Section 907 of the Food and Drug Administration Safety and Innovation Act of 2012 (FDASIA) directed us to develop an action plan outlining recommendations for improving the completeness and quality of data submitted to the FDA that support the agency’s approval of medical products. Specifically, FDASIA requests that the action plan focus on how to improve inclusion of clinical trial data representing demographic subgroups (i.e. sex, race/ethnicity and age) on product labeling and public information sources. FDASIA also requests that the action plan indicate the product types affected by each recommendation set forth.
To address the law, we wrote an action plan with 3 overarching priorities:
- Data Quality - improve the completeness and quality of demographic subgroup data collection, reporting and analysis;
- Subgroup Participation - identify barriers to subgroup enrollment in clinical trials and employ strategies to encourage greater participation; and,
- Data Transparency - make demographic subgroup data more available and transparent.
However, we are not legally mandated to address all of the issues involving demographic subgroups in the action plan. Below are responses to some common comments and questions that we’ve received from stakeholders, but couldn’t address within the action plan.
1. Does the FDA require that animals of both sexes be included in non-clinical studies?
Yes, we expect product sponsors to include animals of both sexes when conducting non-clinical studies for drugs or biologics. Sponsors should also document information such as age, sex, the source or laboratory where the animals came from, and the immediate source of any cell line and its origins. Additionally, sponsors should record the animals’ characteristics per Good Laboratory Practices (GLP) regulations (21 CFR part 58 and ICHM3(R2)). For animal studies of medical devices, the difference between species often has a greater influence than the difference between sexes, so our recommendations will vary depending on product area.
Sponsors can include only one sex in clinical trial data if the investigational product is intended for use in only one sex (e.g. prostate cancer, ovarian cancer, or certain inherited diseases). In this case, the supporting non-clinical studies can include animals of the sex representing the target clinical population. In addition, the sponsor should provide us with their rationale as to why animals of only one sex will be used during studies.
2. What actions does the FDA intend to take related to the use of clinical data outside of the U.S. and how it will be generalized to the U.S. population?
Although there were comments from the open docket and public meeting about using clinical data collected outside the United States and its relevance, we do not anticipate developing any action items on this issue. When trials are conducted in accordance with good clinical practice (GCP), we review these data and may also require additional data from the U.S. population. We are required to consider clinical trial data submitted as part of pre-market applications, which include data from studies performed outside of the United States, and to consider the data’s relevance to the U.S. population.
We recognize that we live in a global economy where many clinical trials increasingly include multi-national subjects, or are performed solely outside of the United States. There are many reasons that trials may be multinational, including that the companies are multinational and want to gain access to many markets. There are also diseases that occur more frequently or solely outside the United States, so it is necessary to study them where they occur. Impeding sponsors’ abilities to conduct clinical trials outside of the United States would almost certainly stifle innovation and increase the cost of medical products.
Data collected from American participants in multinational trials can analyzed if we can justify why such a study would be relevant to the U.S. population. We would need to have sound scientific reasons for why the study could not rely solely on data collected outside of the United States. For example, suppose an angiotensin-converting-enzyme (ACE) inhibitor drug is studied outside of the United States and there are insufficient numbers of African Americans in the study. Because ACE inhibitors are known not to be as efficacious in that population, a U.S.-based study including more African Americans might be requested.
3. What is the role of data collected in Phase 1 and 2 trials?
Data from these early trials is essential for identifying a medical product’s safety, dosing, and proof of concept. For drugs, this early knowledge is also important for identifying how it may affect more specific groups of people.
Early studies can demonstrate how a drug is metabolized or broken down by the body, which may be different base on sex, race, or ethnicity. When we see significant differences, we can then provide specific dosing and administration recommendations for these groups.
4. Are women excluded from clinical trials?
No, women are not excluded from participating in clinical trials as a matter of policy. Overall, women comprise a large portion of clinical trials participants. For example, a 2001 GAO report (GAO-01754 Women in Clinical Drug Trials) showed that women represented more than half of study participants in all of the new drug applications. In some cases, study enrollment criteria – while important for defining a population for whom benefits are expected to outweigh risks – may unintentionally exclude certain demographic groups, thereby limiting useful information about product performance in diverse populations.
5. Women represent more than half of the U.S. population. Does the FDA require that at least half of each clinical trial population include women?
There is no single "right" target number or percentage for subgroup analysis. For the large majority of cases, we do not require any specific numbers for subgroups. The FDA may be able to determine if a medical product is safe and effective for both women and men with enrollment of women at a level less than half and others may be more than half. Statistically, it is possible to determine that a medical product is safe and effective for both men and women in a study with fewer patients than it would be to compare how a medical product performs in women vs. how it performs in men.
If during the course of our review, we need more information about women to make an approval decision, then we can require sponsors to conduct additional clinical trials or other types of studies. In addition, we continually evaluate safety reports on FDA-regulated products after they are on the market and being used by the wider public.
6. What does the FDA mean by enrollment consistent with disease prevalence or incidence?
The FDA believes that enrollment should reflect the patients most likely to use a medical product. Sponsors should design clinical trials using available information from the relevant medical literature, clinical knowledge base, and health statistics for the disease.
7. Does the FDA require that clinical study data address differences in demographic subgroups?
We do require that each sponsor provide sufficient data to evaluate the safety and efficacy for the intended population. Therefore, if there are questions related to the safety and efficacy of the medical product in the intended population, then we may ask for additional data to address these questions.
We also assess and combine data to examine sex differences, called a “pooled analysis.” This type of analysis allows us to detect both clinically relevant and statistical differences between sexes.
If a subgroup is known to have a significantly different response than the rest of the population, or if a specific claim is sought in a certain subgroup, additional analyses may be needed.
8. How does the FDA consider data from analyses of demographic subgroups during product review? How will this apply in the post-market setting?
We carefully determine how the available clinical data can be generalized, in other words, be applied safely, to subgroups. Therefore, we ask that sponsors provide subgroup analyses for safety and effectiveness, and we then assess and evaluate that data in our overall review of the product. When necessary, the FDA requires additional studies to address areas of concern identified in pre- or post-market clinical databases.
9. How can I easily access information about demographic subgroups?
When we consider information on demographic subgroups to be relevant to how a product is prescribed or used, we can request that it is included in a product’s label. To make demographic subgroup information more easily accessible, we will instead post on a webpage for the public, advocacy groups, researchers, industry, health care providers, and anyone who wants to access it.
You will also find information gathered from many drug trials on a single webpage, while clinical prescribing and use information will remain available for providers and patients in a product’s label.
10. What should a patient do if they are prescribed a treatment that might not have been tested in their particular demographic subgroup?
Patients should speak to their health care providers about the risks and benefits of taking the prescribed therapy. Patients should not stop any prescribed treatment without first consulting with their health care provider.
 FDASIA Sec. 907 directed the FDA to develop an action plan outlining “recommendations for improving the completeness and quality of analyses of data on demographic subgroups in summaries of product safety and effectiveness data and in labeling; on the inclusion of such data, or the lack of availability of such data, in labeling; and on improving the public availability of such data to patients, health care professionals, and researchers.” For more information: http://www.gpo.gov/fdsys/pkg/PLAW-112publ144/pdf/PLAW-112publ144.pdf.