FDA offers a status check on its diversity in research provisions, one year post-FDORA

At a workshop on FDA’s implementation of new statutory requirements for diversity in clinical research, agency and industry representatives gave a status update on implementation. So far, the key takeaway was that both industry and regulators are still identifying best practices on setting enrollment targets in their research plans.

Diversity in clinical research: Guidance and statute

• In April 2022, the FDA issued a new draft guidance document on diversity in clinical research programs. As AgencyIQ discussed at the time, the life sciences industry had long been expecting such a guidance from the FDA. While the primary developer of the draft guidance was the FDA’s Oncology Center of Excellence (OCE) as part of its “Project Equity,” the FDA’s three life sciences product centers “also contributed to this collaborative effort,” according to a statement from the FDA. The recommendations in the guidance are intended to apply to a broad scope of products regulated by the agency – including drugs, biological products and medical devices, which fall under the purviews of the Centers for Drug and Biologics Evaluation and Research (CDER and CBER), and the Center for Devices and Radiological Health (CDRH). The guidance followed significant concerns from regulators, industry, clinicians and researchers that historically underrepresented groups, particularly racial and ethnic minorities, were not being adequately included in research populations for life sciences products.
• This guidance joined others in the FDA’s catalog that focus on diversity in research, including its 2022 guidance on inclusion of older adults in cancer clinical trials and its November 2020 guidance on “ Enhancing the Diversity of Clinical Trial Populations – Eligibility Criteria, Enrollment Practices, and Trial Designs.” However, unlike those two earlier guidance documents, the diversity action plan guidance document would not focus solely on making research programs more accessible to a broader range of potential participants, but also on actively directing sponsors to set (and meet) goals for enrolling a defined population that meet certain demographic criteria.
• The diversity plan guidance in a nutshell: In the April 2022 draft guidance, the FDA outlined a policy whereby sponsors of certain products would be required to submit a Race and Ethnicity Diversity Plan that outlines and justifies their approach to recruiting and retaining a “representative” research population. The recommendations apply to a wide range of application types, including Investigational New Drugs (INDs) or Investigational Device Exemptions (IDEs), as well as any product “for which clinical studies are intended to support a marketing submission.”
• The Food and Drug Omnibus Reform Act (FDORA) gave FDA new statutory authority and responsibilities. FDORA was passed into law in December 2022 as part of the end-of-year 2023 federal spending bill. The provisions in FDORA included a specific section on “Clinical Trial Diversity and Modernization,” (found at Subtitle F, Chapter 1). The first section of that chapter is section 3601, “Diversity Action Plans for Clinical Studies.” At a high level, this section of the law amends the Federal Food, Drug and Cosmetic (FD&C) Act to expressly grant the FDA the statutory authority to require diversity action plans for both drug products and medical devices. Per the statutory language, sponsors looking to run a phase 3 or “other pivotal” study for a drug would need to submit the diversity action plan with that protocol; for devices, the diversity action plan would need to be submitted with an IDE or, if an IDE is not required, “in” the marketing application (i.e., 510(k) pre-market notification, De Novo classification request or Pre-Market Approval (PMA) application).
• The law directed the FDA to “update or issue guidance” on diversity action plans. This includes the specific format and content of the plans, such as the rationale for enrollment goals (e.g., estimated prevalence of the condition of interest, any relevant pharmacokinetic or pharmacogenomic data, demographic factors) and how they intend to meet those goals, as well as operational and process factors such as how they should be submitted and how the “action plan” may be updated over time.
• The status of this policymaking so far: OCE associate director LOLA FASHOYIN-AJE had previously confirmed that the agency is currently accepting and reviewing diversity plans, “which are very aligned with our current draft guidance on diversity plans for race and ethnicity,” but was careful to specify that these are separate from the diversity plans that would be statutorily required under FDORA, which the agency has “not yet defined in guidance.” As AgencyIQ has previously discussed, there are several additional considerations under the FDORA authorities that are not addressed in the original 2022 draft guidance, which the agency would have to sort out through additional guidance development (e.g., waivers).
• That said, the diversity plans (DPs) under the April 2022 guidance and the diversity action plans (DAPs) that can be statutorily required are different things.
• FDORA also directed the FDA to convene a meeting. Section 3603 of the legislation directed the agency to hold a public workshop “to solicit input from stakeholders on increasing the enrollment of historically underrepresented populations in clinical studies and encouraging clinical study participation that reflects the prevalence of the disease or condition among demographic subgroups,” as well as related issues like dissemination of information, enrollment goals, and the approaches to supporting inclusion. The deadline for that meeting was one year post-FDORA enactment – in effect, by December 2023. The two-day workshop was convened November 29-30, 2023, and was cohosted by the Clinical Trials Transformation Institute (CTTI).

FDA and industry implementation, one year in: A focus on enrollment target setting and baseline data

• OCE published a report on the one-year experience with voluntary DPs. Discussed at the meeting by Fashoyin-Aje, the report outlines what the agency has seen so far in DPs that were voluntarily submitted to the agency specifically for oncology drugs. While this does limit the sample somewhat, it appears that these are the bulk of the DPs submitted to the drug center so far: “of the 91 DPs submitted to CDER during the evaluation period” of April 2022-April 2023, “76 (84%) were submitted to the oncology divisions.”
• This provided some insight into what the FDA looks for in a DP, and its process for review. Per the report, “DP content completeness was assessed on five components specified,” which are included in the 2022 draft guidance: (1) disease overview; (2) development program scope; (3) enrollment goals; (4) measurements to achieve enrollment goals; and (5) status of meeting enrollment goals. So far, it seems that most sponsors submitting DPs are submitting completed plans, but the FDA found about 13% were “missing one or more core components.” However, it’s not entirely clear what action the FDA or sponsor would be expected to take to address or flag those deficiencies. The report does indicate that there is some back-and-forth between the agency and sponsors on their DPs: “FDA provided feedback on the sponsors’ proposals for 31 of 76 DPs” in the sample, or a little over 40%. The feedback was primarily – 90% – focused on enrollment goals, although feedback on enrollment monitoring and strategies to enhance accrual to meet those goals were also fairly common (29% each). The “average time from DP receipt to FDA sending comments to Sponsors was 61 days,” per the report. As would be expected, the bulk of DPs so far are for later-stage studies (phase 2/3), while about 11% are for early development stages, 4% are for post-market settings and 1% are in marketing applications.
• This report only addresses oncology products in CDER: What about CBER and CDRH? So far, we don’t have much insight into lessons learned or the processes put in place regarding DPs from the other two product Centers, or when to expect any kind of similar report.
• Fashoyin-Aje cited the need for actual implementation of diversity efforts, rather than continued discussions about why it’s so important: “We don’t need to relitigate this every time, or elaborate on a laundry list of challenges and barriers to conducting more inclusive research,” she stated. Instead, “we just have a ton of work to do to regain trust, and at the trial level, central to the study should be the ascertainment of the characteristics, values, experiences of the populations affected…. And this should be done before pen is put to paper to write a clinical study protocol.”Fashoyin-Aje urged developers – and regulators – to move beyond discussions of the impact of a lack of diversity and into full implementation of new practices, at an early stage of the development process.
• Policy-wise, this is still a work in progress. As discussed above, the FDA has yet to define DAPs – as would be required under statute – in guidance. Further, “there’s apparent tension between long standing practices of accepting foreign data to expedite drug development and a new emphasis on diversity and representativeness in clinical trials,” acknowledged Fashoyin-Aje. Going forward, there could be new opportunities to continue including non-U.S. data, but from historically underrepresented regions that may more closely mirror the population of the U.S., she explained, including Central and South America and Africa. As MARC THEORET (Deputy Director, OCE), put it later in the meeting, “one of the key opportunities is by increasing enrollment in Sub Saharan Africa, for example, [and] in Latin America. I think that represents a great opportunity to help increase the representativeness of the clinical study population, [when] thinking about generalizability.” However, this would involve significant build-up of trial infrastructure in these locations.
• Currently, less than half of the registered clinical trials report out race for all five racial groups. FDA defines racial categories based on the Office of Management and Budget (OMB) Directive 15, which requires five minimum categories (white, Black or African American, American Indian or Alaska Native, Asian, and Native American or Other Pacific Islander, while “ethnicity” is defined as Hispanic or Latino, or not Hispanic or Latino). However, according to a review conducted in 2022 of the trials registered on clinicaltrials.gov presented at the workshop by Stanford Medicine’s ELDRIN LEWIS, “only 46% actually report all five groups,” and in general “typically it was going to be white or non-white.” Further, “what often happens in trials is everyone is assumed to be white,” said Lewis, when the data fields indicating race are not filled out, especially in non-U.S. countries. Lewis called for increased standardization in the way that race and ethnicity are defined in research datasets, but also called for incentives to actually fill out the data fields, including specific guidance on doing so or requiring grant proposals to include a specific plan to capture these data.
• So far, it appears that setting baseline enrollment target goals is a point of interest (and contention) for industry and the agency. At one panel session focused on establishing clinical study enrollment goals, panelists from industry outlined their feedback and perspective. MICHAEL REED (GSK) provided some insight into his organization’s work on DPs so far, noting that “there’s been some feedback from various diversity plan responses that I’ve heard, as well as received ourselves, about base-level translatability.” In particular, he said, “that idea is difficult if we’re using different base sizes, difficult if we’re using different data sources, difficult if our endpoints don’t match” with what other researchers are doing or the sources from which they are analyzing data. “And so the coordination that we might be able to do, across [or] perhaps within therapeutic areas, for example, would be remarkable,” said Reed. It was apparent from his presentation that there still isn’t a one-size-fits-all approach to defining population for enrollment targets, or even necessarily the best practice for sourcing the demographic or epidemiological data.
• Still, establishing targets is currently the focus, and even “like, the number one thing to do,” said MEGHAN MACKENZIE (Genentech). However, the best way to do this is still a concern, coming with questions such as “what is the right prevalence, [or] what is the prevalence of the disease in different indications,” and then finding the right source of that data, “so we have epidemiologist[s] interrogating the data,” said Mackenzie.

Moving beyond setting enrollment targets

• Experts at the meeting pointed to data gaps even further downstream. Currently, there is a practical and operational focus on moving trial infrastructure from more traditional sites (e.g., academic medical centers) to new locations and methods that can help bolster contact with underrepresented populations – such as leveraging community health centers or investing in digital research methods. However, panelists at the meeting pointed to a serious outstanding gap: “screen failures,” or individuals who were given the opportunity to participate and chose not to. Identifying the reasons why individuals who meet criteria to participate but don’t follow through can highlight additional barriers that won’t be addressed by simply moving trial sites or recruitment to locations, like community health centers, or methods, such as digital research methods, that could capture a more diverse populations.
• Per SCOTT HALPERN (University of Pennsylvania Perelman School of Medicine), this could improve insight into where, exactly, breaks are happening, and who is given the opportunity to consent, rather than simply who ends up participating. BOBBI CHAPMAN (Abiomed) further discussed this point, saying that in her review of the screen failure logs, “the number one reason for Blacks and women, as well as Asians, to not be enrolled was lack of informed consent.” Only once you know who was approached can you identify additional barriers to their participation, Chapman explained, “and then you can develop strategies [on] how to diversity enrollment.” For her part, “I think we’re learning, as an industry sponsor, the value of being able to capture that additional data with your screen failure logs,” she said, and applying these types of lessons to other data gaps, such as longer-term data on who may be lost to follow-up and why.
• Investment and infrastructure in industry: Implementing the FDA’s policies on standardizing diversity information and expectation in research has necessitated some internal and structural organization. DOOTI ROY (Boehringer Ingelheim) described what that has looked like for her company, explaining that BI implemented its own “Diversity Plan Working Group. “This cross functional group of colleagues had established business practice and treated diversity plan templates to ensure that our teams were working similarly across therapeutic areas and development programs. And as this was something new for us, they also created mechanisms for sharing the completed diversity plans and the lessons we learn from the interactions with the agency,” she explained. This means that all teams within the company are working from the same diversity template when building research programs, and that the template includes fields for each area of consideration in the FDA’s guidance. Per Roy, “it’s basically a work in progress,” as the plans are being developed and translated into practice, and the company is still receiving iterative feedback from the FDA.
• A big concern for researchers, IRBs and regulators: Diversity beyond race and ethnicity. This includes demographics such as people with intellectual and developmental disabilities (IDD), mental illness, or contexts such as pregnancy and lactation – notably, all things that can co-occur and also be present in racial and ethnic minorities. The FDA/CTTI workshop focused several panels on the high-level challenges and barriers to conducting more inclusive research specifically for these groups, but offered limited policy solutions or proposals, besides a generalized focus on building on a system that assumes inclusion, rather than exclusion. From a policy perspective, many of the questions about research inclusivity for people with IDD or mental illness relate to informed consent requirements, which are extremely slow to update.

Analysis and what’s next

• The actual impact “will take some time to see,” acknowledged Fashoyin-Aje. “The impact, obviously, from these diversity plans will take some time to see in the clinical trials that reflect that prospective planning. But I think the impact is already being already seen and felt in the way that companies are standing up teams that have the expertise to really evaluate real world data, getting your logic data and figure out ways to define benchmarks for enrollment across the different demographic groups,” she said.
• In effect: We are still in the organizational stages of implementation, with both industry and regulators setting up the infrastructure to implement the research diversity mandates.
• Going forward, “What I want industry to do is listen to us,” said ILEANA PINA (Medical Officer, CDRH). From Pina’s perspective, regulators are looking for industry to have a plan, and the ability to justify their plan, when they are submitting their research program to the agency. For now, though, it doesn’t appear that there is an established best practice on what that plan looks like – or what the single source of truth is for quantifying prevalence per specific population group or what, specifically, the enrollment targets should be based on that epidemiological data.
• We are still waiting for more information, including resources on DAPs (as defined in statute) and best practices on developing (and carrying out) these plans. Notably, Fashoyin-Aje said at an event earlier in 2023 that FDA “plan[ned] to publish” an information resource on diversity plans (under the 2022 guidance, not the statutory requirement) to help provide information on what the agency was looking for – at this point, it’s not clear if she was referring to the report cited above from OCE.

To contact the author of this item, please email Laura DiAngelo ( [email protected]).
To contact the editor of this item, please email Kari Oakes ( [email protected]).

Key Documents and Dates

• Discussing Approaches to Enhance Clinical Study Diversity Public Workshop