Posted 21 March 2016
By Zachary Brennan
More than a dozen drug developers, including Pfizer, Genentech, Sanofi and GlaxoSmithKline, and industry groups are calling on the US Food and Drug Administration (FDA) to ensure that a clinical outcome assessment (COA) pilot program includes more useful information for researchers.
The pilot project, first proposed by FDA in January, is intended to help industry with a starting point for considering how certain COAs might be used in clinical trials and early drug development.
FDA proposed that the online compendium includes six columns of information for at least 20 treatments for cardiovascular, ophthalmology, psychiatry, oncology and other conditions. The proposed columns would include information on:
- Disease/Condition, which includes any relevant FDA disease-specific guidance.
- Indication and/or Claim(s) Description, which lists key elements, either labeled or qualified, though for ongoing COA qualification projects, targeted labeling or promotional claim(s) may not be yet known and may be described as to be determined.
- Outcome of Interest, which describes what was assessed (labeled) or could be assessed (in a qualification program) by COAs.
- COA Type, which lists a labeled, qualified, or ongoing qualification project clinical outcome assessment name and/or description, and includes the clinical outcome assessment type (i.e., a patient-reported outcome, observer-reported outcome, clinician-reported outcome, or performance outcome).
- COA Context of Use, which describes circumstances under which the outcomes of interest and the COAs that have been used (i.e., labeled) or are targeted for use (i.e., they have been qualified or are part of an ongoing qualification).
- COA Qualification Information, which lists ongoing and completed COA qualification project information, if applicable
FDA warns that the list is not comprehensive or intended to replace either existing disease-specific guidance or key interactions with FDA concerning drug development, such as pre-IND meetings. The COA also shouldn’t be misconstrued as the sole (or primary) determinant of a clinical benefit in a clinical trial.
Industry group BIO set the tone for other companies’ comments, as, like others, the group generally supports the pilot project, but notes that for most COAs, it “provides neither sufficient information nor guidance to aid in the selection of appropriate endpoints for clinical trials. Furthermore, COAs, even if listed in the COA Compendium, will still need to be validated for a specific context of use, a process which will involve consultation with and advice from the relevant Office of New Drug review division early in drug development.”
The industry’s other major lobbying group, PhRMA, meanwhile calls on FDA to provide “more insight into the regulatory perspective on existing COAs” that might be able to expand a COA’s context of use or make it “fit-for-purpose in another therapeutic area.”
PhRMA also encourages FDA to explore new ways to use COAs that are not included in the pilot version of the COA Compendium, particularly COAs that FDA believes to be no longer fit-for-purpose or otherwise inappropriate for ongoing or planned drug development programs.
As far as the individual companies, Roche subsidiary Genentech calls on FDA to ensure that subsequent iterations of the compendium make a clear distinction between COAs intended to support conversations between sponsors and the agency on future drug development programs, and those that are intended to encourage collaboration in the development of clinical outcome assessments for unmet measurement needs, such as those that are undergoing DDT (drug development tool) qualification.
Pfizer’s Carol Haley, director of US Regulatory Policy, took a more critical position on the proposed pilot, saying that if it were to proceed as planned it would serve “mainly as a convenient organization of information that is already publicly available and can already be accessed and analysed by drug developers and the research community.
“While it is helpful to have all this information in one place, much of the information will not provide sufficient transparency to inform sponsors' internal decision-making about development prior to discussions with the agency, particularly in the absence of a procedure for having discussions with the agency about the use of COAs,” Haley added.
GSK also warned that the COA Compendium “may have the unintended consequence of discouraging the development of novel measures,” that could improve on older COAs, including those which may lack sufficient evidence that they are well-defined and reliable.
Botox manufacturer Allergan called for the inclusion of more granular information in the various columns, and Sanofi adds that in order to understand how COAs have been used, it is important to have information on “how the endpoint was constructed, how frequently the COA was administered and how the COA data was analyzed.”
Clinical Outcome Assessment Compendium Comments