Session 2: Advancing Measurement Strategies for Clinical Outcome Assessment Tools

Session 2 was chaired by Ashley Slagle, MS, PhD, COA qualification scientific coordinator, Study Endpoints and Labeling Development, Office of New Drugs, CDER.

The session was divided into several parts:

  1. Background on Qualification and Previously Labeled COAs: Ashley Slagle
  2. Overview of Compendium: Focus on Previously Labeled COAs: Ann Marie Trentacosti
  3. Industry Perspective on Standards and Compendium: Jean Paty
  4. PRO Consortium Perspective on Standards and Compendium: Stephen Joel Coons
  5. Panelists’ reactions

Only the presentations of Dr. Slagle, Dr. Trentacosti, and Dr. Coons are reported.


As an introduction, Dr. Slagle presented the pathways for the FDA to review and provide advice on COAs. She reminded the audience that there are two pathways. The first is the traditional way within an individual drug development program. Assessments reviewed and found acceptable in this context have the potential to support drug approval and a product-labeling claim. The second pathway is a newer process within the Drug Development Tool (DDT) qualification program. This is outside of any individual drug development program. Assessments that are reviewed and found acceptable in this context result in qualified assessments. This means that a notice is published in the Federal Register identifying the measure and the clinical context for which it is qualified for use. Tools are intended to be made publicly available for use across multiple drug development programs. In the future, the FDA anticipates there will be tools that are both qualified and in labeling.

Dr. Slagle reminded the audience that “qualification” is a regulatory term, and that it is not necessary that an assessment be qualified to be used to successfully support a drug-labeling claim. She also added that this term, qualification, should not be understood as an equivalent to validation.

She recalled that, for COAs, there are two types of qualification. COAs can be qualified for exploratory use or for use as a primary or secondary endpoint. In the first case, the FDA agrees with the content and there is cross-sectional evidence of some key measurement properties. A full understanding of the tool’s ability to detect change and interpretation of change is not yet available. By qualifying for exploratory use, the FDA makes the instrument publicly available and recommends that the instrument be used in exploratory settings which are typically Phase II trials in order to learn more about the assessment and obtain the needed information to be able to plan for its use in Phase III confirmatory trials. When all measurement properties are tested, evidence will be reviewed to support COA qualification for use in adequate and well-controlled studies as a primary or secondary endpoint measure to support labeling claims.

Then Dr. Slagle ended her address by presenting the continuum of confidence in COAs to support labeling claims with, on the far left, instruments for which there is no supporting evidence for labeling claim and very few regulatory experience and, on the far right, instruments that are qualified as primary or secondary endpoint. See Figure 1.

Figure 1.

The goal is to have confidence in what is being measured which should reflect what is meaningful to the patients. The idea is not to be left wondering whether the drug failed or was the COA insufficient to detect small, meaningful, interpretable changes. This is why, in the spirit of improving outcome measurement, the FDA is willing to discuss outcome assessment as early as possible in drug development.


Proposed Development of Compendium of Clinical Outcome Assessments - Ann Marie Trentacosti, MD, Study Endpoint and Labeling Development, Office of New Drugs, CDER


Dr. Trentacosti reminded the audience that qualification is a long and costly process and presented the goals of the compendium. Because it is unrealistic to qualify all COAs, the first goal is to identify other COAs that could potentially be used to support labeling claims. Another objective is to provide clarity and transparency about the FDA’s current thinking about COAs by making the compendium publically available. A third important feature of the compendium is to encourage development and implementation of COAs that are important to patients (such as PROs).  She then recalled the pathways of the FDA’s review of COAs, the continuum of evidence, and the differences between “previously labeled” COA versus “qualified COA.” See Figure 1.

Figure 1.

Dr. Trentacosti presented what she called the two compendium communication tools: the FDA compendium guidance and the compendium itself, i.e., a table that will be accessed through the FDA website. The guidance will be a collaborative effort between the SEALD and the review divisions. It will describe the selection criteria for inclusion in the compendium (e.g., scientific and regulatory standards for labeled COAs). The guidance will also discuss the limitations of the compendium. Public comments will be possible after the release of the draft. As the guidance, the compendium will also be a collaborative effort between SEALD and review divisions. It will be organized by therapeutic area and disease/condition. Its development will follow two stages. At stage 1, the compendium table will include qualified COAs, ongoing qualification projects, and recently approved labeled COAs, identified from new molecular entity (NME) labeling approved from 2003 and later. Stage 2 will be the expansion stage. See Figure 2.

Figure 2.

The table will be updated periodically in light of scientific and regulatory advancement as well as drug approvals.

Dr. Trentacosti provided two examples; one in the treatment of irritable bowel disease (constipation) and the other in the treatment of advanced non-small-cell lung cancer. See Figures 3 and 4.

Figure 3.

Figure 4.

The table was divided into six columns. The first one addressed the condition; the second, the indication or targeted labeling claim; the third, the outcome of interest; the fourth, the labeled COA; the fifth, the context of use; and the last one, the relevant disease-specific guidance or ongoing qualification efforts. The tables were intentionally short in details so as not to prevent ongoing communication with the industry. This is why primary and secondary endpoints were not included.

Dr. Trentacosti provided a summary of what the compendium is and is not. See Figure 5.

Figure 5.

She listed the limitations of the compendium. A comprehensive table of all COAs in medicine or drug development was not the goal of the initiative. The list was not exhaustive. There may be COAs that are highly relevant and potentially useful, even for the medical conditions listed, but they have not been identified. Alternate assessments that do not appear in the compendium could be considered as well. Some of the COAs listed in the compendium may be protected by proprietary rights.

Dr. Trentacosti ended her presentation with a reference to the next step of the project, i.e., the notification publication and request for comment in the Federal Register concerning the proposed compendium of COAs.


Patient‐Reported Outcome (PRO) Consortium Perspective on the Compendium and Qualification - Stephen Joel Coons, PhD, Executive Director, PRO Consortium Critical Path Institute


After a brief presentation of the mission of the PRO Consortium (i.e., development of qualified, publicly available PRO instruments for use in clinical trials) and the benefits and challenges of the qualification process, Dr. Coons presented the benefits and potential challenges of the compendium.

Benefits include enhancing transparency, encouraging consistency in COA endpoints, and lowering the risk and uncertainty for clinical-trial sponsors.

Potential challenges include the following:

- Are the COA tools sufficiently patient‐focused?

- Some COA tools listed may not be publicly available.

- Initial focus on COA tools rather than outcomes.

- Will it undermine the qualification program?

Dr. Coons stated that, in his opinion, the compendium will not undermine the qualification program.

As far as adding tools to the compendium, Dr. Coons hoped that legacy COAs will be accepted through clinical/scientific consensus (e.g., the myelofibrosis symptom diary).

To conclude, Dr. Coons said that a compendium of COA tools is an important first step and a valuable complement to COA tool qualification. It will help identify COA tool gaps and facilitate the identification of additional outcomes of interest, and will enhance patient‐focused drug development. Additions could be advanced by consortia, professional societies, advocacy groups, etc. However, the level of evidence needed for a sponsor to use a listed COA tool in a clinical trial needs clarity.