After a multi-year gestation period (my notes on this project go back to 2012!!), a comment on the FDA draft HABP/VABP guidance document has now been submitted by the FNIH (Foundation for the NIH) Biomarkers Consortium HABP/VABP Project Team and posted online. Links to the docket and document are below my signature.
The document is substantial and you should plan to review it in detail if you are interested in studies of Noscomial Pneumonia. I have excerpted below my signature the summary recommendations.
At a high level, the key points are that (a) VABP and ventilated HABP may be studied together, (b) non-ventilated HABP is meaningfully different, (c) all-cause mortality at day 14-28 remains a valid endpoint but also (d) the Standardized MedDRA Query (SMQ) for Toxic/Septic Shock can be used to create a “mortality plus” endpoint that incorporates SAEs & AEs from that SMQ.
Interestingly, it turned out that non-ventilated HABP seems amenable to study with the same symptom-based tool that we use for CABP.
Frustratingly (and despite its obvious biological relevance), the available data on improvement in oxygenation were inadequate to support meaningful analysis. We were all eager for a way to validate this as an endpoint, but the required data were not collected in sufficient detail in the studies avaialble for analysis. The project team recommends that such data be gathered in future trials and perhaps this can be revisited.
Overall, this docket submission is expected to be a substantial boost to finalization by FDA of its guidance on conducting HABP-VABP trials. Establishing the acceptable performance of the “mortality plus” strategy is important as this helps ensure an event rate high enough to allow to allow use of the fixed 10% NI margin recommended for HABP-VABP trials.
It’s a work of many hands, but I would be remiss if I failed to say that the entire commmunity owes thanks to George Talbot for his exemplary leadership of this project. Thank you, George!!
All best wishes,
John H. Rex, MD | Chief Medical Officer, F2G Ltd. | Chief Strategy Officer, CARB-X | Expert-in-Residence, Wellcome Trust. Follow me on Twitter: @JohnRex_NewAbx
The FDA docket:https://www.regulations.gov/docket?rpp=10&po=0&D=FDA-2010-D-0589
FNIH’s submission in the docket:https://www.regulations.gov/document?D=FDA-2010-D-0589-0027
Summary of Recommendations
1. VABP and ventilated HABP appear to be similar syndromes that can readily be studied together. Patients with non-ventilated HABP have a lower mortality rate, and this difference should be anticipated if a study seeks to pool data across all three categories of nosocomial pneumonia.
- Enriching clinical trials for HABP and VABP with older patients who have higher APACHE Il scores will increase the rate of ACM. When this enrichment cannot be performed, the ACM event rate may be low enough that a fixed NI margin cannot be justified and an odds ratio approach to analysis may be needed. To avoid this statistical eventuality, sponsors could pre-specify a “mortality-plus” primary outcome parameter by incorporating SAEs and AES from the Toxic/Septic Shock SMQ. The same NI margin can be used for analysis of a 28-day ACM endpoint as for a 28-day endpoint comprising ACM plus AES or SAEs as defined by the Toxic/Septic Shock SMQ.
- For studies of non-ventilated HABP, the data demonstrate that patient symptoms can be collected and occur with a high enough frequency to allow an endpoint similar to that for trials of CABP. Analysis of resolution of symptoms at Study Days 5-7 appears reasonable. Since the data on symptom resolution in non-ventilated HABP support development of a patient-reported outcome (PRO) instrument, which is currently ongoing, the current FNIH effort will not pursue validation of a symptom-based endpoint in nonventilated HABP.
- Sponsors of future trials in these indications should collect the following data:
- For VABP and ventilated HABP, oxygenation and ventilation, especially positive end-expiratory pressure settings;
- For non-ventilated HABP, symptoms and symptom resolution until Days 7—14; and
- Prior antibiotic usage.
- If mortality-plus is not used as a primary endpoint, sponsors of future trials in any of these indications should include secondary analyses of the utility of the mortality-plus endpoint based on AES and SAEs from the Toxic/Septic Shock SMQ.