus survival at 28 days as a composite.
• 2.     Allow the use of PK/PD measures to justify a margin for clinical benefit as determined by extrapolations to no therapy and/or comparing inappropriate to appropriate therapy.
• 3.     Increase the margins to something more reasonable – 15%.  Forget the odds ratio method constraint.
• 4.     Allow the use of up to 24 hours of an antibiotic at the time of enrollment.
• 5.      Discard the proscription against any antibiotic within the previous month.
• 6.     Consider the inclusion of HCAP patients where HAP/VAP pathogen is documented.

The FDA seemed receptive.  I think they were a little surprised by the extent of the response.  They wondered why the issue of feasibility was not discussed at the advisory committee almost two years ago.  I reminded them that it was, by Steve Barriere.  It is clear that on the one hand, they are driven to provide guidance by the IDSA, by industry and by congressional mandate.  But it is also clear that they are so removed from the industry they regulate and from medical practice that they do not understand what is feasible and what is not in today’s world of medicine and clinical trials. I would add that they seem to understand that not developing new antibiotics for Americans is a bad thing.  When I told them that for HAP/VAP I was advising my clients to go to Europe first, they cringed.  They said, tell them to talk to us.  We’re much more flexible in person than we are in our guidance – or words to that effect. In checking with friends involved in such discussions, it seems clear that this is not the case.  I think maybe the top management believes this is true, but on the ground at the team level, the guidance is the guidance.  That has always been my experience in any case.

On the one hand, I believe that the FDA agrees that issuing guidance with infeasible trial design requirements is at best counterproductive.  I think they feel like they made a mistake at least with HAP/VAP if not also with CABP.  Even though our discussion did not touch on the origins of their motivation for proposing specific design considerations, I think that the FDA may feel trapped between industry, IDSA and others on one side, and groups like Public Citizen and Tom Fleming along with the Grassley – Markey crowd on the other.

I fault industry for not being clear and vocal enough during numerous workshops and AIDAC meetings to make clear to the FDA what would and would not be feasible.  Industry has been too timid in their approach to the FDA in this regard. This is especially true when you consider the agency’s naïveté around considerations of feasibility.

Finally, I think it is time for industry to start to provide a new approach to trial design for consideration by the FDA. More on this in the future.